ENERGY EFFICIENCY IN

DOMESTIC APPLIANCES AND LIGHTING

PROCEEDINGS OF THE
4TH INTERNATIONAL CONFERENCE
EEDAL'06
21-23 June 2006, London, United Kingdom

EDITORS
PAOLO BERTOLDI, BENIGNA KISS, BOGDAN ATANASIU

VOLUME 3

Institute for Environment and Sustainability

2006 EUR 22317 EN

The mission of Institute Environment and Sustainability is to provide scientific and technical support of EU policies
for protecting the environment and the EU Strategy for Sustainable Development.

European Commission
Directorate-General Joint Research Centre
Institute Environment and Sustainability

Contact information
Address:TP 450, I-21020 Ispra (VA), Italy
E-mail: paolo.bertoldi@cec.eu.int
Tel.: +39 0332 78 9299
Fax: +39 0332 78 9992

http://energyefficiency.jrc.cec.eu.int
http://www.jrc.cec.eu.int

Legal Notice
Neither the European Commission nor any person acting on behalf of the
Commission is responsible for the use which might be made of this publication.

A great deal of additional information on the European Union is available on the Internet.
It can be accessed through the Europa server
http://europa.eu.int

EUR 22317 EN
ISBN 92-79-02752-2
ISSN 1018-5593
Luxembourg: Office for Official Publications of the European Communities

© European Communities, 2006

Reproduction is authorised provided the source is acknowledged

Printed in Italy
 Standby

959
960
Standby Consumption in Private Homes Socio-economic Studies,
Mapping and Measuring Reduction? What Works: Campaigns or
Hardware Solutions?
Erik Gudbjerg and Kirsten Gram-Hanssen

LokalEnergi A/S and Danish Building Research Institute

Abstract:
The paper will sum up the findings from a standby project carried out in Denmark over a period of 2 years
involving 30 households. The purpose of the project is to verify to what extent standby losses in
household consumption could be affected by direct communication and by use of technical equipment to
reduce the standby losses. The analysis will include studies on how socio-economic status and life style
is affecting the standby consumption.
In Denmark standby losses are estimated to be approximately 970 GWh/year. One consequence of this
consumption is emissions to the environment by approximately 821 million tons of CO2. It is estimated
that standby power is responsible for 1% of the global CO2-emmisions from end uses including transport,
industry, agriculture and waste disposal in general.
Increased use of electronic household devices will also increase standby losses. The average share of
household energy consumption related to standby losses has increased over the recent years, and will
undoubtedly continue to increase over the forthcoming years. However, when the losses for each single
device (TV, video etc.) are considered separately, the standby losses and the savings potential are small.
However, all devices collectively contribute a significant proportion to total electricity consumption. Earlier
campaigns have all focused on changing behaviour of the customers but this most recent study also
examined a range of technical solutions to reduce standby.
This new Danish project focuses on standby losses within a group of households over a period of 12
months. The group, containing of 30 households with 2 adults and 2 children each, has received advice
and information on how to change behaviour in order to reduce standby losses. Furthermore the group
has been equipped with technical equipment as additional means to reduce losses. This equipment
(save-plugs and newly developed equipment with low standby) will be presented as a part of the
presentation.
Energy consumption due to standby losses in the 30 households has been measured (on each piece of
equipment with standby power, some online and some as spot measurements), and this data has been
analysed and is presented on a website where the households can follow their own consumption in real
time. From these measurements the paper will also include descriptions and analysis of consumer
behaviour related to the appliances and devices.

Introduction
First of all this paper deals only with standby consumption in the household sector, and the kinds of
equipment that can be connected with a plug, and primarily entertainment and office equipment (ie
equipment that is wired directly to the power system has not been covered).
In 1993 at the ECEEE conference in France, Sandberg presented a paper on standby consumption
(which he called “leaking electricity”). Standby consumption was more or less neglected, despite
Sandberg'e conclusion that, on the basis of the research he had carried out in Sweden, standby
consumption 15 years later would be projected to be about 400 kWh/household per year in Europe. At the
ECEEE conferences in 2003 and 2005 several papers were presented on the standby issue by
Harrington and Meier, among others Standby consumption has also been discussed in many related
reports and projects, there has been a particular focus on how to define standby consumption, how to
estimate it, and how to influence the producers to realize the importance of designing equipment with low
standby consumption and to produce products with low standby.

961
The IEA has recommended not putting much effort into discussing definitions, even though they have an
important role to play, because they would like to see actions, but rather working for politically initiatives
to reduce standby consumption. An example how efficient these can be is the Executive order issued by
the President of US (Executive Order 13221), claiming that agencies when buying available off-the-shelf
products, shall buy products that cause no more than one Watt when in standby mode. This executive
order has had a very high impact on the total office product side.
But it is also important to realize that there is another side of the coin with respect to a lack of commonly
agreed definitions. Firstly, where there are no agreed definitions for modes then manufacturers have to
cope with a range of different definitions which apply in different countries. This was noted by HP’s
delegate on the “Action on 1 Watt” conference in Copenhagen, March 2005.
Secondly it has been pointed out that the generic term “standby” can refer to many different modes of the
same appliance, therefore it is appropriate to differentiate between different modes for different appliance
types. For instance, a VCR can have several modes such as active standby (waiting to play), delay start
(VCR programmed to record later) and passive standby (shot down but waiting for a remote signal) (Lane
& Wajer 1997). Another example is that computers and printers have a range of different types of sleep
modes depending on usage regimes and user programmable features. A general taxonomy of such so
called 'lopomo', (so called low-power-modes), has been suggested by several authors including Meier
(2005) and another paper at EEDAL (Nordman et al 2006).
There are at least two different ways to determine the total magnitude of standby energy consumption in
households. Either through bottom-up models where market statistics on ownership of appliances and
equipment are combined assumed usage patterns and known power consumption attributes to provide a
model of household standby electricity consumption. This has been done in a US study, which concluded
that approximately half of the total electricity consumption from consumer electronics are consumed while
the appliances are in standby mode (Sanchez et al. 1998). The other way to determine the size of
standby consumption, is to directly measure it. However since there is no common definition of standby
consumption, there is no one commonly agreed approach to determine it. Another American study is
based on spot measurement of electricity consumption in ten households for all appliances in the mode
which the households normally leave the appliances (Ross and Meier, 2002). However, this
measurement approach does not really capture the actual behaviour of the households. The most
accurate way to measure standby consumption is to have continuously measurements of both in-use and
standby consumption of appliances in ordinary households.
The most comprehensive study done so far in this way is probably the EURECO-project including 100
households each in four European countries, the measurements including all appliances and lighting in
the households, measured during one month (Sidler, 2003). However, individual meter used for this
project were unable to measure power consumption levels less than 3 Watts so limited information on
standby can be obtained from this data set for many appliance types. A similar project is currently being
carried out in Sweden by the National Energy Authority, including 400 households, where 90% of the
households are measured during one month and the remaining 10% of the households will be measured
during a period of 12 months (Bennich 2006). The Danish part of the EURECO-project data has been
analysed further, and it has been concluded that standby consumption for consumer electronics in the
Danish households varies between 0 and 1300 kWh per year , on average 8% of the total household
electricity consumption is attributable to standby power . More than 10% of these households had
between 15% and 28% of their total electricity consumption as standby (Gram-Hanssen, 2005). A recent
review study, including both published and unpublished studies, concludes that standby consumption in
developed countries is between 60 and 110 W per household, corresponding to an average of between
4% and 11% of the household total electricity consumption (Meier, 2005). European households are in
the lower end with 50 to 70 W/household and year (although many of the preliminary studies in Europe
only considered a limited range of appliances and equipment when quantifying total standby), with
countries like Denmark in the upper end (Meier, 2005). This suggests that Denmark does not necessarily
have an inherently worse standby profile than other countries in Europe, but more likely has more
carefully quantified the standby for all appliances and equipment. One of the sources for the conclusion
are based on preliminary results from the study presented in this paper.
The question of how to limit standby consumption is an issue raised in most of the reports and articles
and many concentrate on how to influence the producers of consumer goods to develop products with
low standby consumption attributes. It is generally agreed that there is no technical problems in achieving
this, and the question is, which political tools are the most effective (IEA, 2001). However, it might also be

962
relevant to analyse to what extent it is possible to persuade consumers to turn reduce standby of their
existing appliances and equipment, either through the modification of their behaviour or with the help of
technical devices.
This project been investigated the extent to which it is possible to limit standby consumption in consumer-
oriented approaches, and which of these are the most effective ones: either different types of
communication or the installation of technical devices.
The results presented in this paper is based on a project supported by the Danish association utilities,
and with participants from Danish Building Research Institute, Keep Focus, The Danish Electricity Saving
Trust and Lokalenergi. The project had a total budget of 286 000 Euro.

Method
The overall objective of the project was to monitor the standby electricity consumption in 30 households
during a whole year with the help of on-line measurement, while they where exposed to different kinds of
program measures aimed at changing their standby consumption. To assess each of the different
program measures, 3 phases were monitored during the project:
1 Reference period, where the standby consumption is quantified prior to implementation of any
measures.
2 Communication period, where the families are subjected to different types of communication,
including posted leaflets, visits from an energy adviser and the possibility of following their own
metered standby power consumption on a webpage.
3 Technology period, where different types of technical devices to reduce standby power (eg devices
which make it easier to manually turn off products or products are automatically turned off), has been
handed out. Technical devices include auto-saver plugs for television and PC, remote control or time
switches for power boards.
Due to the limited resources available, the focus has been on the consumption of entertainment and
office products. There have also been spot measurements of standby consumption on other products
such as cordless phones, touch lamps, cell-phone chargers etc. Through the online measurement of the
entertainment and office products it was possible to determine when different groups of appliances were
either turned off, in standby or in use. The recorded electricity consumption data has been analysed from
the perspective of information about the families such as number of members, age, income etc.
For this project standby been defined as consumption of the appliance when it is not actively used by one
or more persons.
The families were selected with a focus on the following characteristics: family living in single-family
houses, generally with children still living in the household, families with equipment which have significant
standby power. The reason for this decision was that the main target of the project was not to estimate
the size and composition of the standby consumption for all households, but to understand what works
regarding behaviour and reducing standby losses.
The families were selected with the help of the database of the local utility, in all 270 families were
contacted. Among the criteria was a total electricity consumption above 5 000 kWh/year.
One consequence of the selection criteria was that the participating families on average had a higher total
income than average families in Denmark. This is not surprising due to the fact that the project only
consists of families living in single-family houses. The participating families' electricity consumption is also
above average.

The online measurements

The main contractor for the installations of equipment for the on-line measurements has been the
company Keep-Focus. Their data server has also served as host for the data collection and for the web
page where the participants could follow their own patterns of standby power consumption.
Typically four data loggers where used in each house. Since a large part of the entertainment equipment
and office products are used together, the procedure has been to connect all entertainments products in a
room to one channel of the data logger, all the office products to another channel, and so on. The data
was logged with a time resolution of one hour. Logging data in 30 homes over one year includes
challenges; loggers fail, people move there equipment around within the home, they buy new pieces of

963
equipment and forget to tell the project staff about it. The project staff has tried to “repair” such incidents
as soon as they were discovered in order to secure high quality and consistent data.

The phases
The communication phase was designed with the purpose to test different communication measures. The
measures that where selected for test were:
Information meeting – cancelled due to the fact that too few people signed up for participation.

Direct mail together with folders on “Efficient use of electricity” and “How to reduce standby
consumption”

Home visits from a “standby consultant”, with the purpose to visualize the standby consumption,
with the help of metering of standby consumption of all products included in the project. The
metering results were mailed to the families a few days after the visit together with a list of
proposals on how to “kill” standby consumption.

The last activity included access to a personal web page where the families could follow their own
standby consumption from all the data loggers installed in their homes.

The technical phase was designed to test how much standby consumption could be reduced with the help
of so called “standby killers” such as:

Master – slave auto power saving plug banks, that automatically power down any equipment (VCR,
DVD, TV etc) when the unit connected to a master plug is switched off.

USB auto-power saving-plug banks, which automatically power down any connected equipment
(monitor, printer, etc.) when the computer is switched off

Remote controlled powerboards

Timer controlled powerborads

Interviews
At the end of the metering phase, 10 families where selected among the 30 for interviews. The criteria for
selection were those where standby consumption had been influenced during the communication phase,
during the technical phase, or where there was no change of standby consumption at all. The purpose of
the interviews was to assess the families' experiences from participating in the standby project and to
correlate their standby behaviour and the families' opinions on standby consumption.

Results
This results commence with a description of the standby consumption in the reference period, which was
the period before any actions was taken to influence the families in the project. This is followed by an
analysis of how the families use there appliances and how the different actions influenced the standby
consumption.
The size of the standby consumption can be assessed in relation to the total electricity consumption of
the families. The standby consumption can also be related to different types of consumption such as
entertainment or office, the age of the families, total income of a family.

964
When relating standby consumption to the total electricity consumption of the household, it can be
concluded that there is no unique relationship. From the available data it can be calculated that on
average 9% of the total consumption is standby consumption, covering a spread from as little as 2% to as
much as 18%, corresponding to 120 kWh/year to 980 kWh/year or an equivalent continuous load of 17 W
to 130 W, with an average of 67 W. This is nearly of the same level as assessed in other projects, such
as EURECO. Due to the limited resources available for the project it has only been possible to online
meter 65% of the standby consumption from office and entertainment equipment in the private homes,
the rest is estimated with help of spot measuring.
There appears to be no clear connection between family type and the magnitude of standby energy
consumption. With the help of spot metering on cell-phone chargers, cordless phones etc. together with
online metering results, it can be concluded that standby consumption related to home office equipment is
on average 37 % of the total standby consumption for the families monitored.
It was found that standby consumption increased with increasing total household income for the families
in the study. But because of limited material it is not possible to show any significant results of these
types of correlation.
The metering results from the reference period clearly show that the majority of the families have
significant standby energy consumption for entertainment equipment. This is generally equal to or larger
than the energy consumption in on mode. However, the results show, not surprisingly, that families with
teenagers living at home (which tend to have their own entertainment equipment) have higher energy
consumption in on mode than in standby mode(s). This is also reflected in the use of office products,
where families with teenagers clearly have longer periods and higher energy consumption in on mode
than in standby mode(s).
On average for the sample monitored, 42% of the measured electricity consumption for entertainment
and office products was in standby mode(s), while 58% was when the equipment was in use.
The following 3 figures illustrate the results for the reference period. Figure 1 shows the consumption as
part of total electricity consumption in the 30 households, arranged into three groups; middle aged (older
without children living in the household) (to the left in the figure), families with smaller children (in the
middle of the figure), and families with teenagers living in the household (to the right in the figure).
Figure 1 show the electricity consumption pr. year for entertainment equipment, distribution in use and in
standby mode.

965
 1000

900

800

700

600
kWh

500

400

300

200

100

0
a11 (1/1 )
a1 (2/2)
a2 (2/1)

a7 (2/1)

a10 (3/3)

a12 (1/1)
a13 (2/2)

b1 (1/1)
b2 (2/0)
b3 (2/ 2)
b4 (2/2)
b5 (2/2)
b6 (3/3)

c1 (2/2)
c2 (2/2)
c3 (2/0)
c4 (4/3)
c5 (2/0)
c6 (3/1)
c7 (1/1)
c8 (4/1)

c10 (3/1)
c11 (2/2)
a3 (4/2)
a4 (2/1)
a5 (2/1)
a6 (1/1)

a8 (1/1)
a9 (2/2)

c9 (4/4)
Middle-aged without children Famil ies with young children Fami lies w ith elder children

Anual standby consumption Anual consumption in use

Numbers in brackets(x,y): x: number of television sets in the household, y: number of television sets measured.
Figure 1: The electricity consumption pr. year for entertainment equipment, distribution in use and
in standby mode

966
Figure 2 shows the annual electricity consumption for office equipment, distribution in use and in standby
mode.

1400

1200

1000

800
kWh

600

400

200

0
a10 (5/1)
a11 (2/1)
a12 (1/1)
a13 (2/0)

c10 (2/2)
c11 (2/1)
a1 (1/0)
a2 (0/0)
a3 (1/1)
a4 (1/1)
a5 (1/1)
a6 (1/1)
a7 (2/2)
a8 (1/1)
a9 (2/1)

b1 (1/1)
b2 (1/1)
b3 (1/1)
b4 (1/1)
b5 (4/4)
b6 (1/1)

c1 (2/2)
c2 (2/2)
c3 (3/2)
c4 (3/3)
c5 (2/2)
c6 (3/2)
c7 (2/1)
c8 (3/1)
c9 (3/2)
Middle-aged without children Families with young children Families with older children

Annual standby consumption Annual consumption in use

Numbers in parentheses (x,y): x: number of computers in the home, y: number of computers measured.
Figure 2: The annually electricity consumption for office equipment, distribution in use and in
standby mode.

As described earlier, different actions were taken in the communication phase of the project to trace
changes in behavior among the families.
The communication phase was followed by a technology phase, where energy advisers again visited the
families to offer them technical devices to reduce standby consumption. The average standby
consumption has been calculated in five steps, in order to assess which of the initiatives have
measurable effects.
Figure 3 shows that sending out written material and leaflets have almost no effect. Similarly access to a
homepage where the families' standby consumption was shown (on the basis of online measurements)
also appeared to have little effect. A real change in the behaviour of the families and thereby in the
standby consumption was observed when an energy consultant visited the individual families. With the
help of a small electronic meter, which was used to take spot measurements on each piece of equipment
in the household, the consultant was also able to demonstrate how much electricity is wasted as standby
consumption. Among the tools the consultant had at their disposal were samples of “standby killers”,
which were used to inform the families of the possibilities to reduce their standby consumption through
technology. The visit by the consultant where the standby consumption was visualize together with a
presentation of standby killers resulted in a reduction of standby consumption by about one third
compared to the reference consumption . The last action was a visit by an energy consultant who
installed “standby killers” to make it easier to switch of standby, this resulted in an additional reduction of
standby losses to a level that was about one third of the original reference. The conclusions from figure 3
seem rather simple: public campaigns and continuous information on standby consumption are of minor
or no use, while individual advice and technical devices result in real reductions in standby. Looking back
it is clear that it could have been very interesting to check the effect of standby killers alone.

967
 50

45

40

35

30
W

25

20

15

10

0
Reference period After written material After visit of energy 3 months after password After handing out technical
advisor devices

Figure 3. Development in the average (measured) standby consumption in different phases of the
project, 3 month after password, means when the families got access to their personal part of the
webpage.

But before jumping to conclusions, it is worth looking in more detail at the information collected through
the project.
The first question to discuss is to what extent the reference period represents the 'normal' standby
consumption or whether the fact that the measurement devices have been installed has altered the user
habits. Two things indicate that the reference period is representative of the normal standby consumption
levels: first from the interviews it was concluded that many of the families were conscious about not
changing their habits initially as they understood that the installation of the meters was to establish their
normal behaviour. Secondly, it was observed later in the project that attempts to change the households
habits were not very successful, which suggests that just putting measurement devices in the household
is not likely to influence their behaviour in any significant way. There where not registered a clear trend in
the development of the whole house consumption.
The next question relates to the influence from the written material, including a leaflet sent to the families
one to two months after start of project. As seen in figure 4, there is almost no positive effect on standby
consumption from this initiative.

From the interviews with the ten families selected for in depth interviews, we learned that some of the
families were confused about the material received. For example, some indicated that they did not know
whether they were still in the reference period or whether they were supposed to react to the material.
Thus in this phase the families' behaviour is clearly influenced by being part of the project. The effect of
the written material has been qualitatively evaluated through interviews in the following ways: some of the
families considered it to be one of the many small inputs pushing them towards doing something, others
(a majority) read the material carefully but only because they where part of the project and others just
threw it away without reading. The project thus does not indicate any significant effect of sending out
written material.
Figure 3 shows the average of how the families reacted to different initiatives of the project. However, this
average includes some major differences between the families. In some families, most of the decrease in

968
standby consumption occurred immediately after the first visit of the energy adviser, whereas in other
families there was almost no reduction in this phase of the project. The families who responded to the first
visit, have something in common: they consider standby consumption to be wasteful, mainly of money but
for some also of natural recourses, and they also found it very easy to change their behaviour. In some
families they simply started to use the on-off button of the appliances instead of the remote control, in
some cases they rearranged the cords and plugs making it easier to turn of many appliances on the
socket outlet with a single action. In these cases one person appeared to take an "evening walk" through
the house, turning of lighting and also equipment with some standby consumption. Some families even
bought “standby killers” in order to make it easier to turn of standby consumption.

The families who did not respond to the first visit also have something in common. Typically they did not
find the amount of energy or money they use for standby alarming, although it was comparable to those
families that found it wasteful. In context, even though the electricity price is high in Denmark ( 22 Euro
cents per kWh) , the cost of an average yearly electricity bill is a small share of the total household
budget. Danish families, like the ones in this project, typically spend 1.5 times more money on
communication and entertainment such as cable TV and telephone, so raising the relative importance of
standby is a challenge, also from the perspective that standby is only “10 %” of the total bill. ( They will
have an total electricity bill in the size of 1100 Euro and a total “commucation” bill in the size of 1600
Euro)
In the technical phase “standby killers” were handed out to the families half a year later. The delay was
intentional, and primarily because we wanted to establish whether communication had a long-time effect
or whether it would fade away in time., and due to the fact that there was a summer season to pass,
where normally the focus in Denmark are more related to outdoor activities. Assessments of the data
clearly indicate that, although small, the changes in the habits resulting from the communication phase
were lasting; this was also supported by the interviews. During the interview there was focus on whether
the family members thought they would continue their new habits, for instance by turning off standby
every evening before going to bed. A typical answer to this question was: 'Of course, because it has
become a routine which I just do without thinking about it'.
In the technical phase, an energy adviser walked through the house together with the family members of
the household, and proposed different types of “standby killers” for different types of standby
consumption. Among the families that did not respond to the actions in the communication phase, the
technical phase was more impressive, especially where the energy consultant installed the “standby
killers”. Further analysis has shown that standby consumption from entertainment equipment, to a larger
extent than from computers, was influenced by communication, whereas standby consumption from office
equipment was more influenced by the use of “standby killers”. The unanswered question is: How to get
people to buy an install “standby killers” if they are not part of a project. There are also questions about
how they could maintain them, because the users may not see if a unit breaks down.
Generally it can be concluded that difficulties in reducing standby consumption primarily occurs in
households with older (teenage) children. This may lead to a conclusion that teenagers do not want to
save energy, however this might not be quite fair to the teenagers. Other research has shown that
teenagers actually do consume more electricity than adults (Gram-Hanssen, Kofod and Petersen 2004).
However, this is a result of social pressure from friends as well as from parents who want their children to
be up-to date with regard to electronic possessions (Gram-Hanssen 2005).
Another question can be how much is possible to reduce the standby consumption with the existing
technology and the modern way of life. Families with older children typically use both television sets and
computers independently of each other in both time and space. However, often these appliances are
connected with each other for internet access or for satellite television. Since family members use the
different appliances in different rooms at different times of the day, it is very difficult to turn off, for
instance, routers and satellite dishes. Another challenge to reductions in household electricity
consumption is the new fashion with the main entertainment unit in the living room and satellite speakers
in other rooms.

Conclusion
The project shows that it is possible to reduce standby consumption in ordinary households to one third of
the normal level, by means of communication and help from technical devices. This is valid for an

969
average of the thirty households participating in the project. However, analysis also show that some
households rather easily could reduce all their standby consumption just by becoming aware of it,
whereas others households mainly reduce their standby consumption by means of technical devices.

Whether a family finds it easy to reduce standby consumption or not depends on on the one hand
whether the family in general is interested in energy savings, and on the other hand whether technical
details, such as how easy it is to turn of the standby either at the appliance or by reaching the socket
outlet in the wall.

One third of the standby consumption of the 30 households was not affected by any of the project efforts.
This was primarily a result of different appliances being connected to each other and used by different
persons at different times. These results emphasize first of all the need for improved pressure on the
producers to develop electronics with much lower standby consumption, for the kinds of appliances which
may be difficult to turn on and off. Secondly it also reveals that many people would like to turn off the
appliance, should it be possible to do so. This is often hindered by the fact that both socket outlet and
appliances not being equipped with a switch, which seams very inconvenient. It is a problem for the
producers to solve.
In all you may say that this project only deals with the top of the standby “iceberg” because the focus is
on office and entertainment, there are also all the white goods in the homes and other products. Another
thing that need to be addressed is the way the products are used the way they power down and so on.
A new test on SetTop-boxes for digital TV, carried out by the Danish Electricity Saving Trust indicates that
there are still a great need for for focus on standby consumption, the standby consumption varies from
4.4 watt (Grundig) to 14.4 watt (Digiality) for products that delivers the same thing.

Even though the production side may look as the most promising and obvious way to solve the standby
consumptions problem, this is a long term solution. However, even progress towards such long term
goals should be started as soon as possible. Until this has been achieved, households have to be
considered as relevant actors too. This project suggests a continued effort by raising awareness of the
public together with development and marketing of “standby killers” will have some impact. Even though
the same success rate achieved in this project can not be expected for a wider program with lower
resources per household, still it has been shown that parts of the standby consumption quite easily can
be reduced.

References

Gram-Hanssen, K. 2005. Husholdningers elforbrug - hvem burger hvor meget, til hvad og
hvorfor?(Households electricity consumption – who use how much, for what and why?). SBI 2005:12.
Hørsholm: Danish Building Research Institute.
Gram-Hanssen, K (2005) ‘Teenage consumption of Information and communication technologies’.
Proceedings of the 2005 European Council for an Energy Efficient Economy.
Gram-Hanssen, K.; Kofod C.; Nærvig Petersen, K (2004) Different Everyday Lives - Different Patterns of
Electricity Use. Proceedings of the 2004 American Council for an Energy Efficient Economy
Summerstudy in Buildings. Washington, D. C.: ACEEE
Harrington, L. and Holt, S. 2003. Australia's contribution on standby power. In Proceedings of ECEEE
2003 Summer study – Time to turn down energy demand. Paris: European Council for Energy
Efficient Economy.
(IEA) International Energy Agensy, 2001. Things that go blip in the night. International Energy Agency,
Paris.
Lane, K and Wajer, B.H. 1997. Standby consumption in TVs and VCRs: Lessons for other equipment with
'leaking electricity'. In Proceedings of ECEEE 1997 Summer study. Paris: European Council for
Energy Efficient Economy.
Meier, A. 2005. Standby: where are we now? In Proceedings of ECEEE 2005 Summer study – What
works and who delivers?. Paris: European Council for Energy Efficient Economy.
Ross, J.P. and Meier, A. 2002. Measurements of whole-house standby power consumption in California

970
 homes. Energy (27) 861-868.
Sandberg, E. 1993. Electronic Home Equipment – Leaking Electricity. 1993 ECEEE summer study. The
Energy Efficiency Challenge for Europe, Rungstedgard, Denmark. European Council for Energy
Efficient Economy
Sidler, O. 2001. End-use metering campaign in 400 households of the European community.
http://sidler.club.fr/
Gudbjerg E. the size and structure of standby consumption, 1 watt conference CHP. 2005
Bennich P, 2006, First report on household equipment electrical consumption, Swedish National Energy
Authority (400 households).

971
972
Standby: The Next Generation
Hans-Paul Siderius, Bob Harrison, Michael Jäkel, Jan Viegand

SenterNovem, UK-MTP, Abakus, Danish Energy Authority

Abstract
Measurements reveal that about 4% to 11% of residential electricity consumption is due to standby.
Organised efforts to measure and reduce standby power consumption have been under way for more
than a decade. However, the nature of standby and the policies to address standby power
consumption need to change. With regard to policies, it has been the assumption for many years that
a voluntary approach is faster, more flexible and less costly than a regulatory approach. However,
experiences with voluntary agreements have not always lived up to expectations and revealed
drawbacks with voluntary agreements which aim to reduce standby power. Regarding the nature of
standby, fewer products now have a single simple standby mode, (e.g. the remote control of a TV) as
the advent of advanced electronics and complex control systems now means that there are often
several more complex modes where the functionality can change in real time.
This paper describes the next generation of approaches to deal with standby power: on one hand the
horizontal approach, which targets the simple standby mode for as many products as possible. On the
other hand the power management approach tries to deal with complex standby situations to ensure
that the product moves into the lowest power consumption state possible for the required task.
Both approaches will be illustrated with examples, from the EU ecodesign framework directive, which
is being considered for use as a regulatory framework for a horizontal measure on standby, and from
the EU Code of Conduct for Digital TV Services Systems which is a voluntary measure for a more
complex product.

Introduction
It is well known that standby power consumption in European homes is about 60 W per home [1],
which corresponds to about 10% of residential electricity use. Although a qualifying caveat should be
noted in that such estimates tend to be based on the measured standby power requirement of
household appliances and generalized usage pattern surveys. Large scale, pan European surveys
measuring the actual duty cycle of each household appliance are rare. Estimates of household
standby power consumption tend to assume that the appliances are selected to standby or off by the
user when the main function is finished. Ad hoc surveys on specific devices such as set top boxes,
DVD players, washing machines, dishwashers etc. show that this assumption may be very misleading
in the context of estimates of energy wasted when appliances are not performing their main function.
In this paper we will present and discuss three issues that are important for keeping a grip on standby
consumption;
• The first issue is the nature of standby. The nature of standby has evolved from one simple
standby mode, e.g. the remote control of a television, to the very complex standby modes of e.g.
personal computers and set-top boxes. In more complex products the focus needs to be on
reducing duty cycle power consumption through power management mechanisms that
automatically control device power to the minimum needed to sustain the functions required at a
given time. This discussion also puts various definitions of standby, e.g. by IEC62301, in
perspective.
• The second is the issue of the approaches needed to address standby. It will be argued that for
the simple standby mode a horizontal approach targeting as many products as possible is
appropriate, whereas the power management approach should mainly deal with complex standby
situations, but may also have an essential basic application in ensuring that simple standby is
automatically achieved wherever possible.
• The third issue builds on the foregoing items and discusses which policy approach is appropriate:
mandatory or voluntary. Both policies will be illustrated with examples. Finally conclusions and
recommendations are presented.

973
The evolving nature of standby; or why standby consumption will increase
In order to illustrate the evolving nature of standby the following simple typology of products is used
[2]: the “on/off” product, the “standby” product and the “networked” product.

The “on/off” product

The “on/off” product is the most simple case: either the product is off (performing no function except
for being there and heating the room if energy is dissipated) or the product is on (performing one of
the main functions). However, even this simple case immediately raises one of the pitfalls for policy
makers: the description of the states “off” and “on” in general does not indicate anything about
whether a specific product has an off mode (e.g. hardwired smoke detectors do not have an off
mode), nor the time that a product spends in the on mode (e.g. electrical (alarm) clocks in the on-
1
mode all the time) .
Despite the trends identified in the next sections, a significant number of household products can be
classified as on/off products: vacuum cleaners, lighting, cold appliances, storage water heaters.
Furthermore, products with a standby mode and even networked products can have an off mode.
From an energy saving perspective these products are interesting for policy makers if they spend time
in the off mode and if the off mode has a power consumption > 0 Watt. An example are halogen
lamps with a transformer where the on/off switch is on the secondary side. On the other hand, cold
2
appliances and storage water heaters are always on and do not spend time in the off mode.

The “standby” product

The “standby” product denotes the classical standby situation, where the product is performing some
function(s) not being one of the main functions of the product, e.g. enabling (remote) control, waiting
for a user command, internal timer/clock, clock display or other indicator. Also in this situation the
“functionality” perspective offers guidance to distinguish whether a product is in standby or in on
mode. Compare e.g. an alarm system with a garage door opener. At a first glance it would seem that
both products are most of the time in standby because they are waiting for “input”, e.g. from a motion
detector, respective a user command to open the garage door. However, in case of the alarm system
providing security is the main function and therefore the system is always on, whereas in case of the
garage door opener the main function is opening the door and therefore the product is mostly in
3
standby, waiting for a user command .
Sometimes the “standby” products may also have an off switch, i.e. meaning that they can be
switched into the off mode (performing no function). The classical example is the (European) CRT
television which has both a remote control and an on/off switch. However there are few products with
these three simple modes these days.
This type of standby could be denoted as “simple” standby, because the procedure to enter (one of)
the standby mode(s) or to switch the product from (one of) the standby mode(s) to the on mode is
simple, i.e. can be determined internally and needs only recognize one type of (user) input.
This type of standby has evolved from a few products in the 80s to many products nowadays. Not
only consumer electronics, office equipment and microwave ovens, but also washing machines,
driers, dishwashers, rice cookers, garage door openers, etc.

The “networked” product

4
The “networked” product could be called an extension of the standby product, because:
a) the product not only can be controlled by the user, but also by other external sources, e.g. a
service provider of other products in the network, and
b) the product can communicate with external sources, e.g. a central heating boiler calling a repair
service.
Because this type of product is – by definition – connected to a network, the network protocol is an
important technical issue.
The consequences of the network connection is that the product becomes more complex in several
ways. Firstly, networked products can be controlled from various sources, not only the user but also

1
Not to mention the philosophical debate whether a product should do something “useful” in the on-mode; and that if a product
is not used by anybody it is not on and therefore it’s power consumption should be (near) zero.
2
Since we have defined “on” in a functional way, i.e. providing one of the main services that the product is acquired for, cold
appliances and storage water heaters are always on, even though their power consumption in on mode will vary greatly
because of components (compressor, heating element) are turned off and on as a result of temperature controls.
3
From this example it does not follow that efficiency improvement in alarm systems are not useful.
4
And as such it can also have an off mode and/or simple standby mode(s).

974
other appliances and network providers. Secondly, the software can be upgraded through the network
which means that the functionality of the product can change during product life. New hardware
architectures take full advantage of these features. Thirdly, data needed for the functioning of the
product can be updated through the network or is not stored at the product at all but in a network
attached storage.
This complexity also strengthens the dependency upon the network connection, which in turn focuses
the attention of appliance and network manufacturers and service providers to provide faster and
more reliable network connections. In the case of PCs this has lead to the concept of “server based
computing” where the PC can’t function (i.e. provide functions like word processing, spreadsheets and
e-mail) without a network connection. This concept was established for a long time, but the practical
application depended on the availability of network connections with sufficient speed, e.g. DSL.
While these products are still limited in number in the typical household at this stage, this is certainly a
trend that is likely to be ubiquitous in the future and hence it is critical to set up measures to ensure
that energy management becomes an integral part of these products in the future.

Overview of products and consequences

The table below provides some examples of products in the 3 categories and some trends.

Overview of products (examples and trends)

on/off standby networked
external power supply
lighting lighting with remote control
white goods white goods with remote control cold appliances with display and
network connection
vacuum cleaners robotized vacuum cleaners
televisions televisions with integrated set-
top boxes (decoders)
VCR, DVD recorders DVD recorders with internet
connection
digital television adapters set-top boxes
personal computers
stereo equipment streaming clients (internet)
printers
monitors
stoves with clock
cordless phones
garage door openers

From the table above it can be concluded that the classical standby product is going to disappear
from the market, or at least will become of less importance, because these products will be replaced
by networked products. Networked products are the next generation products. However, traditional
policy instruments fail with these type of products. Firstly because a single standby mode cannot be
defined and if it could be defined, it is doubtful whether the product will spend any time in it. Finally,
these products might not have a standby mode at all and they may be in on mode all the time. More
probably there will be a sliding scale from off to fully-functional on where it is a question of definition
when the appliance is going from a standby to on mode. I.e. how much “on” should the appliance be
before it is defined as “on”. The target is to have the minimum energy consumption possible in all the
modes as a result of good power management.
Finally the discussion in this section provides clearance on the ongoing discussion on the definition of
standby. First it is clear from the foregoing that no single definition of standby mode exists that
captures all the modes discussed in the typology above. Second, the definition of standby in
5
IEC62301 is ambiguous. For some products it will be the off mode, whereas for other products it will
be the stanby mode (with the lowest power consumption). Furthermore, this definion captures –
deliberately – only one mode, whereas we have seen that several modes can be important. Third,

5
Clause 3.1: Standby mode: lowest power consumption mode which cannot be switched off (influenced) by the user and that
may persist for an indefinite time when an appliance is connected to the main electricity supply and used in accordance with the
manufacturer’s instructions.

975
from an energy efficiency policy perspective it is important that all modes (off, standby and on) are
covered and that policy is not restricted by definitions.

Approaches to deal with standby

From the analysis in the foregoing section we conclude that approaches to deal with standby need to
match the type of product. In this section we present approaches to deal with the different product
types. The emphasis in this section will be on technical solutions regarding the product. It is our
opinion that behavioral solutions to the standby problem, e.g. asking consumers to switch off the
appliance, are at least less effective than technical solutions. The reasons for the product types are
the following.
For those on/off products where switching off is an option, the off mode power consumption can not
be influenced by the user unless the user unplugs the external power source of the product or an
(extra) hard switch is added which performs the same function. Given the fact that this applies to
many dozens of products in the home (e.g. all products with external power supplies), it is in our
opinion too optimistic to expect users to users to unplug all devices when not in use. Moreover,
technical solutions are possible that reduce off mode consumption to 0.1 W or lower and therefore the
behaviour of the user is irrelevant from an energy efficiency point of view.
For classical standby products which can be switched off by the user, it has already been remarked
that this action negates the standby function. Although some consumers may not mind this, it is not
an acceptable option in terms of functionality in some cases. Furthermore, if such products are
(gradually) disappearing from the market, a behavioral solution will become less important.
Because the networked product is dependant on the network connection and given that the network
connection is lost when switching off the product, a behavioral approach for networked products is not
a viable solution. Furthermore, the consumer does not always know whether a product can be
switched off or to standby, because other products might need the services of the product.

Dealing with the off mode

For products that have an off mode, the off mode power consumption should be minimized and close
to 0 W. The rationale for this goal is simple: avoid wasting energy. If the product is in the off mode and
– by definition – is performing no function, it should also use no energy. Note that this does not
indicate anything about how long the product is in the off mode, nor that the product should have an
off mode.
From a technical point of view, the reason for power consumption in the off mode is the on/off switch
of the product being on the secondary side of the power supply, or the product having an external
power supply that can be disconnected from the product while the power supply is still connected to
the mains (e.g. mobile phones). In both cases the off mode consumption consists of losses in the
power supply. Technical solutions exist to reduce the off mode power consumption to 0.1 W or less.

Minimal standby power consumption

First it should be noted that "standby" mode generally offers some level of functionality to the user. So
equipping a product with an on/off switch is not considered to be a solution towards minimal standby
power consumption because in the off mode the standby functionality is lost. This does not mean that
“standby” products could not have an on/off switch, but with a standby power consumption of 1 W or
less, the rationale for having an on/off switch is less clear.
Solutions for standby mode should minimize standby power consumption. In general the following
options exist: decrease the number of components that are powered, increase the efficiency of the
components that are needed for the standby function(s), and add special “standby” components, e.g.
a small efficient standby power supply for sensors or clock functions.
For simple standby functionality that applies to the typical “standby” product, e.g. a clock display or
waiting for a signal from the remote control, solutions that require 1 W or less are available (technical
solutions to less than 0.1W are available in many cases). However, this depends on the level of
functionality required and whether there are other functions such as battery charging etc.

Power management for the networked product

For the networked product the goal is to optimize/minimize total energy consumption. In many cases
the solution is to introduce one or several modes (with different levels of functionality) and power
management. Like the standby product, equipping a networked product with an on/off switch is not
considered to be a acceptable solution for the users, because in the off mode the standby
functionality (including network connection) is lost. However, if the lowest level of standby power for

976
networked products that is likely to persist for long periods is going to be higher than the 1 Watt level,
an on/off switch may be a good option for those consumers that do not always need the functionality
of the networked product and therefore want to switch off the product at their discretion (although in
reality this may be rarely used in many installations).
Power management is the key feature to optimize total energy consumption of networked products.
Power management can be defined as a function of a product that ensures – without user
interference – that the product is always in the state with the lowest power consumption related to the
6
required functionality . Technically speaking this means that if the product is not or only partly used,
unused parts are powered down as far as possible, and only those parts are powered that are needed
to detect the need for an increased level of functionality for the product. This requires the following
activities from the product:
• monitoring activity levels of parts of the equipment (devices)
• decision rules to enter a certain state: demands for a higher state, falling to a lower state where
current level of functionality is not required
• execute state transitions and monitor the result

Power management is a concept that is broader than standby, because also in the on mode power
management can be beneficial. Examples of power management are the cd-player that switches into
standby after the disc is finished (and unless the user has activated a repeat function) or the digital
television that switches from a low standby mode to a mode where it can update electronic
programme guide data and switches back afterwards.

Standby: the next generation policies are needed

In this section we will translate the approaches of the foregoing section into policies. In general, a
variety of policies exist for increasing energy efficiency of appliances: financial instruments (subsidies,
tax rebates), informational instruments (labelling, product information via internet) and “product”
instruments (e.g. minimum efficiency standards). Regarding standby we will focus in this paper on
product instruments, where – either voluntary or mandatory – products that do not meet certain
standards or rules are not brought on the market. The reason for this restriction is the following. In
general standby consumption of a single product is relatively low compared to the total energy
consumption of a television or washing machine. The impact of standby originates from the total
volume of products that have standby consumption. An example: 45 million DECT cordless phone kits
with on average 2 handsets per kit in Europe have an associated 90 million linear external power
supplies. One power supply uses more than 1.5W in standby, this could viably be less than 0.3W, so
cumulatively more than 1TWh of energy is wasted per annum. This means that informational
instruments (directed at the buyers/consumers) will have only limited impact, while bearing the full
cost of them. A typical home may have 50 to 100 products which may potentially have some power
consumption in other than on mode - in most cases the amounts of power are small and the
transaction costs of obtaining comparative information on differences in standby and taking this into
account during the product purchase is not viable. Also financial instruments targeted at buyers would
be restricted to relatively small amounts (because relative to the savings) and therefore have little
effect. An alternative could be financial instruments targeted at manufacturers, which may be a
problem when manufacture also or mainly occurs outside the territory which is financing the measure.
Therefore in this section we will consider only (voluntary and mandatory) product instruments.

Voluntary versus mandatory instruments

To date in Europe a voluntary approach has been implemented to reduce off mode and standby mode
power consumption of selected products, starting with the official communication from the
Commission in 1999 [4]. It was assumed that a voluntary approach was faster, more flexible and less
costly. In previous years several voluntary agreements have been negotiated with industry, most
7
notably on consumer electronics and external power supplies . Furthermore, the ecodesign directive
which was passed in 2005 [3] now explicitly opens the future possibility for voluntary initiatives as an
option to forestall (mandatory) implementing measures.

6
Although power management is discussed here with regard to networked products, it can be also useful to optimize energy
consumption of other types of products.
7
Energy Star programme for office equipment.

977
Experiences so far with self-regulation regarding energy efficiency of products in Europe reveal that
the advantages mentioned above need to be put into perspective. The following table describes a
number of self-regulation initiatives.

Overview of self-regulation in Europe

Self-regulation Products covered Level 2008 (lowest Representativeness
standby mode) (market coverage)
Code of conduct external Single (output) voltage 0.3 W – 0.5 W < 50 % (coverage is good
power supplies external ac-dc and ac-ac for mobile telephones and
power supplies in the laptop computers)
range between 0.3 W
and 150 W.
Code of conduct Digital Set-top boxes, adapter 2 W – 6 W ?
TV Service Systems boxes, IDTV
Code of conduct Broadband equipment not yet decided n.a.
broadband equipment
(planned)
Industry Self- CRT-TVs 1W 70 %
Commitment to improve non CRT-TVs 1W 50 %
the energy performance DVD-players 1W 50 %
of household consumer
electronic products sold
in the EU (EICTA)
Energy Star System units (PCs)* 2W currently >80 %
(* planned in new monitors 1W new specifications: > 60 %
specifications) printers* 1W
copiers* 1W
CECED Unilateral cold appliances covered by duty cycle >80 %
Industry Commitments washing machines not covered n.a.
dishwashers not covered n.a.

From this table we draw the following conclusions.

• Not all products groups are covered by self-regulation.
• For those products where a voluntary agreement has been developed, market coverage is
generally much less then 100 %. A reason for the incomplete market coverage is that for some
products (most) manufacturers are outside the EU; these manufacturers might be not aware of
the voluntary agreements or might not care. If the share of manufacturers outside the agreement
is (too) large, then other manufacturers are less or not willing to sign up because it provides them
a small disadvantage, which is relevant for products where every € cent counts. Thus, voluntary
measures do not create a “level playing field”.
• Finally, voluntary initiatives are re-active, i.e. negotiations only start when the new product variant
has been on the market for some time.

Regarding the on/off product and the standby type of product mandatory measures are preferable.
Learning from the experience with current voluntary agreements, a pro-active and horizontal
mandatory approach should be followed, meaning that a measure should include as many product
groups as possible. Since standby is a basic feature and independent of the functionality of the
product in other modes, mandatory measures can be pro-active. This means that the measures
should also apply to product variations that at the time of the measure coming into force were not (yet)
on the market. TVs provide a good example: a mandatory standby level for CRT TVs can also be
used for LCD and Plasma TVs. A pro-active approach provides a clear guideline for product
development at manufacturers, because whatever product they develop it should comply with the
standard.
However such an approach can not yet be used for networked products. The two main reasons are
that no general applicable standby modes can be defined for these type of products, or the (simple)
8
standby mode that could be defined is not relevant . Secondly, the concept of power management as

8
However, because of the horizontal approach mandatory measures will also cover networked products. So in case (some of)
these products have a simple standby mode they should comply with these measures.

978
such is still too general to be put into law. In the next sections the mandatory approach is elaborated
upon by the example of a horizontal implementing measure for standby in the framework of the
ecodesign directive [3], whereas the voluntary approach for networked products is illustrated by the
EU Code of Conduct for Digital TV Service Systems [5].

Simple standby and off: a horizontal implementing measure

Standby is one of the priority items regarding energy efficiency and explicitly mentioned in the
ecodesign directive, to be covered by an implementing measure. However, an implementing measure
on standby differs in a number of aspects from an implementing measure on a specific product. In this
section first several specific aspects regarding standby related to the ecodesign directive will be
discussed. Then will follow considerations in preparing a draft implementing measure and criteria for
an implementing measure.
Specific aspects regarding standby
Standby consumption is a specific ecodesign requirement. Standby consumption is not an Energy
using Product (EuP) itself, but a characteristic of an EuP. In relation to an implementing measure for
standby, two questions have to be answered:
a) Is a horizontal implementing measure, i.e. an implementing measure covering one (or more)
requirement(s) for – in principle – all EuPs, possible within the current framework directive?
b) Once a horizontal implementing measure is adopted, are other (horizontal) implementing
measures allowed for the EuPs for which the horizontal measure is adopted?

re a): According to Annex VII, article 1, the implementing measure can specify more than one type of
EuP. Article 2 of the Annex VII leaves the choice to specify one or more specific ecodesign
requirements. However, the criteria for EuP, the considerations whether to prepare a draft
implementing measure, the rules for preparing a draft implementing measure and criteria for
implementing measures require analysis per EuP.
re b): The framework directive does not forbid several implementing measures for the same EuP, so
we assume that other implementing measures are allowed.
The purpose of having a horizontal measure on standby is to reduce the energy input in the standby
mode. However, product definition plays an important in various criteria, e.g. the volume of sales and
trade. Annex VII, article 1, requires the exact definition of the types of EuPs covered. How exactly
should this be defined? For a horizontal measure on standby it seems reasonable that the definition
only discriminates between product types where different standby behaviour is defined.
So, in principle the products covered are defined by the type of standby consumption they have (ie
the functionality of the modes). A horizontal implementing measure on standby should cover both the
9
following modes :
• the “off” mode (the product is connected to the mains, no user function is fulfilled)
• the simple or lowest power “standby” mode (product is connected to the mains, is not in the off
mode and not in a main functional mode)

Criteria Energy using Products; preparing a draft.

A horizontal measure on standby could cover in principle all electrical products in homes and offices
that have standby consumption. Article 15(2) provides the criteria that have to be met for an EuP to
be covered by an implementing measure:
a) volume of sales and trade of more than 200 000 units a year
b) significant environmental impact, considering the quantities placed on the market
c) significant potential for improvement without entailing excessive costs; subcriteria:
• absence of other relevant EU legislation
• failure of market forces to address the issue
• a wide disparity in the environmental performance of EuP available on the market

In conclusion this means that implementing measures will be put in place for mass produced energy
using products that have a significant environmental impact and a significant potential for
improvement not yet targeted by other measures.
Both environmental issues and current relevant self-regulation lead to the conclusion that an
implementing measure is needed. The environmental argument points to standby as one of the

9
If applicable. As indicated before not all products have both an off mode and a standby mode. The definition of the “lowest
power standby” is inspired by IEC63201 and provides a solution for products that have more than 1 standby mode.

979
priorities to deal with. Analysing current self-regulation reveals that the estimated market coverage is
on average roughly 50% and therefore is not complying with the criterion of representativeness as
specified in Annex VIII of the Ecodesign directive. The wide disparity must be seen in the perspective
of the low power levels of standby consumption: from 2 W to 1 W is a 50% improvement. The impact
results from the large quantity of products on the market, not from the absolute value of the savings
per product.
Since the implementing measure on reduction of standby power consumption is a specific ecodesign
requirement (and a horizontal program measure), a general life cycle analysis would be sufficient. The
rationale for having a horizontal measure for standby power consumption results from the possible
improvements in energy efficiency for this mode in general, and not on decisions regarding each
product based on an life cycle analysis.
It might be sufficient here to state that standby consumption in general can be reduced while other
environmental aspects are not affected – or even improved (e.g. less material use in case of switch
mode power supplies).
Criteria for an implementing measure (Article 15(5))
Implementing measures shall meet all the following criteria:
a) there shall be no significant negative impact on the functionality of the product, from the
perspective of the user;
b) health, safety and the environment shall not be adversely affected;
c) there shall be no significant negative impact on consumers in particular as regards the
affordability and the life cycle cost of the product;
d) there shall be no significant negative impact on industry’s competitiveness;
e) in principle, the setting of an ecodesign requirement shall not have the consequence of
imposing proprietary technology on manufacturers;
f) no excessive administrative burden shall be imposed on manufacturers.

In general the environment and consumers benefit from lower power consumption and thus lower
energy consumption and – in most cases – lower carbon dioxide emissions. Studies show that
reduction of standby power in general can be achieved at no or little extra costs, e.g. a maximum of €
1 per product (consumer price). With a reduction of 0.5 W to 1 W in standby power consumption,
these costs can be recovered 2 to 4 years (assuming an average electricity price of 0,10 €/kWh). In
practice the costs for consumers will be (near) zero, since they will be “absorbed” in the redesign of
the product.
The impact on competitiveness requires special attention. In the foregoing section a remark was
made on impact on costs. Many – or maybe most – of the products or components (e.g. power
supplies) that would be affected by a horizontal implementing measure on standby are produced
outside the EU. In that case the factor of competitiveness is less of a problem for EU industry because
only a small number of manufacturers may be involved. Because an implementing measure is
mandatory for every product, it creates a level playing field; in fact minimum efficiency levels could
foster innovation and provide competitive advantages for those manufacturers that find clever ways to
meet the criteria with less costs.
Since for the EU a level playing field would be created, the question of competitiveness is related to
the question whether other parts of the world have mandatory measures for standby. In several other
parts of the world (Australia, Korea) mandatory measure are in place or planned. If other parts in the
world do not have mandatory measures, either – depending on the costs – manufacturers will run
specific lines for products that meet the EU implementing measure (and which may cost slightly more)
or manufacturers will change all of their production. In the first situation Europe will benefit from the
efficient products, where other parts of the world may get the balance of non-complying (cheaper)
products. In the other case – when other parts in the world have mandatory measures – it could be
the other way around: Europe will get the less efficient (cheaper) products. Therefore, where other
parts of the world have mandatory measures, an EU implementing measure is an important protective
measure.
Several techniques exist to achieve a one Watt standby solution, so no proprietary technology is
imposed on manufacturers. For power supplies see e.g. the websites of several manufacturers of
(components of) power supplies. In fact, in most cases technical solutions to reduce standby to a
fraction of a Watt are possible.
The administrative burden for manufacturers depends on the way the conformity check is organized.
Since the ecodesign requirements are integrated into the CE marking process, the administrative
burden should be minimal.

980
Conclusion
The ecodesign directive can be used to make an implementing measure on standby (and off mode)
mode consumption for various product groups of energy using products.

Networked products: EU Code of Conduct for Digital TV Service Systems

The Television Broadcasting sector of consumer electronics is currently undergoing radical changes.
The rapid development of a major communication network to support digital television is
complemented by continuous developments in the functionality of the reception hardware, giving the
consumer:
• Major improvements in the quality of the audiovisual presentation of broadcast services.
• For the first time, full interactivity with the content and source of the signals.
• A combined entertainment and communication platform with access to the full Internet or to
“walled garden” information services. So digital television finally closes the perceived gap
between the “lean forward” solo-working tool of the PC and the “lean back” group entertainment
device of the TV. This convergence will be the catalyst for significant lifestyle changes in all
levels of society, not just the information rich.

The technology supporting these changes is developing at an unprecedented rate. One consequence
of this is that the relatively slow and costly, manufacturing and marketing cycle of the mass-produced
TV cannot viably accommodate the accompanying rapid changes in the technical specification of the
hardware. An independent signal interface and data processing platform, the STB (Set Top Box) has
been the preferred manufacturing and market distribution solution. This device readily interfaces with
existing and developing TVs and display systems and allows the rapid modification of functionality
specifications in high volume production. The STB is an good example of a networked product
because it is connected to a network and not only the user but also the service provider need to
control the box, e.g. for software updates and security reasons.
The downside of this solution is that the existing voluntary agreement and labelling mechanisms for
energy efficient domestic electronic products are too slow to keep up with STB development and
could potentially hamper that development. In 1997 a European Commission working group identified
the digital service system STB as the domestic electronic device with the largest potential to increase
the energy consumption requirements of European Households [6]. Research into proposed
development showed that by 2010, the STB could push domestic electronic energy consumption in
Europe above that of fridges and freezers (especially as these are decreasing over time as a result of
energy labelling and efficiency standards programs). With 150 million of these boxes across the EU -
equivalent to one per household – the annual electricity requirement for digital service systems could
be around 60 TWh (close to the total electricity consumption of Denmark for all sectors). This
electricity would also release 24 MtCO2, which would have a significant impact on the EU’s ability to
meet its overall Kyoto CO2 reduction target.
To challenge and resolve this problem the European Commission set up a working group of the key
stakeholders in digital service system development – manufacturers, silicon providers, service
providers – and energy agency specialists to discuss and specify a (voluntary) Code of Conduct that,
amongst others, contains power consumption targets for various types of STB.
This activity has become an excellent example of a product policy initiative that united stakeholders
early enough to impact the design process before the product became ubiquitous. Furthermore the
Code of Conduct has inspired other parts of the world to set efficiency standards for STB; to support
international cooperation a “community of practice” has been set up by the Australian Greenhouse
Office.
The Content of the Code of Conduct
The principal aim of the working group is to reduce the energy consumption of the STB through the
setting of agreed, practicable power requirement targets in a defined development timescale. To that
end, a voluntary agreement or Code of Conduct was devised which Europe’s principal STB and TV
10
manufacturers and a major Service Provider, B-Sky-B currently support .
The Signatories of this Code of Conduct would make all reasonable efforts to:
• Achieve the power consumption targets set out in bi-annualy reviewed tables for new stand-
alone products and for digital TVs with built-in IRD placed on the market after agreed dates.

10
For a full list of signatories see: http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm

981
• Support and contribute to the development and acceptance through an ad-hoc Task Force of the
Common Power Management Guidelines.
• Co-operate with the European Commission and Member State authorities in an annual review of
the scope of the Code of Conduct and the power consumption targets for two years ahead.
• Facilitate and encourage consumers to adopt energy efficient practices in connection with the
use of digital TV services.
• Co-operate with the European Commission and Member States in monitoring the effectiveness
of this Code of Conduct.
• Ensure that procurement specifications for digital TV services, systems, equipment and
components are compliant with this Code of Conduct.

In this context, the Code of Conduct identifies a key tool to the achievement of significant energy
efficiency targets in digital service system platforms: the development of effective power management
in the silicon for the principal functional blocks.

Service Provider Needs and STB Power Architecture; Power Management

Effective power management to maximise energy efficiency can only be achieved after consideration
of the permitted operating modes. The user may only be aware of the states on and standby but the
functionality in these states may vary greatly depending on the requirements of the service provider
and the delivery medium - cable, satellite, terrestrial or DSL.
In the on state, digital set top boxes provide the basic function of decoding of television pictures and
many have on-screen, interactive information services. Other services such as electronic shopping, e-
mail delivery, Internet access, games and software download may also be available. As the digital TV
market develops, service providers are keen to offer further premium services which place additional
demands on the hardware and increase power consumption. These services could include TV
recording, video on demand, telephony, home networking and automation with wireless interface to
peripheral devices. Broadband platforms for new broadcast and communication networks based on
wired and wireless LAN (Local Area Network) will add to this energy requirement load. To counteract
the resulting increased energy demand and follow the Code of Conduct, close attention needs to be
given to efficient power conversion, distribution and the management of power usage depending on
the function required. More specific designers and stakeholders should:
• Choose the lowest power standby mode consistent with service provider requirements. Decide
whether adequate power management of each circuit block can be achieved by control of the
silicon/software itself or if switching of power rails is needed.
• Consider power consumption and in-built power management features when choosing silicon.
Involve software designers from the outset so that energy efficient software architecture and
power management are incorporated in the early design. Check if any third-party software, which
may be used for the operating system or conditional access, supports power management.
• Ensure that power to peripheral ports and devices can be turned off when not required.
• Encourage the rapid standardisation of “intelligent” interconnectivity to ensure that any
component in the home entertainment and communication network automatically adopts the
lowest power requirement for the level of activity required.

Results of the Code of Conduct; relation with ecodesign

The EU Code of Conduct is an important platform for promoting energy efficiency in digital TV services
in Europe. The Code of Conduct has already reduced the energy consumption of STBs, even if these
offer many more features and services. (see figure 1).

982
 Power consumption set-top boxes

25

20

15
W

10

0
2001 2002 2003 2004 2005

Standby passive Standby active On

Figure 1: Power consumption STB (as reported by signatories of the Code of Conduct)

However, up to now the Code of Conduct fails to address manufacturers of simple, low cost STB, so
called digital TV adapters for legacy analogue TVs and VCRs. In many cases there products do not
have an off mode and the power consumption in the standby mode is almost equal to the power
consumption in the on mode, due to poor power management. Since the standby mode in these
adapters is of the simple type, a mandatory implementing measure within the framework of the
ecodesign directive could be used to regulate the digital TV adapter. The same measure should also
target the on mode consumption, since it can be expected that in many cases the adapter is left on,
e.g. because it is also needed when recording programmed broadcasts with a VCR.

Conclusions and discussion

In this paper we have argued why standby consumption is no longer a simple issue for many products
and will become even more complex in the (near) future where more products will evolve into
networked products. In our opinion this will be the case for many of the current “classical” standby
products, e.g. televisions, video recording appliances, but also for new standby products, e.g. fridges
with a screen and internet connection, central heating boilers that call the service centre if
maintenance is needed. On the other hand simple standby products, including products with off mode
(mostly with external power supplies), are also expected to grow in number, although in practice these
products might never be in the off mode.
The current generation of policies, which mostly deal with the simple type of standby (i.e. what we
have called the on/off products and the standby products) on a voluntary basis, are not adequate to
solve the standby problem for many current product types and perhaps the majority of future products.
We have suggested that for the simple type of standby a mandatory horizontal approach should be
followed, whereas for the complex standby (found in networked products) voluntary policies targeted
at power management should be followed (see table below).

Policy options for reducing standby consumption

voluntary X
X
(power management)
Policies
mandatory X
(horizontal)
simple complex
(including off)
Type of standby
X: current prevailing policy
X: suggested policies

983
A horizontal measure should cover both off mode (product connected to the mains, no function
fulfilled) and the simple standby mode (product connected to the mains, not in off mode and no main
function fulfilled). The maximum power level for off mode could be set at < 0.3 W and the maximum
level for the standby mode at 1 W. The ecodesign directive could be used for such a mandatory
horizontal measure. Such a measure would cover also the off mode and the simple standby mode of
networked products, if applicable.
Effective power management in complex products or where a product function is shared with other
products, will demand the unprecedented involvement of a wide cross section of world-wide
manufacturing industry. The object will be to establish basic protocols for all energy using products to
automatically communicate their function and required status in a network or chain of products. The
translation of this work into standards that are dynamically reviewed with product development is,
from experience to date, impracticable in a mandatory regime.
An aspect of power management that may perhaps be mandated in the support of the simple standby
state would simply be the requirement that when a product recognizes that its primary function is
complete and no other subsidiary functions are essential, it automatically goes to the lowest (simple
standby) power state.
From the foregoing conclusions, the following items remain open. Firstly, a horizontal measure that
covers off and standby mode does not automatically mean that products covered should have an off
and/or standby mode. For many product types, several modes may be present (and which ones are
covered?) and similar products have different configurations. Care is required to ensure that
elimination of a particular mode for a product is not an option to avoid regulatory requirements.
Secondly, extrapolating from the discussion on the voluntary approach regarding the simple standby,
it is likely that a voluntary approach regarding the complex standby will not cover all of the market.
Thus, the question arises whether in due time power management should and can be made
mandatory.

Acknowledgement
The authors thank Lloyd Harrington for his comments on the first draft of this paper.

References
[1] Meier, A. Standby: where are we now? Proceedings of the ECEEE 2005 Summer Study,
Mandelieu, pgs 847-854
[2] Siderius, Hans-Paul. Overview of Technical Solutions to Reduce Standby Power Consumption.
rd
3 International Workshop on Standby Power, Tokyo, 7/8 February 2001
[3] Directive 2005/32/EC of the European Parliament and of the Council of 6 July 2005
establishing a framework for the setting of ecodesign requirements for energy-using products
and amending Council Directive 92/42/EEC and Directives 96/57/EC and 2000/55/EC of the
European Parliament and of the Council
[4] European Commission, Communication from the Commission to the Council and the European
Parliament on Policy Instruments to Reduce Stand-by Losses of Consumer Electronic
Equipment, COM(1999)120 final, Brussels, 15.03.1999
[5] Code of Conduct on Energy Efficiency of Digital TV Service Systems – Version 2a, Ispra, 24
November 2004; available from: http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm
[6] Molinder, O. “Study on Miscellaneous Standby Power Consumption of Household Equipment
light-audio-white goods).” In Proceedings of First International Conference on Energy Efficiency
in Household Appliances, Florence, Italy, 1997

984
Quantification of Standby in Australia and Trends in Standby for
New Products
Lloyd Harrington1, Jack Brown1, Paul Ryan2
1
Energy Efficient Strategies, Australia
2
EnergyConsult, Australia

Abstract
Standby power consumption of household products in a range of low power modes has been on the
policy agenda for around 10 years and is an area of growing concern. Australia has been active in
both national policy development issues and in the development of an international test method for
the measurement of standby power.
To support the development of sound long term policies, a range of data collection efforts have been
commissioned, including the measurement of all products present in a sample of Australian
households in 2000 and 2005. The low power attributes of a wide range of products have also been
measured in retail outlets.
Data from a range of sources suggests that the trend in standby power consumption for some
products is improving over time. However, there are a number of products that have very poor
standby attributes such as set-top boxes and hard drive DVD recorders, integrated stereos computers
and telephony equipment. Major appliances of course are of interest but their total contribution to
household standby is only very modest.
One of the most alarming findings was that the number of devices connected to the mains and which
used some power when not performing its main function climbed from 21 in 2000 to 27 in 2005 – this
is an increase of 35%.

Background
The issue of standby power has been discussed in energy policy circles now for nearly 10 years.
Interestingly, the power consumption of appliances and equipment when not in use (i.e. while NOT
1
performing a primary function ) was not a policy consideration in the early 1990s, probably in part
because few products used power in these modes at the time, but also due to ignorance. It was not
until some laboratories noticed in the mid 1990’s that some products were in fact using small amounts
of power when off or in “standby mode” that the issue was canvassed more widely and further
investigations were undertaken. Through the late 1990’s the issue was raised within the IEA and a
series of three international workshops were held up to 2001. This culminated with the publication of
the report Things That Go Blip In The Night [1].
In the late 1990’s only anecdotal information was available on the likely magnitude of standby power –
almost no data from the wide range of equipment installed in households had been collected in any
country. Theoretical estimates ranged from a few percent of household electricity consumption to 10%
or more. There were growing concerns about the issue but there was little data on which to base firm
policy actions.

Policy Development is Australia

In order to provide a sound basis for future policy decisions, Australian state and federal governments
commissioned a survey in 2000 of some 64 households to establish the magnitude of standby power
in Australia and to help identify key products and potential problems. The survey measured every plug
load present and documented the attributes of some 2400 products actually installed in these homes.
The report, which was published in early 2001 [2], found that “standby” (in a very broad sense)
accounted for around 11% of residential electricity consumption in Australia or an average of about 90
Watts. While this figure included a number of end uses and low power modes that were not

1
Many products, particularly those with electronic controls, have a number of possible states other than the primary function
and some have several primary functions.

985
traditionally regarded as “standby” by some commentators, it was clear that the issue was of
significant concern and required a coordinated government response.
In August 2000, the Council of Commonwealth, State and Territory Ministers in charge of energy
matters, endorsed a program of work to lead Australia towards achieving the goal of “One Watt” for all
2
consumer appliances and office equipment . They agreed to develop policies designed to ensure the
maximum standby power of all household products manufactured in or imported into Australia is One
Watt. Australia was the first national government to agree to this formal target. This statement of
principle sent a clear message to industry and provided a coherent structure for a diverse range of
policies designed to combat excessive standby energy consumption. In 2005 Korea also announced
mandatory 1 Watt targets for a wide range of products [3]. 1 Watt targets are now common in many
programs that cover standby like the Energy Star program.
The Commonwealth Government in Australia went further and announced in October 2001 a policy to
purchase only equipment that complies with the US Environmental Protection “ENERGY STAR”
standard, where it is available and fit for the purpose.
International work on a test method for the measurement of standby power progressed within the IEC
which culminated in the publication of the test procedure IEC 62301 Household electrical appliances -
Measurement of standby power in mid 2005 [4]. This test method provides general approaches,
methodology and equipment required to measure power consumption for a range of low power modes
for most product types. Australia contributed to this work through the provision of the Convener of the
IEC TC59 working group and active participation at all stages of the project. The IEC test procedure
does not define relevant low power modes for any equipment types – this is a task for the relevant
product committees. While there has been some progress in this area, some manufacturers are
resisting further developments at the product level which will hamper the development of uniform
product mode definitions into the future.
The development of a ten-year strategy is another key element of the 2000 commitment by the
Australian Government to stamp out unnecessary standby power. The document titled Money isn’t all
you’re Saving – Australia’s standby power strategy 2002 – 2012 [5] is the culmination of considerable
industry and community consultation and the plan:
• sets out a long-term plan for the measures to combat excessive standby consumption;
• identifies some 40 product types initially targeted in period 2003 and 2004 and the process for
identifying future products;
• develops the procedures for setting standby targets; and
• outlines the potential sanctions that could be applied should suppliers not meet targets.
Some 30 product profiles have now been publicly released. These profiles identify a range of potential
standby program measures for each product type and set voluntary standby targets. Many product
types are to be assessed against their voluntary standby target in the period 2006 to 2008. The
profiles can be found on www.energyrating.gov.au
The balance of this paper documents the efforts to date to quantify standby energy consumption of
products already installed in residential homes as well as the attributes of new products offered for
sale in Australia. This data is crucial in the assessment of whether products have achieved the
voluntary targets set by government and whether more stringent actions (such as regulation) are
warranted for some products.

Data Collection Efforts in Australia

As noted above, the first intrusive survey of Australian homes was conducted in 2000. The survey
covered 64 homes and involved direct power measurements of some 2,400 products in a range of
modes using a power meter that was accurate down to 0.1 Watt. A second intrusive survey was
conducted in late 2005 and the sample was expanded to cover 120 houses and data on 8,000
individual products in various modes was recorded and measured [6].
From 2001, Australia has been tracking the low power mode power consumption of new products on
display in retailers. The aim of this work is to benchmark standby performance and trend for each of
the major product types which can be used to assess progress by manufacturers towards voluntary
standby targets that have been set by governments in Australia. A database of some 5,500 products
has been accumulated over the years, with measurements in up to 4 different modes for some
product types. This dataset now one of the most comprehensive sets of standby measurements in the
world [7].

2
While a notional target of 1 Watt across the board has been adopted, actual targets vary by product and mode, depending on
what is technically feasible based on good design practices.

986
The key results from these two data collection methodologies is presented below. In the context of
this paper, the term standby power is intended to be very broad and covers the energy consumption
of a wide range of products in a wide range of low power modes.

Intrusive Surveys 2000 & 2005

The intrusive residential surveys were undertaken to quantify the standby power of products in
Australian households. Several modes were measured for most products. This survey was primarily
concerned with quantification of the stock of products currently installed in households. The general
objective was to quantify the magnitude of standby power used by all appliances and equipment
found in Australian residential households in 2000 & 2005, in their applicable low power modes.
The average standby consumption for 2005 was found to be 92.2 Watts per household, or some 807
kWh per year. This is around 10.7% of Australia’s residential electricity consumption in 2005. This
level of residential standby power consumption was estimated to cost Australian consumers
approximately AU $950 million (at an average retail tariff of 15c/kWh) which will result in nearly 6.5
million tonnes of carbon dioxide in 2005.
The total number of products that run on mains power in a typical house in 2005 was found to be 67
(although this ranged from as few as 16 to as many as 136 items per house). During the survey,
about 28% (19) of all products were in fact found to be unplugged at the time of the survey. These
products were assumed to contribute no power to the total average standby power estimated for all
households.
Of the 48 or so products that were found to be plugged in at the time of the survey in 2005, on
average around 56% (27) used some low mode power when plugged in and not performing their main
function. This is an increase from 21 in 2000. More discussion on usage is included below.
The figure of 92.2 Watts per house could be seen as alarming. However, this is comparable to a study
of homes in California which found 108 Watts (but used a different methodology) [8]. However, if all
products were in fact plugged in and left in their highest recorded non operating mode, then this
power consumption would be as high as 178.3 Watts per household. If all products found had a
maximum non operating mode power consumption of 1 Watt or less, then this power consumption
would fall to 32.3 Watts per household.
The most important product groups in terms of their total contribution to standby are computers and
peripherals, home entertainment equipment (including televisions and set-top boxes), major
appliances and other office equipment. An important issue not addressed by this study is the need for
a consistent classification of modes and product types to enable comparison between different
studies in different countries. This issue is discussed by Nordman et al. [9].

Table 1: Contribution to Total Household Standby from Major Product Groups

Product Group Contribution to Number of Items Average Watts
Total Standby W per Household per Item
Major Appliances 11.8 8.5 1.4
Telephones other office equipment 7.1 2.0 3.6
Other home entertainment 19.6 5.0 3.9
Set-top boxes 3.5 0.3 12.1
Televisions 6.2 1.7 3.6
Computer and peripherals 28.1 5.4 5.2
Monitoring and continuous 8.7 7.6 1.1
appliances
External power supplies 3.4 4.7 0.7
Other items with a standby mode 3.9 9.7 0.4
Products unplugged 0 18.7 N/A
Products with no standby mode 0 3.1 N/A
Total 92.2 67.7
Source: [6]. In this context, total standby takes into account the number of products and the modes in which
these products were found or are normally left. 20% of computers are left on continuously.

987
The expected contribution to the total standby power for each of the major product groups is
illustrated in Table 1. Other important parameters such as the number of items per household for each
of the main product groups and the average standby power consumption per item within the product
groups is also shown. When these parameters are considered, the power consumed per item and the
number of items is quite critical for computers and home entertainment equipment. Set-top boxes also
have a large power consumption per item, even though the total contribution to overall standby is still
low due to their low penetration. This is a product that is likely to have a rapid growth penetration over
the next 5 years. The average power per item is also reasonably high for televisions and telephones
and other office equipment.
More detail is shown in Table 2 including ownership data and age of selected items. The mode in
which the product was found is not reported here as this can be one of several modes (see below).
Based on current information, the growth of standby is of the order of 3% per annum, although there
is some uncertainty surrounding this estimate.

Assumed Usage Patterns

The intrusive surveys assumed a simplistic usage pattern for standby modes – the mode in which the
product was actually found was recorded and used for analysis purposes and this was confirmed with
the house owner whether this was the normal state in which the product was left. Where a product
was found in “on mode” at the time of the survey, the householder was questioned on the normal
state or mode when not in use. The detailed survey report gives information about each product and
3
the modes in which products were found. About 28% of products were found to be off at the mains or
unplugged.
Many products have several possible non use modes and these are fully documented in the study. A
simplistic assumption is that the hours of active use for most products is relatively small. One of the
main exceptions is televisions, where viewing hours range from 20 to 40 hours per week and
computers – many appear to be left on for very long hours even if not in active use (about 20% are
left running continuously). Many home entertainment products were left in active standby mode in the
2005 survey, indicating the potential value of power management or auto-off functions for these
products.

3
Unlike the USA and Europe, Australian power outlets have a switch which controls power to each product which is plugged in.

988
 Table 2: Summary of Key Standby Attributes – 2005 Intrusive Survey, Australia
Product Type On Mode Active Passive Off Mode Ownership Age
(Watts) Standby Standby (Watts) (items/ (Years)
(Watts) (Watts) household)
Amplifiers NA 22.4 3.4 0.2 0.36 8.7
Answering Machines NA 3.1 NA NA 0.17 NA
AV Receivers NA 65.3 3.1 0.3 0.13 3.2
Breadmakers NA 1.7 NA NA 0.18 NA
Clock Radios NA 2.1 NA NA 1.33 NA
Clothes Dryers NA 3.3 NA 0.2 0.63 10.3
Clothes Washers NA 5.8 NA 1.9 0.95 7.0
Computer Speakers NA 4.1 6.0 2.2 0.83 3.9
Computers 82.2 NA 35.5 3.5 1.25 3.5
Cordless Phone Base NA 3.3 2.4 NA 0.98 NA
Stations
CRT Monitors 61.7 NA 8.2 1.9 0.86 4.6
CRT Televisions 66.9 NA 7.4 0.1 2.07 7.7
Set-Top Boxes 13.3 NA 10.7 0.0 0.32 1.8
Dishwashers NA 2.8 NA 0.8 0.57 6.6
DVD Players NA 9.0 2.6 0.0 0.82 1.8
DVD Recorders NA 26.5 4.9 NA 0.08 0.7
External Power Supplies NA NA 1.2 NA 6.86 NA
Facsimiles NA 5.9 NA NA 0.26 NA
Game Consoles NA 26.7 1.9 1.4 0.33 3.7
Inkjet Printers NA 4.6 NA 1.9 0.89 4.4
Integrated Stereos NA 18.1 6.5 1.8 0.97 5.9
Laptops 34.1 NA 16.5 9.2 0.54 2.8
LCD Monitors 29.3 NA 2.6 1.0 0.37 1.4
Modems NA 5.9 4.4 2.4 0.68 NA
Multifunction Devices NA 11.2 NA 5.5 0.23 2.3
Non Convection NA NA 3.0 NA 0.77 NA
Microwaves
Portable Stereos NA 5.8 2.2 2.3 0.63 5.3
Radios NA NA 1.5 NA 0.46 NA
Remote Garage Door NA 4.1 NA NA 0.30 NA
Openers
Scanners NA 8.3 NA 0.9 0.25 4.5
Smoke Alarms NA 0.4 NA NA 1.23 NA
VCRs NA 10.5 5.1 1.7 1.11 6.7
Source: [6]

If total annual energy consumption was to be accurately estimated, much more detail on the possible
modes and length of time in these modes would need to be collected for each product. Surveys using
diaries are notoriously unreliable and there is likely to be limited value in this approach (it is also very
onerous on the participant). A more valuable approach would be to use individual end use meters to
track mode usage for different products. If this approach is adopted, it is important to measure the
power consumption of each mode prior to monitoring as it can be difficult to surmise behaviour from

989
power data where information is not known about the product or its possible modes. Special meters
which are capable of measuring low power modes also need to be used in this case.

Store Surveys 2001-2006

The methodology used has been to visit a number of large retail outlets and measure the relevant
modes for all products that are on display for sale. Measurements are made on 500 to 1,000 new
products during each survey. Data is recorded on brand, model, price, details of any energy labels or
markings, features, product type and any other relevant features. Power consumption in defined
modes is recorded, mostly passive standby and off modes, but also active standby and on mode in
some cases.
Initially, the surveys were conducted once annually, but since 2003, two to three surveys have been
conducted per year. This has resulted in a database of more than 5,500 new products offered for sale
in the period.
When all products considered together, each of the modes measured has remained fairly stable over
the period. The results are summarised in Table 3. It should be noted that the number and mix of
products measured in each year of the survey were somewhat different so the results need to be
taken as indicative and trends within each product need to be examined separately. So while this is a
very coarse measure, it does show some interesting overall trends.

Table 3: Summary of average consumption across all products

Mode/Survey 2001 2002 Jan 2003 2003/04 2004/05 Nov 2005
Total readings off 258 380 330 925 782 296
Average off 1.3W 0.8W 0.9W 0.8W 0.7W 1.1W
Total readings passive 440 397 325 682 737 295
Average passive 5.8W 4.1W 3.4W 3.6W 3.7W 3.9W
Total readings active 101 210 216 527 665 333
Average active 11.4W 15.6W 13.9W 14.3W 15.2W 14.7W
Total readings delay start 71 51 28
Average delay start 4.1W 3.5W 4.1W
Source: [7]

Figure 1 shows the distribution of measurements taken for all products in passive standby mode. All
six years of data are presented on the graph. The graph clearly shows that the distribution of
measurements since 2001 has shifted from the higher end (greater than 3W) to measurements less
than 3W. In 2001, 35% of all measurements taken in passive standby were less than 3W, while in
November 2005, 55% of measurements were recorded under 3W. The latest results show 23% of
products are now less than 1W in passive standby compared with only 8% in 2001. The change in
distribution further supports that passive standby has declined in the products measured since 2001.
The other findings of interest were that off mode power consumption continues to persist at a
significant level (of the order of 1 Watt) and that that fewer and fewer products appear to have an off
mode present over time.
Results indicate that there is an opportunity for many products to improve energy consumption in low
power modes within product groups. For the products measured, there was generally a wide variance
in power consumption in off mode and passive standby mode without any difference in performance
or functionality between these products. This tends to suggest that there are substantial opportunities
for manufacturers to reduce low power mode consumption, probably at low marginal cost. Details on
the findings of the surveys can be found in EnergyConsult 2005 [7].

990
 50%

45%
2001 n= 405
Percentage of appliances measured at load

2002 n= 397
40% 2003Jan n= 325
2003-04 n= 686
35%
2004-05 n= 737
2005-06 n= 295

30%

25%

20%

15%

10%

5%

0%
<0.05 <1 <2 <3 <4 <5 <6 <7 <8 <9 <10 <20 <50 >50
Power of Appliance in the specified mode - Watts

Figure 1: Distribution of “passive standby mode” power – all products

Source: [7]

Trends in Standby for Key Products

The following section analyses the data trends on standby for selected products and modes
combining data from the 2000 [2] and 2005 [6] intrusive surveys and the store surveys conducted
from 2001 to 2005 [7]. This draws on a database of more than 16,000 products. It is important to note
that the attributes of the older stock (say pre 1995) as recorded in 2000 may be different to the
attributes recorded in 2005. For some product types, the sample is quite small so these need to be
seen as indicative. Generally, data from the store survey is the most reliable trend line as each point
consists of a large number of readings taken in store for that year.

Desktop Computers – Off Mode

Personal computers primarily had what appeared to be hard off switches installed in most products up
until the mid 1990’s so they had effectively zero energy consumption when off. By 2001 virtually no
computers had zero power consumption in off mode. Data suggests that power consumption in off
mode peaked at 4 Watts and is now declining to around 2 Watts in 2005. The sample of houses in the
intrusive survey found 1.25 desktop computers per house. It is important to note that this trend is
more a function of the design of the power supplies now commonly used in desktop computers that
than the computers themselves. The other issue is that in Australia, there are a large number of small
companies that assemble computers using parts from a wide range of OEM suppliers. This makes
dealing with this particular product difficult in terms of any voluntary or mandatory industry based
commitments.

991
 10.0

9.0

8.0
Off Mode Power Consumption (Watts)

7.0

6.0

5.0

4.0

3.0

2.0

1.0

0.0
1991 1993 1995 1997 1999 2001 2003 2005
Year Purchased

2000 Survey Off N=77 2005 Survey: Off N=150 2000 Survey Average 2005 Survey Average Store Survey Averages

Figure 2: Trends in Computer Off Mode

VCRs – passive standby mode

While VCRs are perhaps an obsolete technology, they are ubiquitous and ownership levels in
Australia are around 1.11 per household. Greatly reduced prices mean that the product is likely to
linger for many years.

14.0
Passive Standby Power Consumption (Watts)

12.0

10.0

8.0

6.0

4.0

2.0

0.0
1993 1995 1997 1999 2001 2003 2005
Year Purchased
2000 Survey Passive N=91 2000 Survey Average 2005 Survey: Passive N=133 2005 Survey Average Store Survey Averages

Figure 3: Trends in VCR Passive Standby Mode

Passive standby mode is the most important for a VCR – this allows activation by remote control and
clock circuits are active. Virtually no models have an option for off mode. In the mid 1990’s passive
standby mode was around 8W and this has declined to about 3 Watts in 2005 (based on the store
survey data).

992
Integrated Stereos – Passive Standby Mode
Integrated stereos are common devices with an ownership of around 0.97 per household in Australia.
These are products with a single power cord that provide functions such as radio, tape player and CD
player and are not designed to be portable. They usually have separate speakers. In the mid to late
1990’s passive standby for many of these products averaged around 10 Watts. This has declined to
around 4 Watts in 2005.
Despite improvements in the average passive standby mode power, there is still a very wide
distribution on the power consumption in this mode – the distribution is almost bi-modal and split into
high and low power products – the data below is from store surveys 2001 to 2005 for integrated
stereos. The improvement in average power has been a result of an increase in products < 1 Watt
and a decline in products over 10 Watts. So while the improvement over time is to be applauded,
there are still many products on that have very poor passive standby attributes.

20.0

18.0
Passive Standby Power Consumption (Watts)

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0

0.0
1993 1995 1997 1999 2001 2003 2005
Year Purchased
2000 Survey Passive N=50 2000 Survey Average 2005 Survey: Passive N=103 2005 Survey Average Store Survey Averages

Figure 4: Trends in Integrated Stereo Passive Standby Mode

993
 50%

45%
2001 n= 66
Percentage of appliances measured at load

2002 n= 51
40% 2003Jan n= 32
2003-04 n= 51
35%
2004-05 n= 65
2005-06 n= 25

30%

25%

20%

15%

10%

5%

0%
<0.05 <1 <2 <3 <4 <5 <6 <7 <8 <9 <10 <20 <50 >50
Power of Appliance in the specified mode - Watts

Figure 5: Distribution of Integrated Stereo Passive Standby Mode Power

Televisions – Passive Standby Mode

Televisions are common devices with an ownership of around 2.0 per household in Australia. In the
mid to late 1990’s passive standby for many products averaged around 14 Watts. This has declined to
around 4 Watts in 2005. Average new products in Europe are somewhat better than this (less than 2
Watts). Many new products still consume more than 6 Watts in passive standby mode. Most
televisions in Australia have an off switch and most use 0.0 Watts when off.
20.0

18.0
Passive Standby Power Consumption (Watts)

16.0

14.0

12.0

10.0

8.0

6.0

4.0

2.0

0.0
1989 1991 1993 1995 1997 1999 2001 2003 2005
Year Purchased
2000 Survey Passive N=86 2000 Survey Average 2005 Survey: Passive N=183 2005 Survey Average Store Survey Averages

Figure 6: Trends in Television Passive Standby Mode

Clothes Washers – Off Mode

Prior to the mid 1990, few if any clothes washers used any power when off as most had electro-
mechanical controls and switches. From the mid 1990’s machines with electronic controls appeared in
the stock. The average power consumed up to around 2001 was about 2.5 Watts in off mode. Since

994
2003 there has been a marked decline in average power to below 1 Watt in off mode. It is interesting
to note that while the average off mode power from 1990 to 2002 was probably 2 to 3 Watts, there
were a lot of machines that used 0.0 Watts and a lot of machines that used 5 to 10 Watts. The worst
machines were removed in about 2002, possibly in response to the Australian government’s targets
for clothes washers that were announced in 2003.
There is an important issue regarding clothes washers and dishwashers. Most European products
have an off switch, which usually disconnects power to most parts of the machines and drops the
power consumption close to zero Watts. However to achieve this, the consumer must manually turn
the machine off when the cycle has been completed and the load removed. Disturbingly, during the
intrusive survey in 2005 around 40% of the European front loading machines were found left in active
standby mode when not in use (ie the users did not manually turn the machines off when the washing
cycle was completed). These machines persist in active standby mode indefinitely with the current
machine designs. This means that the off mode power consumption for many European machines is
probably not all that relevant. A more relevant figure is the active standby power consumption (which
in many cases is likely to be similar to end of cycle mode). In 2005 this averaged about 4 Watts,
although individual values varied from 1 Watt to 7 Watts in this mode. This is an example of a mode
which is impossible to accurately determine from store or house measurements.
All non European machines automatically revert to off mode automatically at the end of the cycle, so
only off mode measurements for these machines are relevant.
10.0

9.0

8.0
Off Mode Power Consumption (Watts)

7.0

6.0

5.0

4.0

3.0

2.0

1.0

0.0
1991 1993 1995 1997 1999 2001 2003 2005
Year Purchased

2000 Survey Off N=61 2005 Survey: Off N=112 2000 Survey Average 2005 Survey Average Store Survey Averages

Figure 7: Trends in Clothes Washer Off Mode

Other Products of Particular Concern

The 2005 intrusive survey identified a number of other products that are of concern with respect to
standby. The are summarised below:
• Clock radios: 2.1 Watts, ownership 1.33, trend: no changes expected.
• Computer peripherals: range from 2 Watts to 6 Watts, multiple items per house (printers
modems, speakers, switches/hubs, scanners etc.).
• Digital set-top boxes: 13.3 Watts, ownership 0.32, trend: ownership likely to soar.
• DVD recorders: 4.9 Watts, ownership 0.08, trend: ownership increasing as they replace
VCRs, much higher power with hard drives (more common) – blurred divide with set-top
boxes and DVD burners as digital tuners become available.
• External power supplies: 1.2 Watts (no load), ownership around 7 per house (3 as separate
devices not always connected to a product – usually battery charging – e.g. mobile phone).
• Laptop computers: unclear standby measurements (measurements confounded by battery
charge state, probably in the range 2 Watts to 6 Watts when fully charged), ownership 0.54
and set to increase.

995
 • Multi-function devices: 11.2 Watts (active mode), ownership 0.23 and increasing.
• Remote control garage door openers: 4.1 Watts, ownership 0.30 and increasing.
• Smoke alarms: 0.4 Watts, ownership 1.23 and increasing rapidly.

Conclusions and Lessons Learned

Data from a range of sources suggests that the trend in standby power consumption for some
products is improving over time. This partly due to natural improvements in technology (availability of
low power components and modules) but also in part due to increasing pressure from a range of
government programs around the world. Many of the products with standby attributes can be
considered global commodities and as such, pressure in one or markets will have impacts elsewhere.
So there is some good news.
Australia has initiated a multi-pronged strategy for data collection. This is necessary given the diverse
range of products found in homes. Store surveys provide a record of the types of new products
offered for sale over time and the distribution of standby attributes for these products. It also allows
accurate estimates of year to year trends in standby attributes which are important for modeling and
impact evaluation. However, store surveys have some limitations – there are some products which
cannot be measured in store (most notably devices which are hard wired) and some modes which
cannot be reached in a store environment (eg end of cycle mode for clothes washers or dishwashers
or modes that take some time to reach – eg auto power-down modes). Consequently, an alternative
data collection approach needs to be adopted for such products and modes where these are
important. Typically laboratory measurements are necessary, which can be slow and expensive. For
ubiquitous products, an alternative strategy is to mandate declaration of the relevant standby values
as part of the regulation of products for energy labelling and/or MEPS.
The inclusion of surveys of real houses from time to time is seen as a critical part of the data
collection process as it provides information on the range and type of equipment actually installed and
used in typical houses. Importantly, it can provide warning of products with high standby power that
have not identified in other processes. Household surveys also provide the opportunity to record the
age distribution of products (this is usually only indicative) and the modes in which products are left
when not in use. However, household surveys also have limitations on the measurement of hard
wired devices. Also, there are always some devices that are hidden or inaccessible, so while the data
collection for household surveys can be comprehensive, it is rarely complete. Such surveys are also
quite an imposition on the household owners (it can take more than 3 hours per house), so good
communication with potential survey households with regard to expectations is important.
The data from Australia data suggests that there are a number of products that have very poor
standby attributes (particularly in the home entertainment area). Some devices such as set-top boxes
and hard drive DVD recorders with a digital tuner have quite high power consumption in most modes
and ownership for these products are set to grow rapidly (particularly as analogue television
broadcasts are phased out and VCRs disappear). Some integrated stereos have also been very poor,
but the worst products are being removed from the market. There also a multiplicity of other devices in
the home entertainment area (around 5 per house) which contribute on average 4 Watts each.
The other product of some concern are computers. The increase in ownership is extraordinary (with
about than 1.8 per house in the 2005 survey sample, although recent national surveys suggest a
lower figure) and this is set to increase. Presumably similar trends are being experienced in other
countries. There are several aspects of concern. Firstly, the survey participants indicated that around
20% of the desktop computers included in the survey were never switched off (they were operating as
home servers or network points). This has a substantial energy impact as the on mode energy
consumption for an average desktop computer is more than 80 Watts (compared to an average of 3.5
Watts in off mode), assuming the screen goes into passive standby mode (2.6 Watts for LCD screen,
8.2 Watts for CRT monitors). The second issue of concern is the number of associated computer
peripherals. Items such as modems, routers, switches/hubs, inkjet and laser printers, scanners and
speakers are all common (average of 5.4 devices per house, at an average of 5.2 Watts per device).
There is also an increasing blur between computers and entertainment devices.
Telephony equipment is also of some concern – cordless phones are now found in nearly every home
and the power consumption of this equipment is typically 3 Watts per device (average of 2 devices per
house).
Major appliances of course have some standby attributes of concern but their total contribution to
household standby is generally only very modest. The user behaviour with respect to European
dishwashers and clothes washers will have an impact on the household standby (whether or not the
user manually switches the products off when not in use). Also, there is very little data on air

996
conditioners available from the surveys in Australia – the most common products are split systems
which are hard wired in the household supply so standby measurements are not possible in most
situations. Australia has recently regulated to mandate the supply of data on power consumption in
passive standby and off modes and crankcase heaters for air conditioners as part of the energy
labelling and MEPS program, so more data should be available in the coming years.
One of the most alarming findings was that the number of devices connected to the mains and which
used some power when not performing its main function climbed from 21 in 2000 to 27 in 2005 – this
is an increase of 35%.
Some countries have undertaken limited surveys of households to determine standby power
consumption. Many of these have concluded that standby is lower than the values determined for
Australia (around 11% of residential electricity consumption). Perhaps this is because there are fewer
device installed in such countries or their standby attributes are better. However a more likely
explanation is that such surveys have concentrated on major appliances which are the obvious initial
focus for standby power while ignoring the rapid increase in products that are associated with digital
age. Appliances and home entertainment are only part of the total standby story, so it is important to
recognize contributions from a wide range of products.

References
[1] Things That Go Blip In The Night, International Energy Agency, Paris, 2001.
[2] Quantification of residential standby power consumption in Australia: Results of recent survey
work, Lloyd Harrington of Energy Efficient Strategies and Paula Kleverlaan of Energy Consult,
for NAEEEC, April 2001. Available from www.energyrating.gov.au in the electronic library.
[3] Standby Korea 2010: Korea’s 1-Watt Plan, Korea Energy Management Corporation and
Ministry of Commerce, Industry and Energy, Korea. See www.mocie.go.kr and
www.kemco.or.kr
[4] IEC62301: Household electrical appliances - Measurement of standby power, published by the
International Electrotechnical Commission, Geneva, Edition 1, June 2005. See www.iec.ch
[5] Money isn’t all you’re saving: Australia’s standby power strategy 2002 – 2012, by the Ministerial
Council on Energy forming part of the National Appliance and Equipment Energy Efficiency
Program, November 2002, NAEEEC Report 2002/12. Available from www.energyrating.gov.au
in the electronic library.
[6] 2005 Intrusive Residential Standby Survey Report, prepared by Energy Efficient Strategies for
Equipment Energy Efficiency Committee, Australia. Report 2006/02. See
www.energyrating.gov.au
[7] Appliance Standby Power Consumption - Store Survey 2005/2006, prepared by EnergyConsult
for Equipment Energy Efficiency Committee, Australia, Report 2005/23. See
www.energyrating.gov.au
[8] Developing and Testing Low Power Mode Measurement Methods, Lawrence Berkeley National
Laboratory, prepared for California Energy Commission, Public Interest Energy Research
Program, September 2004, Report P500-04-057. Available from http://energy.ca.gov
[9] Electronics Come of Age: A Taxonomy for Miscellaneous and Low Power Products, paper by
Bruce Nordman, Marla McWhinney of Lawrence Berkeley National Laboratory, presented to
EEDAL, London , June 2006.

997
998
Electronics Come of Age: A Taxonomy for Miscellaneous and Low
Power Products
Bruce Nordman, Marla McWhinney

Lawrence Berkeley National Laboratory

Abstract
Most energy end uses such as space conditioning or water heating are apparently well-defined in
what is included, and have terminology that derives from the professionals who work in the relevant
field. The topic of “miscellaneous” consumption lacks such clarity for historical and practical reasons.
As this end use grows in size and interest for the energy community, the confusion and ambiguity
around the topic is an increasing barrier to progress. This paper provides definitions for key terms
and concepts with the intent that that future work can be more correctly and consistently reported and
interpreted. In addition, it provides a taxonomy of product types and categories, which covers both
residential and commercial miscellaneous consumption. A key element is identification of
“electronics” as a distinct energy end use. Finally, products are identified as to whether they
commonly have a low-power mode, and product types that have such modes within the traditional end
uses are also listed.

Introduction
Background
When building energy efficiency science emerged in the 1970s, attention was naturally drawn to the
end uses that were the largest, most closely related to building services, potentially interactive with
the building shell, and most easily characterized in traditional physics efficiency terms. These were
climate control (HVAC), lighting, water heating, and major appliances (esp. refrigerators); we call
these the “traditional” end uses. Other consumption was typically relegated to “Miscellaneous” or
“Other”. In monitoring projects, it was common to meter the whole building and the traditional end
uses, then calculate the residual for the miscellaneous loads.
Over the last 30 years, the traditional end uses have become more efficient (at least per unit floor
area and service delivered) and the absolute amount of miscellaneous electricity consumption has
risen. The number and types of miscellaneous products found in buildings has also increased. The
result is a significant increase in the percent of electricity use in the “other” category. For highly
efficient new houses, it can exceed 50%. There is increasing interest in this topic area, but no
consistent set of terminology on which to base policy and analysis. Since so many product types are
covered, definitions are sorely needed to avoid confusion.

Scope
This taxonomy covers only electricity used by mains-voltage (115 and 230 V for the U.S.) AC
products. It does not include any gas consumption (standby or otherwise), but does include the
electricity consumption of primarily gas products. It addresses primarily residential and commercial
1
buildings , but not exclusively. It does not cover low-voltage DC-powered products (e.g. USB or
Power Over Ethernet), though the consumption of the latter are included in the products that provide
the low-voltage DC such as computers and powered USB hubs.

Approach
In this project we surveyed current taxonomies of miscellaneous or low power products to draw
inspiration for the approach and specific naming and categorization. We had previously reviewed
issues related to power modes [1] and brought the results of that analysis to the taxonomy. We

1
One of the principal sources for miscellaneous products, [4], included educational buildings in addition to strictly commercial
ones. Most of the sources used are residential, so that our coverage of commercial miscellaneous products is considerably
less comprehensive.

999
combined the past approaches with knowledge of how this taxonomy could be used in the future to
2
craft a system with the primary goals of consistency and clarity .
Sources for the taxonomy included studies that focused on measurement methods ([1] and [2]),
particular types of electronic devices ([3], [4], [5], [6], [7], [8], [9], [10], [11], [12], [13], and [14]),
standby power ([5], [15], [16], [17], [18], [19], and [20]), appliances ([15], [21], [22], and [23]),
miscellaneous products ([1], [4], [24], [25], and [26]), and energy consumption in buildings generally
([27], [28], [29], and [30]).

Key Terms
Definitions
Following are proposed definitions (“low power”, “product”, and “product type” taken or adapted from
that specified in [1]).
3
Low Power Mode . Any mode in which a product is not performing any of its principal functions.
Some products have more than one principal function. When feasible, low-power modes shall be
categorized into on, sleep, and off modes. Disconnected is not a low power mode.

Standby. The minimum power mode of a product, or more formally, “the lowest power consumption
mode which cannot be switched off (influenced) by the user and that may persist for an indefinite time
when an appliance is connected to the main electricity supply and used in accordance with the
manufacturer’s instructions.” [2].

Product. A piece of equipment that can be powered directly from mains power. This covers a
specific instance or model number.
4
Product Type . A general category of product within which there is a sufficient amount of common
functionality, modes, and behavior.

Electronics. Devices whose primary function is Information (obtaining, storing, managing, or

presenting).

Traditional End Uses. HVAC, Lighting, Water Heating, and Major Appliances.

Major End Use. The Traditional end uses plus Electronics.

Miscellaneous. Any product type not included in any of the major end uses. “Other” should be taken
as a synonym for Miscellaneous.

Overlooked Products. Devices that perform the function of a traditional end use but are not usually
included in that end use’s total. We have these included in the Miscellaneous end use.
5
Plug Loads . A product powered by means of an ordinary AC plug (e.g. 100 V, 115 V or 230 V).

Hard-Wired Loads. A product with a direct-wired connection to an AC source. These can have
switches or timers between the product and the AC source.

Category. A group of product types within a single end use that share common functionality or are
otherwise logically related.

2
The full version of the paper includes an appendix with additional information including a description of each of the source
taxonomies. A spreadsheet of all of the raw data is also available.
3
This definition is expansive as it includes “ready” modes that are forms of “on”. It also is intended to include the powered
mode of single-mode products.
4
This is the core term of interest to energy professionals, e.g. “dishwasher” or “TV, CRT”. Product types are differentiated “by
capacity or other performance-related features that provide utility to the consumer and affect efficiency.” (U.S. Department of
Energy (DOE). 1991. Energy Conservation Program for Consumer Products: Final Rule Regarding Energy Conservation
Standards for Three Types of Consumer Products. Code of Federal Regulations, Title 10 Part 430, Vol. 56, No. 93, May 14,
1991. Docket No. CE-RM-88-101).
5
This is often understood to exclude product types included in major end uses.

1000
Key Topics

Electronics. In future building science policy and analysis, there should always be a separate end
use of “Electronics” [3] — covering products “… whose primary function is Information”. Most
products in this end use were previously categorized as consumer electronics or office equipment.
The electronics end use does not include electronic components of principally non-electronic devices
(e.g. dishwasher controls) unless it is a separately powered device (powered from mains AC or from
6
DC other than from the non-electronic device ). Electronic products can have significant non-
electronic components. For example, a laser printer contains large heating loads but retains
information presentation as the principal function. Electronics were formally part of miscellaneous but
the magnitude of electronics consumption in both the residential and commercial sectors and
consistency of function (information) across electronics makes it logical to recognize it as a distinct
7
major end use .
There are some products that can be reasonably argued to be within or outside of the electronics
category, e.g. smoke and CO detectors (which provide information about potential fire status), and
exit signs (which provide directional information rather than illumination in the usual sense). In both of
these cases we did not include the product type in the electronics end use as their status in the
infrastructure and lighting categories (respectively) seemed clearly established. In addition, at least at
present, they do not usually connect to other electronic devices. By contrast, we have included the
entire security category in electronics as the amount of information is larger and increasing and the
security devices are also increasingly linked to clearly electronic devices.

Miscellaneous. Miscellaneous is taken to be all building energy consumption (on the load side of the
utility meter) that is not covered by any of the major end uses. This includes “overlooked” products —
those that serve the broad functions of the major end uses but are not usually included in estimates
(e.g. ceiling fans, humidifiers, space heaters). Per above, electronic products as a whole comprise a
distinct end use and can no longer be mixed in with miscellaneous products.
Miscellaneous products span the range from the very small to the very large, both in electricity
consumption and physical size. Examples range from staplers to pool pumps.

Low Power Modes. Low power mode consumption was reviewed in great detail in [1]. That paper
took an expansive view of the concept, including “ready” modes and consumption of single-mode
products. It is useful to look at these modes collectively across many product types. The purpose
and potential efficiency, of low power mode consumption is often similar across products whose
active functions are quite different. Note that per IEC 62301, “standby” is defined with respect to is
power level and in practice can occur in any operating mode (On, Sleep, or Off).

The Taxonomy
Naming and Categorization
Traditional energy analysis deals with a small number of end uses with a modest number of product
types within each end use. Allocation of product types to end uses is almost always obvious. Naming
and categories are by no means trivial but are clearly driven by functional and capacity factors. By
contrast, miscellaneous products lack clarity in categories, product types, and naming.
The number of different miscellaneous product types is large. For example, an assessment of just
eight California houses, [1] found 108 different product types among only products with low power
modes (that is, not including miscellaneous products without low power modes). An audit of 16
commercial buildings [4] found 321 different product types without even assessing closets, plenums,
basements, or attics.
Many past studies have listed and categorized product types, but the naming and grouping has
generally been ad hoc. We reviewed the listing and naming of products in 28 different studies (listed
in the References), each of which offered a system of product grouping that was typically secondary
to the study scope.

6
For example, a 24 VAC thermostat powered from a furnace is not considered electronics, but an electronic HVAC control
powered by mains AC or by USB would be considered to be electronics.
7
“Office equipment” is now better characterized as IT as those products can and often are used for non-office purposes,
particularly as used in the home.

1001
Structure
This taxonomy covers two distinct but significantly overlapping topics: miscellaneous and low-
power mode product types. Miscellaneous is the electricity end use which includes all products not
8
included in other end uses (including arcane uses such as utility meter and wire resistance losses ).
The other end uses are the traditional ones — HVAC, lighting, water heating, major appliances —
plus the new end use of electronics.
For each of the traditional end uses, there are products that are usually not counted as part of them
but are performing the end use’s function. Examples are portable fans, task lights, point of use water
heaters, and wine refrigerators (for HVAC, lighting, water heating, and major appliances respectively).
The overall structure is shown in Table 1. The taxonomy is divided into two major end uses plus the
traditional end uses. Each end use is divided into a number of categories. There is one category for
each traditional end use to accommodate low power mode consumption of these products —
products without low power mode consumption in the traditional end uses are not included. There is
also a category within miscellaneous corresponding to each traditional end use for the overlooked
products. Most electronics are in categories that were formerly consumer electronics or information
technology. Everything else is truly miscellaneous.

Table 1: Overall structure of the taxonomy

End Use Categories
Electronics Audio, Cash exchange, Computer, Display, Imaging, Networking, Peripherals,
Security, Set-top, Telephony, Video
Miscellaneous Business equipment, HVAC, Commercial kitchen equipment, Electric housewares,
Hobby/leisure, Infrastructure, Lighting, Major Appliances, Medical (lab, exam, and
specialty), Other, Outdoor Appliances, Personal Care, Power, Transportation, Utility,
Water heating
Traditional HVAC, Lighting, Major Appliance, Water heating
Note. Italicized categories occur rarely if ever in residential buildings. Overlooked categories are underlined.

The degree to which product types are comprehensive or split into many pieces is a combination of a
number of factors, including the similarity of the ultimate function provided, the key technology
employed (e.g. inkjet vs. laser printers), capacity/size, power levels, usage patterns, stock-wide
consumption, and ease of disaggregation. Some studies will find it useful or necessary to split some
product types or to combine several. Over time, this taxonomy will change, as technologies and the
stock of products evolve. The taxonomy reflects products in use rather than for sale.
In our review of relevant literature, we encountered numerous taxonomies of product categories.
Table 2 provides an overview of some categorizations we found during our review. The order of the
categories has been changed to provide some correspondence across these examples.
The variety and scope of the categories included varied widely. For example, office and consumer
electronics were treated in many different ways, with office equipment variously characterized as IT,
treated as a separate equipment category, or wrapped into the consumer electronics category as a
whole. Different studies included or dropped entire categories, and the naming showed wide variation.
The current version of the taxonomy is presented in Table 3.

8
The loss from electrical resistance of wiring within buildings is not due to the use of a “product” per se, but accounting for it as
such is the best way to include it in total building consumption.

1002
Table 2: Product categories used in five sample studies
Australia Lebot Sanchez Rosen Ross
IT IT Electronics Office Computer
Entertainment Video Video Entertainment
Hi-Fi Set-top
Audio
Telephone Systems Telephony Communication
Major Appliances Cooking Motor Miscellaneous
Small Appliances Miscellaneous Heating
Lighting
Australia: Ministerial Council on Energy Forming Part of the National Greenhouse Strategy. Money Isn't All
You're Saving. Australia's Standby Power Strategy 2002-2012. Lebot: Lebot, Benoit et al. Global Implications
of Standby Energy Use. Sanchez: Sanchez, M. et al. Miscellaneous Electricity Use in the U.S. Residential
Sector. Rosen Rosen, K. et al. National Energy Use of Consumer Electronics in at the end of the Twentieth
Century. Ross: Ross, JP and A. Meier. Whole House Measurements of Standby Consumption.

Table 3: The Miscellaneous and Low Power Taxonomy: May 10, 2006 version (Residential)
Electronics

Audio Computer display, plasma External drive

screen Speakers, computer
Amplifier Game console, portable
Audio minisystem Projector, slide Security
Cassette deck Projector, video
CD player Television, large CRT Security system
CD player, portable Television, LCD
Charger, digital music player Television, plasma Set-top
Equalizer (audio) Television, rear projection
Home theatre system Television, standard CRT Set-top box, analog cable
Karaoke machine Television/VCR Set-top box, digital cable
Musical keyboard Set-top box, digital cable with
Radio, table Imaging PVR
Receiver (audio) Set-top box, internet
Speakers, powered Copier Set-top box, satellite
Speakers, wireless (base Fax, inkjet Set-top box, satellite with PVR
station) Fax, laser
Speakers, wireless (speakers) Fax, thermal Telephony
Stereo, portable Multi-function device, inkjet
Subwoofer Multi-function device, laser Answering machine
Tuner Printer, inkjet Caller ID unit
Turntable (audio) Printer, laser Charger, mobile phone
Printer, photo Phone
Computer Scanner, flatbed Phone, conference
Phone, corded
Computer, desktop Networking Phone, cordless
Computer, integrated-CRT Phone, cordless with
Computer, integrated-LCD Hub, ethernet answering machine
Computer, notebook Hub, USB
Dock, notebook Modem, cable Video
Game console Modem, DSL
Game console with internet Modem, POTS Charger, still camera
connectivity PVR Charger, video camera
Router, ethernet DVD player
Wireless access point DVD recorder
Display VCR
Peripherals VCR/DVD
Computer display, CRT Videocassette rewinder
Computer display, LCD CD recorder
Dock, PDA

1003
Miscellaneous

Business equipment Air cleaner, mounted Waterbed

Air cleaner, portable
Adding machine Air conditioning, evaporative cooler Outdoor Appliances
Pencil sharpener Ceiling fan
Shredder Dehumidifier Charger, hedge trimmer
Stapler Exhaust fan Charger, weed trimmer
Typewriter Fan, portable Coil, snow melting
Fan, rangehood Grill, outdoor
Electric housewares Fan, whole house Lawn mower
Fan, window Timer, irrigation
Automatic griddles Furnace fans
Blanket Heating, fireplace electric Personal Care
Blender Humidifier
Bread maker Space heater, portable (electric) Air freshener
Broiler Space heater, portable (non- Curling iron
Clock electric) Hair dryer
Clock, radio Heat lamp
Coffee grinder Infrastructure Home medical equipment
Coffee maker, residential Massager
Corn popper, air Breaker, AFI Shaver
Corn popper, hot oil Breaker, GFCI Toothbrush
Deep fryer, residential Detector, carbon monoxide Water softener
Espresso maker, residential Detector, smoke
Food processor Doorbell Power
Food slicer Garage door opener
Frying pan GFCI outlet External power supply
Hand mixer Infant monitor, receiver Power strip
Heating pad Infant monitor, transmitter Power supply
Hot plate (kitchen) Utility meter Surge protector
Iron Wire losses Timer
Juicer Uninterruptible power supply
Kettle Lighting
Knife Transportation
Mug warmer Dimming switch
Oven, microwave Emergency light, interior Auto engine heater
Pasta maker (commercial) Car, wheelchair or golf cart
Rice maker Grow lamps
Sewing machine Lamp, decorative Utility
Slow cooker Lights, holiday
Stand mixer Low voltage landscape Bicycle light
Toaster Motion sensor, exterior Charger, battery
Toaster oven Motion sensor, interior Floor polisher
Vacuum, central Night light, interior Pet fence
Vacuum, rechargeable Photosensors, exterior Power tool
Vacuum, standard Timer, exterior Power tool, cordless
Waffle iron Timer, interior Pump, sump
Pump, well
Hobby/leisure Major Appliance
Water heating
Aquarium Garbage disposal
Kiln Refrigerator, wine cooler Water heating, instantaneous
Pool Trash compactor single point of use
Sauna, electric Water dispenser, bottled Water heating, point of use
Spa/hot tub tank
Other
HVAC
Fountain, indoor

1004
Traditional End Uses

HVAC Dishwasher
Lighting, residential Freezer
Air conditioning, central Oven, electric
Air conditioning, heat pump Major Appliance Oven, gas
Air conditioning, room/wall Refrigerator
Heating, boiler Clothes dryer, electric
Heating, furnace baseboard, Clothes dryer, gas Water heating
floor or wall unit Clothes washer, horizontal
Heating, furnace central axis Water heating, electric
Heating, heat pump Clothes washer, standard Water heating, gas
Cooktop, electric Water heating, heat pump
Lighting Cooktop, gas Water heating, other

Product Type Naming

We brought a variety of principles to the process of selecting names for product types.
• Strive for brevity, e.g. dropping “electric” from “electric knife” since non-powered knives would
obviously not have energy consumption. Also use common acronyms like “TV”, “CD”.
• Give preference to names used in ordinary language. Avoid brand names.
• Use commas to distinguish related types of products, e.g. “TV, CRT” and “TV, LCD” or
“Clothes dryer, electric” and “Clothes dryer, gas”. Related types need not be in the same
category.
• Use “/” within lists rather than commas. Use only one comma; if more needed use
parentheses and “/”s).
• Use parentheses to denote product types that share a common name, e.g. “Amplifier
(network)” and “Amplifier (audio)”.
• Distinguish between products that can be run off of integral rechargeable batteries
(“rechargeable products”) from those that cannot (“non-rechargeable products”). Non-
rechargeable products includes both those that can be run from AC or generic batteries and
those that can only be powered by AC (e.g. some audio minisystems); this distinction is not
considered significant for the purposes of the taxonomy. It is not considered significant
whether a battery powerable product can be used while being AC powered (e.g. many
shavers, mobile phones) from those that can’t (e.g. power tools, some vacuums).
• “Portable” as a suffix is to indicate if something can be easily moved, not whether it can be
powered by batteries.

Product Types in the Traditional End Uses

We relied primarily on RECS category definitions to determine which products were grouped into the
traditional end uses, though we have modified these, and will be seeking clarification on how the
RECS definitions are applied. Our lists are shown in Table 4.
For HVAC and Water heating, we used the RECS definitions directly. RECS includes many products
in “appliances” that seem to not merit the “major” name and are not likely included in most energy
analysis estimates of appliances. Examples include: evaporative coolers, spas, and waterbed
heaters. For lighting, RECS includes “Energy used to supply electricity to light bulbs inside and
outside of the housing unit. All types of light bulbs are included: incandescent, fluorescent, compact
fluorescent, halogen, and high-intensity-discharge (HID),” and includes lighting in the general
“appliance” category. We use a narrower definition. “permanently installed” includes products that
are hardwired as well as those that have plugs but are screwed down or otherwise firmly attached.

1005
Table 4: Products Types included in the Traditional End Uses
End Use Included
HVAC Boiler, heat pump, central furnace, heater (baseboard, floor, or wall unit —any
fuel type), central air conditioning, heat pump and air conditioner (room or wall
unit)
Water heating Tank units (electric, gas, heat pump, and other fuel types, such as LPG)
Major appliances Clothes dryer, clothes washer, dishwasher, refrigerator, freezer, oven, and
cooktops
Lighting Permanently installed fixtures (interior), permanently installed fixtures (exterior),
floor lamps, table lamps

Product Type Attributes

In [1], products types were categorized as one of four types for the likelihood of having low-power
modes as defined in Table 5 (from [1]). Estimates of energy consumption need to take into account
the fraction of the time that examples of each product type actually have a low power mode at all.
This affects understandings of power levels and usage patterns. The taxonomy spreadsheet
indicates which product types sometimes or always have low power modes.

Table 5: Attributes of Product Types

Product Type Characteristic Examples
Classification
Always Always (or nearly always) have low-power modes. Cordless telephone
Sometimes Some examples have low-power modes; others do Ceiling fan, radio, toaster
not.
Never Never [or rarely] have low-power modes Corded power tools
Excluded Not included in this measurement procedure, Refrigerator
regardless of whether or not they have low-power
mode consumption.
Note: The concept of an excluded product type does not apply to the taxonomy.

Details
In some cases the principal function is not absolutely clear. For example, is a clock radio mainly a
clock or mainly a radio? While the energy use, cost, and physical size may be driven more by the
radio feature, we judged that these are foremost used as clocks and only secondarily as radios. This
does not preclude an audio system from having a clock on its display; these will be judged to be
principally audio devices with the clock only incidental.
For audio equipment, an “Audio minisystem” (or compact audio system) is one in which the speakers
sold with the product can be physically detached from the main electronics. This is in contrast to a
“Stereo, portable” which has integral non-detachable speakers. In the comments, a “Rack” system is
a set of separately-powered components that are often sold as a unit that match and readily stack.
This provides the equivalent of separately purchased independent components.

Future Development
We have included commercial products found in studies we reviewed, though few of the studies had
the commercial sector as a focus. At present, only the Roberson report and Energy Star commercial
kitchen products are included. A key missing area is much of commercial building infrastructure.
As the taxonomy evolves over time, it should be republished with dated versions and clear differences
identified from previous ones (this version is dated May 10, 2006).

Conclusions
A major result of this taxonomy is that electronics should be considered a major end use on its own,
distinct from miscellaneous products. While products in the traditional end uses are not a particular
focus of this project, it is necessary to know what precisely is in them to know what is or is not in the
miscellaneous category. Products can be named in a reasonably consistent manner and put into
categories that well serve many purposes — no taxonomy is ideal for all purposes (for example,
hardwired products are not a distinct category as the fact of being hardwired is not related to the
function which is our criteria for assigning to categories). This taxonomy is not the ultimate — and it

1006
will always evolve over time — but this should get the energy efficiency community most of the way
towards a consistent framework for naming and categorizing these products.

References
[1] Nordman B. Developing and testing low power mode measurement methods. Prepared for
the California Energy Commission Public Interest Research Program. 2004.
[2] International Electrotechnical Commission. IEC 62301 Ed 1: Measurement of Standby Power.
2005.
[3] Nordman B. Electronics: The New Major Electricity End Use. ACEEE Emerging Technologies
Summit (San Francisco, USA, 14-15 October 2004).
[4] Roberson J. et al. After hours status of office equipment and inventory of miscellaneous plug
load equipment. Lawrence Berkeley National Laboratory. 2004. LBNL-43729-Revised.
[5] Floyd D. and Webber C.A. Leaking Electricity: Individual Field Measurement of Consumer
Electronics. August. Proc. of the 1998 ACEEE Summer Study on Energy Efficiency in
Buildings. Ed.: American Council for an Energy-Efficient Economy, Washington:1998.
[6] Kawamoto K., Koomey J G., Nordman B., Brown R.E., Piette, M.A., and Meier A.K. Electricity
Used by Office Equipment and Network Equipment in the U.S.: Detailed Report and
Appendices. Lawrence Berkeley National Laboratory. 2001. LBNL-45917.
[7] Nordman B. and Meier A. Energy Consumption of Home Information Technology. Lawrence
Berkeley National Laboratory. 2004. LBNL-53500.
[8] Rosen K. and Meier A. Energy Use of Televisions and Video Cassette Recorders in the U.S.
Lawrence Berkeley National Laboratory. 1999. LBNL-42393.
[9] Rosen K. and Meier A. Energy Use of Home Audio Products in the U.S. Lawrence Berkeley
National Laboratory. 1999. LBNL-43468.
[10] Rosen K., Meier A. and Zandelin S. National Energy Use of Consumer Electronics in at the
end of the Twentieth Century. Lawrence Berkeley National Laboratory. 2000. LBNL-46212.
[11] Rosen, K., Meier A. and Zandelin S. National Energy Use of Consumer Electronics in 1999.
Lawrence Berkeley National Laboratory. 2000. LBNL-45988.
[12] Rosen K., Meier A. and Zandelin S. Energy Use of Set-top Boxes and Telephony Products in
the U.S. Lawrence Berkeley National Laboratory. 2001. LBNL-45305.
[13] Roth K., Goldstein R. and Kleinman J. Energy Consumption by Office and
Telecommunications Equipment in Commercial Buildings, Volume I: Energy Consumption
Baseline. Prepared for the U.S. Department of Energy, Office of Building Technology State
and Community Programs. 2002.
[14] Sustainable Solutions Pty Ltd. A Study of Home Entertainment Equipment Operational
Energy Use Issues, Final Report. Prepared for the Australian Greenhouse Office. 2003.
Report No: 2003/03.
[15] Energy Efficient Strategies and Energy Consult. Appliance Standby Power Consumption:
Store Survey 2004. 2004. MAEEEC Report 2003/04.
[16] Huber W. Standby Power Consumption in U.S. Residences. Lawrence Berkeley National
Laboratory. 1997. LBNL-41107.
[17] Lebot B., Meier A. and Anglade A. 2000. Global Implications of Standby Energy Use. Proc. of
the 2000 ACEEE Summer Study on Energy Efficiency in Buildings. Ed.: American Council for
an Energy-Efficient Economy, Washington: 2000. (Also published as Lawrence Berkeley
National Laboratory Report LBNL-46019).
[18] Ministerial Council on Energy Forming Part of the National Greenhouse Strategy. Money Isn't
All You're Saving. Australia's Standby Power Strategy 2002-2012. National Appliance and
Equipment Energy Efficiency Program. 2002. Report No. 2002/12.
[19] Rainer L., Greenberg S. and Meier A. You wont find these leaks with a blower door: the latest
in leaking electricity in homes. Proc. of the 1996 ACEEE Summer Study on Energy Efficiency
in Buildings. Ed.: American Council for an Energy-Efficient Economy, Washington: 1996.
Volume 1, pp 1.187-1.191.
[20] Ross J.P. and Meier A. Whole House Measurements of Standby Consumption. Lawrence
Berkeley National Laboratory. 2000. LBNL-45967.
[21] Appliance Magazine. Statistical Review. May 2000.
[22] KEMA-XENERY Itron RoperASW. California Statewide Residential Appliance Survey Final
Report. Prepared for the California Energy Commission. 2004.

1007
[23] RLW Analytics, Inc. California statewide residential lighting and appliance saturation study.
2000.
[24] Meier A., Rainer L. and Greenberg S. Miscellaneous Electrical Energy Use In Homes. In
Energy Vol. 17 (5), pp 509-518. 1992.
[25] Sanchez M., Koomey J., Moezzi M., Meier A. and Huber W. Miscellaneous Electricity Use in
the U.S. Residential Sector. Lawrence Berkeley National Laboratory. 1998. LBNL-40295.
[26] Zogg R. and Alberino D. Electricity consumption by Small End Uses in Residential Buildings.
Prepared for Office of Building Technology State and Community Programs. U.S. Department
of Energy. 1998.
[27] U.S. Department of Energy. Commercial buildings energy consumption survey (CBECS).
Energy Information Administration. 2001.
[28] Davis Energy Group. Spot The Big Spenders. Prepared for Pacific Gas & Electric. 1999.
[29] Kreith F. and West R. CRC Handbook of Energy Efficiency. CRC Press Inc. Chapter 10
"Electrical Energy Management in Buildings". 1997
[30] U.S. Department of Energy. Residential energy consumption survey (RECS). Energy
Information Administration. 2001.

1008
 Smart IC
and Power Supplies

1009
1010
Active and Passive Harmonic Compensation in Household
Appliances
Edson Adriano Vendrusculo1, José Antenor Pomilio2, Gilberto De Martino
Jannuzzi3
1,2
School of Electrical and Computer Engineering, State University of
Campinas,
3
School of Mechanical Engineering, State University of Campinas, and
International Energy Initiative Latin American Office –IEI-LA

Abstract
This paper points to the possible energy efficiency improvement in the electrical power distribution
grid due to the necessity of applying compensation in consumer electronic equipments. This
“necessity” has arisen as a consequence of distortions in the mains supply that triggered the
compensation strategies to force compliance with power quality standards. Technically, power quality
is the term used to encompass overall concerns about distortions in the mains supply. Taking power
quality issues into consideration, ordinary figures of merit are the measurement of the power factor
(PF) and the total harmonic distortion (THD) and, therefore, are focus in this paper. To improve the
power quality indexes active and passive compensation have currently been used. It is shown that the
passive, which is the conventional approach, presents drawbacks that may fall it into disuse. As a
result, an active compensation strategy is investigated for comparing purposes. The employment of
an active compensation strategy implies that an electronic converter should be used to drive the
household appliance. In fact, on driving a compressor used in refrigeration its efficiency may be
improved in the range of 30% to 40%. In the other hand, it is quite important to limit the harmonic
emissions of the electronic converter as it amounts to losses in the power line, transformers, and
cables are reduced since they are proportional to the square of the current. In this article a popular
(low cost) one-door model refrigerator was chosen as the base-case for analysis purpose. It has been
one of the most sold models and receives the one-star classification according to technical standard
ISO7371.

I. Introduction
The improvement of electrical appliances in general concentrates industry-wide efforts to deal with the
worldwide tendency to seek efficiency in energy consumption. Evidently, it should be mandatory in the
future while natural resources have in recent years been noticeably decreasing.
Higher efficiency may be achieved in household appliances by electronic control, based on feedback
from appropriate sensors of crucial variables such as temperature in refrigerators, motor speed in
washing machines and so on. While embedded electronic is actually an inherent characteristic, for
example in audio and video equipments, it has been emerging in recent years in HVAC (heating,
ventilation, and air conditioning) applications. HVAC accounts for significant percentage of electric
energy consumed in residential areas. Most of these applications utilize single-phase induction
machines for driving fans and compressors. Increase in the energy efficiency of these devices of up to
30% could be achieved by introduction of adjustable or multi-speed electronic drives [1]. In
refrigeration segment specifically, the variable speed compressors serve the possibility to adjust the
refrigeration capacity according to the load by controlling the motor speed.
However, not only to the appliance itself the increase in efficiency also has impact on the public low
voltage supply (i.e. the public 230V ac mains supply) due to the embedded electronic. A ‘public’
supply is one that is shared between more than one organization or household [2].
This paper focuses on the impact on the mains supply due to refrigerators with compressors driven by
power electronic converters (adjustable speed drives).
The aforementioned impact arises from the voltage distortions in the public voltage supply as a
consequence of the harmonic currents generated by the front-end stage of the power converters. The
harmonic currents are power quality-related problems, which may be minimized by using active and
passive compensation strategies.

1011
The objective of this article is also to thoroughly analyze these compensation strategies faced with
power quality requirements. A popular (low cost) one-door model refrigerator was chosen as the
base-case for analysis purpose, i.e., it is active and passive compensated. It has been one of the
most sold models and receives the one-star classification according to technical standard ISO7371.
This article is organized in the following way. Initially, a concise introduction of voltage distortions
related to power quality issues, namely total harmonic distortion and reactive power, is given. Section
III presents the electric characteristics of the original refrigerator, i.e. without applying power electronic
apparatus for speed control. Section IV deals with the non-linear load models and discusses passive
compensation and its drawbacks, section V discuss the alternative active compensation, which give
possible way to improve the refrigerator energetic efficiency without depreciating mains supply
distribution network efficiency and finally section VI presents some details about the current emissions
technical standard IEC61000-3-2.
One should keep on mind that active compensation minimizes THD (total harmonic distortion) and Q
(reactive power) quantities and therefore, increases the energetic efficiency of the mains supply
distribution network due to the reduction in the drawn currents.

II. Voltage distortions – power quality issues

Conventional power converters with rectifier-capacitor front-end have distorted input current waveform
with high harmonic content. Harmonic is defined as “sinusoidal component of a periodic wave or
quantity having a frequency that is an integral multiple of the fundamental frequency” [3]. Therefore,
harmonic is the presence of current with the frequency of a multiple of fundamental current in the
current of the system [3]. For example, Figure 1 (bottom) shows a waveform with 60Hz fundamental
frequency and third (180Hz), fifth (300Hz), seventh (420Hz) and nineteenth (540Hz) harmonics.
For reference purposes, a pure sinusoidal waveform with only fundamental frequency of 60Hz,
therefore free of harmonics, is shown in Figure 1 (top). The current waveform rich in harmonics shown
in Figure 1 (bottom) has been recorded in the variable speed drive with typical rectifier-capacitor front-
end stage shown in Figure 2 [2].

A sine-wave current 60 Hz

A typical non-linear current from a

single-phase rectifier-capacitor converter 60 180 300 420 540 Hz

Figure 1: Comparison of two waveforms and the corresponding spectral content

The rectifier provides unregulated dc voltage rails to the adjustable speed drive. Since the dc storage
capacitor is loaded at the peaks of the ac supply waveform the current consumption is discontinuous
and non sinusoidal as seen in Figure 1 (bottom). For this reasons these load type has a nonlinear
behavior.
The Figure 2 sketches out an approach to voltage distortions in the mains supply network at the point
of common coupling (PCC) due to harmonics drawn by the rectifier. PCC is the site at which other

1012
costumer’s loads are connected to and, therefore, may be affected by the voltage distortions. It is
quite fair to infer from Figure 2 that the “flat-topped” voltage waveform at the PCC is polluted by
harmonics. This distortion (“flat-top”) is consequence of the harmonic currents flow through the
unavoidable impedances associated with the mains supply. As the harmonics currents flow through
the mains supply impedance, they give rise potential differences that distort the waveshape of the
supply voltage at PCC.

Resulting mains voltage

waveform provide to
other equipment

A pure sine
wave is
generated

A non-linear load draws non-

sinusoidal current from the mains
supply
Figure 2: Voltage distortion at PCC due to current harmonics flow through mains impedance.

Typical harmonics quantities in single-phase rectifiers are illustrated in Figure 3, where they are
expressed as a percent ratio of the fundamental current [4]. Third harmonic is 73.2%, fifth harmonic is
36.6% of the fundamental one and so on. A particular problem with single-phase rectifier-capacitor
rd th th
power converters is that they emit significant levels of “triplen” harmonics (i.e. 3 , 9 , 15 , etc.), which
are a particular nuisance because they add linearly in neutral conductors (no phase cancellation) and
in zero-phase transformer flux, and they cause additional (and sometimes unexpected) heating of
cables and transformers [2].
80%
73,2%

70%

60%

50%

40% 36,6%

30%

20%

10% 8,1%
5,7%
4,1% 2,9%
0,8% 0,4%
0%
3 5 7 9 11 13 15 17
harmonics
Figure 3: Typical current harmonics in single-phase rectifier with capacitive filtering.

For the metering and comparison of harmonic contents of waveforms, a parameter has been defined
as a total harmonic distortion (THD). THD has been defined for both current and voltage as follows
[3]:

1013
 ∑ Xh
2
THDx = 100 (1)
X1
where x is the current (i) or voltage (v) quantities and Xh is the current (Ih) or voltage (Vh) harmonics.
The subscript character 1 indicates the fundamental component, i.e., the pure sinusoidal component.
Thus, it is now possible to calculate the THDi by (1) to be 82.59% for the harmonic spectrum in Figure
3. To put it simply, the voltage distortion at PCC (THDv) could be taken into account only to the
rectifier current flow. However, it is a tricky question because there is no certainty about the mains
impedance value. As a consequence it is usually considered that a THDv that exceeds 4% is a cause
of concern, and one that exceeds 8% is cause of alarm as it is likely to cause significant problems [2].
Note that the current components at harmonic frequencies do not contribute to the average (real)
power (P) drawn from the sinusoidal ac mains supply in Figure 2. In fact the THDi is related to the
power factor (PF) as follow:
1
PF = cosφ1 (2)
1 + THDi 2

The dimensionless power factor is a measure of how effectively the load draws the real power (P),
which represents the rate of useful work being performed. The harmonic currents drawn from
equipment’s ac mains supply have a negligible effect on its power consumption, measured in Watts,
and contribute to the power factor decreasing. Typically single-phase rectifier-capacitor has a power
factor lower than 0.65, i.e., quite far from the ideal of 1.0. As a consequence it’s assumed that there is
a reactive power (Q) flow in the network. Reactive power is energy produced for maintenance of the
network and is not produced for end-use consumption. The cos φ1 factor in (2) quantifies the
displacement between the voltage at PCC and the current fundamental component. Power converters
have a nearly unity displacement factor.
The power quality problems related with Q and THD may be kept to a minimum by using active and
passive compensation strategies.
In addition to the voltage distortions, the effect of harmonics can be noticeable in many ways such as
low voltage notching, communication systems (telephone) interference and high voltages and
currents in case of resonance. Further, harmonics may cause relay protection misoperations,
equipment failures and high overall system losses [5].
Regard to energy efficiency, the harmonics give rise to overall losses due to the “skin effect” in copper
and the increased eddy current losses in silicon steel used in mains distribution transformers at
frequencies above 50/60Hz, the thermal losses in conductors and transformers that carry harmonic
currents increase and they run hotter. In a larger installation with a lot of single-phase electronic loads
the excessive harmonic currents (hence overheating) can be as much as 1.7 times greater than the
highest phase current, due to the “triplen” harmonics that flow uncancelled in the neutral conductor
[2].

III. Base-Case Refrigerator with no Variable Capacity

Energy efficient motors in the past were designed to reduce internal losses, which in turn reduce
energy use. Enhanced engineering design and improved materials used in the manufacturing of
motors have resulted in energy efficient motors that generate less heat, use fewer lubricants and last
longer. However, it is shown in Figure 4 that current harmonics and reactive power were inherent due
nonlinear electromagnetic characteristic of the motor/compressor. In this case, the current distortion
depends on the design of the motor and varies with the voltage level [6].
Fig. 4 shows the measured current, voltage and power of the base-case refrigerator. The refrigerator
manufacturer’s assures a target consumption of 26.6 kWh/month in order to pay attention to the
1
PROCEL energetic efficiency labeling requirements. Experimental results concerning power quality
issues are the measured reactive power (Q) of 131VAr and the power factor (PF) of 0,64. These

1
With regards to refrigerators the National Institute of Metrology, Standards and Industry Quality (INMETRO) holds a labeling
program in partnership with the National Energy Conservation Program (PROCEL). Information to consumers on the energy
consumption of household refrigerators is provided through two labels displayed on refrigerators available at the retail market.
The Brazilian energy-efficiency label follows the European design showing energy efficiency class on a scale from A to G, for A
being the best performing appliances. Every appliance achieving class A efficiency level fits in PROCEL and is endorsed by
means of the specific label.

1014
results point to a non-energetic efficient appliance regard to the mains supply distribution network
since it draws 131VAr of reactive power to 108.5W of active (real) power. Figure 4 highlights the
reactive characteristic in the perceptive displacement between voltage and current waveforms which
is cos φ1= 0.58, i.e. the waveforms are not in phase (see displacement factor considerations in section
II). Based on the low distortion in the current waveform (low THDi), the low power factor (PF) of 0.65
in equation (2) is essentially linked to the high cos φ1. This conclusion is corroborated by current
spectrum in Figure 5 where the low amplitude of harmonics 3rd (6,2%), 5th (5%) e 7th (1,5%) may be
compared to the fundamental 60Hz.
Further, the voltage waveform at PCC in Figure 4 is with a minor distortion, i.e. THDv< 0.5% since in
this case the current harmonics do not have a prominent content.
If controlling techniques based on electronic resources are to be used for achieving higher energy
efficiency the use of rectifiers with dc smoothing filters (capacitors) become a common feature. As a
consequence the impact on mains supply distribution network changes from reactive power, for
original refrigerator, to the prevailing current harmonic components. Figure 6 shows the expected
current at input of the rectifier as discussed in section II.
The current is strongly distorted reaching the as high as THDi=108%. The factor power remains
around 0.66 and displacement factor of 0.97 is closer to unity, therefore, reducing reactive power
demand. On the other hand, the increases of the current harmonics cause voltage distortions at PCC,
which is THDv=3.2%. Accordingly, passive or active compensation should be applied to mitigate
harmonics. So, load models are evaluated providing insight into the requirements for the compensator
design.

Voltage, 127V (rms)

Tensão
Current, 1.33A (rms)
Corrente

Potência

Instantaneous power

Figure 4: Voltage, current and instantaneous power Figure 5: Current spectrum. Horizontal
of a refrigerator with no variable capacity. scale: 125Hz/div. Vertical scale:0.2A/div.
Horizontal scale: 5ms/div. Vertical scale: current,
2A/div; voltage, 100V/div and power, 250W/div.

Voltage Current

Power

Figure 6: Mains voltage, current and

instantaneous power in a rectifier-capacitor
front-end power converter.

1015
IV. Nonlinear Load-Type and Passive Compensation
The power quality problems are primarily due to voltage distortion caused by nonlinear behavior of the
loads as seen in Figure 6. Simulation models have represented two types of nonlinear load, i.e,
current harmonic source and voltage harmonic source. The models differ in the rectifier-filtering
element, which changes from a capacitor in the voltage source to an inductor in the current source,
respectively connected in parallel and series with the load. A rectifier with smoothing dc capacitors
behaves like a harmonic voltage source, rather than a harmonic current source [7]. Accordingly, this
type of nonlinear load has to be characterized as a harmonic voltage source and, therefore, the
current source is out of interest in this article. Figure 7 shows the rectifier and its equivalent model,
which has been used for analysis purposes.
Zo

V V
o
Harmonic source

f
Figure 7: Rectifier-capacitor power converter (left) and its harmonic equivalent circuit (right)

The conventional approach to passive harmonic compensation has been the parallel (or shunt) LC
th th
filter in which shunt low impedance branches traditionally consisting of 5 and 7 tuned LC series-
resonant filters are connected in parallel to a nonlinear load [7]. The Figures 8 and 9 shows
respectively the shunt LC filter connected at PCC and its equivalent circuit.

Point of Common Coupling (PCC)

Ii Zi IL Ii Zi I L Zo
PCC
Vi V
o
Vi Zf Zf

If If
ac mains shunt LC Rectifier-capacitor ac mains shunt LC Harmonic
supply filter front-end stage of supply filter voltage
adjustable speed source
driver
Figure 8: Basic principles of shunt LC filter for Figure 9: Equivalent circuit
harmonic voltage source compensation.

The compensation performance of a shunt LC filter for a harmonic voltage source load can be
expressed as:
Ii Zf
= (3)
Vo Z o Z i + Z o Z f + Z i Z f
It is clear from Fig. 9 and (3) that the compensation characteristics depend on the load impedance Zo
as well as the source impedance Zi. If Zo is zero, then (3) is reduced to =1/Zi, which means the
parallel filter is useless. On the other hand, if Zi is zero, then (3) is reduced to Ii/Vo =1/Zo, which means
the parallel filter is useless as well.
Experimental results of a circuit including a linear (inductive-resistive) and a nonlinear load (rectifier-
capacitor) connected at PCC are shown in Figure 10. In this case the former demands reactive power
th
and the latter is a harmonic-rich load. A 5 harmonic tuned LC filter is connected at the PCC,
consequently there is no phase shift between current and voltage. At this point the displacement
factor is close to “1” and the LC filter capacitance compensates the reactive power. However, the
current is noticeably distorted (has harmonics).

1016
 th
Table I indicates the PCC voltage harmonics. As expected, the shunt filter reduces de 5 component
th rd rd
(and also the 7 ) but increases the 3 and 9 harmonics. This unexpected effect can be understood
by taking into account the non-linear behavior of the load, i.e., the load reacts to changes in the
th
voltage waveform. The voltage THDv is decreased from 4.46% to 4.05% mainly due to the 5
harmonic reduction. The harmonics amplitude and THDi at the mains supply current (Ii in Figure 9) are
th
listed in the Table II. Likewise, the 5 tuned harmonic filter reduces this component but increases the
rd
3 harmonic and consequently the THDi as well. The fundamental component is reduced due to the
reactive power compensation.

PCC Voltage Current

Figure 10: Voltage and current distortions at the

PCC caused by reactive and harmonic-rich loads
th
with passive compensation ( 5 tuned LC shunt
passive filter). Horizontal scale: 5ms/div. Vertical
scale: current, 2A/div and voltage, 50V/div.

Table 1: Voltage harmonics (peak

value) at the point of common
coupling (PCC)
Harmonic With 5th Without
order harmonic the filter [%]
filter [%]
3 2.92 2.2
5 0.83 2.58
7 1.73 2.03
9 1.27 1.08
11 0.66 0.69
THD (%) 4.05 4.46

Table 2: Harmonics (peak value) of

the ac mains supply current
Harmonic With 5th Without
order harmonic the filter [A]
filter [A]
1 2.2 2.68
3 0.904 0.671
5 0.157 0.477
7 0.232 0.268
9 0.132 0.11
11 0.056 0.058
THD (%) 43.7 32.8

Review of these results shows that the passive compensation does not avoid the mains supply
distribution system of carrying harmonic currents and, therefore, decrease its energy efficiency
(discussed in item II). It is quite clear that the improvement of the energy efficiency is closely linked to
the power quality of the system.

1017
V. Active Compensation
An active front-end stage should be used in order to reduce harmonic currents of adjustable speed
drives. Accordingly, power switches (transistors) must be used rather than diode rectifiers or
semiconductor-controlled rectifiers (usually termed thyristor) since transistors allow full controlling of
the current. Most common switches are insulated gate bipolar transistors (IGBT) due to its high range
of application and as a consequence low cost.
In this article two, a high and a low switching frequency, active front-end stages are evaluated.

1) High switching frequency active front-end stage

Figure 11 illustrates an alternative to a high frequency active front-end stage for adjustable speed
drives [8]. A split capacitor bank forms the dc link. The switches Ta and Tb form the front-end rectifier.
The inverter to convert the dc-link voltage to a balanced two-phase output with adjustable voltage and
frequency features is configured with four switches Tc to Tf, respectively.

AC mains
supply
Vm

Voc

active front-end stage Motor/compressor drive Motor/compressor,

(ac – dc converter) (dc – ac converter) main and auxiliary windings
Figure 11: Active front-end stage applied to adjustable speed drive.

The single-phase AC supply is rectified by the front-end rectifier switches Ta and Tb. The split
capacitor bank (C1 and C2) in the dc link is charged through the diodes present in Ta and Tb. The
switches Ta and Tb are operated on a PWM pattern synchronized to the ac mains to shape the input
current to provide a unity power factor. The filter inductor Lm aids in filtering higher order current
harmonics.
The fundamental component of the voltage at points “0” and “c” is V0c,1, which is essentially the
reflected voltage due to the PWM operation of Ta and Tb. A sinusoidal PWM technique or an
advanced PWM technique that selectively eliminates several lower order harmonics may be used to
control Ta and Tb. In this article the former has been evaluated and therefore the harmonic spectrum
has only high harmonic frequencies, which would be mitigated through (electromagnectic
2
interference) EMI filters .
Results from simulation in PSIM/SIMCAD software are following shown. Current in such a front-end
rd
stage is shown in Figure 12 to its respective spectrum in Figure 13. The 3 harmonic (180Hz) is
thirtieth times lower than the fundamental at 60Hz, which is sufficient to comply with harmonic current
standards such as IEC 61000-3-2.
Since harmonic current and reactive power are the main focus in this article only concise information
about control is provided as follow: two nested control loops are used, the inner guarantees the
sinusoidal current shape and the high power factor, and the outer keeps the total dc voltage with a

2
EMI filters are designed for complying with technical standards that limits frequencies amplitude in the range of 150kHz up to
higher frequencies.

1018
stable value. An equalization loop may be also needed to equally share the voltage in the split
capacitors C1 and C2.

60Hz

180Hz

Figure 12: Simulated voltage and current in Figure 13: Ac mains supply current spectrum.
the ac mains supply. Switching frequency of Horizontal scale range: 20 to 300Hz.Vertical
10kHz scale in dB.

Results from an experimental prototype under test are shown in Figure 14, designed for output power
of 220W, total dc voltage of 300V, C1=C2=450µF and Lm=2mH. Although the voltage distortion looks
satisfactory a closer examination reveals that control design could be improved in order to phase
voltage and current. The voltage in capacitors could be better balanced through applying an
equalization loop.

Means values: Split capacitors

165V and 150V voltage

ac mains supply
voltage and current
RMS values: 60V
and 2.38 A.

Figure 14: Experimental results. Vertical scale: voltage

(50V/div) and current (5A/div). Horizontal scale: 5ms/div.

Even though, the active high-frequency front-end rectifier satisfactory raises the energy efficiency at
PCC, due to the fast switching inside the drive, there's a risk of electromagnetic emissions, which can
take the form of conductive and radiating interference. It does not overall affect efficiency but, in turn,
interfere in other appliances work. International regulations set limits on both low- and high-frequency
emissions. With the use of filters, screening, and suitable mechanical construction inside the drive
cabinet, it's possible to meet the electromagnetic compatibility (EMC) standards. Other suitable
alternative is to use line-frequency commutated auxiliary AC switch in traditional passive rectifiers as
is going to be discussed below.

1019
b) Front-end stage coupled to line-frequency commutated AC switch
The line-frequency commutated AC switch coupled to split capacitors shown in Figure 15 is able to
greatly improve both power factor and output voltage regulation of rectifiers with passive L-C filters [9].
As seen the AC switch has bi-directional current flow capability. It is getting through the arrangement
of the diodes and the uni-directional switch.
The switch is turned on with some delay regard to the line voltage zero crossing. It allows having
current flow start close to the voltage zero crossing and therefore reducing the displacement between
voltage and current as illustrated in Figure 16. This auxiliary switch starts a resonance phenomenon
between inductor L and both capacitors C1 and C2. During the resonant interval the input current
splits almost equally between the capacitors. This circuit has a boost action introduced by the
commutation cell, which through compensation of the voltage drop across the input filter inductor
allows the achievement of dc voltages higher than the peak of the ac mains supply voltage.

Bi-directional AC switch

C1

L
C2

Figure 15: Line-frequency commutated Figure 16: AC mains supply voltage and current
front-end rectifier waveform. Total link CC voltage and switch
pulses.

Despite of its slightly distortion the current waveform in Figure 16 comply with the limits for harmonic
current emissions set by IEC 61000-3-2 technical standard.

VI. Current emissions technical standard (IEC 61000-3-2)

Compliance with IEC61000-3-2 is now a requirement (for all equipment within its scope) for
conformity with the Electromagnetic Compatibility Directive [10].
The standard establishes four classes of equipment, each with their own harmonic emissions limit.
™ Class B for portable equipments
™ Class C for lighting equipment, including dimmers
™ Class D for Personal Computers (and their display monitors) and TV receivers, with a
“specified power” less than or equal to 600W.
™ Class A for everything else, particularly balanced three-phase equipment.
Adjustable speed drives for compressors/motors in refrigerators are class A limited due to the front-
end rectifier. The emissions limits EN61000-3-2 Ed.2:2000 for each class are listed in Table III.

1020
Table 3: EN61000-3-2 current harmonics limits
Harmonic order Max current Max current Max current Max current
‘n’ Class A Class B (% of fundamental Class D
current) (but no more than Class A)
2 1.08 Amps 1.62A 2% Not specified
3 2.30A 3.45A 30λ% 3.4mA/Watt
4 0.43A 0.645A Not specified Not specified
5 1.14A 1.71A 10% 1.9mA/Watt
6 0.30A 0.45A Not specified Not specified
7 0.77A 1.155A 7% 1.0mA/Watt
8 ≤ n ≤ 40 0.23 (8/n) A 0.345 (8/n) A Not specified Not specified
(even)
9 0.40A 0.6A 5% 0.5mA/Watt
11 0.33A 0.495A 3% Not specified
13 0.21A 0.315A 3% 0.35mA/Watt
15 ≤ n ≤ 39 0.15 (15/n) A 0.225 (15/n) A 3% (3.85/n) mA/Watt
(odd)
where λ is the circuit power factor and n is the harmonic order.

The harmonic limits are absolute values for Classes A and B, whatever the input power. The Class C
limits are expressed as percentages of the 60Hz current consumed, and for Class D they are a set of
sliding values proportional to the mains power consumed. For equipment with an input rating greater
than 600W the Class A and Class B limits, being fixed, become proportionately more difficult to meet
as the mains input power increases.
Compared to the Class A the current draw by the line-frequency commutated front-end rectifier is
within the current emissions limits as shown in Figure 17.

2,5

1,5
[A]rms

Class A

0,5

0
3 5 7 9 11 13 15 17 19 21 23
harmonics
Figure 17: Line-commutated rectifier complies with IEC 61000-3-2 technical standard.

VII. Conclusions
It has been shown that power quality indexes, such as Q and THD, and the energy efficiency of the
mains supply distribution network are closely linked. Thus, a design concerning with these indexes
avoids depreciating the network overall efficiency while improves the energy efficiency of household
appliances.

1021
Power converters with rectifier-capacitor front-end stages are usually embedded in household
appliances for achieving higher energy efficiency since electronic resources allow implementing
control techniques.
Nevertheless rectifiers do not determine the current drawn from network, otherwise, it imposes the
voltage at the PCC. As a consequence, the conventional approach to passive harmonic
compensation, i.e. shunt filters, cannot effectively filter out the resulting current harmonics.
If one is interested in reducing current harmonics at PCC, the best alternative according to (3), is to
increase both, the series impedances to the mains supply (Zi) and the load series impedance (Zo). In
conclusion the individual compensation is unfeasible for rectifier type loads because they already
have a capacitive fundamental displacement factor. Thus, active solutions considering high and low
switching frequency have been investigated for applying in an adjustable speed drive.
The presented results showed that both active solutions comply with technical standards, although
the line-frequency commuted converter is very close to the limits of the standard. It helps to choose
better solution, however a more careful analysis should be done taking into account the performance,
cost and volume of a prototype. It would also be interesting to note that the solution based in the line-
th th th
frequency commutation complies with IEC61000-3-2 despite of drawing low amplitude 3 , 5 , 7 , and
other harmonics. That is quite different from the high switching frequency based solution, which
essentially transfers energy at the high switching frequency.So, strictly taking into consideration the
power quality indexes the high-frequency based solution will not significantly decrease the energy
efficiency of the mains supply distribution network.

VIII. References
[1] Chomat M. and Lipo T.A. Adjustable-speed drive with single-phase induction machine for
nd
HVAC application. 32 Power Electronics Specialists Conference - PESC (17 June - 21 June
2001). Volume 3, pp. 1446 – 1451.
[2] Keith Armstrong. EN 61000-3-2 Limits for Harmonics. Can be downloaded at:
http://www.reo.co.uk/guides
[3] Emadi A., Nasiri A and Bekiarov S. B. Uninterruptible power supplies and active filters. 2005 by
CRC Press LLC Ed. ISBN 0-8493-3035-1
[4] Mohan N., Undeland T.M. and Robbins W.P. Power Electronics- Converters, Applications and
Design. 1995 by John Wiley & Sons Ed. ISBN 0-741-58408-8.
[5] Mihirig A. Harmonic Study Analysis Guidelines for Industrial Power Systems. Electricity Today
Magazine. Issue 3, 2001. Can be downloaded at: http://www.electricity-
today.com/et/issue0301/i03_harmonic.htm
[6] Pomilio J. A. and Deckmann S. M. Characterization And Compensation For Harmonics And
th
Reactive Power Of Residential And Commercial Loads. 8 Brazilian Power Electronics
Conference – COBEP (June 2005). Recife, Brazil, pp 599-604.
[7] Peng F.Z. Harmonic Sources and Filtering Approaches. IEEE Industry Applications Magazine _
July/August 2001.pp 18-25.
[8] Enjeti P.N. and Rahman A. A New Single-phase to Three-phase Converter with Active Input
Current Shaping for Low Cost ac Motor Drives. IEEE Transactions On Industry Applications,
o
Vol. 29, N . 4, July/August 1993.
[9] Spiazzi. G., Martins E. da S. and Pomilio. J.A. A Simple Line-Frequency Commutation Cell
Improving Power Factor and Voltage Regulation of Rectifiers with Passive L-C Filter. IEEE
Power Electronics Specialists Conference, PESC (17 June -21 June 2001). Vancouver,
Canada, pp. 724-729.
[10] European Union Directive 89/336/EEC of 3 May of 1989 on Electromagnetic Compatibility. Can
be downloaded at: http://europa.eu.int/comm/enterprise/electr_equipment/emc/index.htm

1022
Increasing Efficiency in Appliances, Office Equipment and LED
Lighting
Douglas Bailey

Power Integrations

Abstract
New energy efficiency regulations with tight standards are driving original equipment manufacturers
(OEMs) to make their products operate more efficiently. Markets such as domestic appliances, small
office/home office equipment and consumer electronics are very cost competitive, so the means used
to increase the efficiency of the products sold in those markets must be affordable for OEMs to
implement. Highly integrated power conversion ICs already exist, making it relatively painless for
OEMs to quickly, easily, and cost effectively design energy-efficient power supplies for their products.
When this fact is understood, many OEMs are motivated to bring their products into compliance since
that can give them a distinct advantage over other manufacturers who’s products do not yet comply
with the new standards.
This paper looks at some power budget aspects of power supplies for products that range from
washing machines and LED lighting to ink-jet printers, and shows how their power supplies can be
radically simplified and made compliant to all of the current (and proposed) worldwide energy
efficiency standards, by redesigning them around highly integrated power conversion ICs.

Introduction
Various governmental and standardization bodies around the world, committed to protecting our
dwindling energy resources, have enacted power supply regulations to help save energy and further
minimize power supply costs. However, if the power supply efficiency regulations focus on making
power conversion more efficient, but the system-level product misuses the converted power, any
potential energy savings may be lost. For this reason, new regulations are emerging—such as the
ENERGY STAR “imaging products” and “computing equipment” specifications—that define the
number of watts required to power standard system-level functions. This presents the designer with
more design-for-efficiency decisions and choices. The designer can choose whether to derive energy
savings from the power supply, from the functional system, or from a combination of both. Applying
this “power budget” approach to standby power usage has resulted in the introduction of many
efficient products (popularized by the 1-Watt Standby movement). This paper looks at a few of the
more common applications where a power budget approach was used to reduce power consumption
in standby modes and shows how the derived savings can be applied to other modes.

White Goods
Washing machines and white goods, in general, require small amounts of low-voltage DC power to
operate their controls and status displays. Typically, the physical enclosure of the machine provides
sufficient isolation, enabling the use of a non-isolated type of power supply that is both easy-to-
implement and inexpensive. One of the most popular types is the post-regulated capacitor-dropper
(see Figure 1), which uses a minimum number of discrete components to provide the small amounts
of current needed to drive the system’s electronics. However, the efficiency of a cap-dropper is about
30 percent at the full load of 50 mA, making the conversion of power for the control systems extremely
inefficient. In standby, the efficiency drops to 18 percent. In the example shown in Figure 1, the 360
mW of power absorbed by the microcontroller and other electronics was costing a continuous 2 W of
power drain. This equates to the energy of one wash cycle every two weeks, or about 10 percent of
the total energy usage of the washer over a one-year period. To achieve a 1 W power budget on the
standby power of this same appliance, the designer would have to redesign with a more efficient
power supply technology. The use of a non-isolated buck (see Figure 2) can provide 720 mW of
power for just 1W of input power (see Figure 3), and free 360 mW of power for use in driving the
electronics. Put in practical terms, the benefit of using the non-isolated buck in the redesign is this:

1023
any consumer wishing to use less power in standby than in wash mode can limit laundry days to a
bachelor-approved “once-a-month.”

50%

40%
Efficiency

30%

20%

10%

0%
0,00 0,20 0,40 0,60 0,80 1,00
Wout [W]

Figure 1: Efficiency curve of a typical, post-regulated, capacitor-

dropper power supply.

1, 2
Figure 2: A non-isolated, buck power supply designed to replace
capacitor-dropper supplies.

Figure 3: Efficiency curve of the non-isolated buck power supply.

1
See LinkSwitch-TN datasheet.
2
See LinkSwitch-TN Design Example Report 92.

1024
Printers
The power used in inkjet printers varies from ≤ 1 W in standby mode to 70-to-80 W in peak-power
mode (generally when the paper feed motors are engaged). Given this large dynamic range, the
power budget for a printer is best viewed as the average of operation over a substantial period.
Figure 3 shows the power usage of a typical inkjet printer. The standby-mode functions are well
defined, and can be generally summarized as follows: check the “ON” switch for activity, and keep the
“Power Connected” LED illuminated. However, printers also have a sleep mode that permits fast
power-up when a print-job is issued: this mode needs to be considered when using a power-budget
approach to design. To maximize the use of available power across the printer’s entire power range,
the power supply needs to deliver efficient power from standby to full peak-power modes. Figure 4 is
the block diagram of such a power supply, designed using an integrated switched-mode IC with peak-
power capability. Figure 5 is the efficiency curve of the supply depicted in Figure 4.

3
Figure 4: A power supply designed around a switched-mode IC with peak-
power capability.

Figure 5: Efficiency obtainable using a switched-mode IC with peak-power capability.

3
See PeakSwitch datasheet.

1025
 2
Power delivery, proportional to P = 0.5 • LI f, where L = transformer inductance, I = current limit, f =
frequency), is predefined for the design. The current limit, which is defined in the ON/OFF control
scheme, is also predefined to one of 4 levels. The integrated switch-mode IC with peak-power
capability changes average switching frequency to provide variable amounts of power across the
range required by the printer, as shown in Figures 5 and 6.

Figure 6: Peak-power management using intelligent, switched-mode IC.

LED Lighting
California’s new Title 24 brought advancements in efficiency to residential lighting installations. The
regulation requires that, under most circumstances, high-efficiency lighting be installed on the hard
switch points in every room. The target for lighting efficiency is 40 lumens per watt, which is out of the
range of most incandescent technologies, making fluorescent or LED lighting a necessity. The
current standard, which ignores the ballast losses of the incandescent light, is likely to change to
address the need to manage these losses. LED use in general lighting is beginning to gain
momentum, and can achieve 40 lumens per watt when a highly efficient power conversion technology
is used. A luminaire using LED sources with a conversion efficiency of 50 lumens per watt requires a
power conversion efficiency of at least 80 percent to deliver 40 lumens per watt (input power to output
light). For a 15 W luminaire, this requires losses in the luminaire to remain at or below 3.75 W.
Figure 7 shows an example of an LED lighting source designed using this power budget approach.

1026
 4, 5
Figure 7: Energy efficient power supply for an LED lighting source.

Summary
New regulations are forcing engineers to apply pre-defined power budgets to their design. Optimizing
for power conversion efficiency in the supply unit is no longer regarded as the only means to reduced
energy usage. Given the tighter regulations governing system-level power usage, the designer must
now look at ways to derive energy savings from the power supply, from the functional system, and
from a combination of both. Taking a power budget approach to design results in more power-
efficient products in standby mode and can lead to new orders of efficiency in no-load and peak-
power modes, as well.

References
®
[1] Power Integrations, LinkSwitch -TN data sheet, revision G, March 2005.
http://www.powerint.com/PDFFiles/lnk302_304-306.pdf
[2] Power Integrations, Design Example Report 92 (DER-92) revison 1, August 2005.
http://www.powerint.com/appcircuits.htm
[3] Power Integrations, PeakSwitch™ data sheet, revision A, March 2006.
http://www.powerint.com/PDFFiles/pks603-606.pdf
[4] Power Integrations, TOPSwitch®-GX data sheet, revision O, November 2005.
http://www.powerint.com/PDFFiles/top242-250.pdf
[5] Power Integrations, Design Example Report 100 (DER-100) revison 2, December 2005.
http://www.powerint.com/appcircuits.htm

4
See TOPSwitch-GX datasheet.
5
See TOPSwitch-GX Design Example Report 100.

1027
1028
Monitoring and Control Systems to Manage Energy Use in US
Homes
Michael Breton1, Todd Brady2, Eric Williams3, H. Scott Matthews3
1
Intel Corporation, Sacramento, California, USA
2
Intel Corporation, Phoenix, California, USA
3
Dept. of Civil and Environmental Engineering, Carnegie Mellon University,
Pittsburgh, Pennsylvania, USA

Abstract
Energy use in homes represents 21% of US total energy demand in 2004. Managing this sector is an
important priority for addressing global warming, conserving resources and improving energy security.
Much energy is wasted in delivering energy services not actually used by residents. IT-enabled
monitoring and control technologies have played an important role in eliminating similar inefficiencies
in other sectors, they could have an important role in the home as well. The technology level of
energy control in most residences is at least 20 years old, with simple programmable thermostats still
in only about a quarter of US homes. Networked thermostats, power meters and switches, and zone
heating are technologies that can that can monitor energy and enable control for delivery only where
and when needed. In addition to direct energy savings, there is also the potential to reduce indirect
needs for energy infrastructure through peak shifting, or redistributing of electricity demand more
evenly throughout the day. In addition to surveying these energy management issues, this article also
relates the experience of a pilot project setting up monitoring/control systems in three Sacramento
homes. The design specifications of these systems combine capabilities for web-based monitoring
and control and peak shifting via pre-cooling, and load shedding. The pilot has shown that such a
monitoring and control system satisfying the design parameters can be implemented via mainly off-
the-shelf parts. Much work remains to be done however, to develop low-cost user friendly systems
attractive to typical homeowners.

Introduction
Mitigating energy use in homes is an important challenge. Residential energy use accounted for 21%
of US total energy demand in 2004, and grew a total of 16% over the last 10 years [1]. Improvements
in residential energy efficiency have been offset by the growth in average home size, rising from 1400
2 2 2 2
ft (130m ) in 1970 to 2200 ft (204m ) today, as well the proliferation of additional electrical
appliances [2]. Considering growth in telecommuting, the use of e-commerce, and digital home
entertainment, it is likely that time spent at home (and thus energy use) is increasing, making the
residential sector an even more important target to mitigate demand. In addition, housing is one of the
more inefficient energy sectors both in terms of technology and management.
On the technology side, while there have been improvements in areas such as insulation and furnace
efficiency, major innovations such as heat pumps and solar/wind generation remain at the fringes.
Inefficient transformers and power supplies continue to cause significant losses in small appliances.
While there is clearly much to be done to improve these “hard” technologies, there are also vast
opportunities to lower energy use through better energy monitoring and control, and increased
awareness. A typical example of the former is heating and cooling residences when no one is at
home. Improved control and management of energy use at home using information technology has
great potential to reduce residential energy consumption. Networkable sensors, meters, and switches
continue to decline in price and improve in sophistication. These elements can be integrated into
energy monitoring and control systems that inform residents of how and where energy is being used
and provide automation of many actions to mitigate consumption.

Energy use in residences

To set the context for the discussion and analysis, in this section we review the structure of energy
use in US homes. Table 1 shows average use in space heating, air conditioning (AC), water heating
and appliances in US residences [2]. This data comes from the Residential Energy Consumption

1029
Survey (RECS) of the US Department of Energy (DOE), which is the central source of such
information. Energy use in column 1 is what DOE terms as “primary energy”, which adds the direct
energy of fossil fuels and the secondary energy needed to produce electricity (a factor of 3.03
6
megajoules (MJ = 10 Joules) of input energy per MJ of electricity). Energy use per household is
obtained by dividing total energy use in each category by the total number of households, 107 million.
Note that not all homes have air conditioners (26 million do not), the average energy per household
with an air-conditioner is 24,500 MJ, considerably higher than the figure in Table I.
The share that appliances account for in total energy use continues to increase over time, and indeed
the rise in total energy demand in residences is largely driven by the use of more appliances. We also
mention that yearly expenditures on utilities are increasing rapidly due to rises in natural gas and
electricity prices.

Table 1: 2001 average energy use in a US residence [2]

Energy use per Energy Expenditures per
household (MJ) Share household ($)

Space heating 53,000 29% 480

AC 19,000 10% 197
Water heating 23,000 13% 203
Appliances 88,000 48% 670
Total 183,000 1,550

Applications of monitoring and control systems for managing residential

energy use
In this section we overview major aspects of residential energy use which can be addressed through
monitoring and control technologies.

A. Delivering heating/cooling when and where needed

A substantial amount of energy is used for heating and cooling residences in unoccupied homes and
rooms, and to a lesser degree, lighting. RECS data on heating suggests that on average thermostat
settings when people are not at home are nearly the same as when they are. This suggests that the
practice of leaving heating (and presumably cooling as well) full on when not at home is common.
Also, most homes in the US have central heating/cooling, systems that make it difficult, if not
impossible, to differentially climate control in different areas of the home. Thus substantial energy is
used to heat and cool unoccupied rooms. For example, keeping an entire home climate controlled at
night when only bedrooms are occupied represents an obvious opportunity for energy savings.
Space heating, and to a lesser extent air conditioning, are usually controlled from one central
thermostat. This often leads to substantial temperature variations through different parts of the house.
In particular, the common placement of the central thermostat on the first floor results in upstairs
temperatures that will generally be higher than the downstairs setting. These area variations in
temperature lead to inefficient use of energy due to “oversetting” of thermostats.

B. Enhanced awareness of energy use leading to behavioral shifts

Energy monitoring systems can influence behavior by making residents more aware of the economic
and environmental implications of their home energy decisions. Given that the only information source
on energy use in the home for most people are monthly utility bills, its structure remains a black box.
This makes choices of actions to mitigate energy use difficult. Imagine a system in which the
breakdown of real-time electricity use of heating/cooling equipment, water heaters, and a variety of
appliances are viewable through via computer or television with an easily understood graphical
interface. Armed with such information, consumers could make much better choices in decisions of
what equipment to purchase and how to use it. In terms of purchase decisions, if consumers know
how much electricity is being used by different equipment as well the possibility to reduce this use by
opting with a different model, this information could influence purchase decisions. Also, there are a
variety of energy-related behaviors, some discussed earlier such as leaving heating and cooling on
when not at home, that consumers might reconsider given knowledge of the cost implications.
There are pilot projects examining the effect on behavioral changes induced by a monitoring system.
For example, a recent Japanese study found that energy savings of up to 10% per household were
achieved with a monitoring system providing real-time information on heating, cooling and select
appliances [3]. While we know of no equivalent trial in the US, there is a pilot project in California

1030
which explores residential response to increases in the peak price of electricity, to be discussed in the
next section.

C. Peak shifting: precooling and load shedding

There is also the possibility to use monitoring and control systems to mitigate peak-time electricity
use. Demand for electricity varies substantially on daily and seasonal time scales, peaking during the
afternoon of summer days. In warmer climates, the peak can be twice higher than the typical
baseline. The need to provide additional electricity during this peak makes utility construct a power
generation and distribution infrastructure larger than what is generally needed. There are
environmental impacts associated with building and maintaining this additional infrastructure. For
example, for a typical nuclear plant, the energy embodied in construction, operation and management
of the plant is about 0.2 megajoules per kWh of electricity generated [4,5]. In addition, high peak
demand makes it more difficult for utilities to maintain stability, increasing risks of brownouts and
blackouts.
One strategy to manage peak demand is known as peak shifting, which involves redistributing the
time an energy service is delivered away from the time of peak demand. For the nuclear plant above,
the energy investment required implies that one can spend up to an additional 5% of electricity on
peak shifting, and still realize a net energy benefit due to the reduced need for power infrastructure.
One way to peak shift is through precooling, which involves overcooling a building before the heat of
the day, thus reducing AC electricity use during the peak period. While peak use goes down,
precooling is expected to increase total electricity use [6]. This is because temperature differentials
inside and outside are higher for precooling, increasing the rate of heat leaking into the house and
lowering the coefficient of performance for A/C. However, given the secondary energy use and other
environmental impacts associated with the infrastructure to deliver peak electricity, there clearly exists
a crossover point where the energy saved through avoided infrastructure exceeds the additional
electricity demand.
In additional to environmental considerations, pre-cooling presents significant cost savings
opportunities to the consumer. Many utilities offer rate-saving plans to encourage consumers to more
energy during off-peak times and use less energy during peak periods. This is because the
incremental cost to the utility of electricity only needed at peak times is relatively high. Dependent on
the size of the home and climate, these savings can run from $10-$100 each month.
Another strategy for peak shifting is known as load shedding. This refers to the capacity for utilities to
restrict electricity use during peak times. There are a variety of schemes to implement this idea,
clearly they all depend on the utility being able to remotely control some aspect of resident’s energy
system. Another strategy is to substantially increase the price electricity during critical peak hours to
motivate consumers to reduce energy use during those times. [
A recent case study in California explores both automated load shedding and shifts in behavior of
residents in response to peak pricing [7]. One group of residences were equipped with automation
that shut off HVAC systems during peak times, while another was left to manually change behavior
given prior information on what days peak pricing would be implemented. Both groups displayed
reductions in peak use, -4% to -13% for the manual group and -25% to -41% for the automatic group.
However, there was no clear pattern of reduction of total electricity use, with net use increasing in
some cases and decreasing in others. This is despite the fact that 5-20% of residents reported taking
actions that would reduce overall energy consumption, such as turning off lights, and reduced laundry
water temperature) [8]. It is possible that the additional electricity burden of precooling tended to
cancel these beneficial actions.
The adoption of monitoring and control system in homes remains limited. The mainstay control
system currently used in homes is the programmable thermostat, but only 27% of residences have
these [2]. RECS survey results also indicate little difference between thermostat settings during day
when occupied, unoccupied, and during sleeping hours. This suggests that most programmable
thermostats are not actually being used. In terms of monitoring, homeowners generally have no
information on energy use beyond the monthly utility bill. To sum up, it would seem that the last
several decades of progress in information technology has yet to be applied in the monitoring and
control of home energy.

Overview of residential monitoring and control systems

There are many possible monitoring and control systems that could be built for residential energy
consumers with current technology. The structural elements of such a system are:
a. Wired (especially through power line) and wireless network architecture/protocols

1031
b. Programmable and networked thermostats
c. PC/input-output device hub
d. Controllable/networked vents (zone heating/cooling)
e. Sensors (temperature, flows (e.g. of natural gas))
f. Software systems to monitor and control home utilities via hub

However, while there are products available on the market in each of these areas, it is fair to say that
the challenge of how to optimize each component and integrate into a system attractive to a typical
homeowner has yet to be achieved. Working towards such a system is the objective of the case study
described in following sections.

System overview
From here we introduce a pilot project initiated in February 2005 between the Intel Folsom Innovation
Centre and the Sacramento Municipal Utilities District (SMUD). Three equivalent monitoring and
control systems were developed and deployed in Sacramento, California, USA. These systems were
designed to combine capabilities for web-based monitoring and control, offpeak pre-cooling, and load
shedding. In all three homes gas is only used for heat and hot water. Heat in not used over the
summer months of the pilot period. Major electric users are A/C, dryer, and a home pool pump for
one home.
The target of the web-based monitoring and control system is to allow users to view and control
energy data via a home URL with a graphical user interface. The off-peak precooling system is
essentially a programmed thermostat schedule implemented on days when peak demand is expected
to be high. The load shedding system target is an implementation of On Demand Load Shedding, in
which the utility would publish (on the internet) a request to shed load. The automation system in the
digital home periodically checks and based on the utilities demand notice, notifies the homeowner
(email, SMS message to their cell phone, or voice in the home) that it is responding to the utilities
request and then “up-tick” the home’s thermostat by 2-4 degrees. The homeowner could override if
desired. When the demand period has ended, the automation system would set the thermostat back
to its normal set point. Participation in the program would result in lower energy costs for the
consumer.
The central platform consisted of a computer connected to the home’s primary entertainment display
(TV) for viewing purposes. Home Automation and other specialized software were installed to collect,
normalize, store, and display information from temperature sensing devices, as well as a whole house
energy meter. Interfaces included radio frequency reading of the whole house energy meter and a
command and control interface for HVAC. Internet access was used to transfer data to centralized
project owners as well as to provide remote access to data and control systems. Wireless and
powerline technologies were used for meter and control system communications.
The systems used in this study used Intel-based PCs running Windows Media Center edition. This
operating system (OS) was chosen due to its built in navigation, user interface, and display
technologies. This same OS supports advanced features such as personal video recordings, home
music and video playback. However the programs used would run on any Windows-equipped PC. In
general, though, a goal of the project was to demonstrate how energy awareness and home
automation features could be added to existing home computer systems. Since Windows Media
Center computers are specifically designed to already be “always on”, e.g., to make personal
television broadcast recordings on demand, we did not consider the electricity burden of the central
computing system in this analysis. In reality, the marginal electricity use of any of the home
automation or control subsystems would be relevant, but we perceive these marginal burdens to be
small.
The system is composed of several key components, which are summarized below.

A. Energy Meters
The existing traditional analog meters for residences were replaced with wireless digital meters
attached to each home’s electrical panel and are of a type now commonly used in new homes in the
Sacramento area. A major motivation for utilities to switch to wireless meters is that monthly use for
billing purposes can be collected remotely, realizing significant savings in labor costs. The specific
type of meter used broadcasts its reading every 2-3 minutes and can be received via a special
wireless reader attached to the PC. The reader connects via a standard USB port and is accessed
via terminal emulation software.

1032
B. Temperature Sensing
In the scope of this project, only one device type was needed in the home control space, HVAC.
HVAC was controlled with Residential Control Systems thermostats using the X10 protocol for
powerline communications. Typical X10 reliability issues such as low signal strength and failures due
to noise on the powerline were encountered and resolved by installing phase couplers and noise
filters. Next generation networking technologies such as Universal Powerline Bus or Z-Wave would
have prevented these communication issues, but such devices were not commercially available when
the project started. We expect that networking capabilities enabled by such technologies will be
needed to implement the kinds of advanced features desired in next generation monitoring and
control systems. Inside the home, on wall thermostat displays showed unit status and allowed for
manual control. A separate thermostat was used to record outdoor temperature. At the PC, a USB
based X10 interface was used to communicate with the thermostats over the powerline.

C. Software
Several commercial and custom software packages were used to build out the entire system.
• HAL2000 from Home Automated Living (HAL) was used to provide the core home automation
solution. This system provides voice, web, PDA, touch screen and Media Center interfaces into all
aspects of home control. In the case of this project, it was used to record temperate data from all
thermostats in the home, including outside temperature, as well as to control the set points and
modes of the cooling system, and was specifically used for implementation of the pre-cooling
schedules in the home.
• To gather data, a custom application was written to log temperature data from the thermostats
in the home every 15 minutes.
• A terminal emulator was used to communicate with the digital meter reader. This application
also records date and time of the meter reads as they were written to the text file.
• For the On Utility Demand Load Shedding scenario, a custom application using freeware
components and the HAL interface was used to demonstrate how HAL could “watch” a utility’s
website for critical load condition notification and then take load shedding action.
• Additionally, custom web pages were developed for display within Media Center using the
Media Center 2005 SDK and a 3rd party XML based Flash charting system (FusionCharts) to create
the “My Energy” portal providing current and historical views in to the homes energy consumption,
cost, and temperatures

Figure 1 shows the resulting interface showing results for hour by hour electricity use for one home on
August 23, 2005, which had a peak outside temperature of 93degrees C. The steep increase in use
after 10A.M. reflects electricity use for A/C.

D. Hardware and Software Costs

This section summarizes the costs of the system deployed in the homes in the case study testbed.
Each home used a Windows Media Center-equipped PC ($1500), one x-10 interface ($40), two $250
thermostats, automation software ($400), and $50 of miscellaneous costs, totaling slightly less than
$2,500 per house. As mentioned previously, although a Windows Media Center PC was used in this
pilot, the software suite would run on most Windows XP desktop computers, commonly found in
homes today, albeit with less of an integrated experience. Doing so would reduce the cost to less
than $1000 per home. It is important to note that the project goal was to build prototypes with off-the
shelf technology rather than to find components and technology at scale or minimum costs. In reality,
the costs at large-scale adoption could likely be in the $200 or less range (excluding PC).

1033
Figure 1: Screenshot of “MyEnergy”: thermostat set at 78 degrees F (26C), max outdoor
temperature 95 degrees F (35C)

Results
Functionality of the web-based monitoring and control, pre-cooling and load shedding were
established. Test runs on the energy effects of precooling were carried out for several summer weeks.
The precooling schedule was as follows:
• 8am - set both thermostats to 73 degrees
• 2pm - set upstairs to 85 degrees & downstairs 78 degrees
• 8pm - set upstairs to 78 degrees

A. Precooling Results
Precooling results for the residences studied showed that overcooling the house during 8-12:00
effectively shifted the peak of AC electricity use to that period. This can be seen from Figure 2, which
displays electricity use for one home on August 24, 2005, which had a maximum outside temperature
of 90 degrees F. Consumers would save on electricity expenditures because they would be
purchasing it at a less expensive time. In terms of energy use, it seems that the net increase in
electricity use for precooling is small to at least up to the 90 degree range for outdoor temperature.
These results are tentative, however, and it must be emphasized that insulation, windows and other
aspects of the residence likely cause strong variation in the degree of extra energy needed for
precooling. Also, if precooling can be at least partly achieved through passive means, the energy
balance could become positive.

1034
Figure 2: Screenshot of “MyEnergy”: precooling schedule, max outdoor temperature 90
degrees F (32 C)

Overall, the pilot has shown that such a monitoring and control system satisfying the design
parameters is viable via mostly off-the-shelf parts. Non-PC costs for each house were approximately
$1000 using off the shelf components. As previously stated, the project goal was to build prototypes
with off-the shelf technology rather than to find components and technology at scale or minimum
costs. In reality, the costs at large-scale adoption could likely be in the $200 or less range (not
including PC). Given possible energy savings, this presents a significant business opportunity Thus
there is possibly a market for IT solutions that have the added benefit of managing household energy
consumption (and the associated environmental benefits). Intel and SMUD both have interest in
continuing the pilot for additional summer months, as well in adding more participating households. In
addition it is desired to add the capacity to monitor end-loads of major appliances as well as non-
electricity loads, such as natural gas consumption.

B. PC Energy Usage
Although not the focus of this study, the energy consumption of the PC itself should not be ignored. A
PC with limited or no power management can consume a considerable amount of power by itself. In
this study, it is not expected that the additional functionality of home control will significantly increase
electricity use by the PC. This is because the system used is compatible with the power standby
mode. Also, PC electricity use is already included in the measurement of the home’s overall energy
usage. With this said, future work should assess the life cycle energy footprint of an entire home
control system, including additional control and networking equipment. Also, as mentioned above,
Media Center edition PCs are already in wide use due to their media management capabilities. When
setting up a PC-based home control system, power consumption may be minimized by choosing an
energy efficient PC and by enabling power management (sleep, stand-by) of the device. The US
Energy Star program is in the process of updating its specifications for computers. In the future,
computers carrying the Energy Star label will be required to consume less power when on, asleep
and in stand-by power configurations.

Looking ahead
There is a growing trend toward integrating computers and related IT equipment into the home,
particularly for home entertainment purposes. As such, the opportunity exists to expand the features
of these products to include home management, including the optimization of heating, cooling and
lighting. Such integration offers the possibility of future benefits to the consumer in terms of
convenience, cost savings, and reductions in energy use at a reasonable incremental cost. As part of
ongoing R&D research in this area, Intel and other IT manufacturers are continuing to explore such
possibilities.

1035
References
[1] Annual Energy Outlook (2004), Energy Information Administration, US Department of Energy,
www.eia.doe.gov
[2] RECS (2001) Residential Energy Consumption Survey, Energy Information Administration: US
Department of Energy, http://www.eia.doe.gov/emeu/recs/contents.html
[3] Ueno, T., R. Inada, O. Saeki, K. Tsuji (2005) Effectiveness of displaying energy consumption
data in residential houses: Analysis of how the residents respond, ECEEE 2005 Summer Study
– What Works & Who Delivers, 1289-1299
[4] Projected Costs of Generating Electricity (1998) International Energy Agency, Paris
[5] Economic Input-output Life Cycle Assessment model (2006) Green Design Initiative, Carnegie
Mellon University, www.eiolca.net
[6] G. Henze (2005) Energy and Cost Minimal Control of Active and Passive Building , J. Solar
Energy Engineering Thermal Storage Inventory 127, 343-351
[7] Herter K., P. McAuliffe, and A. Rosenfeld (2005) Observed Temperature Effects on Hourly
Residential Electric Load Reduction in Response to an Experimental Critical Peak Pricing Tariff,
Lawrence Berkeley National Laboratory, Berkeley, California, LBNL-58956
[8] SPP End-Of-Summer Survey Report (2004), Momentum Market Intelligence, Portland Oregon

1036
How Small Devices are Having a Big Impact on U.S. Utility Bills
Andrew Fanara1, Robin Clark2, Rebecca Duff2, Mehernaz Polad2
1
U.S. EPA
2
ICF International

Abstract
Energy consumption attributed to electronic devices in the typical U.S. home has more than doubled
since 1980 and is expected to continue to grow at a rate nearly double the forecasted growth rate for
1
residential electricity end use. The breadth of these devices also grows continuously, driven by
technological innovation designed to meet surging consumer demand and changing lifestyles.
While the traditional sources behind this increasing energy consumption trend are office equipment
and consumer electronics, other miscellaneous devices, such as power tools, portable appliances,
and personal care products contribute as well.
The growth in electricity consumption within the typical home from miscellaneous end uses is in
addition to the rising costs of fuel, such as gasoline for automobiles and oil for home heating. These
costs are intensifying, straining consumer budgets while adding to the climate change burden. To
address these concerns, consumers, more than ever, need relevant information about the growing
array of miscellaneous products and their energy consumption in order to make smart buying
decisions.
As a result, new opportunities exist for ENERGY STAR to highlight existing electronics products that
are efficient across multiple modes of operation, including "active" or "on" mode, and to address non-
traditional miscellaneous products. As each opportunity to address multiple modes of operation and
product types presents itself, it brings with it the need to overcome a variety of technical challenges
while designing relevant policy options that will benefit consumers.

Introduction
The energy consumed by the typical U.S. home has more than doubled since 1980, according to the
U.S. Department of Energy’s (DOE) Energy Information Administration (EIA), and it is expected to
continue to increase. The growing quantity of electrical products found in homes contributes
significantly to this growth of energy consumption. While this trend is due in part to the proliferation of
home computer equipment and consumer electronics, other devices, such as power tools, portable
appliances, and personal care products, contribute as well. The individual and collective energy
consumed by miscellaneous devices are worthy of scrutiny as consumer appetites for these devices
drive their quantity, quality and usage patterns.
The growth in residential electricity consumption comes at a time when the United States is
experiencing rising prices for electricity, natural gas, and heating oil, used primarily in the densely
populated northeastern United States. Along with utility bills, car centric homeowners also feel the
pain from spiking gasoline prices. In the short run, many home owners find these times confusing and
frustrating, as they search for a way to lessen the bulging energy bills that have cut into discretionary
spending. Neil Elliott of the American Council for an Energy Efficiency Economy likens the situation to
being in an energy straitjacket. Few options for the nation as a whole exist in the short run, especially
for people who have awakened to find they are vulnerable to rising energy prices. Adding to their
anxiety is a growing awareness of the connection between the energy they use to support their
2
lifestyle and global climate change, over which they feel they have little control.
Sustainable pathways forward to break out of the straitjacket have been identified, but they will unfold
over many years. In the short run however, consumers can begin to reduce their energy bills by
3
understanding and managing the amount of energy consumed by miscellaneous devices. This paper
will provide data and background information to define and substantiate the emergence of

1
Per the Annual Energy Outlook 2005, the growth rate for residential miscellaneous electricity consumption is 3.7% for 2005 –
2010 while the growth rate for residential electricity consumption is 2%.
2
Despite rising gas prices, according to the U.S. Bureau of Economic Analysis, the total percentage of personal income spent
on gasoline today is 3%, as opposed to 5% in 1981.
3
In the California Independent System Operator area, for example, efficiency measures reduced energy consumption by 6.7%
from 2000 to 2001.

1037
miscellaneous products as a new major category of residential end-use energy consumption. The
miscellaneous category will be compared to other major end-uses in terms of their characteristics in
the typical residential setting. Lastly, the paper will identify available policy options which are being
pursued by governments looking to address the growing energy demand from miscellaneous
products.

Background
What are the “Miscellaneous” Products?
EIA has traditionally tracked five main sources of energy consumption in the residential sector:
• Lighting;
• Major appliances (white goods);
• Water heating;
• Air conditioning; and,
• Space heating.
Miscellaneous electricity consumption constitutes all the energy consumed that is not directly a result
of the use of the above sources. Examples of these miscellaneous products are extremely varied and
include televisions, computers, mobile phones, small home appliances, such as toasters, coffee
makers, baby monitors, and home security systems to name a few.

For the better part of the last 50 years, the utility bills for the typical U.S. home were dominated by the
energy consumed by key products representing the five major categories listed above. Collective
ownership of these products were hallmarks of growing economic prosperity, both at a national level
and for individual households. Against the backdrop of abundant and inexpensive energy,
homeowners traditionally gave little thought to the energy these products consumed. By the late
1970s, a growing environmental movement and higher energy prices drove several states to launch
the first appliance efficiency standards. These were followed by the U.S. government’s mandatory
standards in 1987, as legislated by the National Appliance Energy Conservation Act. These initial
standards took hold as a legitimate and cost effective means to ensure greater efficiency and paved
the way for new standards to be developed and implemented for a variety of products such as
furnaces and boilers, electric motors, lighting, pool heaters, and water heaters.

How Much Energy is Consumed by “Miscellaneous” Products?

Even as the energy attributed to miscellaneous products continued to grow, it was still outweighed in
absolute terms by the five main categories listed above. However, the trend is clear – a plethora of
new energy consuming devices now populate the typical home. The collective energy consumed by
this new category constitutes a new major source of end-use consumption, and individually, is larger
than several of the five major residential end-use categories.
Based on a 1995 EIA report, DOE estimates that in 1980, residential miscellaneous electricity
consumption totaled 2.25 quads (Buildings and Energy in the 1980’s, EIA). 18% of this consumption,
4
or 0.4 quads, was due to “other” or miscellaneous consumption. Electronics products, grouped into
the miscellaneous category, accounted for only about 5% of home electric consumption. Electronics
includes products such as consumer electronics and IT equipment, motor includes products such as
pool pumps, well pumps, and fans and heating includes products used to heat water-beds, hot-tubs,
pools, and other such products. (The breakdown of the miscellaneous category into electronics,
motor, and fan consumption was based on an analysis by Marla Sanchez, Lawrence Berkeley
National Lab, and John Cymbalsky, EIA.) The remaining 1.85 quads were attributed to the other five
main categories. Figures 1 and 2, below, illustrate how the 2.25 quads were divided among the
primary categories.
By 2005 total residential electricity consumption doubled to 4.5 quads, according to EIA. Of this,
approximately 28%, or 1.3 quads, could be attributed to ‘other’ or miscellaneous energy consumption.
The allocation of the remaining 3.4 quads was: appliances (20%); lighting (18%); air conditioning
(15%); space heating (11%); and, water heating (8%). Electronics products accounted for about 13%
of total home electric consumption; almost three times the level in 1980 (Building Energy in the 1980’s

4
Lawrence Berkeley National Lab’s estimate for miscellaneous consumption in 1980 (as a percentage of EIA’s ‘appliance’
category) is based on the 1980 consumption estimates for the heating, motor and electronics categories in Sanchez et al.,
1998.

1038
 5
and Annual Energy Outlook 2005, EIA). Figures 1 and 2, below, illustrate how the 4.5 quads were
divided among the primary categories.
By 2015, EIA projects residential electricity consumption to increase 20% from 2005 levels, to 5.4
quads. Lighting will still account for about 18% of the total, but space heating, water heating, air
conditioning and appliances will all consume smaller percentages of energy than they did in 2005. On
the other hand, the total miscellaneous category is projected to grow to 34%, or 1.8 quads.
Electronics products alone will account for 18% of total home electric consumption. Figures 1 and 2,
below, illustrate how the 5.4 quads are divided among the primary categories. Other studies, such as
a report prepared for DOE entitled U.S. Residential Information Technology Energy Consumption in
2005 and 2010 (TIAX LLC), have concluded that the major drivers of residential energy consumption
growth rates will be IT equipment, much of which has aggressive usage patterns.

5
Electricity Consumed (Quads)

Other
Space Heating
Water Heating
3
Air Conditioning
Lighting
Appliances

0
1980 2005 2015
Year

Figure 1: Total Residential Electricity Consumption for 1980, 2005 and 2015 (Projected)
Figure 1 Note: In 1980, EIA grouped lighting into the appliances category. Lighting has since been broken
out separately as its own category, as reflected in the data for 2005 and 2015. (Source: EIA’s Building and
Energy in the 1980’s, June 1995; Sanchez et al,1998 )

5
LBNL combined certain categories within EIA’s 2005 data so comparisons could be drawn to 1980 data: for space heating,
EIA’s space heating and furnace fans categories were combined; for appliances, EIA’s refrigerators, freezers, cooking, dryers,
clothes washers and dishwashers categories were combined; for other, EIA’s other, TV and PC categories were combined.

1039
 2

1.8

1.6

1.4
Electricity Consumed (Quads)

1.2

Heating
1 Motor
Electronics
0.8

0.6

0.4

0.2

0
1980 2005 2015
Year

Figure 2: Breakdown of ‘Other’ Residential Electricity Consumption in 1980, 2005 and 2015
(Projected)
Figure 2 Note: Heating in Figure 2 does not refer to space heating; rather, it reflects heating for water-beds, hot-
tubs, pools, and other such products. Motor in Figure 2 refers to fans and pumps. (Source: EIA’s Annual Energy
Outlook 2005)

Several studies, such as Whole-House Measurements of Standby Power Consumption (Ross, J.P.
and Meier, A.) and Developing and Testing Low Power Mode Measurement Methods (Nordman, B.
and McMahon, J.E.) have documented the breadth of miscellaneous products in the typical U.S.
home. Hundreds of products found in homes (and in some commercial buildings) that use electricity
have been identified. Many are commonly recognized such as televisions and coffee-makers, while
others are less obvious, such as air cleaners and garage door openers. Often, certain rooms have
their own unique devices, such as electric tooth brushes and shavers in bathrooms, coffee-makers
and toasters in kitchens, and power tools in basements.
It is unknown how some of the newer miscellaneous products entering the market will affect future
energy demands. For example, sales of mobile phones and game consoles are growing quite fast and
are quickly becoming modern essentials, but have yet to reach market saturation. According to The
NPD Group, mobile phone sales to U.S. consumers reached 34.8 million units in the first quarter of
2006; an increase of more than 11% compared to the same period in 2005. In spite of the growth in
sales, U.S. market penetration for mobile phones in mid-2005 was still only at about 65%; nowhere
6
near countries such as Germany and Switzerland, which were at 90% market penetration. The U.S.
still has a ways to go before reaching market saturation for these products. Alternatively, the market
for products such as dehumidifiers is mature and more regional in nature with significantly fewer units
being shipped. According to Appliance Magazine, an average of 972,685 units shipped annually
from 2000 – 2003 in the U.S. TVs on the other hand, have more than 90% penetration in U.S. homes
and shipments are once again growing briskly due to new technologies and features. In fact,
according to iSuppli’s TV Systems Market Tracker, shipments of TVs to the U.S. in 2005 increased
14% over 2004, going from 25.6 million units to 29.28 million units.
Not surprisingly, sales of consumer and IT electronics have the fastest rates of growth, due in part to
their rapid obsolescence. According to the Consumer Electronics Association (CEA), sales of
consumer electronics rose 11.5% from 2004 levels to $125.9 billion in 2005. For 2006, CEA expects
7.5% growth and sales of $135.4 billion. Many consumer electronics products such as televisions or
cordless telephones have been prevalent for decades, whereas others, such as cable modems, have

6
Data is from research conducted by Merrill Lynch in June 2005.

1040
only recently emerged in volume. Digital cameras, which are usually plugged in to an electricity
source to be recharged, have rapidly displaced their traditional single-lens-reflex (SLR) counterparts,
only to now be overtaken in sales by camera phones. iSuppli estimates more than 57 million camera
phones were sold in 2005, representing 46% of all handset sales. By 2007, InfoTrends predicts 109
million unit sales of camera phones, or 71% of all cell phone handset sales.
As compared to the other major common household products, miscellaneous products are more
diverse and populous. They are also typically designed to perform a unique and sometimes
infrequent task. Such a narrow usage pattern contrasts with general purpose lighting and cooling
products, for example, which can be used for thousands of hours per year.

Why Care About Miscellaneous Energy Consumption?

As the name implies, the source of miscellaneous energy consumption can be nebulous, and its
growth and impact on utility bills obscured, given the disproportionate attention paid to white goods,
lighting, and the other primary energy usage categories. In some cases, the small physical size of
certain miscellaneous devices obscures their importance as a group, making the tracking and
inventory of their energy consumption challenging. While this is slowly changing, EIA has wrestled
with the means to account for miscellaneous energy consumption in its Residential Energy
Consumption (RECs) survey, with the exception of computers.
Slowing the energy consumption growth of miscellaneous products may require a different mix of
actions from policy-makers and a persistent commitment to efficiency on the part of manufacturers.
Increases in the prices of all forms of energy have begun to raise consumer awareness. In some
cases, consumers do a fairly good job of not wasting energy because they understand the connection
between a product and the energy it consumes. For example, consumers have become conditioned
to flip the lights off when they leave a room. But it is probably unrealistic to expect that most home
occupants would unplug a multitude of devices, which they may think are off in the first place since
they are not “being used.” Nor is central control of these devices an option. While programmable
thermostats exist as a controlling device for products in the air conditioning and space heating
categories, they have not always been found to save energy; a lack of consumer-friendly
programmable thermostats has sometimes perversely caused more energy to be used than saved
(Haiah et al, 2004).
Quantifying the size and growth of miscellaneous energy consumption is necessary to understand the
impact on the home energy budget. This paper reveals that the energy used by miscellaneous
devices is comparable in magnitude to the energy used individually by the five major end-use product
categories. The growth of energy attributed to miscellaneous products has expanded the amount of
funds families must contribute to their home energy budget. These actions have obvious implications
for disposable income, which are well documented. On a positive note, the energy consumed by the
major end-use categories has not experienced as much growth due the presence of mandatory
standards, which are often coupled with consumer education, mandatory labeling and improved home
7
building codes.
Feature-rich products such as televisions and computers, which serve communication, entertainment
and work purposes, have appreciably improved the quality of life to such an extent that many
consumers view them as indispensable. Many U.S. consumers would no sooner sacrifice TV or high-
speed Internet access, than they would sacrifice hot water and refrigerated food. In fact, as
consumers cut back on non-essentials to conserve money, such as dining out and entertainment, they
are spending more time at home, which translates into more hours watching television, playing video-
games, or using the computer. Homebuilders recognize this trend. According to a 2004 builder
survey conducted in the U.S. by Parks Associates, approximately 45% of homebuilders offer built-in
home theatres as an option in new, single-family homes.
In 1980, the average U.S. household spent $1,280 on utility bills (adjusted for inflation and not
including water usage). Of this amount, approximately $230 was for the electricity to power
miscellaneous products. Today, the average household utility bill is approximately $1,500 per year
with $420 spent to power miscellaneous products. Add in the average annual household gasoline
expenditures of $2,327, and this raises the estimated home energy budget to $3,800. Miscellaneous
consumption would still represent 11% of this total.

7
It should be noted that the intent here is to raise awareness of the unique nature and characteristics of the miscellaneous
product category; not discount the importance of efficiency for the other major end-use products.

1041
Miscellaneous plug loads have also emerged in the commercial sector. DOE estimates that for
certain types of buildings, such as commercial office-buildings or schools, miscellaneous plug loads
are equal to half of lighting levels. Figure 3 shows a breakdown of the electricity consumption
attributed to the primary product categories in the U.S. commercial sector for 2005 and projected for
2015 (Annual Energy Outlook 2005, EIA).

6.00

5.00

Other Uses
Cooki ng
4.00
Venti l ation
Water Heati ng
Space Heating
3.00
Ref iger ati on
Of f ice Equi pment (PC)
Space Cool ing
2.00
Of f ice Equi pment (non-PC)
Lighti ng

1.00

0.00
2005 2015

Y ear

Figure 3: Percentage Breakdown of U.S. Building Sector Energy Consumption

Figure 3 is derived from EIA’s Annual Energy Outlook 2005.

It should be noted that the building types in the commercial sector are more diverse than the
residential sector. Universities, hospitals, industrial sites, etc have different types of functions that may
require different levels of miscellaneous plug loads. There are some common miscellaneous devices,
however, found in both the residential and commercial building types. But more study is required to
determine whether common policies can be used to address the miscellaneous plug loads found in
both sectors.
An analysis of projected overall growth of residential electricity consumption from 1980 to 2015
indicates that the miscellaneous category will contribute significantly to this growth. Conversely, the
major end-use categories (e.g., heating, cooling, etc.) are mature and there is a greater probability
that their energy consumption will be kept in check due in large part to mandatory minimum efficiency
8
standards. Products within the major end-use categories have also been through many design
cycles, and so their efficiency has increased over time as technological advancements are made. For
example, although the number of lighting fixtures per American home has increased along with the
size of the typical home, several technical advancements and market efforts are minimizing the
overall growth in energy use. The improved quality of compact fluorescent lamps in recent years has
increased consumer adoption of screw-in lamps and fixtures that use fluorescent technology. In
addition, residential fixture manufacturers are more accepting of the technology and most major
manufacturers are offering decorative fluorescent lines (Vrabel, 2006). Use of high quality fluorescent
lighting in a new home is an emerging trend due to efforts by EPA, electric utilities, and the adoption
of state energy codes in California, such as Title 24. Additionally, the average new home in the U.S.
is being built to the International Energy Conservation Code building code, which ensures greater
efficiency compared to homes from a decade ago. An improved building envelope will use energy for
heating and cooling more optimally.

8
These standards are even more important given that the average total square footage of the typical U.S. home expanded to
approximately 2,527 feet in 2001, from 2,278 feet in 1993, per DOE’s 2001 Residential Energy Consumption Survey.

1042
What is Being Done to Address the Energy Use of Miscellaneous Products?
Miscellaneous products, as with most products, have not ordinarily been designed to save energy as
a primary function. Where a product incorporates energy management, action has been needed by
the user to initiate this feature or it has been included as a safety precaution. In the case of
programmable thermostats, they are marketed primarily to provide comfort and convenience, and
secondarily to reduce energy cost. For most miscellaneous products, the options to control energy
consumption range from the non-existent to intelligent engineering designs for ‘smarter’ operation.
Some miscellaneous devices allow for passive or active means to manage energy use. Televisions,
for example, are easily put into standby mode via a remote control. Computers, on the other hand,
offer several means to implement power management. Other products, such as air cleaners or
dehumidifiers provide occupancy or other sensors to manage energy. Products utilizing lithium ion
battery chemistry have intelligence designed into the product circuitry to prevent over-charging, but
this is primarily a safety feature that has a side benefit of energy management.
Given the diversity of energy management features found in miscellaneous products, policymakers
face a challenge to find a common thread upon which to base uniform requirements to reduce energy
waste. In many cases, policies can be devised to encourage greater energy efficiency in these
miscellaneous products, but the greater challenge is ensuring that they remain relevant given the
changing mix of technology, consumer lifestyles, and the need to be cost effective.

Policy Pathways for Miscellaneous Products

As indicated earlier in the paper, standard making organizations have focused on more traditional and
high profile energy consuming products over the last few decades (e.g., lighting, water heating, etc.).
Products in these categories offer significant savings per unit and are relatively simple to characterize
with little variance from one model to the next in terms of key components. As a result, standards for
products in these areas have been embraced and implemented across the U.S. and globally.
Meanwhile, the miscellaneous category has been growing, in both numbers and energy consumption,
and slipping under the radar screen of energy modelers. These products have either been left out of
the energy consumption equation completely or not addressed in their entirety (i.e., only low power
mode requirements have been addressed). At the same time that consumers are being educated
about the range of efficiencies and opportunities presented by more traditional products, and thus
associating them with rising energy costs and high utility bills, we as consumers and policymakers are
missing a growing contributor to the energy used in the home.
Over the last decade, the ENERGY STAR voluntary labeling program has sought out these
miscellaneous products and taken a number of different approaches to developing specifications to
address this category. Efforts began in the early- to mid-1990s with the more visible electronics
products such as TVs and computers. These products represented large and widespread markets
that were ripe for the type of market differentiation that the ENERGY STAR mark provides. More
recently, however, EPA wanted to both expand the scope of products covered by these specifications
and address additional operational modes. To conserve limited resources and maximize the energy
savings potential, EPA identified three approaches, which are described in further detail below. The
first approach was to address the energy being consumed by numerous miscellaneous products
through a common thread: the power supply. A second approach, designed for those products where
it was not feasible to isolate or appropriate to focus solely on the power supply, was to develop
specific metrics to directly address Active mode for the entire product system. And finally, a third
approach was to work with international stakeholders to harmonize test procedures and specifications
as much as possible, given that a number of products in the miscellaneous category are globally
manufactured and sold.

Power Supplies: The Common Thread

Focusing on the power supply provides enormous energy savings potential because:
• They have broad application in finished electronic products. EPA estimates there are
more than 10 billion power supplies are in use worldwide.
• Many current power supply designs are only 30-60 percent efficient, but efficiencies of 90
percent or more are feasible.

1043
The following is a brief discussion of EPA’s three-pronged strategy to transform the power supply
market.

Phase I: External Power Supplies (EPSs)

The first step in EPA’s power supply strategy was to provide market-based incentives to improve the
efficiency of EPSs. EPSs represented the best near term opportunity because 1) they are physically
separate from their end-use devices and thus easier to measure, and 2) more research has been
devoted to understanding their current and potential efficiencies. By pairing an end-use device with a
highly efficient EPS, one can significantly reduce the energy consumption of the end-use device
without investing in costly or time consuming product redesigns.
To prepare an EPS specification, EPA supported the development of one internationally recognized
test procedure and then compiled a data set with over 600 EPS models tested in Australia, China, and
the United States. Figure 4, below, illustrates the wide range of Active Mode efficiencies found in
EPSs with a nameplate output power of 10 watts or less within EPA’s data set. For example, within
EPA’s data set, EPSs with a nameplate output power of 3 watts had average Active efficiencies of
31.9 percent to 67.8 percent and No-Load consumption of 1.9 to 0.2 watts.

90.0%

80.0%

70.0%

60.0%
Efficiency

50.0%

Greater Than 1 to 49 Watts

40.0%
Efficiency ≥ 0.09Ln(Nameplate Output Power) + 0.49

30.0%

20.0%

Average Active Efficiency Data

10.0%
1 Watt and Lower Tier 1 Active Specification
Efficiency ≥ 0.49 * Nameplate Output Power
0.0%
0 1 2 3 4 5 6 7 8 9 10
Nameplate Output Power (W)

Figure 4: Range of Active Mode Efficiencies in EPA’s Data Set

Figure 4 is derived from data collected during the development of EPA’s Version 1.0 ENERGY STAR
specification for external power supplies. A point above the line denotes an EPS model that exceeds minimum
efficiency levels.

Figure 5, below, illustrates the wide range of No Load efficiencies found in EPSs within EPA’s data
set. The final ENERGY STAR specification for EPSs, which took effect in January 2005, includes
requirements for both Active (tested at 100 percent, 75 percent, 50 percent, and 25 percent of rated
current output) and No-Load Modes and recognizes approximately the top 25 percent of models.

1044
 4.0

3.5

3.0
Measured No-Load Power (ac watts)

2.5

Measured No-Load Data

2.0 No load max power Tier 1
No load max power Tier 2

1.5

1.0

0.5

0.0
0 25 50 75 100 125 150 175

Nameplate Output Power (dc watts)

Figure 5: Range of No Load Efficiencies in EPA’s Data Set

Figure 5 is derived from data collected during the development of EPA’s Version 1.0 ENERGY STAR
specification for external power supplies.

The ENERGY STAR specification for EPSs is designed to highlight those models with an efficient ac-
dc or ac-ac conversion process. A broad array of single voltage external ac-dc and ac-ac power
supplies, including those used to power computer and consumer electronics such as laptops, digital
cameras, monitors, CD players, cell phones and cordless phones, are covered by this specification.
As of May 2006, 26 manufacturers have joined ENERGY STAR and qualified more than 255 models.
EPSs that earn the ENERGY STAR mark are on average 35 percent more efficient than conventional
models. EPA projects a potential U.S. energy bill savings of $636 million, electricity savings of over 9
billion kWh, and prevention of 5.13 million metric tons of carbon dioxide pollution from 2005 to 2015
due to the specification.
In addition to recognizing efficient EPSs, EPA is 1) requiring ENERGY STAR EPSs in its latest
electronic product specifications (e.g., ENERGY STAR for telephony, imaging equipment, etc.), and 2)
allowing end-use product manufacturers, whose products otherwise might not be eligible for ENERGY
STAR, to join the program and promote their end-use devices that use ENERGY STAR qualified
EPSs (e.g., digital cameras, wireless routers, etc.). In this way, EPA is not only striving to increase
the supply of efficient EPSs, but is also taking steps to stimulate demand for them.

Phase II: Battery Charging Systems (BCSs)

Soon after the launch of the EPS program, EPA began work on another specification to extend the
brand to BCSs—the components used to recharge a wide variety of cordless products. BCSs found in
small household appliances, personal care products, and power tools were specifically excluded
under the EPS specification in order to allow time for the development of an appropriate test
procedure and efficiency metric.
Based on a careful review of the energy used in more than 100 BCSs, EPA decided to focus on
Nonactive modes (i.e., Battery Maintenance and Standby) because these modes offer significant
potential for energy savings (70-75% of the energy is used after the battery is fully charged) and can
be consistently measured through a robust and easy-to-use test method (see “Test Method for
Determining the Energy Performance of Battery Charging Systems” on the ENERGY STAR Web site).
The ENERGY STAR specification for BCSs allows models to qualify if they do not exceed a maximum

1045
 9
Nonactive Energy Ratio based on the nominal battery voltage. By avoiding explicit requirements for
Battery Maintenance and Standby Modes, EPA allows manufacturers to choose the most efficient
design(s) for the overall operation of the product and takes into account products that do not have a
Standby Mode.
The ENERGY STAR specification for BCSs took effect in January 2006. Consistent with the ENERGY
STAR guiding principles, the final specification represented the top 24.8 percent of data points from
EPA’s data set. Care was taken to ensure that a wide array of models and manufacturers would be
eligible to qualify as ENERGY STAR under the BCS specification, as illustrated in Figure 6 below.

70
Non-Active Energy Ratio

60

50

40

30

20

10

0
0 5 10 15 20 25 30
Nominal Battery Voltage (Vb)

Power Tools Spec UBC Floor Care Personal Care Other

Figure 6: BCS Specification by Product Type

Figure 6 is derived from data collected during the development of EPA’s Version 1.0 ENERGY STAR
specification for battery charging systems.

In the United States alone, more efficient battery chargers have the potential to cut energy
consumption by more than 1 billion kWh per year, saving Americans more than $100 million annually
while preventing the release of more than 1 million tons of greenhouse gas emissions.
The BCS specification is intended to complement the EPS specification to comprehensively cover a
variety of different miscellaneous products and ultimately impact their power conversion efficiency and
standby power consumption. The last phase of the power supply strategy will allow EPA to reap even
greater environmental savings.

Phase III: Internal Power Supplies (IPSs)

Following successful development and launch of the EPS and BCS specifications, the U.S. EPA has
now shifted focus to the development of a test procedure for IPSs. Given the complex nature of these
devices, however, the U.S. EPA recognizes that the same efficiency levels will not work for all IPSs
and their end-use products. The U.S. EPA has therefore decided to follow a product-specific
approach with IPSs, based on the end-use products for which they are designed. The first end-use
product specification that will incorporate an IPS requirement is the new ENERGY STAR specification
for computers.

Since its inception in 1994, the ENERGY STAR computer specification has sought to bring energy
savings to consumers through low power mode requirements. However, with enabling rates reaching
9
Nonactive Energy Ratio (ER) is the ratio of the accumulated nonactive energy (Ea) divided by the battery energy (Eb).
Accumulated Nonactive Energy (Ea) is the energy, in watt-hours (Wh), consumed by the battery charger in battery maintenance
and standby modes of operation over a defined period. Battery Energy (Eb) is the energy, in watt-hours (Wh), deliverable by
the battery under known discharge conditions.

1046
only 11% and computers spending most of their time outside of these low power states, EPA realized
that to guarantee significant savings, energy consumed during active mode would need to be
addressed. One way to reduce active power is through the internal power supply.
While beginning revisions to the ENERGY STAR computer specification in 2004, EPA was also
supporting another initiative created to help develop a market for more efficient internal power
supplies. The utility-funded 80 PLUS program offers incentives to computer manufacturers that
incorporate 80%+ efficient internal power supplies into their computer and server designs. The
program also provides power supply manufacturers a vehicle to certify and market their more efficient
power supplies to interested buyers. To date, 11 power supply manufacturers, three of which hold a
significant share of the market, have certified internal power supply models through the 80 PLUS
program.
With a variety of models and brands now available in the marketplace, EPA felt confident that
including an internal power supply requirement under the new ENERGY STAR computer specification
was both technically feasible and cost effective for manufacturers to incorporate. Therefore, in
addition to including more challenging low power mode and power management requirements, EPA is
requiring a minimum 80% internal power supply efficiency for ENERGY STAR qualification.
Replacing a standard efficiency internal power supply (70% efficient) with an 80% efficient model
could save 50 kWh/year for each computer and more than 200 kWh over the expected four-year
lifetime of the computer.
In addition to minimum power supply efficiency, EPA is also including a Power Factor (PF) of 0.9.
While PF may not contribute to the efficiency of the power supply, studies show that it does help to
reduce the overall plug load of the building by minimizing harmonic distortion. Use of high efficiency
power supplies will result in direct reduction of energy use, due to lower power consumption. Power
factor correction adds another 12 to 21% to the resulting energy savings, based on the cable lengths
typically found in residential and commercial buildings (40 feet and 100 feet, respectively). Many
power supply manufacturers are designing PF into their high efficiency models already. The new
ENERGY STAR computer specification goes into effect on July 1, 2007, which includes the 80%
efficiency and 0.9 PF internal power supply requirement.

Other Approaches to Addressing Active Mode

Another approach that EPA is increasingly adopting to address the power consumption of products in
the miscellaneous category, particularly when revising existing product specifications, is to set new
criteria that focus on reducing the Active Mode power consumption of these products. This approach
is more complex than EPA’s power supply strategy (and may take more time to design and
implement), but allows manufacturers more design flexibility when meeting the energy efficiency
criteria.

Computer Monitors
The new ENERGY STAR computer monitor specification, which took effect on January 1, 2005, was
EPA’s first foray into the development and implementation of an Active Mode specification for a
product within the miscellaneous category. In the early part of this decade, market penetration of
ENERGY STAR qualified computer monitors exceeded 90% and there were limited additional energy
savings to be gained by merely lowering the Sleep Mode levels specified for ENERGY STAR
qualification. EPA had to pursue an Active Mode specification for these products in order to both gain
additional carbon savings and ensure that the ENERGY STAR continued to be a differentiator in the
marketplace for the computer monitor product category.
A primary reason why Active Mode had not been addressed for this category under previous
ENERGY STAR specifications was because no test procedure existed to measure the active mode
power consumption of computer monitors, irrespective of display technology. As such, EPA worked
closely with interested stakeholders to develop a sound test procedure to measure the power
consumption of computer monitors in Active, Sleep and Off Modes. Once finalized, manufacturers
were requested to test their latest models using this test procedure. EPA received data on
approximately 270 monitor models, all of which were available or being introduced to market as the
specification was being developed, and only these data points were used when setting the Active,
Sleep, and Off Mode levels for the new ENERGY STAR specification. Due to the new specification
alone, EPA projects a potential U.S. energy bill savings of almost $590 million, electricity savings of 8
billion kWh, and 1.2 million metric tons of carbon (MMTC) avoided over the time period of 2005 to
2015.

1047
Imaging Equipment
A similar approach was used when developing the new ENERGY STAR specification for imaging
equipment, scheduled to take effect on April 1, 2007. When EPA initiated the specification revision
cycle for imaging equipment in 2003, the specifications had been in effect for up to seven years and
ENERGY STAR qualified printers, copiers, and fax machines accounted for 92 to 99 percent of units
sold in 2000 (Gartner 2001). The high market penetration levels alone suggested that a review of
ENERGY STAR performance specifications was warranted. Further, as imaging equipment products
increased in speed and functionality, Active Mode contributed to a greater portion of total product
energy use.
When EPA began to revise the imaging equipment specifications, the need to address Active mode
for some products quickly became apparent. This was accomplished with the “typical electricity
consumption” (TEC) approach that considers the electricity consumed by imaging equipment during
its entire duty cycle. This method for assessing product energy efficiency was received favorably by
many stakeholders and at the time, demonstrated a forward-thinking approach to the development of
an energy-efficiency specification. The revised imaging equipment specification will save U.S.
consumers more than $3 billion over the next five years and avoid greenhouse gas emissions
equivalent to taking more than four million cars off the road.

Televisions
EPA is following a similar approach in revising the current ENERGY STAR specification for
televisions. When first developed, the ENERGY STAR specification for televisions focused on
Standby Mode due to the vast amount of time these products spent in standby mode and the millions
of televisions in use in U.S. homes. In 1998 when the specification was introduced, EPA estimated
that ENERGY STAR qualified televisions would use about 20 percent less energy in a year than
comparable televisions. However, Active power consumption is now becoming increasingly important
due to changes in product technology and usage patterns that result in increased energy
consumption. Such changes include:
• The advent of new display technologies, some of which may use significantly more
energy than their traditional counterparts;
• The trend towards larger screen sizes;
• The marketing of televisions as part of “home theater packages,” which may be used in
conjunction with a variety of audio and video devices, increasing overall system energy
consumption;
• The burgeoning availability of new cable and satellite programming content, leading to
increased television viewing; and,
• The growth in sales of game consoles, meaning there is an increase in the number of
hours a typical television operates each day.
Today, TVs account for about four percent of annual residential electricity use in the U.S. – enough to
power all of the homes in New York State for an entire year (Natural Resources Defense Council,
February 2006).
EPA is working with stakeholders and a number of key governments, including Australia and the
European Union, to develop a single, harmonized global test procedure for televisions. This test
procedure, when completed, will be used by each of the government entities to implement policies to
encourage the sale of more energy efficient televisions, thereby giving consumers a means to factor
power consumption into the purchasing decision.

International Harmonization of Test Procedures and Specifications

As various governments and other entities develop standards for these miscellaneous products,
harmonization will be a key element to successful implementation. Manufacturers are designing
these products for global distribution and are looking for ways to streamline their certification
processes. Through harmonization, countries can combine their resources and technical knowledge
to create test procedures and specifications that will be applicable worldwide. For countries that have
yet to begin addressing these product types, global test procedures and specifications will provide for
ease of adoption and implementation. As an added benefit, countries will have access to a large data
set already developed and maintained by those governments that have already made the decision to
align their test procedures and specifications. Policymakers representing the European Union,
Canada, Australia, Japan, Taiwan, New Zealand and China are considering the inclusion of
miscellaneous products such as computers, TVs, EPSs, and BCSs in their voluntary and/or regulatory
standards and working very closely with EPA to harmonize their respective approaches. Through

1048
initiatives such as the Asia Pacific Partnership, the U.S. is also working with fast-growing markets
India and South Korea. This brings the total population that could benefit from harmonization of test
procedures and specifications to well over 3 billion.
An important step taken by the U.S. EPA and its country partners towards global harmonization of test
procedures is that products are now required to be tested to the appropriate conditions of intended
markets where the products are to be sold as ENERGY STAR qualified. EPA has implemented this
policy in specifications moving forward, particularly for products in the home electronics and office
equipment categories, because energy consumption values may vary according to the input
voltage/frequency combination. In some instances, particularly for those product-categories that have
a 1-watt Off Mode specification, the variance caused by testing at market-specific input
voltage/frequency combinations is enough to qualify products in certain markets and not in others. As
the ENERGY STAR program develops an increasingly international scope, EPA has determined that
it is important to confirm that products meet specifications at the representative market conditions
where the products are sold. This ensures that consumer expectations are satisfied when they
purchase products that carry the ENERGY STAR mark.

Conclusions
As illustrated throughout the paper, the energy consumed by miscellaneous products continues to
grow. Through their ongoing work to understand and control this energy consumption, U.S. EPA has
learned some key lessons. Most importantly, it is possible to develop and implement measures to
mitigate the amount of energy consumed by the miscellaneous products category, as many of the
products have common elements such as power supplies. However, often a combination of
measures works most effectively, such as both voluntary and mandatory standards, as well as Market
Transformation tools such as consumer education campaigns, government procurement, rebates,
and/or tax incentives. These measures aid in raising awareness of the energy consumed by
miscellaneous products, allowing increased information to be shared with consumers and
empowering them to make more informed purchasing decisions. There is still a lot of additional work
that needs to be done to continue addressing the power consumption of miscellaneous products,
especially in terms of increased data collection to better understand this product category. However,
due to the global nature of many of the products in this category, international cooperation is vital as it
brings together the critical mass needed to effect change.

References
[1] “Another Record Year for CE Sales in 2005; ’06 Forecast Bright.” This Week in Consumer
Electronics online edition at www.twice.com, January 30, 2006.
[2] California Energy Commission, Developing and Testing Low Power Mode Measurement
Methods. September 2004. Report No. P500-04-057.
[3] California Energy Commission, Energy Efficiency and Conservation: Trends and Policy
Issues. May 2003. Report No. 100-03-008F.
[4] Energy Information Administration (EIA), Annual Energy Outlook 2005. January 2005.
DOE/EIA Report No. 0383(2005). Available for download at:
http://www.eia.doe.gov/oiaf/aeo/.
[5] Energy Information Administration (EIA), Buildings and Energy in the 1980’s. June 1995.
DOE/EIA Report No. 0555(95)/1.
[6] ENERGY STAR Product Development Web site: www.energystar.gov/productdevelopment.
[7] Fortenberry, Brian. Epri Solutions. Koomey, Jonathan G. Lawrence Berkeley National Lab,
Assessment of the Impacts of Power Factor Correction in Computer Power Supplies on
Commercial Building Line Losses.
[8] Haiah, Peterson, Reeves, and Hirsch. November 2004. Programmable Thermostats
Installed into Residential Buildings, Predicting Energy Saving Using Occupant Behavior
Simulation.
[9] iSuppli, “LCD TV Shipments Soar – Will They Bring Early Demise of CRT TVs,” – Television
Systems Market Tracker, Q4 2005.

[10] Marla Sanchez, Lawrence Berkeley National Lab, telephone conversation, May 18, 2006.
[11] Natural Resources Defense Council (NRDC), Tuning in to Energy Efficiency: Prospects for
Saving Energy in Televisions, February 2006.

1049
[12] Nordman, Bruce, Meier, Alan, Piette, Mary Ann. PC and Monitor Night Status: Power
Management Enabling and Manual Turn-off. Published in Proceedings of the ACEEE 2000
Summer Study on Energy Efficiency in Buildings. August. Lawrence Berkeley National Lab,
Berkeley CA.
[13] Parks Associates Web site:
http://www.parksassociates.com/research/reports/tocs/2005/builder_survey.htm
[14] Paul Vrabel, ICF International, telephone conversation, May 15, 2006.
[15] Ross, J. P. UC Berkeley. Meier, Alan. Lawrence Berkeley National Lab, Berkeley CA.
Whole-House Measurements of Standby Power Consumption. Published in Proceedings of
The Second International Conference on Energy Efficiency in Household Appliances, Naples
(Italy), September 2000.
[16] Sanchez, Marla C., Koomey, Jonathan G., Moezzi, Mithra M., Meier, Alan, Huber, Wolfgang.
1998. Miscellaneous electricity use in the U.S. homes. Published in Proceedings of the
ACEEE 1998 Summer Study on Energy Efficiency in Buildings. August. Lawrence Berkeley
National Lab, Berkeley CA.
[17] Scoblete, Greg. “Cam Phones May Boost DSC Sales.” This Week in Consumer Electronics
online edition at www.twice.com, February 13, 2006.

1050
 IT Equipment

1051
1052
Environmental Impacts of Computers in Belgium
Karine THOLLIER

ICEDD asbl

Abstract:
In the framework of a study focussing on four product categories (namely housing, packaging, cars
and computers), we modeled their annual impacts in terms of GHG emissions, raw material use and
waste production over their life cycle for the years 1990 – 2010. The results showed that though the
impacts of computers in households are small, they are increasing at an amazing pace (+ 150% GHG
emissions in 2010 compared to 1990). Moreover, about half of the impacts occur in the production
phase, due to the manufacturing of chips.
Different strategies could be envisaged in order to decrease these impacts, in the framework of a
product policy. Concerning computers, 5 strategies were identified: stabilising the demand,
substituting products, increasing the lifespan of computers, improving consumers’ behavior and
increasing recycling. The strategy that rises as the most important one is increasing the lifespan of
computers, which of course decreases the impacts due to manufacturing. Targeting the demand
could also lead to significant improvement.
In a last step, key stakeholders were consulted in order to gain insight on which practical measures
could be taken in Belgium. Increasing the lifespan of computers, which appeared as the top option,
still has a negative image in the industry, while there are already some initiatives of repairing for reuse
in Belgium by the social economy. Targeting the demand is foreseen as extremely difficult by all
stakeholders, in particular due to the association between computer ownership and social status.
Increasing services is seen as the best possibility.

Introduction
Attention is shifting more towards examining the potential of policies oriented at improving products
and consumption patterns. This, because the sector-based, mostly ‘process’-oriented environmental
policies seem insufficient to reach the objectives of sustainable development. The perspective of
“Integrated” in Integrated Product Policy (IPP) is the consideration of the entire product’s life cycle
chain and the consideration of multiple environmental problems to avoid adverse effects and shifting
to other impact types.
Climate change, resources use and waste management are major challenges for policy makers and
society in general. These three environmental issues constitute the core issues of the project
"Integrating climate, waste and resource policies through a product policy" that started in 2002 and
ended in 2004. This project was carried out within the scientific support plan for a sustainable
development policy financed by the Belgian Science Policy.
After a first step aimed at the identification of key products for a product policy that would intend to
reduce all three impacts simultaneously, the second step of the project studied in detail four product
categories (cars, household packaging, housing and computers & paper), with the view to evaluating
theoretical potentials of reduction of greenhouse gas emissions, energy use, waste production and
material resource use. The last phase of the project aimed at contributing to establish a framework for
the evaluation of product policy measures by stakeholders.
For the study of computers, we covered both computers in households and computers in federal
administrations (as an example of an office activity). In this paper, we present the results for
computers in households only, since the methodology and scope of these two parts were very
different.

1053
Methodology
Determination of theoretical impact reduction potentials
Elaboration of an analytical tool
The implementation of product policy requires an appropriate approach to take into account both the
different life cycle stages of products with their own environmental impacts, and the different
strategies with their ability to tackle one or several life cycle stages. In the framework of this study, we
have developed and used an analytical tool that aims at linking product life cycle stages with the
emissions, emission reduction potentials and relevant policies and measures. It summarises how
different improvement strategies of products can decrease their greenhouse gases emissions, waste
production and resource use at different life cycle stages, thus making computers more efficient
during their production and use. It resulted in a series of strategies that can have an impact on one or
more life cycle stage; for each of these strategies, independent parameters are varied in order to
reflect the potential effect of targeted policies and measures on the product characteristics.
For computers, the strategies envisaged were as follows:
Strategy 1 : Changes in final demand for the considered function: This strategy aims at changing the
demand of consumers for a particular. It would, of course, influence all life cycle stages, but is not
often envisaged in environmental programmes, as it influences consumption and therefore production
of goods and is often seen as damaging to the economy and the quality of life.
Strategy 2 : Substitution by products fulfilling the same function: The final function can be fulfilled by
different products. The design of the product (shape, material composition) can influence its life cycle
impacts. For those products for which the impacts from the use phase are important, the shift from
lower to higher energy efficiency or from higher to lower carbon-content fuel can also contribute to
reducing the life cycle impacts. In the case of computers, we considered desktop PC with CRT and
LCD screens, as well as laptops. Shifting from CRT to LCD , or from desktop to laptop, are two
possibilities to fulfil this strategy.
Strategy 3 : Product reuse: Product reuse allows doubling (or more) the product’s use phase, thus
delaying the time when it will become a final waste. Therefore it mainly influences the waste phase
and the production phase (because less new products are consumed), but also the use phase of
products if their impacts vary over the time. Product reuse is mostly not considered in actual climate
policies, but more in waste policies.
Strategy 5 : Rational use of the product: Properly using a product has a significant influence on its
impacts during the use phase and can also extend the product life span. It is only partially considered
in the current environmental policies, mainly through information and education measures. However,
information is increasingly considered as insufficient to change consumers’ behaviours [1]. For
example, enabling power management is more likely to be successful if this is a compulsory setting
on the equipment than if users are informed and encouraged to enable it. Measures envisaged in this
strategy should thus focus also on technology, and not rely on users to voluntarily change their
behaviour.
Strategy 7 : Increasing end-of-life products recycling: The increase of product recycling has a mostly
beneficial influence on energy used/generated in the waste phase, but also on the material production
and product manufacturing through the use of recycled materials or parts. In general, recycled
materials need less energy for transformation. Product recycling is not envisaged as such in climate
policies, but is mostly implemented through waste policies.
These strategies are product-oriented and do not target the production phase of computers (except
the design phase). The improvement of the production phase was included in the business as usual
scenario. Scenarios 4 and 6 (namely decreasing the product lifespan and change of composition of
the product) were nor relevant in the framework I-of the study of computers. Another strategy had to
be left out, though interesting: the decrease of transportation distances. Indeed , in the case of
computers, parts are manufactured world-wide and assembled somewhere else, leading to probably
important impacts in terms of fuel consumption from transportation. However this production and
assembly network is so large and unknown that we could not envisage to quantify this strategy.

Impacts from product consumption

A product-oriented policy would potentially aim at curbing not only impacts from products bought each
year, but also impacts from products already used in the country (existing stock).

1054
This means that both the existing stock of products and the products put on the market had to be
considered in the study, especially in the development of scenarios on impacts from products,
including impacts produced in Belgium but also impacts produced abroad.
As the study aimed at evaluating to what extent a product policy would contribute to the simultaneous
fulfilment of the three objectives to reduce GHG emissions, waste and pressure on raw materials,
impacts and impact reductions have been calculated in such a way as to fit with the following
environmental objectives:
• regarding greenhouse gas emissions, the emission reduction targets agreed in the Kyoto Protocol
are expressed as a percentage reduction of the average annual national emissions over the period
2008-2012 compared to the 1990 level.
• regarding waste, several types of objectives are defined in the regional waste policy documents,
based on per capita annual emission ceiling and minimum recycling or reuse rates in a year.
As a result, for GHG emissions and for waste, annual emissions need to be estimated and especially
"domestic" emissions.
Regarding material resources, there is no quantified objective.
Based on this approach, we calculated the three impacts for each phase:
• GHG emissions, in kg CO2 equivalent
• raw material use. This impact category indicator is related to extraction of scarce minerals and
fossil fuels. The Abiotic (= non-organic) Depletion Factor (ADF) is determined for each extraction of
minerals and fossil fuels based on the remaining reserves and rate of extraction. It is based on using
the equation, Production/(Ultimate Reserve)2 and comparing this to the result for Antimony (Sb),
which is used as the reference case. The reference unit for abiotic depletion is therefore kg Sb
equivalent.
• waste, in kg of final waste, i.e. waste after treatment of the disposed product, thus going to landfill.

Calculation of annual impacts and scenarios of impacts from products

For all four product groups studied in the project, we have developed a common methodology for the
construction of alternative scenarios. The development of this methodology has been guided by two
main concerns, which are to:
• reflect the annual environmental impacts from existing and new products used in Belgium from
1990 through 2010, with a clear distinction between impacts from the disposal phase of existing
products, the use phase of existing products and the production phase of new products;
• quantify the individual impacts from each general improvement strategy as well as the impacts
from the combination of all these general improvement strategies.

It starts with the building of a “business as usual” scenario (BAU) that reflects an evolutionary case
where the current policies do not change in the future. The BAU scenario is based on the different
data sets and on several assumptions regarding future trends. Alternative scenarios were then built
by changing key parameters, following each of the strategies mentioned above.

Stakeholder consultation on possible policies and measures

Once the theoretical improvement potentials were determined, we envisioned how a product policy
could help– at least partly – to achieve them. In order to identify measures, the expertise of
stakeholders was required. Indeed, existing policies and measures in Belgium are not evaluated after
completion (ex-post evaluation) nor beforehand (ex ante evaluation). This gives us very few
knowledge on the consequences of these measures.
We reviewed a series of consultation methods based on the following pragmatic criteria:
• Usability for all stakeholders
• Equity between stakeholders
• Amount of time required from stakeholders
• Possibility to express their views
• Quantitative exploitability of the results

We also evaluated cost and total time requirement for the research team.

1055
Following the evaluation of existing methods, we chose to develop a method based on the Delphi
principles [2], though not aiming at reaching consensus but rather at mapping opinions. This approach
is sometimes referred to as “policy Delphi”[3].
Using quantitative notation ranges allows the determination of mean values for each question, as well
as the statistical distribution. The Delphi method has been used for a long time. and allows structured
answers and removes traditional bias occurring in interactive settings. Consequently, it is a good
method to obtain stakeholders’ views while obtaining quantitative inputs to the model.

First consultation round

A Delphi questionnaire was supplied to the stakeholders willing to cooperate that asked them to
evaluate possible measures on a series of criteria:
• Technical feasibility: is technology / knowledge adequately available to implement this measure?
• Political feasibility: is there political willingness (or no political impediments at least) to set up,
implement and monitor this measure?
• Effectiveness: can this measure result in a substantial reduction of at least one environmental
impact (GHG emissions, resource use, waste production)?
• Acceptance by the stakeholder: would the measure be well accepted by the stakeholder?
• Market acceptance: will the affected market react positively to this measure?
• Cost for industry: will this measure substantially increase the costs in the industry concerned?
• Cost for policy: will this measure require the allocation of a substantial amount of money from the
policy level for its implementation and monitoring?
• Cost for users: will this measure substantially increase the total cost of ownership (purchase price
+ run price) for the user?

The questionnaires were filled in through the Internet. In both rounds, we had to be very careful about
the phrasing of questions and of measures. Stakeholders had the opportunity to enter new measures,
as well as to rephrase the measures. We also differentiated the criteria to evaluate depending on the
stakeholder in order to require less time from them.

Findings from round one : reorientation of round two

We had several findings from the first round on the general methodology for this consultation.
First, although we limited the number of criteria and the number of measures to limit the time
requirement from stakeholders, we had several negative reactions to this approach. Indeed, most
stakeholders would have preferred to have all measures and all objectives, and select themselves the
ones for which they feel they have enough expertise.
Second, the number of criteria was found to be too large to enable an easy assessment; moreover,
some criteria were found either to be redundant or to have too broad a definition that would have
required to split them up into other criteria. For the second round, we thus reduced the criteria to the
following ones: acceptance by users (user perception and user costs); acceptance by industry (costs
for industry and social costs), and efficiency.
Last, but not least, we found that the comments given by each stakeholder were more useful than the
actual figures of evaluation of the different criteria in order to evaluate measures. However, the results
altogether did not enable a common ground required to carry out the identification of priority
measures.
Based on these findings, we decided to reorient the second round to deepen the comments from all
stakeholders, with the aim to present the variety of opinions of the different stakeholders on the
subject of product-oriented policies and measures.

Hypothesis for the modeling of computers

For household computers, the distinction was made between laptops, desktops with CRT monitors
and desktops with LCD monitors. The data sources and main hypotheses used for these different
products follow.

1056
BAU data
Sales and market shares
We have few data on market shares of these different products. However, we can estimate for 2001
that laptops market share is ca. 20%, desktop with CRT screens, 65% and desktop with LCD screens
15%. These assumptions were used in the DSM study by Fraunhofer Institute [4]. In 2005 we had
data from GfK on market shares of laptops [5] (46% in 2005); although we would tend to think that
they are more likely to be sold to professional consumers. The rapid decrease of prices encourages
the buying of laptops (but probably not always as a replacement of a desktop) and LCD screens. We
assumed that after 2005, the sales of computers in Belgium will continue to increase (after the
stabilisation of sales for 2003 – 2004), as it seems to be the case already for 2005.
It is very difficult to get reliable data on computer stock in households before that date. Actually,
computers were only integrated in the INS annual survey on household consumption in 1995. For that
year, the stock of computer is 1,28 million units. The share of the different products is not known, but
it is not likely that laptops were often used by households in 1995, and LCD screens for computers
were not widespread either. Therefore it will be assumed that in 1995, desktops with CRT screens
accounted for 100% of the sales. The stock of computers in 1990 is unknown ; we assumed that the
ownership rate was at best 20% (thus 785 722 units).
Production phase
For the composition data, we based our calculations on EPA [6] and Handy and Harman Electronic
Materials Corp [7].
Data on impacts from production of materials were extracted from SIMAPRO. However, for
computers, the production and assembly stages of chips are of high relevance [8], we thus used data
provided by E. Williams.
Use phase
Knowing the power draw of computers is more difficult than it seems. Indeed, on central units, the
wattage found in the technical specifications represents the maximum power draw, but by no means
the average power draw. The maximum power is reached when all electricity consuming parts of the
central units (CPU, hard disk, fan, drivers, etc) run at the same time, which is almost never the case.
Therefore, it is necessary to measure the average power draw in use. Variations in power draw
between different models also complicates developing an average number for the installed base. It is
also more difficult to predict the future power consumption, even if it is assumed that due to higher
computing speed and the increasing number of peripherals (CD writer, , etc), this average
consumption will increase slightly. The growth in graphic card power draw also contributes, to a large
extent, to this increase of average power draw..
For displays, the task is easier since technical specifications mention an average power consumption
while in use, although these specifications may not be precise (e.g. depending on the contrast, colour
level of the screen etc). In the future, due to the increased size of screens, the energy consumption is
likely to increase slightly Although CRT-based monitors’ active mode power draw tends to increase
moderately as screen size grows, the active mode power draw of LCD monitors increases rapidly with
increased screen size. Eichhammer et al [4] estimated some figures from 2000 onwards that we used
for our modelling and extrapolated back to 1990. This source of data was also used to derive values
on the on and stand-by hours.
Disposal phase
For 2000, OVAM (the competent body for waste in the Flemish region) reported a selective recovery
rate of 95 % in 2001. In this value the computers sent abroad are also included (39%). These
computers are sent abroad mainly for charity purposes, but may also contain illegal electronic waste.
They will be taken into account as other end of life computers afterwards, but it should be noted that
the vast majority of these computers, after a possible second use in developing countries, are likely to
end up in landfills under very bad conditions.
Scenario assumptions
- Scenario 1 : change in final demand for the function: In this scenario, we assume that the
penetration rate of computers in households remain constant after 2005.
- Scenario 2 substitution of products fulfilling the same demand: In this scenario, we assume
an increased market share of LCD screens (to 35% in 2010) and laptops (to 65% in 2010) and a
decrease of CRT screens down to 0% in 2010.
- Scenario 3 product reuse: Computers can be re-used easily as long as they meet the needs of
the customer and as long as they are not too damaged when they are disposed of. Indeed, most
computers are replaced because they are outdated, not because they are not working. In these
conditions, they can fulfil a second use. We assumed a second life span of the same length as
the first one, and these computers are assumed to replace new ones (this is optimistic). Scenario

1057
 5 rational use of the product: In the BAU scenario, both the time of usage and the time in
standby mode are increasing. For this scenario we assumed that they remain constant after 2005,
thus 370 hours in on-mode, 12250 hours in stand-by mode.
- Scenario 7 increasing end-of-life recycling: We assume 100% selective collection rate and
100% recycling for all materials after 2005.
- Scenario synthesis : this scenario combines the different assumptions made for each scenario.

Results of modeling
In this section, we present the results of the life cycle modelling of computers including monitors in
households for the 3 impacts (GHG emissions, raw material use, and waste production) for the period
1990 – 2010.

Greenhouse gas emissions

Greenhouse gas emissions mainly cover CO2 emissions due to energy use during the whole life
cycle. Although semi-conductor manufacturing uses PFCs, HFs and SF6 as plasma etchers, these
emissions are still not well quantified and difficult to link to functional units of computers. GHG
emissions presented here include only CO2 emissions and should thus be seen as minimum values.

CO2 emissions

300

250
BAU
kton CO2 / yr

200 scenario 1
scenario 2
150 scenario 3
scenario 5
100 scenario 7
scenario synthesis
50

0
1990 1995 2000 2005 2010

Figure 1: GHG emissions from computers – 1990 to 2010

Figure 1 shows that GHG emissions are expected to level-off for the period 2005 – 2010. Indeed,
even though computer sales are increasing again, the rapid shift towards LCD screens and laptops
should compensate for this.
CO2 emissions come in almost equal proportions from the production phase and the use phase. But
only the use phase occurs in Belgium because most of the production phase is abroad. Moreover,
only half of the emissions from electricity take place in Belgium, since the other half is due to fuel
1
extraction . Thus, only 25% of the CO2 emissions shown above take place in Belgium.
Figure 2 clearly shows the equal levels of impacts for production and use.

1
Source : SIMAPRO database, fuel extraction

1058
 CO2 emissions in 2010

250

200
kton CO2

150
disposal phase
use phase
production phase
100

50

0
BAU scenario 1 scenario 2 scenario 3 scenario 5 scenario 7 scenario
synthesis

Figure 2: CO2 emissions due to consumption of household computers in Belgium– 2010 :

impacts per life cycle stage and per strategy

As can be seen from the figures above, the different strategies could decrease GHG emissions in a
very significant way: 64% reduction from the BAU scenario in 2010 for the combination of the different
strategies. The main strategies that would be effective are: substituting products i.e. increasing sales
of LCD screens and laptops (scenario 2), product reuse (scenario 3) and targeting the demand
(scenario 1).
These strategies target different life cycle stages. Increasing the trend towards LCD screens and
laptops would thus result in decreased GHG emissions for both the production and the use phase.
2
Indeed, in our model , manufacturing laptops and LCD screens require less material, because they
are lighter . This also applies to scenario 1.
In contrast, scenario 3 (reuse) only results in a decrease of GHG emissions in the production phase.
These correspond to avoided GHG emissions of new computers that would otherwise have been
manufactured.
Raw materials
Raw materials used by computers come from two main sources: materials used to manufacture the
computer; and energy used both to manufacture and to use the computer. Indeed, about 95% of
these impacts come from electricity consumption in the use phase. This is due to the electricity mix of
Belgium, which uses about 65% of nuclear energy in its electricity production. Although it does not
result in high GHG emissions, it still uses a large amount of raw materials due to the extraction of
fuels (including uranium).
Therefore, raw material use is expected to level-off until 2010, showing a similar curve as GHG
emissions (also linked to energy).

2
However, other studies indicate that this hypothesis could be flawed, laptops being much more complicated to manufacture
than desktops [9]

1059
 Raw material use

3,0

2,5
kton eq SB / yr

BAU
2,0 scenario 1
scenario 3
1,5 scenario 5
scenario 7
1,0
scenario synthesis

0,5

0,0
1990 1995 2000 2005 2010

Figure 3: Raw material use of computers – 1990 to 2010

Here the share of the use phase is highly predominant, as is shown in Figure 4, thus encouraging
strategies mainly targeting the use phase : scenario 2 (substitution), 1 (targeting the demand) and to a
lesser degree scenario 5 (rational use of the product) are the most promising. On the contrary,
reusing computers could have an adverse effect on raw material use, since it results in a longer
lifespan of old, less effective equipment (mainly due to CRT screens).

raw material use in 2010

2,5

2
kton eq Sb

disposal phase
1,5 use phase
production phase

0,5

0
BAU scenario 1 scenario 2 scenario 3 scenario 5 scenario 7 scenario
synthesis

Figure 4: Raw material use of computers in 2010 per life cycle stage and per strategy

Waste production
Belgium is a leading country in Europe in terms of product take-back and recycling. Product take-back
was installed through the Recupel taxation scheme that started in beginning 2001, thus 2 years
3
before the European WEEE directive .
A computer’s average lifespan of 5 years was assumed for the period 1990 - 2005, decreasing to 4
years in 2010. Indeed, most computers are replaced not because they are not working anymore, but
because they are obsolete and do not have sufficient capacities to support ever more memory- and
speed-demanding applications.
The model used here was a very simple one. The Gaussian distribution of waste production was not
taken into account for two reasons: it was difficult to integrate in Excel modelling and, in the case of
3
Directive 2002/96/CE.

1060
computers, this Gaussian distribution is not very accurate. Indeed, people tend to store their
computer, before either reselling it, or finally dispose of it (usually when it is too old) [10].

Ultimate waste

1,4
1,2
1 BAU
scenario 1
0,8
kton / yr

scenario 2
scenario 3
0,6 scenario 4
0,4 scenario 5
scenario 7
0,2 scenario synthesis

0
-0,21990 1995 2000 2005 2010

Figure 5: Waste production of computers : 1990 – 2010

Figure 5 shows the waste production calculated in this study. It shows that although Belgium is
already doing a lot of efforts, waste from computers could be lowered further, especially by improving
the recycling processes. It should be noted that these figures do not account for waste production in
the production and distribution phases. This is thus an under-estimation of actual impacts of
computers over their life cycle, and will be discussed later on.

Policies and measures

The results of this project underlined the importance of impacts in the production phase. Thus,
computers should be considered not only as products that consume energy in the use phase
(electricity), but also products resulting from very energy-intensive processes that are thus worthy of
careful treatment as a “waste”, and where reuse is a policy option that could lead to substantial
environmental benefits.
We presented to key stakeholders a list of possible improvement measures taken from literature or
foreign experience. The comments from each stakeholder for each strategy are presented hereunder,
with a view of showing all opinions on these matters.

Scenario 1 : reduction of final demand

• Industry emphasised the need for a product policy to change consumption patterns, but not
consumption levels. The development of services versus computers (e.g., through the increase of
computer centres) was not seen as a good option due to possible side-impacts on traffic. However,
in order to decrease the “digital gap”, the industry is favourable to the presence of free PC access in
town halls.
• For the actors from the social economy (non-profit organisations specialised in the collection and
reselling of electrical equipment) (selling second-hand computers), it is important to inform
consumers about the impacts of computers when they buy one. Especially, information on the cost
of upgrading vs. the cost of a new computer.
• For environmental NGOs, decreasing demand would be possible by subsidizing specialised small
and medium enterprises (SMEs)which could bring to households the use of software.
• For consumer organisations, the symbolism of computers is now so important that targeting the
final demand would be extremely difficult.
• Researchers in sociology also think that the symbolism of computers is now so important that
targeting the final demand would be extremely difficult.
Agreements on this strategy
There seems to be an agreement with the difficulty of the task of controlling the demand for
computers. Increasing services seems the only possible solution, but a further analysis of the impacts
associated with them, as well as the means to promote them, would be necessary.

1061
Scenario 2: product substitution
• Concerning this strategy, the industry underlines the fact that such a substitution must be carried
out carefully, taking into account the whole life cycle impacts of the different products as well as a
cost / benefit analysis of the change. Attention must also be paid to side-effects of such changes,
e.g. concerning the security of the product. Concerning possible measures, ecolabels are not seen
as a good measure since it is not a compulsory label, thus disadvantaging other environment friendly
products. They also comment that international discussions have already started to develop the
Energy Star label.
• For consumer organisations, the promotion of energy efficient computers should preferably take
place through compulsory labels rather than ecolabels, because consumers are not receptive to
labels. For the same reason, developing existing labels should be preferred to creating new ones.
• For researchers, rather than developing new labels, they prefer field agreements.
Agreements on this strategy
There is a general agreement on the uselessness of promoting ecolabels for computers (they are
considered to be ineffective labels). Stakeholders agree that a better solution to promote the use of
energy-efficient equipment would be the development of existing compulsory labels, or by field
agreements.
Scenario 3: product reuse
• For industry, reuse can be an attractive option if scientific evidence suggests that it is an effective
and environmentally friendly solution, and provided that it can be implemented at a reasonable cost,
without detracting from the technical characteristics required for the intended application. The
limitations with respect to reuse make it very important to continue encouraging research and
development (R&D) to improve the eco-efficiency of computers. Industry maintains that the
promotion of second-hand computers is not always the best solution, since new PCs are more
energy efficient than older ones
• For the actors from the social economy, computers “cost” a lot during their production in terms of
environment. Increasing their life span is therefore logical. They are also in favour of open-source
software, which enables the consumers to have a flexible system which can evolve as they wish.
They also underline that reused computers do not directly compete with new ones on the market; on
the contrary, they fulfil different needs (e.g. basic needs in terms of speed and power) in a
responsible manner. Finally, they agree with the need to increase the upgradeability of computers,
which should be carried out in the assembly phase.
• Consumer organisations warn about a possible rebound effect, where people who cannot afford a
new computer could possibly buy a used one, thus increasing the penetration rate of computers in
households. Also, the network of second-hand equipment sales should be associated with the new
equipment network, so that salesmen can point the consumer towards the best choice depending on
needs and constraints. But in this case, second-hand equipment should be interesting to sell for the
new equipment distribution network. They also suggest to develop a tool to know the characteristics
of the computer fulfilling the consumer’s needs. This tool could be supplied through the internet and
sales points.
• For environmental NGOs, in order to lengthen the life span of computers, PC design is very
important (for maintenance in particular).
Agreements on this strategy
In this strategy there is a clear opposition between the industry and the social economy actors.
However, this strategy is particularly important for computers, as about half of their CO2 emissions
occur in the production phase. Further research and market studies would be necessary to determine
if second-hand computers compete with new ones on the market place, and if developing them would
lead to a rebound effect. It should be emphasised that developing the market for second-hand
computers could possibly lead to the creation of new jobs in Belgium.
Scenario 5 : improvement of product use
Many comments from researchers underlined the necessity to help consumers to use computers
more efficiently. For example, sociologists believe that stand-by enabling should be set up
automatically and less easily changed by users. Also, the benefits for consumers should be clearly
stated and given to them, in order to increase the efficiency of communicating.
Scenario 7 : material recycling
• Industry is of the opinion that 1) no measures can be imposed which make the use of recycled
material obligatory and 2) market forces themselves must determine where the raw materials yielded
via recycling can be most effectively deployed.
• For the social economy actors, this strategy is an “end-of-pipe” measure that is only efficient if
accompanied by priority measures linked to prevention and reuse. Experience for white goods

1062
 shows that monetarisation of flows inhibits reuse and limits the freedom of stakeholders in terms of
private initiatives. Moreover, the IT sector works in a business to business view, thus old equipment
often find solutions in the framework of leasing contracts or sales with / without take-back of old
equipment.
• In general, environmental NGOs insist on the fact that public authorities must orient industrial
policy depending on determined social objectives. It should be remembered that economic
instruments would be very useful to send the necessary signals to reorient production and
consumption patterns. These instruments are more and more refused by industry delegates. On the
contrary, they support awareness measures that have minor impact. Public authorities should
overcome these wishes and target public interest.
Agreements on this strategy
For this strategy, it is difficult to make conclusions. The industry is against regulations in the field of
recycling, and the social economy thinks this should not be a preferred strategy (but rather reuse).

Discussion

Producing computers
When having a look at the calculations of impacts over the life cycle of computers, we notice that the
impacts occurring from the production phase mainly stem from the core process of chip
manufacturing. This manufacturing consists in building a series of layers on top of a silicone wafer,
therefore laying the network of transistors and diodes that give a chip its functionality. This process
requires large amounts of chemicals, water, and energy [8], actually about 500 times the weight of a
chip is required in a series of raw materials (including fossil fuels) in order to manufacture it. As a
comparison, a car requires twice its weight of fossil fuels to be produced [11].
The GHG emissions estimated by our model is actually an underestimation, since it does not take into
account PFCs use as a solvent. In 1993, an estimation of PFC use in the semiconductor industry
showed that about as much global warming impact occur due to PFC use as to electricity
consumption during the production [8]. PFCs have since been phased out to some extent, but in an
un-quantified amount.
Besides waste from end-of-life disposal, waste during production also occurs. In 1997, the Electronics
Industry Association of Japan reported 15 kg of waste occurring during the production of a desktop
unit [12]. A large majority consists in acid and alkali wastes.
A favoured way of reducing impacts from computer production is to replace part of the production of
new computers by the reuse (with possible upgrade) of older ones. This strategy would lead to a
decrease of GHG emissions occurring from the production of computers; however, it should be
coupled with the replacement of old CRT screens by LCD screens; otherwise, electricity use in the
use phase is likely to increase compared to the BAU situation.
This proposition was one of the biggest sources of disagreement in the consultation phase of the
study. Industries are reluctant to envisage this strategy that they consider threatening to the economic
health of their sector. Open-source software, that can be a good option in order to solve the problem
of software equipment on used computers, is also considered with caution. Other actors emphasise
the social impacts of reusing computers. One of the main negative effects could be the rebound
effect, where used computers would not substitute to new ones but actually increase the penetration
rate of computers on the Belgian market. Moreover, given the social status that is linked to owning the
latest, most trendy type of computer, this could be hard to achieve.
While reuse thus appears as an approach to be taken cautiously in households, it stills can be a good
option in schools and offices, where decisions are taken more rationally than in households.
Concerning the stabilisation of the demand for computers, all stakeholders are doubtful on this
possibility. The only way that could possibly work would be to increase the offer of computer as a
service, thus developing internet cafés, computer renting etc. How this strategy could be implemented
would require a more in-depth analysis of social representations of computers.

Using computers
The trend of GHG emissions and raw material use is to remain level until 2010. Indeed, although an
increase of sales of computers in Belgium can be expected and begins to be noticed, the trend
towards more “flat screens” (LCD screens) and laptops is decreasing the average consumption per
unit. A CRT screen consumes on the average 60W; a LCD screen, 15 W; a laptop, about 20W [4].
The unknown factor in these figures is the market shares of different screen sizes : indeed, the larger
the screen, the higher the electricity consumption. 17” screens seem now the largest category of

1063
screens on the market, having largely replaced 15” screens, and even larger screens appear on the
market. The trend for the years to come may influence slightly these power consumption figures.
Besides the question of CRT versus LCD screens, the issue of energy efficient equipment should be
tackled. Energy efficient equipment can today be identified by either the European Ecolabel, or the
Energy Star label. The former is actually not available on the Belgian market; the latter has outdated
requirements and does not really identify energy-efficient equipment [13]. Stakeholders are not willing
to develop new labels; indeed, there is a general state of confusion on labels in Belgium, consumers
generally not recognising them or not knowing what they mean [14]. Creating new labels would only
lead to increased confusion. The priority thus seems to update the Energy Star requirements at the
international level. This has been undergone at the European level where legislation has been taken
4
for display units (CRT and LCD screens) and criteria are under revision for computers.
Using computers in an efficient manner is also an option. It has small effects on total GHG emissions
(electricity being mainly produced from nuclear in Belgium), but higher impacts in terms of raw
material use. It requires using the stand-by mode of both the monitor and the central unit, but also
shutting down if the equipment is expected to stay unused for a longer period. This would require
improving the ergonomics of the stand-by and shut-down management options, which are largely
unused by inexperienced users.

End-of-life computers
Belgium is already ahead of existing legislation related to electrical and electronic equipment. This
means that additional improvement will be difficult to get. However, several strategies could be
beneficial.
First, a “mechanical” decrease of waste production would occur both from the reduction of the
demand and from the substitution of products. LCD screens and laptops being much lighter, their
development could contribute to waste reduction in terms of quantities.
Second, reusing computers would decrease the production of waste at a significant degree; indeed, if
all computers had a second life of the same length as the first one, this would halve yearly waste
production from end-of-life products. The feasibility of this option is yet to be assessed, 10 years of
working life being considerable for a computer.
Last, increasing recycling targets would enable a further decrease of waste production. This should
be envisaged only when reuse is not feasible (not-repairable equipment for instance). However, the
network of EEE collection in Belgium (especially for large equipment such as computers) is now well
developed; the critical step would therefore be the development of recycling facilities and their
treatment capacities.
It should be underlined that this study does not take into consideration a phenomenon that is currently
being investigated at the federal level: some computers are exported as second-hand computers to
developing countries, but they are actually only EEE waste. Waste treatment in developing countries
is of course not as efficient as in Belgium, resulting in very harmful exposure to pollutants for the
population [7].

Conclusion
Computers are not often seen as polluting equipment. However, they have non-negligible impacts in
terms of GHG emissions, raw material use and waste production. Production in particular is a very
polluting phase, and it produces as much greenhouse gas emissions as the CP itself during its whole
lifespan; however since it does not occur in Belgium the only possibility to influence this phase is a
product-oriented policy aiming at favouring products with the smallest impacts on the environment
over their life cycle. A preferred way of doing so is to encourage the reuse of computers; this would
require the development of the social economy network, thus also offering the possibility to create
jobs. It would also have the benefit of reducing computer waste.
The impacts of the use phase could be mitigated by increasing the proportion of LCD screens and
laptops. They could also be addressed by updating the most widely known energy efficiency label to
include an active mode power draw component, Energy Star, in order to make it a real commercial
incentive for low electricity consumption.
Finally, it is worth underlining the fact that this sector is changing at an increasing pace. Technology
developments occur at such a rate that it makes modelling and policy design very difficult. For this
reason, it is important to involve the industries manufacturing or reselling computers for reuse (social
economy network). Belgium is probably not the best country for this to happen, given its small impact

4
Commission Decision 2005/42/EC

1064
on the computer production market, but it is necessary to begin raising awareness and finding
solutions at a more global level.

References
[1] Wallenborn,G. , Rousseau,C. , Thollier K, Residential energy consumption: Necessity and
possibility of segmenting population, to be published
[2] T.J. Gordon, 1994, The Delphi Method, in AC/UNU Millennium Project : Futures Research
Methodologies
[3] M. Turoff, 2002, The Delphi methods : techniques and applications
[4] FhI, 2003, Gestion de la demande d’énergie dans le cadre des efforts à accomplir par la Belgique
pour réduire ses émissions de gaz à effet de serre
[6] EPA, 2002, Desktop computer display : a life cycle assessment
[5] GfK Consulting, http://www.gfkbenelux.nl/fbi/flexpage/flexpage.asp?id=4454
[7] Handy and Harman Electronic Materials Corp, 2002, in « Exporting Harm : the high-tech trashing
of Asia
[8] The 1.7 Kilogram Microchip: Energy and Material Use in the Production of Semiconductor Devices.
E. Williams, R. Ayres, and M. Heller. Environmental Science & Technology 36 (24). 5504-5510.
Dec. 15 (2002)
[9] Williams, E, Kuehr R., Computers and the environment : understanding and managing their
impacts, 2003
[10] Matthews, H. Scott, Francis C McMichiael, Chris T Hendrickson and Deanna Hart, 1997,
Disposition and end-of-life options for personal computers, Carnegie Mellon University Green
design initiative technical report
[11] Mac Lean H., and L. Lave, 1998, A life-cycle model of an automobile. Environmental Science and
Technology 32(13): 322A – 329A
[12] EIAJ, 1997, Environmental vision for the electronics parts industry
[13] Schaeppi B. and Ritter H., 2003, Energy Star : Actual market relevance and prospects for the
future – An Austrian market study, EEDAL 03 proceedings
[14] CRIOC, 2004, Label écologique européen: quels impacts sur les choix de consommation ?

1065
1066
Residential IT Energy Consumption in the U.S.
Kurt Roth1, Ratcharit Ponoum1, Fred Goldstein2
1
TIAX LLC
2
Ionary Consulting

Abstract
TIAX carried out a study to quantify the energy consumption of U.S. residential Information
Technology (IT) equipment in 2005 and develop scenario-based projections for 2010 [1]. This paper
presents the key findings of this study for 2005. The project team identified ten key equipment types
and evaluated their energy consumption in 2005. Prior studies suffered from a dearth of data about
PC and monitor usage patterns and these devices account for a majority of residential IT energy
consumption. Consequently, TIAX independently commissioned a survey to assess residential IT
usage patterns in 2005 that posed a dozen questions to people in 1,000 demographically-
representative households about the usage patterns for up to three computers per household. This
survey yielded more accurate estimates of PC and monitor use than prior studies, which markedly
improved the accuracy of the estimate of total residential IT energy consumption. Overall, U.S.
residential IT equipment consumed about 42TWh of electricity in 2005, or about 0.46 quads of
primary energy. This translates into approximately 3% of residential electricity consumption and 1%
of U.S. electricity consumption and about 0.45% and 2.4% of U.S. and residential primary energy
consumption, respectively. Desktop PCs and monitors account for about two-thirds of the energy
consumed by the key equipment types. The current study’s estimate of total residential IT energy
consumption is much higher than prior estimates, primarily because the new usage data indicate that
PCs and monitors spend much more time in active mode.

Introduction
Over the past decade, the widespread commercialization of the Internet, increased integration of IT in
peoples’ lives, and consistently large gains in capability and functionality coupled with consistent
decreases in equipment costs have resulted in a dramatic increase in the use of information
technology (IT). Consequently, residential IT equipment has begun to have an impact on residential
electricity and energy consumption. An earlier study estimated that residential IT equipment
consumed 16.5TWh of electricity (0.2 quads primary) in the U.S. in 2001 [2]. Desktop PCs and their
monitors accounted for more than 75% of the total. In the context of residential buildings, they found
1
that residential IT accounted for approximately 1% of both total electricity and primary energy
th
consumption circa 2001. This suggests that U.S. residential IT energy consumption was about 1/6
of commercial IT energy consumption (based on [3]).
Although the earlier study indicates that residential IT accounted for a relatively small portion of U.S.
residential energy consumption in 2001, recent trends suggest that it may have increased
substantially since 2001 due to an increased number of PCs and other IT equipment in residences
and the possibility of increased utilization (e.g., due to greater deployment of broadband Internet
connections and home networks). In addition, the data used to estimate PC and monitor usage
patterns had limitations that make estimates based upon the data have a high degree of uncertainty.
To support its strategic planning efforts, the U.S. Department of Energy, Building Technologies
Program (DOE/BT), contracted TIAX to develop an estimate of U.S. residential IT equipment energy
consumption in 2005 and scenario-based projections for 2010. To realize those goals, TIAX and
DOE/BT decided upon the following approach to the project:
1. Generate a list of equipment types and collect existing data from literature.
2. Develop a preliminary estimate of national energy consumption for each equipment type.
3. Select 5 to 10 equipment types for further evaluation, based upon preliminary calculations
and perceived growth in future energy consumption.

1
Primary energy, as opposed to site energy, takes into account the energy consumed at the electric plant to generate the
electricity. On average, each kWh of electricity produced in Y2000 consumed 11,030 Btu [4].

1067
4. Develop refined bottom-up estimates of national energy consumption of each selected
equipment type, for Y2005 and three scenario-based projections for Y2010.
5. Compare the current results with the results of other studies.
6. Publish the findings in a report, including feedback from government and industry.

This report describes the methodology, results, findings, and recommendations of the evaluation of
residential IT energy consumption in 2005.

Approach
Equipment Analyzed

A wide range of energy-consuming IT equipment is found in homes. Moreover, the continued

integration of IT into more aspects of home life suggests that the proliferation of new kinds of devices
will only continue to grow in the future. The definition of what equipment falls under the scope of this
study, i.e., are defined as IT equipment, is somewhat arbitrary. For instance, IT could, potentially,
encompass a broad range of different equipment types, ranging from office equipment to telephony to
many consumer electronics (e.g., televisions, personal audio products, etc.). Notably, a growing
portion of a wide variety of electricity-consuming devices are incorporating IT functionalities, such as
on-board computing and networked connectivity. Due to the scope limitations of this study, however,
the project team could only consider and model the energy consumption of a limited subset of
equipment types. Table 1 shows the equipment considered for evaluation in the current study.

Table 1: Potential and Selected Key Equipment Types

Potential Key Equipment (Key equipment types in BOLD)
o Broadband Access Devices (cable and DSL modems, satellite)
o Cellular/Wireless Phone
o Copy Machines
o Desktop PCs
o Digital Set-Top Boxes with Personal Video Recorder (PVR)
o Facsimile Machines
o Fiber Optic Terminal (w/ Fiber to the home [FTTH])
o Home Routers (wired and wireless)
o Home Servers
o Inkjet Printers
o Laptop PCs
o Laser Printers
o Monitors
o Multi-Function Devices (MFDs)
o Personal Digital Assistants (PDAs)
o Scanners
o Uninterruptible Power Supplies (UPS)
o Voice Over IP Adaptor (VoIP)

To focus the analysis on the equipment that accounts for most residential IT energy consumption, the
team selected up to ten key equipment types for evaluation from many potential equipment types.
The selections reflected preliminary estimates of candidate equipment types’ energy consumption in
2005, as well their projected impact in 2010. In addition, the selections had a moderate bias for
devices that had not been evaluated by prior studies. Subsequently, the team analyzed the energy
consumption for each key equipment type in 2005.

Energy Consumption Calculation Methodology

Figure 1 shows the basic methodology used to develop the annual energy consumption (AEC)
estimates.

1068
 Device Annual Unit
Mode Annual Usage, Power,
Electricity Consumption,
by Mode by Mode by Mode

x = Device Unit
Active Tactive P active UECactive Residential
Electricity
Consumption Stock

x =

M
Sleep Tsleep P sleep UECsleep = UEC X S = AEC

x =
Off Toff Poff UECoff

Figure 1: Annual Energy Consumption Methodology Source: [3]

For each equipment type analyzed, the project team calculated the average unit energy consumption
(UEC, in kWh) of a single device for an entire year. The UEC equals the sum of the products of the
approximate number of hours that each device operates in a residential setting in each power modes
and the power draw in each mode. The product of the estimated residential stock (i.e., installed base)
and UEC yields the total annual energy consumption (AEC, in TWh) for that equipment type.
Reference [3] describes the calculation methodology in greater detail.
The following sections describe our approach to developing values for the different components of
AEC calculations.

Residential Equipment Stock

Residential building equipment stock simply means the number of devices in use in residential
buildings. Stock estimates primarily came from published estimates, such as industry market reports,
the U.S. Department of Energy’s Energy Information Administration (EIA) Residential Energy
2
Consumption Survey (RECS ), and other research reports. Overall, residential stock estimates
appear to have the smallest uncertainty of all three components of device AEC calculations. Table 2
displays the stock estimates for the different key equipment types; the full report [1] provides details
about the stock estimates, as well as the sources used to develop each.

Table 2: Residential Stock Estimates in 2005 for Key Equipment Types.

Sources: See [1].
Equipment Type Stock Estimate
[millions]
Broadband Access Devices 32
Desktop PCs 85
Inkjet Printers 75
Laptop PCs 36
Monitors 85
Multi-Function Devices (MFDs) 13
Digital Set-Top Boxes with Personal Video Recorder (PVR) 10
Uninterruptible Power Supplies (UPS) 8.5
Voice Over IP Adaptor (VoIP) 1.5
Home Routers (wired and wireless) 15

2
EIA characterizes the RECS as: “a national statistical survey that collects energy-related data for occupied primary housing
units. In the 2001 RECS, data were collected from 4,822 households in housing units statistically selected to represent the
107.0 million housing units in the United States.” More information about RECS can be found at:
http://www.eia.doe.gov/emeu/recs/ .

1069
Usage Patterns
A device’s usage pattern refers to the number of hours per week that, on average, a device operates
in a given mode. Most equipment types have several different modes that are typically condensed
into three distinct modes (see Table 3). The analysis also uses two other modes for certain
equipment types. Inkjet printers and MFDs have a “ready to print” mode instead of an active mode,
3
as printing accounts for a small portion of residential inkjet UEC . In addition, digital set-top boxes
with PVR and VoIP adaptors both have an on-ready mode for when the device is on but not providing
user functionality in lieu of a sleep mode. In many cases, power management (PM) strategies (such
®
as the voluntary EnergyStar program operated by the U.S. Department of Energy and the U.S.
Environmental Protection Agency) and their degree of implementation have a major impact on the
amount of time spent in the active and sleep modes.

Table 3: Office Equipment Usage Modes

Mode Type Description Example
Active Device carrying out intended operation • Monitor displays image
• Copier printing
Sleep Device not ready to carry out intended • Monitor powered down but on
operation, but on • Copier powered down but on

Off Device not turned on but plugged in • Monitor off, plugged in

• Copier off, plugged in

Table 4 displays the usage patterns estimated for the different key equipment types; the full report [1]
provides details about the usage patterns, as well as the sources used to develop each estimate.

Table 4: Usage Patterns Estimates in 2005 for Key Equipment Types

Equipment Type Average Annual Hours by Mode
Active Sleep Off
Broadband Access Devices 8,760 0 0
Desktop PCs 2950 350 5460
Inkjet Printers 50* 1610 7100
Laptop PCs 2370 930 5460
Monitors 1860 880 6020
Multi-Function Devices (MFDs) 50 1610 7100
Digital Set-Top Boxes with Personal Video Recorder (PVR) 2890 5870** 0
Uninterruptible Power Supplies (UPS) 2500 1070 5180
Voice Over IP Adaptor (VoIP) 360 8400** 0
Home Routers (wired and wireless) 8760 0 0
Sources: See [1]. *Ready to print for inkjet printers and MFDs. **On-ready mode for set-top boxes and
VoIP adaptors.

In contrast to commercial building IT equipment, very few measurements exist for residential IT
usage. Most prior studies of residential IT energy consumption have used informed estimates for
usage by mode ([2], [5]). One survey, the EIA Residential Energy Consumption Survey (RECS),
asked respondents to estimate weekly PC usage and reported results in very broad time bands (e.g.,
2-15 hours, 16-40 hours, 41+ hours [6]). Although these data were the best available prior to this
study, they provide only a general feel for PC active mode usage and do not directly address time
spent in off or sleep modes.
Because desktop PCs and monitors appear to dominate residential IT energy consumption (per [2]
4
[5]), TIAX decided to commission a phone survey to develop more refined and up-to-date estimates
5
for PC and monitor usage by mode (henceforth referred to as the “TIAX Survey”). The TIAX survey

3
Assuming that an inkjet printer consumes an additional 30W while printing (than in ready mode) and prints five pages per
minute, printing a single page consumes approximately 0.1W-h. If a residential inkjet printer printed 1,000 pages per year,
printing would contribute approximately 1kWh to inkjet printer UEC.
4
This survey was funded in its entirety by TIAX LLC, i.e., no government funds were used to carry out the survey.
5
An appendix in the full report [1] contains the complete survey.

1070
posed twelve questions to 1,000 demographically-representative households about the usage
patterns of up to three PCs and monitors used most often in each household. The current study uses
usage data for PCs and monitors derived from the TIAX survey.
The team developed two models to translate the TIAX survey responses into residential PC and
monitor daily weekday and weekend usage patterns. Together, they provide upper and lower bounds
for PC and monitor usage. The current usage estimates presented for PCs and monitors use the
lower-bound approach (i.e., “OffModel” in Appendix A); applying the upper-bound approach would
increase desktop PC and monitor AEC by approximately 30% and 25%, respectively. Appendix A
discusses the two models in more detail.
Power management (PM)-enabled rates can have a significant impact on the UEC of a given device,
particularly if the device remains on overnight. Consequently, the TIAX survey posed questions to
ascertain the PM-enabled rates for PCs and monitors. Based on responses to the question: “If the
computer monitor of the computer is left on, after one hour or more of no use, does it continue to
display the same image, display a screensaver, or go blank?,“ residential monitors have a PM-
enabled rate of about 60%. This rate is consistent with the estimate of reference [2] and similar to
values from surveys of monitors used in commercial buildings [7].
The survey data also suggest that the residential desktop PC power management (PM) enabling rate
appears to be about 20%, which is higher than the commercial sector rate (roughly 6% [7]). This
estimate has appreciable uncertainty. The project team attempted to develop a meaningful question
that users could readily answer to determine the PM-enabled status of PCs and solicited feedback
from several people outside of TIAX. Ultimately, PM-enabled rates were derived from respondents’
answers to question: “If members of your household leave the computer on and do not use the
computer for one or more hours, how long does it take for the computer to respond to moving the
mouse or typing on the keyboard?”. Potential responses included:

01 Almost instantaneously
02 Within a few seconds
03 In about ten or more seconds
99 DON’T KNOW

The thought was that a PC that took more time to respond to input after a significant period of non-
use would have a higher probability of being in sleep mode. Consequently, the analysis assumes that
6
a PC that responded “In about ten or more seconds” had its power management enabled . If,
however, “Within a few seconds” and “In about ten or more seconds” both denoted PM-enabled, the
PM-enabled rate rises to about 50% and desktop PC AEC decreases by about 20%.
As one researcher points out, responses to the question may correlate more with monitor PM-enabled
status, i.e., people are aware if the monitor displays an image or not, and may prevent accurate
characterization of the PCs’ PM-enabled status. Other measures associated with PM, such as hard
drives spun down, do not necessarily mean that the measure that typically achieves the largest
reduction in power draw (powering down the CPU) has occurred [8]. On the other hand, it is not clear
that most users would be aware of these events and could provide meaningful responses to a
telephone survey. Consequently, the approach taken provides a general sense of PM-enabled status
but has appreciably more uncertainty than metered data. Overall, device usage patterns have the
greatest uncertainty of any component of the AEC calculations.

Power Draw by Mode

The AEC estimates rely upon power draw estimates for all of the relevant operational modes for each
equipment type analyzed. For each mode, the power draw value represents the best estimate for the
average power draw of all of the different devices included in a single equipment type. This estimate
assumes that annual usage by mode does not vary appreciably with power draw by mode, e.g., that
desktop PCs draw 120W in active mode do not spend appreciably more hours in active mode per
year than desktop PCs that draw 50W in active mode. This simplification is not demonstrably true;
however, in general, the magnitude of the error introduced by this assumption is likely on the order of
or less than that of the magnitude of other uncertainties in usage patterns.
Whenever possible, the active mode power draw values reflect actual power draw measurements as
opposed to the device rated power draw. Rated power draws represent the maximum power that the

6
All “DON’T KNOW” responses were excluded from the PM-enabled percentage calculation.

1071
device’s power supply can handle and often exceed typical active power draw values by at least a
factor of three (see, for example, [3]).
Table 5 displays the power draw by mode estimates for the different key equipment types; the full
report [1] provides details about estimates, including the sources used to develop each estimate.

Table 5: Power Draw by Mode Estimates in 2005 for Key Equipment Types
Equipment Type Average Power Draw by Mode
Active Sleep Off
Broadband Access Devices 6 n/a n/a
Desktop PCs 75 4 2
Inkjet Printers 13* 5 2
Laptop PCs 25 2 2
7
Monitors 45 2 1
Multi-Function Devices (MFDs) 19 11 7
Digital Set-Top Boxes with Personal Video Recorder (PVR) 27 25* n/a
Uninterruptible Power Supplies (UPS) 9 6 6
Voice Over IP Adaptor (VoIP) 6 4 n/a
Home Routers (wired and wireless) 6 n/a n/a
Sources: See [1]. n/a = not applicable. *Ready to print for inkjet printers and MFDs. **On-ready mode
for set-top boxes and VoIP adaptors.

Unit Energy Consumption

Table 6 summarizes the UEC values for the key equipment types.

Table 6: UEC Values for Key Equipment Types in 2005

Equipment Type UEC [kWh]
Broadband Access Devices 53
Desktop PCs 234
Inkjet Printers 23
Laptop PCs 72
Monitors 91
Multi-Function Devices (MFDs) 68
Digital Set-Top Boxes with Personal Video Recorder (PVR) 225
Uninterruptible Power Supplies (UPS) 60
Voice Over IP Adaptor (VoIP) 36
Home Routers (wired and wireless) 53

Results and Discussion

In 2005, U.S. residential IT equipment was estimated to consume 42TWh (see Figure 2), or about 3%
of residential electricity consumption and 1% of U.S. electricity consumption [10]. For comparison
8
sake, residential IT is projected to account for about 1.4% of electricity consumed in Germany in
2005 [11]. Translated into primary energy, residential IT accounts for 0.46 quads, which equals
9
approximately 0.45% of U.S. primary energy consumption in 2005 [10].

7
Monitor average power draw values based on the following installed base distribution: 55% 17-inch CRT, 20% 15-inch LCD,
30% 17-inch LCD [9].
8
This total includes additional devices not included in this study, including: Mobile phone chargers, personal digital assistants
(PDAs), scanners, matrix printers, copiers, “active boxes,” ISDN boxes, and assumes that 50% of set-top boxes and satellite
receivers are digital devices in 2005.
9
Assuming that each kWh of electricity requires the consumption of 10,913 Btus on average to generate, transmit, and
distribute [4].

1072
 Total = 42TWh

Router 0.8
Broadband
VoIP Adapter Other 3.9 Access 1.7
0.1
UPS 0.5
PVR/DVR 2.2
MFDs 0.9

Desktop PC
Monitors 7.7 19.9

Laptop PC 2.6
Inkjet Printers
1.7

Figure 2: U.S. Residential IT Energy Consumption in 2005 (in TWh)

In 2005, desktop PCs and monitors account for about 65% of residential IT energy consumption.
“Other” devices included come from reference [2], with scanners accounting for around ¾ of the total.
As the authors of that report note, the value for scanners appears to be quite high. In general, the
AEC of at least three of the four “other” devices is probably decreasing as MFDs supplant scanners,
facsimile machines, and copy machines.

Comparison to Other Studies

The current estimate of residential IT energy consumption is more than twice as high as the Y2001
estimate of reference [2]. In addition, the current estimate for the sum of desktop PC, laptop PC, and
monitor AEC is about 35% greater than that used in the Annual Energy Outlook (AEO [10]; see Figure
3). Overall, increased residential IT usage estimates for both PCs and monitors, based on the TIAX
Survey, drive the higher levels of total AEC. Specifically, the TIAX Survey suggests that PCs and
monitors spend at least 100% and 60% more hours, respectively, in active mode than estimated in
10
reference [3] (based on [6] ). This occurs because approximately 20% of PCs remain on overnight
on a given day. On the other hand, the current study finds that 2005 residential IT energy
11
consumption is less than half of projections of commercial IT energy consumption from reference
[3].

10
For comparison sake, the active mode hours per year estimates for desktop PCs and monitors are more than four and five
times greater than those projected for Germany in 2005 [11].
11
The values shown for [3] equal the projections for the key equipment types divided by 0.9 to include energy attributed to other
(i.e., non-key) devices.

1073
 140 130

120
1 09
97
100 92
AEC [TWh]

80

60
42
40
22
16
20

AEO - 2005^
Ubiquitous -
Residential -
Commercial -

Residential -

Commercial,

Commercial,
Commercial,

PC Reigns -

Residential
Greening -
2005**
2001#

2005*

2005*

2005*
2000*

*R e fe re n c e [3 ]
# R e fe re n c e [2 ]

**C u rre n t S tu d y
^ R e fe re n c e [9 ]; in c lu d e s o n ly P C s a n d M o n ito rs
Figure 3: Comparison of Current U.S. Residential IT Energy Consumption Estimate with Other
Studies

The authors believe that the current study provides a more accurate estimate of residential IT energy
consumption than RECS for several reasons. First, the TIAX Survey was carried out in March of
2005, so it provides more recent data than the RECS. Second, the TIAX Survey posed more targeted
questions about residential PC and monitor usage than RECS, including questions about individual
(versus all) devices’ daily usage on both weekdays and weekends, nighttime status, daytime status
when not in use, and PM-enabled status. Third, the TIAX Survey elicited information about the usage
of up to three computers and their monitors for each household, whereas the RECS poses a single
question about the number of hours that all of the PCs are on during day. Fourth, the TIAX Survey
allowed a much wider range of responses than the RECS, e.g., respondents provided numerical
estimates of both typical weekday and weekend residential IT active use hours instead of the broad
ranges used in the RECS, i.e., less than 2 hours, 2 to 15 hours, 16 to 40 hours, 41 to 167 hours, and
168 hours per week [6].

Conclusions and Recommendations

Conclusions
TIAX carried out a study to quantify the energy consumption of Residential Information Technology
(IT) equipment in 2005 and develop scenario-based projections for 2010. The study identified ten key
equipment types and evaluated their energy consumption in 2005 and for each 2010 scenario.
Overall, the key residential IT equipment analyzed consumed about 42TWh of electricity in 2005, or
about 0.40 quads of primary energy. This translates into approximately 2.8% of residential electricity
consumption and 1% of U.S. electricity consumption and about 0.39% and 2% of U.S. and residential
primary energy consumption, respectively. Desktop PCs and monitors account for about two-thirds of
the energy consumed by the key equipment types.
This represents a dramatic increase in residential IT energy consumption relative to prior estimates of
residential IT energy consumption, i.e., more than twice the 16.5 TWh estimated for 2001 [2] and
about 35% greater than the Annual Energy Outlook projection for 2005 for PCs and monitors [10].
Relative to these two prior studies, the current study estimates that PCs and monitors spend much
more time per week in active mode, i.e., approximately 100% and 60% more, respectively. This
occurs because a non-trivial minority of PCs remain on overnight at least some portion of the time.
The current study uses usage data derived from a survey of 1,000 demographically-representative
households of residential PC and monitor usage patterns carried out by TIAX in 2005. This survey
yielded much more accurate estimates of PC and monitor usage. Because desktop PCs, laptop PCs,

1074
and monitors account for a majority of residential IT energy consumption, this markedly improved the
accuracy of the estimate of total residential IT energy consumption.

Recommendations
This study yields two distinct recommendations, one programmatic and the other informational.
Programmatically, this study reinforces the strong link between power management (PM)-enabling
rates and residential IT energy consumption. The TIAX Survey suggests that residential PCs may
have higher PM-enabling rates than commercial PCs (approximately 20% as compared to likely less
than 10% [10]), while monitor PM-enabled rates are around 60%. Nonetheless, increasing PM-
enabled rates both devices has a large energy saving potential that may well increase in the future as
always-on operation of residential IT equipment becomes even more common. Consequently,
programmatic actions that can increase the PM-enabled rate of PCs and monitors have a large
energy saving potential. This study also points out the need to collect good data about residential PC
and monitor usage. To date, almost all usage data have come from survey responses or reflect
researchers’ rough estimates. The finding of this study that residential IT now accounts for about 3%
of residential electricity consumption emphasizes the need to collect representative, high-quality data
about IT usage for desktop PCs, preferably via monitoring of equipment power draw for a useful
period of time (e.g., two weeks or more).

Acknowledgements
The authors wish to thank the U.S. Department of Energy, Building Technologies Program, for
funding the study upon which this paper is based. In addition, Mr. Lew Pratsch, Mr. Sam Johnson, Dr.
James Brodrick of U.S. Department of Energy provided day-to-day oversight of this assignment and
helped to shape the approach, execution, and documentation.

References
[1] TIAX. U.S. Residential Information Technology Energy Consumption in 2005 and 2010. Final
Report by TIAX LLC to the U.S. Department of Energy, Building Technologies Program. March,
2006. Can be downloaded at: http://www.tiaxllc.com/aboutus/abo_news_bytype_reports.
[2] Nordman B. and Meier A. Energy Consumption of Home Information Technology. Lawrence
Berkeley National Laboratory Report, LBNL-5350. July, 2004.
[3] ADL. Energy Consumption by Office and Telecommunications Equipment in Commercial Buildings
– Volume I: Energy Consumption Baseline. Final Report by Arthur D. Little, Inc., to the U.S.
Department of Energy, Office of Building Technology, State and Community Programs.
January, 2002. Can be downloaded at:
http://www.eere.energy.gov/buildings/documents/pdfs/office_telecom-vol1_final.pdf.
[4] BTS. 2004 Buildings Energy Databook. U.S. Department of Energy, Energy Efficiency and
Renewable Energy. August, 2004. Can be downloaded at:
http://btscoredatabook.eren.doe.gov/ .
[5] Kawamoto, K., Koomey J., Nordman, B., Brown, R., Piette, M.A., Ting, M. and Meier A. Electricity
Used by Office Equipment and Network Equipment in the U.S.: Detailed Report and
Appendices. Lawrence Berkeley National Laboratory Report, LBNL-45917. February, 2001.
Can be downloaded at: http://enduse.lbl.gov/Info/LBNL-45917b.pdf .
[6] EIA. Residential Energy Consumption Surveys: Home Office Equipment Surveys. U.S. Department
of Energy, Energy Information Administration (EIA). Can be downloaded at:
ftp://ftp.eia.doe.gov/pub/consumption/residential/2001hc_tables/homeofc_household2001.pdf.
[7] Roberson, J.A., Webber, C.A., McWhinney, M.C., Brown, R.E., Pinckard, M.J. and Busch, J.F.
After-Hours Power Status of Office Equipment and Inventory of Miscellaneous Plug-Load
Equipment. Lawrence Berkeley National Laboratory Final Report, LBNL-53729. January,
2004. Can be downloaded at: http://enduse.lbl.gov/info/LBNL-53729.pdf .
[8] Carrie Webber: personal communication. Lawrence Berkeley National Laboratory.
[9] iSuppli, 2005, “Computer Monitor Historical and Projected Sales and Inventory Data,” Provided by
P. Semenza to TIAX LLC, October.
[10] EIA. Annual Energy Outlook 2005 with Projections to 2025. U.S. Department of Energy, Energy
Information Administration Report, DOE/EIA-0383. 2005. Can be downloaded at:
http://www.eia.doe.gov/oiaf/aeo/ .
[11] Cremer, C., Eichhammer, W. Friedewald, M. Georgieff, P. Rieth-Hoerst, S. Schlomann, B. Zoche,
P. Aebischer, B. and Huser, A. Energy Consumption of Information and Communication

1075
 Technology (ICT) in Germany up to 2010. Final Report to the German Federal Ministry of
Economics and Labour, Project Number 28/01. January, 2003. Can be downloaded at:
http://www.isi.fhg.de/e/eng/publikation/online/iuk/iuk-e.htm .

Appendix A – Usage Models to Interpret the TIAX Survey

The team developed two models to translate the TIAX survey responses into residential PC and
monitor daily weekday and weekend usage patterns. Together, they provide upper and lower bounds
for PC and monitor usage. The first model, referred to as the OffModel, assumes that the user turns
off his/her PC outside the timeframe that the PC has the potential to be routinely used. The second,
known as the OnModel, makes the opposite assumption, that is, that the PC is left on. The survey
explicitly asked how many hours each PC (and, thus, its monitor) is actively used on both weekdays
and weekend days and both models directly incorporate this data. Using the data from the TIAX
Survey, each model also calculates the hours each PC and its monitor spend active-unused,
sleeping, and off per week based on a used-hours allocation algorithm, model-specific assumptions,
and a weighting mechanism that reflects the average likelihood that the PC and its monitor will be
turned off during the day when not in use and overnight. Specifically, both models assume that a PC
is most likely to be continuously used from 7pm to 10pm during a typical weekday and 3pm to 5pm for
weekend days (for the sake of simplicity, the discussion will focus on weekday hours). In all cases,
monitors were assumed to have the same basic usage patterns as their PCs, albeit using the monitor
PM-enabled and night status information instead of that for the PC.
On Model
The usage pattern for the OnModel first allocates active-used time in the 7pm-10pm window. If
additional active-used hours remain, the model subsequently allocates a half hour of usage at 7am,
i.e., it assumes that the PC is turned on first thing in the morning. It then allocates remaining active-
used hours by adding a 0.5-hour active-unused break (assuming that users often take a break from
12
their PCs instead of continuously using them ), followed by up to a 2-hour active-used period. If
active-used hours still remain, the algorithm repeats this 0.5-hour active-unused 2-hour active-used
pattern as many times as necessary, working back to 7am until all the active-used hours have been
allocated (see Figure A-1). If additional hours remain, the algorithm converts the 0.5-hour active-
unused periods to active-used to account for the remaining active-used hours (beginning with the half
hour closest to the 7pm-10pm time window, i.e., 6:30-7pm). Finally, if yet more time remains, time
during the 10-11pm period can be assigned to active-used.
Used

2 3 1 night

Unused

7am 7pm 10pm

Figure A-1: Illustrative Application of the OnModel for Active-Used Hours

Once turned on in the morning, the model uses the response to the question about how often users
leave the PC on during the day if it is not used for more than half an hour to determine its daytime
status. If a PC is never left on, the model assumes that the PC is turned off until the additional active-
used hours later during the day (case ONA). If a PC is always left on during the day, the model
assumes that the PC remains in active-unused for an hour. Subsequently, any time between that
point in time and the time of the later active-used hours begin is allocated to active-unused and
13
sleeping based on the devices PM-enabled status (the ONB case). The response to the question
about how often users leave the PC on during the day if it is not used for more than half an hour is
used to calculate the weighting between the ONA and ONB cases and calculate the average usage by
mode (see Table A-1). For example if the response is that the person “often” leaves their PC on
during the day if not used for more than half an hour, the ONB case would receive a 75% weighting
and the ONA case a 25% weighting.

12
Implicit in this assumption is that the PC does not have sufficient time during the 0.5-hour breaks to enter sleep mode, an
assumption generally consistent with the sleep-mode settings for new desktop PCs.
13
In essence, the model assumes that a PC spends 1 hour in active-unused before sleeping.

1076
Table A-1: OnModel Weighting
Based on Survey Response
Survey Response OnModel
Weighting
Always 100%
Almost always 90%
Often 75%
About half the time 50%
Occasionally 25%
Rarely 10%
Never 0%
Source: Reference [1].

The OnModel also consists of two weighted sub-models based on whether or not the PC is turned off
at night. If the PC remains on at night, the hours will either be active-unused or sleep, depending on
14
the PCs PM-enabled status (ON-NightA). If the PC is turned off, it is off (ON-NightB). The weighting
assigned to the two night models, ON-NightA and ON-NightB, is determined by the response to the
question about how often users typically turn off their PC at night and used to calculate the average
usage by mode.
For weekend usage, the algorithm first allocates time from 3pm to 5pm and then propagates the 0.5-
hours active-unused-2-hours used allocation sequence before 3pm. If additional active-used hours
still remain, it then begins to convert the ½-hour active-unused periods to active-used (as with the
weekday model) and, subsequently, assigns hours continuously (i.e., without ½-hour active-unused
periods) after 5pm to active-used until all of the active-used hours have been allocated. In the
OnModel, the day is assumed to begin at 9am (versus 7am on weekdays). The nighttime calculation
procedure does not change.
OffModel
The OffModel uses both the OnModel and another model, depending on the night status of the
device. If the device is always turned off at night, the OffModel assumes that the device is not turned
on at 7am (or 9am on weekends). Instead, it first allocates active-used hours in the 7pm-10pm (or
3pm-5pm) window and any additional active-used hours via the ½-hour active-unused 2-hours active-
used allocation sequence described above, followed by conversion of the ½-hour active-unused
periods to active-used and, if necessary, addition of the 10-11pm period (case OFFA). That is, all
remaining daytime hour are off hours (see Figure A-2). If the device is always left on at night, the
OffModel uses the OnModel to calculate the usage pattern (case OFFB). The weighting assigned to
cases OFFA and OFFB is determined by the response to the question about how often users typically
turn off their PC at night and used to calculate the average usage by mode.
Used

2 1 night

Unused

7am 7pm 10pm

Figure A-2: Illustrative Application of the OFFA Case of the OffModel for Active-Used Hours

The following example illustrates how the model works. The model allocates hours as follows for a
person responding that she typically used her PC four hours a day and always (100% weight to
OffModel) turned off her PC at night. First, the OffModel allocates three hours to the 7-10pm period.
Next, it allocates one hour to the 5:30pm-6:30pm period, assigning ½ hour to active-unused between
6:30pm and 7pm. Finally, it allocates all remaining hours to off. In total, the model allocates four
hours active-used, ½ hour active-unused, and 19.5 hours off per day.
If, however, she responded that she turned off her PC at night occasionally (25%) and rarely (10%)
left her PC on after not using the PC for more than half an hour during the day, and her PC was found

14
It is also worthwhile to point out that in the case where PC usage is less than 3 hours per day, e.g., the active-use hours do
not fill the 7pm to 10pm window, is not filled entirely, the algorithms assume that the PC is turned off immediately after the last
used hour if the PC is turned off at night. For example, if a PC is used for 1 hour per day and the user turns it off for the night,
the PC is off starting at 8pm and remains so throughout the night.

1077
to have PM enabled, the calculation becomes more complex. In that case, the model calculates daily
usage in two ways and weights them to derive the usage estimate (see Figure A-3 and Table A-2).
For the 25% of the days when the PC is turned off at night, the usage calculation uses the process
described in the prior paragraph and receives a 25% weighting (Figure A-3a). For the 75% of the
days when the PC remains on at night, the model allocates three active-used hours to the 7-10pm
period. Next, it allocates another active-used ½ hour to 7am-7:30am and the remaining active-used ½
hour to 6pm through 6:30pm, with ½ hour spent in active-unused between 6:30pm and 7pm. The
PM-enabled response determines whether the nighttime hours are allocated to sleep or active-
unused. Since PM was enabled, the model assumes that the PC enters sleep mode after one hour
from 10pm to 11pm active-unused and allocates the rest of the time from 11pm to 7am to sleep. The
allocation of the daytime hours between 7:30am and 6pm depends on her response as to how often
she left her PC on when it had not been used for more than half an hour during the day. Based on
her “rarely” response (10%), the model calculates two usage scenarios and weighs each. For the
10% of the time the PC is left on, the model allocates one hour, from 7:30am to 8:30am, to active-
unused and the remaining 9.5 hours from 8:30am to 6pm to sleep (because PM was enabled; Figure
A-3b). For the 90% of the time that the PC is turned off, the model allocates all 10.5 hours to off
(Figure A-3c). Applying the appropriate weightings, the model calculates the following daily usage for
the 75% of the days that the PC remained on at night: 4 hours active-used, 1.6 hours active-unused,
8.95 hours sleep, and 9.45 hours off.

On-used On-unused Sleep

40
30
Off at Night: 3a)20
10
25% Weight 0
0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00

0
0
On During Day:
3b)0
10% weight 0
0
On at Night: 0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00
75% Weight
0
0
Off During Day: 3c)0
90% weight 0
0
0:00 2:00 4:00 6:00 8:00 10:00 12:00 14:00 16:00 18:00 20:00 22:00

Figure A-3. Example of Usage Hours Allocation Model

Subsequently, the model weights the outputs from those two scenarios appropriately (10% and 90%)
and sums them to obtain the average daily usage for the 75% of the days when the PC remains on at
night. Finally, the model weights the model usage output for the 75% of the days when the PC
remains on at night and that for the 25% of the days when the PC is turned off at night and sums
them to obtain the average daily usage estimate for that PC (see Table A-2).

Table A-2. Sample Translation of Survey Responses to a Usage Pattern

Survey Response Total Hours per Day
Weighting Active- Active-
Sleep Off
Used Unused
Off at Night – 25% of Time 0.25 4 0.5 0 19.5
On During Day – 0.075
4 2.5 17.5 0
10% (=0.75*0.1)
On at Night –
Off During Day – 0.675
75% of Time 4 1.5 8 10.5
90% (=0.75*0.9)
TOTAL 0.75 4 1.6 8.95 9.45
TOTAL 4 1.33 6.71 11.96

1078
The Power Challenge - Intel’s Holistic Approach to Power
Management

Kevin Fisher, Todd Brady

Intel Corporation

Abstract
Driven by Moore's Law, semiconductor manufacturers such as Intel are able to continually produce
new innovative products that deliver increasing levels of performance and other user-valued
capabilities. However, as more transistors are packed into a smaller area, power density increases,
creating challenges for cooling and thermal management. Efficient power delivery and thermal
management are critical as systems become smaller and more capable with every generation of new
computing and communication products.
In 1983, the Intel® 286 microprocessor consisted of 134,000 transistors. Intel microprocessors today
can contain over 1 Billion transistors (see Figure 1). Similarly, the computing power of the PC has
increased by factors far exceeding 1,000X since the early 1980s. If the power consumed by an
average PC had increased at the same rate, each one would require a 250-300 kilowatt (kW) power
supply. Instead, for an average PC, the power consumed has stayed largely the same over the last 20
years despite dramatic improvements in PC performance and computing ability [1, 2].

Figure 1: Growth in the number of transistors in an Intel CPU

2
The active power for a CMOS device is defined as: P = CV f, where P = active power needed for
switching, C = total capacitance being switched, V = operating voltage and f = switching frequency.
Although the switching frequency has increased dramatically over the past decade, Intel has focussed
on driving both the voltage and capacitance down through improvements in manufacturing
technologies. As a result, energy consumed per CMOS logic switching has decreased about 300X
since the early 1990s [3].
Intel is a leader in developing innovative solutions to address and resolve power challenges. To be
successful, Intel has taken a holistic approach toward power management by tackling the challenge at
all levels: micro/macro architecture, silicon and circuit design & manufacturing, packaging, platform
design, software optimization, and ecosystem enabling. Building on its past, but further emphasizing
the future, in August 2005, Intel’s CEO Paul Otellini announced a new product direction for the
company that put even more emphasis on energy efficiency.

1079
1 Intel Power Management achievements: Recent history

1.1 Mobile
In 2003, Intel released its Pentium® M processor which when combined with the Intel® 855 chipset
family and the Intel® PRO/Wireless 2100 Network Connection made up the key building blocks of the
TM
Intel® Centrino™ mobile technology. Intel® Centrino mobile technology improves both performance
and battery life over previous mobile processors. Energy efficiency has improved by over 30% using
industry standard benchmarks [11].
In 2005, Intel released mobile processors with Enhanced Intel SpeedStep® technology and Intel®
Mobile Voltage Positioning. Both technologies helped to minimize the power consumption of the
mobile processor. Enhanced Intel SpeedStep® technology enables real-time dynamic switching of
the voltage and frequency between two performance modes based on processor demand. Intel®
Mobile Voltage Positioning (Intel® MVP IV) dynamically lowers voltage, based on processor activity.
In partnership with the Mobile PC Extended Battery Life Working Group (www.eblwg.org), Intel led a
successful effort to increase the energy efficiency of LCD’s by about 40%. LCD screens are the
largest source of power consumption in a notebook PC (~30-40% of the total power). This work
successfully reduced the energy consumption of the screen from ~5 watts to 3 watts or less.

1.1 Desktop
When evaluating the total system power of a desktop PC, it can be seen that the processor consumes
only about 10% of the total power. The video display devices and power supplies tend to dominate. If
a CRT monitor is used, the monitor and power supply alone can account for up to 75% of the total
desktop system power. If a LCD monitor is used, this value drops to about 50% of total system power
[1].
Through research which started in 2000, Intel was able to show that the desktop PC power supply
was a major source of energy inefficiencies for the system (some power supplies were as low as 50%
efficient and most were designed to give optimum performance at full load). As a result of these
findings Intel issued an update to its Power Supply Design Guidelines to include minimum energy
efficiency targets for power supplies at 3 loadings – 20%, 50% and full load. As a result of these
efforts, typical power supply efficiencies today are of the order of 80%.
Intel has a long history of actively working to improve the power management of PCs through work in
industry groups to develop open industry specifications for power management. As an initial founder
of Advanced Power Management (APM) and the follow-on Advanced Configuration and Power
Interface (ACPI), Intel has helped develop and promote the use of sleep states to reduce overall
system power consumption. In September 2004, ACPI version 3.0 was made available to the public.
In 2005, a new feature was introduced into desktop (and server) products as a means of reducing
platform idle power consumption. Enhanced autohalt (c1E) is a low power state entered into when the
processor executes the HALT instuction. Since introduction, the autohalt feature has been continually
optimized to enabling lower power states for the microprocessor.
There are many other examples of contributions from Intel to enable more power efficient desktop
system designs such as: Intels integrated graphics chipset product line and the enabling of new form
factor standards (actually Intels enabling of new form factors covers many market segments such as:
mobile, ultramobile, workstation mobile, desktop, gaming machine, workstation, server, etc).

1.2 Servers
Servers present their own unique challenges for power management integration. Server systems can
have multiple processors, significantly large memory, redundancies, multiple networking cards and
hard drives. A server architecture, design and usage model is significantly different than a desktop or
notebook PC. A server may be used locally or remotely with one or even millions of users.
Availability and response time to an uncertain frequency of requests for service is paramount. These
requirements pose challenges for energy efficiency of servers and the facilities that host them.
In 2004, Intel launched its first server processor products with Enhanced Intel Speed Step®
technology to support Demand Based Switching (DBS). DBS minimizes power consumption of the
server system by dynamically changing the processor performance states. The performance states
are changed based on demand for computing power and/or utilization. For systems using DBS,
energy savings of up to 24% can be realized [12]. As illustrated in Figure 2, energy savings are
greatest when processor utilization is less than 50%, decreasing as the processor approaches full
utilization. Since most servers are utilized much less than 100% under typical operating conditions,
DBS has the potential for significant energy savings.

1080
 400

Without DBS
System Power (Watts)

350

300

250

200
With DBS

150
0% 20% 40% 60% 80% 100%
% CPU Utilization

Figure 2: Effect of DBS on System Power Consumption

Intel introduced low voltage versions of its Intel® Xeon® processors in 2005. These processors can
be used in server rack and blade designs where space is constrained and power-efficiency is a
priority. These processors incorporate Demand Based Switching (DBS) technology.
Technologies such as “autohalt (c1E)” (as mentioned under desktop PC above) and improved power
supply efficiencies have also been introduced to server based products.

1.3 Silicon
In August 2005, Intel introduced its 65 nm manufacturing technology. 65 nm technologies allow
printing of individual circuit lines on a semiconductor device at widths smaller than that of a virus. The
gate within the transistor is even smaller, with a width of 35 nm and a thickness of 1.2 nm or 5 atomic
layers [4]. At such small sizes, leaking current, which grows exponentially as the size of the transistor
shrinks, becomes a problem. If steps are not taken to control it, leakage current can become a barrier
to practical device operation [5].

1.3.1 Strained Silicon

Intel first introduced strained silicon into its 90 nm technology and is now used to manufacture many
of Intel’s state of the art semiconductor products, such as the Intel® Core™ Duo processor. Strained
silicon uses a region of silicon with built in stress, or strain, to increase the speed of the current flow
across the transistor. By stressing or straining the silicon crystal lattice, electrons can flow with less
resistance. Figure 3 illustrates this point. The result of such a technology is a 5-25X reduction in
leakage current and 10-25% improvement in transistor current [5, 6, 7].

Figure 3: Illustration of Strained Silicon Benefits [8]

1081
1.3.2 Sleep Transistors
Another method for reducing leakage current is to turn off, or put to sleep, the transistors of the silicon
device which are idle or not actively in use. For example, Static Random Access Memory (SRAM)
makes up a significant portion of Intel 65 nm microprocessors. SRAM is used to cache data and
instructions. Sleep transistors can be used to shut off blocks of SRAM that are idle, saving energy by
stopping the leakage current in these sections of the microprocessor [9].

1.3.3 High K Dielectric

In order to further reduce leakage at the gate, Intel has developed new thicker gate material termed
“high-k dielectric.” The high-k material reduces leakage by 100 times over existing silicon dioxide
materials [10]. Figure 4 is an image of both the traditional silicon dioxide gate and the new high-k
gate material.

Figure 4: SEM image of existing SiO2 gate material and

new “high-k” gate material.

1.4 Power-Optimized microarchitecture

Intel continually strives to eliminate redundancy at the microarchitecture level by identifying frequent
instruction sequences, extensively optimizing them, and storing them for later reuse. For example,
Intel® NetBurst® microarchitecture has an advanced form of an instruction cache called the
Execution Trace Cache, which stores already-decoded machine instructions or micro-ops for future
reuse. Hyper-Threading Technology is another Intel microarchitecture that has been increasing
performance without impacting the power envelope.

1.5 Offices and factories

In its offices and factories, Intel completed over 20 energy improvement projects in 2005. These
projects resulted in savings of over 20 million kwhrs of electricity and nearly 2 million therms of natural
gas. These results have been achieved through the use of improved controls, heat recovery, and
other conservation techniques. The 2005 projects are part of an ongoing multi-year effort that has
resulted in savings of over 200 million kwhrs of electricity and approximately 5 million therms of
natural gas. Due to these projects, Intel’s energy used per unit of product manufactured has declined
significantly over the last several years, well ahead of our publicly stated goal to reduce normalized
consumption 4% per year. Intel has now begun working with the suppliers to drive improved efficiency
in the manufacturing tools used in production. We believe progress in this area will complement the
work begun on facility systems and continue to drive further improvements in the overall energy
efficiency of the manufacturing process. More information is available at:
http://www.intel.com/intel/other/ehs/perform.htm

2 Intel Power Management achievements: Current achievements

In the microprocessor world, performance usually refers to the amount of time it takes to execute a
given application or task or the ability to run multiple applications or tasks within a given period of
time. However, true performance is a combination of both clock frequency (GHz) and Instructions
Executed per Clock Cycle (IPC).
• Performance = Frequency x Instructions per Clock Cycle (IPC)
Therefore it is possible to increase performance by increasing clock frequency or instructions per
clock cycle or both. Today, Intel is focusing on delivering optimal performance together with improved

1082
energy efficiency, eg to take into account the amount of power the process will consume to generate
the performance needed for a specific task.
As noted in the introduction, power consumption is related to the dynamic capacitance required to
maintain IPC efficiency, times the square of the voltage that the transistors and I/O buffers are
supplied with, times the frequency that the transistors and signals are switched at
• Power consumed = Capacitance x Voltage x Voltage x Frequency (P = CV f).
2

By taking into account both performance and power equations, designers can carefully balanced and
therefore optimise performance and power efficiency.

2.1 Intel® Core™ microarchitecture: setting new standards for energy efficient performance
The move to multi-core processing has opened the door to many other micro-architectural innovations
that further improve performance and energy efficiency. Intel Core microarchitecture is one such
state-of-the-art micro-architectural update that has been designed to deliver increased performance
combined with superior energy efficiency. The Intel Core microarchitecture is a new foundation for
Intel architecture based desktop, mobile, and mainstream server multi-core processors and is
expected to start shipping in Q3 2006.

2.1.1 Intel Wide Dynamic Execution

Intel Wide Dynamic Execution enables delivery of more instructions per clock cycle to improve
execution time and energy efficiency. Every execution core is wider allowing each call to fetch,
dispatch, execute and return up to four full instructions simultaneously (previous technologies could
handle up to three instructions at a time).

2.1.2 Intel Intelligent Power Capability

Intel Intelligent Power Capability is a set of capabilities designed to reduce power consumption and
design requirements. This feature manages the runtime power consumption of all the processor’s
execution cores. It includes an advanced power gating capability that allows logic control to turn on
individual processor logic subsystems only if and when they are needed. Additionally, many buses
and arrays are split so that data required in some modes of operation can be put in a low-power state
when not needed. In the past implementing power gating has been challenging because of the power
consumed in the powering down and ramping back up, as well as the need to maintain system
responsiveness when returning to full power. Through Intel Intelligent Power Capability, Intel has
been able to satisfy these concerns, ensuring both significant power savings without sacrificing
responsiveness.

2.1.3 Intel® Advanced Smart Cache

The Intel® Advanced Smart Cache is a multi-core optimized cache that significantly reduces latency
to frequently used data, thus improving performance and efficiency by increasing the probability that
each execution core of a dual-core processor can access data from a higher-performance, more
efficient cache subsystem.

2.1.4 Intel® Smart Memory Access

Intel® Smart Memory Access improves system performance by optimizing the use of the available
data bandwidth from the memory subsystem and hiding the latency of memory accesses. Intel Smart
Memory Access includes an important new capability called "memory disambiguation," which
increases the efficiency of out-of-order processing by providing the execution cores with the built-in
intelligence to speculatively load data for instructions that are about to execute before all previous
store instructions are executed.

2.1.5 Intel® Advanced Digital Media Boost

Intel® Advanced Digital Media Boost is a feature that significantly improves performance when
executing Streaming SIMD Extension (SSE/SSE2/SSE3) instructions. They accelerate a broad range
of applications, including video, speech and image, photo processing, encryption, financial,
engineering and scientific applications. The Intel Advanced Digital Media Boost feature enables these
128-bit instructions to be completely executed at a throughput rate of one per clock cycle, effectively
doubling, on a per clock basis, the speed of execution for these instructions as compared to previous
generations.

1083
2.2 Platform-Scalable Architectures
The new Intel Core microarchitecture will provide a solid foundation for new server, desktop, and
mobile platforms.

2.2.1 Server Platforms

Servers can deliver significantly greater compute performance and compute density. Intel is
developing a DP Server processor optimized for dual-core based on the new, state of the art, Intel
Core microarchitecture, codenamed Woodcrest. The Woodcrest processor is targeted for introduction
in the third quarter of 2006 and will work within the Bensley server platform and the Glidewell
workstation platform. The Bensley server platform and the Glidewell platform are targeted for
introduction in the second quarter of 2006.
Intel will also deliver a quad-core (4 full execution cores) processor to the DP Server segment based
upon this new microarchitecture, codenamed Clovertown. Clovertown is targeted for introduction in
the first quarter of 2007, on the Bensley and Glidewell platforms.
For the MP server segment, Intel is also developing a MP Server processor optimized for quad-core
based on this new microarchitecture, codenamed Tigerton. The Tigerton processor is targeted for
introduction in 2007 and will work within the Caneland server platform.

2.2.2 Desktop Platform

Desktops can deliver greater compute performance as well as ultra-quiet, sleek and low-power
designs.
Intel is developing a desktop-optimized, dual-core processor based on the new, state of the art, Intel
Core microarchitecture, codenamed Conroe. The Conroe processor will work within the 2006 Digital
Home platform codenamed Bridge Creek, and the 2006 Digital Office platform, codenamed Averill.
Conroe is targeted for introduction in the third quarter of 2006.
Intel will also deliver a quad-core (4 full execution cores) processor to the high-end desktop based
upon this new microarchitecture, codenamed Kentsfield. Kentsfield is targeted for introduction in the
first quarter of 2007.

Mobile Platform
Laptop users can take advantage of the increased multi-core compute capability within the mobile
form factors. Intel is developing a mobility-optimized, dual-core processor based on the new, state of
the art, Intel Core microarchitecture, codenamed Merom. The Merom processor will work within the
Intel® Centrino Duo® mobile technology-based platform and Merom is targeted for introduction to
align with the 2006 holiday buying season.

3 Intel Power Management achievements: Looking further ahead

3.1 Low Power on Intel Architecture Project
In addition to technologies and products on the market today to improve energy efficiency, Intel is
actively researching future technologies for possible future applications. For example, in the consumer
and business marketplaces, smaller devices are proliferating. Smart phones, notebooks, and micro
PCs are the leading edge of a wave of new devices designed for communication and entertainment.
As part of Intel's vision of architectural innovation for convergence, the Intel Systems Technology
Labs is working to accelerate these next-generation technologies and products.
The Low Power on Intel® Architecture (LPIA) project of STL focuses on researching and developing
low-power technology building blocks for future Intel® architecture-based platforms. Key learning’s
from this research will lay the groundwork as Intel product development groups move toward low
power on Intel architecture.
Using the research platform, the LPIA team is conducting research to better understand and reduce
the thermal and physical demands of computing technology. A critical focus area is extending battery
life for portable devices. In addition, the team is developing power management polices and metrics
for future Intel architecture-based platforms.
Intel is also performing system-level profiling and benchmarking, spanning from the power source to
power distribution and on to power consumers. Power-smart platforms will extend battery life and
enhance user experiences by applying best-in-class power management technologies. Focused on
system software policy management, the STL LPIA project is:
• Researching system-level power states and aggressive power-management policies
• Developing power metrics to calibrate power management in handheld devices
• Focusing future efforts on close cooperation with OS vendors for implementation

1084
3.2 Tri-gate Transistors
Intel has also developed a novel three-dimensional design that helps make transistors that scale,
perform, and address the current leakage problem seen in smaller dimension planar transistors. Tri-
gate fully depleted substrate transistors have a raised plateau like gate structure with two vertical
walls and a horizontal wall of gate electrode. This 3D structure improves the drive current while the
depleted substrate reduces the leakage current when the transistor is in the off state. Reducing the
leakage current in the off state translates to increased battery life in mobile devices.
Intel believes that these new discoveries can be integrated into an economical and high-volume
manufacturing process to address the power and heat increases in increasingly smaller transistors.

3.3 Intel and QinetiQ Collaborate On Transistor Research

Researchers from the two companies have successfully built 'quantum well' transistors by integrating
a new transistor material, pioneered by QinetiQ called indium antimonide (InSb). InSb is made up of
elements found in the III and V columns of the periodic table. Transistors made of this material enable
research devices to operate at very low voltages, while still rapidly switching and consuming little
power. The research results obtained from the quantum well transistors research showed a 10x lower
power consumption for the same performance, or conversely a 3x improvement in transistor
performance for the same power consumption, as compared to today's traditional transistors.

4 Summary
Intels efforts in power management go back more than a decade (though this paper has focused on
more recent activities). For example Intel was one of 13 companies to receive the EPA’s first Energy
Star Computer Awards back in 1994. Intel is an initial founder of Advanced Power Management
(APM) and the follow-on Advanced Configuration and Power Interface (ACPI), Intel helped develop
and promote the use of sleep states to reduce overall system power. In response to Energy Star’s
computer energy-efficiency specification, Intel developed in 2001 the new Instantly Available PC
Power Management to improve sleep-state power management.
Intel today works closely with regulatory bodies such as the US EPA and EU/EC in driving Energy
Star and other WW regulatory standards to improve computing platform energy efficiency. Today Intel
is one of 20 companies working with the DOE and EPA to help define the new Energy Star
specifications.
This paper has described a number of Intels recent past, present and future activities aimed at
improving the energy efficiency of computer devices. These efforts can be summarised in the table
below:

1085
 Recent history Today Future
Servers • enhanced autohalt (c1E) • Intel® Core™ • Low Power on
• Enhanced Intel Speed Step® microarchitecture Intel® Architecture
technology • Intel Wide Dynamic (LPIA)
• Demand Based Switching Execution • Tri-gate
• low voltage versions of its Intel® • Intel Intelligent Transistors
Xeon® processors Power Capability • Intel and QinetiQ
Desktop • Power Supply Design Guidelines • Intel® Advanced Collaborate On
• Intel® Active Management Smart Cache Transistor
Technology • Intel® Smart Research
• ACPI version 3.0 Memory Access • Ever smaller
• New form factors • Intel® Advanced transistors (22nm)
• Integrated functionality Digital Media Boost • Etc
• enhanced autohalt (c1E)
• Intel® Centrino mobile
TM
Mobile
technology
• Enhanced Intel SpeedStep®
technology
• Intel® Mobile Voltage
Positioning
• Mobile PC Extended Battery Life
Working Group
Silicon • 65 nm manufacturing technology
• Strained Silicon
• Sleep Transistors
• High K Dielectric
• Intel® NetBurst®
microarchitecture
• Hyper-Threading Technology

“Lead the industry in performance per watt across all market segments” is one of Intels strategies
within its 2006 strategic objectives for 2006.
Intel observes that energy efficiency demand from users is most pronounced for notebook and server
products. Notebooks, because of battery life demands and servers due to high end data center
energy demands. Data centers increasingly want to add more computing performance. Doing so
requires more energy efficient products in order to effectively cool and stay within the power budget of
the datacenter.
Energy efficiency has not been a major market driver for desktop computers. However, factors such
as acoustics and smaller form factors are beginning to drive this market need. Despite this, as can be
seen in this paper, Intel is investing heavily in continually improving the energy efficiency of desktop
products.

5 Conclusions

Moore’s Law will continue to drive advances in semiconductor manufacturing. Intel’s manufacturing
process roadmap predicts the development and use of 22 nm technology by 2011. Such
advancement will continue to make technical challenges such as leakage current even more
pronounced. As a result, Intel has made power management a top priority in its technology
roadmaps.
To be successful in addressing the power challenges of the PC, one must take a holistic approach
toward power management, tackling the challenge at all levels: micro/macro architecture, silicon and
circuit design & manufacturing, packaging, platform design, software optimization, and ecosystem
enabling. Intel has taken such an approach and has made significant progress toward meeting these
challenges.
• The need for raw computing performance has evolved into a drive for energy-efficient
performance to meet people's expanding demands for more capabilities and higher
performance – whether for smaller devices, longer battery life, or greater power savings. Intel

1086
 is driving innovations in computing multi-core architectures to deliver new levels of
performance, capabilities and energy efficiency.
• The Intel® Core™ microarchitecture, Intel's new foundation for delivering even greater energy
efficient performance, is expected to deliver significant performance gains and power
reductions in desktop (Conroe) and server (Woodcrest) processors and to extend the strong
energy-efficient performance leadership of the Core Duo processor.
• Supporting the new multi-core architecture are Intel's unparalleled manufacturing capacity
and the most energy-efficient performance CPU transistors in the world.
• Intel delivers energy-efficient performance advances across its architecture, silicon, platforms
and software to help the industry's leading companies create new uses, build new markets,
and meet the evolving needs of people and businesses worldwide.
Delivering energy-efficient performance requires a holistic effort across all common platform
components – processors, chipsets, hard drives, power supplies, graphics cards, memory
subsystems, displays, BIOS, software and more. Intel’s manage these components as a collective
system. This creates a platform whose components work together to deliver performance when
required and to conserve resources when one or more individual resources are not needed.
Building energy efficient products in energy efficient buildings is without question high on Intels
agenda. Technology realities and market demand are two of the factors determining our strategic
research agenda direction on energy efficiency.

References
http://www.intel.com/technology/eep/index.htm

[1] T. Brady, K. Fisher, “Intel’s Technology Contributions to Energy Efficiency of IT Products”,

International Conference on Improving Energy Efficiency in Commercial Buildings, Frankfurt,
Germany, April 2004.
[2] D. Cole, “Energy Consumption and Personal Computers,” Chapter 7 in Computers and the
Environment, R. Kuehr and E. Williams, Eds., Kluwer Academic Publishers, 2003, pp. 136-138.
[3] C. Calwell, C. Hershberg, and D. Hiller, “Forging Ahead with Desktop PC Power Supply Efficiency
Improvements,” Intel Technology Symposium, September 8, 2004.
[4] M. Bohr, “Nanotechnology Goals and Challenges for Electronic Applications,” IEEE Transactions
on Nanotechnology, Vol 1, No. 1, p.56, March 2002.
[5] K. Mistry et al., “Delaying Forever: Uniaxial Strained Silicon Transistors in a 90nm CMOS
technology,” Syposium VLSI Technology, pp.50-51, June 2004.
[6] T. Ghani et al., “A 90 nm High Volume Manufacturing Logic Technology Featuring Novel 45 nm
Gate Length Strained Silicon CMOS Transistors,” IEEE International Electron Devices Meeting
(IEDM), Dec 2003.
[7] P. Bai et al., “A 65nm Logic Technology Featuring 35nm Gate Lengths, Enhanced Channel
2
Strain, 8 Cu Interconnect Layers, Low-k ILD and 0.57 mm SRAM Cell,” IEEE International
Electron Devices Meeting (IEDM), Dec 2004.
[8] Source of graphic: www.intel.com/technology/silicon/si12031.htm
[9] “Designing for Power – Intel Leadership in Power Efficient Silicon and System Design,” 2004,
www.intel.com/technology.
[10] R. Chau, “Advanced Metal Gate/High-k Dielectric Stacks for High Performance CMOS
th
Transistors,” American Vacuum Society 5 International Conference on Micoelectronics and
Interfaces, March 2004.
[11] Mobile Mark 2002 benchmark data. Available at:
http://www.intel.com/performance/mobile/centrino_mobile_experience.htm
[12] D. Bodas, “New Server Power-Management Technologies Address Power and Cooling
Challenges,” Technology@Intel Magazine, September 2003.
http://www.intel.com/update/contents/sv09031.htm

1087
1088
New Approaches to Energy-Efficiency Specifications: Considering
Typical Electricity Consumption
Darcy Martinez1, Bruce Nordman2
1
ICF International
2
Lawrence Berkeley National Laboratory

Abstract
®
The current ENERGY STAR specifications for imaging equipment address electricity consumption
only in low-power modes. As products increase in functionality and speed, Active mode contributes to
a greater portion of total product energy use. Additionally, consumer dissatisfaction with long recovery
times can result in decreased use of energy-saving features. When EPA began to revise the imaging
equipment specifications, the need to address Active mode for some products quickly became
apparent. This was accomplished with the “typical electricity consumption” (TEC) approach that
considers the electricity consumed by imaging equipment during its entire duty cycle. This method for
assessing product energy efficiency has been received favorably by many stakeholders and
demonstrates a forward-thinking approach to the development of an energy-efficiency specification.
This paper reviews the process of creating the TEC method and resulting test procedure and
specification levels. The authors present the key considerations for developing this innovative
approach, including:
• Determining the universe of covered products;
• Harmonizing the approach with international standards;
• Developing the test method;
• Accounting for international usage patterns;
• Collecting new TEC data;
• Setting appropriate specification levels; and
• Securing stakeholder support.
The paper reviews specific lessons learned during this process and includes detailed examples that
illustrate the vision for this new method.

The Roots of ENERGY STAR Imaging Equipment

In 1992, EPA introduced the first ENERGY STAR product specification for computers and monitors.
The suite of products covered by ENERGY STAR expanded over the next five years to include
imaging equipment such as printers, fax machines, copiers, multifunction devices, and scanners. The
backbone of these and all ENERGY STAR specifications are performance criteria, which are
developed in consultation with industry stakeholders to address a product’s energy or power
consumption. The early ENERGY STAR imaging equipment specifications were generally
characterized by performance criteria that addressed a product’s power consumption in low-power
modes (e.g. Sleep or Off), and the default time in which the product’s low-power modes were
activated. For example, the Version 1.0 ENERGY STAR copier specification launched in 1995
required that low-speed copiers enter an Off mode of fewer than five watts within 30 minutes of
completing the last copy job.
Targeting low-power mode in imaging equipment was logical for EPA. Manufacturers were eager to
participate in ENERGY STAR and there was available technology to employ power-saving modes in
imaging products. By the time the copier specification had been in effect for one and one-half years,
EPA had signed partnership agreements with copier manufacturers that produced more than 90
percent of the copiers sold in the United States, and had more than 100 models on the list of
ENERGY STAR qualified products. Low-power mode consumption was the low hanging fruit that
allowed EPA to engage industry enthusiastically in a new voluntary program that had achievable
requirements and demonstrated value to partners and consumers.

1089
Developing a New Approach
Factors Suggesting the Need for Change

ENERGY STAR adds value to a product category by assuring consumers that qualified models are
more energy efficient than alternatives, allowing them to express a preference. The differentiation
allows manufacturers to compete to satisfy consumer demand for efficient products, which will spur
long-term market transformation and maximum energy savings over the long term. To achieve market
differentiation, EPA sets a specification that not all products in the market can meet. Typically,
approximately 25 percent of models will perform at a level sufficient to qualify when the specification is
introduced.
In the beginning of 2003, EPA initiated a cycle of specification revision for imaging equipment. At that
time, the specifications had been in effect for up to seven years and ENERGY STAR qualified
printers, copiers, and fax machines accounted for 92 to 99 percent of units sold in 2000 (Gartner
2001). The high market penetration levels alone suggested that a review of ENERGY STAR
performance specifications was warranted. In addition, revising the specifications allowed EPA to
consider the following:
• Power and/or energy consumed in active modes;
• New technologies and functionalities that had entered the marketplace or become more
prominent (e.g., color, digital, and multifunction capability);
• Consistency in requirements and terminology across imaging equipment and other ENERGY
STAR products;
• Harmonization with other domestic and international organizations; and
• Streamlining of the product development process for manufacturers, as they tend to be
involved in multiple imaging product categories.
Historically, ENERGY STAR and other international, energy-efficiency programs were focused on
power-saving modes, such as Standby. Later, partly due to the success of reducing energy
consumption in low-power modes, increased attention was paid to Active and Ready modes. The
ENERGY STAR Version 4.0 computer monitor specification is the first set of electronics product
criteria to include active mode.

Choosing a New Specification Approach – What Should It Accomplish?

At the outset of the specification revision process, EPA’s consultants found a number of factors that
shaped the approach: field data indicating long default times to low-power modes in many office
equipment products, particularly in copiers and MFDs (Nordman et al. 1998); concerns over low
power-management enabling rates (Roberson et al. 2004; Webber et al. 2001); and the apparent
opportunity to achieve energy savings beyond that available in low-power modes. These factors
prompted EPA to consider a new direction for some imaging equipment products. Information
received from some partners and international stakeholders affirmed the importance of addressing the
Active and Ready mode consumption of certain products.
Rather than extending the existing system to specific criteria for Active and Ready modes, EPA
proposed assessing energy efficiency through a product’s entire duty cycle, by covering all states and
activities. It was intended that the TEC approach would attain the following objectives:
• Relevance and Longevity - The new specification should resume the differentiation the
ENERGY STAR mark brings to the marketplace and the reasonably attainable goal it
provides for manufacturers. The TEC approach also provides a general framework that does
not impede long-term technical innovation.
• Harmonization - The definitions, measurement methods, and criteria levels should be
harmonized with existing international standards and test procedures as much as possible.
• Simplicity - Simplicity in the TEC test procedure makes it less onerous and expensive to
conduct and increases the transparency of the process and results.
• Universality - Products should be tested with a similar method where possible, which should
result in a clearer, more consistent set of specifications across product types.

The Duty Cycle Approach

The TEC test procedure and calculation method cover all states and activities. TEC continues to
provide an incentive to partners to minimize the energy consumption of products in low-power modes
while rewarding equivalent progress in Active and Ready. Rather than amending the existing

1090
specifications by adding energy efficiency criteria for one or more new modes, the duty cycle
approach sums the consumption of all modes. This avoids the need to define multiple power states,
particularly as the combination of modes is not consistent across TEC products.
As energy specifications mature, it seems likely that other product types would benefit from a duty
cycle approach, particularly if the approach can reasonably reflect typical usage. For computers, the
increment of Active energy consumption over Idle levels is quite small; however, the distribution of
time spent in Idle, Sleep, and Off could be incorporated into an energy efficiency approach,
particularly to the degree that capability while in Sleep (especially network connectivity) will drive how
often Sleep is used. Thus, a duty cycle approach could allow for different Sleep capabilities, but
require more Idle time to be included in TEC calculations for products with less network connectivity in
Sleep. As cars begin to consume increasing amounts of energy while idle, particularly for uses such
as air conditioning and electronics, a duty cycle approach could incorporate this in addition to various
driving conditions.

Establishing the Universe of Covered Products

An important, initial step in creation of the TEC approach was deciding which products would benefit
most from this duty-cycle method. Theoretically, all products could be evaluated to TEC. The use of a
single approach for all products could result in more straight forward criteria, and make it easier for
consumers to compare the energy efficiency of various products. However, this was a new approach
that required development of a test method and collection of data, and the testing and reporting is
more burdensome than the existing system. Using TEC is most critical when a significant portion of a
product’s energy demand is in Active/Ready modes and users tend to disable or lengthen time to
power management settings.
Based on this, EPA determined that standard-size copiers, printers, fax machines, multifunction
devices (MFDs), and digital duplicators using electrophotographic (EP), direct thermal, dye
sublimation, solid ink, and thermal transfer marking technologies were best suited to the TEC
approach. These marking technologies use heat-intensive processes in transferring images to the
media, which causes Active and Ready modes to dominate energy consumption and potential
savings. Additionally, field data shows that products using these technologies are more subject to
power-management disabling due to longer recovery times (Nordman et al. 1998).
The TEC structure is designed for the preceding, standard-size products, both monochrome and
color. Small-format products and scanners have different usage patterns, and there is no indication of
a recovery time problem. In addition, these differing usage patterns would require a modified test
procedure. Ink jet products were not considered for the first iteration of TEC as they use little power in
Ready mode and lack long recovery times. Digital duplicators were selected for consideration under
the TEC approach based on high productivity and functional similarities to traditional copiers and
printers.
Much of the rationale for addressing products by TEC applies to large-format EP products as well.
These were not included due to the newness of the approach, the relatively small aggregate energy
consumption involved, and the paucity of models and data in this sector. However, at some future
time, it may be appropriate to bring these products under the TEC framework. For other product types
(most notably the large category of standard-sized inkjet), the additional energy consumption in Active
does not seem large enough to merit the additional complexity of the TEC approach.

Test Procedure Creation

Once EPA had established the set of imaging equipment products to be addressed by TEC, the next
step was to develop a test procedure. The TEC test procedure would specify a series of events to
apply to each piece of equipment as well as the recording of accumulated energy use during each
step. As consultants to EPA, ICF and LBNL would then develop a calculation to apply to the results to
arrive at a figure of energy use over time (expressed in kWh/week).
The primary purpose of the TEC test procedure is to provide a consistent method of measuring and
comparing the relative energy efficiency of similar products. It is important to emphasize that the
procedure is not intended to precisely replicate real-life operating patterns, in part, since this will vary
by country and specific use. In addition, the procedure is not intended to cover all aspects of product
usage, but only those which substantially affect the TEC result.

1091
Review of Existing Test Procedures

ENERGY STAR consultants ICF and LBNL reviewed existing test procedures to identify structure,
principles, and components that could be used in or adapted for the TEC test. ASTM’s “Standard Test
Method for Determining Energy Consumption of Copier and Copier-Duplicating Equipment” provides a
procedure by which copiers, copier-duplicators, accessories, and similar office imaging equipment
may be rated for energy consumption. The TEC test procedure draws from its overall structure and
calculation approach. The International Electrotechnical Commission’s IEC 62301 informed the test
conditions and parameters of the TEC test. International harmonization reduces the testing burden on
manufacturers as well as the time associated with the procedure’s design.

Test Parameters

Next, ICF and LBNL proposed parameters to specify how products should be configured and tested.
Following are a few key parameters from the TEC test procedure and an examination of how EPA
arrived at a conclusion.

Testing in simplex mode

In the initial draft of the procedure, it was specified that the test be performed in duplex mode for
machines that are duplex-capable. Several stakeholders expressed concern that performing the test
in duplex mode could be problematic. Since not all products that fall under the TEC approach have
duplex capability, testing in duplex could mean that products of the same speed would be tested
differently. In addition, limited data were obtained that indicated there were not appreciable
differences in simplex and duplex imaging energy use for current products, so that no meaningful
change in the TEC result was at stake. Allowing simplex output alleviated these concerns and
provided for greater consistency across all products tested under the TEC approach.

Testing in monochrome mode

Color-capable products are to be tested making monochrome images, unless they are incapable of
doing so. EPA proposed monochrome-only testing on the assumption that energy efficiency of units
tested making monochrome images is highly correlated to their efficiency producing color. Some
stakeholders argued that color imaging should be part of the test and would affect the results, both in
general, and specifically, how serial and parallel color printers appear in comparison to each other.
Disagreement persisted over several meetings and it became clear that only empirical data could
resolve the issue. An additional color job was added to a draft of the TEC test procedure, and several
stakeholders submitted data, representing 16 products. Making up the dataset were 12 parallel EP
color printers and four serial EP color printers, with most in the 30-40 images per minute (ipm) speed
range. For the 12 parallel models, the energy consumption for monochrome and color jobs was
virtually identical. For the four serial models, color imaging was notably more energy-intensive than
monochrome. EPA excluded four models from the 16-model dataset that were only instantaneous
power measurements, and ranked the remaining 12 according to the calculated TEC result while
printing in monochrome. These printers were then re-ranked according to their TEC result using only
color imaging. For three of the serial units, the ranking for color printing as compared to monochrome
printing changed slightly. However, a test involving 100% color imaging, as implied through color-only
ranking, is not realistic. For example, one proposed ASTM test procedure for color-capable products
includes approximately equal rates of monochrome and color imaging in the job tables. When the 12-
printer dataset from above was re-ranked with half monochrome and half color imaging, there is only
one very small difference from the monochrome-only ranking, resulting from a serial machine. Thus, a
monochrome-only ranking essentially provides the same result as a mixed ranking.

Testing as-shipped

The product shall be configured as-shipped and recommended for use, particularly for key parameters
such as power-management default-delay times and resolution. Many TEC products have hardware
components that can be added or removed, software settings that can be adjusted by users or service
technicians, or settings that may be determined by incoming print jobs. The procedure’s requirement
for testing “as shipped and recommended” ensures transparency in test results and that users can
readily achieve them in normal use. Testing a product other than as-shipped would offer the

1092
opportunity to “game” the system. This provision reassures manufacturers that other companies are
not submitting skewed data.

Testing to the appropriate market conditions

Products are tested to the appropriate conditions of intended markets. Product testing should be
performed at the relevant market conditions since energy consumption values may vary according to
the input voltage/frequency combination and media type. Testing to market-appropriate voltage and
frequency conditions has been the general procedure for ENERGY STAR testing in the past. As the
ENERGY STAR program develops an increasingly international scope, EPA has determined that it is
important to confirm that products meet the new specification at the representative market conditions
where the products are sold. Parameters of concern are voltage, frequency, media size, and media
basis weight.

Table 1: Testing to Global Conditions

Market Supply Voltage, Frequency Media Size Media Weight
2
North America/Taiwan 115 V, 60 Hz 8.5” x 11” 75 g/m
2
Europe/Australia/New Zealand 230 V, 50 Hz A4 80 g/m
2
Japan 100 V, 50 Hz / 60 Hz A4 64 g/m

Testing in print mode

MFDs are tested in print mode, where possible. Users employ the print function on an MFD more
often than the copy function, and testing both the print and copy functions of an MFD would
complicate and lengthen the testing. If the page rendering process increases consumption, then the
procedure should take that into account. EPA has not seen evidence showing that measuring both
printing and copying would change the results enough to merit the added complexity of the procedure
and calculations. The majority of stakeholders support these conclusions.

Use of a standard test image

EPA did not initially propose a standard test image in the belief that for EP products, any basic image
would require the same energy to produce. Even if very complicated images did use more energy, no
manufacturer would deliberately choose to disadvantage their product during testing by using these.
Some stakeholders felt very strongly about the value of a standard test image, and as there was no
detriment to the procedure’s development or testing burden, this was accommodated using an
existing test image widely used in industry.

Designing the Test

In creating the new TEC test procedure, ICF and LBNL had to assist EPA in determining what actions
the product undergoes, the number of images to be made during active imaging, and how the energy
measurements from the test would be extrapolated to a weekly TEC figure. Throughout all of this, it
was important to address key testing variables that can differ by product speed and/or country/region.
This was especially important, given that the specification will be used globally.

The measurement procedure

The TEC test procedure contains two measurement protocols—one for products assumed not to
utilize an Auto-off function (printers, fax machines, and digital duplicators and MFDs with print
capability), and one for products that do use Auto-off (copiers, and digital duplicators and MFDs
without print capability). For all products, the test pattern consists of measuring:
• Off energy for five minutes or longer;
• Sleep energy for one hour;
• Four, 15-minute “job intervals,” which capture the energy associated with recovery from
Sleep, active imaging, Ready, and possibly Sleep; and
• “Final” energy, which includes energy used from completion of a job interval until the product
reaches its final mode (Sleep or Auto-off).

1093
Figure 1 shows a graphic form of the measurement procedure.

Figure 1: The TEC Measurement Procedure

ICF and LBNL carefully considered what number of jobs was necessary to reliably estimate job
energy while not lengthening the test unnecessarily. The first job incorporates recovery from Sleep
and so requires more energy than the rest. Job 2 is usually greater than job 3 as the thermal
conditions in the fuser have yet to reach a steady state. Examination of early TEC test data made
clear that three jobs would be too few, but that four jobs was sufficient. The average of jobs 2, 3, and
4 is taken as the average for all jobs after job 1 in the calculations.
The job interval is 15 minutes for all products tested under TEC. Some stakeholders suggested that
the job interval should be greater for lower speed products to better reflect Sleep time during the day.
Other stakeholders supported the static 15-minute job interval, noting that for EP products, residual
heat from one job reduces the consumption of successive jobs. EPA decided to retain the 15-minute
job interval in the final TEC procedure because it seemed to be the best single interval to use across
the full range of imaging products.

Defining the job structure

One of the most difficult parts of creating the TEC test procedure was the choice of imaging “job” —
how many originals are presented, how many images of each original are made, and how often a job
is performed. An example job is three images (duplexed) of five originals, every 15 minutes. This
amounts to 15 images per job, 60 per hour, 480 per day, 9,600 per month, and 115,200 per year
(based on eight hours per day of active use, and 20 days of use per month). The number of images
made over a period of time is the “imaging rate.”
The number of images per job is determined by calculations of jobs per day and images per day. The
result reflects the assumption that products with greater imaging speeds typically produce greater
numbers of jobs per day. The calculation of jobs per day was developed in response to stakeholder
comments, which called for the calculated number of jobs per day to increase according to product
speed, generally consistent with the ASTM test for copiers (ASTM 1997). The Job Table numbers in
the TEC test procedure are based on regressions of manufacturers’ monthly rated volumes. EPA took
20% of these figures to be closer to typical usage. In the context of the TEC test procedure, “speed” is
the maximum claimed simplex speed making monochrome images.

Weekly extrapolations

The TEC calculation result could be expressed per day, per week, per month, or per year. EPA
initially proposed a daily TEC result but changed it to weekly in response to strong stakeholder
preference. The energy associated with events in the TEC test procedure is extrapolated to a total
TEC value in kWh per week. The TEC calculations embody two clusters of jobs during the day, with
the unit going to its lowest power mode in between (as during a lunch break), as illustrated in Figure
2. The “lunch” period was added in direct response to international stakeholders who commented that

1094
this slow down time is common. The TEC calculations assume that weekends have no usage and no
manual switching-off is done.

Figure 2: A Typical Day, per TEC Calculations

Figure 2 shows a schematic example of an eight-ipm copier that performs four jobs in morning, four
jobs in afternoon, has two "final" periods and an Auto-off mode for the remainder of the workday and
all of the weekend. The figure is not drawn to scale. Jobs are always 15 minutes apart and in two
clusters. There are always two full “final” periods regardless of the length of these periods. Printers,
digital duplicators and MFDs with print capability, and fax machines use Sleep rather than Auto-off as
the base mode but are otherwise treated the same as copiers.

Establishing TEC Performance Criteria

Data Collection

Once the TEC test procedure was finalized, EPA asked industry to test products and submit the
results for analysis. Stakeholders were given just under four months to complete testing, were
encouraged to test their newest models, and were invited to submit data on products that both could
and could not meet the current ENERGY STAR specifications. In advance, EPA created a data
worksheet to ensure all important data were captured and reported in a consistent format for easy
analysis. Stakeholder participation in testing and data reporting was a critical component to this effort,
as this was a new test procedure and previous data were unavailable.

Initial Conclusions from Data Analysis

As of this paper’s preparation date, ICF and LBNL are in the process of assisting EPA with finalization
of specification. Therefore, the following section on establishing criteria presents the best thinking to
date, although some details of the final energy efficiency criteria may change.

The role of job energy in total TEC

As shown in Figure 3, job energy contributes significantly to the Total TEC. Specifically, job energy
accounts for an increasing percent of total TEC as product speed increases; products with speeds
above 25 ipm always attribute greater than 50 percent of the total TEC to job energy.

1095
Figure 3: Share of TEC Energy Consumption among Key Modes

Use of efficiency formulas

Where possible, EPA attempted to use linear formulas when defining energy efficiency criteria that
consider speed as the determining factor. Many stakeholders expressed a preference for this method,
in particular, to avoid sharp jumps from small changes in product speed when bins of speed ranges
are used. This suggestion is implemented in the Job Table as well as the proposed TEC specification.
The simplest approach to setting a specification line is a linear formula based on product speed. This
works well across large speed ranges, but the imaging equipment specification covers an order of
magnitude in speed and such a large range necessitates more than one single line. At low speeds,
TEC energy is dominated by Sleep/Off energy, which is well correlated with speed. At high speeds,
however, Active energy dominated and is driven by the number of images per week, which varies with
the square of speed. A pair of two linear segments generally seemed adequate for the job, and the
TEC data suggested that their “elbow” should be close to 50 ipm.
Figure 4 shows the TEC data submitted to EPA for monochrome MFDs, contrasting the TEC metric
with product speed in images per minute. As an interim measure, the ENERGY STAR specification
includes an additional allowance for products between 20 and 70 ipm for MFDs. The line was drawn
to ensure a sufficient number of models above and below the line at a variety of speed points, and to
be consistent with the lines for other products (printers, copiers, and color versions of all).

1096
Figure 4: Specification Line: Standard-sized Monochrome MFDs

The effect of parallel requirements

To support other important energy efficiency initiatives, the criteria for ENERGY STAR qualified
imaging equipment will include parallel requirements beyond a target TEC number. As examples,
EPA plans to require standard and optional duplexing capability in various speed segments and will
require that products with an external power supply use one that can meet ENERGY STAR
requirements. As could be expected, these parallel requirements affect which models can meet
ENERGY STAR and have an impact on EPA’s goal to include approximately the top 25 percent of
products on the market at the time the specification is set. To ensure parallel requirements do not
reduce the number of qualified products below the intended level, EPA will consider the number of
products that would fail the parallel requirements before creating formulas to set the TEC criteria.

Engaging Stakeholders
The open participation of industry and other energy-efficiency authorities is crucial to the success of
ENERGY STAR specifications and is comprised of three main components: A) open communication
that ensures everyone involved has equal access to information; B) ensuring that stakeholders’
feedback is considered carefully and regarded in some manner in the specification; and C) providing
sufficient lead time before a specification becomes effective to ensure the levels are attainable. Even
the most refined process will fail if there is a general perception that stakeholder feedback is
disregarded or that timeframes are unreasonable.

Communication

EPA began the imaging equipment specification revision process with an open letter to all interested
parties to explain the upcoming effort and anticipated timeframes. EPA then began meeting with
individual manufacturers to understand concerns about the current specifications and changes they
would like implemented. A Directional Draft (February 10, 2004) preceded a more official first draft
specification and identified objectives, summarized thinking to date, proposed a general specification
framework, presented comments received and responses, shared a timeline, and invited further input.
The Directional Draft contained many placeholders and was a unique opportunity for stakeholders to
comment at the very early stages of the process. The Directional Draft also contained definitions and
terminology. The definitions and terminology associated with TEC were circulated early and often for
feedback to establish a common language stakeholders could use when sharing additional feedback.

1097
Since the release of the Directional Draft, EPA has distributed for comment three drafts of the
specification; six drafts of the TEC test procedure; summaries of and responses to all comments
received; numerous interim updates, rationale, and discussion documents; and all data sets upon
which conclusions have been drawn. To further ensure this process was transparent and
collaborative, EPA has made all of the abovementioned documents available on the ENERGY STAR
Product Development Web site at www.energystar.gov/productdevelopment, which is updated
regularly.
EPA also gathered invaluable feedback from stakeholders during many meetings held in the US,
Europe, and Asia. These meetings provided a unique opportunity to work through issues in an open
forum, and for participants to hear the opinions of other stakeholders.

Incorporating feedback

EPA’s goal for the TEC approach was to develop a test procedure that allowed for the relative energy
efficiency of imaging equipment to be measured and compared in a precise, repeatable way, and to
create a specification that recognized approximately the top 25 percent of the market while fairly
accounting for the increased energy required of higher-functionality products. Industry representatives
and international program implementers know their products and markets better than anyone else,
and their comments throughout this process contributed to a quality result.
EPA attempted to accommodate all comments that would lead to the best possible specification. As
an example, requests for small changes or additions that would not affect the outcome were not
deliberated extensively in an effort for simplicity. However, other comments did not align with
ENERGY STAR guiding principles and could not be accommodated. Perhaps the most difficult
comments to resolve were those that conflicted with other feedback received, or those that presented
plausible changes whose impact could not be understood immediately. In addressing these last two
categories of comments, EPA attempted to obtain empirical data to support the final decision. This
ensured that all feedback was investigated carefully, and that ultimate decisions were easily
understood by all. The issue of monochrome versus color imaging presented earlier in this document
is an example of where EPA consulted test data to inform a decision.

Sufficient Transition Time

As product specifications come up for revision, EPA is committed to accommodating production

cycles via establishing reasonable effective dates for new requirements. EPA strives to allow a
minimum of nine months transition time between the final specification’s publication and the effective
date.

Conclusions
In summary, the time was right for EPA to address Active mode consumption in imaging equipment
and TEC provided a flexible framework in which EPA could consider Active while achieving other
important goals. The process was methodical and done in collaboration with stakeholders, and should
lead to significantly more energy efficient imaging equipment products. This paper has presented a
number principles that were critical throughout the TEC development process. They include:
• Use of empirical data to drive key decisions;
• Embracing simplicity over strict “correctness” in many cases;
• The need for transparency, in process and result; and
• Cautions, deliberative decisions over time to produce a quality result.

References
[1] ASTM. 1997. Annual Book of ASTM Standards, "F757, Standard Test Method for Determining
Energy Consumption of Copiers and Copier Duplicating Equipment", American Society for
Testing and Materials, Philadelphia, PA.
[2] Better Buys for Business. Retrieved January 23, 2003. “Multifunctional Machines Explained.”
from www.betterbuys.com/FSR/wp2.FDML.
[3] Better Buys for Business. February 1999. The Mid/High-Volume Multifunctional Guide. Guide
No. 110.

1098
[4] Better Buys for Business. October 2001. The High-Volume Copier & Multifunctional Guide.
Guide No. 137.
[5] Buyer’s Laboratory, Inc. Fall 2003. Multifunction Device Specification Guide.
[6] Buyer’s Laboratory, Inc. Winter 2004. Printer (non-impact) Specification Guide,
[7] Buyers Laboratory Inc. Spring 2002. “Multifunctional Specification Guide – Copier-Based
Products.”
[8] Consumer Reports. 2003. “Printers: Buying advice.” Consumer Reports.org. Retrieved
January 16, 2003 from www.consumerreports.org.
[9] ENERGY STAR Product Development Web site.
http://www.energystar.gov/index.cfm?fuseaction=find_a_product.
[10] Hershberg, Craig. 2005, July 11. ENERGY STAR Qualified Imaging Equipment Typical
Electricity Consumption (TEC) Test Procedure. EPA ENERGY STAR Office Equipment
program, Washington DC.
[11] Hershberg, Craig. 2004, Jan 12. ENERGY STAR Qualified Imaging Equipment Specification
Revision: Directional Draft. EPA ENERGY STAR Office Equipment program, Washington DC.
[12] International Electrotechnical Commission, IEC 62301 Ed 1: Measurement of Standby Power,
IEC TC 59, TC59, Working Group 9 59/297/CD Household Electrical Appliances.
[13] Nordman, Bruce, Mary Ann Piette, Brian Pon, and Kris Kinney. 1998. "It's Midnight...Is your
Copier On?: ENERGY STAR Copier Performance." Berkeley, CA: Lawrence Berkeley National
Laboratory. Report No. LBNL-41332.
[14] Roberson, Judy A.; Carrie A. Webber, Marla C. McWhinney, Richard E. Brown, Margaret J.
Pinckard, and John F. Busch. 2004. After-hours Power Status of Office Equipment and
Inventory of Miscellaneous Plug-Load Equipment. LBNL-53729-Revised, May. Lawrence
Berkeley National Laboratory, Berkeley CA.
[15] Webber, Carrie A., Roberson, Judy A., Brown Richard E., Payne, Christopher T., Nordman,
Bruce, Koomey, Jonathan G. 2001. Field Surveys of Office Equipment Operation Patterns.
LBNL-46930. September. Lawrence Berkeley National Lab, Berkeley CA.

1099
1100
Set Top Boxes, Televisions
and Consumer Electronics

1101
1102
The Digital TV Challenge, Sharing Knowledge Between China and
the EU
Li Aizhen1, Matthew Armishaw2, Bob Harrison2
1
China Standard Certification Center,
2
UK Market Transformation Programme

Abstract
Both China and the European Union face significant environmental and technological challenges from
the growth of digital television but are at different stages of development - the EU has a mature yet
growing market, whilst China has comparatively recently embarked on a programme of digital roll-out.
There is potential for the EU to share its experience of building a digital television infrastructure and
for China to play a leading role in developing products with improved environmental performance. As
1
part of that process, the UK Government's Market Transformation Programme (UK MTP) is
2
supporting knowledge exchange activities, in the EU-China Energy Environment Programme (EEP)
aimed at developing a common approach to specifying products, which do less harm to the
environment
In October 2005, a delegation from the principal Chinese standards and testing institutions,
accompanied by some of the leading Chinese set-top box manufacturers, visited the UK under the
auspices of the EEP to study the EU approach to the development of common test, performance, and
interoperability standards for digital TV Set Top Boxes (STBs) and to share an insight into the
Chinese adapter (basic STB) market. This paper will outline the background and objectives for that
visit, and report on the progress that has been made since then in harmonising criteria for energy
efficient STB development. The paper may also encourage similar activities in other product areas
and therefore may be of interest to anyone seeking to explore collaborative activities with non-EU
3
nations, especially those involved in the Marrakech Process .

Introduction
It is universally accepted that the introduction of digital TV reception platforms, particularly set-top
boxes (STBs), has the potential to cause a step increase in the overall energy demand of domestic
electronic products in Europe and in China.
European countries have varying schedules to stop analogue broadcasts completely over the next 10
years and China has ambitious plans for 130 million STBs to be in Chinese homes by the time of the
2008 Olympics. Evidence is already available, from products actually brought to the European
horizontal and vertical markets, to show that the energy efficiency of these STBs could, today, be
improved by 20% (on average) if best practice is applied, conversely the average energy consumption
could increase to more than 50% above best practice if left to market forces.
Although China does have a programme for introducing appliance efficiency standards, digital TV
STBs are not a priority and without external influence, it could take several years before the Chinese
authorities address this issue. In the meantime, when the European Commission and European
Member State governments discuss improved energy efficiency of products with major procurers, the
typical response is that product efficiency is not in their hands but in the hands of the manufacturers
who are predominantly Chinese. Neither side appears to be presenting proactive solutions to change
the current position.
For this reason EEP intends to act as a catalyst that will bring together Chinese manufacturers with
their European counterparts and with major European equipment purchasers with the common
purpose of improving the efficiency of STBs and many other energy using products in volume
production.
EEP organised the highly targeted study tour to the UK for representatives of China’s STB
manufacturing industry, the China Standard Certification Centre and two relevant research

1
http://www.mtprog.com/
2
http://www.eep.org.cn/
3
http://www.un.org/esa/sustdev/sdissues/consumption/Marrakech/conprod10Yconcl.htm

1103
institutions. The purpose of the visit was to engage Chinese stakeholders with their European
counterparts and with large equipment buyers in Europe so that in the future the efficiency of the STB
products imported to Europe would be optimised. This would improve the opportunities for export for
the Chinese manufacturers and mitigate the projected increase in domestic energy demand in Europe
and China that the wide-scale introduction of STBs is expected to promote.
During the Study Tour, the delegation visited several bodies influencing the STB product including
manufacturers, test laboratories and major service providers. Experts from these organizations
shared many valuable experiences on STB design, production, volume procurement, testing and
deployment to a volume domestic market.

Objectives of Study Tour

A primary purpose of the Study Tour was to adequately prepare the Chinese manufactures and other
stakeholders to contribute to the following meetings in Seoul Korea:
st
ƒ International Harmonisation of Set Top Box Specifications, 31 October
ƒ Special meeting of the European Code of Conduct for STBs working group, to engage
4
International participants 31st October
nd th 5
ƒ IEA “Global cooperation on One Watt” (standby power) Conference, 2 to 4 November
The Seoul events intended to start a process for the STB product that would create a universal
method of testing and would "put in one document” a harmonized set of energy efficiency
performance requirements. This would build on discussions between California, Australia, the UK
MTP the US EPA (Energy Star) and the EU Code of Conduct working group on STB. It would also
address an imperative of the leading organizations involved in developing energy efficiency standards
in Europe, notably European Commission DG JRC and the International Energy Agency underlining
the importance of bringing Chinese Manufacturers to Europe and catalyzing their direct involvement in
International standards harmonization work.

Outcome of Study Tour

The STB Study Tour to Europe provided significant help to Chinese STB manufacturers in identifying
marketing opportunities in Europe in the context of the energy performance expectations for this
globally traded product. These manufacturers recognized that the primary message on sustainable
STB design obtained from the tour harmonized with the related requirements for the rest of the world.
With the preparation afforded by the Study Tour the Chinese delegation took an active part in the
discussions at the Seoul event, understood what the International Community wanted to buy in terms
of efficient products and obtained valuable guidelines to their manufacturing and testing.
Most importantly the study tour was the catalyst for the publication of a new set of energy efficiency
criteria for basic digital TV STB cable platforms for the Chinese market that have been drafted in close
cooperation between the CECP and the major Chinese manufacturers. (Appendix 1) The energy
efficient design and manufacturing standards for these products will impact on all STB products
sourced in China.

Overview of Study Tour Components

Advanced Silicon Manufacture.
The general purpose of this component of the tour was to review the very latest Silicon developments
for Energy Efficient STB platforms for Terrestrial, Satellite, Cable and Internet TV services.

Outcome and observations: With the rapid development of Silicon technology, the STB can realize
multiple functions in single device architecture. Each function can be software programmed to match
the requirement of specifiers (for example the device could be updated by software to provide MPEG4
decoding and HD decoding standard from an original MPEG2 decoding and SD encoding
specification without recourse to new Silicon design and fabrication). This allows the rapid
development of new functions on existing silicon and allows remarkable levels of power management.
Figure 1 provides a schematic overview of this Silicon architecture
The silicon functions are enhanced in speed, and energy efficiency by the progressive transition from
90nm to 32nm technology. Usually, continually evolving functionality, adversely affects the minimum

4
http://www.energyrating.gov.au/pubs/2005-stb-program.pdf
5
http://www.standbyforum.co.kr/and http://www.action1watt.com/

1104
power requirement of STBs and the start up time from low (standby) power states. This new Silicon
technology has readily programmed power management architecture. This can achieve very low
minimum (standby) power levels and yet allow effective control of the transition between the low
power standby state and the fully active on mode, for the whole system architecture, to guarantee full
function within 2 or 3 seconds. The Integration of Silicon design group know-how ensures that lessons
learned from mobile phone power management migrates to Silicon for STBs, and continuously
improves the potential of this power management. The power requirement of this new silicon concept
is further reduced by a dramatic improvement in DC to DC conversion efficiency for the supporting
power supply source Silicon and power distribution regulators. Typical conversion efficiency of up to
95% is now commercially viable in the volume production of low cost products.

Major Digital Television Interactive Service System Provider

The purpose of this component of the study tour was to discuss the practical problems of
broadcasting digital TV interactive services to a range of millions of set top boxes covering six years
of technological change. Sky are noted for their long term support of the European Code of Conduct
for STBs and discussions were also scheduled to cover the specification and procurement of energy
efficient STBs to meet Code of Conduct criteria in the context of challenging commercial financial
imperatives. Other topics for discussion were to include, the implications on energy requirement of the
new generation of High Definition Television STBs and energy efficient solutions to multiple access
Satellite TV (SMATV). The latter is a common solution to Satellite TV reception for Chinese homes.

Outcomes and Observations: The over–arching criterion for China or any emerging digital
broadcasting market where there is no well structured dominant Service Provider must be that the
policy makers who influence the major market (Government or Union of States / Authorities) must give
firm leadership, to drive agreements on common standards for DTV Broadcasting and Networking and
for the related reception platforms (STBs –IDTVs) This will catalyse the volume production of products
that are interoperable and that meet good minimum standards of, technical performance, usability and
energy efficiency.
STB power consumption targets must be realistic and technically possible in the time frame of a
Service Providers commercial roadmap When planning such targets, policy makers, manufacturers,
and TV service providers should be involved in cooperative discussion for an agreed set of criteria
and an agreed implementation time frame. Common minimum standards that can be applied to large
volume production runs of STBs will reduce costs and allow a higher investment in energy efficient
design without compromising the commercial viability of the product in a competitive procurement
market or in a retail market.
When specifying the functional and network requirements of an STB, the Service Provider must
consider not only the cost, but also the ease of use and attraction of the features from the consumer
viewpoint. This is particularly important in the pay to view market where there is often no competition
in the type of STB available.
In BskyB’s specification for the competitive procurement of STB’s, each potential supplier is required
to produce an STB that meets a common performance and physical specification and that must
comply with the relevant energy efficiency criteria of the European Code of Conduct. The technical
solution for this compliance is left to the design ingenuity of the manufacturer. This encourages
innovation in technical solutions in a commercially viable cost framework.
The functions and ergonomic design of the remote control for an STB can influence the users ability
to put the STB, the TV set and other connected devices into standby. In China consideration should
be given to developing an industry standard for remote controls that combines good ergonomic
design with the large energy saving cost benefit to the user and Government of ensuring that all
peripheral products are not left on by mistake when the TV is switched off. A common Industry
Standard for the remote control has the further benefit of significant reductions in volume production
cost. All 10 million BskyB remote controls distributed since 1998 are to a common design philosophy
and specification regardless of the suppliers of the STBs. Independent testing bodies regard the
design as one of the most ergonomically successful (user friendly) ever produced and it should be
studied as the template for Chinese volume production industry standards.
BskyB have experimented with some Subscription Service Integrated Digital TVs (IDTVs) for
distribution through the retail market. In their opinion the IDTV is unlikely to ever be a suitable
reception platform for subscription services because of the difficulty of protecting the encryption

1105
techniques of CA in a remote cam module. Even if this could be resolved, the manufacturing and
subsequent retail cost of the IDTV is comparatively very high because software and middleware
specification variations between Service Providers make large production runs of one IDTV chassis
type impossible. The resultant impact on manufacturing cost makes the product uncompetitive in its
class.
Without a step change in Silicon architecture, HDTV STBs will introduce much higher energy
consumption. In current Silicon architecture (now up to 10 years old) particular problems exist with HD
tuner power requirement and MPEG-4 processing. Some development time will be needed to improve
the energy efficiency of HDTV STBs’ and in discussion it was noted that the European Code of
Conduct working group is regarding the impact of HDTV as a priority issue for International
stakeholders.
There is a potential for significant power management advances in a new generation of HDTV STBs.
The new HDMI interconnection standard for these platforms can recognize and control the required
power state of peripheral devices connected directly or wirelessly. Sky recognizes that International
standardization work is urgently required to agree common HDMI power management protocol
applications and is steering a high-level working group on other HDMI connector issues to support
this,

British Broadcasting Corporation and Digital Television Group (DTG) Testing Centre

The general objectives of this component of the study tour included, discussions of UK experiences in
the development of free to view terrestrial digital television services and of maintaining the
interoperability and usability of the set top boxes, low cost converters and IDTVs supplied through the
horizontal market for the reception of those services. In the latter context demonstrations and
discussions were scheduled in the DTG Testing Centre

Outcomes and observations: The STB products should conform to at least a minimum specification
and applications delivered to that product in the broadcast data stream should be authored within
strict specifications to ensure long-term interoperability of all installed STB platforms. Ideally published
guidelines for broadcasters and domestic equipment manufacturers should be put in place by a
commercially neutral organization (e.g. DTG, DVB etc,) or regulatory authority working with all
stakeholders. The guidelines and related detailed specifications should be continuously reviewed in
the light of technical and functional innovation.
A test suite and test programme development that can determine if a product conforms to the
guidelines and specification should be implemented by a central body, which is supported by industry,
standards organizations, broadcasters, and service providers. It was considered that the DTG Testing
Centre was a good example of this. In this context it is considered essential that Chinese
manufacturers have access to appropriate test suite equipment and test programmes to ensure that
products meet the technical and interoperability standards of the targeted market. In China the
CEPERI certification body should perhaps consider providing such facilities.

Specialist Consumer Electronic Product Testing Centre

In this component of the study tour discussions on the assessment of, STB ergonomics (usability)
technical performance, operational (subjective) performance and energy efficiency, were planned
around demonstrations of elements of complete test programmes.

Outcomes and Observations: When designing better products, user trials should be launched, and the
ease of use of products should be assessed through an experienced panel of users under the
guidance of a test programme devised and continuously reviewed by an ergonomic expert who is
familiar with the product genre. The object is to ensure that the STB is easy to connect to the TV and
other equipment and has good visual and aural interactivity to support a reliable tuning and
installation routine. This is particularly important in simple low cost products that may be more
commonly used by unskilled disabled and older consumers.
For energy efficiency STB products, both ergonomic testing and user trials should be implemented to
assess the ease of operation of STBs featuring power management. These tests and trials should
ensure that start up delays from standby are not excessive and that the control of the standby state of
the STB, the connected TV and other peripherals is user friendly in achieving low power states.

1106
Discussions led the tour group to related considerations for China. Once minimum energy efficiency,
technical, operational and usability standards are established, the Government Standards bodies
should arrange the continuous testing and approval of all STB products in the market place. Strict test
programmes run by independent test houses should be used. Ideally conforming products should be
given a publicized approval logo. For Government or provincial procurement, only the most energy
efficient approved products should be short-listed for purchase.
Intertek bases its terrestrial digital TV STB tests on the detailed specifications in the ”D-Book”
published by the combined Industry and Broadcasting advisory body in the UK, the DTG. Testing of
Satellite and cable equipment is based on guidelines agreed with equipment manufacturers and
Service Providers.
For STB energy efficiency testing purposes, IEC62087 is the main reference standard and provides a
good guide to general test conditions, simple operating mode definitions and test set up procedures
for simple STBs,
However Many experts, including those at Intertek are warning that the advent of more complex
platforms with multi media, and home networking capabilities will drive the need to urgently review this
standard. In the opinion of Intertek experts, the revised standard should avoid the stricture of
attempting to qualify performance criteria, functionality and usage patterns It was noted that
International cooperation on issues associated with this standard was high on the agenda in the STB
forum associated with the International Standby Power Conference in Korea November 2005.
Chinese manufacturing and product testing experts should support this work.

Leading Manufacturer of Digital TV Reception Platforms

The purpose of this element of the tour included discussions on the key impediments to designing and
manufacturing commercially viable energy efficient STBs in an intensively competitive world market,
the potential impact on energy efficiency of the next generation of high functional specification multi
media home platforms, energy efficiency testing methodologies and the implications for existing
standards.

Outcomes and Observations: STBs are acquiring new advanced functions such as video-on-demand
multi-room networking of separate entertainment streams, complex interactivity and universal
broadcast data access, including the Internet. These developments are driven, mainly by subscription
Service Providers to retain or attract end users. Because end users rarely make purchase decisions
on energy efficiency considerations, these rarely feature in the current commercial development
roadmap of most STBs unless the manufacturer drives that requirement.
Power management, automatically driving the lowest power level for a given functional requirement is
the key to achieving energy savings or mitigating energy requirement increases in the next generation
of high functional specification multi media home platforms. Mature evidence shows that most users
cannot be educated to reliably perform manual power management functions. Surveys have shown
that even in simple STBs, over 50% of users don’t activate the standby mode if it requires a
secondary remote control activity after the TV has been put into standby. So even in these basic
products a simple form of power management such as an auto standby function is desirable.
Continuously revised and internationally agreed test specifications and methodologies are important
for consistent testing results and criteria conformance testing. Pace bases its testing on nationally and
internationally agreed standards and regulations for all types of digital broadcasting and platform
technical criteria. It develops its own standards - compatible test material and very sophisticated
broadcast data stream generating and analysis equipment.
Energy efficiency improvement requires the cooperation of all parties involved in the platform design,
hardware designers, silicon vendors, service providers, software designers, external software
suppliers, CA suppliers, LNB suppliers, the consumers, and government agencies. Pace put
significant effort into communicating with these parties and contributing to many relevant International
standards and national working groups such as the European Code of Conduct working group and
the UK DTG.

Challenges to China Prompted by STB Tour Observations.

At present, there is no coordinated broadcasting and telecom service development plan, or even
cooperation on the subject between the media regulator of China, SARFT, which also oversees cable

1107
television operators and MII, the telecom regulator of China. MII wants to promote an IP-based digital
TV service that would be distributed though the telecommunications companies. This lack of high
level coordinated policy for multi media communication contrasts badly with the European situation
and inhibits long term investment in major Service Systems and related reception platform
manufacturing.
This is compounded by delays in finalising the national DTV standards, STB related standards, and
Card Separate-from TV Receiver standard. These uncertainties make the STB manufacturers
reluctant to progress research and development of new products. With the exception of some large
companies, most manufacturers prefer to wait-and-see. Besides these political issues there are
many key factors to prevent STB production moving to the large-scale production that will allow
China’s Industry to drive forward to a new generation of high specification energy efficient yet
commercially competitive world benchmark products. Observations on the Chinese situation prompted
by this study tour highlight the following problems.
STBs are restricted in design and can’t readily benefit from the economy of production volume. They
are being tied to the limited cost driven requirements of technically uncoordinated and loosely
regulated, cable companies. This situation is aggravated by the fact that the consumer can’t obtain a
better cable STB in the competitive free market and so drive up quality through competition.
Various cable companies monopolise very segmented markets and do not cooperate commercially or
on technical standards. There are many differences on both interface software and Conditional
Access systems. Manufacturing for such an uncoordinated set of market standards weakens the
ability of the manufacturing Industry to excel in core technologies. (This contrasts very badly with the
situation in Europe)
There is often no requirement to develop core technology in OEM procurement contracts from foreign
companies, further weakening manufacturing expertise and innovation.
Lack of the essential foundation of independent working groups equally involving all stakeholders to
drive common standards and testing / conformance regimes, will hold back the excellent potential of
digital TV broadcasting and equipment manufacturing in China.
All these factors will bring uncertainty to STB industry development but may be mitigated by three
factors. Firstly the last 5 years of industry awareness of the long term potential of these products has
fostered an expansion of the expertise resource. Secondly, there will be a positive impact from the
administration system reform that has allowed deepening and vigorous promotion of communication
resources by local government. Thirdly the pull of high quality broadcasts from the 2008 Olympic
Games, will be a good driver for digital TV start-up and will prompt great growth in a short period.

Conclusions and Actions

Under the challenging conditions outlined above it is simply not realistic to expect the Chinese
Manufacturing industry to drive autonomously, excellence and energy efficiency in the STB product.
Many energy-saving opportunities in STBs will be missed if there is no coordination of the many
different players in the technology, communications network, and broadcast industries. Energy
efficiency improvement requires the cooperation of all parties involved, hardware designers, silicon
vendors, service providers, software designers, external software suppliers, CA suppliers, LNB
suppliers, the consumers, and Government agencies. With this cooperation there is the market
potential to easily drive China to the forefront of benchmark STB production. The following actions will
be a catalyst for this key objective.
„ Energy efficiency agencies must provide proactive advice on Energy Efficient STB criteria to
policymakers, like NDRC, MII, SARFT, and local government and support their coordinated
involvement in energy efficient STB procurement or distribution.
„ China must participate in the process of common international standard development and
support those standards through internal market application. Relevant market transformation
measures should be explored.
„ Mechanisms must be established to cooperate with and advise service providers to
coordinate design requirements and thus foster energy efficiency innovation in STBs through
the R&D investment made possible by strong commercial road maps based on continuous
volume production.
„ Hardware and software designers must be given a market environment where they can
achieve radical innovation in power management and cost-effective smart network
interfaces. This innovation combined with international standards cooperation will ensure that

1108
 STBs from Chinese industry will achieve benchmark energy efficiency performance for any
functional state.

Figure 1: New Silicon Architecture allowing software programmable functionality of transport

stream decode and output encode formats for fast to market Energy Efficient STB design.

1109
Appendix 1.

CSC
Technical Specification for Energy Conservation Product Certification for
Digital to Analog Set-top Boxes for General Digital Cable Transmissions
Preface
This Technical Specification is one of the series of the technical Specifications for Energy
Conservation Products. It is especially established in order to, carry through and implement: the Law
of the People's Republic of China on Conserving Energy, energy conserving product certification,
protecting the environment, improving the energy efficiency of digital to analog set-top boxes for
general digital cable transmissions, and driving the technical energy conversation advancement of
Corporations.
This Technical Specification is formulated according to the current conditions of the manufacturing
and use of digital to analog set-top boxes for general digital cable transmissions in China and by
reference to similar foreign technical specifications.
This Technical Specification is proposed by and under the jurisdiction of China Standard Certification
Center (CSC).
And in association with the main drafting units of the Standards: CSC, CNIS, China CEPREI
Laboratory, National Computer Testing Center, Huawei Technology Co., Ltd, Skyworth Group
Company Ltd, Shenzhen Tongzhou Electronic Co., Ltd
Main draftsmen of the Standards: Li Aizhen, Zhang Guoqin, Song Danmei, Zhou Xinghua, Qiao Mu,
Xie Yongmin, Xu Bin.

Technical Specification for Energy Conservation Product Certification for Digital to Analog
Set-top Boxes for General Digital Cable Transmissions

1. Scope
1.1 This specification defines the energy conservation evaluation and testing methods for digital to
analogue set-top boxes for general digital cable transmissions (set-top boxes for short in the following
text) which turn digital signals into analogue signals.
1.2 This specification is applicable only to the basic set-top boxes that turn digital signals of general
cable transmissions into analogue signals.
1.3 This specification isn’t applicable to those set-top boxes that have additional functions such as
interactive return path communication functions.

2. Standard documents quoted

The clauses in the following documents will, through quotation in this specification, become the
clauses of the specification. For all the quoted dated documents any subsequent modification lists
(including the content of erratum) or any modified version shall not be applicable to this specification.
However, the parties that have reached agreements based on the specification are encouraged to
investigate and determine whether later versions of those documents are applicable. For all the
undated documents quoted, their latest versions are still applicable to the specification.
GB/T17975.1-2000

3, Definitions
For the purpose of this specification, the definitions listed below apply.
3.1 On mode
The state in which the set-top box is connected to the mains power source and is performing the main
function.
3.2 Passive standby mode
The set-top box is connected to the mains power source and performing no function other than
maintaining the functions required to switch to on mode from a remote control or local control signal.
3.3 Automatic switch time
The time that elapses before set-top boxes are automatically switched from on mode into passive
standby mode..
3.4 Evaluating values of energy conservation

1110
The technical target is for evaluating the energy conversation performance of set-top boxes. It
includes the power consumption in the state of on mode, the power consumption in the state of
passive standby mode, and the required automatic switch time.

4 Technical requirements
The set-top boxes shall meet the requirements of 4.1, 4.2 and 4.3 at the same time.
4.1 The power consumption in the state of on mode
The power consumption of the set-top boxes in the state of on mode shall not exceed 9 watts.
4.2 The power consumption in the state of passive standby mode
The power consumption in the state of passive standby mode shall not exceed 1 watt.
4.3 The automatic switch time
The time taken by the set-top box to switch automatically from the state of on mode into the state of
passive standby mode, after the last user instruction has been received from remote or local controls,
shall not exceed 4 hours.

(NOTE: If the set top box has received no remote or local control instructions for a period of time
approaching four hours, indicating no user presence or a long programme it should provide a visual
and aural warning that switch off is imminent. If the user does not respond to this message within two
minutes, the device will switch automatically into passive standby mode. The auto switch off function
should be capable of being disabled by the user for recording purposes but should automatically reset
after the user designated recording period)

Test methods
5. Test conditions
5.1.1 Ambient conditions
Ambient temperature: (23±5)ºC
Relative humidity: 45%-75%
Atmospheric pressure: (86-106) kPa.
5.1.2 Power supply
Voltage: 220V(AC)
Frequency: 50Hz
The test sequences shall be carried out at the alternating current mains with steady voltage. The
mains fluctuating of both voltage and frequency shall not exceed ±2%.
5.1.3 Equipment and requirement of the test
A wattmeter (or watt-hour meter) shall be used for all power requirement measurement. The testing
accuracy of the wattmeter shall be more than or equal to 0.01W.
5.1.4 Input signals
5.1.4.1 MPEG-2 standard active image order transport steam shall be provided in accordance with
with GB/T17975.1-2000. The images should be continuously looped without a stop and the basic
cycle is 60s.
5.1.4.2 High / Standard definition testing signals
Color stripe signals

5.1.4.3 Format of image signals

number of Line Field
lines ratio Amplitude
Format of signals scanned frequency（ frequency
alternately ratio
lines kHz） （Hz）
720×576I 2：1 625 15.625 50 4：3

5.1.4.4 Audio testing signals

stereo 1kHz 0dBfs sine-wave signals

1111
Testing methods

5.2.1 Standard testing signal parameters

Bit Rate 2Mbps-15Mbps
Symbol Rate 3.52-6.89MS/s 64QAM
database package size: 188/204byte
modulate mode 64QAM 256QAM
Reeling & Interweave Depth I=12
Rolling & Dropping coefficient a=0.15
Frequency 110-862MHz
Standard Input Voltage 60dBµV
Yawp Equivalent Bandwidth Signal Occupation Bandwidth/ 1+a
Audio output signals of the tested equipment at the maximum.
5.2.2 The power consumption in the state of on mode

Figure 1: The schematic interconnection of test equipment and test sample

All the testing equipments and the tested samples shall be connected according to Fig. 1. The
MPEG-2 transport steam signals generator produces standard active image order，and the whole
testing system shall be set at the standard testing signal parameters, the tested equipments shall be
adjusted to have a natural image in the display.
The test sample shall work continuously for 30 minutes before testing.
The period of measurement for the power consumption of the tested samples shall not be less than
10 minutes.
5.2.3 The automatic switch time
After the test according to 5.2.2, the sample shall be switched to standby and then to on mode No
further operations of the remote controls or local controls should be performed. The time taken from
the start of on mode (set top box producing images and sound) to the point of automatically switching
to standby should be recorded. A note should be made of the time of appearance and the duration of
the switch off visual warning.
5.2.4 The power consumption in the state of passive standby mode
According to 5.2.2, after the test sample switches into passive standby mode and achieves a steady
state, the power requirement should be monitored over a period of not less than 10 minutes. The
Final steady state reading should be recorded. If the power requirement fluctuates in the standby
mode an integrating (watt- hour) meter must be used for a measurement period of 10 minutes from
the onset of a regular fluctuation pattern.

Calculating method for stated power requirement when measuring with an integrating (watt –
hour) power meter.
The power consumption per unit time =
The indicated power consumption (watt-hours) / The duration of the measurement period (hours)
The power consumption per unit time is expressed in Watts (W), to an accuracy of 0.1w.

1112
Australian Mandatory Standards for Consumer Electronic
Equipment
Keith Jones, Paul Ryan, Melissa Damnics

Energy Consult (Australia), Electronics Industry Consulting Service (Australia)

Abstract
In late 2002 the Ministerial Council on Energy in Australia launched a 10-year strategy to deal with
excessive standby. As part of this strategy Australia plans to regulate the maximum in-use and
standby power consumption of digital Set Top Boxes (STB) from October, 2007. In Australia, it is
planned to phase out analog TV from 2010 and an estimated 14 million TVs will require a STB.
Projections of the energy use for these STBs show energy consumption could increase to over 2,000
GWh pa by 2012. Minimum Energy Performance Standards (MEPS) could reduce the business as
usual energy consumption by STBs to as little as 300GWh pa.
STBs were initially identified for voluntary targets for standby power consumption, however industry
groups requested the government consider mandatory requirements. Further, the regulation of in-use
power consumption was identified as a priority due to the large number of STBs that will be potentially
left in this mode.
Australia is also moving to regulate the standby power of a group of consumer electronic equipment
covering the home entertainment products. The proposed regulation may provide potential energy
savings of 1.7 GWh pa and greenhouse emission reductions of 1.7 Mt CO2-e pa by 2020. Home
theatre systems, DVD players/recorders, portable and integrated sound systems are some of the
products included in the broad scope of the proposed requirements.
The paper explores the design of the MEPS, lessons learned from the Australian experience, energy
impacts, test method issues and projected impacts of the proposed regulation

Introduction
In late 2002 the Ministerial Council on Energy in Australia launched a 10-year strategy to deal with
excessive standby. Consumer Electronic equipment was initially identified with voluntary targets for
standby power consumption. STBs used for the conversion of Digital TV broadcasts were identified as
a priority product. Other Audio Video equipment was also identified as needing action on MEPS at the
earliest time possible.
These targets were initially intended to be voluntary, however industry groups requested the
government consider mandatory requirements. The reason for this will be discussed later. Further,
the regulation of in-use power consumption was identified as a priority due to the large number of
STBs that will be potentially left in this mode.
This paper will exam the factors surrounding the decisions associated with the development of these
MEPS. STBs will be considered separately from the wider category of consumer AV equipment due to
some unique circumstances surrounding their penetration into the market and the significant impact
that this has on the energy consumption associated with these products. Australia’s intention for
MEPS on consumer equipment will then be discussed along with the potential impact that these
products will have on energy consumption in Australia. Finally the paper will consider lessons learned
from the development of these MEPS and how this knowledge may be used to assist the
development of future MEPS under consideration.

The Development of Mandatory MEPS for Digital TV Set Top Boxes(STBs)

Background
st
On the 1 January 2001 free to air digital television broadcasts were started in order to replace analog
broadcasts by 2008. This target has recently been under review and the likely start date for the initial
turn off of analog broadcasts is now 2010. Under the Australian broadcasting regime broadcasters
were required to simulcast the same material on the digital channels as they were broadcasting on
the analog channels. In addition they were also required to broadcast by the end of 2003 no less than
20 hours per week of high definition programs.

1113
The consequence of these new broadcasts was the introduction of STBs into the Australian consumer
electronic market. There are 7.6 Million households in Australia(1) and the estimated the number of
TVs per household is around 2 (2) with an expectation that this figure could rise to around 3 by 2012.
If switch over to digital were to be achieved by 2012 this could mean, at worse case for energy
consumption, that some 20 million STBs would be purchased.
Televisions with inbuilt digital tuners have now been introduced into the Australian market and over
time most televisions in the Australian Market will have a digital tuner. In Australia it is estimated that
around 80% of televisions sold annually are replacement televisions. For this reason not all
televisions will need a STB to receive a digital signal. We are still left, however, with the potential for
some 14 Million STBs being needed to fully convert all televisions in Australia to Digital.
In fact the take up of digital television has been less than impressive. To date estimated take up of
DTV is about 17% of households or 1.3M STBs(6). Many models have been developed to forecast
the take up. All have underlying assumptions Fig 1 shows one that was developed by one of the
Authors of this paper which was presented at the 1 Watt conference in Seoul and subsequently
supplied as evidence to an Australian Parliamentary committee review into the take up of DTV(3).
This model forecasts the energy use of STBs under the proposed MEPS as approximately 1,000
GWh pa by 2012.

COMPARISON OF ENERGY CONSUMPTION BETWEEN

DTV AND STB
FOR AUSTRALIAN DTV CONVERSION

TV INCLUDING INBUILT TUNER IN STANDBY STB IN STANDBY PLUS TV WITHOUT DIGITAL TUNER 50% STB LEFT ON PLUS TV WITHOUT DIGITAL TUNER
1,000
900 ASSUMPTIONS
5% GROWTH FOR TV ANNUALLY
800
100% CONVERSION TO DIGITAL BY 2012
700 MEPS LEVELS IN STANDBY 8W ON COMPLIANCE WITH MEPS FROM 2007
NEW TV SALES WILL SHIFT RAPIDLY TO WIDESCREEN
600
500
400
300
200
100
0
2005 2006 2007 2008 2009 2010 2011 2012

Page 7

Figure 1: Energy Consumption Model for STB use due to DTV take up in Australia

An important development in the forecasting of DTV take up is the Parliamentary Committee review
entitled “Digital Television: Who’s buying it” which conducted hearing late in 2005 to establish the
cause of slow take up and to produce recommendations to improve the take up of DTV. This
committee has released 11 recommendations that address the poor consumer proposition for DTV
which in their view will allow a turn off of Analog signals by 2010. One of the recommendations is the
adoption of MEPS for STBs which is a clear indication that in their view the energy consumption of
STBs is an important consideration in the take up of DTV in Australia.

Set Top Box MEPS Voluntary of Mandatory

The proposed MEPS regime for STBs was to be a voluntary scheme. However, when industry was
consulted there was concern raised that a voluntary scheme may not produce the outcomes that the
scheme was designed for. Many suppliers reported that as a matter of cooperate policy they would
comply with official standards whether it was voluntary or not. This had the potential to put them at a
disadvantage compared to companies that did not have such policies. In general it was pointed out
that the companies with such policies were the larger more established brand names such as Sony,
Panasonic, LG and Samsung. In addition these suppliers pointed out that the structure of the market

1114
in Australia meant that there were a large number of suppliers in Australia with few having a large
market share (See Fig 2).

100%

90% 19
26 28
34 33 32 34 33 32 35 34
35 36 38
80% OTHERS
5
2 5 TOSHIBA
70% 6 2
5 TOPFIELD
2 1 7 2 1 2 2
2 2 1 2 THOMSON
6 6 2 6 6 6 2 2 6
60% 7 5 6
2 2 2 1 5 2 TEAC
2 1 2 6
23 15 2 21
13 12 STRONG
50% 16 15 13 12 13 14
19 13 10 SAMSUNG
3 5 5 3
2 3 3 4 6 5 PANASONIC
40% 3 7
7 6 7 7 7 7
6 3 7 7 6 6 6 LG
7 5 5 4
30% 7 6 4 5 7 9 6 LEGEND
7 6 5 10
6 8 6
KAON
9 7 12 11 11 9 9 5 9
20% 10 8 10 DGTEC
8 1
12 4
10 4 9 7 7 7 8 4 7
5 8 9 2
10% 5 4 4
4 3 4 4 4 3 14 3
5 2 11
4 6 7 7 7 6 6 5 5 6 5 6
0%
Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Jan MAT MAT
05 05 05 05 05 05 05 05 05 05 05 06 05 06

Figure 2: STB Brand shares in Australia (4)

From figure 2 it is evident that 66% of the market is supplied by 11 suppliers. The other 34% is
actually supplied by at least 21 identifiable other suppliers. Many of these other suppliers, and indeed
a number of the 11 larger suppliers, are in fact traders who source existing product from various OEM
manufacturers. They have little input or influence into the design of the product and would have a
strong propensity not to follow a policy of voluntary compliance with MEPS.
After much discussion the two main electronic industry associations (AEEMA and CESA) agreed to a
mandatory regime for MEPS in order to maintain a level playing field and ensure that the desired
outcomes of reducing STB energy consumption. This is not to say that industry would probably have
preferred MEPS not be implemented at all. It was a position that if it were to be implemented then a
fair system would have to be mandatory compliance. A similar position was also held for the yet to be
developed MEPS for TV.

Set Top Box MEPS and Energy Saving

As has already been discussed a realistic figure for energy consumption for STBs even if MEPS is
implemented could be as high as 1000GWh by 2012. If MEPS were not implemented, however, the
situation would be considerably worse.
It is now commonly accepted that many STBs are never switched into passive standby mode. If we
make the following assumptions:
• 50% of STBs are never turned off.
• 50% of STBs are used 4 hours per day
• On Power is an average of 23W (7)(Panasonic AVC Networks (Aust) 2005 Survey of STBs)
• Passive Standby is an average of 7W (Panasonic AVC Networks (Aust) 2005 Survey of
STBs)
• 14 Million STBs sold to convert to Digital.
Then the annual energy consumption would be 2000GWh pa. If we introduce MEPS based around
the specifications in tables 1 & 2. We can reduce this energy consumption by half (see Fig 1). If we
add to the MEPS a High Energy Saving a mechanism to ensure the STBs are switched into standby
when not in use the power consumption is reduced by another 70% to 300GWh pa.
It is crucial, therefore, that an effective MEPS regime is established for Australia.

1115
Set Top Box MEPS the process for Implementation

Announcement and Consultation.

The intention to develop a MEPS regime for STBs was announced October 2004. This was followed
by a period of consultation with all stakeholders.. There were several meetings held at which the
scope of the STB MEPS was discussed and feedback was given by various stakeholders. When there
was general agreement on the nature and scope of the STB MEPS it proceeded to the standards
development phase.

Standards Development
The organization for Standards Development and Publishing in Australia is Standards Australia.
Within standards Australia a Committee known as TE1 is responsible for following and where
appropriate implementing the work of IEC committee TC108. Essentially TE1 is concerned with the
safety of household appliances but its work has now been extended to include energy efficiency for
products.
For the development of STB MEPS a working group was formed under TE1 and throughout 2005 met
several times to develop and write the STB MEPS standard. This standard was then released in draft
form for public comment. At the conclusion of the public comment phase a few minor editorial
changes were recommended and the document was then processed for ballot.
It should be noted that some changes were made to the actual committee processes early in 2006.
There was concern expressed by a number of members of TE1 that being safety regulators they may
not have the appropriate expertise to examine and vote on the standard. To correct this, a constituted
committee was formed under TE1 known as TE1-008. The work of the working group was transferred
to this committee early in 2006. Being a constituted committee its members are able to vote on the
document unlike the working group structure which relied on TE1 members for voting.

Regulatory Impact Statement (RIS)

Before any regulation can be applied to ensure the specification within the MEPS standard is met it is
necessary to conduct a Regulatory Impact Study. The aim of this study is to determine the extent of
the problem, what policy approaches were considered and the costs and benefits of the preferred
option (typically Mandatory MEPS regulation). If the RIS is favorable then regulations will be enacted
in each state of Australia to ensure product complies with the MEPS Standard.

The Technical Specification and Scope of the STB MEPS standard

The proposed regulation covers STBs suitable for free-to-air(FTA) broadcast TV and subscription
(pay) TV (STV). Maximum power levels for MEPS are based on the existing requirements used by the
voluntary agreement provided under the European Code of Conduct (CoC) and the mandatory
requirements for digital television adapters in California. The Australian MEPS is tailored to mirror
international requirements, while being moderated to address local industry technical issues. Detailed
consultation was conducted with the local industry and specific requirements were developed to
provide for Australian subscription TV services and high definition broadcasts.
The proposed MEPS includes requirements for passive standby, active standby and in-use modes,
separate requirements for standard definition and high definition STBs as well as free-to-air and
subscription TV services. The maximum power levels for MEPS are based on the power consumption
of a basic platform configuration. The MEPS for a particular configuration of STB is made up of this
maximum power level and an allowance for additional features. Finally, the total allowable power
consumption for a STB is not to exceed a specified amount, regardless of the number and type of
features that are included in the STB. The proposed regulation also includes recognition for those
devices that automatically switch to passive standby after 4 hours of non-use and no user activity or
that utilise a HDMI connection.

The basic requirement for energy consumption is shown in Table 1 and Table 2.

1116
 Table 1: Maximum Power Levels for Standard Definition STBs (From AC supply)
Product Type Passive Standby – Active Standby – On mode – Max
Max Power (W) Max Power (W) Power (W)
MPA/MPL MPL/MPL
Proposed Compliance Date 1 October 2007
STB – FTA Either Option 1 1.0W 8W /15W 8W /15W
or Option 2 2.0W 7W /15W 7W /15W
STB – STV Not applicable 9W /15W Not applicable

Table 2: Maximum Power Levels for High Definition STBs (From AC supply)
Product Type Passive Standby – Active Standby – On mode – Max
Max Power (W) Max Power (W) Power (W)
MPA/MPL MPA/MPL
Proposed Compliance Date 1 October 2007
STB – FTA Either Option 1 1.0W 12W /19W 15W /22W
or Option 2 2.0W 11W /19W 14W /22W
STB – STV Not applicable 13W /19W Not applicable

In addition to the basic allowance MPA the standard allows for additional power consumption
depending on functionality. The extra allowance however must not exceed the Maximum Power
Levels shown in Tables 1 & 2.
The Additional Power Allowances are shown in Table 3.

Table 3: Additional Power Consumption Allowance

Feature Additional power consumption Additional power
(Active Standby Mode) consumption (On Mode:
STB – FTA Only)
SCART Port 1.0W 1.0W
IEEE1394 interface 0.8W 0.8W
Ethernet interface 0.4W 0.4W
100 Mbit
Wireless interface 0.7W 0.7W
SPDIF port 0.1W 0.1W
Serial USB 0.3W 0.3W
interface
Home automation 0.4W 0.4W
interface
Broadband modem 2.0W 2.0W
Extra cable modem 0.7W Not applicable
Extra LNB Not applicable Not applicable
Additional tuner 2.0W 2.0W
Powered remote IR 0.25W Not applicable
receiver

The development of Mandatory MEPS for Home Entertainment Equipment

Background
The government is considering a report that proposes a mandatory maximum standby power (5). The
proposed regulation may provide potential energy savings of 1.7 GWh pa and greenhouse emission
reductions of 1.7 Mt CO2-e pa by 2020 in standby power alone. Over 4 million of these products are
sold annually in Australia and the sales of home entertainment equipment are increasing rapidly.
Home theatre systems, DVD players/recorders, portable and integrated sound systems are some of
the products included in the broad scope of the proposed requirements. It is intended to develop a
horizontal requirement for standby power that will apply to the entire group of products. In this way,
emerging new product categories will be included in the coverage of the proposed regulation without
the need to continually update the definitions.

1117
Depending on the product group, maximum passive standby levels of between 4 and 6 watts are
proposed by 2008 with 1 watt being required for all home entertainment products by 2012. These
levels are based on international standby programs and aim eventually to meet the IEA long term
target. In addition given that recording product such as DVD recorders and PDRs are emerging with
Digital Tuners consideration is being given to treating these in the same way as STBs and Specifying
“On Power” standards as well.
Australia’s standby strategy is part of the IEA plan for standby, which encourages all member nations
to address standby in a coordinated manner, and achieve a common long-term target for 2012 – “1
Watt” maximum for all low power modes. The IEA standby target however relies upon the user to
switch the product off or into standby mode. Evidence from Australian surveys suggests that many
products are not switched into this mode and remain in active standby for extended period of time.
The proposal includes the requirement that the device automatically switches to passive standby after
30 minutes of non-use and no user activity.
Some products already include the auto power down feature as suppliers respond to the challenge
set by various international and national standby programs. The impact of this requirement represents
approximately 75% of the projected impacts and the benefit of auto power down will be more
significant as the proliferation of such devices increases.

Why Develop Mandatory MEPS for Home Entertainment Equipment

Increasing Sales Volume

Home Theatre Equipment is a rapidly expanding product group. As the potential impact of a MEPS
policy is highly dependent on the sales of the products, actual and forecast sales by product group
were developed. Based on actual data from 2000 to 2004, projections were made for the various
categories of equipment, as shown in Fig 3
The sales of home entertainment equipment are related, as some products replace or substitute the
functions of other products, such as the decrease in VCRs which are being replaced by DVD&HD
Recorders. Also, it is likely that sales of DVD Players will be plateau and decrease as both Home
Theatre Systems and DVD&HD Recorders increase. In 2005, it is estimated that approximately 4
million home entertainment products will be sold in Australia.
The sales projection model does not yet consider the impact of product convergence, such as
including DTV tuners in recording product. As more product emerges in this form, it is likely to impact
on the sale of simple STBs. The inclusion of DTV tuners in these products will also further increase
the sales of DVD recorders and Personal Video Recorders (PVR). The savings that have been
predicted as a result of the STB MEPS could be diluted as more sales of STBs are substituted by
PVR/DVD products were at present only standby power is considered. For this reason consideration
is being given to treating recording product with a DTV tuner as a special class of product with the in-
use power also subject to limits. In fact from an energy consumption view this class of product should
be encouraged because it will reduce the actual number of pieces of equipment within the household
that are consuming standby power.

1118
 2 500 000

2 000 000 DVD Players Sales

DVD&HD Recorders

AV Receiver Sales

1 500 000 Integrated Stereos Sales

Sales

Home Theatre Sales

Portable Stereos Sales

1 000 000
Audio Components Sales

Speakers Sales

VCR Sales
500 000

-
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010

Year

Figure 3: Annual Sales Forecast for Home Entertainment Equipment

Potential for Greenhouse Gas Reductions

By implementing MEPS on standby power for home entertainment products it has been estimated
(Ref) that CO2 emissions will be reduced by 970KT pa and 1,200 GWh of electricity will also be saved
by 2020 See Fig 4 and 5. The greatest contribution to this saving is the saving associated with
passive standby power. It is estimated that the contribution to the saving by standby power will be
800MWh pa by 2020.
1.8

BAU GHG Emissions MEPS GHG Emissions

1.6

1.4

1.2
Mt CO2-e/yr

1.0

0.8

0.6

0.4

0.2

0.0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Year

Figure 4: BAU vs. MEPS Policy – GHG Emissions for Home Entertainment Equipment

1119
 Total MEPS Standby Energy Savings

1 400

1 200
Total Passive Standby Savings
Total Active Standby Savings
1 000
Energy Savings (GWh/yr)

800

600

400

200

-
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020
Year

Figure 5: MEPS Potential Energy Savings by Standby Mode

The average standby power consumption by mode for these products is shown in Table 4, with
significant power consumption in active standby mode.

Table 4: Summary of NAEEEC Store Survey Average Measurements 2004/05

Home Entertainment Mean Active Mean Passive Mean Off
Product Standby Standby Mode
AV Receivers 41.2 1.3 0.2
Home Theatre Systems 31.5 2.2 0.1
Integrated Stereos 18.0 4.6 2.4
Portable Stereos 6.7 2.1 1.9
Audio Components 9.0 3.0 0.0
Sub Woofers & Speakers 9.7 4.7 0.6
DVD Players 9.2 2.4 0.0
DVD Recorders 23.0 7.5 NA
Hard Disk Recorders 29.4 7.2 0.0
VCR’s 7.7 2.9 NA
Total for all Hom 19.2 3.5 0.5
Entertainment Products

Table 5: Summary of NAEEEC Store Survey Measurement Ranges 2004/05

Home Entertainment Product Active Standby Passive Standby Off Mode
Max Min Max Min Max Min
Av Receivers 112.3 11.1 6.4 0.2 0.7 0.0
Home Theatre Systems 55.7 12.1 18.1 0.2 0.9 0.0
Integrated Stereos 52.2 5.1 34.9 0.2 4.8 0.0
Portable Stereos 20.5 2.8 5.1 0.7 3.5 0.0
Audio Components 15.8 4.2 5.6 0.4 0.0 0.0
Sub Woofers & Speakers 20.4 0.5 17.8 0.1 10.2 0.0
DVD Players 21.5 4.4 9.8 0.1 0.0 0.0
DVD Recorders 36.5 15.9 24.4 2.8 NA NA
Hard Disk Recorders 36.0 20.8 20.1 2.9 0.0 0.0
VCR’s 11.3 6.1 6.3 1.2 NA NA
Total for all Home Entertainment 112.7 0.5 34.9 0.1 10.2 0.0
Products

1120
These tables demonstrate a large variation in passive standby energy consumption across a range of
products and within product categories. The largest differential was found to be AV receivers in active
standby which varied by over 100W. When analyzed against price no relationship was found for price
vs energy performance indicating that poor performers could be improved without a significant effect
on cost.
It is clear from this discussion the proposed MEPS on standby power for home entertainment product
is necessary given the clear proliferation of sales of these type of products. This category of product
will significantly add to both the future energy requirements and Greenhouse Gas emissions.
It is also obvious from the field surveys that energy efficiency improvements can be made to product
at little or no cost; and that much product already meets or exceeds the requirements imposed by the
MEPS

Consistency with international trends

It is necessary when considering a MEPS regime for Australia that the requirements and levels are
consistent with international thinking. It is clear that, although implementations of schemes may differ,
many other countries have considered these products as having the potential to provide energy
savings. The Australian MEPS regime is entirely consistent with other schemes, as shown in Table 6.

Table 6: Brief Overview of Energy Efficiency Programs

Voluntary Programs Mandatory Programs
Energy Star

GEEA

EICTA

HomeSpeed

Nordic Swan

Energy Boy

IEA 1W

CEC

Top Runner
Program
Type Label Label Agreement Database Label Label Target MEPS Target
AV Receivers
Home Theatre Systems
Integrated Stereo
Portable Stereo
Stereo Components
Sub Woofers
Speakers
DVD Players
DVD Recorders
Hard Disk Recorders
VCRs

These international programs establish a consistency in the approach Australia is taking and what
other countries have done. This would indicate that the Australian requirements are achievable
without any cost impact or limiting available product.

Proposed mandatory MEPS for Home Entertainment Equipment

A broad “horizontal” definition of products is proposed to be targeted by the MEPS and aim to simplify
the issues associated with determining MEPS coverage. The definition proposed is based on the
earlier definition, as follows:
Commercial available consumer equipment that produces, records or assists in producing an
audio or video signal/output.
Products that are specifically covered by other MEPS requirements (such as TVs and set top boxes
and possibly video recording equipment with DTV tuners ), would be excluded from this MEPS.

Stage 1 MEPS - 2008

The MEPS for implementation in 2008 is suggested at 4 watts for all home entertainment products
without video recording capabilities and at 6 watts for those with video recording capabilities.
Additionally all products with an off mode would be required to have consumption less than 0.3W.

1121
Approximately 30% of all models surveyed in 2005 would not meet this MEPS level as demonstrated
in Table 7.

Table 7: 2005 Home Entertainment Products excluded by proposed MEPS Stage 1

Product Passive Passive Off
Standby Standby Mode
≥4W ≥6W ≥0.3W
Av Receiver No of Models 2 - 3
% of Models 4% - 6%
DVD Player No of Models 18 - 0
% of Models 27% - 0%
Home Theatre No of Models 8 - 1
System % of Models 15% - 2%
Stereo No of Models 2 - 0
Component % of Models 25% - 0%
Integrated No of Models 21 - 1
Stereo % of Models 32% - 1%
Portable No of Models 2 - 6
Stereo % of Models 10% - 14%
Sub Woofer No of Models 4 - 7
and Powered % of Models 15% - 13%
Speakers
DVD No of Models - 17 Na1
Recorders % of Models - 40% Na1
Hard Disk No of Models - 10 0
Recorders % of Models - 53% 0%
VCRs No of Models - 1 Na1
% of Models - 4% Na1
Total of all Home Entertainment 106 - 33
Total % of all Home Entertainment 29% - 7%
1
No products were found with off mode

A MEPS level that removes the 30% of the least efficient models surveyed is consistent with the
method used to determine the MEPS levels for other products, such as three-phase air conditioners
and motors, while still somewhat consistent a significant trading partners

Stage 2 MEPS – 2012

It is anticipated that by 2012 the market would be ready to achieve a MEPS level of 1 watt or less in
passive standby for home entertainment equipment. This would be consistent with the levels
expected by international programs. To reduce the impact of active standby all home entertainment
products would be required to power down to passive standby within 30 minutes of inactivity. Video
recording equipment would be exempt from this due to the requirement for equipment to be preset for
future recording. This level would currently exclude 70% of the market. However, given that the EU
agreement requires manufactures to meet the 1 Watt level by 2007, it is not unrealistic to assume the
Australian market will be able to adjust by 2012.

Conclusions
Australia after consultation with industry has adopted a mandatory approach to MEPS for Set Top
Boxes and Home Entertainment Equipment. In the case of STBs the MEPS includes limits on both the
standby and on mode operation. In the case of home entertainment the MEPS are based on standby
power except for the consideration of a specification for on mode in the case of recording devices with
DTV tuners.
The mandatory regime that will be imposed is with agreement with industry and will ensure a
consistent and fair application of the specification.
The MEPS are considered necessary as they will result in significant reductions in CO2 emissions any
energy consumption associated with the use of this equipment.

1122
Care has been taken to ensure that the proposed MEPS are achievable with current and emerging
technology without dramatic cost or supply implications to suppliers. Indeed it has been found in the
studies leading to these MEPS that equipment already exists that meets and even exceeds the limits
specified.
The MEPS covers product categories that are consistent with international programs and the levels
adopted are also consistent with international energy reduction programs.
The road ahead will require ongoing consultation with stakeholders to ensure these standards are
implemented appropriately to provide the energy and greenhouse gas savings.

References
[1] House of Representatives Standing Committee on Communications, Information, Technology
and the Arts. Digital Television – Who’s Buying It? Enquiry into the uptake of Digital Television
in Australia , Commonwealth of Australia, February 2006 Page 29.
[2] ABS 2005, Environmental Issues: People’s Views and Practices, Australian Bureau of Statistics
, catalogue number ABS 4602.0 29 November 2005. Available from www.abs.gov.
[3] House of Representatives Standing Committee on Communications, Information, Technology
and the Arts. Digital Television – Who’s Buying It? Enquiry into the uptake of Digital Television
in Australia , Commonwealth of Australia, February 2006 Page 174.
[4] GFK. Digital Set Top Box Report , January 2006 .
[5] EnergyConsult Pty Ltd. Analysis of the Standby Energy Efficiency Improvements to: Home
Entertainment Equipment , December 2005 .
[6] Digital Broadcasting Australia. www.dba.org.au Members News Letter, January 2006 .
[7] Panasonic AVC Networks (Aust) Survey of STBs 2005
[7] Standards Australia, Draft AS62087 Parts 1 & 2, Public Comment March 2006

1123
1124
The EU Codes of Conduct: What Have They Achieved and What are
the Challenges?
Paolo Bertoldi1, Bob Harrison2, Hans Paul Siderius3, Ken Dale4, Michael Jäkel5
1
European Commission
2
UK Market Transformation Programme
3
SenterNovem
4
ce Micro Technology plc
5
Abakus

Abstract
There is no doubt that developments in digital TV and broadband communication will have a large
impact on residential energy consumption if no adequate policy actions are implemented. The
European Union has established a successful stakeholder forum through its Codes of Conduct for
Energy Efficiency. This forum agrees on power levels for defined operating modes, and provides
further guidelines, e.g. on power management, to achieve desirable efficiency levels.
Two Codes of Conduct are in operation: the Code of Conduct for Digital TV Service Systems (since
2001) and the Code of Conduct for External Power Supplies (since 2003). In 2005, the first draft of a
Code of Conduct for Broadband Equipment was presented at the stakeholder forum. Together, it is
projected that these Codes of Conduct will save 20 TWh per year from 2010.
This paper reports on the first results for the Codes of Conduct for Digital TV Service Systems and
Power Supplies, based on data from participating manufacturers. Analysis of the data shows a
considerable overall improvement in power consumption, not only in the standby modes but also in
the on-mode. Furthermore, the paper shows how discussion with stakeholders based on a common
road map provides guidelines for the continuous development of this innovative policy tool and leads
to the efficient adoption of the new criteria required to mitigate the energy impact of the new features
and functions in the product that are prompted by rapid technological development and market forces.

The Need for the Code of Conduct (CoC) Policy Tool

The market penetration and energy impact of domestic electronic products and appliances is often
subject to step changes driven by new technology. Commercial secrecy about product development
and marketing often makes the accurate prediction of these changes complicated, if not impossible
without the direct engagement and support of the key stakeholders involved in production and
marketing. This engagement is often resisted, or is subject to protracted negotiation, where
mandatory policy tools such as energy labeling and minimum energy performance standards (MEPS)
are introduced because perceived by manufacturers as imposed on them [2]. In addition, at least in
the European context, the legislative process for MEPS and mandatory labels is rather slow, and once
legislation is finally adopted it will take a while for changing it. This is why for fast changing
technologies such as computers and digital set top boxes (STBs), there is a preference in Europe for
voluntary approaches (Energy Star labeling for office equipment, and CoC for STB). As discussed
later in the STB CoC the power level, definitions and allowances needs to be reviewed almost every
year, and these specs can never be set established more than two years in advance (we do not know
how technology, demand and services will evolve), and as well the specs can not last more than 2
years. This rules out the use of MEPS in the European context.
In 1997, a European Commission working group identified the digital service system STB as the
domestic electronic device with the largest potential to increase the energy requirement of European
households. To challenge and resolve this problem, the European Commission set up a working
group of the key stakeholders in digital service system development – STB designers, STB
1
manufacturers, component providers (e.g. Silicon , LNB, Tuners), service providers – and energy
agency specialists. The continuous threat of introducing MEPS was a strong driver to bring key
manufacturers to participate in the CoC development and in the end of sign it.

1
Throughout the text the authors use the term "Silicon" to describe the electronic components within the set-top box –
components, since the principal functional blocks are embedded in LSI silicon chips.

1125
Research into proposed development showed that by 2010, the STB could push domestic electronic
energy consumption in the Europe Union (EU) above that of refrigerators and freezers [3]. With 150
million of these boxes across the EU - equivalent to one per household – the annual electricity
requirement for digital service systems with full functionality and poor power management could be
around 60 TWh (close to the total electricity consumption of Denmark for all sectors). The fast
penetration will be driven by the announced phased put of the analogue signal in 2009. Generation of
this electricity would also release 24 MtCO2, which would have a significant impact on the EU’s ability
to meet its overall Kyoto CO2 reduction target. This increased electricity consumption could be cost
effectively be halved if polices and measures will be in place.
At the same time, in 1997, work also began on external power supplies, which at the time had high no
load losses and very low on mode efficiency. External power supplies were predicted to penetrate the
market due to many new uses (mobile computing, mobile telephones, consumer electronics, etc.). For
this product, the fast technology change was not the main driver to introduce a Code of Conduct as
for STBs. As with many other products in Europe there is a strong preference for voluntary and
negotiated approaches to improve efficiency of end-use equipment [2]. For external power supplies
the voluntary agreement approach was investigated; however, because too many different
2
stakeholders were involved, included OEMs, it was decided to establish a Code of Conduct . The
savings potential for external power supplies was evaluated to be in the order of 10 TWh for the EU,
to be achieved by 2010 if all external power supplies would meet the CoC.

The Basic Mechanism of the Code of Conduct Policy Tool

The new rationale of the CoC policy tool is to proactively engage stakeholders, at the earliest possible
stage of a product’s commercial roadmap, in a voluntary scheme, to mitigate the energy impact of the
product, without stifling product development and commercial objectives. This activity has become an
excellent example of a product policy initiative that has united stakeholders early enough to impact the
design process before the product became ubiquitous.
The following stages can be identified, and they will be discussed in the next sections:
• Stage 1: Identify priority products and set up working groups
• Stage 2: Improved energy efficiency criteria and CoC roadmap
• Stage 3: Achieved outcome, agreed with stakeholders
• Stage 4: Continuous review to identify best practices

Stage 1: Identify Priority Products and Set Up Working Groups

In Stage 1 of the Code of Conduct mechanism, products with a potentially large market and high
energy impact are identified by the European Commission through reports and papers prepared by
expert consultants [4,2]. Products with the highest potential to impact on energy are prioritized, and
approaches are made to key stakeholders to contribute, on a voluntary basis, in working group
discussions on practicable options to mitigate predicted energy impacts. The membership of these
working groups requires adroit planning and negotiation:
• The working group must involve the main manufacturers of a product genre. Manufacturers’
representatives should be in a position to act as a conduit to senior levels of, product design
and marketing.
• The interest and participation of major procurers, such as Service Providers, is essential,
since their endorsement of CoC objectives has a significant influence on product
manufacturers. For example, the early involvement of major procurers of external power
supplies in the Power Supply and Charger CoC Working Group (mobile phone manufacturers)
resulted in the fast uptake of relatively stringent criteria through the use of efficient switching
power supplies.
•
3
A balance of independent experts and industry experts (e.g. from the Silicon Industry) is
required to evaluate the factors qualifying product energy efficiency performance, identify best
practice and define practicable criteria objectives to mitigate the energy impact of the product
in the marketing roadmap. Ideally some of these experts should have involvement in related

2
The difference between voluntary agreements, negotiated agreement and Code of Conducts in described in [2], and in a
forth coming publication by the same author.
3
Throughout the text the authors use the term "silicon" to describe the semiconductor industry components in the product –
since the principal functional blocks are embedded in LSI silicon chips.

1126
 technical working groups of standards bodies and industry association, in order to ensure a
common approach to product testing that supports the discussions in the CoC working group.
• Political representation is important, to ensure that both national and EU-wide energy-
efficiency objectives are fully considered; to provide practical support with CoC working group
research and testing commitments through national agencies; and to catalyze endorsement,
procurement and fiscal support schemes for products that meet the CoC criteria.

Stage 2: Improved Energy -Efficiency Criteria and CoC Roadmap

The principal aim of a working group is to reduce the energy consumption of a product through the
setting of agreed, practicable power requirement targets in a defined development timescale. To that
end, a voluntary Code of Conduct is devised, and Europe’s principal stakeholders in the product
genre are encouraged to support the agreement.
4
The Signatories of the Code of Conduct will make all reasonable efforts to:
• achieve the power consumption targets set for new products placed on the market after an
agreed date, based on an agreed test method;
• support and contribute to the development and acceptance of new criteria based on the
commercially practicable application of “best practice” technology;
• co-operate with the European Commission and Member State authorities in an annual review
of the scope of the Code of Conduct and the power consumption targets for two years ahead;
• facilitate and encourage consumers to adopt energy-efficient practices in connection with the
use of services involving the product;
• co-operate with the European Commission and Member States in monitoring the
effectiveness of the Code of Conduct; and
• ensure that procurement specifications for related services, systems, equipment and
components are compliant with the Code of Conduct.

Stage 3: Achieved Outcome, Agreed with Stakeholders

The achieved and committed objectives of the CoC in a product area are disseminated to a global
audience to promote common international energy efficiency criteria for the product genre. It is
important to notice that for the products covered by the two CoCs, there is a large degree of imports
into the EU (almost 90% fro external power supplies), and also a large similarity with products
commercialized in other economic regions. Therefore there is a strong interest to have similar energy
efficiency polices and programmes in other region of the world, to assumer more compliance to the
CoC levels. To this end Europe has been among the leaders in promoting STB and external power
supplies polices since 1997, and to harmonization of test methods and specifications.

Stage 4. Identified Best Practice for CoC Reviews

This aspect of the CoC mechanism is of particular importance. Expert input to the CoC working group
must continuously review the potential for innovation in the technology to mitigate the energy
demands resulting from increased functionality in the product. On a regular basis, relevant industry
and independent experts are asked to review power requirement criteria for the basic product and for
the power demands of the increased functionality dictated by product marketing.
In this context, the CoC for Digital TV identified a key tool for the achievement of significant
energy efficiency targets in digital service system platforms - the development of effective power
management in the silicon for the principal functional blocks.
The CoC power management task force, originally formed and supported by the UK Market
Transformation Programme, has created a dynamic Silicon development “roadmap” which helps to
qualify future revisions of the power requirement targets in the Code of Conduct. A range of power
requirements in the standby passive, standby active and on modes are assembled for each functional
block or group of blocks for which power can be managed. These values, qualified by the likely uptake
time of the new Silicon or hybrid component (e.g. tuner) are used to agree on the timing and value of
criteria updates. The task force relies on the cooperation of platform designers and silicon suppliers in
the construction and continuous updating of this roadmap of the functional blocks.

4
The list of the Signatories and the Codes of Conduct can be seen at the website of the EU Stand-by Initiative, the European
Actions to Improve Energy Efficiency of Electrical Equipment while either Off or in Standby, at
http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm

1127
A similar approach was also followed with manufacturers of external power supplies, where through
discussions with all stakeholders the no load consumption has been dramatically reduced.

What Has Been Achieved by the Existing Codes of Conduct?

Conformance Results for the Code of Conduct for Digital TV Service Systems

The conformance of set top box products from signatories to the CoC for Digital TV service systems
for the year ending December 2005 was presented to the March 2006 CoC working group. An
overview of the results for 2005 shows declarations from four key signatory manufacturers. Also the
data supplied by a major service provider was included. Conformance criteria for 57 products were
returned. They comprised, 33 TVs with integrated receiver-decoder, 11 stand alone Satellite set top
boxes, 7 stand alone Terrestrial set top boxes and 6 stand alone Cable set top boxes. 90% of these
products complied with the CoC criteria. These products represent a major share of products for the
vertical market (service providers specifying the box functionality to receive their services and in many
cases providing the boxes to their clients), this because of the active engagement of the major service
providers. In additional many other manufacturers and service providers in the EU market
produce/specify products CoC compliant without having signed the CoC and/or reporting the
consumption of their models.

Figure 1: 2005 Declared Conformance Results (averages)

The average power requirements for the on mode and standby active mode for Terrestrial and
Satellite STBs in Figure 1 show the impact of the trend toward Personal Video Recorder (PVR)
functions in these products. Although the power requirement of the added functional characteristics in
general meet the CoC’s 15W criterion for maximum standby active power , the power management
task force has prioritized actions to examine power management options for the hard disc, DVD R/W
and additional tuner/decoder functions of these products to achieve a better energy efficiency in the
standby active mode.
Figure 2 shows the impact of revised criteria on the energy efficiency of STBs in Europe and
underlines the importance of the impact of power management in the context of a massive increase in
platform functionality.

1128
Figure 2: Influence of the CoC on Set-Top Box Power Requirement 2001-2005
(Average Power Requirement – All types STBs)

Conformance Results for the Code of Conduct for Energy Efficiency of External Power
Supplies

The conformance of External Power Supplies from signatories of the CoC for the year ending
December 2005 was presented to the March 2006 meeting of the working group. It should be noted
that signatories to this CoC commit themselves to:
• Design power supplies or components so as to minimize energy consumption of external
power supplies. Those companies who are not responsible for the production of power
supplies shall include the concept of minimization of energy consumption in their purchasing
procedures of power supplies.
• Achieve both the no-load power consumption and on-mode efficiency targets shown (see
Tables 4 and 5) within the time schedule for at least 80% of products for Phase 1, and 90% of
products for Phase 2, for the new models of external power supplies that are introduced on
the market after the indicated date.
For these criteria, the on-mode efficiency is measured at 100% load (i.e. full rated output
current) or declared as the simple arithmetic average of efficiency measurements made at 25%, 50%,
75% and 100% of full rated output current.”
st
An overview of the declared results for the period ending December 31 2005 showed that eight key
signatory manufacturers reported on 130 power supply models. Of these models 92% conformed with
the CoC criteria. Figures 3 and 4 below summarize the conformance details and show the number of
products already meeting the next criteria phase. The coverage is very high in external power
supplies for mobile telephones (the major mobile telephone manufacturers have signed the Code of
Conduct and reported), and notebook computers (again major manufacturers have signed, together
with major power supplies manufacturers). There are also other OEMs (mainly ITC equipment), and
power supply manufacturers that participate in the stakeholder meeting, but have not signed the CoC.
These manufacturers claim that they are meeting the CoC levels.

1129
Figure 3: No Load Power Conformance

Figure 4: Efficiency Conformance

As indicated, the External Power Supplies CoC is rather successful in reducing energy consumption
in mobile telephones (almost 90% of the market) and office equipment (more difficult to evaluate,
estimated by the authors to be around 50%), however the CoC of conduct has been rather
unsuccessful in other products categories, such as simple STB (DTAs), DECT telephones, small
appliances and kitchen tools, consumer electronics.

What are the Challenges for the Existing Codes of Conduct?

The New Generation of Digital TV Reception Platforms.

New technologies are emerging for digital television in Europe particularly to service the fast
developing market for high definition television (HDTV). This is likely to grow significantly in Europe
from 2006 with a predicted five million homes using advanced (mainly Satellite) platforms for HDTV by
2008.
High definition requires higher data rates and new signal processing techniques. Advanced video
Codecs (e.g. MPEG4) for this purpose require more power as do the tuner/demodulators for the DVB

1130
– S2 standard for satellite HDTV signal delivery. This standard delivers a 30% increase in bit rate and
has powerful error correction, but there is a power requirement penalty. This is compounded by the
functional requirement of several tuner/ demodulators to allow local hard disc recording of multiple
signals within the STB platform and decoding of multiple signals for distribution around the home.
The implications for power management are severe. There will be a higher overall power requirement
generated by a large number of circuit blocks. The power management system will need to know the
status of circuit blocks simultaneously supporting internal platform requirements and external network
requirements. The management of network data streams will be complex, and remote command
instructions from home network clients may mean that the box never has a standby active mode as
the current CoC defines it. Intelligent “Auto-Standby functions may be needed. In satellite systems we
will see multiple LNB combining units for single cable routing. These may well be separately powered
and incorporate their own power management.
For the CoC, a task force has been mandated to address the issues raised by the new generation of
platforms. This will not only have to address the implications on power requirement criteria limits of
new functions and circuit blocks but will also have to review the relevance of the existing definitions of
standby, standby active and on modes. Test set up and test methodology for these new platforms will
require a radical review especially in the context of networked signals and single cable multiple LNB
routing. This work will have to be completed in a relatively short time frame since one of Europe’s
leading Service providers, B-Sky-B has stated that 3 million UK subscribers will be using multi-room
networking by 2010.

Figure 5: Network Architecture for the Next Generation of STBs (Courtesy Pace Micro
technology)
The Challenge of Simple Digital TV Converter Boxes

It is likely that more than 200 million simple converters will be in operation in Europe, beyond 2015, to
keep legacy analogue TVs and video recorders in operation in the transition to Digital TV
broadcasting.
These converters will be, in the main, horizontal market products. The impact of cost competition on
these products has already driven energy efficient design to an all-time low. Standby passive power
requirement has moved from less than 1W in 2003 from two major manufacturers to a point where no
manufacturer is currently meeting the CoC criterion of 2W. Conformance testing shows that there is a
5
step increase in the average standby power of these products to an average level of over 6W.

5
Briefing note UK Market Tansformation Programme www.mtprog.com

1131
Activity from CoC working group members to remedy this situation is concentrated on establishing a
set of international criteria for simple converters, with the object of influencing OEM sources of the
product to endorse good energy efficiency criteria. The large majority of simple converters are
imported into the EU, and are manufactures by relatively known manufactures. There is in place a
major efforts to try to make this manufacturers aware of the CoC specifications. Cooperation with
China, the USA and Australia in this area, a very important initiative to further promote the CoC, is
starting to influence production standards. But every effort must still be made by political
representatives in the CoC working group to ensure that national product procurement schemes and
product subsidy mechanisms endorse only the best energy efficiency standards.
Even if good standby passive performance is achieved for the majority of simple converter products in
the market, the problem of mitigating their potentially massive energy impact is still not resolved.
6
Unrelated surveys of user habits show that around 60% of STBs are left on when the TV is put into
standby, if another remote control or a secondary remote control operation is required. The CoC
working group has discussed this problem over several meetings and has recently endorsed an
7
elegant Auto-Standby solution proposed from the Pace Micro Technology . Again every opportunity is
being taken by working group members to drive this solution into an internationally accepted set of
criteria for simple converters.

External Power Supplies - What are the Challenges?

8
Conformance testing for the CoC working group has shown that the most popular external power
supplies, those supplied for mobile phones, all comfortably meet CoC criteria. There are still major
product groups where
Inefficient linear supplies are the norm such as DECT wireless PSTN phones, DAB Radios and of
growing concern because of the spiraling volume of sales, Broadband, Modems, Switchers and
Wireless routers. A priority for the political members of the working group is to convince major
procurers of these products, such as supermarket chains, of the fact that CoC working group
discussions show that the delivered cost at point of sale of a more efficient supply, meeting the CoC
9
criteria, can be the same as or less than the inefficient linear alternative.
A major step forward in further reducing external power supplies consumption will be the introduction
of the Phase 2 levels on1.1.2007, when no lad consumption will be further reduced, and the load
operation efficiency will be harmonized with the Energy Star current specifications.
In addition, both the Digital TV Service CoC and the Broadband CoC requires that products supplied
with external power supplies, have models that are CoC compliant, this should further promote
efficient power supplies. In additional the revised Energy Star specifications for office equipment will
also require compliant external power supplies for the Energy Star labeled products.

Conclusions
The EU Codes of Conduct have served as an important platform for promoting energy efficiency in
Europe. The Code of Conduct for Digital TV Service Systems has already reduced the energy
consumption of STBs, even if these offer many more features and services. Now energy efficiency is
among the design priorities of next-generation STBs and in the procurement specification of service
providers. The full transition to digital TV and enhanced services such as HDTV and home networking
will add further challenges for energy savings and climate change mitigation in the EU.
The revised Code of Conduct for Digital TV Service Systems also covers the ‘simple’ STBs used to
convert free digital TV signals. It is hoped that the Code of Conduct will make an impact on the power
demand of these devices before hundreds of millions of them are sold. Simple’ STBs are now the
principle target for the CoC, however if it proves too difficult to have them adequately covered by the
CoC, MEPs should be introduced as soon as possible. The new European Directive on Eco Design
(also known as EuP) offers an excellent platform for adopting effective MEPS. The only concerns is the

6
BSkyB 2003, UK Consumers’ Association 2000
7
Presentation Ken Dale 30th Nov 2005 STB CoC meeting Draft Proposal for an Auto Standby Standard for STB
http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm
8
Presentation Bob Harrison 25th May 2005 Power Supply CoC meeting UK MTP Conformance testing overview.
http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm
9
Presentation Ben Sutherland Power Integrations at the CoC Power supply meeting:
http://energyefficiency.jrc.cec.eu.int/html/standby_initiative.htm

1132
time requested to arrive to have MEPS in force, and the ambitiousness of the MEPS levels. For more
traditional and technological stable products, imported in large quantities such as external power
supplies and simple STBs, in the authors’ view MEPS will be a more appropriate policy tool for
achieving savings. In any case, the experience made so far with the external power supplies CoC has
been useful to show that efficient power supplies could be used also with mass market end-use
products (e.g. mobile telephone), and that cost-effective technologies exist to meet challenging
efficiency requirements (even with linear technologies).
Last but not least, given the success of the Codes of Conduct in creating a useful stakeholder forum
and in achieving concrete results in a very dynamic technological sector, the European Commission
and national experts have prepared a new Code of Conduct to reduce energy demand in broadband
equipment for the residential sector. This activity has raised an interesting expansion of the CoC to
influence energy-efficiency standards in the Broadband Service Provider network exchanges and road
boxes, where savings in power at the subscriber interface will drive large savings in cooling power.

References
[1] Bertoldi, Paolo et al.. 2002. “Standby Power Use: How Big is the Problem? What Policies and
Technical Solutions Can Address It?”. In Proceedings of the ACEEE 2002 Summer Study on
Energy Efficiency in Industry. Washington, D.C.: American Council for an Energy-Efficient
Economy
[2] Rezessy, Silvia, Paolo Bertoldi and Agneta Persson. 2005. “Are voluntary agreements an effective
energy policy instrument? Insights and experiences from Europe” In: Proceedings of the
ACEEE 2005 Summer Study on Energy Efficiency in Industry. Washington: American Council
for Energy Efficient Economy
[3] Harrison, Robert et al. 2004. “Digital TV and Broadband Communication: Containing the Energy
“Black Hole” With the Innovative Policy Tool of a Code of Conduct” In Proceedings of the ACEEE
2004 Summer Study on Energy Efficiency in Industry. Washington, D.C.: American Council for an
Energy-Efficient Economy
[4] Molinder, Olof 1997. “Study on Miscellaneous Standby Power Consumption of House Equipment
(light-audio-white goods).” In Proceedings of First International Conference on Energy Efficiency
in Household Appliances, Florence, 1997 Italy (Editor P. Bertoldi, Publisher Springer Verlag).

1133
1134
The Need for a New Television Power Measurement Standard
Based on Average Picture Levels (Apls) due to the Emergence of
New Technologies Including LCD and Plasma
Keith Jones1, Robert Harrison2, Jon Fairhurst3
1
Digital CEnergy Australia
2
Intertek ETL SEMKO
3
Sharp Corporation USA

Abstract
Television technology has changed over the last few years with the introduction of new technologies
such as Plasma and LCD technologies. In addition, particularly in developed nations, there has been
a strong shift in the screen sizes of televisions purchased toward larger screens.
This paper reviews this trend and the applicability of a number of standards that have been developed
to measure the power of televisions. A new approach to the measurement of television power is
discussed using the notion of Average Picture Level (APL) as an important component of the test
signal applied to televisions in order to better simulate the signals that are actually applied to
televisions and the resulting video that is displayed.
It is hoped that this contribution will lead to a new standard that is applicable to all television
technologies and reflects a fair and real life use and resulting energy consumption of television.

Background
Over the last several years there has been much discussion at many forums as to the inadequacy of
current test methods for the power consumption of Televisions. In 2005 Dr Larry Weber delivered a
1
paper in San Francisco that highlighted the inadequacy of current test patterns to correctly simulate
the Average Picture Levels (APLs) that are inherent in the video signals being displayed on TV
display devices.
Recently the work on characterizing APLs and the power consumption characteristics that differing
display type exhibit as a result of APL level has been continued in Australia, UK and the USA. This
paper presents the result of this effort and proposes a way forward. Also included in this report is a
comparison of the APL method as opposed to the JEITA method.

The need for a new method

Over the last several years television technology has seen the emergence of Plasma and LCD
technologies which are rapidly replacing CRT technologies in developed countries and will become
the dominant technologies over the coming years (Fig1 and 2). As these technologies are suited to
larger screen sizes and as the market particularly in developed countries moves toward larger screen
sizes there will be a resultant increase in energy consumption. To asses the impact of this power
consumption and to establish programs to ensure efficient use of these products an accurate real life
use measurement of power consumption will be required.

1135
 Global TV Demand

Demand Forecast of TVs by Technology

250

200
Others
150 RPTV
PDP TV
100 LCD TV
CRT TV
50

0
'03 '04 '05 '06 '07 '08 '09 '10

Displaybank, 2006 Forecast

2
Figure 1: Global TV demand and migration to LCD and Plasma

TV Growth Trends

z CAGR、LCD TV：42%、PDP TV：28%、CRT TV：-10%、FPD TV

250

200
Millions Units

150

100

50

0
2004 2005 2006 2007 2008 2009
MD RP 1.6 2.5 3.0 3.4 3.5 3.7
CRT RP 4.3 2.6 1.4 0.6 0.3 0.0
PDP 2.8 5.7 8.9 11.9 14.2 15.3
LCD 8.8 20.3 35.7 52.6 69.6 82.9
CRT 165.9 155.4 141.9 127.1 114.2 102.9

3
Figure 2: Growth of Plasma and LCD Technologies

Figure 3 shows the estimated global power consumption for 2005 and the contributions made from
each technology type.

1136
 Estimated Global Power Consumption 2005

Sales 2005 Screen Size Power Rating Annual GWh

CRT 155400000 51 70 22546.60

LCD 20300000 66 130 5168.13

PDP 5700000 106 300 3219.57

Rear
Pro 5200000 120 180 1798.36

Total 186600000 32732.65

3
Figure 3: Estimated Global Power Consumption 2005

Objectives of a New Test Method

Given the proliferation of display technologies, and the differing nature of their power consumption
characteristics, the primary objective of any new test method must be to develop a test methodology
that provides realistic power consumption measurements for all display types under real world use
conditions. The test method should be a simple as possible to administer.
In addition to the power consumption associated with the pictures being displayed the test method
must also consider the contribution to power consumption of the following factors:
• Audio
• Digital Tuners
• Standby Power
• Energy Saving Features and Settings
Lastly, the new test method must have some way to handle TVs with other devices built in such as
DVD recorders etc.
This report demonstrates that a method that meets the above requirement is indeed possible and
could be developed relatively quickly.

Results of the APL sampling survey from Australia, USA and the UK
APLs have been collected from a variety of sources over the last several months. The objective of the
survey was to gain a better understanding of the nature of the APLs inherent in the video that is then
displayed on a display device. To ensure that representative APLs were being sampled it was
proposed to ensure that a wide variety of program material was sampled over the main platforms of
Terrestrial, Satellite and Cable.
Three aggregated sample sets are shown in Figure 4. They represent 80 samples from NSW in
Australia being a mix of Digital Terrestrial and Satellite. There were 27 samples from Victoria in
Australia being a mix of Analog and Digital Terrestrial and DVD. There are 8 samples from the USA
being a mix of Cable, Satellite and DVD.

1137
 APL Sample Set Comparison

25.00%

20.00%
%of CapturedFrames

15.00%
NSW
VIC
USA
10.00%

5.00%

0.00%
0 20 40 60 80 100 120
APL %

4
Figure 4: APLs sampled from Australia and USA

From a TV power consumption point of view the most important trend that is evident is that only 10%
of APLs exist above a level of 55% brightness. The three curves do vary a little. Analysis shows that
the NSW samples had a larger number of sporting content than the other two samples and this
content was captured in either bright conditions or under lights. The APL distributions (See Appendix
1) demonstrate a higher average for these types of content. This has the tendency to pull the average
of the aggregate higher. In addition the other two samples had more Movie APLs captured. The profile
of the Movie APLs is darker than many of the other APLs (See Appendix 1) captured and would have
pulled down the Aggregate average of those samples. This is also evident from the percentage of
APLs below 15% brightness. For the NSW sample this was 10% for the Victorian and USA samples it
was 22.5% and 24.3% respectively.
The NSW sample has been analysed for this trend and is shown in Figure 5. The other samples are
very similar.
Independent of the category the important point still remains that few APLs exist above 55%
brightness and the effect of category is merely the shifting of the APL distribution higher or lower in
the bottom half of the capture range.
On last comment on the APLs captured. The APLs are shown as non Gamma corrected. CRT TVs
will Gamma correct due to the characteristic of the picture tube. As has been discussed previously
this is indeed where the Gamma factor came from as it was decided at the onset of TV broadcasting
that it was easier to compensate for the response of the CRT display in the camera rather than every
television. A question needs to be asked as to whether Gamma in its own right plays any part in the
power consumption of the differing display technologies. This paper contends that any TV or display
device will exhibit a power consumption dependent upon an input signal level. What any particular TV
does to that signal to account for display characteristics does not change the eventual power
consumption outcome and therefore does not need to be considered.

Average APL by Category

35.0% Documentry Average

AverageNumber of Frames

30.0% News Average

25.0% Austar Movies Average

Cricket Average
20.0%
Drama Average
15.0%
Golf Average
10.0%
Various Sport Average
5.0%
Cooking Average
0.0%
Rugby Average
-5.0% 0 50 100 150
Cartoons Average
APL % Soaps Average

4
Figure 5: NSW Australia average APL by category

1138
In fact, as will be shown when the power consumption characteristics of differing display types is
reviewed, there is evidence of quite complicated processing of the video signal to enhance the picture
quality which are not consistent with pure Gamma correction.

Power Consumption Characteristics of Display Technologies

Along with understanding the nature of APLs to understand the power consumption of TVs it is
necessary to analyse how the TV consume power. In order to do this a test sequence of 20 APL
levels in 5% increments from 0% - 100% was developed. This test sequence of patterns was then
applied to a variety of TV display types and for each type a number of samples were measured. For
the purposes of this study the actual models and brands will not be named.
The TV types measured were Plasma, CRT and LCD. The results are shown in Figures 6,7, and 8.

% of Maximum Power Consumption Vs APL For

CRT

120.0%
% of Maximum Power

100.0% 36 Inch
80.0% 32 Inch
used

60.0% 29 Inch
40.0% 33 Inch
20.0% 36 Inch

0.0% 32 Inch
0 20 40 60 80 100 120 32 Inch

% APL

4
Figure 6: Per Cent of Maximum Power Consumption vs APL for CRT

% of Maximum Power Consumption for Plasma

120.0%
% Maximim Power

100.0%
42 Inch
Consumption

80.0% 42 Inch
60.0% 42 Inch

40.0% 42 Inch
42 Inch
20.0%

0.0%
0 20 40 60 80 100 120
% APL

4
Figure 7: Per Cent of Maximum Power Consumption vs APL for Plasma

1139
 % of Maximum Power Consumption Vs APL for
LCD

120.0%
% of Maximum Power

100.0% 26 Inch NBLM

Consumption

80.0% 26 Inch BLM

60.0% 26 Inch BLM
40.0% 26 Inch BLM
20.0% 40 Inch BLM

0.0%
0 50 100 150
% APL

4
Figure 8: Per Cent of Maximum Power Consumption vs APL for LCD

What is evident from these results is that with the exception of a few TVs the tendency is for
maximum power consumption to be achieved before 50% APL levels. The exceptions also exhibit
high power consumption (Between 60% and 80%). The reason for this is associated with the
operation of what is known as Automatic Brightness Limiting (ABL) in CRT and Plasma technology
and Back Light Modulation (BLM) in LCD technology.
ABL is applied so as to ensure that CRTs and Plasma panels are not overdriven thus causing the
potential for damage and/or reducing the potential life of the product. It acts to limit the brightness of
the display and thus caps the power consumed.
BLM is almost the reverse. LCDs suffer from relatively low contrast ratios. This is caused by black
level bleed where the crystals cannot shut out all the light at low black levels so for these levels of
black they appear grey. To over come this problem, designs have been introduced recently where the
APL is continually monitored and when conditions are right and the picture is darker the backlight is
reduced to give an apparent improvement in black level. This has the effect of reducing power for
those levels of black. The effect of this is seen in Figure 8. Not all LCDs displays have this feature and
for those that don’t, because the backlight is no all the time, there is no variation in power
consumption against APL level.

Review Of JEITA and IEC 62087

For the purposes of this paper only the On Mode power measurement method of the JEITA and IEC
62087 will be considered. In doing this it is not implied that other aspects of the power consumption
for TVs are not important. However, by limiting the discussion at this point only to the On Mode it will
be easier to assess the effect each method has on the final measurement.

IEC 62087
5
IEC 62087 uses a three bar black and white pattern for the measurement of ON Mode Power. This
pattern has an APL of 50%. As has already been discussed, from the power consumption
characteristics of most display types, TVs at this APL level will show a power consumption of 100% of
maximum. This pattern therefore, in effect, will give the same power measurement as a 100% APL
white raster in most cases. Given that the TVs are actually showing pictures with APLs below 50%
ninety percent of the time the suitability of this test pattern must be questioned. This is reflected with
the IEC currently in the process of forming a working group to look at this issue.

JEITA
6
The other standard of interest is the JEITA standard. This uses four patterns. 0% Black, 100% White
Raster, three bar Black and White and Colour bars both with an APL level of 50%. The results of
these measurements are then averaged according to the formula:

((Pw+Pb)/2 + Pc + Pt)/3

1140
Where:
Pw is Power Measured with 100% white
Pb is Power Measured with 0% Black
Pc is Power Measured with Colour Bars
Pt is Power Measured with Three Bars

This method would be more accurate because it is at least considering power consumption at black
level which would effectively reduce the measured power. However, across all our APL samples the
highest 0% APL frame count was 3.3%. The other two being 1.0% and 1.4%. In light of this the
validity of using a 0% pattern and giving it an effective weighting of 16.75% in the calculation would
have to be challenged.
The same could be argued for the 100% white raster although given the figure that 10% of APLs are
greater than 55% and that almost all the TVs sampled did consume 100% of their maximum at that
level it is indeed a closer representative figure.
The method becomes more distorted when we consider the Colour Bars and the Three Bars. These
have an APL of 50% which means most TVs are at or close to full power and their weighting is 66.6%
of the calculation. When in fact at best case less than 20% APLs from our sample reach this level.
In summary then the JEITA method gives effectively a weighting of 83.25% to 100% power
consumption when in fact less than 20% of pictures reach this level and 16.75% weighting to the
power consumed at 0% APL when from our analysis less than 3% of pictures are at this level.

Conclusion
The conclusion is that due to the nature of TV video signals and their APLs and the power
consumption characteristics that neither the existing IEC62087 or the JEITA adequately measure TV
power consumption in a way that reflects real word use. To be relevant the new test method must
take into account the real world APLs. Power consumption of most TVs varies over a similar range of
APLs that actually make up the pictures displayed so to measure power realistically the video used for
the measurement must also have APLs that exist in the 0%-50% region.

A proposed Test Methodology

This proposal is not to be taken as anything more than a discussion point. It has some validity, but,
suffers from a relatively isolated development by a small group of individuals. Test methods need to
be flushed out properly with all stakeholders and in the appropriate forums. It is offered, however, as
an example to establish that a test method based on APL measurements can be developed and can
potentially produce better results than the existing methods.
It also needs to be noted that the results of this work are simulated and not actual measurements. The
simulation is achieved by using the Power Consumption Characteristic curves from the TVs sampled.
As these curves were produced by actual measurements against the APLs it would be expected that
actual measurements would confirm the results. This must be done as further work.
The two biggest challenges in developing a test method based on APLs are:
• Establish an appropriate profile for the representative viewed APLs or a good estimate
• Establish the actual video signal that represents that APL profile. It is possible to use normal
moving pictures that have been selected to replicate the APL profile. It is also possible to
produce a series of test patterns with fixed and moving segments that can also replicate the
APL profile.

APL Profiles
To create an appropriate APL profile the first step is to understand the differing distributions that
different content produces (see Figure 5). Simply aggregating these APLs with equal weighting for
each content type is not going to guarantee the average viewer profile. This will certainly be subject to
more research.
For the purposes of this report the author profiled his own weekly watching habits which provided the
following breakdown. This is not implying a typical viewer but is intended a starting point to develop
the test method.

1141
Table 1: Author’s TV viewing for an average week
Weekly
Content Hrs %
Sport 10 23%
Movies 4 9%
News 12 27%
Drama 12 27%
Documentary 6 14%
Total 44 100%

By applying this weighting to the collected APLs the following profile was generated (Figure 9).

Authors Profile

14.0%
% of Total Frame Coun

12.0%
10.0%
8.0%
Authors Profile
6.0%
4.0%
2.0%
0.0%
0 50 100 150
% APL

Figure 9: Author’s APL viewing profile

This was then used as the basis to model the power consumption of the various TVs that had been
measured for the Power vs APL response.

Estimated APL profile Approach

By considering the nature of the distribution of the APL, only 10% of APLs are more than 55%
brightness. We also know that of the remaining APLs, very few occur below 15% and the majority fall
between 15% and 55%. From this a simple profile has been developed that distributes the APLs as
follows:

Table 2: Estimated simple APL approach

APL Level Weighting
5% 10%
15% 10%
20% 10%
25% 10%
30% 10%
35% 10%
40% 10%
45% 10%
50% 10%
100% 10%

The idea behind this approach is to see if a simpler method may suffice rather than trying to resolve
the more complex issues of what is actually watched.

1142
 Test Simulation
Using the two APL profiles outlined above and the representative APLs for both the JEITA and
IEC62087 test patterns, the power consumption for ON Mode has been simulated and the results are
shown in Table 3.

Table 3: Simulated Power consumption – Comparison of methods

Authors Simple % % %
Model Profile APL Variation JEITA Variation IEC62087 Variation
Plasma 42 Inch 184 192 104.4% 217 117.5% 210 114.0%
Plasma 42 Inch 209 216 103.3% 250 119.6% 250 119.7%
Plasma 42 Inch 270 274 101.7% 300 111.0% 336 124.5%
Plasma 42 Inch 317 318 100.3% 307 97.1% 330 104.2%
Plasma 42 Inch 300 302 100.6% 303 100.8% 330 109.9%
LCD 40 Inch 182 181 99.9% 172 94.5% 191 105.2%
LCD 26 Inch 117 117 99.9% 116 99.1% 118 101.2%
LCD 26 Inch 105 106 100.4% 106 100.6% 110 104.6%
CRT 32 Inch 109 110 101.1% 123 113.3% 129 118.7%
CRT 29 Inch 107 108 100.6% 111 103.4% 119 111.1%
CRT 33 Inch 150 151 100.6% 157 105.0% 169 112.7%
CRT 32 Inch 68 72 105.4% 78 114.5% 72 105.9%
CRT 32 Inch 129 128 99.9% 126 97.9% 132 102.7%
CRT 36 Inch 165 166 100.6% 173 104.5% 182 110.1%
CRT 36 Inch 252 253 100.4% 259 102.7% 274 108.7%

The Power figures are in watts. The % variation is against the Authors Profile. We find that the
correlation between the Authors profile and the Simple APL is very close with no more than a 4.4%
variation and 9 of the TVs within 1%.
The JEITA method as expected does not give such a good result with a maximum deviation of 19.6%
and only 2 TVs within 1%.
The current IEC62087 is the worse with a maximum deviation of 24.5% and no TVs within 1%.

Simple Profile and Test Video

If it can be shown that the simple profile is sufficiently representative of normal viewing profiles using
such criteria as the variation in measurement and it greatly simplifies the creation of a test signal. The
APL levels are not hard to replicate and the actual test patterns can contain moving objects that better
simulate a moving picture. Also the technique is really an extension of existing methods with the
additional patterns within the APL ranges that the sampling has shown video signals to contain. This
does offer a potentially easier route to adoption.

Validation of Results and Further Work

The simulated results will need to be validated with actual power measurements. In addition a wider
range of TVs should be sampled to confirm the existing pool of data on the power consumption
characteristics.
At the beginning of the paper other factors of power consumption within TVs will need to be
considered. Some methodology for this exists within the existing standards and may well be suitable
for use. Others will have to be developed such as how energy saving modes are considered. An
important aspect of this is careful consideration of the differing modes of operation between Off Mode,
Passive Standby through to On Mode. The impact of Audio should also be considered although recent
8
work may suggest that the impact of this on overall power consumption is minimal.
Much work still needs to be done both on validating this work and consideration of other aspects of TV
power consumption. It is hoped however that this work can form the basis for a wider discussion
process with all stakeholder and at the appropriate forums.

1143
References
[1] L. Weber. Challenge of Measuring Annual Energy Consumption of TV Sets. New Paltze New
York.
[2] Jon Fairhurst. Sharp Cooperation USA 2006
[3] K. Jones. Digital CEnergy Australia 2006
[4] K. Jones. Digital CEnergy Australia 2006, Study on the nature of APLs and TV power
consumption Characteristics.
[5] IEC 62087 2002
[6] JEITA PDP05-B05-16

1144
 Appendix 1
APL by differing Sampled Categories and Conditions

Average Broadcast APL

25.0%

20.0%
% of Frames

15.0% Studio n= 28
DVD n= 6
Non Studio n= 23
Sport n= 32
10.0%
Studio Sets n= 21

5.0%

0.0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

Studio 28 9.7% 73.9% 16.3%
DVD 6 29.3% 63.3% 7.4%
Non Studio 23 29.4% 54.3% 16.3%
Sport 32 3.2% 74.3% 22.5%
Studio
Sets 21 46.8% 46.0% 7.2%

Average Broadcast APL

25%

20%
% of Frames

15% Studio n= 29
DVD n= 6
Non Studio n= 23
Sport n= 32
10%
Studio Sets n= 21

5%

0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

Studio 29 9.4% 74.4% 16.2%
DVD 6 29.3% 63.3% 7.4%
Non Studio 23 29.4% 54.3% 16.3%
Sport 32 3.2% 74.3% 22.5%
Studio Sets 21 46.8% 46.0% 7.2%

1145
 Average Studio Broadcast APL

25%

20%
% of Frames

15%
News n= 13
Variety n= 14
Current Affairs n= 2
10%

5%

0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

News 13 8.0% 71.7% 20.3%
Variety 14 10.9% 77.7% 11.4%
Current Affairs 2 8.5% 68.4% 23.0%

Average Studio Sets Broadcast APL

30%

25%

20%
% of Frames

Situation Comedy n= 1
15% Soap n= 5
Drama n= 14

10%

5%

0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

Situation 1 3.9% 92.0% 4.1%
Comedy
Soap 5 51.1% 37.6% 11.3%
Drama 14 49.3% 45.0% 5.7%

1146
 Average Sport Broadcast APL

50%

45%

40%

35%
Basketball n= 1
% of Frames

30% Football n= 1
Golf n= 4
25% Skating n= 1
Soccer n= 4
20%
Cricket n= 15
Rugby n= 6
15%

10%

5%

0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

Basketball 1 0.4% 95.8% 3.8%
Football 1 0.0% 80.7% 19.3%
Golf 4 4.6% 64.4% 31.0%
Skating 1 4.6% 36.8% 58.6%
Soccer 4 3.2% 73.2% 23.6%
Cricket 15 4.5% 72.6% 22.9%
Rugby 6 0.0% 85.3% 14.6%

Average Non-Studio Broadcast APL

18%

16%

14%

12%
% of Frames

10%
Documentry n= 8
Cartoon n= 7
8%

6%

4%

2%

0%
0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100

% APL

Category Sample <20% 20 to 50% >50%

Documentary 8 10.6% 69.0% 20.4%
Cartoon 7 9.0% 59.7% 31.3%

1147
1148
An Approach to the Environmental Analysis of the Eco-design of
Television Devices
Matthew Armishaw1, Bob Harrison2
1
Future Energy Solutions, AEA Technology plc
2
Bob Harrison Associates

Abstract
In 2004/05 AEA Technology plc conducted a project for the EC Joint Research Centre’s Institute for
Prospective and Technological Studies which aimed to provide a reference for the potential
application of eco-design in the consumer electronics sector through a case study dedicated to
television devices, taking into consideration both current and future designs. The definition of
television devices for the project covered current and future screen designs, and considered the
trends towards deconstruction and convergence.
The study assessed environmental impacts by TV screen design technology, both current and future.
It identified:

• The main environmental impacts of televisions

• Performance indicators selected to enable their application to other consumer electronic
products, and indeed other electrical appliances in a generic way.
• How these performance indicators are influenced by TV design.
• Market sales distribution trends and performance indicator variations by TV design
• Alternatives for benchmarking.
• Priorities for eco-design options, and policy options to achieve these improvements.

The project examined the environmental impacts throughout the lifecycle of televisions, including
energy, material resources and hazardous content.
It developed a set of micro and macro performance indicators for greenhouse gas and resource
impacts, and used these to evaluate performance of a variety of screen designs including CRT,
plasma, LCD, OLED and Field Emissive Display types.
Using market data and assumptions about future trends in the EU market, the study quantified the
impacts for alternative technologies and eco-design potential over a period to 2020.

Evaluating the Environmental Impacts

The project assessed the environmental impacts, aspects and product characteristics which are
applicable to existing TV designs using cathode ray tube ( CRT), liquid crystal display ( LCD) and
plasma display (PDP) technologies from 29 sources of analysis including LCA, eco-label and
environmental related studies available in the public domain, and concluded that life-cycle impacts are
dominated by energy in the in-use phase (in-use accounts for 80% of energy consumption according
to Reference no. 1, Bang & Olufssen). TVs share this characteristic with other products that are long-
lived, frequently used and consume electricity in operation. Depending on a TV’s use pattern,
operation accounts for the major share of the energy consumed in-use (as opposed to stand-by
consumption).
However, in the search for indicators which allow description of the relevant product characteristics
and environmental aspects in a measurable and comparable way, it was the aim to define indicators
which have a broad application to other consumer electronic products.
Although energy consumed in the use-phase of a TV is generally considered to have by far the
largest environmental impact with the current dominant TV technology (CRT display TV), there is a
danger in isolating this impact and, furthermore, there is an obvious difficulty in comparing different
kinds of environmental impacts. Lifecycle energy (i.e. that associated with production, use and
recycling) but also virgin resource depletion and the use of hazardous substances have to be taken
into account and included if these indicators are to influence and meaningfully compare the eco-
design of developing and new TV display technologies. Likewise, a system of indicators for the TVs
as well as for other consumer electronic products must consider or reflect a product’s entire life cycle.

1149
A list of indicators which were been considered for consumer electronics products is shown below,
thus including three groups, energy, resources and hazardous substances. These indicators were at a
later stage filtered to highlight the most important ones for television design.

Energy Indicators
• Material production (kg CO2 per product)
• Electronic part production (kg CO2 per product)
• Display device material production (kg CO2 per product)
• Production – finished product (kg CO2 per product)
• In-use power consumption – function of standby power + on-mode power + duty cycle (kg
CO2 per product)
• Product transportation (kg CO2 per product)
• Consumable materials (kg CO2 used in the manufacture rationalised to single product)
• End of life products transportation (kg CO2 per product)
• Recycle / disposal operations (kg CO2 per product)
The unit “kg CO2 per product” was later replaced by maybe the more straightforward “energy
consumption in kWh”.

Resource Indicators
• Recovered resources (by content weight in kg), meaning reused components and re-
cycled materials which are obtained from re-processed components and materials used
at least once.
• Recoverable resources (by content weight in kg), implying resources which has been
used at least once in the TV-product and can be used in other products/processes.
• Disposed resources (by content weight in kg).
• Newly supplied resources (by content weight in kg).
For each of these categories, the materials of which they are composed will also be indicators,
e.g. glass, copper, flame-retarded plastics, silicon etc.

Hazardous Substances Indicators

• Phosphors
• Lead in solder
• Lead
• Mercury
• Flame retardant Tetra-bromo-bis-phenyl A (TBBPA)
• Flame-retardant Poly-brominated diphenyl ethers (PBDE)
• Flame-retardant Poly-brominated biphenyls (PBB)
• Hexa-valent chromium
• Cadmium
• Poly-vinyl chloride
• Bismuth

It was important that the suggested indicators were robust enough to be used for other domestic
electronics products (used peripherally with the TV) as well. The tables below show how the
suggested indicators for energy and resources are relevant for other products. Similarly, all but
mercury and phosphors in the list for hazardous substances will apply to the indicated products.

1150
 Indicator Product
Surround DVD & Hard Disc Games
Set Top Box
Sound System (Player/Recorder) Machine
3
Includes
Material 3 3 DVD or
3
production Includes DVD Silicon
medium software
medium
Electronic part
3 3 3 3
production
Production of
3 3 3 3
finished product
In-Use power
3 3 3 3
consumption
Consumable
materials 3 3 3 3
production
Product
3 3 3 3
transportation
End-of-life
Product 3 3 3 3
transportation
Recycling and
Disposal 3 3 3 3
operations
Others not
related to TV None none none none
product
Display Device
component
n/a n/a n/a n/a
production and
assembly

Micro Indicators for Resource Factor

Indicator Product
Surround DVD & Hard Disc Games Set Top
sound System (Player/Recorder) Machine Box
Recovered
3 3 3 3
Resources
Recoverable
3 3 3 3
Resources
Disposed of
3 3 3 3
Resources
Newly Supplied
3 3 3 3
Resources
Others not
related to TV none none none none
product

Benchmarking the Product

The study detailed the different methodologies for comparing the impact of technology improvements,
and finally opted for a factor system approach. This system has been adopted by many major global
industries in recent years as an effective Eco indicator tool that incorporates not only environmental
aspects but also attempts to add the products value to the community in terms of its lifetime and
functions. The factor system was developed by considering three issues separately: prevention of
global warming, effective utilization of resources and use of non-toxic materials. Factors are

1151
calculated for each of the three areas and, thus, a transparent system is created where different
impacts and causes for impacts are kept separated. As shown below, an environmental factor is
calculated by comparing the product being evaluated with a reference product. The environmental
factor is not a composite index including the impact of all the three areas mentioned above. Instead,
three different environmental factors are proposed, each of them covering one of the three areas but
including the environmental impact over the whole lifecycle. In that sense, the environmental factor is
a compound value.

Eco-Efficiency of the (new) Evaluated Product

Environmental Factor =
Eco-efficiency of the Base (reference) Product

Thus, the comparison is being made through eco-efficiencies, which are individual for the three
separate areas i.e. energy used, resource efficiency and use of hazardous substances. The eco-
efficiencies combine the product’s value to the community in terms of lifetime and functions (i.e. it is
taken into account if the product functions only as a TV or also as a PC, a DVD etc.) with its
environmental impact (by summing up the indicators presented previously) in each area. The
approach does not place any value of the use of the devices, it merely incorporates possible multiple
functions of a product, so that e.g. a device with multiple functions but with higher energy use than a
device with only one function is not treated unfairly in the comparison of products. The general
formula is shown below and the environmental impact (as well as the unit used) would change
depending on if the eco-efficiency is calculated for energy use, resource efficiency or the use of
hazardous substances.

Product life x Number of product functions

Eco - Efficiency =
Environmental Impact

Or

Product life x Number of product functions

GHG efficiency =
Lifecycle GHG emission

Product life x Number of product functions

=
Resource [Newly supplied resources + Resources disposed of]
=
efficiency (i.e. resources that do not re-circulate into the product
over its entire lifecycle)

Product life x Number of product functions

= (Used resources – Recovered resources) + (Used resources – Recoverable
resources)

Forecasted Environmental Performance

The factor system can furthermore be used to illustrate forecasted environmental performance. As it is
considered important that such illustration is adequate for benchmarking as well as for the
prioritisation of environmental indicators for those products that will have the greatest overall
environmental impact, the study has opted to include a sales weighting element. In this approach,
individual technology trends are multiplied with each technology’s proportion of the (sales) market at a
given point in time and, thus, the method shows the contribution of each technology to the overall
performance of the sector. This allows the technologies that will have the greatest current and future
impact to be identified and priorities set accordingly. As an example, the figure below shows the sales
weighted trend of the greenhouse gas factor (GHG) for each technology as well as a trend for the
“total” factor. The trends are derived by multiplying the calculated GHG factor for each screen
technology with forecasted sales.

1152
 Greenhouse Gas Factor by TV Technology
1.20
Sales Weighted Factor

1.00

Total Factor
0.80 CRT
LCD
0.60 Plasma
Projection (MEM)
0.40
FED
OLED

0.20

0.00
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Year
Sales weighted trend of the greenhouse gas factor per screen technology

Establishing The Base Case

Through close co-operation with EICTA, the European trade association for consumer electronics
appliance manufacturers, the project was able to obtain environmental data from a variety of sources
such as manufacturers stated in-use energy consumption, partial bills-of-material, and partial Life
Cycle Analysis (LCA) results. However, very few complete LCAs are available in the public domain
and it was necessary to expand the review to capture computer monitor LCAs (there being little
difference between screen technologies for TVs and computers) as well as eco-labels.
Moreover, any available information covers mostly the CRT technology and throughout the project it
was necessary to resolve partial data sets together in order to derive a composite representative of
non-CRT technologies. For the yet to emerge OLED and FED screens, the project used anecdotal
evidence, industry evidence, and media reported claims of energy and resource usage to estimate
their eventual impact.
Once materials and energy data had been obtained, the factor system allowed the comparison of
different technologies against an established benchmark:

GHG factor GHG efficiency of the evaluated TV

=
(energy factor) GHG efficiency of a base (reference) TV

Resource efficiency of the evaluated TV

Resource factor =
Resource efficiency of a base (reference) TV

As the dominant technology is CRT it was chosen as the reference TV and allowed the relative
impacts of the emerging technologies to be assessed. The following graph shows the projection
obtained for resource efficiency.

1153
 Resource Efficiency Factor by TV Technology

1.80

1.60

1.40
Sales Weighted Factor

Total Factor
1.20 CRT
LCD
Plasma
1.00
Projection (MEM)
FED
0.80 OLED

0.60

0.40

0.20

0.00
2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Year

Assessing Design Improvements

From discussions with manufacturers as well as using market intelligence, a number of potential
improvements to the design of televisions which could result in a reduced environmental impact were
identified. The table below shows a small selection of the 16 improvements which were identified:

Primary 'Area Earliest Date Effective Quantified

Screen
of Effect' Innovation Introductio (100% of Saving/
Technology
(GHG/RE/HSF) n sales) Improvement
380W to 250W
Primary
LCD GHG 2006 2010 (On)
Secondary 70W to 47W (On)
CFL/Discharge Backlighting Primary 45 mg Mercury
LCD HSF 2006 2010
1 Replaced by 'super-bright' Secondary 23 mg Mercury
LEDs

MEM GHG Primary 2006 2010 195W to 125W

410W to 310W
Energy Recovery Circuitry Incorporated (On)
2 Plasma GHG Current 2008
410W to 260W
Plus Additional Effect of LSI (On)
10kg material can
Magnesium Alloy Casings
3 All RE Primary 2006 2008 be recycled that
(fewer barriers to recycling)
currently is not
10kg material can
Metal Sintered Plastics in
4 All RE Primary ? ? be recycled that
Casings
currently is not

For the above design measures for which an impact can be quantified, it is possible to calculate new
values for GHG and resource efficiencies and, consequently, their respective factors. When sales
weighting is applied, the resultant impact of each potential improvement on the EU fleet can be
derived. In this way, the design improvements which have the most significant impact at a market
level can be assessed and prioritised. For example, the following table illustrates that although the
incorporation of energy recovery circuitry in plasma screens (example 5) produces a significant GHG
efficiency improvement in that product, this improvement is minimal at the EU level in the long term
due to the relatively low penetration of plasma technology into the EU market. Whereas innovation
policy might have centred on such measures in the past, this system of assessment allows policy
makers to focus on those measures which will achieve the best overall impact.

1154
 2015
2010
Efficiency Factor Improve-
Improve-ment
Improve- Improve- ment
in 'Fleet
ment ment in 'Fleet
Factor'
Factor'
GHG Measures
Current Laptop
Technology (buck
1 0.4443 0.0100 0.0000 0.2345
regulator) Designed
into PSU in OLED
CFL/Discharge
Backlighting
2 0.3403 0.0076 0.0000 0.0151
Replaced by 'Super-
bright' LEDs in MEMs
Development of
3 Improved OLED 0.1707 0.0038 0.0000 0.0532
Screens
CFL/Discharge
Backlighting
4 0.4321 0.0022 0.3130 0.0551
Replaced by 'super-
bright' LEDs in LCDs
Energy Recovery
5 Circuitry Incorporated 0.0841 0.0019 0.0042 0.0000
into Plasma (Pus LSI)
Rectangular
6 Deflexion Coils in 0.0009 0.0009 0.00 0.00
CRTs
Resource Efficiency Measures
Metal Sintered
7 Plastics/Magnesium 0.0320 0.0036 0.0297 0.0382
Alloy Casings
Rectangular
8 Deflexion Coils in -0.0172 -0.0012 -0.0037 0.0000
CRTs
Reduced CRT Glass
9 0.0729 0.0050 0.0160 0.0000
Envelope Profile

Using performance factors to set overall fleet-based benchmarks

If the factors for each model is multiplied with its share of sales or forecasted share of sales, one gets
a good view of where the most important environmental impacts are and where they will be. Sales
could be the entire TV product output of all manufacturers – the EU ‘fleet’ - and would cover the blend
of different technologies, or could be applied to the sales of a particular manufacturer, group of
manufacturers or design technology. In this way, benchmarks can be established for the whole EU
fleet or for certain manufacturers or products. This allows the consideration and negotiation of
benchmarks, agreements and policy with both individual manufacturers or as a collective through a
representative body. Establishing an EU fleet benchmark for the “average TV”, gives manufacturers
the flexibility to allocate varying benchmarks among themselves, which would not hamper innovation.
This approach has the flexibility to accommodate new technologies entering the market. As long as
the overall benchmarks are not exceeded, manufacturers are free to bring new products onto the
market. However, projecting a benchmark for future years presents a rolling target for manufacturers
that ensures that the market constantly moves forward.
Although the fleet factor method offers significant advantages over alternatives such as simple
minimum environmental performance standards, there are complexities involved in its introduction.
For example, the adoption of a common LCA tool must be resolved with manufacturers.

1155
Compatibility with other EuP Methodologies
The flexibility of the fleet factor method means that it could readily support, or be supported by, the
methodology and tool developed by VHK as part of a separate EuP Pre-preparatory study. In fact,
these seemingly different approaches might compliment each other very well, with the ‘factor’
approach being used to set overall policy and the VHK tool helping to calculate product factors based
on manufacturers inputted data. An ‘interfacing’ tool could easily be created which could allow
manufacturers to securely input their product bills of materials and sales figures, and the software
could then automatically calculate their fleet factor, set future benchmarks, and provide instant impact
data to policy makers.

Conclusion
The methodology of a sales-weighted fleet factor allows manufacturers flexibility across their range of
TV products/technologies - and as such is positive for innovation – whilst allowing environmental
improvements to be assessed at a “macro” level. This allows policy makers to set meaningful targets
for improvements and to monitor them. It takes into account functionality of product can accommodate
new technologies. However, the concept is radical and it may take time for manufacturers and policy
makers to assimilate. It also involves a higher degree of maintenance and would require
manufacturers to collate and submit comprehensive product and fleet data (assurance needed about
confidentiality). It would also require a comprehensive auditing and policing process to verify
manufacturers’ data.

Perhaps, the main advantage of adopting this type of system over straightforward minimum standards
is that it can be applied as a preventative measure rather than a reactive one. This is particularly
beneficial in the consumer electronics market where the rapid process of technology turnover will
increasingly mean that minimum standards, which are difficult to set in advance of a technology’s
growth, are developed too late to have meaningful impact on the environmental impact of a product.

1156
Table of References Consulted in the Project

production
Material

facture
Manu-

In-use

Disposal
Ref
Title
#

1 LCA of a Cathode Ray Tube Television (B&O) 3 3 3 3

2 LCA of a Cathode Ray Tube Monitor 3 3 3 3
Nordic Swan Ecolabel (Audio-Visual
3 2 3 3 3
Equipment)
4 Blue Angel Ecolabel (Monitors) 2 3 3 3
5 Canadian Environmental Choice 2 2 2 2
6 US Green Seal Program 2 2 2 2
Australian Good Environmental Choice
7 3 3 3 3
Programme
8 Korean Environmental Labelling Association ? ? ? ?
9 Taiwan Greenmark 2 3 3 3
10 Environmental Choice New Zealand 2 3 3 3
11 LCA of CRT and LCD Monitors 3 3 3 3
EU Ecolabel – Feasibility Study for Portable
12 2 3 3 3
Computers
13 EU Ecolabel for Personal Computers 3 3 3 3
14 EU Ecolabel for Portable Computers 3 3 3 3
15 EU Ecolabel – Feasibility Study for TVs 2 3 3 3
16 EU Ecolabel for TVs 2 3 3 3
UK Market Transformation Programme TV
17 2 2 3 3
Innovation study
18 Design for Environment Guide 2 3 3 3
Case Study for Calculation of Factor X (eco-
19 efficiency) – Comparing CRT, PDP and LCD 3 3 3 3
TVs
20 European Energy-Based Legislation - EuP 2 2 3 2
European Energy-Base d Legislation – the EU
21 3 3 3 3
Ecolabel
European Energy-Base d Legislation – Industry
22 2 2 3 2
self commitment
European Energy-Base d Legislation – Code of
23 2 2 3 2
conduct for DSS
International Energy Based Legislation – US
24 2 2 3 2
EPA
International Energy Based Legislation –
25 2 2 3 2
Japanese Top Runner
International Energy Based Legislation – -
26 2 2 3 2
GEEA
Life Cycle Assessment of a Colour TV –
27 3 3 3 3
Samsung Electronics Co, 1997
28 PRe Life Cycle Tool 3 3 3 3
29 Product Design Tool 3 3 3 3

1157
1158
CE and IT - Market Continuously Driven by New Technologies and
by the Development of Changing Consumer Approach
Jürgen Boyny

GfK Marketing Services Europe

Consumer Electronic Market today is not anymore only watching TV or

listening music.
Approximately 50 years ago, when consumer started to watch TV or to listen to music, recorded on
vinyl the use of consumer electronic products have been very easy, very concentrated on specific
times and the number of devices, which have been in use have been very limited. Beginning of this
century, 50 years later, the usage of consumer electronics is much more different. Not judging the
quality of content here, the use of consumer electronics has changed into different directions: more
frequent use, at home and mobile outside, more media as for example internet, more content as e.g.
TV programs now are running 24 hours a day. But the most significant change is, that content
became digital.

What is Consumer Electronics in private households?

Basically Consumer Electronics can be differentiated by to different experiences:
• viewing pictures
• listening music
Pictures can be moving pictures as for example TV programs or still pictures, taken by a camera.
Nowadays Digital Still Camera is absolutely accepted as a part of Consumer Electronics. But we can
differentiate also by another criteria: consumer electronics can be passive or active.
In the beginning, all have been passive only. People sat down in front of the TV or in front of a radio
or record player. During the seventies however, Consumer Electronics became active, which means
consumer had the possibility, not only to playback, but to record first of all music with a Cassette
Deck. During the 80ies they new technologies have been developed. Video Tape Recorder came up
and a view years later also Videocameras or Camcorder.
But we have to be aware, basically even today Consumer Electronics is nothing more than viewing
pictures or listening music.

Innovation steps became faster and faster

Even Consumer Electronics can be focused on viewing pictures and listening music, the whole
branch is at least developing only via innovative products. Let’s take a view examples to demonstrate
the innovations of that branch and even the acceleration of the innovations

• In the 50ies TV started with black and white programs

• 17 years later 1967 the programs changed to color
• 1983 Video Tape Recorders, even with different systems have been developed
• 1886 – 3 years later - Camcorder have been introduced in the market
• parallel at that time CD-Player came up – the first step into digitalization
• in the 80ies the products became tiny and portable – Walkman have been introduced
• then in the 90ies a longer time with a brake of new products
• and then – even it seems as in a hurry – beginning and of the 90ies: DVD – Flat-TV – DVD-
Recorder – MP3-Player – Car Navigation

The basket of a retailer, dealing today with consumer electronic products is complete different to the
basket 7 years ago:

1159
 GfK Group Marketing
Retail & Technology
Services Consumer Electronics Germany Jürgen
Jürgen
Boyny
Boyny October
29
07-02-2006
April 2006
2004

Consumer Electronics

TV/Audio/Video*
Value %
1998 Western Europe 2005

Audio Home Systems

Videorecorder Car Audio Flat TV

USB Sticks
11 8
CD-R/RW
Camcorder
12 Car Navigation 25
7 DVD-R/RW 3
Video Tapes
3
4 Car Radios 3
3 Radio Recorder
Projection TV 4
2 Hifi Speakers 4 13
34 2
2 Settop Boxes
4 Color TV

TV+VCR Combos 4 6
DVD-player
5 5 5
Color-TV
CD-Player Memory cards MP3 Player
DVD-Recorder Audio Home
Camcorder Systems

€ 35.2 Bln. € 44,7 Bln. + 27%

Consumer Electronic assortment changing from 1998 to 2005

Also usage of Consumer Electronics has changed

Watching TV today might even be basically the same as 50 years ago, beside the fact, that it is
colored instead of black and white, however, it is more often.
Worldwide personal time of viewing TV programs is more than three and a half hour. More the 230
minutes a day one single person switches on the TV-set. The reason for that is easy to understand.
Beside the fact, that more programs are available (in Germany today app. 50 programs are available
free of charge, 30 years ago there have been only 3 available), content can be received 24 hours a
day.
But also social situation is changing. Number of elder households is increasing, number of
households with unemployed people is in creasing, number of single person households is also
increasing. And all these different points bring person to a situation to watch more TV:

GfK Group Marketing

Retail & Technology
Services Consumer Electronics Germany Jürgen
Jürgen
Boyny
Boyny October
29
07-02-2006
April 2006
2004

Average Daily Viewing Time

Per Individual
(30 global territories included)

in minutes + 13%

229 232
221 221 224 226
209 212
205 207 205

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005

Average Viewing Time per Individual 1985 – 2005

1160
But in total usage of TV has been changed. If we remember back, in former times family and maybe
friends came together in the evening to watch TV, Sitting down in the living room, maybe at 8 p.m.,
having some kind of social family event in front of the TV. Today, maybe a single person is coming
home in the evening after work, no direct contact person is available for conversation. Quite often
automatically the TV-set is switched on. Even that person is doing something complete different,
maybe having a meal in the kitchen, maybe even starting PC to check the emails, which came in. The
TV is mostly running all the time.
GfK Group Marketing
Retail & Technology
Services Consumer Electronics Germany Jürgen
Jürgen
Boyny
Boyny October
29
07-02-2006
April 2006
2004

Color-TV Multi Ownership Index

(2001)

2,00
1,84
1,76
1,80 1,68 1,64 1,62 1,57
1,60 1,54 1,54
1,44
1,40 1,29
1,20
1,00
0,80
0,60
0,40
0,20
0,00
GB I E S EU* A F NL D CH
Households
in mil. 24.4 20.7 12.4 4.2 130.5 3.3 22.6 6.8 33.1 3.0
*Westeurope (9)

TV Multi Ownership 2001

And also number of TV-sets in a household has been increased over the years. In Western Europe
we have had in 2001 a household penetration of TV-sets of 162%, which means app. 55% of the
household own more than one TV-set. This second TV-set might be located in the kitchen, in the
bedroom, in the children room or anywhere else.
GfK Group Marketing
Retail & Technology
Services Consumer Electronics Germany Jürgen
Jürgen
Boyny
Boyny October
29
07-02-2006
April 2006
2004

Location of TV Appliance
Total Ownership (%)

*West Europe (9)

2001
EU* D GB F I E NL S A CH
Living Room
Dining Room
Study
Guest Room 33
Master Bedroom 38
Childrens Room 52 53 51 54
Kitchen 60
67 62
Bathroom 71
Storeroom / Cellar 13
Other Room
1 27
No indication 1
16
8 5 10 1
1 3
2 3 1 1 3 3
2 2 1 2 3
2 17 1 2 3
14 15 8
3 12 17 20 2
3 12
26 4
10 3
9 13 7 7 4 8 10
9 11 3
9 8 9 1 5
11 4
1 2 10 1 13
2 2 31 25 1 2 4 2 2 3 2 1 3 2 2
Total Ownership
205,7 46,5 44,2 32,8 36,1 20,4 10,3 6,7 4,9 3,7
in mil.

Location of TV Appliance

1161
And we have to be aware, the more often usage of TV is only one part of increase use of Consumer
Electronics. Videorecorder came up, DVD-Player and Recorder came up, CD-Playe or Digital Dolby
Audio Receiver and at least the household penetration of PC at home.
Meanwhile we have in Europe 65% of the household, who own a PC. And also additional PC
environment have been added: a printer, a scanner, a digital camera – household penetration
meanwhile 55% - a webcam or a MP3-Player.
Even at the time it seems, that we have a re-launch of traditional consumer electronics, driven by Flat-
TV, DVD and MP3, during the 90ies the market have been grown enormous by the growth of the
information technology:
GfK Group Marketing
Retail & Technology
Services Consumer Electronics Germany Jürgen
Jürgen
Boyny
Boyny October
29
07-02-2006
April 2006
2004

Consumer Electronic Markets Europe*

Development in value 1990 - 2007

31% 30% 30% 32% 34%

35% 35% 35%
TV/Audio/Vide o 41% 39% 39% 40% 40%
45%
51%
59% 57%
63% 5% 5%
Photo 5% 6% 133,7
5% 7% 7% 8% 128,6
5% 123,0 122,5 123,9
7% 7% 6%
5%
6% 114,3 116,5 115,3 116,3
Information
6%
Technology
100,7
7% 49%
7% 50% 51% 49%
49% 46% 44% 43%
Tele communication 7%
43% 46%
88,1 40% 40% 41%
40% 81,0
35% 77,2
73,5
27% 29%
24% 64,6 64,9 66,8
60,2

13% 14% 16% 13% 13% 14% 14% 14% 13% 13%
7% 7% 8% 9% 10% 10% 10%
6%
Bill Euro
90

91

92

93

94

95

96

97

98

99

00

01

02

03

04

05

06

07
19

19

19

19

19

19

19

19

19

19

20

20

20

20

20

20

20

20

* Europe: D/F/GB/I/E/NL/B/CH/A/S

Enlarged market of Consumer Electronics to IT, Telco and Photo

With all these development we learn at least: as far as innovative products, totally new for the
consumers are offered, the consumer will try to catch this train as soon as possible. The consumer
seems at least to be fascinated by the technology, only by the technology, probably not by the
content.

1162
 Heating
and Water Heating

1163
1164
Tomorrow's' Heating Technology in the Light of Eco-Design and
Labelling
Thomas Behringer

Chairman Technical Commission EHI, association of the European Heating

Industry

Abstract
Heating appliances have an important environmental impact mainly due to their energy consumption.
In Europe the variety in households, building types and climate conditions requires specific
responses. State of the art and future heating technologies can provide solutions to these challenges.
Several European directives and a multitude of standards define the legislative framework. The
saving potential of boilers and water heaters is discussed by policy makers, alongside relevant trends
in building insulation levels, fuel switching, integration of renewable energy sources leading to new
and improved technologies, etc.
Yet, the visibility of the product group is low and product life is long, often leading to a lack of
awareness on the potential energy saving. Obviously the economics of Least Life Cycle Costs –the
payback a higher purchase price against lower running costs does not automatically lead to market
transformation. Both conventional and innovative technologies offer advantages to achieve better
efficiencies. Member states have developed policy instruments and the relevance of these will be
discussed in terms of their results. Using the example of a base case representing a typical
household in Europe financial investment, operational costs and energy savings are validated.
Lessons learned from these evaluations may provide for the development or re-evaluation of policy
scenarios.

Who are we?

Founded in 2002, the Association of the European Heating Industry (EHI) represents and promotes
the common interests of 18 market leaders producing heating systems, burners and boilers, and 13
national industrial associations from the EU Member States, Liechtenstein and Switzerland. In
Europe, the heating industry directly employs approx. 120,000 people. Besides big players such as
BALTUR, BAXI, BBT, CARADON IDEAL, DE DIETRICH, ENERTECH, GIANNONI, HONEYWELL,
HOVAL, IMMERGAS, MERLONI TERMO SANITARI, RIELLO, SIEMENS, SIT, VAILLANT,
VIESSMANN, WEISHAUPT and WOLF most companies are small and medium sized enterprises.
Heating industry contributes € 20 billion to the European Union’s GDP.
Approximately one third of energy consumed in the EU is used to heat buildings. As a result, the
heating industry has a key role to play in helping to protect the environment and in developing energy
efficient instruments. Maximising the potential for saving energy and reducing CO2 emissions is
considered a top priority for EHI members. The industry has therefore invested massively in research
and development, to create technically advanced, energy efficient heating systems.

Status quo – Role of heating and hot water production

As mentioned before, roughly one third of energy consumed in the EU is used to heat buildings
including hot water production (in the following text hot water production is generally included in the
term ‘heating’ if not indicated otherwise). Despite the fact that its importance is understood by
European Commission and specialists in some national administration bodies, others often lack
expertise or are simply not interested. The visibility of the boiler may play an important role, as they
are often found in basements and therefore the boiler is ‘forgotten’. Another ‘psychological’ reasons
may be the long and quiet running times - over decades - of those heat generators. The fact that very
few people are confronted with the appliance in day-to-day life makes them not very ‘thrilling’ and less
exciting than e.g. cars.
However, at least once a year private house owners have to pay the heating bill and this may be
cause for some excitement, at least from the financial point of view. As an example in one middle
Europe country, the total expenditure for space and water heating was over € 34 billion (€ 34.239
million) Euro in 2005. This is almost 50% more than 10 years ago. Overall, these households spends

1165
around € 2.300 annually in energy costs with heating and the car each are 40% of that sum. In a
larger single family home of e.g. 150 m² with –as isoften the case in the building stock— sub-optimal
insulation, the heating costs alone can amount to sums of over € 2000,- (e.g. 40 000 kWh) per year,
based on May 2006 energy prices. It is often forgotten that this substantial amount of money has to
be paid just to keep the indoor temperature of a building at a certain level in the cold periods of a
year. And it is too easy just to blame the boiler, but is the total package of building insulation,
consumer behaviour and heating installation that determines the bill. And at current trends of ever
increasing energy prices this bill is very likely to increase in the near future. It s not difficult to imagine
that most households would then run into serious financial trouble…
But it is not only from the financial and thereby social policy point of view that boilers and water
heaters are exciting. It is no secret that there are enormous potentials to save both fuel and CO2 in
the domestic sector across Europe. CO2 emissions are proportional to energy consumption, that is
0,27. kg/kWh for oil and for gas 0,194. kg/kWh. Ways to achieve this in a relative short period of time
and without reducing comfort – which is a precondition here – will be discussed below.

What the customer, legislation and politics wants

There are a large number of requirements and wishes from the different stakeholders affecting the
heating industry and its products. A selection of those requirements, mainly related to legislation, is
listed in the following table.
Parameter Require- EC legislation
ment
Efficiency High • 92/42 Boiler Efficiency Directive (BED)
• 93/76 Limiting of CO2 emissions (SAVE)
• 2002/91 Energy Performance of Buildings
Directive (EPBD)
Emissions (NOx, CO, …) Low • some member states (national) legislation
Lifetime High • 2005/32 Eco-design of energy using products
(EUP) – lifetime is discussed in this context as
well as most other parameters in this table
Include renewable energies High • 2002/91 Energy Performance of Buildings
Directive (EPBD)
• ??? RES-H
Recycling High • 2002/96 Waste of Electric and Electronic
Equipment Directive (WEEE)
Eco-friendly materials High • 2002/95 Restriction of Hazardous Substances
Directive (RoHS)
Noise Low • some member states (national) legislation
Hygiene, Health High • 98/83 Drinking Water Directive (DWD)
Transparency of efficiency High • 92/75 Energy Consumption Labelling Directive
(ECLD)
Price for product Low -
Safe operation High • 89/336 Electromagnetic Compatibility (EMCD)
• 89/392 Machinery Directive (now 98/37)
• 90/396 Gas Appliances Directive (GAD)
• 97/23 Pressure Equipment Directive (PED)
• 89/106 Construction Products Directive (CPD)
• 2001/95 General Product Safety Directive
(GPSD)
Compactness High -

It is clear from the table that the number of regulations for heating products is relatively high and that
for many parameters there is a strong correlation between the product price and the technical
requirements. To complete the picture, the additional requirements that may arise from the
2005/32/EC directive on Eco-design of energy using products (EUP) should be mentioned.
This framework directive will possibly have the biggest influence in the near future, and not just on
heating products. Main drivers for the EUP directive are the Kyoto protocol and European Climate

1166
Change Programme (ECCP), but also several other EU environmental objectives and of course the
objective of a single EU market (Art. 95).
As a first reaction after publication of the draft EUP directive the heating industry argued that such a
directive would just increase the level of regulation. Taking into account that some of the above listed
EC directives such as the BED and ECLD did not yet overcome some of their inherent weaknesses it
was criticized that just a another new directive would be squeezed into the hierarchy of legislation. But
after an intense dialog with the European Commission also chances emerged. If the implementing
measures under the EUP directive could clarify and improve existing directives under one ‘umbrella’,
industry and ultimately the consumer could gain considerably. The single EU market would certainly
be strengthened if
• Gaps of existing legislation would be closed
• Overlapping of existing legislation would be avoided in the future (CPD, GAD, EPBD, BED,
ECLD)
• National relics such as German Ü-marks and Belgian HR+ label would disappear
• Double and multiple efforts such as UK SEDBUK database would be unified / standardized
One could argue that almost every other sector of industry should be affected and has to do his
‘homework’. In that context also the automotive industry with their pretty low ‘well-to-wheel’ efficiency
should be considered. Consider power generation technology with a high potential to improve on
today’s relatively low efficiency. And the energetic improvement of the building stock, not just the
heating installation, should fit that picture. Technically speaking, in the field of heating technology
considerable progress has already been made and quite often physical limits have been achieved.
This will be shown in the next paragraph.

Status quo of technology – what has been already achieved?

Parameter State of the art Remarks
Efficiency Nearly 100% (condensing) The longer the operation times the
shorter the payback time, i.e. some
countries with long heating periods
use more and more condensing
technology, for others in southern
areas the situation may be different
Emissions (NOx, CO, etc) Down to some 10 ppm (best Low level if compared to automotive
practice) and industry sector
Lifetime 15-25 years Corresponds to lifetime in industrial
plants
Include renewable energies Standard Most modern heat generators can
be combined with solar hot water
collectors
Recycling Quote of recycling is high Since lot of valuable materials are
used such as brass and copper
almost all materials will be reused
(recycling)
Eco-friendly materials High See above
Noise Low Products are often installed near
living rooms therefore low noise
level is a design criterion
Hygiene, Health High In most countries boilers heat up
drinking water
Transparency of efficiency Good Today most professional product
documents (data sheets) already
contain efficiency information
Price for product Low Full range from “low budget to
premium products”, on average low,
very strong competition
Safe operation High Safe operation is priority #1 design
parameter (product liability)
Compactness High (if wanted by wall hung boilers
customers)

1167
The table shows the already high level of technology achieved by modern heating technology.
Modern designs, materials and sophisticated controls ensure a reliable, low emission and highly
efficient operation of heating generators.
In other words, the heating industry doesn’t see the technology as a problem in meeting the highest
environmental standards. However, there are several barriers of another nature that may stand in the
way. For instance:
1. The heating market is to a large extend –typically over 70-80%-- a replacement market. And
most customers are just looking for that: A simple 1:1 replacement of the old equipment for
the new equipment. In that context, already the replacement of e.g. a standard boiler by a
condensing boiler is considered a huge step, despite the fact an individual customer could
stand to gain over 30% of energy saving and the economics are obvious.
2. For some legislators –however—this is not deemed enough. It is often felt that also existing
houses should be equipped with the best front-runner technology, imposing the customers by
law to a completely new concept of how they should heat their house or produce their hot
water. This could cause substantial friction in the existing housing market
3. This friction is aggravated by methodological problems with the legal instruments being
developed. For instance, the mandatory energy labelling of water heaters is on the political
agenda since 1992, but has suffered a series of setbacks due to methodology. The evaluation
of tapping patterns, in- and outlet temperatures, etc. across Europe is not a self-evident issue
that has required considerable effort not only by the European Commission but certainly also
by the industry. Another example is posed by combi-boilers and especially indirect cylinders,
where the difficulty lies in having to evaluate a part of an installation, rather than the whole
system. The assessment of auxiliary energy or the problem of tolerances within the narrow A-
G bandwidth are other problems that need to be tackled.
4. Against that background, the latest ideas on labelling of hot water appliances –enlarging the
scope to include also solar water heaters—are surprising and will introduce new problems to
be solved, relating to the different climate zones in the EU and the wide variety of possible
combinations with conventional technology.

The future – what can still be done

Let us now briefly state principal ways to achieve energy and CO2 savings - without reducing comfort:
1. Increase building insulation and make buildings more airtight
2. Adjust all existing heating equipment properly, especially controls
3. Use combined heat and power (CHP) technology to heat the building
4. Replace old and inadequate boilers
5. Use renewable energies for heating

While measure #1 is outside the scope of this paper, it certainly makes sense to take a holistic
approach and combine building insulation with the other measures listed. Energy savings from
insulation of an existing building can lead to energy savings of 50% or more. However, in most cases
the payback time is longer than what can be achieved with improvement of the heating installation.
Just to provide an order of magnitude: A complete insulation (walls, roof, floor, glazing) of the 150 m²
single-family house described earlier would cost more than 30.000 Euro in middle Europe for
technical reasons the insulation thickness is limited to 10-15 cm for existing buildings. At 50% energy
saving, the simple payback period (SPP, i.e. without taking into account interest) would be as high as
30 years.
Concerning measure #2, studies have revealed that typically up to 10% energy saving in a single
installation can be achieved by simply adjusting all the heating equipment properly – especially the
controls of a boiler. To realize this, the heating industry is permanently training installers and is
improving the technology. Examples are better control designs such as self-adjusting electronics, new
sensors to automatically check combustion, Internet remote control of complete boiler operation,
sophisticated measuring devices such as energy efficiency monitoring systems. But in the end,
responsibility is more on installers than on industry. Therefore official support of authorities to better
train installers and increase awareness of both installers and end customers on that topic would be
welcomed.
Regarding measure #3 one can state that most of (big) companies already offer CHP products.
Today’s most reliable technology is the conventional internal combustion engine that is combined with
an electricity generator. Both engine and generator are (heating) water cooled, encapsulated together
and designed to run up to 40.000 hours without replacing the engine. Electrical power output is down

1168
to 1-2 kW for single houses, total system efficiency up to nearly 90%. Bigger units can be used in
multi family houses. Depending on what is taken as a reference situation and in ideal operating
conditions, energy savings can be up to 25% and 30% in CO2 reduction. Unfortunately, this
technology is not yet widespread because of certain installation requirements and relatively high
investment costs. To reach the aforementioned saving potential, the installations need steady
electricity demand and simultaneous heat demand. Not every installation can provide this. Alternative
CHP technologies are Stirling driven generators and fuel cells as most promising future devices.
However, it will take at least another 5 years to make a final statement on the competitiveness of fuel
cells.
Replacement of old boilers (measure #4) is the focus of today’s heating industry. Since in many
northern European countries the number of new buildings per year will decrease, the replacement
market –which is already by far the largest segment—will become even more important. Internal
studies of EHI on Eco-design of conventional boilers were conducted largely using the methodology
developed for the Eco-design of EuP directive studies. These studies on the life cycle of the product
confirmed the dominance of the use phase in terms of environmental impact, i.e. energy needed for
production of boilers is small compared to energy consumed by the boiler over its lifetime. As a
consequence, better efficiency of new boilers is essential. Based on actual replacement figures in the
various member states it was shown that for EU-15 countries1,5% of total heating energy per year
can be saved. Over a 10-year period this amounts to as much as 15% saving. Since these numbers
are based on actual sales figures and realistic customer behaviour it is strongly recommended to
policy makers to support measures in the field of appropriate boiler-replacement.
Regarding the use of renewable energy sources for heating purposes--measure #5-- it is essential
that policy makers combine ambition with a realistic view. It is often said that renewable energy
sources alone will solve all energy problems. And in the wake of such a statement, conventional boiler
technology is immediately dismissed as ‘yesterday’s technology’. Probably these indiscriminate
statements make ‘good politics’, but it would be nice for both the voters and the environment if they
had a basis in reality. It is true that solar thermal collectors can provide users with a fairly good
2
amount of energy saving for hot water production. A solar thermal collector area of typically 5-10 m
can cover 50-60% of total hot water production of a household in Central Europe (depending on
collector technology and quality used). In new buildings and in Southern Europe, where hot water
production may constitute up to 40% of the total heating energy, this is a significant amount. However,
in the existing buildings, constituting the vast majority of the heating market, with their significantly
lower level of insulation, the hot water production constitutes only 10-20% of the total and the use of
solar energy would result only in savings of 5 to 10% of that. This effect is not to be ignored, but it
won't solve all energy problems. If the user wishes to use active solar systems to assist his space
heating things become even more difficult, especially in existing buildings. To realize e.g. 10%
coverage from a solar system for space heating not only a larger collector surface is needed, but also
one or more really big (and costly) heating water storage cylinders are required. As long as no long
term and high energy density system to store heat is available, most people would need considerable
subsidies to install such systems.
Another technology to be mentioned here only briefly is heat pumps that are normally driven by
electric motors. Environmental heat is used in a thermodynamic cycle to achieve total efficiencies up
to 130%. Heat pumps need an environmental heat source such as earth heat collectors, ground
water, geothermal heat sources or simply the outside air. They are normally restricted to modern or
new (insulated) buildings with relatively low heat demand and low heating system temperatures.
The next promising technology also discussed here only briefly is using wood as a fuel. Most
sophisticated technology is the “pellet” oven, using very small pieces of before processed wood.
Generally, wood fired boilers or ovens can substitute fossil fuels. Yet, there are several issues to be
solved: The large-scale supply of biomass is one issue, the long-term behaviour of the appliances is
another. Furthermore, although the technology is beneficial to the CO2 –balance, the emission of
other pollutants (e.g. particulate matter) especially in urban areas is still an issue to be carefully
considered.
All in all, of the measures mentioned above, conventional boiler technology, i.e. high efficiency heat
generators such as modern low temperature or condensing boilers are the most effective solution in
big replacement markets. With respect of measures such as building insulation it is well established
that for the existing building stock the payback time for improving the heating system is much better
than that of insulating the building shell. With respect of the development of renewable energy
sources, the heating industry is fully committed. Therefore if possible, the new boiler should be
combined with solar hot water production since this technology is already mass produced The routes
of heat pumps and CHP will continue to be explored and applied whenever appropriate. But it must be

1169
recognized that not every installation can be easily and immediately adapted to the very latest
technology and that in some cases this technology that looks good on paper still is not fully mature.

Conclusions
The European heating industry is fully committed to building a sustainable and energy-efficient future.
Having said that, the industry would like to urge policy makers to develop a balanced package of
measures to achieve environmental objectives, reflecting the reality sketched in this paper. And this
balance would have to be found in more than one way:
Policy makers should not just focus on restrictive legislative measures, but also develop indispensable
lateral measures regarding
• Installer training, raising the standard of expertise in the field of energy –efficient installations
and controls
• Subsidies and direct rebates to allow the low-income groups to benefit from the energy saving
measures in a responsible way,
• Financial incentives and tax rebates to support house owners facing extra investments to
make their building and installation suitable for the new technologies,
• Appropriate information campaigns to the end users,
• In as much as labelling is used it should be fair and transparent,
• In as much as legal measures are deemed necessary they should be aimed at simplifying the
current and future body of national and EU legislation rather than rendering it more
complicated, and in that context the measures should be directed to truly developing the
single EU-market.

Policy makers should also not just focus on the more spectacular ‘green’ heating alternatives, but
should direct their measures and budgets in function of effectiveness of the technologies for the
society as a whole. And their commitment to these objectives and measures should not be ad-hoc
and short-term, but structural and long-term. This would allow the industry and all other actors –not in
the least the installers—to build an economically and ecologically sustainable strategy for the future.

1170
High Efficiency Circulators for Domestic Central Heating Systems
Niels Bidstrup1, David Seymour2
1
GRUNDFOS Management A/S,
2
Grundfos Pump, UK

Abstract
High efficiency circulators for commercial buildings have been on the market since 2001. Now high
efficiency circulators are also available for domestic central heating systems. These circulators can
save up to 80% electrical energy compared to conventional circulators installed today. This is
achieved by using high efficiency permanent magnet motors and speed control. To increase market
share of circulators with higher efficiency, a labeling scheme has been introduced. Circulators are now
labeled with A-G energy label - well known from white goods and household lamps market. The
labeling scheme came in force in March 2005 and is controlled by a voluntary industry commitment
agreement which is managed by Europump It is estimated that 120 million domestic circulators are
installed in EU25 today. These circulators are responsible for up to 15% of the electricity consumption
for some European households. The average energy efficiency of installed circulators today
corresponds to label “D” or “E”. If those circulators were changed to “A” labeled high efficiency
circulators the electrical energy saving potential in EU25 could be 44 TWh per year, with a reduction
of 17,6 million tonnes CO2 per year. This paper describes how energy labeling in combination with
national schemes can increase the market share of high efficiency circulators for domestic central
heating systems.

1. Background
Energy consumption of circulators has been high on the agenda for the last 10-15 years. The reason
for this is the huge energy saving potential, which could have a significant influence on CO2
emissions. Circulators consume a lot of energy due to the high running hours in some countries and
the very large installed numbers. In SAVE II [1], the installed base in EU15, was estimated at 87
million units. Based on this figure the pump industry estimate the current installed base in EU25 is
120 million. The estimated energy consumption of those circulators is 57 TWh per year, assuming a
heating season of 285 days.
The typical lifetime of a circulator is 10-15 years, but circulators, which are 15-20 years old, are also
operating today. These old circulators are less efficient and are responsible for up to 15% of the
electricity consumption of some European households. By replacing the old circulator it is possible to
save up to 10% of the electricity consumption for these households.

2. High efficiency domestic circulator

High efficiency circulators for commercial buildings have been on the market since 2001 [2]. Since
2005 high efficiency circulators are also available for domestic central heating systems. Figure 1
shows a high efficiency and a standard domestic circulator. A high efficiency domestic circulator looks
similar to a standard domestic circulator from outside and is interchangeable with these. Internally the
induction motor is replaced by a permanent magnet motor and the terminal box contains a full three
phase frequency converter with motor control, system control and intelligent adaptive functions.

1171
Figure 1 High efficiency and standard domestic circulator

A high efficiency circulator uses considerably less energy per year than a standard circulator. This is
not only achieved by higher efficiency in the component i.e. pump and motor, but also due to speed
control. Speed control enables the circulator to adapt to the changing demands in the heating
systems, which would otherwise result in hydraulic loss in control actuators i.e. thermostatic radiator
valves etc.
Speed controlled circulators for domestic heating systems have shown savings up to 50% compared
to a standard circulator. A high efficiency circulator combines speed control with a permanent magnet
motor, which has higher efficiency especially at part load. This results in energy savings up to 80%
compared to the circulators installed today.

3. A-G energy labeling of circulators

Speed controlled circulators have been on the market for approximately 15 years. Despite short pay
back time for the end user (typically 2-3 years in replacement situations) less than 20% of circulators
sold today are speed-controlled. To increase market share of circulators with higher efficiency a
labeling scheme has been introduced. Circulators are now labeled with A-G energy label well known
from white goods and household lamps market. The labeling scheme came in force in March 2005
and is controlled by a voluntary industry commitment agreement [3], which is managed by Europump.
Seven companies with a total market share of more than 80% have signed up to this commitment
agreement so far.

3.1 Classification scheme

A-G Energy labelling of circulators is based on a classification scheme developed by Europump. The
classification scheme is summarized below. A detailed description of the theory behind it is given in
[4] and [5]. The A-G labeling comprises all circulators for heating systems, which fulfill all conditions
below
• Stand alone (not an integral part of a boiler)
• Pump and motor integrated
• Wet runner (glandless)
• Centrifugal pumping
• P1<2500 W

Stand alone circulators with pump and motor integrated are circulators which are sold as a separate
products and not as an integral part of, for example, a boiler. Wet runner means that the rotor is
running in the pumped fluid. Only circulators with a power input P1<2500 [W] (for every head on
double pumps) and based on centrifugal pumping principle are comprised.
Circulators are labeled based on an Energy Efficiency Index (EEI). the EEI is calculated as
PL ,avg
EEI = [−]
Pref
,where
PL,avg: is the average compensated power input
Pref: Reference power input
PL,avg is a weighted averaged power input based on a yearly load profile and compensated for control
error. Pref is the reference power input for a standard circulator at a specific size. By dividing these

1172
two figures the EEI expresses how efficient a specific circulator is compared to the circulators on the
market in 2002. The system is calibrated as shown in table 1

Table 1 Calibration of EEI to A-G energy label

Class Energy Efficiency
Index (EEI)
A EEI < 0.40
B 0.40 ≤ EEI < 0.60
C 0.60 ≤ EEI < 0.80
D 0.80 ≤ EEI < 1.00
E 1.00 ≤ EEI < 1.20
F 1.20 ≤ EEI < 1.40
G 1.40 ≤ EEI
This calibration implies that circulators on the market in 2002 will get a “D” or “E” label (EEI=1.00). To
get an “A” label the average power input must be reduced by at least 60% compared to that.
The label is shown in figure 2.

Figure 2 EU A-G energy label for circulators

Only information about average power consumption is on the proposed label, which is indicated by
the letters and arrows on the label. The yearly energy consumption depends on running hours of the
circulator, which depend on the heating system.
The average energy efficiency of installed circulators today corresponds to label “D” or “E”. If those
circulators were changed to “A” labeled high efficiency circulators the electrical energy saving
potential in EU25 could be 44 TWh per year, a reduction of 17,6 million tonnes CO2 per year.

4. Energy consumptions and savings – an example

Energy consumption and savings vary a lot between installations and depends on circulator age,
dimensioning, setting and running hours. In the following example a comparison between different
circulator options in a specific heating system is made. The idea with this comparison is to show the
energy consumption and savings in an ideal case where pumps are dimensioned and set correctly.
Actual consumption and saving would be higher in most systems, where the pump very often is too
big and/or the setting is too high.
The heating system for this example is defined as
• Dimensioned heat load: 15 kW

1173
 • Dimensioned flow temperature: 75 °C
• Dimensioned return temperature: 60 °C
• Dimensioned pump flow: 860 l/h
• Dimensioned pump head: 2 m

Calculations are based on a heating season of 285 days, which is a typical heating season in mid
EU25.
Four different circulator options are chosen. A “C” labeled fixed speed circulator set on speed 2 (mid
setting), which is 1-10 years old. A new “B” labeled fixed speed circulator set on speed 2 (mid setting).
A high efficiency “A” rated circulator operated as fixed speed circulator in speed 2 (mid setting). This
is an option if uncertainty about boiler minimum flow rate exists in relation with variable speed mode.
The last option is an “A” rated high efficiency circulator operated as variable speed.
In table 2 the annual energy consumptions of the different circulator options are calculated. The load
profile is the same used for EEI calculation. Notice that circulators only operate at 100 % flow 6 % of
the time and are below 50 % for nearly 80 % of the time. This is due to the distribution of outdoor
temperature and the nonlinear characteristics of the heating system.

Table 2 Yearly energy consumption of different circulator options in a specific system

Fixed speed New Fixed speed High efficiency High efficiency
circulator <2005 circulator >2005 circulator circulator
(fixed speed) (variable speed)

UPS 25-40 UPS 25-40 Alpha Pro 25-40 Alpha Pro 25-40
Load Profile Labelled: “C” Labelled: “B” Labelled: “A” Labelled: “A”

Flow Time P1 [W] E P1 E P1 E P1 E

[%] [%] [kWh] [W] [kWh] [W] [kWh] [W] [kWh]
100 6 44 18 34 14 25 10 19 8
75 15 44 45 33 34 25 26 15 15
50 35 42 101 32 76 25 60 11 26
25 44 41 123 31 93 25 76 9 27
∑E 287 ∑E 217 ∑E 171 ∑E 76

The power input shows that a high efficiency circulator only draws 9W at 25% load, where the
corresponding fixed speed circulator draws 41 [W]. Most circulators installed today draw between 60 –
100 W at 25 % load.
The bottom line of the table shows the yearly energy consumption in bold letters. A high efficiency
circulator in variable speed mode consumes only 76 kWh per year instead of 287 kWh per year, which
is a saving of 73 %.
These results are also shown in the figure 3 below, where consumptions in percentages are also
depicted.

1174
 Yearly energy consumption

350

300 287

250
217
E [kWh]

200 171

150

100 76

50
100% 76% 60% 27%
0
Old fixed speed New fixed speed High efficiency High efficiency
(Fixed speed) (variable speed)

Figure 3 Comparison between different circulator options in a specific system

4. Promotion of high efficiency domestic circulators

A-G labeling is first of all a marketing tool directed towards end-users, but labeling can also be used
to set minimum standards and targets, which is easier to monitor and evaluate. From that point of
view labeling is a useful tool for promotional campaigns on national and EU level to accelerate market
transformation towards circulators with higher efficiency. Different national schemes are already using
these possibilities.

4.1 Denmark
Danish ELFOR is running a campaign to change all pumps towards better efficiency. A key element in
this campaign is the “List of Energy Efficient Pumps” [6], which is a website that enables installers and
consultant engineers to identify and select the right energy saving pump for a specific installation.
Only circulators with rating “A”, “B” and “C” are on the energy saving pump list, which means that
labeling is used to set minimum standards. Furthermore the actual labels for the different circulators
are also displayed, which make it easier to select the most efficient circulators for a specific job.
Another campaign is carried out by the Energy Saving Trust in Denmark. This campaign is focused on
domestic circulators and directed towards the end-user. TV commercials and website [7] are used to
increase awareness of old energy wasting domestic circulators and how to select an energy efficient
circulator by using the A-G energy labeling.
It is difficult to measure the impact of different campaigns separately, but today 50 % of domestic
circulators sold are energy saving circulators with the majority belonging to class “A” or “B” and a few
in class “C”. Before A-G labeling and these campaigns only 20% of the circulators was energy savings
circulators belonging to class “B” and “C”.

4.2 UK
In January 2006 the Energy Saving Trust approved the Grundfos “A” and “B” rated circulator as
Energy Saving measures. The labeling scheme and EEI were used to classify the measures and
using the UK SAP assumption of 2000 running hours, the savings were sanctioned as 78 and 52 kWh
respectively. This enabled Grundfos to approach the UK Energy suppliers to secure promotional
funding from the Energy Efficiency Commitment scheme operating in the period 2005 to 2008. These
new measures were particularly attractive as the savings are attributed to electricity not gas and oil.
Grundfos will work with Energy Suppliers to create market transformation in the new and replacement
circulator market.

1175
5. Conclusions
High efficiency “A” labeled circulators are now available for domestic central heating systems. These
circulators can save up to 80% electrical energy compared to conventional circulators installed today.
An example calculation showed that even in an idealized case, where standard circulators were
dimensioned and set correctly, savings of 73% were achieved by using a high efficiency circulator in a
certain situation
The A-G labeling scheme of circulators came in force in March 2005. The labeling is controlled by a
voluntary industry commitment agreement and is managed by Europump. Seven companies with a
total market share of more than 80% have signed up to this commitment agreement today
It is expected that the A-G energy labeling will have a great impact on the circulator market in EU25 in
the coming year, especially if combined with other market transformation programs. In Denmark A-G
labeling is combined with two different promotional programs and the sale of energy efficient
circulators have increased from 20% to 50%.
The average energy efficiency of installed circulators today corresponds to label “D” or “E”. If those
circulators were changed to “A” labeled high efficiency circulators the electrical energy saving
potential in EU25 could be 44 TWh per year, with a reduction of 17,6 million tons CO2 per year.

References
[1] Bidstrup, N., van Elburg, M. and Lane, K., Promotion of Energy Efficiency in Circulation Pumps
especially in Domestic Heating Systems, Study under SAVE II program, June 2001.
[2] Bidstrup, N. , A New Generation of Intelligent Electronically Controlled Circulator Pumps, 2nd
International conference on Improving Electricity Efficiency in Commercial Buildings (IEECB),
Nice 2002
[3] INDUSTRY COMMITMENT – To improve the energy performance of stand-alone circulators
through the setting-up of a classification scheme in relation to energy labeling, Europump 2005
[4] Bidstrup, N., Hunnekuhl, G., Heinrich, H. and Andersen, T., Classification of Circulators,
Europump report, January 2003.
[5] Bidstrup et al. Classification of circulators, Proc. of the EEDAL’03 (Turin, Italy 2003)
[6] http://www.sparepumpe.dk
[7] Http://www.elsparepumpe.dk

1176
Electricity in Non-electric Central Heating Systems
Nick Davies

UK Market Transformation Programme

Abstract
In a house heated by a gas or oil boiler, about 8% of the annual electricity consumption of the
household is required to operate the heating system. This is about the same level of electricity
consumption as the Consumer Electronics Sector (9 TWh/yr in UK). Yet the electrical components of
typical heating systems (fans, pumps, motorised valves, boiler electronics, controls) have received
scant attention in terms of power demand and power economy. Installer choices based on price, ease
of fitting and familiarity also impact on the electricity consumption, but there is little information
available for the consumer to make informed choices. Until recently there was no comprehensive
model, measurement method nor analysis of the prospects for improvement.
This policy gap has been recognised in the European SAVELEC project (completed in June 2005),
and the findings from the project show that worthwhile savings are possible, and could be achieved by
an agreement to cover measurement, labelling, provision of information and a commitment to improve
design. The paper will explore these and other opportunities, particularly with reference to the UK
market (1.5M domestic boiler sales per year).
This paper will summarise the findings of the SAVELEC project, outline the common system
components, the scope for energy improvement and discuss options for the way ahead.

Common heating system electrical components

The common electrical components were identified along with their typical instantaneous power
consumption and estimates of market levels, where available. BRE contacted several UK boiler and
system component manufacturers and reviewed industry datasheets. The other project partners within
this task; GdF, DTI, VTT and OWI presented information from their respective EC countries using a
similar method. Table 1 presents the most common system components for gas and oil fired systems.
Table 1: common system components and power ratings
Component Typical Typical Consumes if Consumes if Consumes if system on
instantaneous instantaneous system off? system on and boiler firing?
power - Watts power - Watts boiler not
BRE (UK) Other partners firing?
Pumps 65 GdF 55, SAVE[*] N Y Y
70
Boiler fans 50 30 N N Y
Zone valves - 2 port, 3 5 6W during state 3 port 5W, 2 Y Y
port change port 0W
Programmers 2 0.4W; 2 – 6VA Y Y Y
Thermostat(s) Note1
Gas valves ~7 ~7 N N Y
Keep hot heaters 15 – 30 25 Y N N
Boiler standby power 8 <15 ’Blue angel’ Y Y Y
Igniters Note 2 2 N N Y for 10 s typically
Oil preheat Note 3 40- 150 N N Y for 50 s on cold start
Oil – pressure type 90 - 200 N N Y
Oil - atomisation type 75 N N Y
Oil igniter Note 2 60 N N Y for ~35 seconds
[*] SAVE is the earlier EU SAVE II Project on the Promotionof Energy Efficiency in Circulation Pumps, especially
in Domestic Heating Systems (Contract No 4.1031/-Z/99-256)
Note 1: bimetallic strip types have a small ‘shunt accelerator heater’ – generally a ¼ watt resistor ~ 250kΩ
resistance to reduce hysteresis. Power dissipation at 230 volts is around 0.2 W so is assumed negligible.
Note 2: Consumption is assumed negligible - gas boilers the duration is 5-10 s, for oil upto 35 s.
Note 3: Light fuel oil is typically preheated to 70°C. Power required varies with the oil flow rate. VTT found the
range to be 54W – 305W for boiler powers 17.8kW-101kW with underground oil tanks. The operating time with
cold start is around 50 seconds. Warm starts do not need preheating. Therefore consumption is assumed
negligible.

1177
There is wide variation in instantaneous power consumption in some instances. Worthy of note is the
large variation of oil fired boiler components with thermal power rating and the consumption of some
components when the system is off.

Estimated system annual electricity consumption

Many of the components have a wide range of instantaneous power ratings. To put this into context
the operating hours need to be considered and this depends on many factors, including system type,
plant size ratio (boiler output/heat load), local climate and length of heating season. Fig.1 shows the
spread of expected annual consumption for six common UK gas systems with the same heating
season of 250 days. The heating and hot water programmes are the same in all cases (CH + DHW 10
hrs for 250 days, DHW 2 hours for 165 days).

Annual electricity consumption

450
400
350
300
kWh/yr

250
200
150
100
50
0
1 1a 2 2a 3 3a 5 5a
System type

Fig.1 Electricity consumption / system type for 6 UK gas systems.

• Type 1 shows a typical cast iron boiler with gravity hot water and poor controls, 1a the effect
of a room thermostat controlling the pump.
• Type 2 shows a natural draught boiler with a fully pumped system and 3 port mid position
valve, type 2a is the same system with two 2 port spring return zone valves.
• Type 3 is the same system as type 2, but with a fan-flue fixed output boiler.
• Type 5 is an instantaneous combination boiler without, and 5a with, room thermostat.

Fig.1 clearly shows the effect of system type on the annual electricity consumption. This typically
varies from 135 kWh for a semi-gravity system through to around 415 kWh for a combination boiler
system with temperature control by TRVs and bypass. Although this shows the simplest systems with
poor controls have the lowest electricity consumption, the thermal efficiency of systems with better
controls is higher.

1178
Annual consumption by system components
In order to prioritise the saving potential, the proportion used by each component becomes important.
Taking system 3a in Fig.1, which uses 241kWh per year, the consumption proportion is shown in
Fig.3. This is indicative of many systems designed to current good practice using widely available
components. The system comprises a regular boiler with fixed speed fan flue, no standby power, fixed
output burner and separate pump and programmer. The heating system is fully pumped with interlock
by room and cylinder thermostats controlling 2 no. 2 port motorised spring return valves. An example
schematic for such a system is shown in Fig.2.

Fan flued
Boiler

Heating 2 DHW 2
port valve port valve

Radiator circuit – sized for

heating load

Fig.2 Simplified UK system 2a with fully pumped circulation and two 2-port motorised valves

Typical electricity consumption system 3a

4%
7%

32% Motorised valve

programmer
pump
gas valve
4% boiler fan
53%

Fig.3 Annual electricity consumption by component % for a typical UK domestic system

Fig.3 shows the pump represents 53% of the electricity consumption (124 kWh/a), the boiler fan 32%
(77kWh/a), programmer 7% (18kWh/a), motorised valves 4% (11kWh/a) and gas valve 4%
(11kWh/a). The following section considers scope for improvement prioritised in order of power
consumption.

1179
Scope for component energy improvement
Pumps
Fig.3 shows the pump typically uses 53% of the system electricity or 124 kWh annually. The earlier
EU SAVE project (see table 1) identified savings via improved efficiency and better control. The report
also suggested conventional pump efficiency was around 18% and that this could be improved with
design changes, in particular the use of permanent magnet motors. Some of these design changes
are now commercialised, and permanent magnet (PM) pumps became available on the market in late
summer 2005. PM pumps offer similar head and flow rate performance but potentially using only 30%
of the electricity of the conventional type.
The PM pump requires power-switching electronics to control the stator field so it is easier to add
speed control based on system pressure and this is included on some designs. Replacing a
conventional 65W pump with a PM pump would use around 25W for similar performance. If this
replaced the UK pump in Figure 2, the energy used would be 48 kWh/a, a 61% saving on the 124
kWh/a used by a conventional pump.

Boiler Fans
Fig.3 shows the boiler fan typically uses 32% of the system electricity or 77 kWh annually. Boiler fans
are typically the radial blower type, driven by a shaded pole motor which are cheap, well proven and
widely available but have poor efficiency and limited speed control. Modulating burners usually have
continuously variable fan speed and hence the ac shaded pole motor becomes less attractive in this
application.
DC motor speed can be varied by efficient power electronics giving better opportunity for reduced
electricity consumption, with brushless motors becoming common. The motor is smaller and inertias
are kept low since permanent magnets are used in the rotor. Less heat is generated in the rotor since
there is no current flow, and any heat that is created is dissipated to the surroundings more quickly
than a brush dc motor as the stator winding is on the outside, often fastened to the casing. Typical
efficiencies are approximately 90 - 95%, and generally efficiencies are 5 -10% higher than for a typical
ac induction motor.
Many fan manufacturers are now offering brushless DC motor fans with on board speed control and
these are being fitted to some boilers. The electricity saving potential is assumed to be in the region of
10 – 20%.

Programmers
Fig.3 shows the programmer typically uses 7% of the system electricity or 18kWh annually. Despite
having small instantaneous power consumption of around 2W, the running hours are continuous. All
controls manufacturers visited make both electromechanical and electronic programmers and the true
power consumption (in Watts) is similar for both for all brands. Although electronic programmers are
often cheaper to produce, the market for electromechanical devices is still significant, particularly for
users with special needs.
Few manufacturers indicated significant energy improvements were imminent or that they were
possible with existing designs. The call for permanently illuminated displays from some user groups
was seen to be an energy retrograde step by the manufacturers, although all agreed if it meant the
controls were better used there would be an overall energy saving because of fuel saved.

Boiler standby power

The boiler modeled in system 3a in Fig.3 has a simple bourdon gauge type thermostat and consumes
no power when heat is not required. Increasingly, many boilers use an electronic control card that
provides the thermostat function along with flame detection, ignition phasing and so forth. These
typically have power supply capability to drive dc fans and therefore have a fairly large power supply
even though the full capability is seldom used. GdF identified the typical standby consumption of such
cards using mains frequency transformers to be 8W.
New generation cards have switched mode power supplies and standby consumption is thought to be
in the region of 1W, in line with other standby power initiatives. It should be noted most of these cards
have capability for advanced thermal control such as weather compensation and separate set points
for heating and domestic hot water. However, manufacturers indicated these functions were rarely
enabled as installers were often unfamiliar with the benefits of such advanced controls. If they were, it
is thought the fuel savings would easily outweigh the electricity penalty.

1180
Motorised valves
There are two common types in service, 2 port and 3 port, usually based on spring return operation in
the UK. The zone thermostat (cylinder or space) energises the valve motor to open the valve that is
held open against spring force. A shaft mounted microswitch switches on the boiler and pump once
the valve is open and the spring shuts the valve once the supply to the motor is removed.
Consumption is a continuous 5 watts when the zone is calling for heat – there is no standby or holding
current when there is no call for heat.
The conventional 3 port valve is particularly wasteful and often consumes power in a holding state
when the heating system is off such as overnight. This energy burden can amount to around 18kWh
pa if a 3 port valve is used in a heating system similar to that of Fig.2. and manufacturers estimate
between 7 – 10 million such three port valves are in service in the UK. The reason for the holding
consumption occurring relates to the construction of the valve and the way it is integrated into the
controls system, detailed in the SAVELEC report.
One strategy for improving the energy efficiency of zone valves would be to avoid spring return and
use motor on, motor off therefore consuming power only during state change. One manufacturer
quotes a 30% electricity saving using this type.

Conclusions
The annual electricity consumption of non-electric heating systems varies considerably according to
component choice and system type and is a significant proportion of domestic electricity consumption.
Within the SAVELEC project the annual electricity consumption of typical UK central heating systems
was simulated and found to vary from 131 kWh/a for a semi-gravity system through to 415 kWh/a for
a combi boiler system, placing it in a similar context to some A rated cold appliances. With system 3a,
which uses 241 kWh/a, the pump represents 53% of the electricity consumption (124 kWh), the boiler
fan 32% (77kWh), programmer 7% (18kWh), motorised valves 4% (11kWh) and gas valve 4%
(11kWh).
Using recent technical developments, this consumption could be reduced to 139 kWh per annum, a
42% saving. The largest saving is attributable to replacing the pump with a PM type. However, there
is currently little available consumer information to enable them to assess how much their system
uses or to make an informed component choice.
One way forward would be to develop a voluntary agreement that manufacturers could participate in
and this has achieved success in standby power in other industries, notably television. Combined with
component, or preferably system, energy consumption measurement and labelling consumers could
then make informed choices. For the scheme to be successful, the industry partners would need to
establish a commitment to improve and set an achievable target.
The agreement would need to develop an agreed measurement procedure and method for converting
instantaneous power into annual consumption. Within SAVELEC, considerable differences occurred
in annual energy consumption between the partner countries due to different climate, heating running
hours and installer preference. One solution to these difficulties is to focus on the biggest consuming
components and treat them in isolation – for example set a maximum standby power for boilers and
programmers, or a specific energy consumption (e.g. watts per litre at fixed head) for pumps.

References:
[1] BOILER SAVELEC - Characterisation and reduction of electrical consumption of central heating
systems and components, EC Contract no. SAVE 4.0131/Z/02-021/2002. June 2005.
[2] Promotion of Energy Efficiency in Circulation Pumps, especially in Domestic Heating Systems
(Contract No 4.1031/-Z/99-256)

1181
1182
European Initiatives on Labelling of Central Heating Gas Boilers
Daniel Hec1, Jean Schweitzer2, Karsten V. Frederiksen2, Terry Williams3, Miguel
Manucas4
1
Marcogaz
2
Danish Gas Technology Centre
3
Advantica
4
Repsol

Abstract
Society is moving towards more information to the consumer. The labelling of white goods has been a
success and shown that the market can change toward the most efficient products.
However boiler are so called “installed products” witch performances are depending on the installation
conditions therefore the labelling of such appliances shall be carefully designed and information
systems shall accompany labels.
The present paper illustrates by examples in the UK, Spain and Denmark how such schemes have
been designed and implemented. The paper also concludes (in the light of the experience gathered)
on the way to design and implement a European scheme.

1. Introduction
CH boilers are a huge market in the EU (5 to 6 million appliances sold every year) and heating
represents more than 20 % of the whole energy consumption in the EU. As a result the impact of CH
boilers on the whole energy consumption and saving is very important for the individual consumer, but
for the whole society as well.
In most countries the final user shows a lack of interest in the choice of a new boiler and the decision
is left to the installer. As a result the boiler chosen is not always selected with due consideration to
energy savings. Despite initiatives from the Commission to regulate the market toward high efficiency
boilers (dir. 92/42) the system proposed (minimum efficiency + star rating = labelling) has not had the
impact expected.
The new ECO design directive provides the opportunity for new measures like labelling and
information systems on boilers. Such initiatives have already been launched in the UK and in
Denmark.
In the UK, the SEDBUK method consists of an Information System and a Database aimed at
informing the end user in a simple way through the Internet, giving consumers, installers and
designers an accurate tool to compare different boilers. At the same time, the application of a label on
the same basis makes it possible to obtain simplified information using the now well-accepted A to G
system of the labelling directive 92/75. This has encouraged all manufacturers to provide the highest
rated energy efficient boilers and the success of this initiative is such that the latest UK Building
Regulations generally only allow for A or B rated boiler types to be installed.
In Denmark, such a label was also introduced and similarly an information system is made available
on the Internet. The difference is that in Denmark the system is interactive and can take into account
the specificity of the installation.
These cases complement each other in the two systems and tools used. The label gives a first
indication to the end user (and a message that “boilers on the market are different!”). Next, the
information system may enable the end user to have a more informed discussion with installers and to
fine-tune the choice of the boiler. The label is clearly targeting the end user, where the information
system has a wider target group (end user + installer + energy adviser + architects etc.).
The impact of such systems has dramatically changed the market mechanism in the two countries,
and the end users become increasingly involved directly or indirectly in the choice of their heating
appliances. Labels have become strong marketing tools in both countries and as a result,
manufacturers have withdrawn some appliances, where the level of efficiency was not attractive
enough for the market.
At the same time, the DGTREN has become more and more interested in introducing such a system
at the European level. The industry has been reflecting on how this could be done and Marcogaz has

1183
issued a paper that gives some indications of what could be the main features of such a system. In
order to go ahead with the implementation of the measures described in the ECO design directive,
DGTREN has recently launched two tenders (on BED and ECO design directive) that should help
clarify the situation.
In the meantime, more countries tired of waiting for a European label have decided to launch a label
of their own, as in Denmark or the UK. At the present time, such a system is also being developed in
Spain. Thanks to the effort of Marcogaz and gas utilities, at present a lot of efforts are being put into
harmonising such individual initiatives.
This paper will give some details on the labelling and information systems that have been developed
and will discuss their impact on the market. The paper will further develop on the industry’s point of
view of how to design such systems at EU level in order to fulfil the requirements of the ECO design
directive and its implementation measures. Among other issues the paper will discuss the relevance
of an EU appliance database, with the design of a European label and information system.

2. Heating with CH boilers; short explanation of the technique

The heating system and its parts (source: www.boilerinfo.org [1])

Figure 1: The main parts of the heating installation

The main parts of a heating installation

1. Boiler/burner combination: the part producing the heat
2. Piping with pumps and valves: the part distributing the heat
3. Radiators, convectors,…: the part emitting the heat to the room
4. Control equipment as room thermostat, outside temperature control: the part controlling room and
water temperature

The central heating system provides heat to all rooms of the dwelling. In the EU this is mostly done
via water (or in a more limited number of cases via heated air). To produce the heat and deliver it
correctly to all rooms the heating system consists of 4 main parts.

Production system
The production system is the part of the installation that produces the heat. In the case of a central
heating system, this is the boiler/burner combination.

Distribution system
The distribution system transfers the heat to the radiators or convectors. It consists of piping with
pumps and valves.

Emission system
The emission of the heat in the room is achieved by radiators, convectors or floor heating.

Control system
The control system is the part of the installation that manages and controls the boiler. There are two
main ways to operate the control: either on indoor temperature or on outdoor temperature. It generally
consists of a thermostat, the radiator valves etc.

1184
The different parts of the heating system interact
Convector, radiator or floor heating requires different water temperatures for their optimal use.
Different water temperatures will be obtained by using different types of control systems and different
emission system sizing. These differences in water temperature will, of course, lead to differences in
boiler efficiency and in distribution losses.

The heating system interacts with the building

The main function of the heating system is to compensate for the heat losses of the building. When
the heat losses of the building decrease, the requirements for the heating system change at the same
time. This has practical consequences on the boiler efficiency as well.

Annual efficiency
Each boiler is characterised by a nominal efficiency, which is measured in the laboratory. As
explained above, once installed, the efficiency of the boiler could depend on many parameters, such
as:
• type of boiler
• design of the water circuit distribution and emitters
• heat demand
• etc.
So in real life the annual efficiency of the same boiler may differ considerably from one installation to
another.
This is the reason why the annual efficiency shall be preferred to the nominal efficiency when
choosing a boiler for a given installation.

BOILSIM
BOILSIM is a sophisticated boiler model that works with simple parameters measured by the
application of boiler standards. It was developed over several years and validated over more than 80
boilers. Today it is used for the basis of labelling (in Denmark) and information systems (Denmark,
Spain).

Figure 2: The principle of the BOILSIM method

3. Overview of the EU boiler market

The size of the EU market is probably about 6 million units a year. The existing boiler population is
some 60 to 80 million. Knowing that the average energy consumption is probably somewhere
between 15.000 and 25.000 kWh a year the heating with boiler represents a considerable amount of
the total energy consumption in the EU. Therefore, the impact of policies on the market could be quite
considerable.

1185
Most of the new boilers sold today are traditional, but there is a clear trend in favour of the condensing
technology. In countries like Denmark or the Netherlands condensing boilers are today the most sold
type of boiler (see below).
The efficiency of most existing boilers older than e.g. 15 years is probably 10 to 15 % lower than the
new condensing boilers available. Therefore, there is a considerable potential of energy saving.
More details on the market data will be available at the end of 2006 as the Commission has ordered a
study that should clarify a number of points, including the market.

4. EU regulations, boiler directive, EPB directive and standards. Impact on the

market
In order to generally improve energy efficiency in Europe, the Commission has recently issued
different instruments like the Directive 2002/91/EC on the Energy Performance of buildings or the
proposed Directive on Eco-design requirements for energy Using Products which will amend the
Directive 92/42/EEC on energy efficiency requirements for new hot water central heating gas boilers.
Furthermore a new Directive on Energy Services is also discussed.
Directive 92/42/EEC provides a star rating system intended to ascertain the energy performance of
boilers. This star rating system has proved not to deliver the expected results and Directive
92/42/EEC will be revised.
There are mainly 4 framework directives that are relevant to boilers:
1. Boiler Directive 92/42 (Council Directive 92/42/EEC of 21 May 1992) on efficiency requirements
for new hot-water boilers fired with liquid or gaseous fuels.
2. Labelling Directive 92/75 (Directive 92/75/EEC Council Directive 92/75/EEC of 22 September
1992) on the indication by labelling and standard product information of the consumption of
energy and other resources by household appliances.
3. EPB Directive 2002/91/EC of the European Parliament and Council, on the energy performance
of buildings.
4. Directive 2005/32/EC on the eco-design of Energy-using Products (EuP).
The boiler directive (1) effect was to remove from the market the boilers with very poor efficiency. The
labelling system (star rating) that was included has never been successful; one reason being that the
system was not adapted to the interlaboratory reproducibility of efficiency measurement. Since then,
there has been a stronger focus on measurement aspects, and laboratories grouped in a network,
LABNET, have been working for the improvement of the situation.
The labelling directive (2) does not apply directly to boilers, but as the eco-design directive (4) makes
the labelling system mandatory, the directive becomes highly relevant.
The EPB directive (3) implementation (in the national building codes) did in many countries result in
direct or indirect new minimum requirements on boiler efficiency. As a result, the condensing
technology appears to be the only choice possible in many countries.
According the present version of the Eco-design Directive (4), central heating boilers shall be subject
to labelling and the Commission has also confirmed that Directive 92/75/EEC on the indication by
labelling and standard product information of the consumption of energy and other resources by
household appliances will apply to central heating boilers. This directive will also apply to domestic
hot water production.

5. The labelling directive. Marcogaz point of view for its application to boilers
Background
In many countries there has been a long tradition of informing the final users and installers about the
real performance of central heating gas boilers and this was done in the past mainly by gas utilities.
For most gas consumers the image "gas" is given through the gas appliance installed. A “bad”
appliance will therefore lead to a poor image of the energy “gas". Therefore, it was and still is
important that the gas utilities are present in this discussion.
Boiler performance is not only dependent upon nominal efficiency but also on the conditions of
installation. Therefore in order to achieve the targeted savings a broad collaboration and action are
needed. Labelling, education and training, inspection, consultation and information should be
organised in collaboration with the Authorities, Energy utilities, manufacturers, designers and
installers.

1186
How to achieve energy savings for central heating gas boilers? Overall organisation
As already mentioned, central heating gas boiler efficiency depends both on the boiler itself and on
the system as well as the users habits. Therefore an action on the boiler is useful but will not
necessarily lead to a significant improvement without consideration being given to the whole of the
central heating installation. So the emphasis should be put on the right choice of the boiler at the
installation design phase.
It should be noted here that gas boilers are commonly sold by installers. The existence and quality of
the information given to the gas appliance installers is therefore a key issue.

Labelling
The labelling of boilers is a first action that aims at informing the public widely about the fact that the
boilers have different efficiencies and that different technologies exist. This label needs to be based
on simple assumptions and be valid for an average installation.
The parameters used for the calculation of the label should be the same: EU CE type testing. The
label shall include information on the boiler efficiency, heat output and electric consumption.
As the label is a simplified information, the choice of the efficiency to be used as basis for the label is
not a determining discussion for MARCOGAZ: it can be part load efficiency or annual efficiency, it
could also be the efficiency for a given load determined with models such as BOILSIM that has
proven to be more accurate than measurements at low load, it can also be the SEDBUK efficiency:
50% of full load + 50% of part load (which halves the uncertainty of the part load efficiency tests).
The label is aiming at fair consumer information and this aspect shall be reflected in the design of the
system (e.g. the same basis shall apply for fuel oil and gas). MARCOGAZ is flexible as to the exact
content of the label as long as there is an agreement on having an information system launched
together with the label. One role of this information system would be to correct from the inaccuracies
of the labelling system that would always be too simple to take into account the particularities of a
given installation.
Whilst the label is primarily for new boilers on the market, it should also be used in a simplified way for
existing boilers. Obsolete boilers can be rated and included on a database which will then allow a
comparison between the existing boiler and a new ‘A’ rated boiler showing the running costs savings
achievable. This has been implemented in the UK with SEDBUK (www.sedbuk.com) and has lead to
increase in the boiler exchange market (currently 800,000 plus per year).

Information system
Besides the labelling there is a strong need of an information campaign aimed at the installers and
possibly the interested end users. This system aims at bringing some additional information and will
take into account the influence of the design, installation, controls, commissioning and operation of
the central heating system. Information relative to energy consumption and performance,
corresponding to annual consumption and covered by standardised test method, should not be mixed
up with comfort criteria such as waiting time or temperature fluctuation at constant water flow. This
system should help the installers to find the best adapted boiler for a given installation. We believe
this system should be implemented on the Web for easy access.
The information system will deal with energy saving, but also with cost. It is important to give to the
user of the system, elements helping to calculate e.g. payback time of a new boiler.
Replacement of old boilers shall be done in consideration to energy and cost savings for the
consumers. The Directive 2002/91/EC related to Energy Performance of Buildings offers a platform
which can be used. The opportunity should be taken to choose an appropriate replacement boiler and
consider central heating system improvements to optimise potential energy savings benefits.

Connection between the different tools needed

In order not to create confusion or contradiction in the data given by a label or by any information
system, basically the same data and same models (if such are to be used – some validated models
exist e.g. for the annual efficiency) shall be used for the label and information system.
In practice, there is a need for a centralised boiler database that would list the boilers tested and data
needed for the labelling and information system. Such a databases already exists as
(www.boilerinfo.org, SEDBUK www.sedbuk.com) , and a similar system shall be further developed at
EU level under the control of Organisations such as the Notified Bodies in order to guarantee the
validity of the data used (see further in section 5).

1187
Need for harmonised testing and measurement (art 7.1 of dir 92/42)
The quality of the data to be used in the information system and especially the measurement
accuracy of laboratories and inter-laboratory reproducibility are key issues for the successful
application of the label and information system.
The various Round Robin Tests (inter-comparison tests) organised between EU accredited
laboratories have all demonstrated that the inter-laboratory reproducibility of test results for the
measurement of efficiency is giving problems. Especially part load (30%) efficiency is a problem and it
is necessary to consider this very seriously when designing both the labelling system and the
information system.
It would be the responsibility of an organisation such as BEDAC to set up the criteria for the
acceptance of the data to be used for the labelling and used in the data base. It is known that a
number of measurement problems can affect the result in such a way that the labelling or information
system cannot be used in practice.
Therefore, specific attention shall be given to this point and it is advised to involve LABNET (network
of laboratories) in the discussion about the criteria for test result acceptance.
The efforts to improve the measurement accuracy and consistency between laboratories should be
continued, and especially in the phase of enlargement of the EU. Already much progress has been
achieved in LABNET. It could be suggested that the laboratories which are providing test data for the
database should work together with the Notified Bodies for the definition of the measurement
requirements and perhaps with Round Robin Tests (inter-comparison) every two years or so.

6. National initiatives for labelling and information. Short description and

impact
The UK
First issued in 1999, the SEDBUK (Seasonal Efficiency of Domestic Boilers in the UK) method
estimates seasonal efficiency as a percentage figure. SEDBUK results are used to rank boilers and
give an accurate comparison of their energy efficiency. SEDBUK is a fuel efficiency value (maximum
100%, using gross calorific values), averaged over one year, for boilers installed in a typical UK
house. It has helped transform the UK market for domestic gas and oil-fired boilers in recent years.
The UK government introduced SEDBUK after a comprehensive 3-year test programme during which
laboratory tests were carried out on 20 different boiler models and these were combined with field
trials of boilers in some 100 premises.
The SEDBUK numbers are listed as percentages that indicate how much of the heat obtained from
burning fuel (gas or oil) is transferred into the heating and hot water system. The figures have been
calculated by combining a series of factors which affect boiler performance once it is installed. This
includes the boiler type, fuel type, the ignition arrangement, any internal water storage, patterns of
usage, and the control system, as well as the results from standard laboratory performance tests.
This laboratory performance data is obtained as part of the manufacturers’ compliance with the
European Boiler Efficiency Directive (BED), to which all boilers are tested at their approvals stage.
Full and part load (30 %) performance tests are carried out to meet the requirements of the directive
and these are fed into the SEDBUK calculations.
A joint industry programme, funded by the UK government and managed by BRE, provided
cooperation between appliance manufacturers, trade associations and the test houses, and this led to
the development of a set of ten equations. The type of boiler, control system, water storage etc.
dictates which equation is relevant to a boiler model. Seven equations are used for natural gas and
LPG boilers with the remaining three equations being used for oil-fired boilers.
Boiler manufacturers submit their efficiency test results, with independent certification, to the manager
of the UK’s Boiler Efficiency Database, who verifies the figures and calculates SEDBUK from them.
The results are entered in the database, where they are used to calculate and certify the energy
performance of dwellings and to show compliance with Building Regulations. They are also published
on the internet for all consumers to access and make decisions on their appropriate choice of
appliance (www.boilers.org.uk).
Boilers are graded into efficiency bands from A to G with A being 90 % or above and G being below
70 %. The database gives brief details of each boiler model with its A-G efficiency band and an
annual running cost indicator. Gas boilers with a SEDBUK efficiency of 90% or more (band A), and oil
boilers with 86% or more (bands A & B), qualify for endorsement and can display an “Energy Saving
Recommended” logo as part of an energy efficiency product promotion scheme. All these activities
are helping to increase public awareness of energy savings benefits.

1188
Further to this, there have been changes made to the UK Building Regulations (part L) in April 2005
such that installers must now only install high-efficiency condensing boilers with SEDBUK rating band
of A or B, unless there are exceptional circumstances that make this impractical or too costly. It is
thought that these exceptions will apply to some 5 % of new boiler installations.

Denmark
An energy labelling scheme for electrical appliances has been used for the consumer imformation for
more than 10 years. The Danish gas companies and Danish Gas Technology Centre decided to use
this well-know design as the basis for development and implementation of a voluntary labelling
scheme for small domestic gas boilers.
The aim of this initiative was to give the user an easy-to-use and fair tool for choosing a new domestic
gas boiler and thus to promote the use of high-efficient boilers.
The labelling scheme was established with the support from the entire gas industry in Denmark; this
was essential since the extended use of the label to a very large extent depends on the broad support
of all parties concerned.
The annual efficiency method and the calculation program BOILSIM have formed the measurement
and calculation basis for the boiler labelling scheme. The advantage of choosing this basis is that
these tools were developed under the collaboration of a wide range of European partners including
manufacturers. Therefore, the method is considered as the standard for determination of boiler annual
efficiency by a large number of EU test laboratories.
The “arrow label” for electrical appliances is already well established among consumers as a tool for
choosing the most energy efficient electrical appliance. This is the reason for choosing the same
design as basis for the design of the layout of the boiler label.
The support of the Danish gas industry has been ensured by involving all parties in the development
and implementation work of the boiler labelling scheme.

Labelling of gas boilers

The arrow label graduated from A-G and with key figures with comparable information on energy
utilization and environmental load of the individual boilers is meant as a decision tool for the
consumer. To further assist the consumer in achieving a high sanitary hot water comfort and energy
optimized operation of the heating system guidelines for choosing the best boiler/hot water tank
combination were developed. Storage technology is the main market for sanitary hot water
production. Figure 3 shows the template of the design of the label.
A detailed description of the calculation method for the energy label is found in the document
“Description of the calculation method for the Danish labelling of gas fired boilers” that can be
downloaded from www.dgc.dk.
Apart from the label, other tools were developed that explain the contents of the label and facilitate
comparison of alternative boilers. Different tools were developed for the consumer and the installer,
respectively. Boiler manufacturers are yet another target group. Apart from the advising tools the
manufacturers need to know what they are required to do in order to have their new boilers included
in the labelling system.

1189
 Energy
Label Logo
Model

Low consumption

A A
B
C
D
E
F
G
Large consumption
Annual electricity consumption XYZ kWh
Annual efficiency XYZ %
Environmental load (NOx) X (Y,Z kg/år)

Choice of hot water tank

Alternatives, hot water tanks Hot water need
Hot water tank 1 (xx litre) Very large (15-18 litre per min.)
Hot water tank 2 (xx litre) Large (12-15 litre per min.)
Hot water tank 3 * (xx litre) Normal (9-12 litre per min.)
Hot water tank 4 (xx litre) Small (6-9 litre per min.)
*The boiler has been tested with a tank of xx litre. Choose a hot water tank to meet your need following
the advice of your gas installer. Please note: If you choose a larger tank than the one tested, your
annual energy consumption may turn out larger, and a smaller tank may lead to a lower consumption

Prepared by DGC in collaboration with the natural gas companies and the Danish Energy Authority.
For further information please contact your local gas company.

Figure 3: The template of the design of the label

Report on the situation after a one-year pilot period

The implementation of the labelling scheme in the market was based on a close cooperation between
DGC and the gas companies, boiler industry, Danish Electricity Saving Trust, National Consumer
Agency and Danish Energy Authority.
After the pilot period it appears that the energy labelling scheme has indeed influenced the boiler
market:
• During this one-year period, the boiler manufacturers have adapted their boilers regarding the
electrical components. Boilers that originally were sold with three-stage pumps have been modified
to be sold with modulating pumps.
• The supply of the best A labelled boilers is increasing at the expense of B-G labelled boilers.
• The gas company show rooms only show A labelled boilers today.
• As the labelling system is a voluntary system, some boilers are still not labelled.
The Danish gas industry has fully supported the labelling system. The boiler manufacturers’
commitment is mainly due to the fact that the labelling system has been prepared for EU
standardisation. Such standardisation may in the long run result in a reduction of the number of boiler
tests in the individual countries.

Spain
The aim of the Spanish administration is to develop a boiler database in order to promote the use of
the most efficient boilers in new installations, normally combi-boilers without storage system or, as
much, a micro-storage (4 litres) device. It is noticeable that, in the case of Spain, the installation of
solar panels for producing hot sanitary water (60 % of annual coverage) will be compulsory from
2006.

1190
The heating requirements in Spain (and other countries such as Portugal) are a bit different compared
with other countries in Europe.
In the first place, the heating requirements are much lower because of the climate, this means that in
most parts of the countries the boilers are working only 4-5 hours/days as an average, but very
dependant of the climate and the type of installation. Therefore, the use of simple methods for
estimating annual efficiency is difficult, because boilers are working most of the time at a high load but
for short periods. This is the reason why condensing boilers, which are common in other countries
nowadays, are not so efficient (they are used in non-condensing way) when used in Spain.
These difficulties have given a strong impulse to the use of more sophisticated tools, such as Boilsim.
Currently there is a common project between the Spanish Administration, gas companies and the
Spanish Boiler Manufacturer Association for developing a new tool that will be partially based on
Boilsim and will allow installers or customers to choose the best boilers. The tool also will take into
account climate and building, not only nominal efficiencies. The new tool is planned to be available on
a public web, and maintained by the Spanish Administration. Because the system is Boilsim-based,
the database will admit data obtained by a certified laboratory of Europe, and for this reason technical
aspects such as repeatability, reproducibility, and intercomparisons between labs are very important.

7. Conclusion
Society is moving towards more information to the consumer and implementation of a single labelling
system is a positive initiative. Information sent out and labels used should be simple, easy to
understand, and should permit the comparison between different energies only if information are
established on the same basis.
Such a system, however, can only be considered together with an information system able to
complete the general information given by discriminative and efficient label. Boiler performances as
other “installed products” depend greatly on the system (installation) and the users’ habits, therefore,
such an information system is needed. The Directive on Energy Performance of Buildings is also
bringing up the discussion of the evaluation of existing installations. It is important to ensure that a
consistent approach can be used which includes new installation of full central heating systems,
existing system improvements, and boiler replacement (including appropriate system improvements)
in order to realise energy saving potential.
The huge boiler population and market leads to the conclusion that only few % change on boiler
efficiency will result in considerable amounts of energy. Therefore, there is a strong focus on boilers in
the EU, but as things are moving slowly, national initiatives have been undertaken as there is a real
need for labelling and information systems.
The national systems implemented show that the labelling systems are moving the markets. In
Denmark the market has moved almost exclusively to “A” boilers.
Experience with the UK labelling system SEDBUK since 2000 has shown that the greatest success
has been the comparison between existing boilers and new ‘A’ rated boilers. By showing the rating of
the old boiler (usually F or G rated) and the running savings possible, customers are exchanging their
old broken down boilers rather than repairing.
It would certainly be useful to extend those initiatives to all countries, but it is urgent to work on the
future EU label as the multiplication of national labels would create an unacceptable difficulty for the
open market of boilers.
Beside the label it would also be useful to have a European information system able to take care of
the individual situation and bring a more accurate figure of energy consumption in the choice phase of
a new boiler.
Finally, the efforts done for the improvement of the interlaboratory reproducibility should be continued
in order not to create doubts on the validity of the future labels and information systems.

References
[1] www.boilerinfo.org - an informative website on Domestic Central Heating Boilers
[2] SEDBUK www.sedbuk.com
[3] LABNET http://labnet.dgc.dk/

1191
Glossary and definitions
BEDAC
Boiler Efficiency Directive Advisory Committee

CEN
The European standardisation organisation

CH
Central heating. A central heating installation consists in a heat generator and a heat distribution
system. It is typically a boiler connected to radiators.

FL
Full load. (see also PL = part load)

GLP
Good Laboratory Practice. The document developed by LABNET is describing in detail the
procedures for testing and measuring efficiency. It is a complementary document to the procedures
described in CEN standards

LABNET
Network of laboratories testing boilers - has about 25 members (http://labnet.dgc.dk/default.htm)

PL
Part load (see also FL = full load).

RRT
Round Robin Test: This is also called Inter-comparison test and it consists in sending the same
reference material (boiler) to different laboratories.

SEDBUK
Seasonal Efficiency of Domestic Boilers in the UK (http://www.sedbuk.com/)

1192
The Emerging European Water Heating Labelling Directives
Bruce Young

UK Market Transformation Programme

Abstract
The Energy Labelling Committee of the European Commission has turned its attention to domestic
water heaters, and development of three directives to deal with them is now in progress. The
directives define 19 product categories, embracing a wide variety of appliances, systems, and fuels.
But, unlike products covered by earlier labelling Directives, the categories behave in very dissimilar
ways, they are manufactured and installed by several separate industries in competition with one
another, and there are competing fuel interests too.
The aim of the new Directives is to supply consumers with information across the whole range of
choices, helping them to understand the different product and system types, as well as differentiating
products of the same type. Comparable measurement of the annual average energy performance of
so many categories presents a formidable technical challenge, and the mandate to create a set of
harmonised standards recognised that by emphasising “... the need to avoid distortion or bias which
would unfairly favour one type of product or industry relative to another”. As an important first step,
the mandate has provided some firm foundations by defining three principal patterns of daily hot water
demand to be used as common loading conditions.
Water heaters also use a variety of fuels, including electricity, gas, oil, and renewable energy sources.
That forces consideration, for the first time, of how fuels should be treated under a single labelling
scheme to indicate not just cost but wider environmental impact.
The paper summarises developments so far, explores some of the complex issues raised, and
discusses options for the way ahead.

Introduction
Hot water service is either the largest or second largest requirement for energy in housing in most
countries of Europe. In the UK it accounts for about 23% of the energy consumed by the housing
stock, and in newly built houses (which have a far lower demand for space heating) it is likely to be
nearly 40%. In both cases it exceeds, by a substantial margin, the energy demand for lighting and
appliances. It might therefore be expected that hot water in dwellings is a prime target for attention in
the energy efficiency policy arena.
Yet there has been little policy intervention at European level so far, and it has been left to national
building codes to impose minimum performance standards. Slow progress may reflect the complexity
of the subject, which arises from:
- the many types of product and system on offer
- the variety of configurations
- integration with space heating systems
- the interaction of separately purchased components
- the selection of fuels, including renewable energy sources
- the different (and competing) industry interests.

Early activities of the European Energy Labelling Regulatory Committee concerning water heaters go
back to 1994, though after September 2000 were concentrated on electric storage water heating only.
A comprehensive history is set out in [11]. More recently the Committee decided, in February 2002,
that energy labels should be introduced for nearly all types of water heaters installed in single family
homes. Later in 2002 the Committee issued a mandate [12] to CEN and CENELEC to develop test
standards, with the provision that they should produce comparable results, should be fair, and should
all use the same daily load patterns. In October 2004 the Committee then issued for discussion the
first drafts of the new Water Heater Energy Labelling Directives.

1193
Scope of the Directives
At the time of writing (February 2006) there are three draft directives for water heaters under
development by the Energy Labelling Regulatory Committee. There is one for electric water heaters
[5], one for gas and oil water heaters and water storage devices [6], and one for solar water heaters
and water storage devices [7]. They cover 4, 9, and 6 product categories respectively, as shown in
Table 1.

Table 1 : Categories of water heater in the Directives

Partial diagram
Directive &
Category description (see Appendix
category
B)
Electric – 1 Electric Storage water Heater A
Electric – 2 Electric Night Storage water heater A
Electric – 3 Electric instantaneous water heater -
Electric – 6 Electric Storage water Heater (with heat pump) i) air, ii) ground C
Gas/oil – 1 Gas Storage water Heater B
Gas/oil – 2 Gas instantaneous water heater E
Gas/oil – 5 Gas fired heat pump water Heater D
Gas/oil – 6 Gas combi boiler E,F,G
Gas/oil – 7 Oil Storage water Heater B
Gas/oil – 8 Oil instantaneous water heater E
Gas/oil – 11 Oil fired heat pump water Heater D
Gas/oil – 12 Oil combi boiler F,G
Gas/oil – 13 Hot water storage tanks (with heat exchanger) B,C,D
Solar – A Solar water heaters with no auxiliary heat source B
Solar water heaters with auxiliary heating by gas or oil (including
Solar – B B with E
combi appliances also designed to provide heating)
Solar – C Solar water heaters with auxiliary heating by electricity C
Solar water heaters with auxiliary heating by means of a heat
exchanger driven by a gas or oil central heating boiler, even if
Solar – D D
they can also be heated by an electrical resistance (immersion
heater)
Solar – E Solar hot water tanks D
Solar Collectors (suitable for use in water heating systems), sold
Solar – F -
separately

The intention of the Directives is to measure the energy performance of nearly all water heating
products and systems designed for a single family home, and label them on a comparable basis.
There are exceptions for systems that are very small, very large, very powerful, or that provide for a
single outlet only. A few uncommon product categories and combinations have been omitted (eg,
water source heat pumps, thermal stores, micro-cogen), but the range is otherwise comprehensive
and a laudable attempt to embrace the very wide selection available to purchasers.

Requirement for comparable performance measures

It is the requirement for comparability across this wide range of products and systems that presents
the greatest challenge. Performance is to be measured by harmonised test standards, where
possible. Mindful of the different fuel and industry interests, the mandate [12] declared in resounding
words that harmonised standards “... need to avoid distortion or bias which would unfairly favour one
type of product or industry relative to another”.
But this important aim will not be easy to achieve. Performance evaluation methods are required that,
in the context of energy labelling, are straightforward to apply, without undue expense, across the
whole range of products and systems on sale. Their purpose is to produce data for an energy label to
inform the purchasing decisions of consumers. If the methods are too difficult they will take a long
time to develop and agree; if too simple, they will be unable to discriminate fairly between similar
products. Performance of domestic water heating systems generally was discussed at EEDAL 2003

1194
[13], and since then the draft water heater directives have been issued with energy classification
scales. The relevant factors affecting the development process are discussed below.

Common load patterns

A restricted working group was convened in 2001 to determine a small number of standard daily hot
water demand patterns (“tapping patterns”) against which all products and systems could be tested.
Tapping patterns strongly influence test results, so there is good reason to impose uniformity rather
than permit different patterns for different products. For example, it might be alleged that certain
patterns favour, say, electric storage water heaters (good for rapid delivery; slow to recover) over
instantaneous water heating by combi boilers (significant start and stop losses; no recovery time).
The restricted working group was informed by recent SAVE projects ([8], [9]), which indicated average
hot water consumption of 36 litres per person per day and 100 litres per household per day in the 142
million homes of the EU. These figures became the basis for two of the tapping patterns defined by
the group, and were later copied into the mandate and draft Directives. It is important to remain aware
that, while reasonably consistent across the different countries, the consumption figures are averages
of very wide individual variations. Also, the energy demand for any given volume of hot water varies
significantly with climate.

Functional and operational characteristics

The product categories covered by the Directives (see Table 1) are not functionally equivalent; for
example, some are relatively simple components (eg, Gas/oil – 13) while others are complex systems
(eg, Solar – D). Operationally they differ too, as they range from “instantaneous” (meaning hot water
production on demand) to storage systems in which heat generation is decoupled from draw-off, with
hybrid options sharing some characteristics of both.
Systems may be constructed in a variety of configurations, and it is impractical to anticipate and
analyse them all for the purpose of energy labelling. Probably the most important distinction is
between those that contain a heat generator and those that don’t.

Range of measured performance

The many product categories and disparate behaviour will lead to a wide range of assessed energy
performance. If the results for all product categories are placed on a single energy classification scale,
comparison between types will be clear, but steps on the scale (bands) may prove to be too coarse to
discriminate between products of the same type. One way to overcome that would be sub-division of
the bands; eg, A1, A2, A3 to discriminate between products of the same type with similar
performance.

Fuel variety
The variety of fuels, including renewable energy sources, introduces a difficulty that did not arise in
earlier labelling directives, as the products were all electric. For water heating, it is very likely that
combinations of two or three fuels will be used in the same system (eg; solar, gas, electricity), and
reliable methods are needed to assess their quantities and relative contribution to an overall energy
rating.
The different fuels have greatly different environmental impact, and the Energy Labelling Regulatory
Committee has not announced its policy on how fuel types affect the measures of performance
required for energy labelling, where the types must be combined and placed on a single energy
classification scale. The treatment of different fuels in the draft Directives has been examined from the
perspective of delivered energy, primary energy, CO2 emissions, and annual fuel costs (using UK
figures), and results are shown in Appendix A. It can be seen that the classification scale chosen for
the Directives is closest to primary energy.

Test standards
The plethora of standards for different hot water products was one of the reasons that impelled the
Regulatory Committee to issue a mandate calling for testing under common conditions. To describe
hot water products, the mandate refers to 14 existing standards. Not all of them include energy
performance measurement requirements, and, where they do, that may not have been their primary
purpose and they may not have set minimum levels. Historically, the standards have been seen as
independent, with no requirement for co-ordination during development.

1195
The mandate now requires CEN and CENELEC to develop test measurement standards for water
heaters with a common loading pattern. This calls for adaptation, or re-writing, of existing standards,
and possibly new ones. They will still be developed by different committees, and effective co-
ordination is necessary if comparable results are to be assured.
Furthermore, the scope of the relevant standards for water heaters has previously been limited to
individual products rather than complete systems in which such products are installed. For an
assessment of the energy performance of complete hot water systems, a rather different approach is
required, more like that adopted for the standards being written for the Energy Performance of
Buildings Directive ([14], [15], [16], [17]).

Industry structure
Water heaters comprise diverse technologies and fuels, and separate industries are responsible for
their design and manufacture – often as part of a range of other products of similar type and
construction unrelated to water heaters. As noted already, standards development is the responsibility
of several unco-ordinated CEN and CENELEC committees, and such committees are dominated by
experts from the particular industries involved. Against this background, the requirement to deliver
comparable performance measurement standards for the full range of products and systems, on
parallel development tracks, will need careful independent monitoring.

Methods of performance evaluation

Although harmonised performance test standards are seen as the preferred option, it is recognised
that they will not always be feasible or even practical. It is relatively easy to design a performance test
standard for a self-contained product such as a hot water tank with an electric heater element, or an
instantaneous gas water heater. The results from such tests can be used to predict annual energy
demand with reasonable confidence. But testing of products that do not include the heat generator,
and are merely component parts of a larger system, is another matter.
Three approaches to performance evaluation are possible, as discussed below.

Laboratory tests
Laboratory tests can be carried out under carefully controlled conditions, to defined levels of accuracy
specified in EN standards, and, if properly designed, should at least provide a fair comparison
between products of the same type. The scope and complexity of tests may vary widely, and it is
helpful to distinguish them on that basis.

(i) Testing of a self-contained product, operating as a water heater

Provided the product under test includes the heat generator and represents behaviour of the whole
water heating system, the performance results can be combined into an average daily or annual
energy figure to be shown on the energy label. If the product does not include the heat generator, of if
there is a second heat generator (eg; a solar collector), then test results of the product alone are
inadequate as a guide to overall system performance.

(ii) Testing of products that are components of larger systems

If the aim is to measure a product intended to be installed as part of a larger system and give it an
energy label of its own, it must be installed in a reference system where the qualities of the other
components have been rigorously defined. There might be a number of possible system
configurations in practice, but only one reference system configuration can be allowed if results are to
be comparable.
For example, a product in category “Gas/oil – 13” (a hot water tank) could be installed and tested in a
system with a boiler of defined efficiency operating under a defined control regime. The weakness of
this approach is that boiler efficiency and behaviour may well be the dominant characteristic from the
point of view of overall energy performance, and the tank would affect the outcome only to a modest
extent. If that were so, the accuracy and repeatability of tests could not be expected to be very high,
and the results may not discriminate between tanks under test with sufficient resolution. A more
serious objection is that the results are irrelevant when the tank is to be installed in a system with a
boiler of entirely different characteristics, and potential customers would then be gravely misled by the
energy label.

1196
(iii) Testing of complete water heating systems
Testing becomes far more complex if it extends to a complete water heating system that is an
assemblage of different components, separately specified and purchased. For example, in product
category “Solar – D” (a water heating system containing a solar collector, boiler, and tank with two
heat exchangers), the interactions between components affect overall performance. Relative sizes
and other characteristics influence the annual energy consumption, and the outcome of the test would
depend on all of:
- size of solar collector
- efficiency of boiler
- volume of solar zone in tank
- total volume of tank
- insulation of tank
- performance of solar heat exchanger within tank
- performance of boiler heat exchanger within tank.
It is difficult to see how testing of a complete system such as “Solar – D” can be specified accurately
and carried out economically for the purpose of energy labelling, and it seems likely that alternative
methods will be adopted.

(iv) Testing of secondary characteristics

Secondary characteristics are properties of the product that affect its energy performance but do not
give enough information, on their own, to say what the energy consumption will be when the product
is installed and used. In the case of hot water tanks, the relevant characteristics would be insulation
and heat exchanger performance, both quoted in their normal units of measurement (not energy). For
products in category “Solar – F” (a solar collector), it would be temperature rise under specified
conditions of solar radiance, inlet water temperature, and flow rate. It is not possible to equate the
measured values to energy consumption directly, but that can be done indirectly by using them in
calculations based on an analytical model.
Testing for secondary characteristics is far cheaper and simpler than testing of complete systems,
and harmonised standards are relatively easy to produce.

System models
An alternative to laboratory tests of whole systems is to develop an analytical system model, in which
system and boiler behaviour is simulated mathematically and measurable characteristics of the
product to be labelled are entered as parameters to the calculation. Such models need not be very
complicated, but even if they are the final application of them may be reduced to a straightforward
procedure or set of equations. The paradigm is SEDBUK (“Seasonal Efficiency of Domestic Boilers in
the UK”), in which EN standard test results for boiler efficiency are converted to an annual average
installed efficiency, making reasonable assumptions about controls, occupancy patterns, climate, etc.
If analytical modelling is adopted as a solution to labelling some water heater product categories, it is
essential to ensure that underlying assumptions are realistic and representative of the typical systems
in which they are likely to be fitted. Identification of typical central heating systems, some integrated
with water heating, has already been carried out for the SAVELEC project [10].
Modelling assumptions must always be clearly stated, and may be subject to challenge and dispute.
For those reasons, development of acceptable models for the more complicated water heater product
categories will probably be a slow process.

Secondary characteristics
Measurement of secondary characteristics, as described under Laboratory tests (iv) above, is a less
ambitious alternative to a full analysis of systems, and can be achieved in a shorter timescale. An
energy label can be based on secondary characteristics alone, though if it is also to show average
energy consumption to assist consumers (rather than designers or installers) further calculations and
a system model are required.

Other relevant Directives

The Boiler Efficiency Directive
The Boiler Efficiency Directive [1] is severely out of date and sets a very low minimum standard that
does not take account of major improvements in boiler design over the last 15 years. Its relevance to

1197
water heating is marginal, as it does not extend to hot water tanks connected to a boiler, nor does it
require measurement of the performance of combination boilers while they are producing
instantaneous domestic hot water. Nevertheless the efficiency test standards used to demonstrate
compliance with the Directive remain relevant for water heating systems with certain boiler types
(regular boilers, but not combination boilers).

The Energy Performance of Buildings Directive (EPBD)

The Directive [3] demands a methodological approach to the design and analysis of new and existing
buildings, with a certification scheme to declare their energy performance. Installed plant, such as
water heaters, is an important contributor to overall building performance and is included in the
methods of analysis now being developed. There are four standards for hot water systems in
buildings – see [14], [15], [16], and [17]. However, their scope is not the same as that for the labelling
directives, and the initial drafts could not be adapted for labelling purposes without substantial further
work.
For the purpose of analysing and certifying buildings, clear identification of the water heating system
and its energy performance would be a great benefit to building assessors. An energy label, including
relevant technical data, would be able to provide that. However, the label would have to be permanent
and visible, which is not among the provisions of the current energy labelling framework Directive [2].

The Energy Using Products Directive (EUP)

The Directive [4] proposes whole life cycle costing as the principal method of analysis. Unlike other
energy using products, it is probable that energy consumption over product lifetime will be the
dominant factor for boilers and water heaters, with embodied energy and disposal costs exerting only
a minor influence on the results.
A study of 14 products to be covered by the Directive has commenced, including boilers and water
heaters, and will take 18 months to complete. It is too early to say how the performance measures for
water heaters under the EUP will correspond with those for the water heater energy labelling
directives.

Conclusions
A large number of product categories have been defined in the draft Water Heater Directives, and
differences between them are so great that separate treatment is needed in almost every case.
Measurement standards for some categories are well advanced and can be introduced in the near
future; others are less advanced but straightforward to complete, while a third group presents
considerable difficulties that have not yet been pursued in sufficient detail.
If the Directives are to be finalised within the near future (a period of two years was forecast by the
Regulatory Committee at its meeting in November 2004), then some means of expediting progress
must be found. Some expedient options worthy of consideration are:
- continue with harmonised performance measurement test standards for the product categories that
are straightforward to measure and label as self-contained water heaters. These include categories
Electric 1 to 3 and Gas/oil 1, 2 and 6, for which the test standards are nearly ready.
- simplify the requirements for product categories that are components of larger systems by adopting
an interim labelling scheme based on secondary characteristics, instead of overall system
performance. These include Gas/oil 13 and Solar E and F. Modelling of overall performance of typical
systems in which they are installed should be pursued, but is unlikely to be completed in a short
period.
- abandon or postpone whole system labelling for product categories such as Electric 6 and Solar B,
D, on the grounds that test methods or system models are extremely complex and will take a long
time to develop and agree.
Other points worthy of further investigation include:
- comparability of performance test standards on separate development paths will be difficult to
assure unless positive action is taken to co-ordinate the work of different standards committees and
critically examine their output.
- placing all product categories on a single energy classification scale makes clear the comparison
between types, but may prove to be too coarse to discriminate between products of the same type.
One way to overcome that would be sub-division of the bands; eg, A1, A2, A3 to discriminate between
products of the same type with similar performance.

1198
- standards already under development for the EPBD contain analytical methods that could be
adapted and used for system models for labelling, if that is to be the chosen method of evaluation of
some product categories.
- energy labelling of water heaters would assist building assessors responsible for certifying buildings
under the EPBD, provided the labelling scheme can be extended to require labels to be permanent
and visible.

References
[1]European Council: Council Directive 92/42/EEC of 21 May 1992 on efficiency requirements for new
hot-water boilers fired with liquid or gaseous fuels, Official Journal of the European
Communities, L167, Volume 35, 22 June 1992
[2]European Council: Council Directive 92/75/EEC of 22 September 1992 on the indication by
labelling and standard product information of the consumption of energy and other resources
by household appliances
[3]European Council: Council Directive 2002/91/EC of 16 December 2002 on the energy performance
of buildings
[4]European Council: Council Directive 2005/32/EC of 6 July 2005 establishing a framework for the
setting of ecodesign requirements for Energy-Using Products and amending Council Directive
92/42/EEC and Directives 96/57/EC and 2000/55/EC of the European Parliament and of the
Council
[5]DRAFT Commission Directive 200X/YY/EC of XXXXXX implementing Council Directive 92/75/EEC
with regard to energy labelling of Electric Water Heater, May 2005
[6]DRAFT Commission Directive 200X/YY/EC of XXXXXX implementing Council Directive 92/75/EEC
with regard to energy labelling of Gas and oil Water Heaters and water storage devices, May
2005
[7]DRAFT Commission Directive 200X/YY/EC of XXXXXX implementing Council Directive 92/75/EEC
with regard to energy labelling of Solar Water Heaters and water storage devices, June 2005
[8]SAVE study, final report: Analysis of energy efficiency of domestic electric storage water heaters,
SAVE-4.1031/E/95-013, March 1998
[9]SAVE study: Framework for a European method to determine the energy consumption for domestic
hot water, SAVE XVII/4.1031/Z/95-052
[10]SAVE study: SAVELEC Characterisation and reduction of electrical consumption of central
heating systems and components, SAVE 4.0131/Z/02-021/2002, June 2005
[11]AFECI: Historical review ELRC views on water heaters: Water heaters as discussed within the
ELRC, Doc 975.01 / Econ 512.01, February 2002
[12]European Commission: Mandate to CEN and CENELEC for the elaboration and adoption of
measurement standards for household appliances: Water-heaters, hot water storage
appliances, and water heating systems: M/324, TREN D1 D(2002), 27 September 2002
[13]Bruce Young: Performance of domestic water heating systems and appliances, Proceedings of
the 3rd International Conference on Energy Efficiency in Domestic Appliances and Lighting,
Turin, October 2003
[14]prEN 15316-3-1 Heating systems in buildings - Method for calculation of system energy
requirements and system efficiencies - Part 3.1 Domestic hot water systems, characterisation
of needs (tapping requirements)
[15]prEN 15316-3-2 Heating systems in buildings - Method for calculation of system energy
requirements and system efficiencies - Part 3-2: Domestic hot water systems, distribution
[16]prEN 15316-3-3 Heating systems in buildings - Method for calculation of system energy
requirements and system efficiencies - Part 3-3: Domestic hot water systems, generation
[17]prEN 15316-4-3 Heating systems in buildings - Method for calculation of system energy
requirements and system efficiencies - Part 4-3: Space heating generation systems, thermal
solar systems

1199
 CO2 emissions Delivered energy

8,000
2,000
G 7,000
Heater Directive

G
1,500 F 6,000 F
E 5,000 E
1,000 D 4,000 D

kWh/yr
3,000 C

kgCO 2/yr
C
500 B 2,000 B

A 1,000 A
0 0
Electricity Oil Gas Electricity Oil Gas

1200
Appendix A : Energy Classification Scales

Annual fuel costs

Primary energy

300
10,000
G
250 G
F 8,000 F
200 E
6,000 E
150 D

€/yr
D
kWh/yr

100 C 4,000 C
B B
50 2,000
A A
0 0
derived from Annex 2 of the May 2005 draft Electric Water Heater Directive and Gas and Oil Water

Electricity Oil Gas Electricity Oil Gas

Appendix B : Types of Water Heater
(A) Electric storage water heater

A
Standing loss label will give good indicator of relative
appliance efficiency. It will not indicate performance
relative to other types.

A label based on standing losses should not claim to

represent “energy efficiency”.

(B) Indirect (C) “Mixed” (D) Twin coil (for solar, etc)

B C D
Coupled to separate
heat generator(s) (eg,
boiler, solar collector).
Both standing loss and
heat exchanger
information are needed.
Overall efficiency
depends largely on the
heat generator(s) and
controls.

(E), (F), (G) Combustion appliances, with or without internal store

E F G
The principal energy
performance
indicator is the fuel
efficiency.

A standing loss label

is less relevant,
though increases in
importance with
larger stores.

No store With store With store

1201
1202
Electricity and Natural Gas Efficiency Improvements for Residential
Gas Furnaces in the U.S.
Alex Lekov, Victor Franco, Steve Meyers, James E. McMahon, Michael McNeil,
Jim Lutz

Lawrence Berkeley National Laboratory

Abstract
This paper presents analysis of the life-cycle costs for individual households and the aggregate
energy and economic impacts from potential energy efficiency improvements in U.S. residential
furnaces. Most homes in the US are heated by a central furnace attached to ducts for distributing
heated air and fueled by natural gas. Electricity consumption by a furnace blower is significant,
comparable to the annual electricity consumption of a major appliance. Since the same blower unit is
also used during the summer to circulate cooled air in centrally air conditioned homes, electricity
savings occur year round. Estimates are provided of the potential electricity savings from more
efficient fans and motors.
Current regulations require new residential gas-fired furnaces (not including mobile home furnaces) to
meet or exceed 78% annual fuel utilization efficiency (AFUE), but in fact nearly all furnaces sold are at
80% AFUE or higher. The possibilities for higher fuel efficiency fall into two groups: more efficient non-
condensing furnaces (81% AFUE) and condensing furnaces (90-96% AFUE). There are also options
to increase the efficiency of the furnace blower. This paper reports the projected national energy and
economic impacts of requiring higher efficiency furnaces in the future. Energy savings vary with
climate, with the result that condensing furnaces offer larger energy savings in colder climates. The
range of impacts for a statistical sample of households and the percent of households with net
savings in life cycle cost are shown.

Background
The residential furnace is an appliance that provides heated air through ductwork to the space being
heated. It is equipped with a blower to circulate air through the duct distribution system. In North
America, most houses are heated by forced air systems. Residential furnaces, for statutory purposes,
are defined as furnaces having a heat input rate of less than 225,000 Btus per hour (66,000 watts). In
the United States, 70% of households have furnaces of this type.
The National Appliance Energy Conservation Act (NAECA) legislation of 1987 established the initial
minimum standards for furnaces and boilers, effective in 1992. Current regulations require new
residential gas-fired furnaces (not including mobile home furnaces) to meet or exceed 78% annual
fuel utilization efficiency (AFUE), and in fact nearly all furnaces sold are at 80% AFUE or higher. In
2000, the U.S. Department of Energy (DOE) identified residential furnaces and boilers as priority
products for an updated standards rulemaking. The analytical approach and results reported here are
part of DOE’s rulemaking process for the Advance Notice of Proposed Rulemaking (ANOPR), which
was issued on July 29, 2004. [1]
Furnaces use electricity in addition to fossil fuel energy for combustion. Most of the electricity is used
by the circulating air blower. The furnace uses electricity for other purposes as well. The combustion
air is pulled through the furnace by a draft inducer fan, and a furnace also has various controls and an
electronic ignitor to assure proper ignition of the fuel.
The circulating air blower of the furnace pushes cool air returning from the inhabited space of the
house past the outside of the heat exchangers and supplies heated air to the house through a system
of ducts. Heat is provided by burning gas and moving combustion products through the inside of the
heat exchangers. The products of combustion are exhausted to the atmosphere through a flue
passage connected to the heat exchangers. The amount of air the blower can force through the
house’s ducts depends on the pressure and flow relations of both the house and the furnace. If the
house has an air-conditioner, as over three-quarters of houses with furnaces in the U.S. do, the
furnace blower and the same ducts will be used to circulate cooled air. To operate properly, air-
conditioners need more airflow than furnaces, so the blower motor is run at a higher speed during air-
conditioning operation.

1203
Residential furnaces are rated with annual fuel utilization efficiency (AFUE) as an efficiency
descriptor. AFUE represents the equipment's performance over an entire year’s heating season. It is
intended to represent the effective annual operating efficiency of a furnace under dynamic conditions.
It includes performance during start-up, steady-state, and cool-down operations. The AFUE is
calculated from performance parameters that are measured under laboratory conditions using the
DOE test procedure. [2] These include a set of temperatures, fuel consumption, and a few other
performance parameters. AFUE does not account for the electricity consumption of the appliance and,
therefore does not include the circulating air and combustion fan power consumption, except to
account for the amount of waste heat produced by these fans.
There are two main types of residential furnaces: weatherized and non-weatherized furnaces.
Weatherized furnaces are generally installed outdoors (often on rooftops), and non-weatherized
furnaces are installed indoors. Manufacturers test non-weatherized furnaces as an isolated
combustion system (ICS), which means it is isolated from the conditioned space where it is located
and the furnace draws combustion and dilution air from the outdoors. Manufacturers test weatherized
furnaces under outdoor conditions. The main difference between a weatherized furnace and a non-
weatherized furnace is that the weatherized furnace is well insulated and has a weather-resistant
external case. The heat loss through the jacket in a weatherized furnace is totally dissipated outside,
resulting in a lower efficiency compared to an equivalent non-weatherized furnace installed indoors.
Non-weatherized gas furnaces can be either non-condensing or condensing. When the flue
temperature is substantially higher than the dew point of the combustion products, the latent heat (the
heat from condensation of water vapor in the combustion products) is lost in the flue. In this case, the
furnace is classified as non-condensing. The AFUE of such furnaces is generally below 83 percent
AFUE. Condensing gas furnaces recover more heat from the combustion products by condensing the
water vapor and can reach efficiencies as high as 96% AFUE.
Mobile home furnaces are a separate class of furnaces, due to three differences. They employ sealed
combustion, pre-heat the combustion air, and have very tight space constraints. Mobile home
furnaces have historically had a lower efficiency standard and were considered as a separate product
in rulemakings in the early 1990s.
Approximately, 2.7 million units of non-weatherized gas furnaces were shipped in the U.S. in 2003,
compared to 0.4 million units of weatherized gas furnaces and 0.14 million units of mobile home gas
furnaces. Thus, we devoted most attention to non-weatherized gas furnaces.

LCC Analysis
Life-cycle costing is a standard engineering economic approach for choosing between alternative
products or designs that provide equal service to the user. It allows for a comparison between
products having different initial and long-term operating costs. The goal of this LCC analysis was to
calculate the LCC for alternative equipment designs in houses that are representative of those in
which new furnaces will be installed. The life-cycle cost consists of two main components: (1) the cost
of buying and installing a furnace, and (2) the operating costs summed over the lifetime of the
equipment, discounted to the present.
To account for the uncertainty and variability in the inputs to the LCC calculation for a given
household and between different households, we used a Monte Carlo simulation. A Monte Carlo
simulation uses a distribution of values to allow for variability and/or uncertainty on inputs for complex
calculations. For each input, there is a distribution of values, with probabilities (weighting) attached to
each value. The simulations sample input values randomly from the probability distributions. For some
variables, such as energy price and climate, the calculations used the values associated with each
sampled household. We used Microsoft Excel spreadsheets with Crystal Ball, an add-on software, to
perform the Monte Carlo analysis.
The LCC analysis estimated furnace energy consumption under field conditions for a sample of
houses that is representative of U.S. homes. These conditions include outdoor climate during the
heating and cooling season which influence the operating hours of the equipment.
We calculated the LCC for a representative sample of houses, one house at a time, using appropriate
values for the inputs each time. We selected a sample of households from the 1997 Residential
Energy Consumption Survey (RECS97). [3] For each sampled household, we estimated the energy
consumption of the furnace, incorporating: (1) baseline design characteristics, and (2) design options
that yield higher efficiencies.
We treated a furnace in a new home differently from one purchased as replacement equipment for
three reasons. First, heating equipment prices are different for new construction and retrofit
applications. Equipment cost for new construction includes a builder markup and does not include

1204
sales tax. Equipment cost for replacement installations includes sales tax and does not include a
builder markup. Second, the financing method (and therefore the discount rate in the LCC calculation)
for new construction is usually a mortgage loan. Financing methods for replacement installations can
take a variety of forms that have different interest rates. Third, new construction tends to be built with
more insulation and more energy-efficient products, compared to houses that receive replacement
installations, and is also concentrated in certain parts of the country. We estimated that 26% of annual
shipments of non-weatherized gas furnaces are installed in new construction.
The change in LCC resulting from a change to higher-efficiency equipment is calculated relative to the
equipment a house would have in the absence of any change in standards (the base case). We used
the distribution of efficiencies in shipments in the year 2000 as the base case. Thus, some houses in
the base case are assumed to purchase higher-efficiency furnaces.

Design Options
We calculated the impacts for furnaces incorporating a variety of design options that increase fuel and
electricity use efficiency. The design options shown in Table 1 were those that met the screening
criteria used in this study.

Table 1: Design Options Considered for Non-Weatherized Gas Furnaces

Design Option Fuel-saving Electricity-
saving
Improved Heat Exchanger Effectiveness X
Condensing Secondary Heat Exchanger X
Modulating Operation X X
Increased Motor Efficiency X
Increased Blower Impeller Efficiency X

Heat exchanger effectiveness can be improved in many ways. Furnace manufacturers optimize the
heat exchanger size and geometry, gas input rate, combustion air delivery system, heat transfer
coefficient and heat exchanger mass, and may apply other enhancements to provide the greatest
comfort, reliability, and safety.
A condensing furnace requires some extra equipment, such as an additional stainless steel heat
exchanger and a condensate drain device. Condensing furnaces also require a different venting
system, since the buoyancy of the flue gases is not sufficient to draw the gases up a regular chimney.
Plastic through-the-wall venting systems are typically used in conjunction with condensing furnaces.
Condensing furnaces present a higher initial cost, but provide significant energy-efficiency gains.
A modulating control is any control that uses either gradual or step-wise adjustment of the furnace
input rate in response to changes in the heating load. Two different types of modulating controls can
be applied to furnaces, two-stage and step control, to decrease fuel and electricity use. Both two-
stage and step modulating gas furnaces are currently available on the market.
Two-stage control refers to a modulating control that cycles a burner between reduced heat input rate
and off or between the maximum heat input rate and off. Two-stage controls are limited to these two
operations. Step modulation can operate at a large number of heat input rate.
Furnaces that operate at substantially reduced output over longer periods of time can provide more
uniform space temperatures, quieter operation, greater efficiency, and reduced emissions. Achieving
these objectives requires that the combustion stoichiometry (the proper fuel/air mixture to assure
clean combustion) be carefully controlled at all firing rates to assure safe operation and minimum
emissions.
Most furnaces sold in the U.S. use forward-curved impellers directly driven by a permanent split
capacitor (PSC) motor. Two design options to improve blower efficiency were considered: 1) an
electronically commutated motor (ECM); and 2) a backward-curved blower with a modified ECM
motor (BC/ECM+). ECM motors have permanent magnets on the rotor. By changing the frequency
and voltage across the stator coils, the speed and torque of the motor can be adjusted. The BC/ECM+
motor operates at a higher speed, has a smaller diameter, and has improved magnets and
electronics. Furnaces with ECM and BC/ECM+ blower motors take advantage of the adjustable speed
and torque of ECM motors to provide constant airflow, regardless of the static pressure. Backward-
curved blowers have different aerodynamic characteristics than forward-curved blowers. For each of
the above designs, the burner operating hours are different, since the furnace efficiency, overall air
moving efficiency, and blower motor electricity consumption are different. Therefore each design
requires a different operating time to provide the same amount of heat to the same house.

1205
Equipment and Installation Costs
The cost of buying and installing a furnace consists of three main elements: the manufacturing cost,
markups in the distribution chain, and the installation cost.
In order to compare the total additional consumer cost of improved equipment efficiency, a baseline
design was defined for each product class. The baseline model establishes the starting point for
analyzing technologies that provide energy-efficiency improvement. Based on the market assessment
and input provided by manufacturers for this study, a baseline model was defined as an appliance
with an efficiency at the minimum level prescribed by EPCA (78 percent AFUE for non-weatherized
gas furnaces), and having commonly available features and technologies.
To estimate the manufacturing cost of alternative furnace designs for this study, several design
options were evaluated that could meet each considered efficiency level. It then selected the design
option(s) it believed manufacturers would most likely implement to achieve a given considered energy
efficiency level. To estimate the manufacturing costs of these design options, this study relied
primarily on a reverse-engineering approach.
Using the manufacturing cost as a base, we applied markups for manufacturers, wholesalers,
contractors, and builders, as well as sales tax. The markups and sales tax was applied depending on
the type of installation (i.e., in new construction or replacement).
The LCC analysis used manufacturing costs from the reverse-engineered cost of the baseline size
furnace. To derive the manufacturing costs for the other sizes, we scaled the reverse-engineered
model costs. To represent the majority of combinations of input capacity and nominal maximum
airflow, we developed generic “virtual” models to represent 25 different combinations of those two
variables. (We refer to these as virtual models because they are not real models on the market.) Each
virtual model had its own cost and energy characteristics. The virtual models include models with the
most commonly-occurring input capacities, with corresponding nominal maximum airflow rates at
static pressure of 124.5 Pa [0.5 inches water gauge].
The installation cost is the cost to the consumer of installing a furnace; it covers all labor associated
with the installation of a new unit or the replacement of an existing one. This includes costs of
changes to the house, such as venting modifications that would be required for the installation. The
estimates of installation costs vary by efficiency level. Installation of 81% AFUE equipment may
require use of more expensive venting systems to prevent problems from condensation. At this
efficiency level, this study estimated that 8% of installations that would require such a venting system.
The size of the equipment, the type of installation, and the installation costs depend on the
households for which the equipment is bought. Characteristics listed in the RECS data set enabled us
to make reasonable assumptions about these factors for each household in the analysis.

Calculating Furnace Energy Use

Estimating the energy consumption of alternative furnaces in the sample houses requires estimating
the heating and cooling loads of each particular house. The loads represent the amount of heating
and cooling required by a house to keep it comfortable for an entire year. The annual house heating
load is the total amount of heat output from the furnace that the house needs for an entire year. This
includes the heat from the burner and the heat generated by the inefficiencies of the blower and the
blower motor. The house heating load was determined for each sampled household from the annual
space heating energy consumption reported in RECS97 and the assumed characteristics of the
existing furnace. The annual house cooling load is the total amount of cooling provided to the house
for the entire cooling season. It includes the cooling provided by the existing air conditioner, and
accounts for the waste heat from the inefficiencies of the blower and blower motor. The house cooling
load was calculated from the cooling energy consumption reported in RECS97 and the assigned
efficiency of the existing air conditioner.
To estimate the energy consumption of furnaces if they used alternate designs rather than the existing
equipment, the LCC analysis used representative virtual model furnaces. These virtual models
incorporate typical features of currently marketed furnaces. One virtual model was created for each of
25 combinations of maximum airflow and input capacity to represent the available range of actual
models. Specifications from actual models were used to determine the specifications for the
corresponding virtual models. The specifications include blower size, motor size, supply air outlet
area, power consumption of the draft inducer and the igniter, and several delay times.
The estimation of heating and cooling loads requires calculation of the electricity consumption of the
furnace blower, since waste heat from the blower and blower motor heats the house. The amount of
waste heat produced depends on the overall efficiency of the blower and blower motor and the
amount of electricity the blower motor consumes. The electricity consumption of a blower motor
depends on the type of motor, the speed at which the motor operates, the external static pressure

1206
difference across the blower, and the airflow through the blower. To calculate blower motor electricity
consumption, the operating conditions (the pressure and air flow) at which a particular furnace in a
particular house will operate were determined. Circulating air blower motor electricity consumption at
full-load steady-state is a function of airflow, external static pressure, and the overall air-moving
efficiency of the furnace.
The blower moves heated air through the house whenever the furnace is on. It also operates in the
cooling season (summer) if the house is air-conditioned. Since the efficiency of the blower will have
different impacts on the overall energy consumption in different seasons, the electricity use calculation
was carried out separately for winter and summer.
Table 2 presents the average energy calculation results from the LCC analysis. These results show 2-
stage modulation reducing gas use but slightly increasing winter electricity use. (In practice, though,
the reduction in gas use may not actually occur, as discussed in the Selected Issues section.) The
reason for the electricity use outcome is that when the blower operates at lower speed, the blower
runs for a longer period.
The 90% condensing furnaces lowers gas use by 11% relative to the 80% AFUE furnace. Note that
these results do not reflect furnace performance of the various design options as it would be under
identical conditions. Rather, the results are influenced by the assignment of equipment to the sample
houses. The ECM option reduces total electricity use by one-third for the 80% AFUE furnace, while
the backward-curved blower with a different ECM motor reduces it by 50%. Note that improving the
efficiency of the blower in a gas furnace reduces electricity consumption, but slightly increases gas
consumption (due to the need to make up for the reduction in heat given off by a more efficient
motor).

Table 2: Average Energy Use for Non-Weatherized Gas Furnaces in the LCC Analysis
Design Options Annual Gas Winter Summer
Use Electricity Electricity
Blower Use Use
AFUE Controls
Motor Type
GJ [MMBtu] kWh kWh
80% single-stage PSC 68.4 [64.9] 475.5 153.9
80% single-stage ECM 69.2 [65.6] 300.2 115.2
80% single-stage BC/ECM+ 69.5 [65.9] 239.3 76.2
80% two-stage PSC 67.0 [63.5] 492.3 153.9
80% two-stage ECM 68.5 [64.9] 247.5 115.2
80% two-stage BC/ECM+ 69.0 [65.4] 201.9 76.2
81% single-stage PSC 67.6 [64.1] 469.8 153.9
90% single-stage PSC 61.1 [57.9] 421.0 153.9
90% single-stage ECM 61.6 [58.4] 277.8 115.2
90% single-stage BC/ECM+ 61.8 [58.6] 223.6 76.2
91% two-stage ECM 60.4 [57.2] 240.0 115.2
91% two-stage BC/ECM+ 60.8 [57.6] 198.6 76.2
92% single-stage PSC 59.8 [56.6] 412.0 153.9
96% step modulation ECM 56.9 [54.0] 226.3 76.2
PSC = permanent split capacitor
ECM = electronically-commutated motor
BC/ECM+ = backward-curved impeller and improved ECM

Other Operating Cost Inputs

In addition to annual energy consumption, calculation of operating costs requires data on the future
prices of natural gas and electricity. We used marginal energy prices to calculate the cost of saved
energy associated with higher-efficiency equipment. Marginal energy prices are the prices consumers
pay for the last unit of energy used. We calculated average and marginal energy prices for each
sample house in 1997 using RECS data. We estimated marginal energy prices from the RECS
monthly billing data by a linear regression of monthly customer bills to monthly customer energy
consumption for each household for which billing data were available. We divided the natural gas
billing data into two seasons: winter and the rest of the year. We estimated the marginal electricity
price for those two seasons as well.
We used the average and marginal prices for 1997 of each sampled house combined with the
forecast annual price changes in EIA’s Annual Energy Outlook 2003 (AEO2003) [4] to arrive at prices
in 2012 and beyond. The projected average residential natural gas price is in the $7.81-7.91/GJ

1207
[$7.40-7.50/MMBtu] range between 2012 and 2019, but then begins to increase at a strong rate,
reaching $8.44/GJ [$8.00/MMBtu] in 2025.
The maintenance cost is the annual cost of maintaining a furnace in working condition. Several
groups of maintenance costs were developed. For the LCC analysis, we assumed a triangular
distribution for maintenance costs to capture the variability of these costs. We assumed a minimum
and maximum of 15% around the average.
The lifetime is the age at which furnaces are retired from service. For non-weatherized gas furnaces,
we used an average lifetime of 20 years, with a range of 10 to 30 years.
We derived the discount rates for the LCC analysis from estimates of the finance cost to purchase
furnaces. Following financial theory, the finance cost of raising funds to purchase furnaces can be
interpreted as: (1) the financial cost of any debt incurred to purchase equipment, principally interest
charges on debt, or (2) the opportunity cost of any equity used to purchase equipment, principally
interest earnings on household equity. Consumers use different methods to purchase equipment for
new and existing homes. Furnaces purchased for new homes are financed with home mortgages.
Furnaces for existing homes (replacement equipment) are purchased using a variety of household
debt and equity sources. We used different discount rates corresponding to the finance cost of new
construction and replacement installations.
We estimated the discount rate for equipment in new housing based on mortgage interest rate data
provided in the Federal Reserve Boards’ Survey of Consumer Finances (SCF). [5] This survey
indicates that mortgage rates carried by homeowners in 1998 averaged 7.9%. After adjusting for
inflation and interest tax deduction, real after-tax interest rates on mortgages averaged 4.2%.
In the residential sector, replacement equipment is usually purchased using cash or some form of
credit. One approach for deriving an average discount rate is to identify the types of credit used to
purchase a given type of equipment (e.g., dealer installment loan, credit card), the associated interest
rates, and the shares of each credit type in total replacement purchases. Such information is difficult
to come by, however, and there are reasons to favor an alternative approach. When a household
makes a major appliance purchase, the short-term effect may be an increase in debt if the purchase is
financed with a dealer loan or credit card, or a decrease in cash if the product is purchased with cash.
However, financial theory suggests that in the medium-term, households should tend to rebalance
their overall equity/debt portfolio to maintain approximately the same relative shares of different
equity/debt classes. According to this line of reasoning, the appropriate opportunity cost (or discount
rate) for purchase of major appliances should reflect a household’s overall equity/debt portfolio, and
not simply the financial or opportunity cost of the debt or equity used to purchase the equipment.
The types of equity and debt likely to be affected by appliance purchases include second mortgages,
credit cards, transaction accounts, certificates of deposit (CDs), U.S. savings bonds, stocks, and
mutual funds. We estimated the shares of each type in the total household equity and debt portfolio
from SCF data and estimated interest or return rates associated with these equity and debt classes
from a variety of sources. The weighted-average real interest rate across all types of household debt
and equity used to purchase replacement furnaces is 6.7%.

Payback Period Inputs

Numerically, the simple payback period (PBP) is the ratio of the increase in purchase (and installation)
price to the decrease in annual operating expenditures (including maintenance). We made the
comparisons based on replacing the base case furnace with a furnace incorporating another design
option. Payback periods are expressed in years. A payback period of three years means that the
increased purchase price for the energy-efficient furnace is equal to three times the value of reduced
operating expenses in the year of purchase; in other words, the increased purchase price is recovered
in three years because of lower operating expenses.
The data inputs to the PBP calculation are the cost of the equipment to the customer and the annual
(first-year) operating expenditures. The PBP calculation uses the same inputs as the LCC analysis,
except that energy price trends and discount rates are not required. Since the PBP is a “simple”
payback, the required energy prices are only for the year in which a new standard is to take effect—in
this case the year 2012.

LCC and PBP Results

For each of the design options, we calculated LCC savings and payback period relative to the base
case equipment assigned to each house, and then averaged the LCC savings and payback period.
The base case assumes, in the absence of new standards, purchase of equipment that best
represents equipment that is sold today. We sampled the AFUE of the base case furnace assigned to
each house from a distribution of AFUEs that is representative of shipments for the year 2000, and is

1208
correlated with climate. Therefore, the base case equipment is not limited to only baseline model
equipment. To some houses, we assigned furnaces that are more efficient than some of the design
options. We considered these as “no impact” cases, since they would not be affected by new
standards.
Table 3 shows LCC and payback results for non-weatherized gas furnaces. Going to 2-stage
modulation results in positive average LCC savings, but, in practice, the reduction in gas use may not
occur, as discussed in the Selected Issues section. The 81% AFUE level shows basically no change
in average LCC. The 90% AFUE condensing furnace has a negative average LCC impact, but it does
have a benefit for houses in colder climates.

Table 3: LCC and PBP Results for Non-Weatherized Gas Furnaces; Fuel Consumption Design
Options
LCC Payback
Design Option Average No Net
Average Net Cost Median Average
(AFUE and technology Savings Impact Benefit
description)* $ $ % of % of % of years years
houses houses houses
80% $9,795 $0 0% 99% 1% 2.1 37.8
80% 2-stage $9,718 $41 33% 27% 40% 8.6 13.5
81% $9,789 -$3 32% 27% 41% 8.8 27.8
90% Condensing $9,917 -$154 56% 26% 18% 17.9 42.5
92% Condensing $9,924 -$166 60% 15% 25% 16.1 41.7
96% Condensing $10,724 -$954 89% 2% 9% 32.3 88.9
Note: Due to the form of the payback calculation, a very small change in operating cost can result in extremely
large paybacks. These extremely large paybacks will skew the average payback. In these cases, median
payback is probably a better indicator.
* All design options include a PSC blower motor, except for 96% AFUE, which includes step modulation and
ECM motor.

Table 4 shows results for the electricity design options. The electronically-commutated motor (ECM)
and BC/ECM+ options have a negative effect on the average LCC. Yet, when these options are used
with two-stage modulation the LCC results are not as negative and in the case of the 80% AFUE 2-
stage modulation with BC/ECM+, the average LCC savings is slightly positive. Therefore, the current
extra cost of these technologies more than offsets the sizable electricity savings.

Table 4: LCC and PBP Results for Non-Weatherized Gas Furnaces; Electricity Consumption
Design Options
LCC Payback
Design Option Average No Net
Average Net Cost Median Average
(AFUE and technology Savings Impact Benefit
description) $ $ % of % of % of years years
houses houses houses
80% PSC $9,795 $0 0% 99% 1% 2.1 37.8
80% ECM $9,873 -$59 60% 27% 14% 23.0 33.7
80% BC/ECM+ $9,822 -$21 51% 27% 23% 17.2 26.3
80% 2-stage, ECM $9,795 -$13 48% 27% 26% 15.4 21.1
80% 2-stage, BC/ECM+ $9,782 $1 45% 27% 28% 14.3 20.9
90% PSC $9,917 -$154 56% 26% 18% 17.9 42.5
90% ECM $10,007 -$226 66% 15% 19% 21.5 47.0
90% BC/ECM+ $9,957 -$180 63% 15% 22% 19.1 42.0
91% 2-stage, ECM $9,898 -$141 58% 15% 26% 16.5 40.6
91% 2-stage, BC/ECM+ $9,878 -$118 58% 15% 27% 16.2 37.8
Note: Due to the form of the payback calculation, a very small change in operating cost can result in extremely
large paybacks. These extremely large paybacks will skew the average payback. In these cases, median
payback is probably a better indicator.

Analysis of National Impacts

This section describes the estimation of national energy savings (NES) and the net value to
consumers from new furnace efficiency standards. For specific efficiency levels, it describes: 1)

1209
cumulative NES in the considered period (2012–2035), and 2) the net present value (NPV) of
efficiency standards for consumers, accounting for products installed in the period considered. The
NPV represents the difference between the present value of operating cost savings and increased
installed costs.

National Energy Savings

We calculated annual NES as the difference between: annual energy consumption (AEC) in the base
case forecast (without new standards), and AEC in a case with new standards. Cumulative energy
savings are the sum over 2012 to 2035 of the annual national energy savings. We calculated the
national annual energy consumption by multiplying the number of existing furnaces (by vintage) by the
unit energy consumption (UEC) (also by vintage).
The UEC is the site energy (natural gas and electricity) consumed by a furnace per year. The annual
gas consumption is directly related to the efficiency AFUE of the unit. Using the energy consumption
calculations described in the LCC section, we determined the national average annual natural gas
consumption that corresponds to each AFUE level. The UES used in the NES model is the average
value from the energy consumption calculations that correspond to each AFUE level.
The NES model considers non-condensing and condensing gas furnaces market segments
separately. The average current AFUE (based on data from the Gas Appliance Manufacturers
Association) is approximately 80 percent for non-condensing furnaces and 93 percent for condensing
types. [6] Most non-condensing furnaces operate at 80 percent AFUE and most condensing furnaces
operate at either 90 or 92 percent AFUE, with just a few percent of market share at other values.
There is a limited amount of historical efficiency data available for furnaces, but the evidence
suggests that there has been little change since the early 1990s within non-condensing and
condensing market segments. Therefore, the base case forecast assumes that current efficiencies
remain constant.
We forecasted the share of condensing furnaces in the base case using the average growth rate in
1991–2000. The share grows from 23% in 2000 to 27.4% in 2012 and 30.4% in 2020.
For non-weatherized gas furnaces, we forecasted shipments as a function of new construction and
expected replacements. In the standards case, the shipments model takes into account that,
increased installed cost of more-efficient gas-fired equipment will cause some customers to purchase
electric rather than gas equipment. Therefore, projected shipments of gas equipment are lower in the
higher-efficiency cases, and there is a corresponding increase in electric heating equipment
shipments.
The stock in a given year is the number of units shipped from earlier years that survive (remain in use)
in the given year. The NES model keeps track of the number of units shipped each year, and the
average UEC of each cohort. The effect of standards is to raise the minimum efficiency to the target
level. The fraction of the market already at or above the standard level is not affected by the
standards. We assumed that the units have an increasing probability of retiring as they age. The
survival function is the probability of survival as a function of years-since-purchase.
In determining national annual energy consumption, we initially calculated the annual energy
consumption at the site. We then calculated primary (source) energy consumption from site energy
consumption by applying a conversion factor to account for losses associated with the generation,
transmission, and distribution of electricity and gas. We used annual site-to-source conversion factors
based on the Lawrence Berkeley National Laboratory (LBNL) version of the National Energy Modeling
System (NEMS), which corresponds to DOE’s Energy Information Administration (EIA’s) Annual
Energy Outlook 2002 (AEO2002). [7] The factors used are marginal values, which represent the
response of the system to an incremental decrease in consumption. Primary energy for electricity
generation is about three times site energy. Natural gas losses include pipeline leakage, pumping
energy, and transportation fuel. Primary energy for natural gas is about 90% of site energy.

Consumer Impacts: Net Present Value

The NPV is the difference between the present value of operating cost savings (PVS), which includes
energy and maintenance costs, and the present value of increased installed costs (PVC), which
includes equipment and installation. We determined PVC for each year from the effective date of the
standard to 2035 and calculated PVS for each year from the effective date of the standard to the last
year when units purchased in 2035 are retired. The NPV of the standards is the sum over all years of
the difference between PVS and PVC.
The average installed cost for the base case forecast and each efficiency level in 2012 comes from
the LCC analysis. Because of the uncertainty concerning future trends in furnace manufacturing, we
assumed no change in average real equipment costs at each efficiency level after 2012.

1210
The total incremental cost of equipment between a standards case forecast and the base case
forecast depends on the average incremental cost of each unit, and on any changes in shipments. In
addition, for the portion of the market expected to switch to electric equipment in the standards case,
we accounted for the cost differential of electric equipment versus a gas furnace and air conditioner
combination.
The annual operating cost savings to consumers are equal to the difference between site annual gas
and electricity consumption in the base case forecast and a standards case forecast, multiplied by the
respective marginal energy price. We accounted for the operating cost of additional electric heating
equipment purchased instead of gas-fired equipment in standards cases.
The savings calculation uses the marginal price for gas and electricity. For the years after 2025, we
applied the average annual growth rate in 2010–2025 for gas and heating oil prices and the average
annual growth rate in 2015–2025 for electricity prices.
The discount factor is the factor by which monetary values in one year are multiplied in order to
determine the present value. We used both a 3 percent and a 7 percent real discount rate in
accordance with the Office of Management and Budget’s (OMB) guidelines. [8] We defined the
present year to be 2001, for consistency with the year in which the manufacturer cost data were
collected.
To illustrate the basic inputs to the NPV calculations, Figure 1 presents the non-discounted annual
installed cost increases and annual operating cost savings at the national level for the 81 percent
AFUE non-weatherized gas furnace (single-stage). The figure also shows the net savings, which is
the difference between the savings and costs for each year. The annual equipment cost is the
increase in equipment price for products purchased each year over the period 2012–2035. The
annual operating cost savings is the savings in operating costs for products operating in each year.
The NPV is the difference between the cumulative annual discounted savings and the cumulative
annual discounted costs.

Consumer Impacts of 81% AFUE Standard

(8% of installations require Category III venting systems)
2012 2022 2032 2042 2052 2062
$0.3

Operating Cost Savings

$0.2 Equipment Cost
Net Savings
Billions of 2001 Dollars

$0.1

$0.0

-$0.1

-$0.2

-$0.3

Figure 1: Non-discounted Annual Installed Cost Increases and Annual Operating Cost Savings
at the National Level for the 81 percent AFUE Non-Weatherized Gas Furnace

NES and NPV Results

The NES model offers a range of possible outputs all of which depend on the inputs used on deriving
the results. Table 5 shows the NES and NPV results for non-weatherized gas furnaces for various
design options. Since few furnaces sold are less than 80% AFUE, there is little to be gained by a

1211
standard at this level. The 81 percent AFUE level has positive energy savings, but the NPV is
negative at 7% discount rate and barely positive at 3%. The 90 percent AFUE level has substantial
energy savings. It has a negative NPV at 7% discount rate, but a large positive NPV at 3%. At the
80% AFUE level, the ECM option has small energy savings, but either slightly negative or positive
NPV.

Table 5: Cumulative National Energy Savings and Consumer Net Present Value for Non-
Weatherized Gas Furnaces
Design Options NES NPV
(EJ [Quads]) (billion 2001 $)
AFUE Controls Motor 3% 7%
Type Discount Rate Discount Rate
80% single-stage PSC 0.03 [0.03] 0.15 0.05
80% two-stage ECM 0.07 [0.06] 0.09 -0.02
81% single-stage PSC 0.46 [0.44] 0.04 -0.29
90% single-stage PSC 4.33 [4.10] 5.11 -0.56
91% two-stage ECM 5.78 [5.48] 3.96 -2.20
96% step modulation ECM 7.54 [7.15] -14.53 -11.61

Selected Issues
Two important issues that arose in the analysis are (1) the limits to improving the efficiency of non-
condensing gas furnaces due to the costs of providing appropriate venting to avoid condensation
problems; and (2) the energy impacts of modulating operation.
As mentioned earlier, installation of 81% AFUE equipment may require use of stainless-steel material
venting systems to prevent problems from condensation. The conditions which determine the type of
venting system are defined based on the operating pressure and temperature in the vent. An 81%
AFUE efficiency level is close to the limit (for non-condensing furnaces) at which the temperature of
the flue gases is sufficiently low to cause condensation in the vent system. U.S. National Fuel Gas
Code (NFGC) [9] venting tables describe the configuration of these systems in terms of length and
diameter of the vents. In the analysis reported here, to insure safe operation, we estimated 8% of
installations of 81% AFUE equipment would require stainless-steel venting system. For this fraction of
the installations, the analysis assigned the appropriate cost.
In the case of modulating furnaces, the DOE test procedure calculates the fuel energy consumption at
maximum input capacity mode, while during the actual operation the modulating furnaces operate
largely in reduced input capacity mode (about 90%-100% of the time). This test procedure assumption
causes overestimation of the calculated fuel energy savings. Therefore, the gas use for modulating
furnaces in actual usage may not decline as shown in Table 2. Note that a currently proposed update
of the ASHRAE test procedure corrects this problem. [10]

Conclusion
Gas furnaces are somewhat unusual in that the technology does not easily permit incremental change
to the AFUE above 80%. The results indicate that for non-weatherized gas furnaces, the 81 percent
AFUE level has positive energy savings, but the NPV is negative at 7% discount rate and barely
positive at 3%. This level shows basically no change (-0.03%) in average LCC.
Achieving significant energy savings requires use of condensing technology, which yields a large
efficiency gain (to 90% or higher AFUE), but has a higher cost. The 90 percent AFUE level has
substantial national energy savings. It has a negative NPV at 7% discount rate, but a large positive
NPV at 3%. The condensing furnace has a negative impact on average LCC, but has a positive LCC
impact for some households (mainly those in colder climates). This result suggests that some States
in cold climates may benefit from establishing a furnace efficiency standard at 90% AFUE.
With respect to electricity efficiency design options, the ECM has a negative effect on the average
LCC. The current extra cost of this technology more than offsets the sizable electricity savings.

1212
References
[1] Title 10, Code of Federal Regulations, Part 430. Energy Conservation Program for Consumer
Products: Energy Conservation Standards for Residential Furnaces and Boilers; Proposed Rule
Furnace and Boiler Advanced Notice of Proposed Rule (ANOPR). July 29, 2004. Washington,
DC. [Docket No. EE–RM/STD–01–350] Can be downloaded at:
http://www.eere.energy.gov/buildings/appliance_standards/residential/furnaces_boilers.html
[2] Title 10, Code of Federal Regulations, Part 430- Energy Conservation Program for Consumer
Products, Appendix N to Subpart B of Part 430-Uniform Test Method for Measuring the Energy
Consumption of Furnaces and Boilers, January 1, 1999. Washington, DC. Chapter II,
Subchapter D. Report No. EE-RM-93-501.
[3] U.S. Department of Energy-Energy Information Administration, A Look at Residential Energy
Consumption in 1997, 1999. Washington, DC. Report No. DOE/EIA-0632(97). Can be
downloaded at: http://www.eia.doe.gov/pub/pdf/consumption/063297.pdf
[4] U.S. Department of Energy-Energy Information Administration. Annual Energy Outlook 2003:
With Projections Through 2025, report no. DOE/EIA-0383(2003). Washington DC: US
Department of Energy; 2003. Can be downloaded at: http:www.eia.dos.gov/oiaf/aeo.
[5] The Federal Reserve Board. Survey of Consumer Finances, 1998. Can be downloaded at:
http://www.federalreserve.gov/pubs/oss/oss2/98/scf98home.html.
[6] Kendall, M., Appendix A - Furnace Shipments; Appendix B - Boiler. Comment # 24 submitted to
Docket Number: EE-RM/STD-01-350 Shipments, April 10, 2002, GAMA. Arlington, VA.
[7] U.S. Department of Energy - Energy Information Administration, Annual Energy Outlook 2002:
With Projections Through 2020, December, 2001. Washington, DC. Report No. DOE/EIA-
0383(2002). Can be downloaded at: http://www.eia.doe.gov/oiaf/aeo.
[8] Office of Management & Budget, Guidelines and Discount Rates for Benefit-Cost Analysis of
Federal Programs, Circular No.94. 1992. See also http://www.whitehouse.gov/omb/circulars/.
[9] National Fire Protection Association, National Fuel Gas Code -1999 Edition, 1999. 1
Batterymarch Park, P.O. Box 9101, Quincy MA. Report No. ANSI Z 223.1-1999.
[10] American Society of Heating Refrigerating and Air-Conditioning Engineers Inc., Method of
Testing for Annual Fuel Utilization Efficiency of Residential Central Furnaces and Boilers,
October, 2003. Atlanta, GA. Report No. BSR/ASHRAE Standard 103-1993R. First Public
Review.

1213
1214
Optimizing Heating Energy in the Domestic Sector
J.J. Bloem, B. Atanasiu

European Commission – DG Joint Research Centre, Institute for Environment

and Sustainability

Abstract
This paper focuses on reducing primary energy use in the domestic sector by changing from
conventional to renewable resources. Starting from an analysis of present available energy
consumption data in the EU-25, the area for higher energy efficiency is identified on hot water
consumption. Introduction of renewable energy technologies in the built environment gives
opportunities to improve the overall energy performance of buildings. In particular the application of
solar energy technology offers a variety of possibilities. The Renewable Electricity Sources Directive
[1] and the Energy Performance of Buildings Directive [2] require all Member States to implement
national regulations within the near future. Other European Directives are under preparation to
stimulate further improvements in energy performance and energy efficiency in the building sector.
Integration of renewable energy sources for heating and electricity in the built environment is also
stimulated through national regulations in a few Member States.

Introduction
The philosophy underlying this study starts from the integral energy performance concept for a
building. All options for energy demand and supply must be considered together if society is to attain
significant levels of energy efficiency and renewable energy deployment. The building sector that
consumes about 40% of the total energy consumption in Europe offers several possibilities to
contribute to achieve higher energy efficiency by introducing distributed renewable energy resources.
The traditional conversion chain from fossil fuel to end-use electricity is composed of a thermal to
electricity step at the power plant. Distribution losses have to be considered also before final
electricity is delivered to the user.
In general, in the household sector an important part of the supplied electricity is used to heat water
for domestic appliances such as hot tap water, washing machines and dishwashers. Integrating solar
energy technology in the built environment will improve overall efficiency and reduce primary energy
since distribution losses do not count for. Moreover solar thermal energy will reduce electricity
demand for water heating and lower the cost for domestic energy use, while at the same time it
contributes to lower greenhouse emissions, supporting the Kyoto protocol. For several Member
States this paper discusses energy and cost analysis based on climate data and electricity prices.
Remarkable is the fact that the present market penetration of solar collectors in Cyprus, Greece,
Austria and Germany do not appear from the economic analysis as the most advantageous countries
for private investment. Electricity prices have not been for private persons the driving force to
purchase a solar collector system. In figure 1 one can find the solar thermal capacity per capita and
the market share in EU-25 countries in 2004. Germany takes almost half of the solar thermal market.
Spanish recent implementation [3] of the EC Directive on the Energy Performance of Buildings
includes an obligation to cover 30-70% of the Domestic Hot Water (DHW) demand with solar thermal
energy. It is expected that this will support the boom in the solar collector market in Europe. In
addition the European certification scheme, the Solar Keymark [4] for solar thermal collectors (EN
12975) and factory made systems (EN 12976) is more and more accepted, both by the industry and
by public authorities.

1215
Figure 1: Source: European Solar Thermal Industry Federation; ESTIF

An important part of electricity consumption in the residential sector is used for heat, being hot water
demand. One of the major problems for statistics is that measured data on heat in the building sector
is hardly available, in comparison to electricity. Most of the information about domestic thermal energy
is derived from empirical methods. Because of this lack of data renewable heating energy knows
several barriers to overcome. Two running EC projects, K4RES-H and THERRA [respectively 5 and 6]
are trying to assess more up-to-date information on renewable energies heat production and
consumption.
First of all it should become visible in statistical data as provided at national level and at European
level by Eurostat. Clear definitions have to be made for thermal energy production and consumption.
Secondly it should become an energy consumption figure and not an energy-saving mean for other
energy resources. In principle one could measure thermal energy as easy as electricity consumption
or telephone usage. In the tertiary building sector it is quit common to express energy consumption in
2
kWh/m /year for space heating and light. A similar approach could be developed for the domestic
sector for different building types and occupancy. Once such an approach has been defined, an
integral energy performance can give clear insight in the different energy flows, energy savings and
use of energy resources.

This paper studies the contribution of households to reduce CO2 emissions and energy consumption
by changing energy resources for electric domestic water heating, moving from conventional to
renewable energy use.

Figure 2: Photo of solar thermal system, rooftop and integrated systems

Energy Consumption in the Residential Sector

To maintain a desired level of comfort people have energy resources and modern technology
available. In Europe about 60% of the energy is used for space heating, 20% for water heating, about
13% for lighting and cooling appliances and another 7% for others things like cooking. Energy
resources are usually electricity, gas, in some countries an important part is coal or oil. In all 25+
European countries on average the electricity consumption in households is around 27% and if the
tertiary building sector is included, 52% [data sources: references 7 and 8].

1216
Taking into account that about 27% of the produced electricity in Europe is consumed in the domestic
sector it offers possibilities for distributed energy resources such as solar energy, to contribute to the
reduction of primary energy sources demand. Solar energy produces electricity and heat close to
where it is consumed. In residential buildings solar thermal technology can be applied very effective in
cases where water is heated only by conventional produced electricity. In a recent study the solar
energies technologies are recognised as the most clean energies [9].

Several reports are available on the subject of domestic hot water energy consumption. A good report
is the analysis of energy efficiency of DESWH, a SAVE report from March 199 [10]. A second report
on EURECO [11] , also a SAVE project on end-use metering in the residential sector, has been used
for this paper also, giving detailed data for electricity consumption. The EURECO report concludes
with potential savings/household by changing appliances that are more efficient in terms of energy
consumption. Potential savings are reported from 20 to 37% for the investigated countries. The
Ecodesign studies [12] by VHK and the market study by BRG Consult in the context of the Boiler
directive revision is expected to supply more quantitative data on the thermal energy use in Europe.

The application of solar thermal technology is

not new. This picture from 1911, shows a solar
water heater installed on the front roof of a
house in Pomona Valley, California, (the panels
are circled above the four windows). Source
Energyquest, California.

Energy Conversion
The conventional conversion from primary energy resource (coal, oil, gas, biomass) to electricity is a
thermo-mechanical process with an efficiency depending on the process. On average the conversion
efficiency is about 47%, see for example UK energy flows in2004 [13] or the the figures for the Dutch
renewable energy monitoring protocol [14]. Taking into account as well the distribution losses of
electricity supply to the consumer for EU-25 which is according to Eurostat around 15% [7], the
electricity consumption efficiency will be around 40% from primary energy supply. This means that for
every toe-equivalent electricity quantity that is consumed in the domestic sector, an equivalent of 2.5
toe primary energy is required .

Figure 3: Electricity production – distribution – end-use

As a distributed renewable energy source, solar energy would contribute strongly to reduce primary
energy consumption and emissions.

Calculation
This study focuses on the domestic use of hot water by appliances and direct use. In general energy
is required to heat the supplied water from 5 – 10 °C up to a desired level of 60°C. The temperature
difference is proportional to the energy consumption. A decrease of conventional energy consumption

1217
can be established by reducing the temperature difference using renewable energy sources. Data
from Eurostat about the number of persons per household reveals 2.5 for EU-15 and about 3.0 for
EU-10. Calculating with an average use of 80 liters of hot water per day per household, an estimate of
electricity consumption per year can be made.
The calculation in this paper assumes a very conservative use of electricity for water heating of 1.5
MWh,th /yr, partly to show results for an average electricity consumption per household and partly to
demonstrate to house-owners the possible results for their specific situation, which can be based on
gas heated combi-boilers. The new Spanish technical building regulation [3] distinguishes for the
minimum solar contribution between electricity and other sources for hot water, depending on five
different climate zones. This figure ranges from 50 to 70% for the building in which electricity being
the energetic source and from 30 to 70% for the other (gas, oil) energetic source.
The problem with Eurostat data is that no solar radiation data is used or available. Concerning solar
electricity from photovoltaic plants the produced MWh are reported by the national offices. Solar
electricity has become very popular in Germany by means of an attractive feed-in tariff, whereas solar
thermal is quite common in Germany, Austria and Greece and is about to boom in Spain. It is
common that solar thermal statistics usually show the solar thermal collector area in square meters.
This has frequently prevented solar thermal from showing up in (renewable) energy statistics, which
are typically given in MW or MWpeak.
2
Solar thermal data are available from the European industry federation (4) and are expressed in m
sold or installed. The International Energy Agency's Solar Heating & Cooling Programme, together
with ESTIF and other major solar thermal trade associations have decided to publish future statistics
in MWth (Megawatt thermal) and have agreed to use a factor of 0.7 kWth/m2 to convert square
meters of collector area into MWth.
Although only the Dutch have supported a non-metering PV subsidy programme in the past, based on
average daily consumption, the feed-in tariff incentive schemes, based an annual production, look to
be the future for solar electricity [3 and 4]. Solar thermal energy production is rarely measured and is
in most countries supported by means of capital incentive schemes.
The intention of this paper is to give an idea of the potential of applying solar thermal energy in the
domestic sector for the situation that electricity is used for producing hot water. Electricity for water
heating is common in all households when considering clothes washing, dish-washers ans sanitary
use. A washing machine and dish-washer consume each, on average 250 kWh/yr per household. In
case of electric hot water production solely by electricity, a household is estimated to consume around
1 to 1.5 MWh/yr [11 and 15] depending on fully use of electricity or electricity assisted hot water
production.

Results
Calculations have been carried out using solar radiation data from the GIS based solar radiation
database [16]. In addition an average household is supposed to be composed of 2.5 persons [7]
consuming 70 liters of hot water / day, for sanitairy use, clothes and dish washers.

2.4. Solar irradiation

Monthly and yearly (figure 4, Table 1) values of solar global irradiance for an optimal inclined surface
facing south [8] have been applied in the calculation for all cases.

1218
Figure 4: Yearly global irradiation at optimal inclination for solar energy applications.
See also http://re.jrc.cec.eu.int/pvgis/pv/

Note that roughly a factor 2 can be applied when Northern Europe is compared with the
Mediterranean area. In practise this means that a house-owner in Scandinavia will need twice more
m2 of solar collectors than in Southern Europe to achieve the same capacity. A further remark has to
be made concerning the optimal inclination because of its definition as the angle the produces the
most energy over the whole year. However during the winter months the low level of solar radiation at
this inclination is not sufficient to fulfil the request for hot water, and therefore the angle of the solar
collectors might be more inclined for more efficiency in the winter than in the summer months.

Table 1: Global irradiance at fixed optimal angle for the following locations:
January July irradiance Year
Location
[Wh/m2/day] [Wh/m2/day] [Wh/m2/day] [kWh/m2/year]
Germany (Munich) 1601 5212 3477 1269
Graz (Austria) 2061 4977 3595 1312
Greece (Athens) 3099 6611 4890 1785
Italy (Milan) 2500 5976 4155 1517
Italy (Calabria) 3277 6877 5211 1902
Spain (Barcelona) 3481 6312 4945 1805
Spain (Malaga) 4157 6831 5460 1993
Sweden 646 5363 3118 1138
Denmark 887 5133 3074 1122
Belgium 957 4845 2980 1088

In principle the optimal inclination should be derived from hot water demand during the winter month
and requires therefore a calculation based on monthly data input. Detailled calculations for a given
load and system dimension can be made for every site in Europe.

1219
Taking into account a utilisation factor of 0.7 for the consumption of hot water produced by solar
collectors, one arrives at the annual need of 1.5 MWh,th based on a conservative use per household
of 70 liters per day for sanitary use, washing machines and dish-washers. Calculations for different
hot water consumption have been carried out as well but the 70 liter figure gives the most informative
information at European scale.

The same amount of hot water produced by electricity would cost the house owner a total for the year
2005 as indicated in table 2. In the calculation for savings on electricity after 10 years is taken into
account a 30% of use of electricity supported hot water heating.

Table 2: Economical analysis

electricty price Costs for excluding 30% Savings4
Location 20051 1.5MWh2 by electricity3 after 10 year
[€/kWh] [€/year] [€/year] [€]
Germany (Munich) 17.85 264.52 185.16 2050
Graz (Austria) 14.13 209.39 146.57 1623
Greece (Athens) 6.88 101.95 71.37 790
Italy (Milan) 19.70 291.93 204.35 2262
Italy (Calabria) 19.70 291.93 204.35 2262
Spain (Barcelona) 10.97 162.56 113.79 1260
Spain (Malaga) 10.97 162.56 113.79 1260
Sweden 13.97 207.02 144.91 1604
Denmark 22.78 337.58 236.31 2616
Belgium 14.81 219.47 153.63 1701
1
Calculations are based on electricity prices including taxes, for each Member State.
2
When electricity is used for heating water, the costs for an average household is given here
3
In particular for the Winter period, other resource have to be utilised for heating water.
4
An increase of 2.5% / year is assumed when calculating the 10 year period of savings on the electricity bill.

Based on these calculations a house-owner may decide to invest in a solar collector installation, with
or without considering financial support by national or regional incentives. Taking into account an
installed system price of 800 - 1000 € / m2 and the size of collectors needed for the local climate, one
may find the following results, Table 3.

Table 3: Payback time

Required Payback time3 Payback
Installation
collector area1 without incentive [yr] time4
Location costs2
[m2] with
[€]
incentive [yr]
Germany (Munich) 2.4 2400 12 9
Graz (Austria) 2.3 2400 15 12
Greece (Athens) 1.7 2000 25 20
Italy (Milan) 2.0 2000 9 7
Italy (Calabria) 1.6 1600 7 5
Spain (Barcelona) 1.7 2000 16 12
Spain (Malaga) 1.5 1600 13 11
Sweden 2.7 2800 17 14
Denmark 2.7 2800 11 9
Belgium 2.8 3000 18 15
1
Calculations are based on the requirement of 1.5 MWh,th/yr.
2
for an average cost of 800 – 1000 € / m2 (note that collector area has been unified at 0.4 m2.)
3
No governmental financial support is considered.
4 2
A capital incentive of 250 € / m is taken into account

The conclusion that can be made from the above calculation is that an Italian houseowner should be
quite interested to invest in a solar collector installation for hot water supply. Even without any
financial support the payback time would be within the warranty and lifetime expectation of the
installation. Note that the calculations are based on conservative figures for utilisation and installation
costs.

1220
The government could put in place incentive schemes based on electricity tax reduction for those
house-owners that substitute electric hot water systems by solar collectors. The reduction could be
based on a proportional part of the tax and the collector area that is installed or on the savings on
electricity consumption (for example up to 1 MWh,th).

Conclusions
Conclusions can be made from three points of view: energetical, economical and political. The house-
owner is most interested in paying less for his energy bill and therefore will be interested in
consuming less energy [17]. He might therefore be looking into options to use energy in a more
efficient way of which changing fuel type is a relative cheap solution. The calculations in this paper
have been based on electricity prices provided from the Eurostat database. The economics might
differ for day/night tariff and regions that have different tax systems. Energy required to heat up water
is energetical seen, more efficient when gas is used. The best would be to apply solar collectors to
support a gas combi system. From political point of view, solar energy in the building sector as a
distributed energy source would be a resource that contributes a lot on reducing emissions and
energy losses from conversion at the powerplants and distribution of electricity. Opportunities will
become available to governments to support industry in developing innovative building components
that integrate these technologies and support investment of house-owners. The conclusion of the
paper is that highest priority therefore should be given to integrate thermal solar energy systems in
the built environment.

References

[1] DIRECTIVE 2001/77/EC Promotion of electricity produced from Renewable Energy Sources in
the internal electricity market.
[2] DIRECTIVE 2002/91/EC on the Energy Performance of Buildings.
[3] Codigo Tecnico de Edificacion; the Spanish New Technical Building Code (Royal Decree
314/2006, 17 March 2006). English translation at www.estif.org
[4] Solar Keymark at www.estif.org
[5] K4RES-H. Key Issuess for Renewable Heat in Europe.
http://www.erec-renewables.org/projects
[6] ThERRA. Thermal energy from renewables – reference and assessment. See
http://europa.eu.int/comm/energy/intelligent
[7] EUROSTAT Pocketbook (2004): Energy, Transport and Environment Indicators, Edition 2004,
EC, ISBN 92-894-7529-3 Luxembourg
[8] Jäger-Waldau A. (ed.). (2004): Status Report 2004; Energy End-use Efficiency and Electricity
from Biomass, Wind and Photovoltaics in the European Union. EUR 21297 EN
[9] Bremer Energie Institute. (2006) Renewable energies – environmental benefits, economic growth
and job creation
[10] DESWH.
[11] EURECO (2002), End-use metering campaign in 400 households of the European Union, SAVE
Programme Contract n° 4•1031/Z/98-267
[12] VHK Ecodesign studies www.ecohotwater.org
[13] www.dti.gov.uk/energy/inform/flowchart.pdf
[14] SenterNovem. Dutch Renewable Energy Monitoring Protocol (2004, english 2006)
[15] CECED (2002).report on Energy Consumption of Domestic Appliances in European Households
[16] JRC (2005) http://re.jrc.cec.eu.int/pvgis/pv/
[17] Bloem J.J., Colli A., Strachan P., (2005) “Evaluation of PV Technology Implementation in the
Building Sector”. PALENC Conference, Santorini, Greece

1221
1222
 Micro CHP

1223
1224
Micro-CHP to Increase Energy Efficiency: Emerging Technologies,
Products and Markets
Jon Slowe

Delta Energy & Environment Affiliation

Abstract
Micro-CHP has the potential to be a disruptive heating technology that decreases carbon dioxide
emissions through increased energy efficiency in households and small businesses. A handful of
products are already being manufactured and offered on a commercial basis, with a raft of product
developers racing to join them. Markets are emerging in Japan, parts of Europe and the U.S.
In 2005, some five manufacturers were selling micro-CHP product (defined as generating 5-kW
electrical and below) on a commercial basis. Around 16,000 units were sold in 2005, representing 31-
MW of generating capacity with a value of approximately €135 million. The DACHS unit –
manufactured by German-based SenerTec, and the ECOWILL unit – developed by Honda, Osaka
Gas, Toho Gas and others – together accounted for over 90 percent of this market in terms of unit
sales.
Although only five companies offer product to customers on what can be termed a commercial basis,
several others express confidence in having products ready for market by 2007 or 2008. Micro-CHP
markets may see exponential growth over the next three to five years. As each micro-CHP installation
is expected to yield between about 0.2 and 0.5 t/CO2/kWe per year, such growth could bring
substantial reductions in CO2 emissions.

Micro-CHP Markets at the Beginning of 2006

Delta estimates that some 16,000 micro-CHP units were sold in 2005, representing some 31-MW of
generating capacity. In this paper, micro-CHP is defined as generating 5-kW or below as broadly this
represents units suitable for mass markets – single family (and small multi-family) homes. This is in
contrast to the less than 50-kW definition under the EC Cogeneration Directive. Japan accounts for
over 75% of these sales through the ECOWILL unit and Yanmar’s Genelight unit. The German market
contributes nearly one fifth of these sales, through the SenerTec DACHS unit and Power Plus
Technolgies’ Ecopower unit. The ECOWILL and SenerTec DACHS unit together account for over
90% of all units sold.
The value of these markets is estimated by Delta to be €135,000,000, up about 20% on the previous
year. Companies with product available for sale and delivery in 2005 are shown in Table 1 below.

Table 1: Global Micro-CHP Product for Sales in 2005

Electrical output
Manufacturer Product (kW)
SenerTec (owned by Baxi) DACHS 5.5
PowerPlus Technology
(owned by Vaillant) Ecopower 4.7
Honda and partners ECOWILL 1.0
Whisper Tech WhisperGen 1.2
Yanmar Genelight 5.0
Source: Delta Energy & Environment

Japanese Markets
In Japan the ECOWILL product continued to sell well, with most major gas companies offering the
product to housing developers and homeowners, with the notable exception of Tokyo Gas. Annual
sales for financial year 2005-6 amounted to over 10,000 units. Yanmar made steady progress with it’s
Genelight product, with sales of a few hundred units. This product is typically sold to small businesses
such as restaurants. Other CHP products in the 6-kWe to 10-kWe range sold in Japan include those
from Aisin Seiki and Sanyo.

1225
All of these products currently sold in the Japanese market are based on internal combustion engines.
Japanese manufacturers have successfully controlled noise and emissions to acceptable levels, and
have extended servicing requirements to long intervals.

European Markets
In Germany SenerTec continues to successfully sell it’s DACHS micro-CHP product, with much of the
sales growth in 2005 coming from the single-family and two-family home market. Annual sales were
more than 2,500 units, with the German market accounting for the vast majority of these sales.
Vaillant, also one of Europe’s top five boiler manufacturers, owns Power Plus Technologies, which
offers the Ecopower micro-CHP product.
The DACHS and Ecopower units are also both built around internal combustion engines. Noise and
emissions, as with the Japanese engines, are controlled down to acceptable levels.
Elsewhere in Europe most other micro-CHP activity focussed on the UK, where electric and gas
retailer E.ON-UK offered Whisper Tech’s WhisperGen micro-CHP unit to households and housing
developers. Current activity is more akin to a market trial rather than a commercial launch. E.ON-UK’s
current objectives appear to be to build a solid base for a mass launch of their system through
establishing an installation and servicing network, and learning about installation, sales and marketing
issues. Actual installations – in the region of hundreds - lagged sales in 2005.
Two new micro-CHP manufacturers – both based in Germany - started taking orders for micro-CHP
products in 2005, although none of these were available for installation until 2006. Otag took orders
for it’s 3-kWe steam-driven LION micro-CHP product, which is designed for single family homes.
Sunmachine sold distribution licenses for it’s 3-kW biomass fuelled Stirling engine in 2005, with units
available for sales in 2006.

North American Markets

In North America activity was more muted in 2005. Vector Cogen, who had brought a 5-kWe micro-
CHP product (built around a Kawasaki internal combustion engine) to market in 2004, closed their
doors to business. Marathon Engine systems continued to plan to bring the Ecopower product to the
U.S, and Climate Energy unveiled a micro-CHP prototype based on Honda’s internal combustion
engine, to be launched in 2006.

Market Prospects for Micro-CHP

Micro-CHP markets promise to take off, possibly with exponential growth in the next four years. The
exact timing that this may happen is not clear, partly due to the uncertainty about product developers
plans to commercialise their micro-CHP products. In this section we analyse the commercial
prospects for micro-CHP in the U.S, Europe and Japan.

Commercial Prospects in Japan

Japanese micro-CHP markets are likely to continue to develop on two fronts. On one hand the market
for the ECOWILL product is expected to continue to grow strongly, particular with Tokyo Gas starting
to sell the unit from the beginning of 2006. We expect to see gas utilities continuing to aggressively
market this product, with this activity partly driven by the threat from electric utilities in marketing all
electric homes.
On the other front, fuel cell systems are currently the focus on intensive research, development and
field testing in Japan. In 2005 Tokyo Gas and others started offering fuel cell micro-CHP systems to
homes through leasing packages. This is not termed as commercial sales, as the activity is more
similar to a large scale field trial rather than the widespread launch number of product for purchase
by customers. By the end of 2005 Tokyo Gas had installed 100 fuel cell micro-CHP systems, and had
the target of reaching 200 installations by the end of March 2006. In total the Government expects
400 such systems to be installed by March 2006.
Tokyo Gas and others are targeting a mass-launch of fuel cell micro-CHP systems in 2008, aiming to
sell thousands of units a year. However further progress is necessary to reduce costs and improve
lifetime in order for these targets to be met. It remains to be seen whether the necessary progress will
in fact be made.
The overall market environment in Japan for micro-CHP is very favourable. Government support,
together with gas utility action and investment, is leading to strong micro-CHP growth and
development of new products – both PEM and solid oxide fuel cells. Japan is likely to continue to be
the leading micro-CHP market in the world for the next few years.

1226
Commercial Prospects in Europe
Micro-CHP markets in Europe currently are dominated by 5-kWe products sold into the German
market. Whilst these markets are showing steady growth, it is unlikely that sales will rise much above
5,000 to 10,000 units a year for these products by 2010.
Perhaps of greater interest is the emerging market for 1-kWe to 3-kWe sized products, designed for
single-family homes. The initial markets for these products are likely to be the UK, Netherlands and
Germany. This market is currently constrained by three issues:
- Product availability
- Engagement by utilities and boiler manufacturer
- Regulatory barriers and issues

A large number of companies are developing micro-CHP product for the European market, with
several of these hoping to commercialise product in 2007 or 2008. Selected developers include
Microgen Energy Ltd, ENATEC, Stirling Systems Ltd, Baxi Group, and Honda (looking to
commercialise their 1-kW engine with a European partner.
Utilities interest in micro-CHP in Europe is still patchy. In the UK E.ON-UK are leading the market by
offering the WhisperGen unit to households. Of the other six major utilities involved in electricity and
gas retail, Centrica have a heads of terms agreement with Microgen, and an agreement with Ceres
Power. There are signs that some of the other four utilities may follow these two utilities in offering
micro-CHP products to customers. But they face significant challenges in developing a brand in the
heating business and building up a installation and servicing infrastructure.
In the Netherlands gas wholesale company Gasunie Trade and Supply is encouraging the
development of the Dutch micro-CHP market. They have sponsored a field trial of WhisperGen units,
involving most Dutch electricity and gas retail companies. They are also working with Microgen to
develop a combi-version of their micro-CHP product. Although the Dutch electricity and gas retailers
are currently very focussed on the opening of electricity and gas residential markets, there are signs
that at least one, and possibly more, are keen to offer micro-CHP products to their customers in the
near future.
The situation amongst German utilities is harder to read. Several have been involved in, and a
number continue to be involved in, fuel cell micro-CHP field trials. However only a few are also testing
other micro-CHP technologies that are arguably nearer to market than fuel cells. With fuel cell micro-
CHP products unlikely to be commercially available much before 2010 (if at all), it will be interesting to
see the degree to which German utilities engage with other forms of micro-CHP.
The engagement of European boiler manufacturers is important as:
- A number of boiler manufacturers are developing their own micro-CHP products, or
are working with technology developers to develop a micro-CHP product.
- Independent micro-CHP product developers (such as Whisper Tech) are likely to
need the manufacturing capability of boiler manufacturers to help them commercialise
their micro-CHP products.
- Some micro-CHP technology developers such as ENATEC have a business model to
license their technology to boiler manufacturers, so they need to see interest from
boiler manufacturers in order for their technology to reach the marketplace.

The degree to which European boiler manufacturers further engage in micro-CHP is not yet clear (all
have at least some micro-CHP activity at present), and this area should be closely watched.
Finally in Europe, regulatory issues need to be resolved in order for micro-CHP market to develop.
There are no major issues, but minor issues are capable of acting as significant barriers to the
development of micro-CHP markets. Activity is already underway in the UK, Netherlands and
Germany to resolve these issues, and it is likely just a matter of time before they are resolved.

Commercial Prospects in North America

The North American market is, generally, a more challenging market for micro-CHP. This is due to low
electricity prices in much of North America, resistance (for example in the form of interconnection
arrangements and tariffs) from electric utility companies, the penetration of low cost warm air furnaces
rather than the boilers found in Europe and Japan, and a greater focus on local air pollutants rather
than greenhouse gas emissions.
However modest micro-CHP activity is likely in the next few years in North America. Climate Energy is
field testing their Honda engine-driven micro-CHP product in winter 2005-6, with a target of launching
commercial sales later in 2006. Marathon Engine Systems, manufacturers of the engine at the heart
of the Ecopower micro-CHP system sold in Europe, are also expected to launch their product in 2006

1227
or 2007. But both of these products are likely to play initially in niche markets, at least for the first few
years of availability.

Carbon Dioxide Savings From Micro-CHP

The CO2 savings that micro-CHP brings are dependent upon a range of assumptions that include:
- Micro-CHP average electricity and thermal efficiencies
- Running hours of the micro-CHP unit
- Average efficiency of the boiler that micro-CHP displaces
- Assumptions about the source of electricity that micro-CHP displaces, the efficiency
of the power station and associated grid loses
Using a proprietary model together with boiler, power plant and grid loses values (using draft
reference values from the implementation of the EU Cogen Directive), and manufacturers and product
developers data (planned and expected), Delta calculates carbon dioxide savings as shown in Figure
1.

1.0

CO2 savings 0.8

per kWe capacity
(t/a) 0.6

0.4

0.2

0.0
Stirling/Rankine
oxide fuel cell

1-kWe Stirling

1-kWe PEM fuel

1-kWe Gas

5-kWe Gas
engine - high
1-kWe Solid

engine - low
efficiency

Engine

Engine
efficiency
1-kWe

cell

Type and capacity of micro-CHP unit

Figure 1: Annual Micro-CHP Carbon Dioxide Savings Compared to Grid Electricity and Boiler
Alternatives. Source: Delta Energy & Environment
Note that the 1-kW units are assumed to be installed in a single family house with the 5-kW gas engine installed
in a multi-family house. Efficiencies (all figures quoted are higher heating values) range from 9% electrical and
74% thermal, to 36% electrical and 38% thermal. The reference values assumed for separate electricity and heat
production are 52.5% efficiency for a new CCGT power plant (assumed for an average ambient temperature of
15oC) and 82% for a gas boiler. Grid loses are 14% for grid-supplied electricity, and 7.5% for electricity exported
back to the grid.

Conclusion
Micro-CHP sales are likely to continue growing strongly over the rest of the decade. From 2008 there
is the chance that an inflexion point may occur in market growth, leading to very rapid growth as new
products suitable for mass-markets become available. This growth may bring substantial reductions in
carbon dioxide emissions.
Japan is likely to continue to be the world’s leading micro-CHP market. The major point to watch here
is whether fuel cells have been sufficiently developed to be widely commercialised in 2008.
In Europe there is more uncertainty. Niche micro-CHP markets are currently showing steady growth,
but there is the potential for explosive growth if new micro-CHP products suitable for single family
homes are brought to market and utilities and boiler manufacturers aggressively push micro-CHP to
customers.
North America has lagged these other two markets, and is expected to continue doing so. Micro-CHP
is expected to establish itself in niche markets over the next few years. More aggressive growth is
possible towards the end of the decade if HVAC manufacturers engage more firmly with micro-CHP,
and other barriers are removed.

1228
Scenarios for Carbon Abatement in Dwellings by Implementation of
Stirling Engine Micro-CHP Systems
David Kane, Marcus Newborough

Energy Academy, Heriot-Watt University, Edinburgh

Abstract
The Building Integrated Micro-Generation Model, a transient thermal and electrical demand estimation
tool, has been developed to predict the performance of four micro-generation systems within a
dwelling of specified construction during four simulation days corresponding to different climatic
conditions. The thermal and electrical demands were estimated for a specified pattern of occupancy,
appliance and domestic hot water usage relating to one of the domestic building variants from the
Carbon Vision Buildings TARBASE programme. Consideration was given to a base case condensing
boiler and three micro-CHP systems of varying electrical capacities (i.e. 0.5kWe, 1.1kWe and 1.8kWe)
and electrical efficiencies (i.e. 18%, 28% and 38% respectively). The results of these simulation
scenarios were quantified. For example, on the overcast winter day, the relative carbon saving
(versus the base case) of each of the respective micro-CHP implementations was 6.4%, 11.7% and
17.7%. The primary aim of this research is to investigate the factors that affect the simulated and
actual performance of micro-CHP systems in terms of carbon emissions. The main factors identified
were transient thermal and electrical demand, micro-CHP system efficiencies and capacities, thermal
and electrical storage capabilities and system control regime. The importance of carbon credit for the
export of electrical output to the national grid to the carbon saving figures was highlighted, as was the
major contribution to carbon saving from electrical import avoidance.

Introduction
In the UK, the domestic sector accounts for 30% of total energy demand, with an average 83% of this
demand for the provision of space heating and domestic hot water, and the remainder for electrical
lights and appliances [1]. Of the approximate 25 million dwellings in the UK, around 17 million are
fitted with domestic gas central-heating (DCH) boilers [2]. The maturity of gas DCH boiler technology
has resulted in efficiencies of up to 93% for gas condensing DCH boilers, leaving minimal possible
carbon savings through further improvements of DCH technology. Therefore, research into carbon
abatement has pursued other avenues, one of which is co-generation of heat and electricity, using a
micro-CHP system. This generates electricity whilst recovering the majority of otherwise wasted heat,
and can reach overall efficiencies in excess of 90% [3]. The carbon saving is primarily attributable to
the reduction in use of centrally-generated network electricity, which has a carbon intensity
(kgCO2/kWh) that is more than twice that of natural gas.
Several candidate technologies for micro-CHP are under development and the ultimate market
potential for countries with extensive natural gas networks is large. One recent study [4] estimated
that the UK market potential for micro-CHP systems based on Stirling Engine prime-mover technology
was 13.5 million units, or around 40% of the housing stock. Hence it is prudent to investigate and
identify the factors affecting the carbon savings achieved by dwelling-integrated micro-CHP systems.
This work has ramifications for the estimation of possible carbon abatement, and the specification of
future micro-CHP systems, as it seeks to illustrate the variation in carbon savings between micro-CHP
system specifications, and target dwelling construction and occupancy.
In this paper, the major factors which influence the effectiveness of the micro-CHP approach are
highlighted. Using the specifically designed, in-development Building Integrated Micro-Generation
(BIM-G) Model, the predicted performance of three generic micro-CHP systems within a specified
dwelling is analysed. The temporal variation of the electrical and thermal demand profiles generated
for the dwelling corresponds to a specified construction, occupancy pattern and appliance utilization
schedule. By evaluation of “real-time” thermal and electrical demand and transient generation
response, on a 5 second basis, dwelling carbon emissions can be calculated with more accuracy, a
hypothesis supported by a recent paper [5].

1229
BIM-G Model Overview
The BIM-G Model is under development as a research tool to produce synthetic thermal and electrical
demand profiles for domestic buildings, at high temporal precision, in order to permit the investigation
of detailed changes in end use and micro-CHP system design. The BIM-G model can be used to
quantify the performance of the system using pre-defined performance metrics. The model utilizes a
bottom-up approach to domestic energy modelling through the use of scripts which specify the nature
and timing of appliance, lighting and DHW usage, dwelling occupancy and thermal comfort
requirements. The model accounts for the transient nature of thermal energy demand in the building
by considering the effects of thermal mass in both the building fabric and space heating distribution
system.
The model itself simplifies the geometry of the dwelling into two 1-dimensional boundaries (separating
the total volume of internal air from the external environment) representing the wall and roof; and a 1-
dimensional boundary separating the internal air from the ground. Heat is exchanged between the
external surface of each element and the environment through convective and radiative processes.
Heat exchange between the internal surface of each element and the internal air is via conductive and
estimated radiative processes. Energy is conducted between surfaces of each element, which acts as
a thermal storage mechanism, governed by the specific heat and volume of each element.

Simulation Scenario Definition

The initial step in this research was to define a set of scenarios, in terms of BIM-G input scripts and
physical dwelling characteristics. The work utilises the domestic variant definitions emerging from the
Carbon Vision TARBASE programme [6], which aims to deliver technological interventions to
appliances, building fabric and energy generation to reduce the carbon footprint of existing buildings
by 50% within a 2030 timeframe.

Physical Dwelling Characteristics

The selected building variant is a detached dwelling, of 180m2 floor area, constructed between 1988
and 1994, conforming to the building regulations of that time. The dwelling is of timber frame
construction, with a 25% glazing/gross external wall ratio and a total ventilation rate (comprising
infiltration and manual ventilation) of 0.76 air changes per hour. The dwelling is assumed to reside in
the area of Oban, on the west coast of Scotland, and all thermal calculations are performed using
corresponding climate data (i.e. external air temperature, and diffuse and direct solar irradiation
measurements).

Household Characteristics and Occupancy Pattern

The household selected to reside in the dwelling is relatively financially prosperous (the relevance of
which is discussed later), and comprises two working parents, one working offspring and a school-
attending child. The scenario defined for this round of simulations represents a typical working day,
characterised by “active occupancy” (i.e. household occupied and occupiers arisen from sleep) for
several hours in the morning, followed by a long period of vacancy, and finally a long period of “active
occupancy” starting from late afternoon, stretching until all occupants retire to bed. Such an
occupancy pattern is utilised by a popular daily steady-state domestic energy estimation model
developed by the Building Research Establishment [7].
The timing of appliance, DHW and manual ventilation events correspond to times of household
occupancy, as do the corresponding metabolic and appliance casual thermal gains.

Appliance Ownership
A range of possibilities exists with respect to the number, variety and age of appliances fitted in a
home. Load signatures vary with appliance type and in some cases depend on age and usage
technique. The aggregate electrical load profile is determined by the transient nature of appliance
(including illumination) usage. For this initial analysis, the dwelling was assigned a set of appliances,
both electrical and DHW related, from which the appliance and DHW usage scripts were composed.
The selection of appliance set was made using ownership data referenced by the assumed socio-
economic status of the household.
The appliance set for this scenario includes an electric oven, gas hob and DHW mixer shower. In
comparison to other possible scenarios, the electrical appliance usage on the simulation day can be
assumed to be low, in comparison to other possible simulation scenarios, as supported by data from a
previous TARBASE study [6].

1230
Climate Data
The climatic data set used with the BIM-G Model was converted from International Weather for
Energy Calculations measurements for Oban, Scotland [8]. For the purposes of this scenario, four
climate day varieties were specified; extreme summer day, extreme winter day with clear skies,
extreme overcast winter day and shoulder day. The spread of climate days allow the micro-CHP
systems to be analysed over a range of space heating requirements, due to variations in external air
temperature, incident solar radiation and thermal energy stored in the building fabric.

Space Heating & DHW Distribution System

As discussed previously, a 1-dimensional space heating distribution system was specified for the
dwelling, in order to estimate the transient nature of thermal supply and demand. This distribution
system entails a heat emitter and pipework, each with a surface area equal to that estimated using an
industry design guide [9] and the Energy Saving Trust’s “Whole House Boiler Sizing Method” [10]. The
volume of space heating water, also estimated by the same method, is heated by the heat-generating
device, i.e. DCH boiler or micro-CHP system. The transient temperature profile of this space heating
water dictates the required heat generator output and space heating input to the dwelling from the
radiator and exposed pipework.
The DHW system comprises a 180-litre DHW tank, insulated appropriately to achieve an average
heat loss of 76 Watts. In specification of the DHW usage profiles for the mixer showers, it is assumed
that the DHW undergoes a 10 degrees Celsius drop between tank and shower water mixer, which
requires a higher draw-off rate than a zero heat loss assumption. For all DHW events, a dead-leg
period of 30 seconds is assumed, to account for sub-requirement temperature water in the DHW
piping which is discarded.
All micro-generation systems defined in this scenario utilise the existing distribution systems and
DHW storage tank.

Micro-Generation System Specifications

Base Case – 20kW Condensing Boiler
In order to quantify carbon emission reduction due to the implementation of micro-CHP to the
dwelling, a reference case must be considered, representative of the dwelling before intervention. In
this case, a standard DCH condensing boiler was specified, with a constant generation efficiency of
88.0%. The maximum thermal output of the boiler is 20kW, and the minimum operating output is 4kW.

Stirling Engine Micro-CHP Systems

The prime mover incorporated in each of the generic micro-CHP systems is a Stirling external
combustion engine, with varying electrical and thermal capacities as detailed in Table 1. As the heat
recovery efficiency remains constant across each system, both the electrical and overall efficiencies
vary, as detailed in Table 1. An auxiliary burner can supply a maximum and minimum of 15kW and
3kW respectively, at a constant generation efficiency of 88.0%. The prime mover is controlled in a
manner where it operates only at full load, i.e. no part load operation, whereas the auxiliary burner, if
called upon, can freely modulate between minimum and full load operation.

Table 1: Electrical and Thermal Capacities and Efficiencies of Generic Micro-CHP Systems
Generation System Electrical Thermal Electrical Electrical
Capacity (kW) Capacity (kW) Efficiency (%) Efficiency (%)
0.5kWe Micro-CHP 0.5 1.97 18.0 88.7
1.1kWe Micro-CHP 1.1 2.44 28.0 90.1
1.8kWe Micro-CHP 1.8 2.53 38..0 91.5

Results & Discussion

Reference Case
The following section displays and discusses the results for the base case scenario, where thermal
energy is supplied from the condensing boiler, and electrical demand solely from the national
electrical grid.
In Table 2, details of the estimated energy demand of the dwelling for each climate day are given,
where the appliance usage and occupancy patterns remain unchanged. Variations in DHW demand
between days are attributable to the proportional assignment of boiler thermal output between space
heating and DHW circuits. As space heating demand, and hence boiler output, increases, small

1231
amounts of thermal surplus (i.e. when demand is below minimum boiler operating output) may be
transferred to the DHW circuit, overshooting the DHW storage tank temperature by an allowable
margin. The increases in electrical demand with total thermal demand are due to increased boiler
firing and circulation pump usage.

Table 2: Estimated Energy Demand of each Climate Day disaggregated by Type

Climate Day Space Heating DHW Demand Total Thermal Electrical
Demand (kWh) (kWh) Demand (kWh) Demand (kWh)
Overcast Winter Day 72.1 15.1 87.2 14.0
Clear Winter Day 62.6 13.9 76.5 13.9
Shoulder Day 37.7 14.0 51.6 13.8
Summer Day 0 14.0 14.0 13.7
NB: Energy Demand estimated as the energy delivered to Space Heating and DHW heat exchangers

The graphs below (Figures 1 and 2) illustrate the transient space heating, DHW and electrical
demand, resultant internal air temperature and external air temperature on the overcast winter day
and shoulder day respectively.

Figure 1: Transient Demand and Internal & External Air Temperatures for Overcast Winter Day

1232
Figure 2: Transient Demand and Internal & External Air Temperatures for Shoulder Day

The primary performance metric of any thermal generation device supplying a space heating system
is the internal air temperature during thermal comfort demand periods, and the difference from
thermal comfort target temperature. The thermal comfort target temperature used during these
simulations is 21°C, where occupants regard any temperature within a +/- 1.5°C band of this
temperature to be acceptable. Therefore, any internal air temperature between 19.5°C and 22.5°C
would fulfil thermal comfort requirements. Furthermore, previous research [11] has identified a band of
temperatures outside this range (16-23°C during winter, 18-25°C during shoulder months) where the
internal air temperature is acceptable to 80% of possible occupants. The thermal comfort demand
periods during this simulation are between 07:00 and 08:30, and 16:30 and 23:00.

Micro-CHP Implementation Cases

The BIM-G Model was used to estimate the transient performance of three micro-CHP systems, with
a spread of electrical output and efficiencies. The system specifications of these units where chosen
for several reasons. The system with 0.5kWe output was chosen to represent a unit with output less
than the average electrical demand (approximately 580 Watts, as calculated on a daily basis). A
1.1kWe unit was designed to match performance information found on a micro-CHP system in current
development; and the 1.8kWe unit was selected to investigate the effects of larger electrical output
systems.
The tables below (Tables 3 – 6) depict the thermal and electrical output of each micro-CHP system on
each climate day type. Additionally, the electrical import and export to the grid is quantified.

Table 3: Estimated System Generation Levels – Shoulder Day

Generation SE Thermal Aux Burner SE Electrical Electrical Electrical
System Generation Thermal Generation Import Export (kWh)
(kWh) Generation (kWh) (kWh) (kWh)
Condensing Boiler 0 51.6 0 13.8 0
0.5kWe SE CHP 7.4 44.0 2.1 12.0 0.3
1.1kWe SE CHP 10.4 41.0 4.6 10.6 1.4
1.8kWe SE CHP 12.6 38.9 7.6 10.2 4.0

1233
Table 4: Estimated System Generation Levels – Summer Day
Generation SE Thermal Aux Burner SE Electrical Electrical Electrical
System Generation Thermal Generation Import Export (kWh)
(kWh) Generation (kWh) (kWh) (kWh)
Condensing Boiler 0 14.0 0 13.7 0
0.5kWe SE CHP 1.8 12.2 0.6 13.3 0.0
1.1kWe SE CHP 2.5 11.5 1.0 12.9 0.2
1.8kWe SE CHP 2.9 11.1 1.7 12.7 0.7

Table 5: Estimated System Generation Levels – Overcast Winter Day

Generation SE Thermal Aux Burner SE Electrical Electrical Electrical
System Generation Thermal Generation Import Export (kWh)
(kWh) Generation (kWh) (kWh) (kWh)
Condensing Boiler 0 87.2 0 14.0 0
0.5kWe SE CHP 16.4 68.9 4.7 9.7 0.4
1.1kWe SE CHP 23.2 62.3 8.9 6.1 2.5
1.8kWe SE CHP 27.9 57.5 17.0 4.9 7.9

Table 6: Estimated System Generation Levels – Clear Winter Day

Generation SE Thermal Aux Burner SE Electrical Electrical Electrical
System Generation Thermal Generation Import Export (kWh)
(kWh) Generation (kWh) (kWh) (kWh)
Condensing Boiler 0 76.5 0 13.8 0
0.5kWe SE CHP 14.1 62.1 2.1 12.0 0.3
1.1kWe SE CHP 19.9 56.4 4.6 10.6 1.4
1.8kWe SE CHP 24.0 52.3 14.6 6.4 7.0

As expected, Stirling Engine thermal and electrical output increases with Stirling Engine electrical
capacity, as does electrical export, whilst electrical import and auxiliary burner thermal output
decreases. The variation in total thermal output is attributable to the maximum and minimum
operating thermal output of each system, and the resulting effects on transient generation.
In the graph below (Figure 3), the carbon emissions from each generation system, on each climate
day, are districtised by source, i.e. total system gas consumption (at an intensity of 0.19kgCO2/kWh)
and electrical import and electrical export (at an intensity of 0.43kgCO2/kWh). In this research, it has
been assumed that full “carbon credit” is given to exported electricity, as it is assumed that it displaces
electrical generation required for other users, at full grid electrical intensity of 0.43kgCO2/kWh.
In Figure 4, there are several trends apparent in the carbon saving results, if the summer climate day
results are discarded. The summer results, in percentage carbon saving terms, appear
disproportionally high in relation to the remaining climate days. This is a consequence of fuel wastage
experience during the boiler “cold start” periods, which is not as apparent during other days with much
higher thermal demand. Ongoing development of the BIM-G model will clarify this effect and it’s
applicability to real-life situations. The carbon saving for any climate day increases with electrical
output of the Stirling Engine, within the size range investigated in this research. This appears to be a
consequence of avoided electrical imports and credited electrical exports, both of which have a
carbon intensity around 2.3 that of the natural gas consumed by the generation systems. The carbon
saving for any generation system increases with the daily thermal demand of the simulation day, a
consequence of longer total prime mover operation time, which in turn increases avoided electrical
imports and credited electrical exports.

1234
Figure 3: Carbon Emissions by source from each Generation System on each Climate Day

Figure 4: % Carbon Savings from Base Case from each Generation System and Climate Day

Conclusions
A transient thermal and electrical demand estimation tool was used to model the performance of four
building integrated micro-generation systems within a specific dwelling during four simulation days
corresponding to different climates. These systems include a base case condensing boiler and three
micro-CHP systems of varying electrical and thermal capacities and efficiencies. The results of these
simulation scenarios were quantified, and carbon emission figures calculated pertaining to selected
operational measurements. The relative carbon savings (versus the base case) of each micro-CHP

1235
implementation scenario were presented for each simulation day to give a first order estimate of their
carbon abatement potential. Although several interesting trends can be identified in these results, the
primary aim of this research was to highlight the factors that affect the simulated and actual
performance of micro-CHP systems in terms of carbon emissions. Further research is required to
quantify the relative effect of each factor on carbon savings, and the cumulative effects on micro-CHP
system sizing and design. In brief, these factors are:
• Magnitude and transient nature of thermal demand, i.e. Space Heating and DHW, including
those factors that directly determine thermal demand, namely:
o Casual gains from appliances and occupants
o Magnitude and timing of thermal comfort requirements
o Exterior climate
o Dwelling construction
• Magnitude and transient nature of electrical demand, including those factors that directly
determine electrical demand, namely:
o Occupancy and appliance use patterns
o Ownership and transient electrical load of appliances
o Exterior climate
• Co-incidence of thermal and electrical demand, and the ratio and transient nature of such co-
incidence
• Ability and efficiency to store excess electrical and/or thermal energy generated on-site
• Ability and/or desire to export electrical generation from the dwelling, and the magnitude of
“carbon credit” assignable to such an export
• Thermal and electrical efficiencies of prime-mover and auxiliary generators, in steady state
and start-up conditions
• Thermal and electrical capacities of prime-mover and auxiliary generators, including minimum
operating power outputs and modulating ability
• Control regime of micro-CHP system, including start-up sequence, electrical and/or thermal
load following ability and technology dependent on/off switching event limitations
Further research is planned to consider each factor in detail, as discussed above, in tandem with
continuing development of the Building Integrated Micro-Generation model.

References
[1] Market Transformation Programme. BNXS28: Did You Know? Energy Facts & Figures.
DEFRA. 2004.
[2] Shorrock L. D. and Utley J. I. Domestic Energy Factfile. BRE Housing Centre. 2003.
[3] Pehnt M. Micro Cogeneration Towards Decentralized Energy Systems. Chapter 1. Micro
Cogeneration Technology. ISBN 3-450-25582-6. 2006.
[4] Crozier-Cole T. and Jones G. The potential market for micro-CHP in the UK. Report P00548.
Energy Saving Trust. London. 2002.
[5] Hawkes A. and Leach M. Impacts of temporal precision in optimisation modelling of micro-
Combined Heat and Power. Energy. 2005.
[6] Peacock A. D. and Newborough M. and Banfill P. F. G. Technology assessment for radically
improving the built asset base. World Renewable Energy Congress. Aberdeen. May 2005.
[7] Anderson B. R. et al. BREDEM 12 Model Description 2001 Update. Building Research
Establishment. 2002.
[8] ASHRAE. International Weather for Energy Calculations (IWEC Weather Files). 2001.
[9] Heating and Ventilating Contractors Association. Guide to Good Practice – Domestic Heating
Specification. 1993. ISBN 0-903783-25-8.
[10] Energy Efficiency Best Practice in Housing. Whole House Boiler Sizing Method for Houses and
Flats. Energy Saving Trust. 2003.
[11] Brager G. S. and de Dear R. A Standard for Natural Ventilation. ASHRAE Journal. October
2000.

1236
Aiming at a 60% Reduction in CO2: Implications for Residential
Lights and Appliances and Micro-generation
Mark Hinnells

Environmental Change Institute, University of Oxford, UK1

Abstract
A number of European Governments, including the UK, have a target for a reduction in CO2
emissions of 60% by 2050 compared to 1990 levels. This paper explores the implications for
residential lights and appliances and for microgeneration which are both mass produced and have the
potential for significant cost reduction and market transformation.
Projected consumption for lights and appliances could be halved through a combination of new
technologies, fuel switching and reduced purchasing of new and energy intensive products. A focus
on energy efficiency is not sufficient.
Microgeneration devices generate heat and or power and are installed in the building or community.
Typical devices include micro-CHP (both Stirling engine and fuel cell), community based CHP,
biomass, heat pumps, solar PV, and solar thermal. Different technologies have potential in different
types of housing. Microgeneration is developing rapidly in the UK. By 2050 the domestic sector could
supply most of its heat and electricity from microgeneration, with conventional heating technologies
(electric heating and gas central heating boilers) almost obsolete.
The cost of such change is discussed. In particular, ‘experience curves’ show that the cost of a new
product is shown to fall in a predictable way with increases in volume. Applying this approach shows
that payback times fall dramatically with significant levels of uptake possible. Policy to deliver such
large changes through Market Transformation is discussed. The development of Energy Services
Companies could help finance the required investment. Personal Carbon Allowances and information
both have a significant role to play.

Introduction
The IPCC recommended that global emissions are reduced to 60% of their 1990 levels by 2050.
Several governments have adopted this target as an objective, including the UK. It follows therefore,
that even though there is great uncertainty over this timeframe given the rate of technical and social
change, policy makers must begin to discover the implications of such radical changes.
In response to this, the 40% House report [1] postulated an internally consistent scenario whereby
emissions from the UK Housing stock could be reduced to 40% of current levels by 2050, even with a
33% increase in the number of dwellings by 2050, as well as an expectation of more heat, more hot
water per person, and a greater penetration of appliances. To achieve this reduction, under this
scenario, requires:
• A large improvement in existing dwellings through refurbishment equivalent of moving from an
average E (SAP rating of 45) to an average A on the label (SAP of 90 or more, close to zero
carbon buildings, requiring all buildings have super-insulated windows, all cavities are filled and
most solid walled properties remaining have external wall insulation, as well as significant
microgeneration). This is
• A zero space heating demand standard for new build housing, together with significant use of
microgeneration
• Significant uptake of Low or Zero Carbon technologies which provide heat and, or, electricity from
devices which are integrated into the building or community (such as CHP, PV, solar thermal,
building integrated wind, heat pumps etc).
• In terms of lights and appliances, the two crucial changes are that all lights are light emitting
diodes (LED) because of their efficiency as well as lighting quality, and all refrigeration appliances
utilise vacuum panel insulation, which reduces heat gain by these appliances to one-fifth of current
levels.

1
The paper is based on work conducted by University of Oxford and published in 40% House, informed by further work for The
Carbon Trust and the Engineering and Physical research Council as part of a project called Building Market Transformation.

1237
Around two-thirds of the carbon saving in this scenario come from energy efficiency measures, and
around one-third came from LZC technologies. Of course, this is not the only way of achieving a
reduction in emissions to 40%, and it is, in no way a forecast. But it illustrates the level of change
needed to achieve this scale of reduction. A set of policies were proposed that could transform the
market and bring this scenario about, based around information, incentives (support for innovation),
and regulation (appliances standards and building standards).
The work begun in 40% House is being continued in more depth in a new project called Building
Market Transformation. BMT is exploring a range of scenarios, and uses sensitivity analysis to assess
the impact of different assumptions for future population, household size, climate etc. BMT has taken
the modeling work from 40% House, and expanded it to produce a range of scenarios, and sensitivity
analysis. The main three scenarios are
• Scenario A – assumes the broad continuation of current policies including the objective of meeting
Kyoto targets as well as incremental changes in technology
• Scenario B – describes a 60% reduction in carbon emissions from housing
• Scenario C – represents an extreme change, with further reductions in carbon emissions beyond
60%
The rest of this paper explores the implications of this work for components in homes that are mass
produced – lights and appliances and low and zero carbon technologies- and which might therefore
be subject to similar market transformation processes. Mass production may bring benefits to in terms
of cost reduction if devices are made in volume (described in literature as technology learning). These
two groups may be very different from building refurbishment and construction which may be much
more site-specific and increasingly needs a whole-house integrated approach.

Lights and appliances

The basic driver is expected to continue to be more households (23.9 to 31.8M homes in 2050, up
33%) and wealthier households. It is possible to envision a near term future where an older, wealthier
population sit in a cooled conservatory in summer, or under a patio heater outside in winter, with a
laptop which gives them via broadband, phone, a television and radio in the corner of the screen,
whilst they browse the stock markets on the internet. None of these products were envisioned even in
work undertaken under the DECADE programme for the UK in 1997 [2].
The other major issue is smaller households. Many of the new households are one person
households because we are marrying later, divorcing earlier, and living longer. In the UK, based on
NHBC quarterly statistics on completions, the number of homes build as detached has reduced from
85,000 to 51,000 a year (a decline of 40%) whilst the number of new flats has increased from just
over 30,000 to just under 60,000 a year (a 95% increase), all in the space of just 4 years from 2000 to
2004. This has dramatic implications for new and in due course, replacement sales of combined
appliances like fridge-freezers and washer-dryers, as well as appliance ownership more generally. It
cannot be assumed that traditional appliance ownership patterns will continue into the future.

There are a range of new technologies. Products are becoming more portable, with wireless
communications and in future wireless power. This has several implications:
• Always-on devices, Power supply and Standby consumption: making devices portable (like
the landline telephone) implies increased ownership (at least in the short term) of batteries and
chargers. Fewer and fewer appliances –even major appliances like washing machines- have an
off-switch, because it isn’t required at low power demand, and the control system consumes under
this level. Transformer based power supplies currently consume 1-7W, but electronic power
supplies with losses of 0.1W are possible. Standby consumption could be a transitional
technology.
• New power sources: fuel cells may power new mobile devices such as phones, MP3 players,
2
PDA’s and laptops by 2007 . The input fuel would thus be gas rather than electricity, with much
lower carbon emissions. In due course, PV may make a comeback for small electronic goods, if
efficiency improves (from 6% to 20% and even 50%), and if costs come down (eg moving away
from silicon based technologies, and using polymers rather than glass as a substraight). PV will
require back-up power sources either in the form of a battery or fuel cell, with both technologies
making significant technical progress.

2
Toshiba, NEC, Hitachi, and Casio have all announced protypes, see for example,
http://www.engadget.com/2004/05/11/casios-laptop-fuel-cell/, and http://news.bbc.co.uk/1/hi/technology/3837585.stm

1238
• Fuel switching could be a real opportunity to reduce carbon in clothes dryers, hobs and ovens,
and even some portable electronic devices with fuel cells. UKDCM2 Scenario A has no fuel
switching, but Scenario C does.
• Electronics design aimed at improved portability, will drive down power consumption both of
standby modes and of associated screens.
• LEDs for lighting: incandescents are around 15-17 lumens per watt, CFLs around 60lpw, and
whilst LEDs are struggling to match CFLs currently, progress is expected to take them to 150
lumens per watt in future. Lighting consumption could fall to a tenth what it currently is.
• VIPs for refrigeration and hot water storage could reduce energy consumption (from standing
losses) to a quarter of what it currently is.

New product groups with changes in technology, population structure and wealth create new
marketing opportunities, for example:
• More and more kitchen appliances, though increased ownership does not necessarily imply
increased consumption. For example, a coffee machine when used displaces a kettle, a sandwich
toaster may displace an ordinary toaster. What it does mean is that the savings from improved
efficiency are harder to capture.
• Home security systems. Often systems have communications and monitoring potential. These may
be combined with home control products, security products with communications capability could
be two way devices not just one way.
• In a warmer climate with a wealthier population, outdoor products may become more significant,
eg patio heaters (patio coolers are reportedly popular in Australia) hot tubs, outdoor lighting and
conservatories.
• In a warmer climate there may also be more cooling in the home.
• increasingly products are merging. Televisions have writeable storage (currently DVDs) together
with software capability for games playing. Home entertainment products and PCs are likely to
come together as a single group. At the same time, phones are merging with portable computers,
and even -with the advent of 3G- with televisions. Televisions are merging with PC’s, given new
USB-based freeview devices for PC’s giving access to all digital television channels. So whilst
main televisions are getting larger, second televisions could be portable or even pocket devices.
The whole area of communications is merging. In practice, this happening quickly and significant
change could be seen before 2010.
Figure 1 below shows 3 Scenarios developed under BMT. In Scenario A, consumption continues to
rise to 120 TWh. The largest contributor to this is home electronics, but new ‘outdoor products’ also
make a contribution. In Scenario B, consumption in 2050 is reduced to 63 TWh and in Scenario C to
54 TWh.

1239
 140
Scenario A
electrical
120 appliances
Scenario B
Useful energy demand (TWh)

electrical
100
appliances
Scenario C
80 electrical
appliances
Scenario A gas
60 appliances

40 Scenario B gas
appliances

20 Scenario C gas
appliances
0

41

44

47
96

99

02

05

08

11

14

17

20

23

26

29

32

35

38

50
20

20

20
19

19

20

20

20

20

20

20

20

20

20

20

20

20

20

20

Figure 1 Three Scenarios for lights and appliances to 2050

A 56% reduction in electricity can be achieved in Scenario C compared to Scenario A by 2050.

Consumer electronics (solid state power supplies, better power management and better screens) and
lighting (LED’s) account for over half of savings. Fuel switching in cooking and wet appliances saves
more electricity than does efficiency, and in outdoor products, fuel switching and avoided high energy
consumers save 14% electricity. Vacuum insulated panels in refrigeration account for 12% electricity.
If policy agreement were forthcoming in time to implement measures by 2010, most of this could be
achieved by 2030, given the current rate of stock turnover. A key issue is that because lights and
appliances are traded good, agreement at EU level would be needed. However development of new
technology is something that the UK could play a key role in.

Low and zero carbon technologies

In the UK, Low and Zero Carbon technologies show a large potential and gas boilers (condensing or
otherwise) could be very much smaller portion of the stock by 2050. The potential for LZC is broadly
categorised into different types of schemes (see Table 1);
• Combustion based opportunities - those that generate heat and may in the process generate
electricity. They would replace a conventional (gas electricity coal or oil) heating system. These
again divide into group systems (community heating) or individual systems (micro CHP, or
biomass). Community heating is predominantly a technology for dense urban communities. Micro
CHP is essentially a suburban or rural technology.
• Rooftop opportunities – these capture wind or sun and may be competing for roofspace. A low
cost replacement opportunity occurs in new build when a roof is installed anyway, or when a roof is
replaced, which may be every 50-100 years, or when a roofspace is converted for living space.
• Rural opportunities – where location dictates availability such as biomass, or heat pumps which
need space and are only really cost effective when a home is not on the gas network.

1240
Table 1 Assumptions for LZC, Scenarios A, B & C
Scenario A Scenario B Scenario C
Dirty fuels • Electric heating is around • Electric heating is around • Electric heating is around
(direct electric heating, solid the same level, the same level, the same level,
fuels) • Coal and oil slow to • Coal and oil slow to • Coal and oil slow to
decline decline decline
combustion based • Total uptake similar to • 7m homes have zero • 7m homes have zero
opportunities condensing boilers, so space heat. space heat
(gas and urban biomass, by 2050, 5% Stirling • Uptake of each • Uptake of each
heat only and CHP) engine 5% fuel cell, 5% technology similar to technology similar to
district heating condensing boiler condensing boiler
• % biomass in micro is uptake, so by 2050 half uptake, so by 2050 just
5%, and in DH is 20% of homes having some over half of homes
form of CHP (15% havesome form of CHP,
Stirling engine, 20% fuel but higher electrical
cell, and 15% district efficiency of fuel cells,
heating). and more biomass
• Biomass is 15% of /energy from waste (5%
stirling engines and 25% Stirling engine, 30% fuel
of Community Heating cell, 20% district heating)
• large scale intervention
means CH good in early
years, with 50% being
biomass or EfW
rooftop opportunities • Ownership grows at half • Ownership grows at the • Ownership grows at the
(PV solar thermal, BIW) the recent growth of same rate as condensing same rate as condensing
condensing boilers until boilers over the last 15 boilers over the last 15
a fifth of roofs have a years, until ownership years, until ownership
device saturates with a third of saturates with half of
• 10% Solar Thermal roofs having installations roofs having installations
• 5% PV – eg • 25% Solar thermal
• 5% solar thermal • 12% solar thermal • 15% PV
• No increase in output • 10% PV • 10% Building Integrated
from roof devices • 7% Building Integrated Wind
Wind • Significant improvements
• Some increase in output in outputs over
incremental
rural opportunities • 3% Heat pumps • 5% heat pumps • 5% heat pumps
(biomass and heat pump) • 3% biomass • 5% biomass • 5% biomass
Total • Gas boilers (heat-only) • Uptake of LZC is higher, • Uptake of LZC reaches
are still the dominant reaching 89% ownership 115% ownership in 2050,
technology for space & in 2050 but with a higher
water heating with low proportion of renewables
uptake of LZC (38% relative to Scenario B
ownership in 2050)

The total proportion of heat and electricity supplied under each scenario is shown in Figure 2.
100%
Scenario C Heat
from LZC
LZC contribution to useful energy demand

90%

80% Scenario C
Electricity from
70% LZC
Scenario B Heat
60% from LZC
50%
Scenario B
40% Electricity from
LZC
30% Scenario A Heat
from LZC
20%
Scenario A
10% Electricity from
LZC
0%
96
99
02
05
08
11
14
17
20
23
26
29
32
35
38
41
44
47
50
19
19
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20

Figure 2 Three scenarios for the percentage of heat and electricity from LZC to 2050

1241
The potential for each technology in BMT has been based on a range of existing studies.

For Community heating, more than 20% of UK homes could be served by Community Heating with
27% of the potential for CHP in London, and 66% in half a dozen major cities [3]. Another study
explored the potential for LZC in London. PB Power, in a study for the London Mayor and
Greenpeace, found that LZC could reduce carbon emissions by 28% by 2025. Of these savings,
district heating made around 90% of the savings because of the density of build in the capital, and
utilising a mix of natural gas, biomass and energy from waste [4]. In the report, Solar thermal, PV, and
micro CHP together saved around 10% of this total. There is nothing new in the potential for district
heating in large UK cities. The Department of Energy ‘Energy Paper’ series explored the theme
several times. Energy papers 20 [5], number 35 [6], and number 53 [7] together identified significant
potential in 9 major conurbations. So whilst Denmark changed its planning regime in the mid-1970s to
stimulate CHP and now supplies around half of homes with CHP, the UK merely talked about it.

The potential for biomass heat only and in larger CHP applications has been explored by Bauen
Woods and Hailes [8], RCEP [9], and Carbon Trust [10]. The potential for Energy from Waste in
large CHP applications is separate [11].

The potential for microCHP was estimated by FaberMaunsell to be around 12million homes [12].
Little overlap was found with community heating because Community heating is a dense urban
technology, and microCHP is a suburban technology, where density is too low to make a heat
network cost effective.

The potential for heat pumps was identified by Hitchen [13]. The potential is limited to homes without
access to gas, and where the disruption could be tolerated, such as new build or significant
refurbishment (eg installation of underfloor heating and trenches outside).

The potential for Building Integrated Wind was estimated as being up to 5TWh by around 2020 [14].
This is of the order of 5% of residential electricity demand, depending on year and scenario.

An integrated study of microgeneration including Building Integrated Wind, PV and solar thermal as
well as CHP below 50 kW was published by DTI [15].
Many of these studies do not take account of the reductions in energy demand through efficiency
improvements, or increased number of dwellings to 2050 (usually only existing build) or changed
energy prices or reductions in cost through technology learning. A key point is that the UK lags behind
many other OECD countries, for example:

District heating with CHP serves less than 0.1% of UK households. In EU-15 23 million people live in
homes served by district heating. In Finland and Denmark around half the population live in schemes
served by district heating.

Heat pumps: In Sweden just under 10% of homes have a heat pump (as well as more than a quarter
of households on district heating).

PV: the UK has supported installation of just over 1000 completed for the UK [16]. Germany has
recently completed a programme of 100,000 solar roofs [17]. Japans PV programme is four times the
size of Germany’s. Japan has a subsidy program goal of increasing PV demand by 400 MW per year
through 2010 and Germany has a goal of 100 MW per year through 2005 [18].

Solar thermal: around 80,000 solar thermal systems are installed in the UK, but over 1m are installed
in the US [19].
Many countries already have higher installations for a given technology than is foreseen in Scenario C
for the UK in 2050. The technology is available, it simply needs the right market framework. The
extent of the bifurcation thus implied in the market place is shown in Figure 3, where the current
conventional sources of energy (gas burned in heat only boilers, and electricity imported from the
network) are fundamentally challenged. The policy framework to bring this potential about is
discussed in a later section.

1242
 700

600

grid_electricity
500
elec_gen_wind
Energy Sources (TWh)

elec_gen_pv
400 elec_gen_chp_DH
elec_gen_chp_fuel_cell
elec_gen_chp_stirling
300
elec_hp
solar_thermal
200 biomass_chp
biomass_boiler

100 gas_chp
gas_boiler
solid/oil
0
96

99

02

05

08

11

14

17

20

23

26

29

32

35

38

41

44

47

50
19

19

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20

20
Figure 3 Changes in energy supply to homes to achieve a 60% reduction in CO2 (Scenario B)

The cost of change

At face value, the scenario proposed appears expensive. However, this is not necessarily the case.
Technology learning is also known as Experience curves and in some literature as ‘learning by doing’.
In common parlance, the phenomena are also known as learning curves. In essence, the theory goes,
for every doubling in global installed capacity or sales, there is a corresponding reduction in costs
which is remarkably consistent for a given technology over successive doublings. If plotted over a
log/log scale (eg price against volume of sales) the relationship becomes linear.
The phenomenon has been observed for a wide variety of products, technologies and industries over
a period of time. It first made significant appearance in the management literature with Boston
Consulting in 1968. Progress ratios average around 82% (i.e. prices decline to 82% of former levels
after each doubling), however, progress ratios have been observed between 60% and 95% (Figure
7), with only one observation of costs increasing with volume (in the airline industry and put down to
changes within the company that led to a loss of learning). The phenomena has been widely used in
energy policy analysis time [20, 21, 22, 23, 24].
Under a 40% House scenario, Hinnells [25] shows that, with very high levels of installation foreseen
under the 40% scenario, capital costs could fall dramatically, and paybacks could fall to less than 5
years. Thus this scenario has plausibility.
Another issue is on whom does the cost fall? Hinnells and Bertoldi [26] show that through the
development of Energy Services Companies (ESCo’s) that the capital cost may not fall on the
householder, but be considered as an investment justified by a return.
To achieve a 60% reduction in CO2, current conventional sources of energy (gas burned in heat only
boilers, and electricity imported from the network) are fundamentally challenged. One of the key
learning issues might not be technical but organisational: a move towards energy services companies
and a focus placed on site integrated solutions or building integrated solutions. In other words,
consumers become generators, and power companies become not suppliers but enablers. Higher
levels of installation of LZC would seem synonymous with the development of ESCos, and
synonymous with a move away from central power generation. This kind of arrangement might foster
a whole-life cost attitude to investment (where payback within the life of a technology at low discount
rates makes it attractive) and a portfolio attitude to investment (eg developing a set of investments
that together give a load curve that matches demand, rather than assessing investments on an
individual basis).

1243
Implementing 40% House: an action Plan timetable to 2050
The following is an indicative Action Plan for achieving scenario C, which would give a 75% reduction
in CO2 by 2050 from the UK Housing stock. The plan focuses on near term actions which are
necessary to avoid missing a 60% cut in 2050. Action needs to start today.
Possible framework incentives include Personal Carbon Trading (PCT) (synonymous with Personal
Carbon Allowances, and investment through Energy Services Companies.

Table 2 Action Plan to 2050

overarching issues Housing LZC Lights and
appliances
Policy - framework incentives eg - Building Regulations - microgeneration - minimum efficiency
framework Personal Carbon Trading - Energy Performance of obligation standards for a wide
(PCT) Buildings Directive - metering range of products
- governance: devolved - government procurement - integrate LZC into - low stand-by
targets and powers - disclosure of energy and building regulations - solid-state power
- EU Directive on energy carbon emissions - integrate LZC into supplies
end-use efficiency & energy Planning
services

Key - integrated delivery - unified rating system for - diversity of - lighting (phasing
issues mechanisms eg Energy new-build and technologies => security out the incandescent
Service Companies refurbishment of supply lamp)
(ESCOs) - ‘whole-home’ - constrain
- metering improvements at point of consumption in
sale/rent instead of consumer
piecemeal grants electronics
2006- - Ofgem consultation on - labels to differentiate - Lower Carbon Building - negotiations for
2009 domestic metering homes by 2007 based on programme launched, energy labels
innovation reports A-G categories (Home £80M capital grants, plus revision commence
- EU Directive on energy Information Pack) accreditation for - roadmaps
end-use efficiency & energy - negotiations to define products. published and cross-
services in force by 2008- single co-ordinated system - Microgeneration sectoral
rigorous testing of of standards for new build strategy includes procurement
technologies to identify best and refurbishment across development of programmes for VIP
smart meter design(s) by all environmental impacts, roadmaps for refrigeration
2010, including interaction incorporating Building technologies. products and
with microgeneration. Regulations - Microgeneration residential LED
- existing small trials of - negotiate with industry on strategy will explore lights and fixtures
smart meters rolled out to move towards monitoring getting full value for initiated
larger scale, including compliance in terms of exported electricity - trials of energy
monitoring of effect on performance in use (including ROCs etc) or display meters for
consumption and peak - headline performance imposing obligation on cooking appliances
demand- 2 year public & standards for 2010, 2015, suppliers. to encourage
stakeholder consultation 2020 published - Technology behavioural change,
process launched in 2007 to - skills & training procurement for linked with smart
identify the optimal PCT programmes for inspectors microchp with the major metering trials
scheme and HIP auditors initiated housing providers (may - negotiations for
- Government establishes an - procurement programmes be at EU or IEA level) to 2015 and 2020
independent Carbon Policy for government estate get it onto the market. minimum standards
Group in 2008 tasked with launched - training programmes commence
setting annual emissions - negotiations initiated for established for installers
reduction targets mandatory free access to - indicative planning
- public sector projects to utility data on consumption requirement: specify
trial ESCos in new-build and - grants/competitions 10% of electricity in new
refurbishment projects established to reward build to be supplied by
- establish a clear framework innovation (based on LZC
of governance & measured performance)
responsibilities for energy
policy, with a strong
emphasis on regional and
local authorities

1244
 overarching issues Housing LZC Lights and
appliances
2010 - smart meters included in - revision to Building - indicative Planning - energy labels
Building Regulations to Regulations - minimum framework and Building revised to be based
cover all new homes standards implemented Regulations: all new on absolute
- 5 year programme to based on performance in homes to have 25% on- consumption rather
retrofit smart meters in all use site renewables than efficiency
households - private rental conditional - indicitave LZC standards
- regional trials of PCT on compliance with obligation on utilities: - support
schemes initiated in 2010 minimum standard 10% by 2020 and 20% programmes for eg
- National PCT scheme - Housing Corporation by 2025 VIPs, LED lighting,
launched in 2012 funding conditional on zero- and fuel switching
- PCT integrate with heat standards opportunities, to
European Emissions - all development on increase market
Trading Scheme for non- government-owned land to share
domestic sector zero-heat standards - minimum standard
- carbon disclosure - incentive programmes for set for bulb sales:
information provided on refurbishment launched sales weighted
sales of domestic and - utility consumption data average of around
transport fuels publicly available 30 lumens per Watt
- ESCos used in all publicly - review status of cooling - 0.1 Watt minimum
funded new build projects to within housing stock and standard for standby
manage energy assess need for action set
infrastructure, including LZC - Building
- TEEC (Tradeable Energy Regulations require
Efficiency Commitment) energy display
implemented at national meters for cooking
level appliances in new
- Local Authorities required build
to establish an address- - retrofit of energy
specific database on energy- display meters for
use in all households by cooking incorporated
2015 into the 5 year smart
meter retrofit
programme
- agree 2015 and
2020 minimum
standards, and
revised labels scales
for a range of goods
2015 - All existing build and new - revision to Building - all new homes required - 5 year programme
build have smart metering Regulations: tighter to generate 50% of to install 2 LED
and billing standard for new-build and demand from LZC fixtures in every
- Carbon Policy Group refurbishment household
reviews national PCT - tighter standard for - minimum
scheme and set next 5 year private-rented sector standards
emissions reduction targets See table 16 implemented for VIP
- All new housing refrigeration
developments to have improved wet
contracted out all energy appliance
services to an ESCO for performance and
delivery. screen technologies
- sales weighted
average efficiency
for bulbs increased
to 60 lumens per
Watt

2020 - Carbon Policy Group - revision to Building - all new homes required - 20 year
reviews national PCT Regulations: tighter to generate 100% of programme to
scheme and set next 5 year standard for new-build and demand from LZC retrofit lighting in all
emissions reduction targets refurbishment households
- target of 10% of all homes - tighter standard for - sales weighted
to be serviced by an ESCo private-rented sector average efficiency
See table 16 for bulbs increased
to 100 lumens per
Watt
- Building
Regulations require
LED lighting in all
fixtures in new build
- second round of
minimum standards
take effect

1245
 overarching issues Housing LZC Lights and
appliances
2025 - Carbon Policy Group - revision to Building - Building Regulations - sales weighted
reviews national PCT Regulations: tighter require that replacement average efficiency
scheme and set next 5 year standard for new-build and boilers or electric heating for bulbs increased
emissions reduction targets refurbishment must use LZC, and to around 150
- tighter standard for replacement of roofs lumens per Watt
private-rented sector requires solar thermal,
See table 16 PV or BIW if conditions
are appropriate
2030 - Carbon Policy Group - revision to Building - total ownership of LZC
reviews national PCT Regulations: tighter across total stock is 50%
scheme and set next 5 year standard for new-build and
emissions reduction targets refurbishment
- tighter standard for
private-rented sector
See table 16
2035 - Carbon Policy Group Review of building - all refrigeration
reviews national PCT regulations products in homes
scheme and set next 5 year have VIP, and high
emissions reduction targets performance energy-
efficient wet
appliances and TV
in all homes
2040 - Carbon Policy Group - ownership of LZC - lighting in all
reviews national PCT average of 1 per home households
scheme and set next 5 year the residential sector converted to LED
emissions reduction targets being zero net importer
of electricity, and
generate 80% of heat on
site

2045 - Carbon Policy Group

reviews national PCT
scheme and set next 5 year
emissions reduction targets
2050 - target of 0.6 tC per capita - average space heat - LZC ownership at DLA consumption
for all energy use (Hillman & across stock is 6,800 kWh 115% reduced by 47%
Fawcett 2004) per annum compared to 1996
- at least half of all homes to
be serviced by an ESCo

Conclusions
This paper explores the implications of a target of a 60% cut in CO2 by 2050 for components in
homes that are mass produced – lights and appliances and low and zero carbon technologies- and
which might therefore be subject to similar market transformation processes, and similar mass
production with cost reduction (described in literature as technology learning). These two groups may
be very different from building refurbishment and construction which may be much more site-specific
and increasingly needs a whole-house integrated approach. Conclusions cover 4 areas:
• Lights and appliances: Consumption could continue to rise by almost a third to 120 TWh even
with present policies to tackle efficiency. The largest contributor to this is home electronics, but
new ‘outdoor products’ also make a contribution. Or consumption in 2050 could be almost halved.
Improved efficiency is not enough, there has to be a focus not merely on improved efficiency but
on reduced consumption. The technology learning effect of EU standards is such that efficiency
standards can be had virtually for free, though perversely, efficient refrigeration appliances without
the volume effects created by standards are likely to remain expensive.
• Low and Zero Carbon Technologies: To achieve a 60% reduction in CO2, current conventional
sources of energy (gas burned in heat only boilers, and electricity imported from the network) are
fundamentally challenged. One of the key learning issues might not be technical but
organisational: a move towards energy services companies and a focus placed on site integrated
solutions or building integrated solutions. Higher levels of installation of LZC would seem
synonymous with the development of ESCos. At the more extreme levels of learning, micro CHP
could in the long term be implemented with little or no cost implication over and above a gas
central heating boiler. The payback period for LZC may be reduced to as low as 5 years.

1246
• Experience curves: Under a 40% House scenario, experience curves suggest dramatic changes
in cost effectiveness, bringing the payback of measures down to reasonable levels, thus making
the scenario plausible.
• Policy: A policy timetable to achieve a 60% reduction is illustrated from today through to 2050.
The scenario could be delivered by a combination of information, incentives and regulation known
as market transformation. Overarching measures may need to include development of Personal
Carbon Trading, development of the market for Energy Services Companies and improved
metering and information provision. It would need to be both integrated and aggressive to achieve
such levels of change.
Whilst the UK has been modeled in detail, the conclusions are likely to fit other economies developed
to a similar extent and with similar space heating demands. Different economies at different stages of
development and with other heating and cooling demands would need more detailed study.

References
[1] Boardman B et al (2005) 40% House. Report from the Environmental Change Institute
University of Oxford. Funded by Tyndall. See www.eci.ox.ac.uk/lowercf/40house.html
[2] Boardman B et al (1997) 2MtC (2 Million Tonnes of carbon) DECADE (Domestic Equipment
and Carbon Dioxide Emissions. ISBN 1-874370-19-2
[3] EST (2003)
www.est.org.uk/uploads/documents/housingbuildings/UK%20CH%20potential%20report_CTF
inal.doc
[4] PB Power (2006) Powering London into the 21st Century. A report by PB Power Energy
Services Division for the Mayor of London and Greenpeace, available from
www.greenpeace.org.uk/contentlookup.cfm?ucidparam=20060316102648&CFID=3335032&
CFTOKEN=31146385
[5] DoE (1977) District heating combined with electricity generation in the United Kingdom,
Energy Paper 20, London HMSO
[6] DoE (1979) Energy Paper series, London HMSO
[7] DoE (1984) Combined Heat and Power District Heating Feasibility Programme: stage 1
Energy Paper 53, London HMSO
[8] Bauen A, Woods J, and Hailes R, (2004) Biopowerswitch. Prepared for WWF International
and Aebiom. http://assets.panda.org/downloads/biomassreportfinal.pdf
[9] RCEP (2004) Biomass as a renewable energy source. Royal Commission on Environmental
Pollution http://www.rcep.org.uk/bioreport.htm
[10] Carbon Trust (2005) Biomass sector review for the Carbon Trust, undertaken by Paul Arwas
Associates.
[11] Institute of Civil Engineers and Renewable Power Association (2005) Quantification of
the potential energy from residuals (EfR) in the UK, commissioned by Institute of Civil
Engineers and Renewable Power Association, undertaken by Oakdene Hollins.
[12] FaberMaunsell et al (2002) Micro-map, mini and micro CHP – Market Assessment and
Development Plan. SAVE. http://www.microchap.info/
[13] Hitchen R (2004) The UK heat pump market. IEA Heat Pump Centre Newsletter 22 (4).
http://www.heatpumpcentre.org
[14] Dutton, Halliday and Blanch (2005) The feasibility of Building-Mounted/Integrated Wind
Turbines: Achieving their potential for carbon emissions reductions. Part funded by the carbon
Trust. Available from www.eru.rl.ac.uk/pdfs/BUWT_final_v004_full.pdf
[15] DTI (2005) Potential for Microgeneration Study and Analysis, Final Report 14th November
2005, commissioned by Energy Saving Trust, and undertaken by Element Energy. See
www.dti.gov.uk/energy/consultations/pdfs/microgeneration-est-report.pdf
[16] EST (2005)
www.est.org.uk/uploads/documents/housingbuildings/InfoForInstallers23_9_05.pdf
[17] IEA (2003) http://www.oja-services.nl/iea-pvps/nsr03/download/deu.pdf
[18] EIA (undated) http://www.eia.doe.gov/cneaf/solar.renewables/rea_issues/solar.html
[19] Crest (website) http://solstice.crest.org/renewables/seia_slrthrm/#Intro, downloaded on
21.3.06
[20] IEA (2000) Experience Curves for Energy Technology
[21] Kamp, L., Smits, R., Andriesse C., (2004) Notions on learning applied to wind turbine
development in the Netherlands and Denmark, Energy Policy 32 (2004) 1625-1637.

1247
[22] McDonald A, and Schrattenholzer L, (2001) Learning rates for energy technologies, Energy
Policy 29 (2001) 255-261.
[23] IEA (2003) Experience Curves: A Tool for Energy Policy Analysis and Design, IEA workshop
Paris: 22-24 January 2003, see www.iea.org/dbtw-
wpd/Textbase/work/workshopdetail.asp?WS_ID=89
[24] PIU (2001h) Technical and economic potential of renewable energy generating technologies :
potentials and cost reductions to 2020 www.strategy.gov.uk/downloads/files/PIUh.pdf
[25] Hinnells M (2005) The cost of a 60% cut in CO2 emissions from homes: what do experience
curves tell us? Paper to BIEE conference, Oxford Sept 05, downloadable from
www.biee.org and click the link 'Downloads - 2005 presentations', scroll down for 22-23
September 2005.
[26] Hinnells M, and Bertoldi P, (2006) Liberating the power of Energy Services in a liberalised
energy market; next steps for the UK. Paper to EEDAL 06 (Energy Efficiency in Domestic
Lights and Appliances, London 2006).

1248
Massive Coordination of Residential Embedded Electricity
Generation and Demand Response using the PowerMatcher
Approach
I.G. Kamphuis, J.K. Kok, C.J. Warmer, M.P.F. Hommelberg

Energy Research Centre of the Netherlands

Sustainable Energy in the Built Environment

Abstract
Different driving forces push the electricity production towards decentralization. The projected
increase of distributed power generation on the residential level with an increasing proportion of
intermittent renewable energy resources poses problems for continuously matching the energy
balance when coordination takes place centrally. On the other hand, new opportunities arise by
intelligent clustering of generators and demand in so-called Virtual Power Plants. Part of the
responsibility for new coordination mechanisms, then, has to be laid locally. To achieve this, the
current electricity infrastructure is expected to evolve into a network of networks (including
ICT(Information and Communication Technology)-networks), in which all system parts communicate
with one another, are aware of each other's context and may influence each other. In this paper, a
multi-agent systems approach, using price signal-vectors from an electronic market is presented as
an appropriate technology needed for massive control and coordination tasks in these future
electricity networks. The PowerMatcher, a market-based control concept for supply and demand
matching (SDM) in electricity networks, is discussed. The results within a simulation study show the
ability to raise the simultaneousness of electricity production and consumption within (local) control
clusters with cogeneration and heat-pumps by exchanging price signals and coordinated allocation
using market algorithms. The control concept, however, can also be applied in other business cases
like reduction of imbalance cost in commercial portfolios or virtual power plant operators, utilizing
distributed generators. Furthermore, a PowerMatcher-based field test configuration with 15 Stirling-
engine powered microCHP's is described, which is currently in operation within a fieldtest in the
Netherlands.

Background
Traditionally, electricity distribution infrastructures are based on a hierarchical, top-down flow and
distribution of power. The infrastructures were designed and economically validated with accounting
models of energy companies that typically had a time horizon of 20 to 50 year. One of the
consequences of liberalisation is that power networks are being utilized with decreasing reserve
capacity and investment capital preferably has a much shorter payback time horizon. This leads to an
increase of smaller capacity installations operating in a distributed manner. Embedding small-scale
renewable energy resources, with intermittent production, on the other hand, poses another challenge
to match supply and demand of electricity in real-time at several levels in the grid. State-of-the-art
information processing hardware and communication technology networks (ICT) form part of the
solution for coordination and concerted control of demand and supply of electricity in these distributed
environments ([1],[2],[3]).
An ongoing change in the worldwide energy supply is this growing penetration of distributed electricity
generation. Distributed Generation (DG) can be defined as a source of electric power connected to
the distribution network or to a customer site (“behind the meter”). This approach is fundamentally
distinct from the traditional central plant model for electricity generation and delivery. Driving forces
behind the growing penetration of DG are [4-7]: environmental concerns, deregulation of the electricity
market, diversification of energy sources, energy autonomy, and energy efficiency.

The growing share of DG in the electricity system may evolve in three distinct stages:
• Accommodation. Distributed generation is accommodated in the current market. Distributed
units are running free, while centralized control of the networks remains in place.

1249
• Decentralization. The share of DG increases. Virtual utilities optimize the services of
decentralized providers through the use of common ICT-systems. Central monitoring and control
is still needed.
• Dispersal. Distributed power takes over the electricity market. Local low-voltage network
segments provide their own supply with limited exchange of energy with the rest of the network.
The central network operator operates more like a coordinating agent between separate systems
rather than controller of the system.
In specific parts of the world there are already signs of the decentralization stage (see for instance
[6],[7].) During the second and third stage of DG growth, the lower parts of the electricity grid are
expected to evolve from a hierarchical top-down controlled structure into a network of networks, in
which a vast number of system parts communicate with each other and influence each other. In this
scenario, the standard paradigm of centralized control, which is used in the current electricity
infrastructure, will no longer be sufficient. The number of system components actively involved in the
coordination task will be huge. Centralized control of such a complex system will reach the limits of
scalability, computational complexity and communication overhead.
This paper describes a novel control concept for automatic matching of demand and supply in
electricity networks with a high share of distributed (co-)generation on the residential level. In this
concept DG, demand response, and electricity storage are integrated using the advanced ICT
technology of distributed control. The control concept opens the possibility to introduce massive scale
distributed coordination additional to the existing central coordination, when the share of
distributed generation increases. The coordination mechanism has the aim to increase the portion of
DG that can be accommodated under normal operational conditions in the power grid.

ICT for distributed coordination

As a result of the electricity evolution, described above, the electricity infrastructure will become more
and more inter-linked with ICT-infrastructure components. The architecture and algorithms of this ICT-
infrastructure must be adapted to the technical structure of the (future) electricity net and the
connected producing and consuming installations, but also to the structure of the liberalized energy
markets. This ICT-architecture and associated algorithms must be designed using a strong system-
wide viewpoint, but must also consider stakes of local actors in the system.
The PowerMatcher approach is based the computer science technologies Multi-agent Systems (MAS)
and Electronic (Virtual) Markets [8-10]. The combination of these two technologies results in a
combination of properties, interesting from the viewpoint of coordination in electricity networks:
• In multi-agent systems a large number of actors are able to interact, in competition or in
cooperation. Local agents focus on the interests of local sub-systems and influence the whole
system via negotiations with other software agents. While the complexity of an individual agent
can be low, the intelligence level of the global system is high. For instance, for a temperature
control system in a dwelling, the deviation from the set-point temperature is a measure for the
steepness of theprice-response.
• Multi-agent systems implement distributed decision-making systems in an open, flexible and
extensible way. Communications between actors can be minimized to a generic and uniform
information exchange. In the PowerMatcher approach, from the appliance only a volume/price
curve is sent to the coordinator/auctioneer; the auctioneer only spreads the allocations of the
resource.
• By combining multi-agent systems with micro-economic principles, coordination using economic
parameters becomes possible. This opens the possibility for the distributed coordination process
to exceed boundaries of ownership. The local agent can be adjusted by the local stakeholder, and
does not fall under the rules and conditions of an intermediate or central authority.
• Using electronic markets a Pareto efficient system emerges, i.e. a system that optimizes on a
global level, while at the local level the interests of all individual actors are optimally balanced
against each other.
Of course, the total resulting system (the electricity infrastructure plus the ICT infrastructure) must be
dependable, since the power grid is a critical asset in the modern society. Most developed countries
currently have a highly dependable electricity supply, and any changes to the system must not
weaken it but rather strengthen it. Further, the system as a whole must be secure, i.e. hardened
against hackers and cheaters.

1250
The PowerMatcher basic concept
The PowerMatcher is a market-based control concept for supply and demand matching (SDM) in
electricity networks with a high share of distributed generation. SDM is concerned with optimally using
the possibilities of electricity producing and consuming devices to alter their operation in order to
increase the over-all match between electricity production and consumption. In the PowerMatcher
method each device is represented by a control agent, which tries to operate the process associated
with the device in an economical optimal way. The electricity consumed or produced by the device is
bought, respectively sold, by the device agent on an electronic exchange market. The supply-demand
mechanism is explained in figure 1. Articulation of the demand response is translated into a demand
curve; ability or uncertainty to shift generation is modeled into a supply curve bid. Supply and demand
meet at an equilibrium point, which gives the price.
The electronic market is implemented in a distributed manner via a tree-structure of so-called SD-
Matchers, as depicted in Figure 2 and adopted from [1,2]. An SD-Matcher matches demand and
supply of a cluster of devices directly below it. The SD-Matcher in the root of the tree performs the
price-forming process; those at intermediate levels aggregate the demand functions of the devices
below them. An SD-Matcher cannot tell whether the instances below it are device agents or
intermediate SD-Matchers, since the communication interface of these are equal. The root SD-
Matcher has one or more associated market mechanism definitions, which define the characteristics
of the markets, such as the time slot length, the time horizon, and a definition of the execution event
(e.g. “every whole quarter of an hour”, “every day at twelve o’ clock”). When an execution event
occurs, the root SD-Matcher sends a request to all directly connected agents to deliver their bids. The
device bids are aggregated at the intermediate matchers and passed on up-wards. The root SD-
Matcher determines the equilibrium price, which is communicated back to the devices. From the
market price and their own bid function each device agent can determine the power allocated to the
device.

Figure 1 Principle of supply-demand matching

SD-Matcher

Price
Bid
Bid
Price

SD-Matcher SD-Matcher

Bid Bid
Bid Bid
Bid Bid
Price Price
Price Price
Price Price

Device Device Device Device Device Device

Figure 2: Hierarchy of supply & demand matchers in the PowerMatcher concept. The SD-
Matchers implement a distributed electronic market.

1251
Device agent types and strategies
From the viewpoint of supply and demand matching, devices can be subdivided according to their
type of controllability into the following classes:
• Stochastic operation devices: devices like solar and wind energy systems of which the power
exchanged with the grid behaves stochastically. In general, the output power of these devices
cannot be controlled, the device agent must accept any market price.
• Shiftable operation devices: batch-type devices whose operation is shiftable within certain
limits, like (domestic) washing and drying processes. Processes that need to run for a certain
amount of time regardless of the exact moment, like, assimilation lights in greenhouses and
ventilation systems in utility buildings. The total demand or supply is fixed over time.
• External resource buffering devices: devices that produce a resource, other than electricity,
that is subject to some kind of buffering. Examples of these devices are heating or cooling
processes, whose operation objective is to keep a certain temperature within two limits. Devices
in this category can both be electricity consumers (electrical heating, heat pump devices) and
producers (combined generation of heat and power).
• Electricity storage devices: conventional batteries or advances technologies like flywheels and
super-capacitors coupled to the grid by a bi-directional connection. The agent bidding strategy is
to buy energy at low prices and sell it later at high prices.
• Freely-controllable devices: devices that are controllable within certain limits (e.g. a diesel
generator). The agent bidding strategy is closely related to the marginal costs of the electricity
production.
• User-action devices: devices whose operation is a direct result of a user action. Domestic
examples are: audio, video, lighting and computers. These devices are comparable to the
stochastic operation devices: their operation is to a great extent unpredictable and the agent must
accept any market price to let them operate.

In all described device categories, agent bidding strategies are aimed at carrying out the specific
process of the device in an economically optimal way, but within the constraints given by the specific
process. Note that this self-interested behavior of local agents causes electricity consumption to shift
towards moments of low electricity prices and production towards moments of high prices. As a result
of this, the emergence of supply and demand matching can be seen on the global system level.
Device constraints and user constraints are to be dealt with and introduced in a very cautious way.
E.g. micro-CHP can't be operated with too many on/off subsequent cycles in order to save overall
operating time. Adequate feedback has to be given to users to meet socio-economic and acceptance
requirements of partly autonomously operating devices. Using this technology may not lead to major
changes in lifestyle of users without any reward.

Simulation Case
In a simulation study the impact of distributed supply and demand matching applied in a residential
area was investigated. In the study, a cluster of 40 houses, all connected to the same segment of a
low-voltage distribution network (an LV-cell) were simulated. Within the LV-cell an exchange agent
implements the root SD-Matcher. The LV-cell is externally connected to a medium voltage network.
Through this connection power can be obtained form and delivered to other parts of the distribution
network.
Each home has a Home Energy Management gateway, which implements the local energy
management strategy of the house. The HEM-box incorporates the intermediate SD-Matcher
functionality, together with energy performance feedback to the user, and the possibility for the user to
set cost and task preferences. The latter makes it possible to set agent parameters of devices without
a user interface. Within the LV-cell an exchange agent implements the root SD-Matcher. The LV-cell
is externally connected to a medium voltage network. Through this connection power can be obtained
form and delivered to other parts of the distribution network. The electricity surplus of the cluster is
delivered to an external electricity supplier, which delivers electricity to the cluster in case of local
shortage. The external supplier can either be a full player on the local electronic market or set tariffs
for delivery and retribution. In the latter case, the external tariffs are not influenced by the local price
formation, and, typically, the retribution price will be lower than the delivery price. Then, the
equilibrium price on the local electronic market will be bounded by the external tariffs. Half of the 40
simulated dwellings are heated by heat pumps (electricity consumers), the other half by micro-CHP

1252
units (small-scale combined heat-power, producers of electricity and heat). The micro-CHPs are also
used for production of hot tap water. Washing machines are operated as shiftable operation devices
with a predefined operational time window; electricity storage is present in the form of batteries;
stochastic operation devices are present in the form of photovoltaic (PV) solar cells and small-scale
wind turbines; and user-action devices are represented as lights.
Half of the 40 simulated dwellings are heated by heat pumps (electricity consumers), the other half by
micro-CHP units (small-scale combined heat-power, producers of electricity and heat). The micro-
CHPs are also used for production of hot tap water. Washing machines are operated as shiftable
operation devices with a predefined operational time window; electricity storage is present in the form
of batteries; stochastic operation devices are present in the form of photovoltaic (PV) solar cells and
small-scale wind turbines; and user-action devices are represented as lights.
Figure 3 and 4 show the result of a typical simulation run for the LV-cell simulation case. In both plots
the total consumption and the total production in the cluster have been summed into a single plotline,
while production is regarded as negative consumption. The top plot shows the reference case in
which all devices are free running. In this case all heating devices are on/off controlled, washing
machines start their operation at the start of their operational time window, and batteries are excluded
due to the absence of a real-time price signal according which they can be operated. In the bottom
plot the SDM-controlled case is shown. Interesting features are:
• Around the 25 15 minutes period there is a peak in electricity demand caused by the
th

simultaneous starting of a number of heatpumps. Although there is also a small peak in local
production at that moment, the greater part of the electricity needed to meet the peak demand is
delivered from the external connection to the mid-voltage network. In the SDM-controlled case the
peak in external feed-in is 30% lower, due to the reaction of different devices to the price peak on
the electronic market at that moment. Consuming device agents shift part of their operation to
other moments in time, producing agents shift as much as production as possible to this moment,
and battery agents react by switching to discharging mode. In this particular case, consumption
reduction accounts for 50% of the peak reduction, battery discharging accounts for 37%, and
production increase causes another 13%. From the viewpoint of electricity distribution systems,
this is an important result. The highest expected peak demand of a low-voltage net segment
determines the capacity the coupling transformer and the network cables or lines. Reducing the
peak demand lowers network investments in case of building new sub-networks, and defers
network reinforcements in case of demand increase in existing nets.
• Introducing supply and demand matching results in a more flat and smooth profile of the electricity
fed in from the mid-voltage network. Fluctuations in local consumption and local production are
damped, and the mutual simultaneousness in the remaining fluctuations is high. The standard
deviation of the feed-in from the MV-net in figure 3 is 58% lower in the SDM-controlled case. This
means that predictability of the cluster as a whole is increased by the automatic matching of
demand and supply.
Figure 4 further nicely illustrates, that the internal price development on the internal device market
acts as a coordination incentive.

1253
Figure 3: The result of a typical simulation run for the LV-cell simulation (see text).

1254
Figure 4: Internal price formation development during the internal market simulation

Figure 5 Price development on the APX-market in 2003 as a function of wallclock time and
daynumber

1255
Figure 6 Imbalance (over/underrealisation compared to the programme) cost per MWh as a function
of time in 2003

Utilizing distributed coordination for operating virtual power plants on markets

The flexibility in micro-CHP operation utilized by clustering in a VPP can be put into value in different
ways. One might think of:
• Trading the output of the VPP on the day-ahead power market (e.g. the Dutch APX or the
Scandinavian NordPool). Figure 5 shows the highly variable price development on the APX
power market in the Netherlands. Actual price peaks were above 2000 Euro/MWh. The peaks
round mid-day in august and at the end of the afternoon during winter provide opportunities
for flexible electricity generating capacity and demand.
• Trading the output of the VPP on the imbalance or spinning reserve market. Figure 6 shows a
similar 3D-plot for the imbalance market in the Netherlands. If demand/generation shiftability
fits into the timeframe of this market, considerable benefits may be gained.
• Support the local distribution system operator (DSO). For instance, by reducing the local peak
demand of low-voltage grid segments, to defer reinforcements in the grid infrastructure (e.g.
substations and cables).

1256
Fieldtests in a Virtual Power Plant setting

ECN and Gasunie Virtual Power Plant Field-test

Clustered operation of 15 micro-CHPs for grid-support and trading

...
Local Local Local Local
VPP-Node VPP-Node VPP-Node VPP-Node

GPRS
Wireless Central VPP-
Communication Controller

Figure 7 Fieldtest configuration with micro-CHPs

Currently a fieldtest with 15 Stirling based CHP's is in the process of rollout (see fig 6). The Dutch
natural gas company Gasunie is undertaking a rollout and measurement program of domestic micro-
CHP installations. A measurement project of approximately 50 microCHP installations is started.
WhisperGen type installations are placed at the premises of people employed by Gasunie and a
number of cooperating Dutch electricity retail companies. The primary goal of this field test is to
monitor the installation in typical Dutch households and to gain user experiences. The current
available micro-CHP systems are heat-demand driven, i.e. they produce heat and electricity at
moments of heat demand, either from room heating or warm tap water usage. Uncoupling the
electricity production from the heat demand can raise the value of produced electricity. This can be
done by utilizing the inherent heat buffering capacity of the building and the warm tap water buffer.
The electricity and heat functions of the CHP can be uncoupled further by adding a device for heat (or
electricity) storage.
In the first half year of 2006, Gasunie and ECN cooperate to add an ICT-infrastructure and local
intelligence to a series of 15 of the before-mentioned 50 microCHP's. Goal of this field test is to
demonstrate the ability to act as a cluster in a Virtual Power Plant (VPP) setting to attribute to a
common control goal. The PowerMatcher control concept forms the core software of this VPP.
The field test VPP is designed to support the local distribution system. Due to the open character of
the control concept, flexible electricity loads (i.e. demand response) can easily be added to the
system in later stages of the field trial. The field test runs under auspices of the Smart Power System-
consortium, a Dutch industrial and research conglomerate developing technology and business cases
for microCHP based virtual power plants.

Conclusion
Various drivers push the production of electrical power in the current electricity infrastructure towards
decentralization. Multi-agent technology and electronic markets form an appropriate technology that
can contribute to a solution to the resulting coordination problem. The PowerMatcher concept
proposed in this article is a market-based control concept for supply and demand matching (SDM) in
electricity networks with a high share of distributed generation.
The presented simulation case shows that this concept is capable of utilizing flexibility in device
operation via a distributed control mechanism. Due to device reactions on price fluctuations, the
simultaneousness between production and consumption of electricity in a sub-network is increased.

1257
As a result, the net import profile of the sub-network is smoothed and peak demand is reduced, which
is desired from a distribution network operational viewpoint. Two field experiments with the technology
show very encouraging results as to the actual implementation and the use of price signals for smooth
concerted operation of the devices due to the market equilibration mechanism. Proper socio-
economic considerations are to be dealt with when constructing utility functions to be used in the
agents and to assure acceptance and minimal lifestyle changes.

References
[1] A large part of the results described here were obtained in the CRISP-project (see
http://crisp.ecn.nl) I. G. Kamphuis, P. Carlsson, C.J. Warmer, J.C.P. Kester and J.K.
Kok,"Distributed Intelligence for Supply/Demand Matching to Improve Embedding of
Distributed Renewable Energy Sources", Proceedings of CRIS 2004 -- 2nd International
Conference on Critical Infrastructures}, Grenoble, October 2004.
[2] I.G. Kamphuis, M. Hommelberg, C.J. Warmer, F.J. Kuijper and J.K. Kok, ``Software agents
for matching of Power Supply and Demand; A fieldtest with a real-time, automated imbalance
reduction system", Proceedings of the International Conference on Future Power Systems
2005, Amsterdam. November, 2005.
[3] ENIRDGnet. "Concepts and Opportunities of Distributed Generation: The Driving European
Forces and Trends", ENIRDGnet project deliverable D3, 2003.
[4] IEA2002, International Energy Agency IEA, "Distributed Generation in Liberalised Electricity
Markets", International Energy Agency, Paris, 2002.
[5] European Commission, Directorate-General for Research (Brussels), "New Era for Electricity
in Europe -- Distributed Generation: Key Issues, Challenges and Proposed Solutions",
European Communities, Office for Official Publications, Luxembourg, 2003.
[6] R. Bitch, "Decentralised Energy Supply Options in Industrialised & Developing Countries",
World Power 2000, Siemens AG, pp. 61-65, 2000.
[7] D. Cohen, "Using Real-Time Web Technology to Manage DE Networks ", Distributed Power
2001, Intertech, Nice, France, 2001.
[8] J.M. Akkermans, J.F. Schreinemakers, J.K. Kok, “Emergence of Control in a Large-Scale
Society of Economic Physical Agents”, Proceedings of the AAMAS’04 Conference, 2004..
[9] Fredrik Ygge and Hans Akkermans, “Power Load Management as a Computational Market”,
Proceedings of ICMAS, 1996.
[10] F. Ygge and J.M. Akkermans, “Resource-Oriented Multi-Commodity Market Algorithms”,
Autonomous Agents and Multi-Agent Systems, Vol 3, pages 53-72, 2000. (Special Issue Best
Papers of ICMAS-98).
[11] P. Carlsson, “Algorithms for Electronic Power Markets”, Ph.D. Thesis, Uppsala University,
Sweden, 2004

1258
Air Conditioning

1259
1260
Definition of Functionalities of Air Conditioners for Better Public
Policies
Philippe Riviere, Jérôme Adnot

Ecole des Mines de Paris

Abstract
To evaluate innovations and potential regulations as regards air-conditioning, it is necessary to know
the exact function to which the system responds, which is not exactly an easy thing in the case of
individual air conditioning. The categories are not perceived by users as they are technically.
Ventilation is mixed with air conditioning, moisture is always present, and Internal Air Quality is a
rising issue. Envelope and equipment interactions are perceived by the user according to many
factors. Finally, last but not least, comfort is still an ambiguous concept. Any public policy has to
understand and respect first the functionalities demanded by the user for equipment and space, and
establish the link between those expectations and Energy Efficiency objectives. Premature judgment
will lead public policies to problems.

Underlying decisions not perceived by households

The citizen of developed country lives more and more in artificial climates which are determined by
market offers, regulations and policies in a very complex way. He (she) tries to perceive its
environment, namely when there is discomfort, but ignores most of the determinants. However he
(she) has a capacity to judge comfort (defined as the absence of any cause of discomfort). Figure 1
shows the system we are analyzing in this paper. The individual suffers from discomfort as a result of
climate through the envelope (which may be altered by building codes). An installer proposes and
installs equipment or the individual buys directly, maybe in an extreme situation. The utilities and
national agencies interfere with equipment. Early in the process, municipal authorities can interfere
with the structural choice of a more or less centralization of system. We are not developing here the
significant issues existing at city scale, in town planning or utility optimization but knowledge exists [1].

Building Codes climate

local emissions
loads Interior conditions + heat
emissions
control
Utility bill RAC
European
Policy

demand
National Agencies
merchant
manufacturer

influence, support installer

determination of national policies

Figure1: The limited perception of air conditioning by the final user

When we speak of “air conditioners” we exclude the central systems (CAC defined hereafter).
Central Air Conditioning systems (CAC) are characterized by a central refrigerating unit
operating together with an air treatment unit and make use of a fluid (air and/or water) to transport
cold to the air conditioned space. They perform other functions than just refrigerating, like controlling
air change, air quality and humidity. Their specifications are determined by engineers or technicians,

1261
who usually design the system and its associated energy performance without any direct influence
from the final customer or user (except for the preliminary limitations on cost).
A ‘Room Air-Conditioner’ (RAC), as opposed to an ‘air-conditioning system’ (CAC), is an
individual appliance that can be bought by a household or a professional, with a direct link between
the customer and the selection of the purchased good – either direct purchase by the household or
through an installer with whom negotiation and specification of the appliance takes place. The existing
results of EERAC [2] and a review of market features are used here as a basis for further definition of
functionality problems.

Technical wording for RAC

Let us first mention systems types that are close to RAC but that the consumer will be able to
separate, because he (she) can separate moisture and air treatment. Dehumidifiers have the same
technical nature than RAC but are not designed to permit temperature control. They seem to have
large market shares, especially in the UK. All RAC perform some dehumidification, in a controlled or
uncontrolled way. Evaporative cooling is an air conditioning process where the evaporation of water
is used to decrease the dry bulb temperature of the air. Some small appliances are manufactured for
the domestic sector (particularly in Greece). Finally, in desiccant cooling systems, before being re-
cooled by evaporative cooling (or a cooling coil), the air is dried. This is only used in industrial
applications.
There are four main classes of individual air conditioners (RAC) that are recognised by the user as
“air conditioner”:

Split units (figure 2). This type of equipment is made of two packaged units (an inside and outside
unit) connected by the refrigerant piping. The inside unit includes the evaporator and a fan, while the
outside unit houses the compressor and the condenser. They can be connected by a fixed copper
connection (made by an installer) or a flexible connection and are then said mobile split.

Figure 2. Diagrammatic representation of a split-

packaged unit.

Multi-split units (figure 3).Multiple inside units are connected to a single outside unit. This is the
rising segment in RAC sales.

Figure 3. Diagrammatic representation of a multi-split-

packaged unit.

Single packaged unit (Window units, like in figure 4). This type of equipment is made of a single
packaged unit, one side of which (the condensing loop) is in contact with outside air to stimulate
condensation while the other side (the evaporator loop) provides direct cooling to the inside air by

1262
forcing air over the evaporator with a fan. The two sides of the appliance are separated by a divider
wall, which is insulated to reduce heat transfer between the sides.

Figure 4. Diagrammatic representation of a single-

packaged unit.

Single-and Dual duct air conditioners. These appliances are fully indoors and reject hot air at the
condenser to the outside via a duct that is passed outdoors. They are generally movable but in order
to operate they must be set close to a window or a door through which a duct eliminates hot air. In
principle, a purpose built hole should be made in the building envelope for the ducting however, in
practice the ducting is often hung through doors and windows, which increases the infiltration of hot
air. This leads to additional thermal losses, and makes it difficult to asses the true thermal
performance of single duct units. The difference between single duct and dual duct units is very large
from the energy point of view because single duct units take air from the space, which in turn will
demand air to other zones or outside, bringing heat (or cold) to the space. Dual duct units take air
from the outside (the same air that will be rejected) avoiding this phenomenon.

Figure 5. Diagrammatic representation of a

single-duct or dual-duct unit.

Should we compare all RAC with the same grading scale in terms of energy
performance?
Labeling categories of the EU Energy Label for air conditioners include the so called “single duct” and
“double duct” units which are recent products for which the functional attributes are not perfectly
defined, and so should be reformulated. In some way the three basic types of air conditioners (split,
packaged, ducted) correspond to geometrical constraints of the users, and that’s why they were fully
respected in the EU regulation (labeling directive) and looser demands were made on less efficient
categories, in contradiction with EU energy efficiency objectives.
Constraints do exist that decide on the type of RAC: no access to roofs or facades for outside
units of split systems, no right to open holes for packaged (door/window) units. The single or double
duct air conditioners is almost the only appliance that may be installed everywhere. The installation
costs become very high for Multi Splits when the user doesn’t accept individual Splits room by room
and wants to hide the external units leading to longer piping, besides internal connections. One can
argue that when you have a hole for a “window unit” you cannot purchase anything else than a
“window unit”, and that performance of window units is to be judged relative to its geometrical
constraints. In a similar way, in city centers the outside units should be prohibited for aesthetic
reasons, so that single and dual duct systems should be favored in labeling categories. The extent of
those constraints and the cost it would represent to introduce a performance level difficult to reach for
one category within its geometrical constraint is an obvious issue. Present RAC regulation considers
those constraints are absolute and that lower demands should be made on less efficient RAC types.

1263
The reversible attribute of some air conditioners cannot be forgotten here, since it’s a rising heating
mode in some countries. It’s a big functional issue. In Japan, the COP considered is already very
often an average on cooling and heating, a point not reached in Europe.

Customer perceptions of purchase and maintenance

About the time of purchase, the questions are : what perception and understanding of the product the
customers has? How far is the product specified when the consumer comes to a professional? Is
really geometry the main constraint of choice? On what could we play to obtain a more reasonable
functional definition?
The usual approach assumes that the choice of the user’s decision is rational (based on number of
kW needed). This is unlikely for individual purchase namely some Single Ducts and for directly
purchased units, as for appliances sold by installers. Over sizing is the general rule.
In the case of air conditioners the purchaser is not rational about the level of residential comfort given
by each type of equipment : the “functional unit” in some cases is just cooling, in other products it
includes a contribution to Interior Air Quality (air filtration, controlled air change). New products will
try to incorporate moisture control, microbiological treatment, perfume, etc. Market will pull customers
in that domain of artificial product features. In a similar direction (what are the new functions, what is
their value?) the variable speed technology provided by “inverters” represent a secondary functional
benefit (overcapacity in order to make an unoccupied room rapidly comfortable?
The other way round an estimate of the lack of amenity due to the lack of a given comfort feature
could be given : noise (namely for “window units”), dry air (related with some diseases, so called “sick
building syndrome”), under sizing allowing temporary temperature increase. Even temperature
obtained by cooling is an amenity that should not necessarily be kept constant, provided that the
economic value of the discomfort (the loss of productivity, e.g.) is lower than the expenses needed to
oversize or overrun the air conditioner.
As very often, maintenance is poorly perceived by the final user. For an individual Air Conditioning
unit, changing the filter and cleaning the ducts are not expensive or long lasting technically but the
consumer is afraid of doing it himself. Usually it requires the visit of a technician with a minimum
charge. There is the possibility of a “contract” paid in advance but that is very expensive. A reported
value = 200 Euros (two visits/year: service defined for reversible use). After some time without
maintenance, energy consumption increases, as much as 20% due to fouling.
Where is the economic trade off between maintenance cost and over consumption? If the European
owner of a RAC is rational, this trade off corresponds to the 20% consumption increase because this
costs ten times less than the usual charge for maintenance. This means it's not economically optimal
to maintain RAC in the present market situation. So that lack of maintenance means a penalty
corresponding to 20% of cost of energy consumption is likely on the field.

The customer is changing: between price elasticity and awareness of global

warming
We are experiencing the effect of price elasticities, not only on total market (already done : it
became an affordable appliance) but between categories. We consider that the complete change of
price positioning of the “single ducts” invalidates the assumption of a limited and temporary market for
“single ducts”. This type of air conditioner was initially positioned as an expensive object (expensive in
Euros/kW not in Euros) for consumers under pressure buying under impulse (baby crying during a
heat wave). Now we see the massive Chinese (namely under foreign license) imports positioned as
first price in DIY shops (about one third of the price of previous equipment).
At the time of use we know that the consumer behavior is psychological, some people having an
“adaptative” behavior within some limits (comfort is a difference between outside and inside). In the
same conditions other people leave the equipment in full operation all the time (a recent survey in the
UK shows operating time as high as 5000 hours [3] for a basic need of –say- 500 hours). There is
something to expect from the new EU inspection if it is followed by efficient behavioral decisions, and
maybe if some related features are included in the products, like meters easy to read. There is a
cultural aspect, because “traditional” Europeans did know the thermal value of clothing, not only its
aesthetical aspect, and adapted clothing to climate. The necessity of cooling in cars (due basically to
noise insulation) and the fact of being used to air conditioning in (almost unregulated in terms of
comfort and thermal capacity) work places, are rising the demand in residences now.
At the time of disposal, it is essential that there is a good recovery of refrigerating fluid, and that
recycling circuits operate. Another issue at that time is to promote the purchase of more efficient

1264
equipments at the time the consumer is ready to change, once again a point where EU inspection
could be a tool.
There are two significant interactions with global policies of which the citizens and elected people
have to be convinced. One is an interaction with the urban policy : air conditioning sales are
increased by the heat island and heat wave effects (see the book “Cooling the city” [2]); air
conditioners spread is unstabilising the electrical grid globally in Spain and Italy and generate strong
regional problems in Greece, Portugal (Algarve black out) and France (PACA region reliability
problems). The regional development is not uniform, and the new structure of the electricity market
scatters responsibilities at peak time, two factors that could be modeled but are presently poorly
known.

Envelope and equipment regulations–banning air conditoning or improving it?

In a large number of climatic zones across the European Union there is a high proportion of buildings
with a clear need for artificial air conditioning, either due to climatic conditions or from a combination
of thermally inefficient building design and/or high internal loads. In such situations greater energy
efficiency is obtained not by preventing the use of A/C equipment but by deploying improved A/C
equipment, better system design and control and through regular maintenance. Accordingly, some
national building codes in the EU aim to raise A/C efficiency through requiring more efficient A/C
systems. Conversely, some building codes try to avoid active A/C energy demand altogether through
the use of bioclimatic architectural practices instead of through optimizing A/C system performance
and this is also sometimes presented as an approach towards raising summer energy efficiency. Yet
others aim at a combined strategy of improving architectural design and raising A/C efficiency.
In fact, the limits of when the use of active A/C is desirable and when it is not, are poorly known. Such
limits can be understood in terms of the physical limits of human comfort, e.g. “Is natural ventilation
enough in a UK office building?", as well as in terms of economic limits e.g. "Active A/C would
improve the comfort conditions in a UK office building but the cost difference is such that it is not
balanced by the gains in productivity it would bring". The two types of limits should be carefully
studied. We are basically facing a regulation problem between two types of solutions. To set up the
rules for an optimal building code, we should internalize discomfort costs, it is to say compare the total
costs of the two solutions:
- With A/C = Initial cost (of building envelope & equipment) + operating cost
- Without A/C = Building envelope initial cost + any discomfort cost

As a result of the current lack of common understanding, there are widely divergent practices and
rationales in European building codes. For example one national regulation requires artificial A/C to
be associated with any ventilation and another excludes it in principle. In the absence of an objective
set of planning criteria, ideological views will often hold sway e.g. with some solar architects struggling
for natural comfort while others will make purely aesthetic decisions, due to which A/C cost will
increase.
The present contradiction will become embarrassing when efforts are made to harmonize the
treatment of energy-efficiency among building codes. To outline the real limits between an effort on
A/C efficiency and the domain of bioclimatic architecture is an urgent task because the application of
the directive « Energy Efficiency in Buildings » opens a window for integrated measures on the
building envelope and equipment. Such an integrated study is made possible by the progress of LCC
analysis which apart from quantifying energy costs can also give a cost to discomfort, much in the
same way as different levels of system reliability are valued in the utility business.

Comfort levels: a rising demand

The comfort level to be reached is thus an image of the quality of the activity which takes place in
the conditioned space. To make rough distinctions, there is an increasing level of comfort when we
move from Free Cooling, with night time over-ventilation, to Partial Cooling, separated from heating
and very suitable for residential activity or non working spaces, to Total Cooling (controlled
temperature) with some ventilation and an uncontrolled change in humidity level, suitable for simple
work spaces, with or without heating, to Total Air Conditioning where temperature and humidity are
controlled and to Advanced Air Conditioning (humidity, indoor air quality and temperature are
controlled) adequate for clean rooms or other buildings with specific technical functions.
To change the level of comfort changes the consumption of energy. To make the comparison of
options possible, it is necessary to give a monetary value to discomfort, or to estimate the cost of

1265
the final equipment, which would give equal comfort. Often the presence of room-by-room air
conditioners is a sign that comfort was incorrectly valued at the time of the building’s design, thereby
obliging the customer to implement local remedies. An optimal building code may avoid this for new
buildings, while growing comfort expectations will still develop the room air conditioner market for
existing buildings. One could develop the following approach of RAC: at least when a RAC is installed
we are certain that AC has a real value and will strongly increase comfort.
The initial and running costs between the extremes may display a variation by a factor 10, but this
type of choice (who/what deserves a certain level of comfort?) is outside the scope of a regulation.
Most comparisons are to be done internally to a comfort level even if we get an idea of the relative
cost of the various possible comfort levels. Due to the climatic conditions and to social habits, each
country displays at a given time a given level of comfort demand, that we will accept as a data
(France or Greece is more a country of Total Cooling than Spain and Italy who have chosen a Total
Air Conditioning trajectory, UK accepts largely Free Cooling and Partial Cooling). The most difficult is
to predict the future level of expectations, and even more the speed of change.
Changes may be quick. So one sees for example that an increase of 1 % of GNP in southern cities
of the European Union leads to a 2.2 % increase of electricity use for air conditioning through
individual room air conditioners available from supermarkets.
The classic concept of thermal comfort is to define temperature and humidity conditions which are
acceptable by most occupants. Comfort depends in fact on outside weather conditions: in the
residential sector for example, it is well-known that an environment with 18°C or 19°C will be felt
comfortable in winter, but can appear cold if the outside temperature is higher than 35°C. The new
standards of “adaptive comfort” [4] relate acceptable inside conditions to outside conditions, but no
RAC behaves like this!

Some conclusions
What appears first is the difference in functionality between equipment. The final table indicates the
coverage of each function by each technique (“YES”)and potential of each technique to develop new
functions (“Target”).

F1 F2 F3 F4 F5 F6 F7 F8 F9 F10

No
Equipment outside
Temperat Ventilatio Adaptativ Adaptativ unit, no
type Dehumidi Humidific Air Reversibl Movable?
ure n e comfort e cooling geometri
fication ation change e heating
control internally target strategy c Mobility?
constrain
t

Window Uncontroll
YES YES YES
(packaged) ed

Controlled
or
Split YES YES Target! Target! YES
uncontroll YES
ed

Controlled
Single or or Target for
poor YES YES Target! Target! YES
double duct uncontroll DD YES
ed

Dehumidifie
Controlled YES
r

Evaporative Poor
poor YES
cooler control

The split and ducted systems seem to have specific advantages that will keep them on the market.
However the controls remain poor. We are far from having “adaptive” equipment that would take into
account outside conditions to build adequate inside conditions, 1- building on the outside temperature
to define an adapted internal demand, 2- playing with controls on moisture and/or air velocity to

1266
generate conditions within an adapted comfort zone like the one shown on a psichrometric chart like
the ASHRAE one.

Figure 6. Diagrammatic representation of an

adapted control strategy aiming at an adapted
comfort target

Figure 6 shows the conceptual way forward : a strategy defining an adaptated comfort target
corresponding to the user and situation, and adaptated control to bring outside conditions within the
comfort range, both taking into account more variables than today. This is likely to be a path of
development for dual ducts and splits, the system types with a potential for energy efficiency and the
highest acceptance on the other functions.

References
[1] M.Santamouris, J. Adnot et al., 2004, book “Cooling the cities” resulting from the Urbacool
study for the Directorate General Transportation-Energy of the Commission of the European
Union , “Presses de l’ecole”, released through Eyrolles publisher
[2] J. Adnot et al., 1999, Energy Efficiency of Room Air-Conditioners (EERAC), report for the
Directorate General Transportation-Energy of the Commission of the European Union, May
1999, cosigned by Eurovent-Certification and CECED
[3] Knight, 2006, Measured Building And AC System Energy Performance: An Empirical
Evaluation Of The Energy Performance Of Air Conditioned Office Buildings In The UK. Gavin
Dunn, Association of Building Engineers; Clarice Bleil de Souza, Andrew Marsh, Ian Knight,
Welsh School of Architecture, ICEEB’06, Frankfurt (26th/27th April 2006)
[4] CEN, 2006, Draft standard prEN 15251, Criteria for the indoor environment including themal,
indoor air quality, light and noise, 2006

1267
1268
Energy Labelling Directive, 2002/96/EC and EN 14511 Standard for
Room Air Conditioners

Yamina Saheb, André Pierrot, Sulé Becirspahic

Eurovent

Abstract
To improve energy efficiency of room air conditioners, the European Commission decided to apply
energy labelling policy through 2002/96/EC Directive. The directive refers to the standard EN 14511
for testing procedures, uncertainty of measurement and tolerances to be used for energy efficiency of
room air conditioners. However, the standard in its annex A, § 4 stipulates that the allowed tolerances
between the claimed and measured efficiency is 15% without taking into account that the difference
between two successive classes is between 6 and 8 % which allows manufacturers to jump easily to
one or even two higher classes.

Introduction
The implementation of Kyoto Protocol has a high priority for the European Union and strong
measures shall be applied to achieve its target fixed for Europe as a reduction of 8% of equivalent
CO2 emission between 2008 and 2012. To meet this important challenge, HVAC industry decided to
be more proactive and to take relevant actions in advance before some mandatory measures are
being decided by the European Commission or National Authorities.
Eurovent air conditioners manufacturers were involved in the SAVE project EERAC concerning
energy efficiency of room air conditioners. This project was finalised at the end of 1999. Using the
data provided by manufacturers and available on Eurovent directories, it has been realised that
energy efficiency of air conditioners presented on the market varies widely: the best units have
sometimes two times higher efficiency than the worst.

Two possible policies have been analysed in detail: energy labelling and minimum efficiency.

Mandatory minimum efficiency has been introduced in many countries outside of Europe. This is the
simplest way to eliminate low efficiency products from the market but it has a very important effect on
the industry and must be applied carefully. In order to avoid such regulation, Eurovent air conditioners
manufacturers prepared a proposal on voluntary minimum efficiency to be supported by Eurovent
certification. Implementation of this proposal started in January 2004 with the elimination of Class G
products.

Energy Labelling as applied with success to many home appliances like domestic refrigerators was an
obvious measure. However the labelling itself means only clear information to buyers – it is expected
that buyers will prefer better equipment and that in this way the global, average efficiency of the
products sold on the market will increase. Labelling Directive for small Air Conditioners has been
prepared by the European Commission and was published in April 2002.

Full mandatory application of this Directive has been fixed for 30 June 2003. However the relevant
test standard needed to serve as the reference document was missing; new revised standard EN
14511 covering all products in the scope of the Directive has not been finalised before May 2004.
Following several discussions with officials from the European Commission, the Labelling Committee
decided to postpone the application till just before the summer 2004.
It was expected from the application of the labelling directive to increase energy efficiency of room air
conditioners. However, EN 14511 in its Annex A, stipulates that the allowed tolerances between
claimed and measured efficiency is 15%. Using this tolerance, each product can jump at least one
class. That means class G doesn’t exist. Cleaver manufacturers and they are so many will never
declare any product in class G or F because a cleaver consumers and they are so many, will always
prefer to buy products with class A, B or C. In addition to this non reasonable tolerance, there is no
obligation of checking by an independent body of the claimed efficiency and class.

1269
Effect of the certification on room air conditioners energy efficiency
During the last years the energy efficiency has become an important issue for all energy consuming
equipment. Eurovent manufacturers followed this trend becoming more active. As the first action, the
energy efficiency of air conditioners became a certified characteristic in 2001. This action added more
clarity and transparency to the users and anticipated the mandatory implementation of labelling
directive for room air conditioners.
However this measure has no direct effect on improving the energy efficiency of equipment. In fact
when certified data from the directory are compared with those of successive years, the global
average efficiency of products appears sometime to even decrease slightly. This apparent negative
effect is just the consequence of correction of wrong, exaggerated data submitted by some
manufacturers, as we can see in Figure1. Each time, when a new measure is applied, we first have a
decrease of efficiency than an increase.

4
EER

0
1994 1996 1998 2000 2002 2004 2006 2008
Year

EER Min EER Max EER Mean

Figure 1: Evolution of EER for room air conditioners

The strongest possible action is the implementation of minimum efficiency levels as it is done by
Eurovent certification. As certify all policy is applied, the lowest efficiency products under the minimum
st
efficiency level could not be certified and they automatically disappeared from the market. At 1
January 2004, the lowest class, Class G, of room air conditioners disappeared from Eurovent
directory. This was the first concrete action implemented in the certification before any mandatory
minimum efficiency introduction.
The consequence of elimination of Class G appeared immediately, the average energy efficiency
increased by 4% from 2003 to 2004 when the increase from 2001 to 2003 was only 1%. The minimum
efficiency is 2,2. Manufacturers are presenting units with higher efficiency and the average efficiency
of all air conditioners is now increasing.
The increase of energy efficiency of air conditioning is technically possible. The problem is not
technical but economical. Manufacturers know how to build more efficient products but as costs are
higher and there is no pressure from the market, lower efficiency products will still be found. Although
a great number of users will be attracted by better products, there will always be buyers looking for
lowest prices.

Effect of EN 14511 on decreasing energy efficiency of room air conditioners

With the Labelling Directive 2002/96/EC and the EN 14511 standard it is expected to improve energy
efficiency of room air conditioners as it has been done for domestic refrigerators. When products are

1270
independently checked and verified by a certification system using some reasonable realistic
tolerance, performance claimed by a manufacturer is closely related to an energy class. However with
15% tolerances for EER and COP, non certified products may jump one, two or in some cases even
three classes still remaining legally in conformity with the standard and therefore with the Directive.
The purpose of the Directive may be completely deformed.
The following tables(1-8) to illustrate this situation for each type of product. As it can be seen for some
types of air conditioners only three classes may exist instead of seven.

Table 1: Air cooled split and multi-split units - cooling mode only
Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,20 - - 3,20 3,68 A - - -
B 3,20 3,68 A 3,00 3,45 A 3,10 3,57 A
C 3,00 3,45 A 2,80 3,22 A 2,90 3,34 A
D 2,80 3,22 A 2,60 2,99 C 2,70 3,11 A
E 2,60 2,99 C 2,40 2,76 D 2,50 2,88 C
F 2,40 2,76 D 2,20 2,53 E 2,30 2,65 D
G 2,20 2,53 E <2,20 - - - - -

Table 2: Air cooled packaged units - cooling mode only

Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,00 - - 3,00 3,45 A - - -
B 3,00 3,45 A 2,80 3,22 A 2,90 3,34 A
C 2,80 3,22 A 2,60 2,99 B 2,70 3,11 A
D 2,60 2,99 B 2,40 2,76 C 2,50 2,88 B
E 2,40 2,76 C 2,20 2,53 D 2,30 2,65 C
F 2,20 2,53 D 2,00 2,30 E 2,10 2,42 D
G 2,00 2,30 E <2,00 - - - - -

Table 3: Water cooled split and multi-split units - cooling mode only
Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,50 - - 3,50 4,03 A - - -
B 3,50 4,03 A 3,30 3,80 A 3,40 3,91 A
C 3,30 3,80 A 3,10 3,57 A 3,20 3,68 A
D 3,10 3,57 A 2,80 3,22 C 2,95 3,39 B
E 2,80 3,22 C 2,50 2,88 D 2,65 3,05 D
F 2,50 2,88 D 2,20 2,53 E 2,35 2,70 E
G 2,20 2,53 E <2,20 - - - - -

Table 4: Water cooled packaged units - cooling mode only

Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >4,40 - - 4,40 5,06 A - - -
B 4,40 5,06 A 4,10 4,72 A 4,25 4,89 A
C 4,10 4,72 A 3,80 4,37 B 3,95 4,54 A
D 3,80 4,37 B 3,50 4,03 C 3,65 4,20 B
E 3,50 4,03 C 3,20 3,68 D 3,35 3,85 C
F 3,20 3,68 D 2,90 3,34 E 3,05 3,51 D
G 2,90 3,34 E <2,90 - - - - -

1271
Table 5: Air cooled split and multi-split units - reverse cycle
Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,60 - - 3,60 4,14 A - - -
B 3,60 4,14 A 3,30 3,80 A 3,45 3,97 A
C 3,30 3,80 A 3,10 3,57 B 3,20 3,68 A
D 3,10 3,57 B 2,80 3,22 C 2,95 3,39 B
E 2,80 3,22 C 2,60 2,99 D 2,70 3,11 C
F 2,60 2,99 D 2,40 2,76 E 2,50 2,88 D
G 2,40 2,76 E <2,40 - - - - -

Table 6: Air cooled packaged units - reverse cycle

Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,40 - - 3,40 3,91 A - - -
B 3,40 3,91 A 3,20 3,68 A 3,30 3,80 A
C 3,20 3,68 A 3,00 3,45 A 3,10 3,57 A
D 3,00 3,45 A 2,60 2,99 D 2,80 3,22 B
E 2,60 2,99 D 2,40 2,76 D 2,50 2,88 D
F 2,40 2,76 D 2,20 2,53 E 2,30 2,65 D
G 2,20 2,53 E <2,20 - - - - -

Table 7: Water cooled split and multi-split units - reverse cycle

Original Upper Final Lower Average Final
Class limit 15% class limit 15% Final class Value 15% classe
A >3,60 - - 3,60 4,14 A - - -
B 3,60 4,14 A 3,40 3,91 A 3,50 4,03 A
C 3,40 3,91 A 3,20 3,68 A 3,30 3,80 A
D 3,20 3,68 A 2,80 3,22 C 3,00 3,45 B
E 2,80 3,22 C 2,60 2,99 D 2,70 3,11 D
F 2,60 2,99 D 2,40 2,76 E 2,50 2,88 D
G 2,40 2,76 E <2,40 - - - - -

Table 8: Water cooled packaged units - reverse cycle

Original Upper Final Lower Average
Class limit 15% class limit 15% Final class Value 15% Final class
A >3,40 - - 3,40 3,91 A - - -
B 3,40 3,91 A 3,20 3,68 A 3,30 3,80 A
C 3,20 3,68 A 3,00 3,45 A 3,10 3,57 A
D 3,00 3,45 A 2,60 2,99 D 2,80 3,22 B
E 2,60 2,99 D 2,40 2,76 D 2,50 2,88 D
F 2,40 2,76 D 2,20 2,53 E 2,30 2,65 D
G 2,20 2,53 E <2,20 - - - - -

Comparison test between three laboratories for three air conditioners

To evaluate the actual uncertainties of measurement using the European standard EN 14511:2004
and to demonstrate that the tolerances specified in Annex A of the standard are not reasonable,
Eurovent organized a round robin test to compare the results obtained by different European
independent laboratories for the measurement of the cooling and heating capacities of air
conditioners and heat pumps, and of their EER and COP.
Three laboratories equipped with calorimetric room have been selected. Testing facility used in each
laboratory was a balanced ambient room type calorimeter.
In the summary below, the laboratories are numbered 1 to 3. Laboratory nº 1 being the reference
laboratory for the purpose of the survey

1272
 ƒ Standard and test method to be used:
The tests have been performed in the same test conditions than for EUROVENT Certification testing,
using the test procedures described in EN 14511-3:2004.
The test method was the calorimeter room method, including for the ducted sample, as specified in
EN 14511-2:2004, Annex A.
To avoid any risk of difference in the installation of the samples in different laboratories, the maximum
length of 7.5 m was used for the purpose of this round robin test.
The samples were delivered to each laboratory without refrigerant charge, and the samples charged
with the quantity of refrigerant given for each sample previously to each test.

ƒ Test conditions:
The test conditions were the standard rating conditions defined in EN 14511-2:2004 and reported in
Table 9

Table 9: Test conditions

Outdoor heat exchanger Indoor heat exchanger
Inlet dry bulb Inlet wet bulb Inlet dry bulb Inlet wet bulb
temperature temperature temperature temperature
°C °C °C °C
cooling mode 35 not required 27 19

heating mode 7 6 20 15 max

Three samples have been selected. The main characteristics of the samples are described in the
following Table 10

Table 10: Samples tested

Claimed
Sample nº Type of unit Mode
capacity
Split non ducted wall
1 Cooling 3,5 kW
mounted air conditioner

Split non ducted wall Cooling 5,0 kW

2
mounted heat pump
Heating 6,2 kW

Split ducted air

3 Cooling 6,6 kW
conditioner

2- Results obtained in each laboratory

SAMPLE 1, COOLING MODE

Laboratory nº 1 2 3
Dry bulb temperature, air inlet, outdoor unit (°C) 35,07 34,99 35,0
Wet bulb temperature, air inlet, outdoor unit (°C) 20,75 18,67 -
Dry bulb temperature, air inlet, indoor unit (°C) 27,13 27,00 27,0
Wet bulb temperature, air inlet, indoor unit (°C) 19,03 18,99 19,0
Atmospheric pressure (kPa) 95,05 97,34 99,6
Voltage (V) 230,7 229,9 230
Total cooling capacity, indoor side (W) 3408 3293 3490
Total cooling capacity, outdoor side (verification) (W) 3334 3306 -
Dehumidifying capacity (W) 977 948 897
Sensible cooling capacity (W) 2431 2345 2593
Sensible heat ratio 0,713 0,71 0,743
Effective power input (W) 1223 1210 1244
EER 2,79 2,72 2,81

1273
 SAMPLE 2, COOLING MODE
Laboratory nº 1 2 3
Dry bulb temperature, air inlet, outdoor unit (°C) 34,96 34,96 35,0
Wet bulb temperature, air inlet, outdoor unit (°C) 22,27 20,02 -
Dry bulb temperature, air inlet, indoor unit (°C) 27,10 27,00 27,0
Wet bulb temperature, air inlet, indoor unit (°C) 19,03 18,99 19,0
Atmospheric pressure (kPa) 94,87 96,43 100,3
Voltage (V) 230,4 230,0 230
Total cooling capacity, indoor side (W) 4962 4927 5091
Total cooling capacity, outdoor side (verification) (W) 5016 5837 -
Dehumidifying capacity (W) 1128 1194 889
Sensible cooling capacity (W) 3834 3733 4202
Sensible heat ratio 0,773 0,76 0,825
Effective power input (W) 1593 1602 1611
EER 3,11 3,08 3,16

SAMPLE 3, COOLING MODE

Laboratory nº 1 2 3
Dry bulb temperature, air inlet, outdoor unit (°C) 35,06 35,03 34,8
Wet bulb temperature, air inlet, outdoor unit (°C) 21,21 18,99 -
Dry bulb temperature, air inlet, indoor unit (°C) 27,08 27,00 27,0
Wet bulb temperature, air inlet, indoor unit (°C) 19,04 18,99 19,0
Atmospheric pressure (kPa) 93,65 96,93 100,2
Voltage (V) 230,7 230,0 230
Total cooling capacity, measured indoor side (W) 6254 6239 6271
Total cooling capacity, outdoor side (verification) (W) 6422 6216 -
Dehumidifying capacity (W) 1790 1620 1587
Sensible cooling capacity (W) 4464 4619 4684
Sensible heat ratio 0,714 0,74 0,747
Available static pressure (Pa) 44 47 56
Total power input (W) 2489 2481 2527
Total cooling capacity, including corrections for the fans a b
6293 6281 6321
(final result) (W)
a b
Effective power input (W) 2450 2439 2477
EER 2,57 2,51 2,58 2,55
a
no correction for the fan has been done by this laboratory
b
estimated by the reference laboratory

SAMPLE 2, HEATING MODE

Laboratory nº 1 2 3
Dry bulb temperature, air inlet, outdoor unit (°C) 7,11 7,02 7,0
Wet bulb temperature, air inlet, outdoor unit (°C) 6,06 5,99 6,1
Dry bulb temperature, air inlet, indoor unit (°C) 20,13 19,99 20,0
Wet bulb temperature, air inlet, indoor unit (°C) 12,33 12,27 -
Atmospheric pressure (kPa) 94,20 97,85 99,2
Voltage (V) 230,6 229,8 230
c
Heating capacity, outdoor room (W) 6028 6196 -
c c
Heating capacity, indoor room (W) 6132 6188 6143
Effective power input (W) 1808 1867 1752
a b b
COP 3,33 3,31 3,51
a
calculated with the outdoor side measurement
b
calculated with the indoor side measurement
c
final result for the heating capacity

1274
 ƒ Analysis of the results
The aim of the study being the evaluation of the overall uncertainty of measurement of the capacities
and efficiencies, the analysis is focussed on these parameters.
In the following page, we show an abstract of the results for the capacities and efficiencies, together with
the deviation of each individual measurement with reference to the medium value of each parameter.
The main conclusion is the small differences between the results of the three laboratories.
The biggest deviation from the medium value for the capacities is 3,1 %, and 3,7 % for the efficiencies.
These deviations include not only the uncertainty of the measurement itself, but also the effect of the
differences in the installation of the samples.
The conclusion of the survey is that the uncertainty of the measurement of cooling and heating
capacities of an air conditioner or a heat pump by a recognised independent laboratory is fully within the
maximum uncertainty required by the testing standard EN 14511-3, which is 5 %, and that the
uncertainty of the measurement of the efficiencies is also inside this tolerance.

1275
 COOLING COOLING COOLING
MEDIUM STANDARD STANDARD DEVIATION DEVIATION DEVIATION
SAMPLE CAPACITY CAPACITY CAPACITY
VALUE DEVIATION DEVIATION % LAB Nº 1 (%) LAB Nº 2 (%) LAB Nº 3 (%)
LAB Nº 1 LAB Nº 2 LAB Nº 3

1 3408 3293 3490 3397 99,0 2,9 0,3 -3,1 2,7

2 4962 4927 5091 4993 86,4 1,7 -0,6 -1,3 2,0

3 6293 6281 6321 6298 20,5 0,3 -0,1 -0,3 0,4

EER LAB Nº EER LAB Nº EER LAB Nº MEDIUM STANDARD STANDARD DEVIATION DEVIATION DEVIATION
SAMPLE
1 2 3 VALUE DEVIATION DEVIATION % LAB Nº 1 (%) LAB Nº 2 (%) LAB Nº 3 (%)

1 2,79 2,72 2,81 2,77 0,05 1,7 0,6 -1,9 1,3

2 3,11 3,08 3,16 3,12 0,04 1,3 -0,2 -1,2 1,4
3 2,57 2,58 2,55 2,57 0,02 0,6 0,1 0,5 -0,6

HEATING HEATING HEATING

MEDIUM STANDARD STANDARD DEVIATION DEVIATION DEVIATION
SAMPLE CAPACITY CAPACITY CAPACITY
VALUE DEVIATION DEVIATION % LAB Nº 1 (%) LAB Nº 2 (%) LAB Nº 3 (%)
LAB Nº 1 LAB Nº 2 LAB Nº 3

2 6028 6188 6143 6120 82,5 1,3 -1,5 1,1 0,4

COP LAB Nº COP LAB Nº COP LAB Nº MEDIUM STANDARD STANDARD DEVIATION DEVIATION DEVIATION
SAMPLE
1 2 3 VALUE DEVIATION DEVIATION % LAB Nº 1 (%) LAB Nº 2 (%) LAB Nº 3 (%)

2 3,33 3,31 3,51 3,38 0,11 3,3 -1,6 -2,2 3,7

1276
Conclusion
Implementation of energy labelling policy for small air conditioners as a first step of a general energy
efficiency policy can be considered as a good starting point. However in order to achieve the purpose of
this action, the European Commission should ask urgently CEN to revise the tolerances in Annex A of
En14511 standard.
As shown by the survey above, the only reasonable tolerance is the one that will avoid to any product to
jump from one class to another one. 8% seems to be a good compromise
In addition, the declaration of energy class is a good initiative but not a sufficient one. Europe needs to
impose the checking of the characteristics claimed by the manufacturers by an independent body.
As Eurovent, we believe that without strong mandatory measures, some manufacturers will declare
whatever they want.
On the other hand in order to reach Kyoto Protocol objectives and avoid becoming dustbin for the
international industry, it is urgent to have a real energy policy for air conditioning systems. Europe has to
combine tEnergy Labelling policy with minimum energy efficiency and to set up a schedule for the
elimination of lowest efficient products.
Eurovent has been studying different possible schedules to eliminate low efficiency classes, for example
the products with capacity under 4 kW, which represents more than 80% of the residential market and
have a short life cycle, lowest efficiencies, E, F can be already eliminated in 2008.
By these measures, products with less efficiency will not sold in Europe and the average installed EER
in Europe will increase quickly.

1277
1278
China Cools with Tighter RAC Standards
Jiang Lin, Gregory Rosenquist

Lawrence Berkeley National Lab

Abstract
After boiling summer brought brown-out to most part of the country in 2004, China announced a new
set of minimum energy efficiency standards for room air conditioners in September 2004, with the first
tier going into effect on March 1, 2005 and the reach standard taking effect on January 1, 2009. This
represents a milestone in China’s standard setting process since the reach standard levels are
significantly more stringent than previous standards for other appliances. This paper first analyzes
cost-effectiveness of China’s new standards for room air conditioners, and then attempts to evaluate
the impact of the new standards on energy savings, electric generation capacity, and CO2 emissions
reductions.

Introduction

Since 2002, China has experienced wide spread power shortage, leading to the rationing of power in
24 out of 31 provinces across country in 2004. While most observers point to the strong economic
growth in China as the primary cause for such shortage, incremental air-conditioning load is also a
leading contributor. It is estimated that about 30 million new room air conditioners were installed in
2004 along, adding roughly 20 GW in peak capacity, which easily eclipse the generating capacity of
the Three Gorges Dam. In major cities along the eastern seashore such as Shanghai, air-
conditioning load accounts for 40% of the peak summer load.
It is therefore understandable that China decided to update its minimum energy efficiency standard
(MEPS) for room air conditioners. The revised standard was published in September, 2004, with the
first tier going into effect on March 1, 2005 and the reach standard taking effect on January 1, 2009.
This represents a milestone in China’s standard setting process since the reach standard levels are
significantly more stringent than previous standards for other appliances. This paper first reviews the
requirements of the new Chinese standards for air conditioners, and then analyzes its cost-
effectiveness, and finally attempts to evaluate the impact of the new standards on energy savings,
electric generation capacity, and CO2 emissions reductions.

China’s New Standard for Room Air Conditioners

China’s new standard for air conditioners sets two tiers of performance requirements, with the first tier
going into effect on March 1, 2005, and the second tier going into effect on January 1, 2009. The
2005 requirements are listed in Table 1.
The measured value of energy efficiency ratio (EER) of room air conditioners must be greater than or
equal to the values shown in Table 1.

Table 1: Energy Efficiency Ratios (EER)

Category Rated Cooling Capacity (CC) W EER W/W
Single-package - 2.30
CC ≤4500 2.60
4500<CC≤71 2.50
7100 < CC ≤ 14000 2.40

In addition to setting the minimum requirement, China’s new AC standard also include classification
requirements for the newly established Energy Information Label, as well as the certification
requirement for CECP’s Energy Conservation Label.

Rating requirements (measured in EER) for specific Energy Efficiency Grades are listed in Table 2.

1279
Table 2: Energy Efficiency Grade Specifications
Energy Efficiency Grade EER W/W
Category Rated Cooling Capacity (CC) W 5 4 3 2 1
Single-package - 2.30 2.50 2.70 2.90 3.10
Split CC ≤4500 2.60 2.80 3.00 3.20 3.40
4500<CC≤7100 2.50 2.70 2.90 3.10 3.30
7100 < CC ≤ 14000 2.40 2.60 2.80 3.00 3.20

For China’s voluntary energy endorsement label (managed by China Standard Certification Center),
the EER requirements of room air conditioners must be greater than or equal to the values shown in
Table 3.

Table 3: Energy Efficiency Specification

Category Rated Cooling Capacity (CC) W EER W/W
Single-package - 2.90
Split CC ≤4500 3.20
4500<CC≤7100 3.10
7100 < CC ≤ 14 3.00

Moreover, a tighter standard goes into effects on 1 January 2009. The requirements of the 2009
standard for room air conditioners are listed in Table 4.

Table 4: Energy Efficiency Ratios (EER) in 2009

Category Rated Cooling Capacity (CC) W EER W/W
Single-package 2.90
Split CC ≤4500 3.20
4500<CC≤7100 3.10
7100 < CC ≤ 14000 3.00

In order to compare the stringency of the Chinese standards, a collection of standards around the
world are presented below in Tables 5 through 9.

Table 5: U.S. and Canadian Air Conditioner Standards

Category EER W/W
Window 2.87
Central 3.37
*Converted from seasonal energy efficiency (SEER) of 13. SEER is in units of Btu/Whr.
1
Table 6: South Korean Air Conditioner Standards
Category Rated Cooling Capacity (CC) W EER W/W
Window - 2.88
Split CC<4000W 3.37
4.0 kW ≤ RCC < 10.0 kW 2.97
10.0 kW ≤ RCC < 17.5 kW 2.76

Table 7: Singapore Air Conditioner Standards

Category EER W/W
Window 2.73

Table 8: European Union Labeling Requirements

Energy Efficiency Grade EER W/W
Category G F E D C B A
Window <2 2.0 2.2 2.4 2.6 2.8 >3.0
Split <2.2 2.2 2.4 2.6 2.8 3.0 >3.2

1
As of March 2006.

1280
Table 9: Japanese Air Conditioner Standards
Category Rated Cooling Capacity W/W EER W/W
AC-only Heat pump
Window
- 2.67 2.85
AC-only Heat pump
CC<2500W 3.64 5.27
2500-3200W 3.64 4.90
Split
3200-4000W 3.08 3.65
4000-7100W 2.91 3.17
>7100W 2.81 3.10

Where the standards cover products in several categories, the most dominant product type in China
today is the split air conditioners with a cooling capacity smaller than 4500 Watts. Therefore it can be
seen that the 2009 Chinese requirement for this product group will be higher than the EU label A, and
only trails the requirements of Japan and South Korea. While the US standard is technically more
stringent, it is set for residential central system, which carries significant duct losses.
For window air conditioners, the 2009 Chinese standard would be the highest with a minimum EER of
2.9, compared to 2.87 in the US, 2.85 in Japan, and 2.88 in South Korea.

Life-Cycle Cost Analysis of Air Conditioner Standards

One of the factors used to judge the economic feasibility of air conditioner standards is consumer life-
cycle costs (LCC) savings. If a more efficient air conditioner, i.e., a unit that meets possible new
standards, provides consumer LCC savings relative to a minimally compliant product, the more
efficient unit is generally assessed as being economically feasible.
The LCC is the sum of the purchase cost (PC) and the present value of operating expenses (OC)
discounted over the lifetime (N) of the equipment. If operating expenses are constant over time, the
LCC simplifies to

LCC = PC + PWF • OC,

where the present worth factor (PWF) is dependent on the discount rate (r) and the equipment lifetime
(N) and is defined as

N
PWF = 1/r • [ 1 – 1/(1+r) ].

The LCC analysis described below was conducted with a 6 percent real discount rate and an
2
equipment lifetime of 12.5 years.

In order to assess the economic feasibility of increased Chinese air conditioner standards, an LCC
analysis was performed on two different sized split system room air-conditioning heat pumps: 3500
Watt (W) cooling capacity and 7100 Watt (W) cooling capacity (Rosenquist and Lin, 2005). The
above cooling capacity units were determined to be the most popular sized air-conditioning units in
China. Table 10 summarizes the physical characteristics of the two representative units (i.e., baseline
units) for which the LCC analysis was conducted. By defining baseline units, more efficient designs,
otherwise known as design options, can be evaluated to determine if standards are economically
feasible. Note that the rated EERs of the two baseline units have efficiencies roughly equal to the
current minimum EER standards (see Table 1). At the time the LCC analysis was conducted, many
air conditioner models in China had efficiencies that were roughly equal to the current minimum
standards.

2
The lifetime assumption is based on industry estimate of average lifetime of ACs in China. No benchmark discount rate is
available from China. However, interest rate for loans over 5 years was 6.12% by the end of 2005.

1281
Table 10: Baseline Characteristics of 3500 W and 7100 W Split System Room Air-Conditioning
Heat Pumps
Heat Pump Split (single
evaporator)
Rated Cooling Capacity, CC (W) 3500 7100
Rated EER (W/W) 2.57 2.55
Refrigerant R-22 R-22
Short Tube
Flow Control Device Short Tube (assumed)
(assumed)
2
Face area (m ) 0.206 0.303
Evaporator Fin type Hydrophilic Slit Fin Hydrophilic Slit Fin
Tube type Grooved Grooved
2
Face area (m ) 0.381 0.502
Condenser Fin type Hydrophilic Slit Fin Hydrophilic Slit Fin
Tube type Grooved Grooved
Manufacturer Panasonic Wanbao Copeland
Model Number 2K23S225BUA ZR34KH-PFJ-522
Type Rotary Scroll
Compressor
Cooling capacity (W) 4000 8200
3
Displacement (cm ) 21.4 46.1
Efficiency (W/W) 2.81 3.02
3
Air volume (m/h ) 463 1014
Evaporator-side Power Input (W) 50 80
Type PSC PSC
Fan 3
Air volume (m/h ) 1400 (assumed) 2040 (assumed)
Condenser-side Power Input (W) 75 (assumed) 150 (assumed)
Type PSC PSC

Tables 11 and 12 show the cost-efficiency and LCC results for the 3500 W and 7100 W baseline units.
Specific design options were added to the baseline units and their impact on manufacturer cost,
consumer retail price, cooling capacity, EER, annual energy consumption (AEC), operating cost (OC),
payback period (i.e., the ratio of the change in consumer retail price over the change in operating
cost), and LCC were determined. Note that in Tables 11 and 12, only a limited number of design
options were evaluated. For example, increases in evaporator size were not considered so as to
prevent any significant changes to the indoor cabinet. In addition relatively small increases in
condenser size were considered to limit the size of the outdoor cabinet.

Table 11: Cost-Efficiency and LCC Results for 3500 W Split System Room Air-Conditioning
Heat Pump
Manufacturer Cost
Retail Cooling Payback
Incr. Total EER AEC* OC** LCC
No. Design Option Price Capacity Period
Yuan Yuan W/W kWh/yr Yuan/yr Yuan
Yuan Watts Years
0 Baseline - - 2600 3299 2.57 1148 689 - 8541
1 0 +3.0 EER Compressor 40 40 2660 3307 2.75 1074 644 1.3 8214
2 1 +3.16 EER Compressor 61 101 2751 3313 2.90 1018 611 1.9 8015
2
3 2 + 0.419 m Condenser 56 157 2834 3325 2.96 997 598 2.6 7992
4 3 +Cond^ Fan Motor +10% 20 177 2864 3325 2.98 992 595 2.8 7995
5 4 +Evap^ Fan Motor +10% 20 197 2894 3329 2.99 987 592 3.0 7997
6 5 +Cond^ Fan Motor +20% 20 217 2923 3329 3.02 978 587 3.2 7982
7 6 +Evap^ Fan Motor +20% 20 237 2953 3333 3.03 973 584 3.4 7989
* AEC based on annual operating hours of 895.
** OC based on electricity price of 0.6 yuan/kWh.
^ Cond = Condenser; Evap = Evaporator

1282
Table 12: Cost-Efficiency and LCC Results for 7100 W Split System Room Air-Conditioning
Heat Pump
Manufacturer Cost
Retail Cooling Payback LCC
Incr. Total EER AEC* OC**
No. Design Option Price Capacity Period
Yuan Yuan W/W kWh/yr Yuan/yr Yuan
Yuan Watts Years
17,25
0 Baseline - - 5000 6814 2.57 2369 1421 -
4
0 + 3.25 EER 16,66
1 177 177 5264 6857 2.77 2203 1322 2.7
Compressor 1
2 16,55
2 1 + 0.551 m Condenser 70 247 5368 6884 2.82 2163 1298 3.0
7
2 +Cond^ Fan Motor 16,54
3 30 277 5413 6884 2.83 2153 1292 3.2
+10% 8
3 +Evap^ Fan Motor 16,54
4 30 307 5458 6891 2.84 2144 1286 3.4
+10% 8
4 +Cond^ Fan Motor 16,54
5 30 337 5502 6891 2.86 2135 1281 3.6
+20% 8
5 +Evap^ Fan Motor 16,55
6 30 367 5547 6897 2.87 2128 1277 3.8
+20% 6
* AEC based on annual operating hours of 895.
** OC based on electricity price of 0.6 yuan/kWh.
^ Cond = Condenser; Evap = Evaporator

Figures 1 and 2 show the LCC as a function of EER for the 3500 W and 7100 W air-conditioning heat
pump units. Because a limited set of design options were evaluated, the maximum efficiency points
are below the new set of standards that are to become effective in China in 2009 (see Table 3). Thus,
the LCC analysis does not reveal whether the new set of standards, 3.2 EER for the 3500 W unit and
3.1 EER for the 7100 W unit, are economically feasible. Although the LCC analysis does not analyze
the 2009 standard levels, the maximum efficiencies, 3.0 EER for the 3500 W unit and 2.9 EER for the
7100 W unit, do yield LCC savings. Because the maximum efficiencies analyzed are close to the new
standard levels, it seems reasonable to expect that the new standard levels would be economically
feasible.

1283
 8600

8500
Life-Cycle Cost (Yuan)

8400

8300

8200

8100

8000

7900
2.5 2.6 2.7 2.8 2.9 3.0 3.1
EER (W/W)

Figure 1: LCC Results for 3500 W Split System Air-Conditioning Heat Pump

17300

17200

17100
Life-Cycle Cost (Yuan)

17000

16900

16800

16700

16600

16500
2.5 2.6 2.7 2.8 2.9 3.0
EER (W/W)

Figure 2: LCC Results for 7100 W Split System Air-Conditioning Unit

1284
Energy Savings from the New Standard
Given that China is one of the largest consumer markets for room air conditioners in the world, the
impacts of China’s new standards will be very large. According to a recent report (Lin, 2005), the
savings are likely to be 10 TWh of electricity and 2.7 million tons of carbon by 2010, 46 TWh of
electricity and 12.3 million tons of carbon by 2020 (Table 13). Three quarters of the expected savings
by 2020 stem from the more stringent reach standard to go in effect in 2009.

The peak demand reduction calculation depends on an accurate assessment of the coincident peak
factor for room air conditioners in China. The reductions are likely to be about 4.5 GW in 2010 and
20.4 GW by 2020. The 2020 peak demand reduction estimate exceeds the 17 GW capacity of the
Three Gorges Dam.

Table 13. Expected Energy Savings from the Room Air Conditioner Standard
Energy Carbon Peak
Stock Shipments Savings CO2 Savings Savings Savings*
Year millions millions million kWh million tonnes million tonnes GW
2000 67 17.3
2001 85 18.3
2002 106 21.6
2003 133 28.2
2004 161 29.6
2005 189 31.0 714 0.7 0.2 0.3
2006 218 32.6 1,464 1.4 0.4 0.6
2007 248 34.2 2,251 2.2 0.6 1.0
2008 278 35.9 3,077 3.0 0.8 1.4
2009 308 37.7 6,587 6.4 1.8 2.9
2010 337 38.5 10,166 9.9 2.7 4.5
2011 365 39.3 13,817 13.5 3.7 6.1
2012 390 40.0 17,541 17.1 4.7 7.7
2013 414 40.8 21,339 20.9 5.7 9.4
2014 435 41.7 25,188 24.6 6.7 11.1
2015 453 42.5 29,034 28.4 7.7 12.8
2016 469 43.3 32,837 32.1 8.8 14.5
2017 484 44.2 36,573 35.7 9.7 16.1
2018 497 45.1 40,142 39.2 10.7 17.7
2019 510 46.0 43,360 42.4 11.6 19.1
2020 522 46.9 46,113 45.1 12.3 20.4
Cumulative
Total 755 330.201 323 88 -

Conclusion
China’s recent efforts to regulate the efficiency of room air conditioners have been shown to yield
significant consumer economic savings as well as national benefits of avoided construction of large
number of power plants and avoided emissions of GHG and other local pollutants. Relative to air
conditioners that meet China’s previous set of air conditioner standards, air conditioner designs with
efficiencies that are just below the standards that are to become effective in 2009 yield consumer life-
cycle cost savings ranging from approximately 550 yuan for 3500 W cooling capacity units to 700
yuan for 7100 W cooling capacity units. The benefits of air conditioner standards are not limited to
consumers. Air conditioner standards (both the current standards and the 2009 standards) are
projected to yield cumulative national energy savings of over 330 billion kWh. The resulting reduction

1285
in national energy consumption is projected to lower power plant carbon emissions by 88 million
tonnes.

References

[1] Energy Conservation Center of Japan, Top Runner Program,

http://www.eccj.or.jp/top_runner/index_contents_e.html
[2] European Commission, 2002, Commission Directive 2002/31/EC, Official Journal of
European Commission, 3-4-2002.
[3] AQSIQ (China National Administration of Quality Supervision, Inspection, and Quarantine),
2004, The minimum allowable values of the energy efficiency and Energy efficiency grades
for room air conditioners, GB12021.3-2004, Beijing, China.
[4] Lin J. Project Completion Report on China’s Room AC Reach Standard. LBNL-57387. March,
2005. Berkeley, California.
[5] Rosenquist G and Lin J. Chinese Room Air-Conditioning Heat Pumps: An Engineering and
Life-Cycle Cost Analysis of 3500 W and 7100 W Cooling Capacity Units. LBNL-57992. 2005.
Berkeley, California.
[6] US DOE, Energy Conservation Program for Consumer Products: Central Air Conditioners and
Heat Pumps Energy Conservation Standards; Final Rule Finding of No Significant Impact
Energy Conservation Program for Consumer Products; Notice, 10 CFR Part 430,
http://www.eere.energy.gov/buildings/appliance_standards/residential/pdfs/central_ac_hp_fin
alrule.pdf

1286
An Experimental Comparison of Energy Efficiency Indicators, EER
and SEER in Residential Air-Conditioners
Ji Young JANG1, Se-Yoon OH1, Chan Ho SONG1, Ho Seon CHOI1 and Simon JIN2
1
Digital Appliance Laboratory, LG Electronics*
2
Digital Appliance Company Research Lab., LG Electronics**

Abstract
A new energy-saving technology was developed for residential air-conditioners, and energy efficiency
indicators, EER and SEER were compared between a conventional variable speed air-conditioner and
a new one including the new technology. The new air-conditioner, called MPS (Multi Power System)
consists of two constant speed compressors, an evaporator, a condenser, an accumulator and an
expansion device. The system is operated in a three different mode depending on a thermal load: 1)
full operation (two compressors work) 2) medium operation (lager one of the two comp. works) 3) low
operation (smaller one of the two comp. works). Thus, the cooling capacity was modulated in three
steps, and the energy was saved in the same way as that in the variable speed compressor driven air-
conditioner. That is, when the thermal load was low, the compressors did not start and stop
frequently, and start-up loss could be minimized. The energy consumption was reduced dramatically
when the thermal load was minimal, and it led to SEER (Btu/hr / W) improvement. Because of this
point, SEER 16.2 of the MPS was almost the same as SEER 16.8 of the inverter compressor driven
air-conditioner, although the MPS showed lower EER 3.32 (W/W) than EER 3.54 of the variable
compressor one. With MPS technology, it was shown that the energy saving could be achieved
without adopting expensive technologies.

1. Introduction
There have been tremendous needs in air-conditioner industries to develop products protecting
environment and achieving energy conservation. This has caused considerable attention toward
capacity modulation that could save energy more than conventional on-off systems. There are so
many capacity modulated systems nowadays and inverter type system has been in general use
although its higher cost and more complex drive logic. In this article, a simpler and more efficient
system, MPS, was introduced and compared with a inverter type air-conditioner with respect to their
energy efficiency indicators, EER and SEER, with an emphasis on the SEER. ARI standard 210/240[1]
was adopted in this article for SEER measurements. European SEER, ESEER has been mentioned in
some documents such as EECCAC 2003 final report[2] and BS standard 14825:2003. But the basic
concept of ESEER measurements is the same as that of SEER, and the ARI standard was followed.
ESEER comparison would be carried out in our next studies.

2. Advantages of Capacity Modulation

Variable capacity modulation has been one of the most effective methods to reduce energy
consumptions of air-conditioners. It is due to its several advantages. First, when the thermal load is
lower than the design load, the variable capacity modulation reduces the on-off cycling loss. The on-
off cycling loss is that when compressor stops, refrigerants in high-pressure side migrate to low-
pressure side, and energy loss occurs. When compressor starts, there is a possibility of liquid
compression, which is detrimental to compressors [3]. Also, energy consumes while building up high,
and low pressures in the air-conditioner. It takes about ten minutes to reach steady running state, and
the system runs at a lower efficient condition during this period. As shown in Figure 1, the constant
speed system repeatedly starts and stops, depending on the room temperature setting. But the
capacity modulation system minimizes on-off operation by the capacity modulating system.
Second, in lower cooling load condition, the capacity modulation system reduces the amount of
refrigerant pumped out of the compressor, and it leads that the condensing pressure is reduced and
that the evaporative pressure is increased. Thus, the system efficiency increases, in the capacity
modulation system.
The benefit of the capacity modulation system could not be represented with the index, EER, even
though the system offers a lot better benefit and saves energy consumption.

1287
 On/off Run On/off High/low Run
cycling continuous cycling cycling continuous

Cooling capacity
Cooling capacity

Cap Hig
acity h C ap
Low acit
Cap y
acit
y

oa
d ad
Lo
ingL l ing
ol o
Co Co

Outdoor temperature (℃) Outdoor temperature (℃)

Figure 1: Operation of on/off system (left) & modulated system (right)

3. EER & SEER

EER (Energy Efficiency Ratio) is the ratio of the total amount of produced cold heat (W,Btu/hr) divided
by the consumed electric energy (W) at a standard rating condition. ARI 210-240.
Qtotal
(1) EER =
Etotal
SEER (Seasonal Energy Efficiency Ratio) is the ratio of the total cooling of an air-conditioner in BTU’s
during its normal usage period for cooling (not to exceed 12 months) divided by the total electric
energy input in watt-hours during the same period.
Q (T j )
(2) ∑ N
SEER =
E (T j )
∑ N

According to ARI 210/240 standard, four tests named A, B, C, D are required to obtain the SEER of
one or two speed system. From the C, D test, the CD value which is used in calculation of SEER can
be obtained and it can be usually omissible when user knows CD value already. Then 0.25 times CD
value is used generally. The driving range of two-speed (or two compressors) system can be
expressed as Figure 2, in three stages such as low load (case 1: only low stage on/off operation) ,
middle load (case 2: exchange low and high stage repeatedly), and high load (case 3: only continuous
high stage operation). The values of Q and E could be obtained according to ARI STANDARD
210/240-2003 A5.1.3.1 ~ A5.1.3.4 and SEER could be calculated using eq.(2). BL (Building Load)
expresses the cooling load and k means the operation step (1: low, 2: high) in Figure 2. Both of EER
& SEER values describe the efficiency of the system, but EER would not be a representative of the
year round energy performance of the component, nor of the seasonal energy performance of the air-
conditioners.

Btu/ h

Qssk = 2
BL (T j )
Case 3

Essk = 2

Qssk =1

Essk =1

Case 1 Case 2

82F 95F Tj

Figure 2: Performance of cooling with

compressor capacity modulation

1288
4. Main Effects of SEER
The SEER of two-speed modulation system can be obtained from A,B test at ARI standard rating
condition, Table 4, by each stage (low, high) operation. Therefore, four tests were required.
To examine the effect of these four test factors, 24 full factorial analysis was performed with a
MINITAB® release 14 (commercial statistics s/w, Minitab Inc). The standard performance data was
obtained by measurement at ARI A,B test with our 5RT grade unitary system that would be mentioned
at chapter 6 (Measurement Set Up). Table 2 shows each 2-level of four-factors was designed to ±5%
variation of its standard performance. Performance means the cooling capacity and power
consumption. Figure 3 shows the Pareto Chart (a Pareto Chart is used to graphically summarize and
display the relative importance of the differences between groups of data) of 24 full factorial analysis
and low stage at B test condition was identified as a dominant factor in increasing SEER. It shows the
same result with both cooling capacity and power consumption. It means that PLF (Part Load Factor)
is very important to actual energy efficiency. It means that the efficiency at the low cooling load is
more crucial than that at the standard cooling load condition.
4
Table 1: Factors & Levels (2 )

Table 2: 24 DOE table

(a) Cooling capacity effect (b) Power consumption effect

4
Figure 3: Pareto Chart of 2 simulation result

1289
Figure 4: The distribution of
fractional hours in temperature
range (ARI)

Figure 4 shows the distribution of fractional hours in temperature range stipulated in ARI standard. It
also shows that the PLF at low cooling loads is more important in SEER than those at higher cooling
loads. The MPS system focused on this point, and the smaller compressor size of the two
compressors was optimized to give the best PLF[4]. It resulted in the tremendous SEER improvement.

5. Multi Power System (MPS)

There are many ways in achieving capacity modulation, such as Multiple cylinder, By-pass, Pole
change compressor, and Frequency change. Each of these methods has merits and demerits, and is
used at different operating conditions. MPS is one of the methods of capacity modulation to improve
SEER, with merits of simple structures and low costs. Two compressors method, which is typical
example of multiple cylinder method, is shown in Figure 5. When the cooling load is high, all
compressors operate to meet with high cooling load, and if the cooling load is low, only one
compressor goes in working condition because the cooling load is low. Moreover, if capacities of two
compressors are different, one operation step can be added. This type is expected to be high
efficiency because by-pass loss in only one compressor operation is avoided in comparison with any
other by-pass modulation method.[5] The test of two-compressor system has been carried out under
the same conditions and only the combination of each compressor’s capacity is varied.

Figure 5: Two-compressor system

The expansion valve was optimally controlled to correspond with the compressors. The variation of
efficiency corresponding to capacity modulation ratio(R) is shown in Figure 6. The capacity
modulation ratio means the relative performance of the one out of two compressors when total
capacity of the two is set to 100%. As for the equal condition, the smaller capacity compressor
records the higher COP as by the left side of Figure 6. The equ.(3) and (4) for the system
performance corresponding to compressor capacities has derived and we found that the optimum
capacity modulation ratio makes the maximum SEER. Those equations were curve-fitted from the
cooling capacity and the power consumption data.

Q = − 0 . 0135 × R 2 + 2 . 5864 × R − 23 . 519

(3)

W com = − 0 . 0049 × R 2 + 1 . 8229 × R − 32 . 969

(4)

1290
Where the COP was obtained from the equ.(5), and the power consumption of the fan was measured.

Q
COP =
Wcomp + W fan
(5)

From the above result of the system performance corresponding to capacity modulation, the optimal
capacity ratio of the compressors has been determined. Although the smaller capacity compressor
records the higher COP as stated above, the driving range of good efficiency under smaller capacity
compressor operating would be reduced relatively if the compressor capacity ratio is too small. This
can bring the bad effect to SEER as shown in the right side of Figure 6. SEER is given with the each
capacity ratio by KS standards (KSC9306 appendix 5.3) and the optimal capacity ratio has been
found at the minimal load operation of 40% as a result shown in Figure 6.

Compressor Capacity (%) Compressor Capacity Ratio (%)

Figure 6. Performance of cooling with compressor capacity modulation

6. Measurement Set Up
To verify the efficiency performance, MPS system with two-AC fixed type rotary was designed as
Figure 7. Compressor capacity ratio was set up about 40% because of increasing SEER as result in
Figure 6. A common accumulator was adapted for protection against liquid back and electronic
expansion valve was used to control the mass flow at each 3 stage (low, middle, high) operation and
super heat. It would be very simple structure and there is no need to equip a special circuit such as
inverter drive. And it could be substitution with low cost comparatively instead of a hugely capable and
expensive one compressor. But there are some cautions that should be attributable to two-
compressor driving mechanism such as oil return, oil equality, pipe vibrations for reliability. All tests
have been progressed with our psychrometric type calorimeter that was satisfied with ISO standard
and Figure 8 shows its details. A universal air-conditioner that consisted of DC rotary type
compressor was selected as another sample to compare with MPS. All the tests were performed
according to the ARI standard, with the EER measurement under condition A, and the SEER
measurement under conditions A and B in table 4.
4
Table 3. 2 DOE table

1291
Table 4: Tested condition (ARI standard)

(a) The schematic diagram (b) Compressor set up

Figure 7: MPS system with two-compressors

Ve r i f i c a t i o n w i t h ETL 100
samp le test & satisfied
w it hin ± 2% ac c urac y
Accuracy (%)

98
Lower Limit

Calrorimeter Range
Control target
INDOOR OUTDOOR 96
Dry- bulb (℃) 15 ~ 45 - 20 ~ 60
Wet- bulb (RH%) 20 ~ 95 20 ~ 95
Air flow (CMM) 5 ~ 100
Cooling (Kcal/ hr) 1,000 ~ 25,000 94
6,000 12,000 18,000
Heating (Kcal/ hr) 1,500 ~ 32,500
Cooling Capacity (kcal/hr)

Figure 8: Details of facility that was used this study

7. Efficiency Test and Results

EER & SEER test has been studied with different type of air-conditioner as follows;
[1] Inverter system with one BLDC rotary compressor
[2] MPS with two-fixed AC rotary compressors and it operates 2-stage driving (low, high)
[3] MPS with two-fixed AC rotary compressors and it operates 3-stage driving (low, middle, high)
[4]* MPS with two-fixed AC rotary compressors but it operates only 1-stage driving (high)
( *This can be regarded as just one-compressor system because it always operates full capacity. )

Table 5 shows the test results, EER and SEER, and Figure 9 describes it as graph. From the test
results, the SEER 16.8 of Inverter system was highest value at this measurement and its EER 3.54
was good to the other system. But the SEER 16.2 of 3 stage MPS system was also good and similar
to Inverter system although its standard EER 3.32 was equal to one-compressor system. Moreover,
the SEER 12.1 of one-compressor system was very weak value compared to the other capacity
modulated system. These results explain the importance of countermeasures to part load condition.
As shown in Figure 9, under the lower stage operation at any temperature condition, system EER was
more increased than compared to its higher stage operation. It was caused by cycle effects that the
heat exchange performance would be increased when the compressor capacity was reduced with the
same size of heat exchangers. This mechanism was also shown in Figure 6 and this is the key factor
of why the MPS system’s SEER is always higher than those of one-compressor system.

1292
Table 5: Test result
Standard EER SEER
Type of Sys.
(W/ W) (Btu- hr / W)
1 Comp Sys. (AC Rotary) 3. 32 12. 1
MPS Sys. (AC Rotary) with 2 stage 3. 32 15. 8
MPS Sys. (AC Rotary) with 3 stage 3. 32 16. 2
Inverter (BLDC Rotary) 3. 54 16. 8

(a) EER of each operation stage (b) EER & SEER of each system
Figure 9: EER & SEER test result of each type sys.

The SEER of MPS system can closely approach Inverter system when lower stage operation should
be optimized to part load condition. Moreover, MPS system could be very effective air-conditioner
because of its lower cost than any other capacity modulated system such as Inverter type. This can
be another advantage of MPS system.

8. Conclusion
In this study, the EER and SEER were compared between the MPS and the variable compressor
driven air-conditioner. The ARI standard 210/240 with A and B tests with high and low stage operation,
was adopted in SEER measurements. The factorial analysis was carried out to find out the relative
importance of the factors, which contribute to the SEER.
(1) From the results of 24 full factorial analysis, the low stage at B condition was found to be the
most dominant factor in increasing the SEER of MPS system. It means that the PLF is the major
factor in the actual reduction of the year round power consumption.
(2) The MPS air-conditioner with 3-stage operation was almost as good as the BLDC type inverter
system with SEER 16.2 although it showed a low value of EER, 3.32, the same as that of the
single compressor system. Thus, there could be an air-conditioner with low EER, but with high
SEER.
Above results shows that the EER value cannot represent the actual energy consumption and the
concept of the SEER should be implemented to save energy cost. From the view point of year round
power consumption, this is the advantage of MPS system with its lower cost merits compared to
Inverter system.

1293
References
[1] ARI STANDARD 210/240. 2003 STANDARD for UNITARY AIR-CONDITIONING AND AIR
SOURCE HEAT PUMP EQUIPMENT. AIR-Conditioning & Refrigeration Institute.
[2] FINAL REPORT-APRIL 2003. Energy Efficiency and Certification of Central Air Conditioners
(EECCAC). D.G. Transportation-Energy (DGTREN) of the Commission of the E.U. vol. I~III.
[3] Sano,T., 1999, Capacity control in residential air conditioners, Refrigeration(Japanese), vol. 74,
no. 863: p.329-354
[4] Chan-ho SONG, et al, The Assessment of SEER relating to capacity modulation in the air
conditioner with two compressors. International Refrigeration and Air Conditioning Conference
at Purdue, July 12-15, 2004. R111. p1-8.
[5] C.M. KIM et al, The experimental study on the comparison in performance of the system which
used modulated compressors, Proceedings of the SAREK 2001 summer annual conference(III),
p.1114-1120

1294
Strategies for Improving HVAC Efficiency with Quality Installation
and Service
Robert Mowris, P.E., Ean Jones, B.S., Ann Jones, B.S.

Robert Mowris & Associates

Abstract
Residential and commercial air conditioning use the largest share of electricity demand in the United
States with approximately 33% or 344 GW and 313 TWh. Space heating uses 5.08 quadrillion Btu
per year or 57.3% of total residential and commercial gas consumption in the US. There are
approximately 93 million air conditioners and 35 million furnaces in the US. Each year 6 million new
air conditioners and 3.5 million new furnaces are installed. Research shows 50 to 70% of heating,
ventilating, and air conditioning systems have improper refrigerant charge and airflow, leaky ducts,
over-sized units, mismatched coils, or improper maintenance/operation causing them to be 10 to 40%
less efficient than if they received quality installation or service (QIS). Significant energy and peak
demand savings can be achieved with following measures: proper refrigerant charge/airflow, duct
testing/sealing, cleaning condenser coils, proper-sized coils, matching coils, economizer
maintenance, and cool roofs/attics. The historic market barriers to HVAC QIS measures include:
organizational practices, high start-up costs, service availability, performance uncertainty, and lack of
information. Innovative strategies are required to overcome these market barriers such as customer
education, marketing, incentives, standards, labels, and verification service providers to train and
equip HVAC technicians to deliver QIS measures. This paper provides an overview of energy savings,
market barriers, and strategies in the US to improve HVAC efficiency with QIS and transform the
market.

Introduction
Air conditioning uses the largest share of electricity demand in the United States with approximately
33% or 344 GW of total residential and commercial consumption [1]. Space heating uses the largest
share of gas in the US with approximately 57.3% or 5.08 quadrillion British thermal units (Btu) per
year of total residential and commercial gas consumption. 1 Annual air conditioning electricity
consumption is approximately 161 TWh for residential and 152 TWh for commercial. Annual space
heating consumption is approximately 3.32 quadrillion Btu for residential and 1.76 quadrillion Btu for
commercial. There are approximately 93 million air conditioners and 35 million furnaces in the US.
Each year 6 million new air conditioners and 3.5 million new furnaces are installed [2]. Energy
efficiency programs have historically provided incentives to encourage customers to purchase high
efficiency equipment to reduce heating, ventilating, and air conditioning (HVAC) energy use. This only
captures a small portion of potential savings. Research shows 50 to 70% of HVAC systems have
improper refrigerant charge and airflow, leaky ducts, over-sized units, mismatched coils, or improper
maintenance/operation causing them to be 10 to 40% less efficient than if they received quality
installation or service [3, 4, 5, 6].
Significant energy and peak demand savings can be achieved with following QIS measures: proper
refrigerant charge/airflow, duct testing/sealing, cleaning condenser coils, proper-sized coils, matching
coils, economizer maintenance, and cool roofs/attics. This paper provides a brief description of each
measure along with field measurements and supporting information. The paper provides an overview
of market barriers to HVAC QIS and how these market barriers are addressed by third-party
verification service providers. The paper discusses program implementation strategies to improve
HVAC efficiency with QIS including customer education, marketing, incentives, standards, labels, and
verification service providers (VSPs) to train and equip HVAC technicians to deliver QIS measures.
The paper also provides an overview of US efforts to address QIS and how these efforts should be
improved to transform the market.

1
The British Thermal Unit (Btu) is the energy required to raise one pound of water one degree Fahrenheit.
1

1295
Energy Savings for HVAC Quality Installation and Service Measures
This section provides descriptions and energy savings associated with the following HVAC QIS
measures: 1) proper refrigerant charge/airflow (RCA), 2) duct testing/sealing, 3) cleaning condenser
coils, 4) proper sized coils, 5) matching indoor/outdoor coils, 6) economizer maintenance, and 7) cool
roofs/attics.

Refrigerant Charge and Airflow

Proper refrigerant charge and airflow improve the efficiency and longevity of split-system and
packaged air conditioning systems. Average energy and peak demand savings for proper RCA are
approximately 9 to 12 percent for kWh and kW based on field measurements and the 2004-2005
Database for Energy Efficiency Resources Update Study [7]. Several studies show approximately 40
to 67 percent of air conditioners suffer from improper RCA, and this reduces efficiency by roughly 10
to 20 percent [2, 8, 9]. A study of commercial units in California found 46% were improperly charged
causing reduced cooling capacity and efficiency [10]. The study found 39% had very low airflow rates
2
(< 300 cfm/ton). The average airflow rate of all units tested was 325 cfm per ton or 20% less than
the airflow generally used to rate efficiency. Improper charge and reduced airflow results in reduced
efficiency and cooling capacity. The average energy impact is 9 to 12 % of annual cooling energy.
Most air conditioning technicians in the United States do not have proper training, equipment, and
verification methods to diagnose and correct improper RCA. Consequently, many new and existing
air conditioners have improper RCA causing reduced efficiency, noisy systems, and premature
compressor failure. To address this problem, the California Energy Commission (CEC) adopted
residential building standards in 2001 (CEC Standards) requiring either the Alternative Calculation
Method (ACM), thermostatic expansion valve (TXV), or proper RCA [11]. Most air conditioners are
installed using the ACM which is a computer method to show compliance with the annual energy
budget requirements of the standards. The standards require inspections by Home Energy Rating
System (HERS) raters to verify installation of TXV or proper RCA under the prescriptive approach or
the ACM. The CEC allows a TXV to substitute for proper RCA based research findings from three
laboratory studies showing improper RCA can be mitigated by a TXV [12, 13, and 14]. The studies
found TXV systems only had an advantage over non-TXV systems when undercharged.
TXV-equipped systems have problems when incorrectly installed leading to a phenomenon known as
“valve hunting.” This can occur when the evaporator coil experiences reduced heat loads caused by
low airflow, dirty or icy coils, and low refrigerant charge [15]. Under these circumstances the TXV can
lose control and successively overfeed and then underfeed refrigerant to the evaporator while
attempting to stabilize control causing reduced capacity and efficiency. Overfeeding liquid to the
evaporator can also damage the compressor. The tendency for hunting can be reduced by correcting
RCA, by relocating the TXV sensing bulb to a better location inside the evaporator coil box, and by
insulating the sensing bulb. Field and factory-installed TXV sensing bulbs are often installed without
insulation, without adequate linear contact, and at incorrect orientations. This practice is not
recommended by manufacturers [16]. Factory-installed TXVs with un-insulated sensing bulbs inside
the evaporator coil box will be influenced by mixed supply-air temperatures which are typically 10-
20°F higher than vapor line temperatures. Field-installed TXVs with un-insulated sensing bulbs
located in attics or garages will be influenced by attic or garage temperatures which are 50 to 80°F
higher than vapor line temperatures. The three laboratory studies measured TXV-equipped air
conditioners with the evaporator coil box, TXV, and well-insulated sensing bulb located in conditioned
space and this is not typical of field conditions.
Field measurements for a 4-ton TXV air conditioner are shown in Figure 1 [2]. The TXV air
conditioner used 5.8 kW when overcharged and 4.8 kW when properly charged for a savings of 1 kW.
3
The energy efficiency ratio (EER) increased by 37 percent from 7.1 EER to 9.7 EER.

2
The “cfm” is defined as cubic feet per minute of airflow. The “ton” is defined as 12,000 Btu per hour of cooling capacity equal
to the rate of extraction of latent heat when one short ton of ice (i.e., 144 Btu per pound) is produced from water at the same
temperature.
3
The energy efficiency ratio or EER is the cooling capacity in thousand Btu per hour (MBtuh) divided by total air conditioner
electric power (kW) including indoor fan, outdoor condensing fan, compressor, and controls. EER is typically measured under
laboratory conditions at 95°F condenser entering air, 80°F drybulb,and 67°F wetbulb evaporator entering air.

1296
 11 11

Energy Efficiency Ratio (EER)

10 10

9 9

Improper Proper
8 Refrigerant Transition Refrigerant 8
Charge and Charge and
Airflow Airflow

7 7
Total AC Power (kW)

6 6

5 5

4 4
0 10 20 30 40 50 60 70 80 90
Time (Minutes)

Figure 1. Measurements of a 4-ton TXV Unit with and without Proper RCA Source: [2]

Measurements of a 10-ton packaged rooftop air conditioner with and without proper RCA are shown
in Figure 2. The 10-ton unit had a dirty/icy evaporator coil and dirty air filters and was overcharged by
14.2 ounces or 7.1 percent of the factory charge. With improper RCA the average efficiency was 5.7
EER, and average power usage was 13 kW. With proper RCA the efficiency improved to 10.3 EER,
and the average power was reduced to 9.5 kW. This is consistent with the ARI rating of 10.3 EER.

44 14
42
40 12
Energy Efficiency Ratio

38
36 10
34
32 Clean Filter and 8
30 Clean Evaporator Coil
28 Proper RCA 6
26
24 Dirty Filter and 4
22 Dirty/Iced Evaporator Coil
20 Improper RCA 2
18
16 0
Total AC kW

14
12 -2
10
8 -4
6
4 -6
2
0 -8
0 5 10 15 20 25 30 35 40 45 50
Time (minutes)

Figure 2. Measurements of 10-ton Packaged Unit with and without Proper RCA Source: [2]

1297
Duct Testing and Sealing
Duct testing and sealing reduces duct leakage and run time and improves efficiency of heating,
ventilating, and air conditioning systems. Average energy and peak demand savings for residential
duct testing and sealing are 5.8 to 8.8% for kWh and 8.4 to 27.6% for therms based on the 2004-2005
DEER Study Update [7]. Research indicates that duct leakage in commercial buildings may be
comparable to residential buildings [18, 19, 20]. Ducts in commercial buildings with simple rooftop
package units are often located in ceiling plenum spaces that are similar to residential attics.
Duct leakage is typically described three ways: 1) Fraction of flow through the HVAC equipment that
is lost, 2) Equivalent hole size, or 3) Leakage flow at a reference pressure. The latter two are often
normalized by either the surface area of the ductwork or the conditioned floor area. The ductwork in
small commercial buildings is leaky by all three metrics. Using the first metric, the work at LBNL
indicates the average supply duct leakage for 25 Constant-Air-Volume (CAV) systems was 26% of the
flow through the HVAC equipment, as compared to average supply-side leakage of 17% in residential
systems [18]. Using the second metric for the 25 systems, LBNL research showed an average
2 2
normalized commercial duct leakage area of 3.7 cm per m of floor area for supply and return ducts.
The comparable residential attic duct-leakage is 1.3 cm per m2 of floor area. These results suggest
2

that duct leakage of some light commercial duct systems can be greater than residential systems.

Cleaning Condenser Coils

Clean condenser coils provide optimal heat transfer from the condensing coil to deliver rated capacity
and efficiency for split-system and packaged air conditioners. Average energy and peak demand
savings for condenser coil cleaning are 8 to 12% based on the 2004-2005 DEER Study [7]. Trane
found an efficiency loss of 27 percent due to conditions of accelerated fouling for multi-row coils
equivalent to 8 years of typical operating conditions or annual fouling of 6.8% for commercial multi row
coils [22]. Trane provided data from a study of two air conditioners operated continuously with
condensers exposed to a dirty factory environment for 18 months, and this was reported as equivalent
to roughly 4 to 8 years of typical operating hours. Performance measurements indicated the air
conditioner with a standard plate fin coil lost 17% of its capacity and 27% of its efficiency.
Field measurements of a packaged air conditioning unit with a leaky Schrader valve and dirty
condenser coil are provided in Table 1 [23]. The AC unit was found to be inoperable due to lack of
refrigerant from a leaky Schrader valve. The leaky valve was repaired and 55 ounces of R-22 was
4
weighed in with a digital scale to achieve proper superheat within ±5°F. Technicians then combed
and cleaned the condensing coil. The cooling capacity and electric power usage were measured
before and after adding refrigerant charge and combing/cleaning the condensing coil. Savings from
combing and cleaning the condenser coil are estimated to be 15% based on the measured EER
improvement (9.0 EER with dirty coil and 10.4 EER with clean coil).

Table 1. Field Measurements of Air Conditioner Efficiency from Cleaning Condensing Coil

Rated Rated Measured Measured

EER Capacity Rated EER Capacity Measured
Description 95 ODT Btuh kW 80 ODT Btuh kW
Leaky Schrader valve no refrigerant charge 8.5 18,000 2.1 0.2 174 1.1
Repaired leaky Schrader valve and added
55 ounces of R-22 refrigerant 9.0 19,000 2.1
Post-retrofit cleaned condenser coil 10.4 21,900 2.1
Source: [23]

Proper Sized Coils (ACCA Manual J)

Proper sized coils (per ACCA Manual J) improve the capacity and efficiency of split-system air
conditioners. Energy and peak demand savings for proper sized evaporator/condenser coils per
ACCA Manual J are 10 to 18 percent based on field studies showing most units are significantly
oversized, resulting in inefficient operation, reduced reliability due to frequent cycling of compressors,
and poor humidity control [24]. Oversized systems waste capital invested in both the HVAC unit and
distribution system. System over-sizing also affects the ability of the system to provide simultaneous
economizer and compressor operation, and exacerbates problems with distribution system fan power,

4
The factory charge is 55 ounces of R-22 refrigerant, for the 1.5 ton Carrier rooftop AC unit Model 585GJ018040.

1298
since larger units are supplied with larger fans. Each time an air conditioner starts, the input energy is
approximately constant, while it takes several minutes to reach full cooling capacity. Oversized units
operate for a shorter cycle, and the startup time is a greater fraction of the total runtime. The startup
losses are also a greater fraction of the total cooling output, reducing overall efficiency. Systems that
are properly sized will run longer during each cycle, and the startup losses are small relative to total
cooling output. In a study of 250 rooftop units conducted for Pacific Gas and Electric Company, the
typical runtime under hot conditions was 6 minutes, with an off-time of 16 minutes [25]. This
represents a 27% runtime fraction with a reduction in unit efficiency of 18%. The system efficiency is
reduced as the runtime decreases. When the unit runs continuously (CLF = 1), the part-load factor is
1.0, indicating no degradation due to cycling. When the unit runs 60% of the time, the CLF is 0.6 and
the unit efficiency is reduced by about 10%. If the unit runs only 30% of the time, the efficiency is
reduced by about 15% [26].

ARI Matching Coils on Split System Air Conditioners

ARI matching coils provide the rated capacity and efficiency for split-system air conditioners. Energy
and peak demand savings for ARI matching coils on split system air conditioners are 5 to 15 percent
based on field measurements of systems with improper matching coils. Even with correct refrigerant
charge many split-system air conditioners do not perform at their rated efficiency due to improperly
matching evaporator and condenser coils. Condensing coil manufacturers cannot guarantee rated
efficiency per Air-Conditioning and Refrigeration Institute (ARI) SEER/EER ratings with evaporator
coils manufactured by independent coil manufacturers that are not listed in the ARI directory as a
proper match for the condensing coil [27]. Field measurements of two new split-system air
conditioners are provided in Table 2 [28]. Pre-EER values were measured with improper RCA, and
post-EER values were measured with proper RCA. The post-EER is 6.5 or 35 percent less than the
10 EER rating due to improper matching coils. This is a common problem with split system units.

Table 2. Field Measurements for Two New Split-System Air Conditioners

Percent
Measured Average Duct Infil. Charge
Cooling Outdoor, Leak cfm Adjust
Rated Capacity Indoor cfm @ Service per
Rated Capacity Post Dry/Wet Airflow @ 25 50 EER EER Adjust Factory
Site EER MBtuh MBtuh Bulb °F cfm Pa Pa Pre Post Oz. Charge
#1 10 51 38.5 105/81/65 1631 19% 1830 3.9 6.5 +98.2 +49.4%
#2 10 51 41.6 105/80/64 1734 12% 1537 5.5 6.5 +12.5 +6.3%
Note: Rated EER values are based on manufacturers’ data. Source: [28]

Economizer Set-up and Maintenance

Economizer set-up and maintenance reduce unnecessary air conditioning when outdoor temperatures
are cool enough to provide free cooling (i.e., reduce compressor use). Average energy and peak
demand savings for commercial economizer set-up and maintenance are 9 to 21% based on the
2004-2005 DEER Study [7]. A study of commercial packaged units in California with economizers
generally found 64% not operating properly [24]. Failure modes included dampers that were stuck or
inoperable (38%), sensor or control failure (46%), or poor operation (16%). The average energy
impact of inoperable economizers is about 37% of the annual cooling energy. The 2005 DEER Study
found annual savings of 21.3%.
Selection of the changeover setpoint has a major influence on the energy savings potential of an
economizer. If the changeover setpoint is set too low, then mechanical cooling will operate
exclusively even when the economizer is capable of meeting the cooling load [24]. Single point
changeover setpoints are selected on the economizer controller according to an A, B, C or D setting.
The selection of the changeover setpoint depends on the climate; humid climates require a lower
setpoint than dry climates. According to the Title 24 Energy Standards, the “A” setpoint is appropriate
for all climates in California. However, observations of single point changeover setpoint selection in
the PIER study behind this Design Guide showed that the “A” setting was rarely used. Manufacturers
may not ship their products with the “A” setting as the default, requiring a field adjustment of the
controller setting. The distribution of economizer control setpoints observed in the field for single point
enthalpy economizers shows only 28% of the systems in the “A” position as required by Title 24.
Most systems were set in the “D” position, which results in the fewest hours of economizer operation.

1299
Many of the economizer problems observed in the field can be avoided through careful selection and
specification of rooftop unit economizer features. The following measures will improve economizer
efficiency and reliability [24].
1. Specify factory-installed and run-tested economizers. The majority of economizers are
installed by the distributor or in the field. Specifying a factory-installed and fully run-tested
economizer can improve reliability.
2. Specify direct drive actuators. Economizers with direct drive actuators and gear driven dampers
can reduce problems with damper linkages that can loosen or fail over time.
3. Specify differential (dual) changeover logic. Differential temperature or enthalpy changeover
logic instead of single point changeover systems eliminates problems with improper setpoint and
maximizes economizer operation.
4. Specify low leakage dampers for outside and return air. Low leakage dampers with blade and
jamb seals will improve economizer effectiveness by limiting return air leakage during economizer
operation and outdoor air infiltration when the unit is switched off.

Cool Roofs and Cool Attics

Cool roofs and cool attics reduce cooling loads and improve cooling capacity and reduce run time for
split-system air conditioners in residential buildings. Average energy and peak demand savings are
10 to 30% based on studies by Lawrence Berkeley National Laboratory and the Florida Solar Energy
Center [29, 30]. Cool attics provide similar savings using radiant barriers or attic ventilation fans.
Solar-powered or conventional attic ventilation fans with thermostat control will reduce solar heat load
transferred to conditioned space and attic temperatures where air conditioner evaporators and ducts
are located. Cool roofs stay cooler in the sun than conventional roofs. Roofs that have high solar
reflectance (high ability to reflect sunlight) and high thermal emittance (high ability to radiate heat)
tend to stay cool in the sun. The same is true of low-emittance roofs with exceptionally high solar
reflectance. Low roof temperatures lessen heat flow from the roof into the building, reducing space
cooling electricity use in conditioned buildings. Since building heat gain through the roof peaks in late
afternoon, when summer electricity use is highest, cool roofs also reduce peak electricity demand.
Prior research indicates savings are greatest for buildings located in climates with long cooling
seasons and short heating seasons, particularly buildings that have distribution ducts in the plenum,
cool-coatable distribution ducts on the roof, and/or low rates of plenum ventilation [30]. Prior studies
measured air-conditioning energy and peak demand savings from cool roofs on nonresidential
buildings in California, Florida, and Texas. Cool roofs typically have measured summertime air-
conditioning energy and peak demand savings of 10–30%, although savings have been as low as 2%.
Cool roofs transfer less heat to the outdoor environment than do conventional roofs. The resulting
lower outside air temperatures can slow urban smog formation and increase human health and
outdoor comfort. Reduced thermal stress may also increase the lifetime of cool roofs, lessening
maintenance and waste. The potential of cool roofs to save cooling electricity has not gone
unnoticed. In its revised standards for commercial and residential buildings, the American Society for
Heating, Refrigeration, and Air-conditioning Engineers (ASHRAE) included provisions to offer credits
in building energy-use budgets for cool roofs [31]. In January 2001, the state of California followed
ASHRAE by adopting standards to offer Title 24 compliance credit to new commercial buildings with
cool roofs [11]. Georgia, Florida, and Chicago also have building codes to encourage cool roofs.

Market Barriers
There are many market barriers to HVAC quality installation and service (market barrier definitions are
from 32). Performance uncertainty is an important barrier since consumers have difficulty evaluating
claims about future benefits associated with unverified energy guide performance labels. Truth in
advertising is important to consumers who assume new units will be installed properly. Unfortunately,
many new air conditioners do not perform as advertised due to improper installation or service, and
this undermines the credibility of the US energy guide labels [33]. At a minimum, the labels should
include a caveat regarding SEER ratings only being valid for air conditioners installed with quality
installation and service according to manufacturers’ specifications. Other important market barriers
include lack of information or knowledge about the importance of quality installation and service in
terms of delivering rated efficiency, reducing noise, and maintaining longer life of air conditioners.
Organizational practices and rules of thumb discourage quality installation such as “add or remove
refrigerant until the suction line is six-pack cold” or “shows 70 psig on the suction side and less than

1300
250 psig on the liquid line.” Service availability for new air conditioners is an important barrier for
manufacturers, distributors, and dealers who are generally not verifying quality installation and service
due to lack of awareness and availability of cost effective and easy-to-use verification services.
These market barriers are addressed by third-party verification service providers such as: Verified™,
Enalasys™, Honeywell Service Assistant™, and CheckMe™ [34, 35, 36, 37]. VSPs offer cost-
effective methods to verify proper RCA, TXVs, duct testing/sealing, and other measures. Verification
software is provided on several platforms: 1) Personal Digital Assistant (PDA), 2) cell-phone
telephony, 3) web-enabled PDA, 4) cell phones, 5) notebook computers, or 6) telephone call-in
systems. VSP programs can be deployed in any language, and since the systems are automated, the
cost per verified unit is low. Verification information is collected and archived on databases where
technician-supplied data is checked for accuracy and can be viewed over the internet by consumers,
inspectors, dealers, and program managers. The VSP randomly inspects jobs to ensure quality
results. One third-party VSP provides clearly identifiable Verified™ labels and locking, double-
sealing, laser-etched Schrader caps, with tamper-proof keys for technicians. Locking caps are
designed to maintain proper RCA for the life of the air conditioner. This is important since air
conditioning systems are made of welded copper pipe and Schrader valves are the weakest link. Air
conditioners vibrate and this causes Schrader valve cores to loosen over time and leak refrigerant.
Most air conditioners have easily removable Schrader caps without integral “O-ring” seals. Safety is
another reason why locking Schrader caps are important as evidenced by the deaths of two
teenagers in Southern California due to inhalation of refrigerant as an intoxicant [38].

Strategies
A number of intervention strategies are required to improve HVAC efficiency with quality installation
and service such as third-party verification service providers, customer education, marketing, training,
incentives, standards, and labels. Third-party verification service providers are required to train and
equip HVAC technicians to deliver and verify quality installation and service. Customer education,
standards, and labels are important to create demand for QIS. Incentives will help motivate interest,
but are insufficient by themselves to deliver HVAC QIS and transform the market. Consumers
generally assume their air conditioners are properly installed. Current efficiency standards do not
mention the importance of QIS, and California building standards allow a TXV to substitute for proper
RCA to receive the same compliance credit. Therefore, most consumers and builders do not
understand the value of proper RCA. Research studies show HVAC dealers lack interest, training,
equipment, and methods to perform proper installation and service measures such as RCA and duct
testing/sealing measures. To develop a robust set of supply-side market actors, Verification Service
Providers (VSPs) must recruit, train, and equip local HVAC dealers to deliver HVAC QIS measures.
Several utility programs in the United States offer verification service provider incentive programs.
Most programs do not include the new construction market and most programs are implemented
through only one VSP. This is a problem for HVAC dealers who are trained and equipped to perform
HVAC QIS with a different VSP. Switching to a different VSP to participate in a program creates
unnecessary barriers and can cost thousands of dollars per technician. Having a different program in
each utility service area creates problems for larger HVAC contractors who have an established VSP
relationship. Classroom training on quality installation will not be effective without VSP involvement to
increase participation and help transform the market for third-party verification. With greater
participation, there will be more demand and competition for QIS. Competition will expand the market
and drive down the incremental measure cost to the point where QIS is “standard practice” and
incentives can eventually be withdrawn (i.e., exit strategy).
Utilities and government agencies should consider implementing comprehensive and consistent
HVAC programs targeting new and existing residential and commercial market segments. The
following measures should be considered: 1) proper refrigerant charge/airflow, 2) duct testing/sealing,
3) cleaning condenser coils, 4) proper sized coils, 5) matching indoor/outdoor coils, 6) economizer
maintenance, and 7) cool roofs/attics. VSPs and Home Energy Rating System providers should work
together to recruit, train, and equip HVAC contractors to help transform the market for third-party
verification of quality installation and service for both new and existing construction. Programs should
consider internet or database registration and permanent labels for identification and facilitation of
evaluation, measurement, and verification inspections. Locking Schrader caps should be promoted
for RCA measures to help maintain efficiency, promote public health and safety, encourage proper
refrigerant management practices, and prevent further stratospheric ozone depletion [39]. Programs

1301
should work with manufacturers to incorporate HVAC quality installation and service standards within
warranty requirements, ASHRAE, and the International Standards Organization Technical Committee
86 (ISO, refrigeration and air conditioning, www.iso.org).

Conclusions
Energy efficiency programs have historically provided incentives to encourage customers to purchase
high efficiency equipment to reduce HVAC energy use, but this only captured a small portion of
potential savings. Research shows 50 to 70% of HVAC systems have improper refrigerant charge
and airflow, leaky ducts, over-sized units, mismatched coils, or improper maintenance/operation
causing them to be 10 to 50% less efficient than if they received quality installation or service. With
approximately 93 million air conditioners and 35 million furnaces in the US and 6 million new air
conditioners and 3.5 million new furnaces installed each year, the estimated potential energy savings
from HVAC QIS are significant. These savings can be achieved through a number of intervention
strategies aimed at downstream, midstream, and upstream market actors including: education,
marketing, training, incentives, standards, and labels. One of the most important strategies for
success is developing and supporting a robust supply-side verification service provider network to
train and equip HVAC technicians to deliver and verify quality installation and service. Utilities and
government agencies should encourage manufacturers, distributors, and HVAC dealers to work with
VSPs to improve HVAC efficiency with QIS. Utilities and government agencies should also motivate
consumers to demand HVAC QIS through education, marketing, incentives, standards, and labels.

References
[1] Energy Information Agency (EIA). 2001. Residential Energy Consumption Survey. Available
online: http://www.eia.doe.gov/emeu/recs/recs2001. Energy Information Agency (EIA). 1999.
Commercial Building Energy Consumption Survey. Available online:
http://www.eia.doe.gov/emeu/cbecs/tables.
[2] American Refrigeration Institute (ARI). 2004. ARI Statistical Release of Unitary Heating and
Cooling Section's Reports of U.S. Manufacturers Shipments. Arlington, Va.: Available online:
http://www.ari.org/sr/2003. Gas Appliance Manufacturers Association 2004.
http://www.gamanet.org/gama.
[3] R. Mowris, Blankenship, A., Jones, E., Field Measurements of TXV and non-TXV Air
Conditioners, ACEEE 2004 Summer Study, Washington, DC.: ACEEE, 2004.
[4] Hammarlund, J., Proctor, J., Kast, G., and Ward, T. 1992. “Enhancing the Performance of
HVAC and Distribution Systems in Residential New Construction.” 1992 ACEEE Summer
Study, 2: 85-87. Washington, D.C.: ACEEE.
[5] Neme, C., Nadel, S., and Proctor, J. 1998. National Energy Savings Potential from Addressing
HVAC Installation Problems, Vermont Energy Investment Corporation, prepared for US
Environmental Protection Agency.
[6] Parker, D., Cummings, J., Meier, A., Home 1993. Will Duct Repairs Reduce Cooling Load?
Berkeley, Calif.: Home Energy Magazine.
[7] Itron, Inc. 2004-2005 Database for Energy Efficiency Resources (DEER) Update Study Final
Report. 2005. Prepared for Southern California Edison Company. Washington, DC. Available
online: http://eega.cpuc.ca.gov/deer/.
[8] Palani, M., O’Neal, D., and Haberl, J. 1992. The Effect of Reduced Evaporator Air Flow on the
Performance of a Residential Central Air Conditioner, The Eighth Symposium on Improving
Building Systems in Hot and Humid Climates.
[9] Rodriguez, A. 1995. The Effect of Refrigerant Charge, Duct Leakage, and Evaporator Air Flow
on the High Temperature Performance of Air Conditioners and Heat Pumps, Palo Alto, Calif.:
Electric Power Research Institute.
[10] Jacobs, P. Small HVAC System Design Guide, prepared for the California Energy Commission,
500-03-082-A12, prepared by Architectural Energy Corporation, Boulder, CO. October 2003.
[11] California Energy Commission (CEC). 2001 Energy Efficiency Standards. Report. 132. CEC
P400-01-024. June 1, 2001, Sacramento, Calif.: California Energy Commission.
[12] Davis, R. 2001a. Influence of the Expansion Device on Performance of a Residential Split-
System Air Conditioner. Report No.: 491-01.4. San Francisco, Calif. Pacific Gas and Electric.

1302
[13] Davis, R. 2001b. Influence of Expansion Device and Refrigerant Charge on the Performance of
a Residential Split-System Air Conditioner using R-410a Refrigerant. Report No.: 491-01.7. San
Francisco, Calif.: Pacific Gas and Electric.
[14] Farzad, M., O’Neal, D. 1993. “Influence of the Expansion Device on Air Conditioner System
Performance Characteristics Under a Range of Charging Conditions.” Paper 3622. ASHRAE
Transactions. Atlanta, Ga.: American Society of Heating Refrigerating and Air-Conditioning
Engineers.
[15] Tomczyk, J. 1995. Troubleshooting and Servicing Modern Air Conditioning and Refrigeration
Systems. ESCO Press. Mt. Prospect, Ill.: Educational Standards Corporation.
[16] Advanced Distributor Products (ADP). 2003. TXV Installation Instructions. 0991710-01 Rev 1,
October 03. Stone Mountain, Ga.: Online: www.adpnow.com. AllStyle Coil Company, L.P.
(Allstyle). 2001. Evaporator Coil Installation Instructions. Brittmore, Texas. Carrier Corporation.
2002. Installation Instructions: Thermostatic Expansion Valve Kit. KAATX, KHATX (R22),
KSATX (R410a). Syracuse, N.Y. Emerson Climate Technologies, Inc. 1998. Installation
Instructions Expansion Valve Kits TXV153 & TXV355. Lewisburg, Tenn.
[17] O’Neal, D., Farzad, M. 1990. “The Effect of Improper Refrigerant Charging on Performance of
an Air Conditioner with Capillary Tube Expansion.” Energy and Buildings 14: 363-371.
[18] W. Delp, N. Matson, D. Dickerhoff, D. Wang, R. Diamond, M. Modera "Field Investigation of
Duct System Performance in California Light Commercial Buildings", ASHRAE Trans. 104(II)
1998, June 1998.
[19] T. Xu, F. Carrie, D. Dickerhoff, W. Fisk, J. McWilliams, D. Wang, and M. Modera, ‘‘Performance
of Distribution Systems in Large Commercial Buildings’’, Energy and Buildings (2000),
Lawrence Berkeley National Laboratory Report, June 2000, LBNL-44331.
[20] M. Modera, O. Brzozowski, D. Dickerhoff, W. Delp, W. Fisk, R Levinson, D. Wang, , “Sealing
Ducts in Large Commercial Buildings with Aerosolized Sealant Particles”, (2000), Lawrence
Berkeley National Laboratory Report LBNL-42414 October, 1998.
[21] ASHRAE Standard 152P. Method of Test for Determining the Design and Seasonal Efficiencies
of Residential Thermal Distribution Systems. 2001.
[22] Trane, 1990. Spine Fin™: The Technology of Heat Transfer, The Trane Company, American
Standard Inc., Pub #14-4900-1. Braun, R.H., Problem and Solution to Plugging of a Finned-
Tube Cooling Coil in an Air Handler, ASHRAE Transactions Vol. 92. Pt. 1 pp. 385-398, 1986
[23] R. Mowris, Blankenship, A., Jones, E., Evaluation Measurement and Verification Report for the
Mobile Energy Clinic Program #118-02, Prepared for ADM, Inc. 2004.
[24] Small HVAC System Design Guide, prepared for the California Energy Commission, 500-03-
082-A12, prepared by Architectural Energy Corporation, Boulder, CO. October 2003.
[25] Felts, D. 1998. Pacific Gas and Electric Company Roof Top Unit Performance Analysis Tool
Program—Final Report. San Francisco, CA. Pacific Gas and Electric Company.
[26] Henderson, H., Y. Huang, D. Parker, 1999. Residential Equipment Part Load Curves for Use in
DOE-2, Berkeley, CA. Lawrence Berkeley National Laboratory, LBNL-42175.
[27] The Air-Conditioning and Refrigeration Institute's On-Line Directories of Certified Equipment are
available online at http://www.ariprimenet.org.
[28] R. Mowris, Blankenship, A., Jones, E., Measurement & Verification Report for the Residential
Ground Source Heat Pump Program, Prepared for Redding Electric Utility, 2004
[29] Parker, D., Sherwin, J., "Comparative Summer Attic Thermal Performance of Six Roof
Constructions," The 1998 ASHRAE Annual Meeting, Toronto, Canada, June 20-24, 1998.
[30] Levinson, R., Akbari, H., Konopacki, S., Bretz, S., “Inclusion of Cool Roofs in Nonresidential
Title24 Prescriptive Requirements,” Energy Policy 33 (2005) 151–17
[31] ASHRAE Standard 90.1-2001: Energy Standards for Buildings Except Low-rise Residential
Buildings and ASHRAE 90.2-2001: Energy-Efficient Design of Low-rise Residential Buildings.
[32] Eto, J. Prahl, R., Schlegel, J. 1996. A Scoping Study on Energy Efficiency Market
Transformation by California Utility DSM Programs. LBNL-39058. Berkeley, Calif.: Lawrence
Berkeley National Laboratory.
[33] United States Federal Trade Commission (USFTC) 1996. Appliance Labeling Rule for Central
Air Conditioners and Heat Pumps. 16 CFR Part 305. Authorized by the Energy Policy and
Conservation Act, Subchapter III, Part A, 42 U.S.C. 6291 et seq. 52 FR 46894, 1987, as
amended at 54 FR 28034, 1989. http://www.ftc.gov/bcp/conline/edcams/eande/index.html.
[34] Verified™, Inc. 2004. Olympic Valley, Calif.: Ve rified™ Inc. Online: http://www.verified-rca.com.
[35] Enalasys. 2004. Calexico, Calif.: Enalasys. Online: http://www.enalasys.com.

1303
[36] Honeywell 2004. Honeywell HVAC Service Assistant. Minneapolis, Minn.: Honeywell
Corporation. Available online: www.honeywell.com/building/components.
[37] Proctor, J. 2004. CheckMe!TM, San Rafael, Calif.: Proctor Engineering Group. Available online:
http://www.proctoreng.com/checkme/checkme.html.
[38] Los Angeles Times. 2004. Woman and Boy Found Dead in La Puente Home Pool (“drowned
after inhaling refrigerant and losing use of their limbs”). 5-07-04. Los Angeles, California.
[39] Global Environment & Technology Foundation. The State of Stratospheric Ozone Depletion.
Available online: www.getf.org/cecs/Ozone_Study.pdf.

1304
 Posters

1305
1306
The Panpower Transformer
Jörgen Ekelöf, Alan Ericsson

PanPower AB, Sweden

Abstract
Target:
To develop an energy- and cost-efficient linear power supply transformer that will meet future
requirements on low no-load (standby) power consumption and high efficiency rate during load.
Method:
To combine the advantages and to eliminate the disadvantages of existing technologies, the EI-core
transformer and the Toroidal-core transformer. The low production cost of the EI-core transformer is
combined with the superior electrical performance of the toroidal-core transformer. Winding of a
straight bobbin which is formed to a circular shape after the winding is completed and after that
insertion of the electro-steel core-band into the toroidal bobbin.
Result:
• The PanPower toroidal (circular) shaped high performance linear power supply transformer
which easily can be produced in small sizes down to 1VA output or even less.
• A linear power supply transformer that will meet future requirements on low no-load (standby)
power losses and high efficiency rate during operation.
• A simple and fast manufacturing process in combination with less materiel use resulting in
lower total cost than even the conventional EI-core transformer.
• A low cost and energy efficient linear power supply transformer with small size and low weight
generally available for all transformer manufacturers.
Conclusions:
Future requirements on reduced no-load (standby) power losses and increased efficiency rate during
operation will be unachievable for the today most commonly used transformer, the EI-core
transformer. The superior conventional toroidal transformer is too expensive. A new linear transformer
technology is needed to meet future market and legal demands. The new high performance linear
power supply transformer, the PanPower transformer meets future requirements on energy saving at
very low cost to the benefit of the environment and the end user.

General situation
During the last decade the marked for external power supplies and battery chargers has in a literal
sense exploded. Mobile phones, portable A/V-products, laptop PC’s, numerous of IT products, power
tools etc. are powered by external power supplies “EPS”.
Many of these EPS-units suffer from poor performance both in terms of far too high standby power
consumption and too low efficiency rate in operation mode.
A large number of these EPS are permanently connected to the mains, also when not in use, resulting
in energy waste.
Also a large number of internal power supplies “IPS” suffer from poor performance with especially too
high standby power consumption.
Energy saving is on the agenda almost everywhere and future regulations on reduction of the standby
losses and increased efficiency are expected to come all over the world.

Power supply technologies

Most EPS and IPS units are based upon either the linear technology or the switching technology.
Both technologies have their advantages and disadvantages.
During the last decades almost all R&D-work on power supplies has been focused on the switch
mode power supply (also called electronic transformers) while very little or almost nothing has been
done for improvement of the linear power supply (conventional transformers).
The result is large quantities of poor linear EPS/IPS units still in use causing energy waste.
Already several years ago PanPower AB understood the intolerable situation and decided to seek for
ways for improvement of the linear transformer.

1307
Linear technology
Why use linear technology
In certain sizes the linear power supply (EPS & IPS) is a low cost alternative for the benefit of the end
consumer.
The linear transformer is robust and solid making it very durable, reliable and almost immune against
incoming transients. Suitable for markets with poor mains network and severe (i.e tropical) weather
conditions.
For AC/AC operation linear is the only low cost technology available.

Types of linear power supplies

Almost all linear EPS units are based upon two types of transformers, the most commonly used, the
EI-core transformer (>95%) or the less used Toroidal-core transformer (<5%).

EI Core Toroidal Core

air-gaps resulting in lower performance

orientation of the magnetic flux in the core material

orientation of the magnetic domains in the core material

To use the core material in the best way, the magnetic flux and the magnetic domains shall have
same orientation.

Characteristics of the EI-core transformer

• Low production cost, simple and fast production.
• Poor geometrical configuration of the iron-core leading to energy waste.
• Air-gaps in the iron-core also leading to energy waste.
• High standby energy loss.
• Low efficiency rate in operation mode.
• Robust against transients, incoming disturbances.
• Large size and weight.
• Difficult to meet future requirements on energy saving.

1308
Characteristics of the Toroidal-core transformer
• High production cost.
• Not suitable for high volume mass production because of time-consuming production.
• Perfect geometrical configuration of the iron-core and no air-gaps resulting in low energy
waste.
• Lower standby energy loss.
• Higher efficiency rate in operation mode.
• Robust against transients, incoming disturbances.
• Smaller size and lower weight.
• The winding method using a shuttle making production of small size transformers <15VA
impossible.
• Will meet future requirements on energy saving.

The PanPower linear transformer

Overall targets
• To develop a new generation of low cost energy efficient linear transformers available for the
open market and which shall replace the EI-core transformer.
• The transformer shall allow very fast and simple high volume mass production.
• The transformer shall be suitable for both manual and automatic production.
• The transformer shall meet future market requirements in terms of both low standby energy
loss and high efficiency rate in operation mode. The target shall be to meet the requirements
primarily in the Energy Star and as far as possibly in the European Code of Conduct
programs for EPS. Reference to below table.
• The transformer shall be possible to produce in small sizes down to around 1 VA.
• Weight and size shall be less than what is typical today, limited use of raw materials and no
material waste in production.
• Lower price compared with existing technologies.
• The targets shall normally be possible to reach even when using standard low cost non-
oriented silicon-steel.

Energy Star and Code of Conduct no-load (standby) power loss and efficiency rate limits.
Output power Max no load power consumption Efficiency rate
Energy Star Code of Conduct
1 VA 0.5 W 0.3 W 49 %
2 VA 0.5 W 0.3 W 55 %
5 VA 0.5 W 0.3 W 63 %
10 VA 0.5 W 0.3 W 70 %
15 VA 0.75 W 0.3 W 73 %

Method
Because of the energy saving targets PanPower AB decided to focus on the toroidal-core technology.
The general idea has been to make use of and combine the advantages of the EI-core and Toroidal-
core transformers and to eliminate all their disadvantages.
The target was to keep and even reduce the low production cost of the EI-core transformer, to
overcome the time-consuming production method of the Toroidal-core transformer and to make
production of small size transformers, commonly used in EPS, possible.

1309
By deleting the conventional toroidal winding method using a shuttle and replace it with conventional
winding of a straight bobbin, the production speed can be increased more than 100 times compared
with the production time of a conventional toroidal transformer and it will allow production of small
sized transformers.
To make this possible we had to develop a completely new production method for the toroidal-core
transformer which can be described as follows:

1. To make a straight winding on a slave-bobbin using a conventional high speed winding

machine. There will be separate bobbins for the primary and secondary windings.

2. To transfer the straight winding from the slave-bobbin to a toroidal shaped master-
bobbin. The master-bobbin is divided in two parts, one for the primary and one for the
secondary windings.

3. The two master-bobbin halves are connected to each other to form a complete toroid.

4. The electro-steel core-band is inserted into the cavity of the two bobbin halves to make a
complete toroidal transformer.

Result: The PanPower Transformer

• Simple high speed winding of a straight bobbin.
• Simple high speed production process, low cost production.
• Allowing simple near to market production.
• Possible to produce in small sizes down to 1 VA or even less.

1310
• The window area (centre hole) can be reduced to a minimum resulting in smaller size, less
material use and lower energy loss.
• Low standby energy loss and high efficiency rate in operation mode.
• No stamping of core material, no material waste.
• Small size and low weight.
• Low price, even lower than for the EI-core transformer.
• A robust, solid, low cost transformer with a long lifecycle and which will meet future global
requirements on energy savings, in most cases even when using standard non-oriented low
cost silicon-steel.
• An energy efficient low cost linear transformer for the benefit of the environment and the end
consumer.
• A new generation of linear power supply transformers replacing the EI-core transformer.
•
nd
The PanPower transformer is expected to go into first mass production during the 2 half of
2006 and will be available for the open market on license basis.

1. MASTER BOBBIN 2. SLAVE BOBBIN

1311
3. SLAVE BOBBIN WITH WINDING 4. INSERTION OF SLAVE BOBBIN
WITH WINDING INTO MASTER BOBBIN

5. INSERTION OF STEEL CORE-BAND 6. THE COMPLETE PANPOWER

INTO THE MASTER BOBBIN CAVITY TRANSFORMER

1312
Contributions and Expectations of Energy Efficiancy Correlated
with Sustainable Developement

Laurentia Predescu1, Adriana Predescu2

1
Romanian Energy Regulatory Authority – ANRE, Romania
2
University Politehnica of Bucharest, Romania

Abstract
The paper intends to present some important strategical actions for energy efficiency improvement,
correlated with the principles of sustainable development. Also, Romania is now a country in the
process of integration in the European Union.
Considering all these, Romania undergoes a process of harmonization of national legislation with the
Community provisions, with a view to fulfill the commitments assumed for accession to European
Union.
The Law 199/2000 has been enacted, regarding the efficient use of energy. This law intends to create
the necessary legal framework for establishing and applying a national policy of efficient use of
energy, according to the provisions of the Treaty of Energy Charter.
As for the energy used in the domestic sector, its evolution was remarkable in the last years. The
domestic customers can choose from a large range of tariffs, unique all over the country.
Economic agents can also choose their favorite tariff, according to their energy behavior, in order to
obtain the lowest average return prices.
Another principle, that suits both energy efficiency improvement and a sustainable development, is to
supply consulting services to customers. The cost of electricity strongly depends on the moment and
type of energy consumption. Customers’ satisfaction is not the only goal. The company also gains
customers’ loyalty, which in the long run leads to a long partnership.
On a competitive energy market, elements like energy efficiency and sustainable development are
major attributes for ensuring the market evolution in a socio-economical environment where all
participants to market can benefit.

Introduction
Romania’s integration in the European Union requires the achievement of significant progress in the
field of efficient energy use. By using energy in an efficient manner, the energy consumptions can be
reduced, thus allowing the use of less primary resources, which leads to an increase of energy supply
security and the support of economical and social development, decreasing the level of financial
efforts related to import of resources.
The basic elements of the process of Romania’s accession to the European Union are the adoption
and transposition of the "acquis communautaire". When adopting the "acquis communautaire", the
following three components were considered:
a. adopting the "acquis communautaire" and drawing the secondary legislation;

b. implementation of national law by creating an institutional framework;

c. performing market monitoring activities.

The approximation of laws is limited to the essential requirements which place the products on the
Community market. These requirements should be met in order to allow the products to benefit from
the right to free movement in the Community market [1].

1. Legal framework
The Romanian energy legislation has been substantially enriched when the Law No. 199/2000 on
efficient energy use, republished, was enacted. The law has in view to create a legal framework that
is necessary in order to draw and apply a national policy for efficient energy use, according to the
provisions of the Energy Charter Treaty, to the Energy Charter Protocol on energy efficiency and
related environmental aspects and to the principles that underlie the sustainable development.

1313
The Law No. 199/2000, republished, defines as a main goal of the national policy of efficient energy
use obtaining the maximum benefit in the whole energy chain of generation, conversion, storage,
transmission, distribution and consumption of different forms of energy. This law has been made in
order to eliminate two types of restrictions noticed in the promotion of energy services
• lack of a legal basis for investments made by energy companies at the end-users, by means
of applying DSM principles; in this regard, the Article 14 (1)(c) sets into a legal context the
information, consulting and financing actions, as well as the execution of works for increasing
the efficient use of fuels and energy in the end-users installations;
• lack of incentives for ESCO-type companies, that would make up for the negative effect of the
low energy prices, which is a reason why private investments in this sector are quite
unattractive; this is why Article 18 introduces a series of fiscal facilities, later cancelled by Law
No. 414/2002.[2]
The Romanian law, in harmonization with the European law, sets the following principles:
ƒ The labels and the card that accompany the electrical home appliances has to state the
energy consumption and the energy efficiency class;
ƒ The producers of electrical home appliances, their authorized representatives and the
importers of such appliances have the responsibility to ensure the appropriate technical
documentation.

The National Strategy for energy efficiency and the action plan related to it, approved by the
Government Decision No. 163/2004, provides opening of financing for national energy efficiency
programs.
The Romanian Energy Regulatory Authority – ANRE – set up a legal framework based on the
respect of the energy efficiency, even if doesn’t have directly competences for improvement energy
efficiency.
ANRE elaborates, establishes and perform control for application of the national obligatory
settlements necessary to work energy market efficiently, competitively, transparently and to assure
protection of the customers.
Also according to legal arrangements, the useful efficiently of the electricity stays on the base to
establish the prices and the tariffs of the electricity and the heat produced through co-generation.
At the same time, EU Directives with directly implications in energy efficiency, were transposed in
Romanian legislation, in principle by the following:
- The Commercial Code of electricity market approved by ANRE Order no. 25/2004, which
establishes fundamental elements, rules and mechanisms which refer to tariffs calculation
and relationships between participants of the market.
- Government Decision 443/2003 regarding to promotion production electricity by renewable
sources.

According to the legal provisions, the Romanian Agency for Energy Conservation -ARCE- is entitled
to monitor the application of energy efficiency legislation, having precise duties regarding information,
checking, control, but also the obligation to warn economic agents and other categories of energy
end-users who don’t follow the specific provisions in this field.
In order to improve energy management in industry, the Ministry for Economy and Commerce has
issued the Order MEC No. 245/2002 regarding the approval of the Regulation on the authorization of
natural and legal persons who are entitled to make energy balances and the Regulation for
testification of persons with energy management responsibilities.
Also, the following guides have been drawn and approved:
ƒ Guide for making and analyzing energy balances - Decision No. 56/28.05.2003 of ARCE;
ƒ Guide for preparing and examination in the field of energy balances, with a view to apply the
Regulation on the authorization of natural and legal persons who are entitled to make energy
balances - Decision No. 57/28.05.2003 of ARCE;
ƒ Guide for preparing and examination in the field of energy management, with a view to apply
the Regulation on the authorization of persons with responsibilities in the field of energy
management - Decision No. 58/16.05.2003 of ARCE.

2. Implementation of national law in order to create the institutional framework

With a view to apply the provisions of the proposal of the Directive of the European Parliament and
Council on promotion of energy efficiency at end-users and of energy services, a very important role
is held by the power supply and distribution companies. The following are taken into account:

1314
 • increase of participation of energy companies, without the obligation to supply energy
services themselves or to become ESCOs; the proposal provides acceptable forms of
participation, like sharing the responsibilities, outsourcing, etc;
• actions of energy suppliers and distributors, in order to help the customers to shift their
energy demand towards energy services and to make a step ahead from the simple sale of
kWh;
• directing the national and regional energy efficiency funds to the power companies and to
other suppliers of energy services; opening competition.

The Law No. 199/2000, republished, includes, in Chapter VI, under the title “Obligations for energy
consumers”, the following general provisions, which we quote here in order to emphasize some
particular aspects:
- according to Article 12 (1): „Energy consumers, legal persons, have the obligation to observe
the technical regulations in force regarding the design, manufacture, operation, maintenance,
repair of own installations and of energy receivers, as well as providing them with
measurement and control devices.”
- according to Article 12 (2): „Energy consumers, legal persons, have the obligation to have
their own energy consumption recording and monitoring system and to make available for
authorized institutions all information regarding energy consumptions and energy efficiency
indicators.”
- in the industrial energy sector, energy balances (for both electricity and fuel) are made; these
regard industrial assemblies, machines, sections, enterprises, and highlight the methods to
decrease the energy consumptions. By applying the suggested measures, important energy
savings are achieved.
The Law No. 199/2000, republished, includes specific provisions for economic agents with activities
of fuel and energy generation, transmission or distribution, as well as for the promotion of using solar,
wind, geothermal, biomass and household waste energy [Fig.1].

After agreement with consumers, the producers and suppliers of electricity and thermal energy can
engage in activities of information, consulting, financing, as well as execution of works with a view to
increase the efficiency of use of fuels and energy in their installations.
These general provisions represent the basic elements needed to ensure the security and continuity
of operation of installations, machines and equipments in the energy sector. The design, engineering
and operation norms for all components and functional assemblies are strictly regulated.

Solar termic
12,1% Solar fotovoltaic
Solar termic
Geotermal 0,8%
1,4% Eolian Solar fotovoltaic
16,9% Eolian
Hidro
Microhidro
Biomasă
Geotermal

Microhidro
Biomasă Hidro
0,1%
64,3% 4,3%

Figure 1: Annual potential of renewable energy sources [10].

3. Efficient energy management at end-users, as specific form of energy

service

Demand side management represents a service dedicated by energy supply and distribution
companies to their end-users, with a view to make energy savings, as al alternative to capacity
modernization or replacement, in the case of development of power distribution networks. The energy
services have to be considered by the suppliers as a possibility to modify the old energy sales model,
by shifting from a product sale to a service sale.

1315
As energy efficiency policy started to be considered of public interest, the key of success of the DSM
policy, especially in the context of liberalization of energy market, is the political perception on its
legitimacy. A way of proving this legitimacy is to analyze the distributive impact (who benefits from the
DSM program, who and how much will have to pay for this benefit, what costs will be transferred to
the electricity companies). Either it’s reflected in a grant from the state, or in regulations for DSM
promotion, it is necessary to justify the legitimacy of the support of public interest. [3]
In this context, the role of those who make the energy policy is to prove to the public, to the industry,
to politicians, the fact that it is possible to achieve important energy savings, that energy savings are
competitive in terms of costs with the conventional sources and they can be distributed in a fair
manner. The energy efficiency indicators are the quantitative expression of the following strategic
objectives:
• monitoring of targets referring to energy efficiency and to the programs of CO2 level reduction
at national and international level;
• evaluation of energy efficiency policies and programs;
• planning of future actions, including research & development programs;
• setting the basis for energy strategies;
• comparing the performances achieved on an international scale.

The increase of activities efficiency represents a strategic option for any company. In general context,
efficiency can be appreciated by means of global indicators, which mainly refer to the economic
aspects of the company (specific costs, specific revenues), or to aspects regarding company
marketing ( quality of service or product, company’s market share ).
In the same time, energy efficiency improvement can be appreciated by means of evolution energy
intensity [Fig.2] [11].
Awareness and education of suppliers and customers in terms of energy efficiency and the relation to
quality of electricity represent important options in the process of making efficient the activities,
knowing that these elements motivate the staff and improve the company’s performance.
The theme is approached from both sides: electricity consumer and power supplier.

T h e e v o lu t io n o f e n e r g y
in t e n s it y

0 ,2 5
tep/1000 USD95

0 ,2

0 ,1 5

0 ,1

0 ,0 5

1995 1996 1997 1998 1999 2000

- R o m â n ia - U .E .

Figure 2: The evolution of energy intensity in Romania and the EU.

4. Market control activities

Market control represents an instrument for competent authorities, allowing them to implement the
regulatory actions that adopt European directives. For this, the competent authorities assign
structures responsible for market monitoring and control whether:
ƒ the products are according to the provisions of the national law that applies to them;
ƒ actions are taken to bring the products from nonconformity to conformity;
ƒ fines are applied when the products prove not to be according to the essential requirements
set in the regulatory actions.
In this regard, the Ministry of Economy and Commerce, through the Romanian Agency for Energy
Conservation, applying the government decisions and in collaboration with ANPC (the National
Authority for Customer Protection), monitor and control the observance of provisions of technical
regulations on energy efficiency labeling for electrical home appliances (refrigerators, washing

1316
machines, combined washer/dryers, lamps, dryers with cylindrical drum, dish washers, electrical
ovens, ballasts for fluorescent lamps, air-conditioning appliances.
The labeling of electrical home appliances is made according to their energy efficiency, which is
determined on the base of the power consumption of appliances, measured in the conditions set by
the Romanian norms (harmonized with the European ones). These standards establish methods for
measuring the performances of electrical home appliances, that apply to each category of appliances.
The activity of market monitoring has three main stages [1]:
ƒ dissemination of information until the harmonized legislation entry into force;
ƒ monitoring the conformity of products that enter the market with the requirements of national
law that applies to them;
ƒ taking measures to ensure conformity, if applicable.
In order to monitor the products that enter the market, in an efficient manner, the authorities
responsible for market control have the authority, capacity and resources to:
ƒ visit the production sites and the places where the products are stored or sold, as appropriate;
ƒ take samples and examine/test them;
ƒ ask for any relevant information regarding the product.
The corrective actions depend on the degree of nonconformity, which is set on a case-by-case basis.
The producer or the person responsible for introducing the products on the market has to take steps
to solve nonconformities and to make the product according to the requirements that apply to it.

Conclusions

Romania has to harmonize its energy policy to the one of the European Union. Special attention has
to be given to security, efficiency, environmental protection and consumer protection, using the
mechanisms of competition and a balanced statement regarding energy efficiency. In this regard, the
home appliances have an important role, because they are used on a wide scale.
In a society who wishes to implement the concept of sustainable development, considerable efforts
have to be made not only for obtaining stable energy resources, as well as for increasing the
efficiency of the processes that use these resources.
The energy efficiency of home appliances is a key action of the policy of implementing the Kyoto
Protocol, as well as a subject of many directives and programs. All the countries that adopted the EU
regulations on energy efficiency have lower energy consumptions and reduced their CO2 emissions.
Indeed, the energy efficiency improvement in all stages of conversion of primary energy into useful
energy leads both to the a lower social-economical impact and to a lower impact on environment
associated to the activities in the human society. A great number of economical and political sectors
contribute to the implementation of these actions, from those who bring to best value the primary
energy sources to the political decision factors.
The main characteristic for the present situation of Romanian energy sector is the fact that energy
saving is the cheapest energy resource available, in the context of integrated resource planning.

References
[1] Calugar,C., Transpunerea si aplicarea acquis-ului comunitar in domeniul eficientei energetice –
supraveghere piata. „Simpozionul International de eficienta energetica”, Cluj, 19-21oct. 2004,
pp.52-56. ISBN 973-8329-24-8.
[2] Voronca, M., Rotaru, C., Cruceru, M. Gestiunea eficienta a energiei la consumatorii finali.
„Simpozionul International de eficienta energetica”, Cluj, 19-21oct. 2004, pp. 04-10.ISBN 973-
8329-24-8.
[3] Rotaru,C., Barsan, S. Aspecte privind promovarea serviciilor energetice tip ESCO in Europa si
SUA. Revista Energetica, 2001.
[4] Law No. 199/2000 on efficient energy use , republished.
[5] Law No. 14 / 1997 , pentru ratificatrea Tratatului Cartei Energiei si a Protocolului Cartei Energiei
privind eficienta energetica si aspecte legate de mediu.
[6] Programul National de Aderare a Romaniei la Uniunea Europeana – PNAR.
[7] Directive 2003/54/EC of the European Parliament and of the Council of 26 June 2003 concerning
common rules for internal market in electricity.
[8] Energy Efficiency Indicators for Eastern and Central Europe – Studiu ADEME 2003
[9] Standardul de performanta pentru serviciul de furnizare a energiei electrice – Decizia ANRE nr.
34/1999.

1317
[10]Birsan,S., Aplicarea prevederilor Legii nr.199/2000 privind utilizarea eficienta a energiei,
republicata, in sectorul energiei, Masa Rotunda, „Simpozionul International de eficienta
energetica”, Cluj, 19-21oct. 2004, pp. 183-189.ISBN 973-8329-24-8.
[11] Energy Balances of non-OECD Countries, International Energy Agency, 1999-2002 Edition,
Paris, France, 2002.

1318
Decision Support Model for Energy Companies’ Operational
Environment in the EU New & Candidate Member States
Patlitzianas, D. K, Doukas, H., Papadopoulou G.A, Psarras, J.

Management & Decision Support Systems Laboratory, Department of

Electrical and Computer Engineering, National Technical University of Athens

Abstract
One of the current energy policy’s goals in the European Union (EU) is the exchange of knowledge
and experience between the EU-15 member states and the new as well as the candidate countries in
issues regarding the development of Renewable Energy Sources (RES) and the promotion of Energy
Efficiency (EE). Nowadays, the liberalization of energy markets and the deterioration of the climate, in
combination with the non-stop crude oil price increase, have had a decisive influence on the
development of the above sectors. Indeed, one of the most important elements for the RES and EE
development is the enhancement of energy producers by RES and ESCOs respectively. These
companies’ success is based on the formulation of a modern environment in each EU member state.
However, the environment of the new and candidate member states is less mature than the
environment of the EU-15 member states, as it is still in its development phase. In this context, the
main aim of this paper is to present a “multidimensional” decision support model for the formulation of
modern energy companies’ operational environment, which also incorporates the “new parameters”
that enter the energy market, namely the liberalization and the climate change. This model is used so
as to assess the environment of the energy companies in the fourteen (14) new and candidate
member states of the EU.

Introduction
The development potential of the Renewable Energy Sources (RES) in the new and candidate
member states of the European Union (EU) is high [1]. As a result, the expectations for a significant
increase of RES contribution to the primary energy supply reach 26% in the year 2030 [2], from 10%
in the year 2003 [3] in the overall EU. Moreover, the increase of energy consumption, as well as the
CO2 emissions in these countries is an inevitable outcome of social and economic development. As a
result, it is clear that Energy Efficiency (EE) can be improved in these member states.
Nowadays, the European Commission (EC) aims to facilitate the exchange of knowledge and
experience between the EU-15 member states and the new as well as the candidate countries in order
to enhance RES development and promotion of EE.
The developing countries, which can be benefited from the experience and knowledge of the most
developed ones, are constituted of the ten member states (Cyprus-CY, Czech Republic-CZ, Estonia-
EE, Hungary-HU, Lithuania-LT, Latvia-LV, Malta-MT, Poland-PL, Slovenia-SL, Slovakia-SK), which
joined the European Union (EU) in 2004, Bulgaria-BG and Romania-RO, that are going to join the
European Union in 2007, as well as Turkey-TR and Croatia-HR that have recently started negotiations
with EC.
One of the most important parameters for the RES development is the enhancement of involved
producers. These producers can be either companies deriving from utilities producing energy from
conventional sources that have decided to be activated in the field of RES or Independent Power
Producers (IPPs). These companies can be either newly entering companies or companies already
engaged in the construction and trade of renewable energy technical equipment that have decided to
enter the market as IPPs. In addition to this, Energy Service Companies (ESCOs) have been
developed and their role is crucial for the promotion of EE.
The success of the above energy companies is based on the formulation of a modern environment in
each EU member state. Generally, Chandler [4] outlines that the operational environment of companies
determines substantially the main long-lasting objectives and aims of each company, fires a line of
action and determines the necessary means for the realization of these objectives. Johnson and
Scholes [5] note that the environment directs decisively the activities of a company in the long run. In
addition, Ansoff [6] supports that the existence of the companies’ operational environment is the base
of creation of common lines between the activities of a company.

1319
In this context, each member state needs to formulate an up-to-date energy company’s environment,
which has to be enhanced, giving thus the opportunity to more companies in these member states to
be properly activated.
However, the environment of these member states is less mature than the environment of the EU-15
member states and it is still in a developing phase, due to the lower social acceptance, the public
awareness, the fact that the Kyoto Protocol is not top priority yet as well as the absence of appropriate
national financial sources.
Based on the international literature, a large body of literature examines the external factors of the
energy companies – in terms of policies, regulations and financing support schemes of these states.
For instance, Bechberger in 2003 [7] and Patlitzianas et al. in 2004 [8] presented general renewable
energy overviews of candidate countries. In 2005, Streimikiene [9] described the RES and EE
development in the Baltic States and Reiche [10] presented an investigation of the driving forces for a
further promotion of renewable energies in the accession states in 2006. However, these papers do
not take into consideration the recent developments, regarding the EC’s accession negotiations with
Turkey and Croatia. In addition to this, there are no papers investigating the operational environment
of energy companies in an integrated way as presented above.
The main aim of this paper is to present a “multidimensional” decision making model for the
formulation of modern energy companies’ operational environment, which also incorporates the “new
parameters” that are introduced into the energy market, namely the liberalization [11] and the climate
change [12]. This model is used so as to assess the environment of the energy companies in the
fourteen (14) new and candidate member states of the EU.
In this context, the paper is structured in four sections as follows:
y After an introduction, the second section provides a short description of the adopted model.
y In the third section this model is applied in assessing the energy companies’ environment in
the EU new and candidate member states. In particular, this section describes the inputs,
including concrete facts and figures obtained in the data collection process and the outputs of
the procedure as well as their discussion.
y The last section presents the conclusions, which summarize the main points that have been
brought up in this paper and outlines perspectives for the development of the companies’
environment in the region.

Brief Description of the Model

The description of the energy companies’ operational environment is the aim of the model
(I.M.E.C.O.) that is constituted by the following five components:
Component 1 – Identification
This component concerns the identification, based on the experience, of seventeen actions, in each
one of the environment’s dimensions. In particular, the dimensions (Di) of the environment can be
categorized in four dimensions, the political/ legal, financial, social/ cultural and technological (i=1,2,3
and 4) taking into consideration the literature that is related to the company’s environment and its
strategy [13-16].
Based on the relative literature review [17-21], the necessary actions towards the formulation of a
modern energy companies’ operational environment (Aij) are categorized and illustrated in the
following Table 1.

1320
Table 1: The Actions
Dimension Actions
D1: Political - Legal A1.1 Legislation on the support of energy production from RES
A1.2 Verification system for energy service companies
A1.3 Standardisation of energy services contracts
A1.4 Political support of RES-EE
A1.5 Political promotion of international energy cooperation
D2: Financial A2.1 Economical support of RES project
A2.2 Economical support of energy management projects
A2.3 Economical support of EE projects
A2.4 Promotion / Support of new financing sources
D3: Social – Cultural A3.1 Employment for RES-EE fields
A3.2 Social acceptance for the RES-EE fields
A3.3 Education support actions for RES – EE
A3.4 Development of new energy companies in the region
D4: Technological A4.1 Support actions of R & D on the energy sector
A4.2 Support actions of R & D on new innovative technologies of energy production
by RES
A4.3 Support actions of R & D on new energy efficiency’s technologies
A4.4 Support actions of the commercial exploitation of the research results

Component 2 – Modeling
The second component concerns the modeling of the energy companies’ operational environment, via
the development of a group of appropriate indicators. The development of this component is based on
the use of existing indicators, identified after a detailed literature review. In this framework all the past
efforts made for the development of energy indicators in the RES and EE sectors from EC, OECD,
IEA and APERC [22-29] were detected. In any case, the aim is not the development of new indicators
but the use of the already existing ones, in order to measure on a common basis the necessity of an
intervention.
The selected indicators belong to two basic categories: qualitative and quantitative indicators. In
particular:
y Seventeen (17) Basic Indicators (Bij, i- 1,2,3 and 4, j=1,2,…,a) were selected based on the
appropriate literature survey. These indicators present the most essential information
regarding the diagnosis of the country’s performance, in terms of the energy companies’
environment. The basic indicator is the key means of decision making for the necessity of
taking intervention measures or not.
y Some other indicators create a new pool that includes the Secondary Indicators Sk,
(k=1,2,…,m). The secondary indicators act accessory to the estimation of the weaknesses of
the energy companies’ environment. These indicators are focused on specified issues of the
weaknesses of the energy companies’ environment and describe specific activities for
selected aspects of the sectors they examine. The selected secondary indicators are forty
one (41).
y A third pool of indicators is created, representing the effects that the “New Parameters” of the
companies’ market involve in the decisions of their operational environment’s formulation Nl,
(l=1,2,…,n). The new parameters’ selected indicators are twenty-two (22).
Component 3 – Estimation
The third component concerns the estimation of the necessity for each Action (Aij) of the companies’
environment. Concretely, the Basic Indicator is related to a group of Secondary Indicators. Moreover, a
second group of indicators, reflecting the impact of “new parameters”, is related to each one Basic
Indicator. As a result, there are selected:
y The pool of the correlated indicators of “New Parameters” (Nilx, x=1,2,3,…,b).
y The pool of the correlated Secondary Indicators (Sijy, y=1,2,3,…,c).

In this way, a "pool of decision indicators" is being created, the price control of which portrays the
estimation of the “existence of the necessity or not” for improving the companies’ environment.
Component 4 – Choice
After the estimation of the action’s necessity, the model investigates the intervention choices, based on
the evolution indicator’s DBij values. This indicator illustrates the evolution of the Basic Indicator’s
performance during the past year. The value of the above mentioned indicator is estimated according
to appropriate thresholds and the existence or not of appropriate measures in the last year is

1321
examined. In this context, the continuation of the existing measures (I), their modification (II) or the
formulation of new measures (III) is proposed.
Component 5 – Order
The last component receives as input the results of the previous components, in order to evaluate the
direct actions to be done for the development of energy companies’ environment in each country and
involve a methodology of quantifying multiple qualitative judgments based on the multicriteria decision
making method (Ordered Weighted Average) [17].
The six criteria are selected so as to incorporate all the needs of the companies’ operational
environment as well as the emerging needs and opportunities of the “new parameters”, which
determine the final decision. In addition to this, the member states’ performance to each one of the
criteria is based on a 1-5 order qualitative scale, with “1” illustrating an insignificant progress of the
country regarding the particular criterion, “2” a low, “3” a moderate, “4” a high and “5” a very high
progress of the member state regarding the particular criterion. The criteria are presented in the Table
2:

Table 2: The Criteria

Category Priority Criteria
C1: Contribution to the increase of the RES proportion
Basic C2: Contribution to the increase of the EE
Needs C3: Contribution to the security of supply
C4: Contribution to the sustainable development
New C5: Progress regarding the liberalization of the energy market
Parameters C6: Contribution to the reduction of the greenhouse gases

The algorithm of the model appears in the following Figure 1.

In
Dimension Di
Relation

Action Aij
Actions’ Base

Basic Indicators Bij

(Bi1, Bi2,..., Bia)

Total Indicators
Yes Values’
of New Deviation Βij
Parameters N
Thresholds

No

Total Secondary
Indicators S
Related Indicators Nijx
EXPERIENCE

Progress
Indicators Yes Values’
DΒij
Deviation Νijx
No

Related Indicators Sijy

No
Values’ Deviation
Sijy
Yes

Estimation of Action’s Aij Necessity

Bad Good
Check DΒij

Yes Measures for No No Yes

Measures for
Action’s Implementation
Action’s Implementation

Modification of Formulation of Continuation of

Existing Measures New Measures Existing Measures

Yes
Next Aij
No
Yes
Next Di
No
MULTI CRITERIA

List of Actions
ANALYSIS

Priority
Criteria

Priority List of Actions

Out
Figure 1: The Algorithm of Model

1322
Case Study in the EU New and Candidate Countries
Inputs
The inputs are firstly based on the results of a project funded by the ALTENER programme of the EC
[18]. In addition to this, the collected information was enhanced and updated, through the
implemented events and the initial outputs of the on-going FP6 project, funded by EC [19]. Finally, the
related - implemented events in the region [20] as well as the reports written by EREC (European
Renewable Energy Council) [1] and other relevant sources [21-25] were taken into consideration.
Based on the above sources, some indicative data obtained in the process of data collection for the
indicators used are illustrated in the following Figures.
Based on Figure 2, it is clearly illustrated that in most of the examined countries RES primary
production is mainly based on biomass and on a secondary basis on hydro. Even though
photovoltaics (PVs) and wind options are already considered to be mature enough, their penetration
in the energy market of these countries remains very limited. In addition to this, countries which have
a significant RES primary production due to their resources are Turkey, Poland and Romania.
12000

10000
Hydroelectricity Biomass
Wind PV
RES in Primary Production (kTOE)

8000
Geothermal

6000

4000

2000

0
BG HR HU LT LV MT CZ EE CY PL RO SK SL TR

Figure 2: RES in Primary Production – 2003

In Figure 3 it is clearly depicted that there are a number of countries like Turkey, Poland and
Romania, where RES production corresponds to thousands of ktoe annually. However, this doesn’t
necessarily means that in these same countries the RES percentage in GIC is particularly high, as it
doesn’t exceed ~13%. On the contrary, a number of countries with exceptionally high percentage of
RES contribution in GIC are Latvia, Slovenia, Lithuania and Estonia.

1323
 12000 RES in GIC % RES in GIC 40

35
10000

30

8000
25
(KTOE)

% RES
6000 20

15
4000

10

2000
5

0 0
BG HR HU LT LV MT CZ EE CY PL RO SK SL TR

Figure 3: RES in Gross Inland Consumption – 2003

Moreover, from the following Figure 4 can be outlined that RES percentage in the electricity
generation remains limited in most countries, with the exceptions of Hungary and Latvia that possess
significant biomass and geothermal production respectively.
GWh %
40000 70

35000
60

30000
50

25000

40
GWh

%
20000

30
15000

20
10000

5000
10

0 0
BG HR HU LT LV MT CZ EE CY PL RO SK SL TR

Figure 4: RES in Electricity Generation – 2003

Based on Figure 5 it can be easily concluded that the countries with the greatest penetration of RES
in GIC per 1.000 persons are the ones with the lowest population, like Slovenia, Latvia and Estonia.

1324
 Population RES in GIC/ Population
80000 0,7

70000
0,6

60000
0,5
Population (1.000 per.)

50000

KTIP/1.000 per
0,4

40000

0,3

30000

0,2
20000

0,1
10000

0 0
BG HR HU LT LV MT CZ EE CY PL RO SK SL TR

Figure 5: RES per population – 2003

In addition to this, the increase of energy consumption as well as the CO2 emissions in the majority of
these countries as an inevitable outcome of social and economic development is clearly illustrated in
the Figures 6 and 7. As a result, it is clear that EE could be improved in these member states.
2 Energy Intensity Electricity Intensity 0,25

1,8

1,6 0,2

1,4

1,2 0,15
KTOE / b. Euro

KTOE / b. Euro

0,8 0,1

0,6

0,4 0,05

0,2

0 0
BG EE CY LV LT MT HR HU PL RO SK SL TR CZ

Figure 6: Energy Intensity – 2003

1325
 3

2,5

2
Ktn/bil Euro

1,5

0,5

0
BG EE CY LV LT MT HR HU PL RO SK SL TR CZ

Figure 7: CO2 Intensity (ktn CO2 / billions Euro)

Moreover, there is a great number of countries that will have difficulties in meeting their Kyoto target
for 2010. In this framework the countries that will probably not achieve their goal are Estonia, Latvia,
Lithuania, Slovakia and the Czech Republic. On the other hand, countries that will easily meet their
target are Croatia and Slovenia, while countries that will meet their target with difficulties, as in 2003
they were very far behind, are Bulgaria and Hungary.

2010 (Projection) 2003

20

10

0
BG EE CY LV LT MT HR HU PL RO SK SL TR CZ

-10

-20
%

-30

-40

-50

-60

-70

-80

Figure 8: Distance from the Kyoto Protocol Target

In addition to this, a significant amount of money has recently been devoted to the R & D in the RES
and EE sectors in some of the examined countries, such as Bulgaria, Hungary and Romania.
However, as depicted in Figure 9, these countries are still lagging behind from the EU-15 member
states in terms of technology progress and innovation production, which is also obvious taking into
consideration the lack of domestic manufactures of RES and energy conservation equipment.

1326
 0,70
RES
EE
0,60

0,50
Euros / thousand euros

0,40

0,30

0,20

0,10

0,00
BG EE CY LV LT MT HR HU PL RO SK SL TR CZ

Figure 9: R & D in RES & EE - 2003

Outputs and Discussion

The results of the proposed model’s application in the 10 new and 4 candidate member states are
presented in the following Table 3. In particular, the actions examined are ranked, from the most
necessary to the ones of lowest priority.

Table 3: The Outputs

Member Actions
State
BG A3.2 (I) >A3.3 (III) > A1.1 (II) > A3.4 (II) > A2.1 (I) > A1.5 (I) > A2.2 (I) > A1.3 (I) > A2.3 (II) > A4.4 (III) > A4.1 (I)
> A4.3 (II)
CY A2.4 (III) >A4.4 (III) > A1.2 (I) > A4.3 (II) > A1.3 (III) > A3.3 (I) > A4.2 (I) > A1.5 (I) > A2.1 (I) > A4.1 (II)
CZ A2.2 (II) >A3.4 (III) > A2.1 (I) > A4.2 (I) > A2.4 (I) > A4.3 (II) > A2.3 (II) > A1.4 (I) > A1.1 (I)
EE A1.4 (III) > A2.1 (II) > A3.4 (III) > A4.3 (I) > A2.2 (II) > A2.3 (I) > A2.4 (I) > A4.1 (I) > A3.2 (II) > A1.2 (I)
HR A1.1 (I) > A2.1 (I) > A1.4 (III) > A1.5 (I) > A2.3 (II) > A2.4 (III) > A1.2 (I) > A1.3 (III) > A4.3 (III) > A4.2 (III) > A4.1
(I) > A3.2 (I) > A3.1 (III) > A3.4 (II)
HU A3.1 (I) > A4.2 (I) > A3.4 (III) > A4.4 (II) > A2.1 (I) > A2.3 (III) > A1.5 (II)
LT A2.3 (I) > A2.2 (I) > A2.4 (III) > A1.4 (III) > A1.3 (III) > A3.4 (III) > A4.3 (III) > A2.1 (I) > A1.5 (I) > A3.2 (II) > A3.1
(I)
LV A3.3 (I) >A1.3 (III) > A3.1 (I) > A2.1 (I) > A2.2 (I) > A4.1 (II) > A3.4 (I) > A4.3 (II) > A4.4 (III) > A1.5 (I) > A1.1 (I) >
A2.3 (I) > A2.4 (III) > A3.2 (I)
MT A1.1 (I) >A4.1 (II) > A4.2 (II) > A1.4 (II) > A1.5 (I) > A2.1 (II) > A2.3 (II) > A1.2 (I) > A1.3 (III) > A3.2 (I) > A3.3
(III) > A4.3 (II)
PL A4.2 (I) >A4.3 (II) > A4.1 (I) > A2.1 (I) > A3.2 (III) > A3.4 (III) > A1.4 (III) > A1.3 (III) > A2.3 (I)
RO A1.1 (II) >A1.4 (III) > A1.3 (II) > A3.1 (II) > A3.3 (III) > A3.4 (III) > A2.1 (II) > A2.3 (I) > A2.2 (I) > A3.2 (III) > A4.4
(II) > A2.4 (II) > A1.2 (I) > A1.5 (I)
SK A3.1 (III) > A3.3 (III) > A4.3 (III) > A4.4 (III) > A3.4 (III) > A2.1 (I) > A2.3 (II) > A2.4 (III) > A2.2 (I) > A1.4 (III) >
A1.1 (I)
SL A4.1 (I) >A4.2 (II) > A4.3 (I) > A1.4 (III) > A1.3 (III) > A4.4 (III) > A2.3 (I) > A2.1 (I) > A3.2 (II) > A3.4 (II) > A3.1
(II) > A1.2 (I)
TR A1.1 (II) > A1.4 (III) > A1.5 (I) > A1.2 (II) > A2.1 (II) > A3.4 (III) > A1.3 (III) > A4.2 (II) > A3.1 (II) > A3.2 (III) >
A2.3 (I) > A3.3 (III) > A4.1 (II) > A4.3 (II).

The main points that can be drawn up from the results of the current model’s application are the
following:
y Political - Legal Dimension: In most countries the necessary actions to be regarded as
medium priority concern the political support for the RES and EE sectors as well as the
promotion of energy co-operations. Few countries have to emphasize on the enhancement of
the energy companies’ certification system and the standardization of energy services’
contracts, such as Croatia, Malta, Slovenia and Turkey. Countries which seem to have made
adequate efforts in this dimension of the companies’ environment are Hungary and Poland.
y Financial Dimension: The insufficiency of financial resources for the support of RES, as well
as of EE, is particularly obvious, even if the percentage of RES in the accession members’

1327
 electricity market was above the average of the EU-15 members. In most countries, energy
companies’ environment development is lagging behind in terms of its financial dimension. In
particular, the necessity for subsidies for appropriate projects and energy management
activities is evident in the majority of the countries. Moreover, all countries need to promote
modern financing sources, apart from Cyprus and Turkey, countries in which related activities
have already occurred.
y Social - Cultural Dimension: Actions of this dimension are considered to be of high priority in
Estonia, Hungary, Poland Latvia and Bulgaria. Drawbacks, such as low number of specialised
employees, limited amplification of employment, lack of educational policy and limited
activities for promoting the benefits of RES - energy conservation, create problems of
establishment and acceptance of the energy companies’ concept.
In this context, environmental awareness is developed only in a low level in the EU accession
member states and therefore the related actions can be considered to be of significant
importance, with the exception of Cyprus and Latvia. In addition, in all countries, the necessity
for the related employment support was derived. Furthermore, in Bulgaria, Czech Republic,
Estonia, Hungary, Malta, Poland, Slovenia and Slovakia high priority has to be given in
actions concerning the growth of regional new energy companies’ development. Moreover,
the need to support the education for RES – EE appeared for all countries besides Bulgaria,
Croatia, Poland and Slovenia.
y Technological Dimension: The support of R& D in the energy sector generally seems to be an
action recommended to the majority of the countries, besides Hungary. Especially, necessity
for actions of the R& D support for the new RES innovative technologies is presented in all
countries, apart from Bulgaria and Romania, where significant efforts have been implemented
recently. In addition to this, the need for R& D support of the new EE innovative technologies
in the rest of the countries is evident. This is realistic since, in most accession countries, a
lack of domestic manufactures exists, apart from some specific cases such as Latvia,
Slovenia and Turkey (industries for boilers’ equipment and water turbines). Moreover, the
manufacture of solar collectors is rapidly increased in Cyprus as well as Hungary, Slovakia
and Poland. Finally, the necessity for supporting the commercial exploitation of the research
results for the RES-EE sectors exists in all countries. On the other hand, technology for
specific RES technologies (e.g. hydroelectric energy and biomass) for heating reasons is
constantly developing and already has been used in many of the accession states. With
regard to geothermal energy, Hungary, Turkey and Poland show a relatively satisfactory
technological growth.

Conclusions
After taken into account the outcomes, the key points are presented in the following way:
y Their political/legal environment has not developed yet, compared with the corresponding
environment of the 15 member states but specifically in some cases (such as Slovenia,
Lithuania, Latvia, Estonia, Hungary, Poland and Czech Republic) a significant progress has
been observed. In addition to this, there is a great number of countries that will have
difficulties in meeting their Kyoto target for 2010.
y These states have had a century-long tradition in the utilization of renewable energy, primarily
in biomass and hydropower. RES percentage in the electricity generation remains limited in
most countries, with the exceptions of Hungary and Latvia that possess significant biomass
and geothermal production respectively. As a result, the financial environment of most
accession member states should be enhanced.
y The social/cultural brings a different outcome to one’s attention, especially in the middle
places. Cyprus ranks relatively high, showing their high social RES awareness in that way.
Nevertheless, it can be concluded that countries (Latvia, Estonia, Poland, Lithuania and
Hungary) which are active and evolving carry the crowds along with them. A significant
amount of money has recently been devoted to the R & D in the RES and EE sectors in some
of the examined countries, such as Bulgaria, Hungary and Romania.
y In the technological environment, Poland and Latvia surpass the rest, due to their traditional
technological know-how and an already developed “heavy” industry, followed in a close
distance by Estonia, Hungary, Lithuania and Czech Republic.
As a result, decision making support models, such as the one presented in this paper, are needed to
identify, diagnose, and order the appropriate actions in a consistent way, as well as to assist policy

1328
making and formulate a modern energy companies’ operational environment. In addition, the model’s
procedure assisted the specific decision making problem and the outcome might have been quite
different if different indicators and criteria had been chosen. However, the model’s conceptual can
provide a sufficient framework for supporting other decision making problems.

References
[1] European Commission, Altener Programme, European Renewable Energy Council (EREC).
Renewable Energy in Europe: Building Capacity and Markets. James and James (Science
publishers) Ltd, 2003. Can be downloaded at: http://www.erec-renewables.org
[2] European Commission. European Energy and Transport Trends to 2030, Luxemburg: Office for
Official Publications of the European Communities, 2003. ISBN 92-894-4444-4. Can be
downloaded at: http://www.europa.eu.int/
[3] Eurostat. Energy Balance Sheets data 2002-2003, Luxembourg: Office for Official Publications
of the European Communities, 2005. ISBN 92-894-9786-6. Can be downloaded at:
http://epp.eurostat.cec.eu.int/
[4] Chandler A. A Strategy and Structure: Chapters in the History of the American Industrial
Enterprise, Massachusetts Institute of Technology, 1962. ISBN 0-262-53009-0.
[5] Johnson G. and Scholes K. Exploring Corporate Strategy: Text and Cases. London. Prentice
Hall Europe. fifth edition, 1999. ISBN 0130807400.
[6] Ansoff I. The Concept of Corporate Strategy. London Penguin, 1985.
[7] Bechberger M. And Reiche D. RE in EU-28. Renewable energy policies in an enlarged
European Union. Refocus 2003; 30–34.
[8] Patlitzianas KD. Kagiannas AG. Askounis DT. and Psarras J. The policy perspective for RES
development in the new member states of the EU. Renewable Energy 2005; 30:477–92.
[9] Streimikiene D. and Klevas V. Promotion of renewable energy in Baltic States. 2005.
Renewable and Sustainable Energy Reviews (in press).
[10] Reiche D. Renewable energies in EU-Accession States. Energy Policy 2006; 34:365-75.
[11] Meyer NI. European schemes for promoting renewables in liberalized markets. Energy Policy
2003. 31: 665–76.
[12] Pablo R, Hernandez F, Gual M. The implications of the Kyoto project mechanisms for the
deployment of renewable electricity in Europe. Energy Policy 2005; 33:2010-22.
[13] Porter M. What is Strategy?. Harvard Business Review 1996; 74(6):61–78.
[14] Kenneth R.Α. The Concept of Corporate Strategy. The Strategy Process. Englewood Cliffs:
Prentice Hall. 43-5, 1996.
nd
[15] Coulter M. Strategic Management in Action. 2 ed. Upper Saddle River. NJ: Prentice Hall,
2002. ISBN 0536675880
[16] Lynch R. Corporate Strategy. 3rd ed. London: Pitman, 2003.
[17] Bertoldi P. Rezessy S. and Vine Ed. Energy service companies in European countries: Current
status and a strategy to foster their development. 2005. Energy Policy in press.
[18] Vine Ed. An international survey of the energy service company (ESCO) industry. Energy
Policy 2005; 33(5): 691-704.
[19] KD. Patlitzianas, H. Doukas, and J. Psarras. Designing an Appropriate ESCOs’ Environment in
the Mediterranean. 2005. Submitted for publication in the Management of Environmental
Quality-Emerald, (in press).
[20] Patlitzianas K, Doukas H, and Psarras J. Enhancing Renewable Energy in the Arab States of
the Gulf: Constraints and Efforts. 2005. Energy Policy, in press.
[21] Patlitzianas K, Doukas H, Papadopoulou A, Flamos A, and Psarras J. Τhe energy reform and
climate policies in developing “renewable energy” and “energy services” companies. EightH
International Conference on energy for a clean environment, Lisbon Portugal, 2005.
[22] Organisation for Economic Co-operation and Development. OECD core set of indicators for
environmental performance reviews. OECD Environment Monographs No. 83, OECD, Paris,
France, 1993.
[23] European Commission. Integration – Indicators for Energy. Luxembourg: Office for Official
Publications of the European Communities, 1999
[24] European Commission, ODYSSEE. MURE database on measures for the rational use of
energy. 2005

1329
[25] OXERA - European Commission DG TREN. Electricity Liberalisation Indicators in Europe.
Netherlands Energy Research Foundation, Energy System Analysis and Planning (ESAP),
Centre ATOM, Oxford, England, 2001
[26] Division of Sustainable Development, United Nations Department of Economic and Social
Affairs. Indicators of Sustainable Development - Framework and Methodologies, Background
Paper No.3, United Nations, New York, USA. 2001
[27] Unander F., Schipper L., Marie-Lilliu C. The IEA Energy Indicators - Analysing Emissions on
the Road from Kyoto. International Energy Agency – IEA, Cop 6 the Hague, 13-24 Nov. 2000,
Paris, France.
[28] Asia Pacific Energy Research Center- APERC. Energy Efficiency Indicators – A Study of
Energy Efficiency Indicators for Industry in APEC Economies, Institute of Energy Economics,
Tokio, Japan. 2000
[29] World Energy Council – WEC. Energy Efficiency Policies and Indicators. London - UK, 2001
[30] Yager RR. On ordered weighted averaging aggregation operators in multi-criteria decision-
making. IEEE Trans. Systems, Man Cybernet 1988; 18: 183–90.
[31] European Commission. SRS NET & EEE: Scientific Reference System on New Energy
Technologies, Energy End-Use Efficiency and Energy RTD (Project Ref: 006631) – 1st Periodic
Report. 2006.
[32] Dess Economie ET Politique DE L’Energie, Gestion Des Nouvelles Technologies De l”Energie.
Overview of current Renewables Energy Policies in the EU and Candidate Countries and
Assessment of Successful Programs and Strategies Implemented in some EU Member States.
2002.
[33] EC - DG Joint Research Centre, Institute for Environment and Sustainability. Energy Efficiency
potential in buildings, barriers and ways to finance projects in New Member States and
Candidate Countries. Proceedings of TAIEX – JRC workshop on Scientific Technical Reference
System on Renewable Energy & Use Efficiency. Tallinn, Estonia, 6-8 July 2005.
[34] Woodall Μ. Financing Renewable Energy Projects in Europe. Infrastructure Journal 1997
[35] Ragwitz M. and Miola A. Evidence from RD and D spending for renewable energy sources in
the EU. Renewable Energy 2005; 30(11): 1635-1647.

1330
Energy Efficiency in Romanian Residential Sector – Present Situation
and Perspectives
Camelia Burlacu

S.C. ELECTRICA SERV S.A. Bucharest, Romania

Abstract
From the energy consumption point of view, the residential sector is the second largest sector in Romania.
Yet, the use of almost all types of energy (electricity, thermal energy, oil petroleum products, natural gas,
coal) used in households decreased in the last years, following the same trend in energy consumption as
all other sectors. This is the case of households owned by the people (unfortunately to many, today) with
the income less than the ones of so-called “middle class”. Of course, people with greater income, affording
greater energy consumption, improved their way of living having new greater houses, including by buying
and using many household appliances.
However, there is a great potential for efficient use of energy in Romanian homes taking into account the
followings: there are over 8,1 millions of households within almost 4,85 millions of buildings; over 50% of
them are more than 40 years old, 37% have the length of service between 20 and 40 years and 10% are
less than 20 years old.
Furthermore, great steps were also made in Romanian legislation regarding energy efficiency, most of
them being in accordance with European legislation (EC Directives).
The paper presents official data about the present situation, the savings potential, what was done
(programs, projects, grants), what is to be done (politics, objectives, goals, actions plans, information
campaigns), different scenarios (needed financial resources, estimated impact) in order to obtain efficient
use of energy in Romanian residential sector.
Because energy efficiency is not only a great challenge but also a must when Romania wants to be an
equal partner within European Union.

Present situation
General aspects
Romania is slowly recovering from an economical declining period, which was also reflected in the
reduction of energy consumption specific to transition process.
In the last decade of the 20th century, energy efficiency policy and programs were not implemented on a
large scale in Romania.
However, in the last years, having in mind the goal of integration in European Union, energy efficiency
became an important part of Romanian national energy policy.

Some basic statistics

According to the 2002 national census, Romania has a population of about 21,68 millions inhabitants and
a dwellings stock of about 8,15 millions.
It results the average number of persons per household of 2,7, of which:
only 1,1 are active persons (40,8% );
1,6 are inactive persons (59,2%), of which: 0,6 are retired people (23,4%); 0,5 are pupils and
students (17,4%); 0,5 are other persons (unemployed people, such as housewives etc.) (18,4%).

The average net salary is about 230 EUR / month.

From the total number of 8,15 millions dwellings:
- 53% are older than 40 years;
- 37% have the length of service between 20 and 40 years;
- 10% are under 20 years old.

In the urban zones, the most of dwellings are in block of flats (81,5%) and the villages 91% represent
singular dwellings (houses).
From the property point of view, 97,5% of total number of dwellings is in private property, owner-occupied,
and only 2,5% is owned by the state.

1331
Energy consumption and energy intensity

According to [24], for residential sector, in 2005, there was registered an energy final consumption of about
7,97 thousands equivalent tones of petroleum which means about 32% of total energy final consumption in
Romania.

In the ranges, the situation is as follows:

- electricity (9,2%);
- thermal energy (34,7%);
- petroleum (19%);
- natural gas (35,2%);
- coal (0,1%);
- others (renewable resources or offal – 1,8%).

In the residential sector, thermal energy is used for heating, preparing warm water and food cooking.
Speaking in general, the efficiency of this thermal energy use is only 43% (63% in Bucharest – the capital
of Romania).
Because of poor conditions of some centralized heating systems and the lack of metering systems at all
block of flats or at individual levels, many consumers, especially from towns areas, preferred individual
heating systems using natural gas.
As for energy intensity in Romanian residential sector, this is over 8 times bigger than the mean value from
European Union countries, varying between 6,6 (compared to Great Britain or Germany) and 12,8
(compared to Spain). Mainly, this results from the less efficiency of district heating and poor thermal
insulation of most apartments situated in block of flats.

Legislation regarding energy efficiency

Important data regarding basic legislation

February 1997. Ratification of "Energy Charta Treaty and the Protocol regarding Energy Efficiency and
Environment Aspects"
According to [1], these contain stipulations such as:
- correlation of energy efficiency policies of different countries;
- development of cooperation between countries, by energy efficiency programs;
- estimation of energy efficiency potential;
- creation of specialized national structures for energy efficiency;
- application of energy efficiency measures;
- facilitation in introduction on market of energy efficiency technologies and services;
- new approaches and methods for financing energy efficiency investments.

November 2000. Law regarding efficient use of energy

This law, [2], was created for application of Energy Charta Treaty. Besides above-mentioned stipulations,
accent is put on the followings:
- complying of international conventions in which Romania is part of;
- Romanian Agency for Energy Conservation (ARCE in Romanian language) is the national authority
for energy efficiency;
- watching the market in order that technical regulations on energy efficiency are respected;
- compulsoriness of using high energy efficiency materials and equipments in all domains of activity;
- compulsoriness of making energy surveys (by authorized persons or firms) for economical firms;
- possibility to obtain exemption from taxes for equipments to be used in energy efficiency projects;
- granting of loans (with 25% lower interests from banks) for energy efficiency projects;
- financial stipulations for activities that lead to energy efficiency increase;
- firms with energy consumption greater than 1000 equivalent tones of petroleum / year and local
administrative authorities for towns with more than 20000 inhabitants are obliged to have special
energy efficiency programs;
- fining the minor offences of law.

Methodological Norms for application of this law were created in 2002 (according to [9]). In these norms, an
entire chapter refers to firms of energy management and services. A special accent is put on energy
efficiency indicators for this kind of projects.

1332
February 2001. Ratification of the "Kyoto Protocol for United Nations Convention on Climate Changes"
(Signed in 1997)
So, according to [3], Romania is committing itself in:
- application and/or elaboration of energy efficiency policies;
- cooperation with other countries;
- changing information.

June 2003. Guidebook for training and testing specialists for energy administration
In this norm, according to [15], the following fact is stressed:
Professional experiences, required for those who want to be authorized for energy survey, differ depending
on category. This category differs in accordance with:
- survey type (electrical, thermal, complex);
- analyzed equipments power.

June 2003. Guidebook for training and testing specialists for activities regarding energy surveys
In this guidebook, according to [16], a distinct domain is present: “Labeling and standardization. Energy
Efficiency”

June 2003. Monitoring procedure for activities regarding energy surveys

According to [17], it is stipulated that energy surveys give the necessary information about following aspects:
- energy consumptions;
- situation of equipments operating and control;
- applicable measures for making optimum the energy parameters.

July 2003. Energy Law

In this law, according to [18], it is written that activities within energy sector must to be carried out to fulfill
some basic objectives, one of them being the following: “transparency of tariffs, prices and taxes for
electricity within a tariff policy, the goal being the increasing of energy efficiency in production, transport,
distribution and use of electricity”.

August 2003. Energy Road Map for Romania

According to [19], the basic scenario regarding energy efficiency is that in which total energy intensity must
be reduced with 30÷50% until 2015, by means of changing high-energy consumption technologies,
restructuring national economy. Alternative scenario foresees a reduction of only 25%, based on a slow
development of Romanian economy caused by influences of global economy.

February 2004. National Strategy regarding Energy Efficiency

It is according to [22].

National authorities

Romanian Agency for Energy Conservation (ARCE in Romanian language)

According to [11], among its main attributes there are the followings:
- drawing up norms and technical settlements for energy efficiency increasing of equipments, outfits
used for energy production, transport, distribution and consumption in buildings and other fields of
activity;
- technical assessment and approval of projects investments from the special fund of energy system
development;
- free of charge expert advice in drawing up and carrying out.

Romanian Fund for Energy Efficiency (FREE in Romanian language)

It was created, according to [10], for financing projects from two sources: internal (Romania) and external
(international bank, especially).

Specific legislation for energy efficiency domestic appliances

Between 2001 and 2003, a series of energy efficiency norms for different domestic appliances (for
household use) appeared by Governmental Decisions. There are referring to settling energy efficiency,
energy and ecological labeling requirements for: refrigerators ([20]), clothes-washing machines ([4]),
combined clothes-washing and drying machines ([5]), electric clothes driers with cylinder ([7]), dish-

1333
washing machines ([8]), electric ovens ([12]), air-conditioners ([14]), electric lamps ([6]), ballasts for
fluorescent lighting sources ([21]).
It must be noticed that in these norms there are references to each applicable European Norm (in most
cases – adopted as Romanian Standard).
In December 2005, a law regarding energy performances of buildings ([23]) was promulgated, containing
aspects regarding:
- construction;
- heating and cooling;
- ventilation;
- lighting;
- energy consumption.

Some examples regarding Romania’s participation in different energy efficiency

programmes

Agreements for grants

These were concluded between Romania and the following institutions:

- International Bank for Reconstruction and Development (in 2002);
- European Commission (PHARE 1 Program – in 2000): energy efficiency improving in district
heating systems;
- European Commission (PHARE 2 Program – in 2002);
- European Commission (in 2003): technical assistance for preparing the projects for environmental
sector in some towns from Romania – Baia Mare, Botosani, Drobeta Turnu Severin, Galati, Deva,
Hunedoara).

Participation on multi-national level

1998÷2002 – ALTENER Program for renewable energy resources in Europe;
1998÷2002 – SAVE Program for energy efficiency in Europe

Cooperation between Romania and only one other country

France. SIENE Symposium on energy and environment theme (in Romania: Cluj Napoca in 1997,
Timisoara in 1999)
The Netherlands (in 1998). Economic cooperation
Switzerland (in 1999). Projects referring to thermal energy in Buzau Town and Pascani Town, Romania
Norway (in 2001). Modernization of district heating system in Fagaras Town, Romania
Sweden (in 2001). Cooperation regarding energy and environment
Bulgaria (in 2002). Regional development of energy policy
Germany (in 2002). Cooperation in the field of energy

Opportunities
Taking into account all the above-mentioned data and experience from different programs, there are some
ways of energy efficiency in residential sector:
- thermal rehabilitation of buildings (dwellings);
- increasing the efficiency of heating systems, lighting systems;
- using efficient "white goods".
The needed investment is about 167 EUR for an annual economy of 1 equivalent tones of petroleum. It
results a total investment of 6,17 billions EUR.

Perspectives
Within its Energy Strategy on Medium Term and Energy Road Map, it is emphasized that, when it
becomes member of European Union, Romania assimilated and applies European Community Acquis,
Romanian market becoming a part of the united Europe large market.
Romanian energy policy has to be synchronized with the European Union one. Accents must be put on
safety, efficiency, environment and consumers’ protection, using competition mechanisms and an
equilibrated settlement of energy efficiency. Not to forget the information campaigns that must be carried
out in mass media, shops, schools.

1334
For the residential sector, it is estimated that the maximum value of annual economy of final energy is
about 3,6 thousands equivalent tones of petroleum, which represents an economy of primary resources of
4,3 thousands equivalent tones of petroleum.
Considering 133 EUR for 1 equivalent tone of petroleum, it results an economy of 0,57 billions EUR / year
for primary resources acquisition. Not to mention the impact on the environment.

Conclusions
Better use of energy leads to economical development. So, DSM ("Demand Side Management") is not
anymore an option, but a necessity, especially for Romania, which has a large potential for this. Especially
in residential sector that represents an important sector regarding energy consumption.
In now a days Romania, one of the most important conditions of progress is to reduce so-called "energy
intensity" and, in order to do this, efficient use of energy is "a must".
In order to gain its place among high-developed countries, Romania has to develop its economy further
on. For this, the focus is on following objectives:
- providing for energy resources and safety;
- energy efficiency;
- utilization of renewable resources;
- environment protection.
One can see that energy efficiency has an important role. Specific actions have to be done for a better
preparation in carrying out conditions required by the European Community Acquis in this domain.
Romania is also hoping that international community will support its efforts by different means (grants,
know-how, sharing experience).
Because today Romania wants to let, to the future generations, a country with high life standard, less
pollution and better use of different forms of energy.

References
[1] Romanian Parliament, Law no.14 regarding ratification of Energy Charta Treaty and the Protocol
regarding Energy Efficiency and Environment Aspects, Official Gazette of Romania, Section 1
no.26 / 18 February 1997
[2] Romanian Parliament, Law no.199 / 2000 regarding efficient use of energy, modified and completed
in 2002 and in 2005, Official Gazette of Romania, Section 1 no.734 / 8 October 2002 and no.291 / 31
March 2006
[3] Romanian Parliament, Law no.3 regarding ratification of the Kyoto Protocol for United Nations
Convention on Climate Changes, Official Gazette of Romania, Section 1 no.81 / 16 February 2001
[4] Romanian Government, Decision no.598 regarding labelling and energy efficiency of clothes-
washing machines for domestic use, Official Gazette of Romania, Section 1 no.377 / 11 July 2001
[5] Romanian Government, Decision no.671 regarding efficiency and energy labelling of combined
clothes-washing and drying machines for domestic use, Official Gazette of Romania, Section 1
no.445 / 8 August 2001
[6] Romanian Government, Decision no.1056 regarding efficiency and energy labelling of electric lamps
for domestic use, Official Gazette of Romania, Section 1 no.727 / 15 November 2001
[7] Romanian Government, Decision no.1274 regarding labelling and energy efficiency of electric
clothes driers with cylinder for domestic use, Official Gazette of Romania, Section 1 no.845 / 28
December 2001
[8] Romanian Government, Decision no.27 regarding labelling and energy efficiency of dish-washing
machines for domestic use, Official Gazette of Romania, Section 1 no.68 / 30 January 2002
[9] Romanian Government, Decision no.393 regarding the Methodological Norms for Law no.199 /
2000, Official Gazette of Romania, Section 1 no.292 / 30 April 2002
[10] Romanian Parliament, Law no.287 regarding Romanian Fund for Energy Efficiency, Official
Gazette of Romania, Section 1 no.344 / 23 May 2002
[11] Romanian Government, Decision no.941 regarding Romanian Agency for Energy Conservation,
Official Gazette of Romania, Section 1 no.673 / 11 September 2002
[12] Romanian Government, Decision no.1117 regarding energy labelling electric ovens for domestic
use, Official Gazette of Romania, Section 1 no.785 / 29 October 2002
[13] Romanian Government – Ministry of Industry and Resources, Decision no.48 regarding Ministerial
Group for national strategy of energy efficiency , Official Gazette of Romania, Section 1 no.156 /
11 March 2003

1335
[14] Romanian Government, Decision no.407 regarding labelling and energy efficiency of air-conditioners
for domestic use, Official Gazette of Romania, Section 1 no.267 / 17 April 2003
[15] Romanian Agency for Energy Conservation, Decision no.57 regarding Guidebook for training and
testing specialists for energy administration, Official Gazette of Romania, Section 1 no.425 / 17 June
2003
[16] Romanian Agency for Energy Conservation, Decision no.58 regarding Guidebook for training and
testing specialists for activities regarding energy surveys, Official Gazette of Romania, Section 1
no.423 / 17 June 2003
[17] Romanian Agency for Energy Conservation, Decision no.59 regarding Monitoring procedure for
activities regarding energy surveys, Official Gazette of Romania, Section 1 no.426 / 18 June 2003
[18] Romanian Parliament, Law no.318 regarding energy, Official Gazette of Romania, Section 1 no.511 /
16 July 2003
[19] Romanian Government, Decision no.890 regarding Energy Road Map for Romania, Official Gazette
of Romania, Section 1 no.581 / 14 August 2003
[20] Romanian Government, Decision no.1039 regarding labelling and energy efficiency of refrigerators
for domestic use, Official Gazette of Romania, Section 1 no.643 / 10 September 2003
[21] Romanian Government, Decision no.1160 regarding energy efficiency of ballasts for fluorescent
lighting sources, Official Gazette of Romania, Section 1 no.716 / 14 October 2003
[22] Romanian Government, Decision no.163 regarding the approval of National Strategy regarding
Energy Efficiency, Official Gazette of Romania, Section 1 no.160 / 24 February 2004
[23] Romanian Parliament, Law no372 regarding Energy Efficiency of Buildings, Official Gazette of
Romania, Section 1 no.1144 / 19 December 2005
[24] Romanian National Institute of Statistics, Romanian Statistical Yearbook 2005, Bucharest, 2006

1336
The Bottom Line of Green is Blue: How ENERGY STAR is
Transforming Home-Building While Generating Economic and
Environmental Impacts
Lisa Surprenant, Michael Mernick, David Meisegeier

ICF International

Abstract
In the US, the traction gained by the ENERGY STAR residential energy efficiency program and its
highly-recognizable cyan-blue logo illustrates how a voluntary program can stimulate innovation,
transform markets, and conserve greenhouse emissions. From light bulbs to heating equipment,
American homes are becoming more efficient as a result. When the EU instituted home energy
ratings, the business of residential energy auditing began to emerge. As this industry takes hold,
learning the lessons of ENERGY STAR is critical.
In 2006, the US Environmental Protection Agency’s (EPA) ENERGY STAR Guidelines for New
Homes will align with recent changes in federal residential energy efficiency standards. An estimated
2,000,000 new housing starts annually could be affected by these changes. Reliance on “branding” is
one way to achieve continued program participation, yet alignment of state programs with these new
guidelines is no easy task.
This paper describes ENERGY STAR’s outreach aimed at ensuring builder and rater retention,
identifies the impacts on the stakeholder sectors, and details innovations like the Indoor Air Package
(under pilot in 3 states) and the highly-contentious Thermal Bypass Checklist that assures builders
address areas prone to envelope inefficiency. This research compares the consequences and
emissions reductions of more energy-efficient homes across eight climate zones and fifty states,
analyzes the barriers faced and concludes with programmatic solutions and lessons learned as these
radical changes to a national program (and mindset) seek to green the American dream.

Background
According to conservative estimates, the United States could save up to 0.70 quads by 2010 by
implementing simple energy efficiency measures in buildings, if mandated by specific policy changes.
And forty billion pounds of CO2 could be prevented from entering the atmosphere if only 10 percent of
US homes were able to meet ENERGY STAR’s guidelines for new construction.
Today, 90 percent of new US homes are found in large developments in suburban settings and the
2000 US Census found that the majority of Americans now reside in those suburban areas. Urban
sprawl and the development of all rural areas are worsening. For example, from 1985 to 1997, the
US population grew 16 percent while the area of land developed grew 47 percent. In addition, the
average single-family home size has increased more than 700 square feet since 1970.
Inefficiency, increasing home size, and urban sprawl are three factors that have led to an
unsustainable situation. To impact the energy use of all new homes and prevent the resultant
environmental degradation due to ever-increasing home sizes and escalating urban sprawl, the US
Environmental Protection Agency (EPA) began its new homes initiative in 1995.
The ENERGY STAR for New Homes program continues today as a voluntary program under the EPA
ENERGY STAR suite of programs aimed at environmental preservation through market
transformation. This program has had the aggressive goal of transforming the U.S. housing market to
energy efficient construction in the residential sector. What began as a modest initiative in 1995,
today has gained traction in every state in the union through a network of more than 90 utility
partners, 2,900 builders, and 360 providers and rater (verification) organizations.
To achieve the sweeping market transformation in the residential sector envisioned by the ENERGY
STAR for New Homes program, the EPA worked with an implementation team at ICF International for
marketing, sales, and technical support services to builders, subcontractors, and broader international
stakeholders including Home Energy Raters (HERs), mortgage lenders, utilities, and product
manufacturers. This singular team construction has lent the ENERGY STAR for New Homes
program the robustness of a government-backed program and the corresponding budget security

1337
coupled with the market-based approach of a consulting firm that has expertise in all relevant areas
from product labeling to home-building.

Figure 1: ENERGY STAR

Qualified Home Label

For a home to be labeled as ENERGY STAR qualified, the New Homes team has worked through
builders and third-party verifiers (Home Energy Raters or HERs raters) to authenticate the energy
efficiency of each ENERGY STAR home. Figure 1 shows the ENERGY STAR qualified homes label.
Figure 2 illustrates how homes may be labeled under the ENERGY STAR process.

The Strategy
The ENERGY STAR new homes team works through the vehicle of ‘account management’ to provide
technical and market support to all target regions and target groups identified. The team lends
support to all relevant stakeholders in that target area. For example, there may not be a utility
program, or a rating infrastructure in a particular target area. The team will work with the utility (or
provider) to train raters and assist them in starting their businesses providing building science
consulting to local builders who seek to participate in the ENERGY STAR New Homes program.
Builders, too, must be convinced to participate in the program. As shown in Figure 2, the builder
begins this process by signing a Partnership Agreement with EPA.
Thus far, the ENERGY STAR for New Homes program has targeted large production builders (those
who build more than 300 homes a year) as its primary catalysts in the marketplace. Small builders,
too, are welcomed to the program, however, their impact is limited by their size. Through a program
of sales training, marketing support, co-branding, and “advertising partnerships” with local builders,
the ENERGY STAR program has flourished, with nearly 3,000 builders participating presently.
Through training, capacity-building, advertising support, co-branding with the ENERGY STAR logo,
and a host of other support mechanisms, the ENERGY STAR for New Homes program has labeled
more than 500,000 homes across the United States as having achieved the targeted energy efficiency
levels. Table 1 shows how the energy efficiency of the typical US home has increased since 1993.

1338
Figure 2: Process of Rating a Home as ENERGY STAR Qualified

1339
Key Features of the Program
The key features and activities of the ENERGY STAR for New Homes program include developing
regional strategic plans for supporting ENERGY STAR New Homes within the various stakeholders in
each region, identifying, recruiting, and cultivating program champions (like utilities or providers) in
each local market considered of interest to the goal of nationwide market transformation. Additionally,
the ENERGY STAR for New Homes team has worked to foster strategic alliances between partners,
conducting and/or facilitating meetings, seminars and workshops, and developing tools and materials
aimed at strengthening recruiting and account management. Since its inception in 1991, the
ENERGY STAR program has conserved emissions equal to the removal of 150,000 cars from US
highways and has the goal to prevent carbon emissions of 9 million metric tons by 2012. The
ENERGY STAR for New Homes program expects this initiative will account for another 1,000,000
new homes being built efficiently by the end of the decade. So the need to retain program partners is
clear.

Table 1: Increase in Energy Efficiency of Various Home Building Components 1993-2006

Hot Climate Mixed Climate Cold Climate
2006 2006 2006
1993 ENERGY 1993 ENERGY 1993 ENERGY
MEC STAR MEC STAR MEC STAR

Shell Information
Wall Insulation (R-Value) 13 13 21 13 21 19
Door (R-Value) 5 1.33 5 2.5 5 7.75
Attic Insulation (R-Value) 30 30 38 38 49 49
Slab Insulation (R-Value) 0 0 5 10 5 10
Window U-Value (U-Value) 0.68 0.55 0.49 0.4 0.33 0.4
Window SHGC (SHGC) 0.38 0.35 0.58 0.45 0.88 0.55
Infiltration (nach) 0.35 0.46 0.46 0.4 0.55 0.33
Systems Information
Cooling Efficiency (SEER) 10 14 10 13 10 13
Heating Eff. (AFUE) 78 80 78 90 78 90
Heating Eff. (HPSF) n/a 8.2 6.8 8.5 6.8 8.5
Duct R-value (R-Value) 5 6 5 6 5 6
Duct Loss (cfm/cfa)* ~8 4 ~8 4 ~8 4
Thermostat Type Man. Prog. Man. Prog. Man. Prog.
Water Heater Fuel
Typ Gas Gas Gas Gas Gas Gas
Water Heater
Efficiency (EF) 0.56 0.61 0.56 0.61 0.56 0.61
Evaluation Location
City Phoenix, AZ Baltimore, MD Minneapolis, MN

Present Challenges in the ENERGY STAR Program

As markets for energy efficiency mature (e.g., the US market began with the 1973 Energy
Conservation Act) various elements must adapt to keep pace with technological and practical
changes. During the past few years in the US, the Home Energy Raters (HERS) guidelines were
amended to be based on the International Energy Conservation Code (IECC) beginning in 2006.
The national requirements for air conditioners, too, were changing, to require a 13 SEER (a more
efficient unit). Additionally, the IECC became the standard for performance over the older Model
Energy Code (MEC). Some of those code changes are that the insulation requirements are now
defined by the 2004 IECC not MEC 93; the operating assumptions are now matched with the 2004
IECC; and equipment efficiencies in the Reference Home were updated to reflect air
conditioning/heating industry changes like the aforementioned upgrade from a 10 to a 13 SEER air
conditioner, from 6.8 to 7.7 HSPF for heat pumps, and increased water heater efficiency. Even the
HERs Index was changing to reflect a much different system of measurement where “zero” energy
use was the preferred score (0) and 100 now constituted a code-built home (the least efficient). And
there were even revisions to the Reference Home used by the IECC upon which to base this index.
In short, the industry was changing and the EPA guidelines for ENERGY STAR qualified homes

1340
needed revision to reflect these shifts.
The ENERGY STAR team began modeling the impacts and effects of various scenarios, using
Department of Energy (DOE-2) modeling software. These changes to the guidelines were announced
in early 2005, with the request from EPA to builders, raters, utilities, and providers to submit written
comments back to the ENERGY STAR team. Those comments were received, reviewed, and
incorporated into the final guidelines issued in December 2005.
In 2006, the ENERGY STAR for New Homes program guidelines will go into effect, to align all
stakeholders with recent changes in federal residential energy efficiency standards, making it more
important than ever to right-size equipment, seal the building envelope, ensure clean indoor air
quality, and secure emissions reduction. An estimated 2,000,000 new housing starts annually could
be affected by these changes in the ENERGY STAR guidelines. Continued compliance was seen as
key to ongoing success of the program.
Reliance on “branding” was seen as one way to achieve continued program participation, however,
alignment of state programs with these new guidelines remains a challenge. In some states,
equipment manufacturers pose a potential obstacle to widespread implementation of this long-
standing and highly successful voluntary program (as manufacturers of more efficient equipment fail
to release pricing information so critical to builders seeking to incorporate this equipment into their
procurement strategies, for example). The single-most important challenge to the ENERGY STAR
New Homes program is still how to ensure retention of key partners (and their trust) as this well-
known program transformation is being rolled-out to thousands of implementing agencies across the
US in the coming months.

Strategy to Effectuate These Changes

The goals of the new guidelines were to respond to industry changes, use tried-and-true energy
efficiency improvements to make those changes, balance energy efficiency with cost-effectiveness,
integrate more ENERGY STAR qualified products into the homes and promote their use, and create
equivalent construction improvement requirements across the country. This last goal is likely the
most interesting and relevant to the EU market, where each country has its own standards for
construction and not unlike the US states, often has very different regional and climate-sensitive
approaches to home-building.
The new ENERGY STAR guidelines are effective immediately, but contain a grandfather clause,
allowing homes that have been permitted before July 1, 2006 or that have been enrolled in a utility
program before December 31, 2005, to continue to be labeled using the old guidelines until January 1,
2007. And there are essentially two paths by which a home may be labeled ENERGY STAR. The
first is the “performance path” that allows any efficiency of equipment to be used as long as the overall
performance target is reached. The lower the efficiency of the equipment, the more upgrades need to
be made. The second path is the “builder option path” otherwise known as the BOP. The BOP
prescribes the efficiency levels of equipment needed to comply. Both paths require the
accompaniment of a completed “Thermal Bypass Inspection Checklist”.
Since the United States covers such a vast landmass, there was the need to divide the country into
“climate zones”. There are eight climate zones used by the present EPA guidelines, as well as
numerous other agencies. Different climate zones may use the Builder Options Package (BOP) to
comply with the ENERGY STAR guidelines. That BOP includes heating and cooling equipment
ratings, energy efficient windows specifications, maximum duct leakage and insulation requirements,
maximum envelope leakage requirements, and completion of the mandatory Thermal Bypass
Inspection Checklist (TBC).
Different climate zones may opt to use the Performance Path to comply with the ENERGY STAR
guidelines for new homes. In that case, builders in various climate zones would ensure envelope
tightness using the Thermal Bypass Inspection Checklist, ensure maximum duct leakage was not
exceeded, and include at least one ENERGY STAR qualified product category within the home to
achieve a HERs Index, as specified by each climate zone. Figure 3 illustrates the various HERs
Indices required in each climate zone.

1341
Figure 3: Climate Zones and HERs Indices

Thermal Bypass Inspection Checklist

Completion of this thermal bypass checklist is mandatory for either path taken by a builder in any
climate zone. This checklist covers the sixteen areas known to cause energy inefficiency in homes.
Preventing thermal bypass in those areas, the ICF modeling showed, will result in home energy
efficiency that is within the limits acceptable to EPA. Each builder must use a HERs rater to verify
that the builder has, in fact, avoided thermal bypass in these problem areas. The checklist provoked
immediate and strong reaction from builders and raters alike after it was formulated. The comments
from utilities, state energy offices, HERs providers and raters, as well as interested stakeholders were
considered by the ENERGY STAR team, acted upon, and then posted on the ENERGY STAR Web
site to ensure transparency of all actions resulting from this feedback. In general, stakeholders were
in favor of the checklist however some worried that it might increase costs or require additional site
visits by raters. Upon resolution of these concerns, the checklist was agreed and approved,
becoming part of the new guidelines to builders.
To prove that thermal bypass has not occurred in the home’s construction, the rater will test each
home, or may sample “batches” of homes. And because some inspections cannot be completed in
one or two visits by a rater, the builder is allowed to check up to four items on the list. Completion of
this checklist ensures that the home buyer receives a third-party quality check of the home; and the
builder gains the advantage of fewer “callbacks” and reduced warranty costs for each home sold. The
subtler effect of this checklist may be a harmonization of construction practices for these problematic
areas in residential construction, such that more efficient homes become the norm.
The ENERGY STAR program has used numerous business tactics and market leverage to reach
builders, raters, realtors, and manufacturers. Consequently, there has been an increase in the
visibility of concepts like quality assurance and quality control. And in the marketplace, consumers
are increasingly requesting “green” homes, as witnessed through green programs like Environments
for Living, USGBC LEED-H, and Southface’s Earthcraft initiative. EPA foresaw this increased focus
and developed the timely Thermal Bypass Inspection Checklist (in 2006) and the Indoor Air Quality
Package to answer the need for quality construction as well as address the need for baseline “green”
using ENERGY STAR guidelines as the threshold.

1342
Another market impact has been that utility-funded programs continue to rely on selling ENERGY
STAR to their customer-base through rebates and have not yet begun to sell the “added value.” This
gap in sales strategy has made it difficult for utility-based programs to help consumers understand not
just the “added value” aspects but also the “added cost” requirements. Yet builder partners around
the country have become sensitized to the need to reduce callbacks and insurance claims to remain
competitive in an increasingly competitive homebuilding market; so the climate for going “green” using
ENERGY STAR program has continued to be successful. And in part, the cooperative advertising
builders are able to use, that depicts them as part of the ENERGY STAR family carrying this cyan-
blue logo has continued to hold sway.

Present Challenges in the EU Residential Program

The EU Energy Performance of Buildings Directive (EPDB) became law in January 2003. Like the US
regulatory mechanisms mandating energy efficiency, the EU Directive has far-reaching implications
for the owners, operators, and developers of all buildings—homes and offices alike. European
experts estimate that more than US$17 billion (13 bl Euro) in potential costs savings could be
captured by 2015 if energy saving measures (like insulation, window replacement, etc.) were
undertaken in all building stock. In the residential sector, the average EU household could easily save
a range of 200 to 1,000 Euros per year, according to its consumption level.
Not unlike the ENERGY STAR new homes program, the EU directives require that new homes be
benchmarked using a whole building energy performance calculation and be certified by a “suitably
qualified and/or accredited expert in an independent manner”. (The EU version of a HERs rater, the
third-party verifier). The major difference between the EU and the US approach to achieving energy
efficient homes and ratings is that the US ENERGY STAR program is voluntary and provides
substantial cost-effective rationale and support to the process of building and rating energy efficient
homes. The EU approach is via a mandatory mechanism—leaving the pathway for “getting to” energy
efficient homes (and the rating of those homes) largely up to the builders and raters.
The next significant difference lies in the approach to market transformation. ENERGY STAR’s
consumer outreach was aimed at generating a “market pull” for energy efficient homes, thereby
creating business opportunities for both the sales of new energy efficient homes (by ENERGY STAR
qualified builders) as well as a ready market for the services of the certified raters (by the ENERGY
STAR qualified builders who seek to have their new homes labeled and who will hire the raters to do
so).

Rating Scales in the UK

The two EU countries that presently exhibit the most developed home energy ratings infrastructure
are Denmark and the UK. In the UK, there are two energy scales for rating homes. The first,
Standard Assessment Procedure (SAP) utilizes a scale from 1-120, with 120 being excellent. A SAP
rating is required for all new buildings like homes, as well as those existing ones being retrofit. Like
the changes to the US HERs index, in 2005, the SAP 2005 will be introduced (effective April 2006) as
the basis for the implementation of the Energy Performance in Buildings Directive.
Similar to the ENERGY STAR’s more stringent guidelines being rolled out in 2006, the SAP 2005
differs from SAP 2001 in that it takes account of building integrated renewable energy systems and
fixed internal lighting. And to make it more difficult for raters, the SAP scale will change to be 1 to
150, not 120.The second method of home energy rating afoot in the UK is the National Home Energy
Rating (NHER) scale. The National Home Energy Rating (NHER) has a 0 -10 scale and provides an
estimate of the amount of carbon dioxide emitted each year as a result of the home’s energy use.

1343
The two energy scales (SAP and NHER) measure slightly different things: the SAP looks at the fixed
elements of the home and is the same wherever the property is located in the UK. (All homes built to
the same design should have the same SAP). Whereas, the NHER includes various location-specific
elements (including whether the home is South-facing or sheltered from wind by other buildings),
occupancy patterns and cooking. And so the NHER reflects actual running costs—much like the US
HERs index. However, unlike the UK system, the US system standardized on one scale, the HERs
index (which, as previously mentioned, has recently been amended to reflect current industry
conditions). Standardization and agreement on one system of scoring is highly recommended for the
EU, to prevent misalignment of energy efficiency labeling of similar homes in different countries.

The UK Homes Label

Both the SAP and the NHER scales produce a label for rating homes in the UK and are based on the
BREDEM (Building Research Establishment Domestic Energy Model). The UK label uses a
logarithmic scale (e.g., increasing an NHER rating by one point will reduce the energy bills by a more
or less constant percentage). Not only are the US and UK (EU) systems different, but the home
labels (certificates) too, differ. The ENERGY STAR label (as a voluntary program) specifies that the
home was rated, the date, and the rater’s name. The UK building certificate (as part of a mandatory
program) goes much further in that the certificate shows the watts per square meter of energy used,
the kg of CO2 per square meter of net area per year, the rated performance of each mechanical
component, and the expected performance summary for the home (or building).
There are many features of the building certificate that are more detailed than the ENERGY STAR
label. While it is unlikely that the voluntary ENERGY STAR label would benefit from adding these
features, the Energy Guide label (also used in the US) to denote actual energy use of equipment
might benefit from comparison with the UK building certificate which incorporates such innovations as
“decibel levels” of the operating equipment. As US consumers become more informed about
ergonomic stressors like noise pollution, learning from the EU experience is also recommended.

The UK Rating Infrastructure

Not unlike the US HERs raters, the UK NHER raters are trained by a central organization (National
Energy Foundation), a registered charity that also developed the National Home Energy Rating
Scheme for British homes. In the US, all raters are certified through the organization known as the
Residential Energy Services Network (RESNET). Although RESNET began as an organization
serving the mortgage industry, today it provides a critical link to the rating infrastructure of the US.
The ICF team assisted RESNET five years ago to revise the RESNET Vision Statement to more
closely align with this growing rating industry. ICF continues to participate on RESNET’s technical
and procedural committees in an advisory capacity. RESNET’s vision has changed since its inception
and its role has changed, too. Now RESNET certifies raters and works closely with the ENERGY
STAR team to standardize rater qualifications and training, track rating transactions, and provide
mechanisms ensuring the quality of all home energy ratings.
Unlike the US guidelines, the EU directives require that a valid energy performance certificate be
produced for all dwellings when they are rented out. And while much uncertainty remains over exactly
how the EU Directive will be implemented in the rental sector, the use of the EU home labels is highly
innovative in scope. The EU Directive is similar to the ENERGY STAR for new homes program in that
the EU Directive mandates that an energy certificate be provided for all homes. But it is dissimilar in
that the ENERGY STAR program does not mandate the prominent display of this energy certificate.
Precisely where to locate the ENERGY STAR label on a home remains a point of discussion to this
day and is largely up to the discretion of the HERs rater.
Experts believe the average home in England would currently receive an NHER rating of between 4.5
and 5.5 (with 10 being excellent). A home meeting the current Building Regulations would probably
score higher, perhaps around 8.0 or more, these experts claim. Yet in 2006, a new version of the
NHER will be introduced to model lighting and appliances more accurately. And its scale will also be
extended from 0 to 20! Before these latest alignments, refinements, and adjustments occur, the
lessons learned by the ENERGY STAR for New Homes program may be instructive.

1344
Conclusions and Lessons Learned
Probably the most important benefits of retaining ENERGY STAR partnerships with builders and the
rating infrastructure have been the constant communication from the field and feedback between EPA
and the various stakeholders who will implement the new guidelines. The back-and-forth dialogue
that allowed partners to vent frustration, vet the new guidelines, and become vested in the new
specifications has led to maintained partnership in key areas. Additionally, this collaboration has
proven beneficial in achieving greater stakeholder “buy-in” of the new guidelines.
The next important aspect has been the ongoing issuance of guidelines updates and increased
partner training in the changes to the guidelines. When the new ENERGY STAR guidelines were
originally changed, a shorter implementation period was envisioned. However, numerous complaints
from stakeholders and a near “crises of confidence” on their part led to a more reasonable timeline,
along with a grandfather clause, allowing builders to make the change in construction once all homes
permitted had been completed.
The presence of a strong rating infrastructure has proven beneficial (and perhaps necessary) to make
inroads into new markets through ENERGY STAR across the US. Support to this rating infrastructure
in terms of training and capacity-building has meant that ENERGY STAR’s best proponents to “sell”
the program to builders were on-board with the new shifts in the guidelines at a time when their
retention was crucial.
The continued improvement trajectory of the ENERGY STAR New Homes program through vehicles
like the Indoor Air Quality Package and the Thermal Bypass Inspection Checklist has increased the
public presence of the ENERGY STAR brand and captured a whole new audience of consumers
searching for “green homes”. As builders relay this new product to their customers, the knowledge
about the integrity, quality factors, and benefits of the purchase of an energy efficient home (and
ENERGY STAR in particular) increase exponentially. And the reduced callbacks and warranty costs
to builders help reinforce their decision to remain ENERGY STAR partners and practitioners.
Homebuyers the world-over seek answers to immediate questions and want to avoid searching
through onerous and technically-challenging information to get those answers. Over the years,
consumers have “looked for the ENERGY STAR label” as in indicator of quality. The one-stop shop
of ENERGY STAR provides US consumers with a central clearinghouse for information on how to
make more informed choices. Much attention has been given by the EPA team to the challenge of
simplifying the buying decisions for consumers and making program participation by industry
stakeholders economically-justifiable as well as pragmatic from an implementation standpoint.
It is important to note that the ENERGY STAR program has been built upon market transformation
principles. The ENERGY STAR program has been adopted by utilities, state energy offices, and
program champions across the US. Some locales and states have even made ENERGY STAR the
threshold for their residential building codes. Yet as the program matured and the market began to be
transformed, the ENERGY STAR team (both within EPA and at ICF) found they needed to carefully
target which markets were “ready” for ENERGY STAR, which messages resonated with new partners,
which tactics needed revision, and how to best nurture the program through the identification and use
of champions or “market mavens”. As the EU program grows, similar strategy adjustments will no
doubt be required.
As the US housing market changes and US builders attempt to “go green” with energy efficiency,
there is real opportunity in the decades-old ENERGY STAR brand. As home buyers have higher
expectations, do more research, and are increasingly time-poor, prospective home buyers search the
Internet and look through the websites of many builders before they ever encounter a salesman.
Savvy builders realize this, and will continue to follow this known brand. As the EU rating
infrastructure grows alongside the implementation of the new directives for homes, European
consumers will benefit from the trail blazed by this decades-old program and be able to foreshadow
the coming adjustments to the SAP and NHER home energy rating systems by seeing that the bottom
line of green is probably blue.

References
[1] Alliance to Save Energy, Vision 2010, Washington, DC, January 2005.
[2] Biedermann, Horst. Excerpt from “Home is Where the Heat Is—Creating a Policy Climate for
Better Buildings” Washington DC, March 2005.

1345
[3] EREC, “Joint Declaration for a European Directive to Promote Renewable Heating and
Cooling” April 2004.
[4] European Commission, Directorate General for Energy and Transport, Memo, June 2005.
[5] European Commission, Directorate General for Energy and Transport, Memo “20% energy
savings by 2010”. June 2005.
[6] http://energystar.gov
[7] http://energystar.gov/ia/partners/bldrs_lenders_raters/downloads/SummaryCommentsFinal.pdf
[8] http://homeenergy.org/archive/hem.dis.anl.gov/eehem/99/991104.html
[9] http://www.nher.co.uk/pages/insight/eu-directive.php
[10] http://www.nef.org.uk/energyadvice/erhome.htm
[11] http://www.nef.org.uk/pages/consumer_centre/energy_ratings/php
[12] http://homeenergy.org/archive/hem.dis.anl.gov/eehem/99/990911/html
[13] Official Journal of the European Communities, 1.1/65, “Directive 2002/91/EC of the European
Parliament and of the Council”, Dec.16, 2002
[14] Partnership Secretariat, Energy Efficiency News, “Measuring Up the Home Energy Ratings,
London, England. 10 Feb. 2006
[15] SE2, Energy Efficiency Partnership for Homes, “The Energy Efficiency Commitment Briefing
Paper”, April 2005.
[16] Shatzhin, A., “Sprawling Towards Climate Change”. ICLEI-US. Berkeley, CA. Fall 2004

1346
Promoting actions for lighting energy efficiency and saving in
residential buildings in Romania
Florin Pop1, Dorin Beu1, Călin Ciugudeanu2
1
Technical University of Cluj-Napoca UTC-N, Romania
2
S.C. RomProiect ELECTRO S.R.L., Romania

Abstract
Both in European Union countries and in Romania, the residential sector represents an important
potential for the reduction of energy consumption. The energy consumption in this sector is focused
on lighting and domestic appliances and heating/air conditioning/hot water. The efficient use of
electricity is still a neglected issue, with a lack of the necessary statistic data.
The Lighting Engineering Center of the Technical University of Cluj-Napoca (LEC UTC-N), Romania is
involved in two programs for promoting lighting energy efficiency and energy saving measures in
residential buildings: ENERLIN - European efficient residential lighting initiative, an EIE - SAVE
program to promote Compact Fluorescent Lamps (CFL) in the residential sector, and CREFEN –
Integrated software system for energy efficiency and saving in residential sector, a Romanian
CEEX program.
Market research has indicated that in order to substantially increase the use of CFLs in the residential
sector, it is essential to develop and market attractive and good quality CFLs. The ENERLIN EIE
SAVE program has set out to propose and validate robust scenarios for CFL promotional campaigns
on European, national and regional levels. The aim of the CREFEN project is to achieve an integrated
software system for reducing the energy consumption and promoting an advanced energy
management in residential buildings in Romania. The software applications are focused on the
electric energy efficient use and saving in residential sector.

ENERLIN – European efficient residential lighting initiative

The European Climate Change Programme (ECCP) identified residential lighting as an important area
to CO2 emission reductions. After a considerable number of promotion and rebate schemes, about
135 million CFLs are used today in European homes. However, only 30% of EU households have at
least one CFL, with those households that own them having an average of three or four. The
ENERLIN EIE SAVE program is aiming at promoting to all the stakeholders a quality charter to assure
that the CFL that are marketed and promoted can deliver savings which last overtime and meet the
customer expectations of high quality lighting, and the ultimate objective of the program is to
substantially increase the efficiency of residential lighting in a number of Member States and
Candidate Countries.
In the context of the Kyoto Agreement, the European Community and individual Member States are
looking for cost-effective measures to reduce CO2 emissions and combat climate change. The
residential lighting market is still dominated by inefficient Incandescent Lamps (GSL – General
Service Lamps). Market research has indicated that to achieve durable market transformation and to
substantially increase the use of CFLs in the residential sector, it is essential to develop and market
attractive and goods quality CFLs. The ENERLIN EIE SAVE program proposes to develop and
validate robust scenarios for CFL promotional campaigns in European, national and regional levels.
Concerning energy savings from CFLs, assuming that there is 150 million households in Europe the
energy economy by replacing only one additional 75 W GSL by one 15 W CFL is in the order of 22.5
TWh or 4 MTEP per annum, this corresponds to 1.2 Mtonnes of less CO2 per annum. We should add
at these savings that a high quality CFL has a life span higher than 10,000 work hours, compared to
2000 work hours for a GSL.
The European Union initiated numberless campaigns to promote compact fluorescent lamps with the
purpose of increasing the market share of compact fluorescent lamps at 15%. In this case, the
estimated energy saving would amount to 15 TWh per year. This energy saving is similar to a
reduction of annual CO2 emissions of about 800 kTones CO2.

1347
Objectives of the ENERLIN action. Improving the energy efficiency is a central theme of energy
policy within the European Community, as indicated in the White Paper “An Energy Policy for the
European Union”, since improved energy efficiency meets all the three goals of energy policy, namely
security of supply, competitiveness and protection of the environment. Lighting represents an
important part of building energy consumption in the EU – around 10% of the total electricity
consumption, ranging from 5% (Belgium, Luxemburg) to 15% (Denmark, Netherlands, and also
Japan).
The global electric lighting energy use may be split in four sectors: services 48%, residential 28%,
industrial 16% and street lighting and other 8% [Mills 2002].
Overall electric appliances in households, industry and the tertiary sector represent 40% of the EU’s
total electricity consumption, its generation being one of the most important sources of CO2
emissions. Within the EU, the households and private and public services sector buildings are
important power consumers. In both cases lighting represents a large part of their energy
consumption. Several EU and National Initiatives and Directives tented to promote energy efficient
lighting for services sector buildings. These efforts can be judged as very successful because
nowadays the Compact Fluorescent Lamp (CFL) market share represent 20% of the global European
market whereas the same figure in world scale is limited to 17%. The rate of the households owning a
CFL covers the range from 0.8 CFLs per household in UK to more than 3 CFLs per household in
Denmark; the SAVE projects have found that there is at least room for 8 CFLs per home [Kofod 2002,
Loe & Jones 2002, DELight 1998]. An analysis on the lighting pattern in 100 Danish homes denotes
that the monthly average lighting consumption varies between 5% and 21% of the total respective monthly
consumption, and 24% of the lamps are energy efficient lamps (linear fluorescent lamps or CFLs).
However, the same market analysis from Lighting Companies show that in Western Europe energy
inefficient incandescence lamps (including halogens) still represent 30% of the sales [9]. The bulk of
these inefficient light sources concern the residential sector.

There are several reasons explaining that residential sector still use a large amount of incandescence
lamps:
• It seems difficult to convince individual customers that the payback time is so rapid.
• There are still many customers unaware of the environmental and economic benefits of CFLs
• Low quality (and probably lower cost) CFLs are widely available in the European markets,
the customers buying these devices due to the attractive price are very rapidly disappointed
by the reduced lifetime, bad lumen output due to wrong information from manufactures about
how to replace incandescent lamps, and bad lumen maintenance of these lamps.
• Older generation of CFLs were almost unable to offer to customers an acceptable
ambiance within the residence, this due to poor colour rendering index, limited choice of
colour temperatures, ungracious shapes and aesthetic incompatibility with luminaires; most of
these inconveniences are now overcame but there is still a large part of customer unaware of
that progress.
• For households lighting can be purely practical or a very architectural feature or a
combination of both, therefore energy efficiency is often just one consideration and probably
not the prevailing one.
• CFL lamps are not suitable for applications with short on-off cycles as this reduces lamp
life; therefore it is necessary to educate the customer on how to use them effectively.
• The warm up time of the CFL before full lumen output does that the user should not use in
a staircase or elsewhere where they need the full lumen output immediately.
• The common CFL lamps are very sensitive to voltage variation; of course, in many
countries the mains voltage is very well regulated, but in other countries and especially
eastern European regions the voltage may fluctuate and this is still an important issue for
CFLs.
• Finding luminaires with nice design, suitable for CFLs, is in nearly all countries, a difficult
issue; many nicely designed luminaires for incandescent lamps do not look equally nice if
they are used with CFLs.

Promoting CFLs by using solid argumentation, which answer to the specific individual questions and
fears of the customer and then add imitative measures seems to be the right way to act. It should be
noticed here that “rational” arguments are not appropriate for all customer segments. To reach the last
customer segment we need new CFL solution and new control features that create added value. To
identify all possible reasons of using CFLs, compile them and provide the good answers (scientifically
proved) and then translate them to a clear and understandable argumentation for the non-specialist,

1348
is the main barrier that needs to be addressed. Furthermore, barriers to information about energy-
efficient technologies (including lighting) exist on several levels, each of which has implications for
penetration rates. The most widespread problem in many countries, to varying degrees, is that of a
lack of awareness of energy efficiency. Members of the general public simply cannot define what it
means for a technology to be energy efficient. Information barriers are important to policy makers as
well.
The ultimate objective of this program is to substantially increase the efficiency of residential lighting
in a number of Member States and Candidate Countries, and this can be done by offering them good
arguments necessary to overcome the above cited barrier. It is also important to promote the wide-
scale availability of a high spectrum of low-cost CFLs suiting a wide range of needs including different
sizes, shapes, colour rendering, wattage (particular problem in some counties like Hungary is that the
typical good CFLs are of lower wattage and therefore provide limited illumination levels), and bases.
Furthermore, to achieve successful residential market transformation we should promote that both
light fixture outlets as well as design and specialty stores display their luminaires with CFLs (good and
aesthetic ones) rather than GSL. At the same time the program is aiming at promoting to all the
stakeholders a quality charter to assure that the CFL that are marketed and promoted can deliver
savings which last overtime and meet the customer expectations of high quality lighting.
The major part of the program will design implement and evaluate a common promotional campaigns
for CFLs that meet the European CFL Quality Charter along with dedicated fixtures. These national or
regional campaigns shall be conducted in collaboration with lamp manufacturers, retailers, consumer
and environmental organisations, and electricity utilities. On the other hand the elaboration of the
argumentation should be based as deep as possible to quantitative and scientific arguments. This
last, may lead to the creation or/and use of independent test facilities allowing to examine different
proposed solutions before adopt them in the final argument list.
All the program objectives will lead to a higher market share for the most efficient CFLs and dedicated
luminaires. The main stakeholders concerned by this program are manufacturers’ associations,
consumers’ associations, buyer’s groups, energy agencies and other intermediates, utilities, training
institutes, retailers, installers and other professions. The final beneficiaries will be end-users of equipment
mainly in domestic sector.

ENERLIN Consortium. 14 partners from 14 countries constitute the proposed consortium, covering a
large part of Europe form north to south and from east to west. This is an important issue; because,
concerning lighting the reaction of the individual customers is quite different form a country to the
other (north countries prefer low colour temperatures lamps –hot ambiance- and south countries are
more sensitive to high colour temperatures –cold ambiance-). The consortium includes western
countries with high GDPs compared to eastern countries that they just integrated EU (Poland,
Hungary, Czech Rep., Latvia and Estonia). These countries are in full market transformation at this
moment. Finally, two candidate countries (Bulgaria and Romania) are also members of the
consortium. The ENERLIN consortium is strongly cross-disciplinary including, National or Regional
Energy Agencies (ADENE, KAPE, ENEA, SEC, SEVEn, BE), one ESCO in Belgium, academic
institutions (France – the coordinator CPAT - University Toulouse 3 -, Hungary and Romania), a
values-based consultancy focusing on sustainability (Respect) as well as independent consulting
SMEs (Ekodoma, Energy Saving Bureau). Each partner has solid experience with EU projects
(especially from DG TREN), and strong links with international organisms like CIE and projects like
ELI, other European networks (COST-529) and programs (GreenLight). Some consortium partners
are quite influential for policy-making bodies in both national (regional) and European levels.

CREFEN – Integrated software system for energy efficiency and saving in

residential sector
The CREFEN project aims to creating an integrated software system-tool focused on the applications
concerning the electric energy efficient use and saving in residential sector in Romania. The project
integrates the consumption assessment and prognosis methodologies, consumption scenarios,
consumers’ guidance and training to the advanced technologies, sustainable electric energy
management and the economical, social and environment impact, as well. A special issue is to
develop the necessary databases of equipment and endowments from residential sector, using the
market surveys and questionnaires.
The project aims to develop an advanced modeling and simulation software system-tool of electric
energy consumption in residential sector and of economical effects, to implement an application with
databases, an interactive educational application and electronic book related to the energy efficiency

1349
use in order to influence the consumers’ options in selecting energy efficient appliances for
environmental protection by reducing the CO2 emissions.

The National Strategy in the energy efficiency field adopted by the HG 163/2004 underlines that
the residential sector has a primary energy saving potential at 3.6 millions tones equivalent petrol
through 6.8 million tones of the total final consumers; it means more than 50%. This potential can be
capitalized by the rehabilitation of the buildings thermic insulation, the improvement of the heating and
lighting systems and of the electric domestic appliances. The legislative frame was created and is on
line with the EU acquis, the EU Directives related to labeling of the energy parameters for many
electric appliances.
The Governmental Program on the following years 2005-2008 states the necessity to accomplish
the legislative and institutional frame in order to apply the flexible mechanisms adopted by the Kyoto
Protocol, to pursue the implementation of the technical and economical measures for the reduction of
the gas emission with the greenhouse effect, in accordance with the features of the National Plan for
the Allocation of the Emission Quotas, the development of the National Plan for Climatic Changes
Action, the improvement of the energy efficiency and the promotion of regenerate energies.

The specific objectives of the CREFEN project are:

• Drawing up of scenarios and prognosis of electric energy consumption in residential sector;
• Achievement of an advanced modeling and simulation software system-tool of electric energy
consumption in residential sector and of economical and environmental effects;
• Using a tool for defining the potential of energy saving, prognosis and scenarios of consumption
evolution;
• Improvement of the degree of taking into consideration by the consumers, decision factors and
specialists of the opportunities, advantages of promoting new technologies in electric energy
consumption in residential sector, in the framework of a sustainable development integrated at the
European level;
• Designing and implementation of a web application with databases for domestic and lighting
appliances available on the Romanian market, which include information from the energy label
and sheet;
• Designing and implementation of the software for an interactive system and an electronic book.

The project is connected with energy efficient use according to EU directives from one side and with
the implementation of database applications using web-based technologies for assisting and
influencing the consumers decision in selecting the domestic and lighting appliances from the other
side, that leading to sustainable environment management. The last aim of the project is for
environmental protection by reducing the CO2 emissions.
The software application architecture will be a modular one, with the possibility of its extension with
new functionalities, without perturbing the other components or requiring the reorganization of the
system data.

LEC UTC-N work in the CREFEN project aims: (1) to analyze the Electric Lighting component in the
energy balance of the dwelling; (2) to elaborate a simplified mathematical model for calculation of the
inside electric lighting; (3) to present the IT system to technical background users groups (students,
designers, dealers and retailers); (4) to contribute with the chapter Electric and Natural Lighting of the
design specification for IT system; (5) to analyze the direct energetic and educational benefits and
evaluate the importance and the impact of the improvement of the domestic users education; (6) to
design consumption evaluation scenarios based on the evolution of the technological performances of
the new lighting equipments (for 5 and 10 years); (7) to print an informing flyer concerning the electric
and natural lighting component; (8) to promote the project achievements by workshops, the ‘Ingineria
Iluminatului’ (Lighting Engineering) journal and the ILUMINAT international conference.

Analysis of electric lighting energy consumption in the residential sector in

Romania.
The statistic data [10] for the period 2000 – 2004 allow us to determine the variation of total
household consumption – Figure 1 -, total number of household consumers – Figure 2 -, average
consumption per household consumer – Figure 3 – and of the specific consumption per m2 – Figure 4.

1350
Figure 1: Variation of total household consumption in Romania between
2000-2004 [MWh]

Figure 2: Variation of the total number of household consumers in Romania

between 2000-2004

1351
Figure 3: Average energy consumption between 2000-2004 [kWh/household
consumer/year]

At the level of the EU, according to “Energy efficiency indicators in Europe” – Odysee [1], the
residential consumption in 2003 was of 2533 kWh/household/year. We remark that in Romania this
consumption is of about 40% of the EU level.

Figure 4: Household consumption per m2 in Romania

(kWh/m2) – the average value of 37.39 m2 per household in
Romania has been considered.

We currently have little information with respect to the electric lighting contribution to the total energy
consumption of the households in Romania. As estimation, we may count on the results of the study
[4], according to which the electric lighting consumption represents around 25% of the total electric
energy consumption of the studied households.
The analysis of the presented data allows us to estimate a few characteristics of electric energy
consumption of households – Table 1.

1352
Table 1: Characteristics of electric lighting energy consumption in households in Romania.
Year 2000 2001 2002 2003 2004
Total annual electric
[MWh] 1,938,000 2,028,000 1,996,000 2,028,000 2,036,000
lighting consumption
Annual electric
[kWh/household
lighting consumption 246 257 253 256 255
/year]
per household
Annual electric
lighting consumption [kWh/m2/year] 6.58 6.87 6.76 6.84 6.83
per m2
Annual cost of
electric lighting [Euro/household
38.57 39.57 39.19 39.47 39.42
consumption per /year]
household
Note: The total annual household consumption in Romania has been determined considering the sales of
Electrical National Company towards the household consumers and the average contribution of the consumption
on the lighting circuits (25%); The annual electric lighting household consumption per m2 was determined based
2
on the average household surface in Romania in 2002 – 37.39 m /household; An average exchange rate of 3.50
RON/Euro.

A comparison of the information presented in [5] to the data presented in this study is shown in Table 2.

Table 2: Characteristics of electric lighting energy consumption in the residential sector in

Romania – comparative study based on data in [1] and [5].
DELight 1998 UTC-N 2005
Electricity consumption
Total electricity consumption (TWh/an) 60.0 (1996) 33.8
Residential electricity consumption (TWh/an) 7.1 8.001
Residential lighting electricity consumption (TWh/an) n.a. (1996) ~2.036
Household lighting – Information
Household lighting electric energy consumption (kWh/an) n.a. (1995) ~255.3
Number of lamps per household 9 n.a.
Average number of CFL per household 0.006 (1995) n.a.
Household ratio using CFL 0.5 (1995) n.a.
Average number of CFL per household using CFL 1.1 (1995) n.a.
Number of luminaires per household 5.5 n.a.
Households – Information
Number of households (x106) 7.78 (1995) 7.97
Number of persons per household 2.91 (1994) 2.63 (2002)
Average surface (m2) n.a. 37.39
Prices 1996 (ecu) 2005 (Euro)
Price of electric energy per kWh 0 – 50 kWh 0.008 0.38
50 – 100 kWh 0.019
>100 kWh 0.041 0.0921
ecu Euro
Price of GSl 0.3 0.43
Price of CFL 13.2 4.3
Note: The average exchange rate of 3.50 RON/Euro has been used

The increase in the annual household electric lighting energy consumption in 2004 vs. 2000, of
3.61%, correlated with the continuous increase of the number of household subscribers in Romania of
1.24% (2004 vs. 2000) – Figure 2, determine the necessity to adopt measures of energy efficiency in
the household lighting of Romania.
The total energy consumption dropped from 60 TWh/year in 1996 [5] to 33.8 TWh/year in 2005,
according to the data received from the national company of electricity. For the same period, the
weight of the residential consumption increased significantly, partly due to the massive reduction of
industrial consumption after 1989, but as well, to a continuous increase of the number of household
subscribers.

1353
Figure 5 shows the spreading of different types of lamps used in household lighting in a few countries
in Europe – New Member States and Candidate Countries. [2]

Figure 5: Residential electric lighting in a few countries in Europe – spreading of different

types of lamps, Source JRC

The cost of CFLs mainly depend on their life span, the cheapest having a life of 3 years, the typical
cost in Romania being between 2.9 – 4.3 Euro, and the more expensive ones have a life span of 8-10
years and a price of 8.6 – 11.4 Euro. Since the electric energy consumption of these lamps is much
lower and has a much smaller cost (only 20% of the cost of GSL, the cost of the initial investment may
be recovered in 3 – 12 months, depending on the cost of the CFL, after which up to 8.0 – 9.0 Euro per
lamp per year is saved.
The use rate of CFLs is from 0.8 units per household in Great Britain up to over 3 units per household
in Denmark. Projects from the SAVE programme consider as a reasonable upper limit the use of up to
8 units per household. An analysis of the residential lighting, realized in 100 households in Denmark,
shows a lighting consumption of between 5% and 21% of the total monthly electric energy
consumption of the household and the use of 24% saving lamps – linear fluorescent lamps and
compact fluorescent lamps.
During November 2005 a study has been realized using feed-back reply forms concerning the usage
degree of GSL and CFLs in households in Western Romania. We received 295 replies, namely 220
apartments (with 1–4 rooms) and 75 houses (with 2–more than 7 rooms). The light source equipment
in these households is presented in Table 3.

Table 3: Light source usage statistics for GSL and CFLs in Romanian households.
Installed
Household GSL CFL
power
Type No. Units Average Units Average kW
Apartment 220 2624 11.98 367 1.67 0.770
Single-family
75 1088 14.51 196 2.61 1.028
house
Total 295 3712 12.58 563 1.91 0.835

The variation of the number of lamps of type GSL and CFL on household type, based on the number
of rooms (living room and bedrooms), is represented in Figure 1 – for apartments, respectively Figure
2 – for single-family houses.

1354
Figure 6: Variation of the number of lamps on apartments with the number of the
rooms (living room and bedrooms)

Figure 7: Variation of the number of lamps on single-family houses with the number of
the rooms (living room and bedrooms)

The installed lighting power has an average value of 0.835 kW/household.

From the analysis of the data presented above we may conclude that the CFL energy saving lamps
are bought by people with high earnings, which own houses.
We consider the equipping degree with CFLs of approximately two units per household is high, and
this denotes the interest of Romanian consumers for buying energy saving lamps.

Conclusions
The estimative values of total electric energy consumption and the total lighting energy consumption
in the residential sector, presented as a conclusions of our study, are of 255.3 kWh/household/year,
and 25% of the total household electric energy consumption, values that fit in the limits presented by
the specialty literature.

1355
The mounting of a single CFL in each household of Romania would lead to a decrease of the
household electric energy consumption of around 45,246 MWh/year. The estimation has been
realized on a theoretical evaluation, based on 2,036,000 MWh/year (the household electric lighting
energy consumption - 2004) divided to 9 (average number of lamps per household in Romania (Table
2) and then to 5 (the ratio between the electric energy consumption of a CFL and a GSL with the
same luminous flux). This value corresponds to a saving in the CO2 emissions of about 2.5 kTones
CO2
(1 kWh = 0.0536 kg CO2 according to the average values considered for European countries).

Other than achieving important electric energy savings and cuts in costs and in polluted emissions,
the introduction of efficient lighting technologies in the case of household consumers presents a
different, important advantage, namely the reduction of the maximal absorbent power in the morning
and evening consumption peaks.
The predicable economic impact of this study will be established by the adoption of policies towards
an electric energy consumption reduction, both locally and nationally. It is essential to increase the
awareness of the energy efficiency problem both by users and by the electric energy providers, in
order to reduce the consumption peaks that are specifically due to lighting.

Acknowledgments
This work was prepared with financial support from the European EIE - SAVE and Romanian CEEX
programmes.

References
[1] Alexandru Adriana, coordinator. CREFEN – Informatic integrated system for energy efficiency
and saving in residential sector – CEEX programme, Contract C608/2005, 2005-2008.
[2] Atanasiu B. and Bertoldi P. Report on Electricity End Use Consumption in New MS and CC in
Tertiary and Residential Sectors, July 2005, JRC - Ispra.
[3] Berrutto V. and Bertoldi P. European Commission energy-efficient lighting initiatives. Proc. of
the 25th Session of the CIE (San Diego, USA, 25 June - 2 July 2003). ISBN 3901-90621-5.
[4] Beu D., coordinator. Study concerning the energy efficiency of the residential electric
appliances – SEEC – Universitatea Tehnica Cluj-Napoca (RO), grant Gr 6113/2000
[5] Environmental Change Unit. Domestic Efficient Lighting (DELight), University of Oxford (UK),
1998. ISBN 1-874370-20-6.
[6] European Commission. Commission Directive 98/11/EC of 27 January 1998 implementing
Council Directive 92/75/EEC with regard to energy labelling of household lamps, Official
Journal L71, 10.03.1998, pp. 01-08.
[7] Lewis J.O., coordinator. EnerBuild RTD Network - FP5 programme, 2001-2003.
[8] Pop F. Energy efficiency in lighting between regulations and reality, Proceedings of Workshops
Budapest, Prague, Warsaw, Newly Associated States EnerBuild RTD, 2003, ISBN 80-239-
0742-5, pp. 38-42
[9] Zissis G., coordinator. European Efficient Residential Lighting Initiative - ENERLIN, EIE
“Intelligent Energy – Europe” programmeGrant EIE/05/176/SI2.419666, 2006-2008.
[10] Statistic data 2003. Romanian Statistic National Institute.

1356
Lighting Efficiency in China
Ken Tiedemann

BC Hydro

Abstract
Lighting accounts for about 13% of the electricity used in China, and lighting is probably the largest
and most rapidly growing use for electricity in China. Over the decade from 1988 to 1988, it is
estimated that electricity used for lighting in China increased from 44 TWh per year to 152 TWh per
year. Over the past ten years, China has become the largest producer as well as the largest exporter
of lighting products. The efficiency and quality of Chinese lighting products has major impacts on
global energy use, sustainability and greenhouse gas emissions. The China Green Lights program
was established to address barriers to the development, production and sale of high quality and
efficient products in China. This paper examines the growth of lighting products and estimates the
impact of efficient lighting products on electricity use and carbon dioxide emissions.

Introduction
Lighting accounts for about 13% of the electricity used in China, and lighting is probably the largest
and most rapidly growing use for electricity in China. Over the decade from 1988 to 1988, it is
estimated that electricity used for lighting in China increased from 44 TWh per year to 152 TWh per
year, and lighting use has increased rapidly in the residential, commercial and industrial sectors. Key
factors leading to increased lighting electricity consumption include: increased residential use of
lighting with rising incomes; large numbers of commercial buildings built to international standards;
better illuminated factories; and intensive development of street and large area lighting [1, 2, 3].
Over the past ten years, China has become the largest producer as well as the largest exporter of
lighting products. The efficiency and quality of Chinese lighting products has major impacts on global
energy use, sustainability and greenhouse gas emissions. Initial exports of Chinese lighting products,
including CFLs in particular, often had unreasonably high rates of premature failure and of premature
decline in lighting efficacy. This appeared to be due to use of low quality glass and electrical
components combined with poorly training workers and inadequate attention to quality control.
The China Green Lights program was established in October 1996, with the assistance of the United
Nations Development program, to address barriers to the development, production and sale of high
quality and efficient lighting products in China. This initial program invested some US $36 million to
improve both the demand side and the supply side of the Chinese lighting market. Key planned
program outcomes were increased purchases and market share of CFLs and efficient fluorescent
tubes, reduced electricity consumption and reduced greenhouse gas emissions.
This paper uses quantitative methods to examine the impact of the China Green Lights program on
production of lighting products, electricity use and emission in China. The approach used here draws
on the small but developing literature on quantitative analysis of demand side management programs,
which emphasizes the use of econometric models to overcome the need to estimate market effects
such as free riders and spillover.
Several previous studies have used econometric methods to analyze the impact of market
transformation programs. Duke and Kammen [4] found that accounting for feedback between the
demand response and production response for electronic ballasts increases the consumer benefit
cost ratio. Horowitz [5] found that coordinated national electronic ballast programs were more cost
effective than local efforts. Horowitz and Haeri [6] found that the cost of energy efficiency investments
was fully capitalized in housing prices and that purchasing an energy efficient house was cost
effective. Jaffe and Stavins [7] found that insulation levels in new residential housing appropriately
reflect energy prices.
An outline of the paper is as follows. The next section provides a review of the methodology used in
this study. This is followed by individual sections covering production, exports and sales of
incandescent lamps, halogen lamps, low voltage lamps, standard fluorescent tubes, efficient
fluorescent tubes, compact fluorescent lamps and other lamps. The penultimate section provides
estimates of reduction in electricity use and carbon dioxide emissions due to the China Green Lights
program. The last section offers the study’s conclusions.

1357
Model and Estimation
It is convenient to view a single lamp or tube market (such as the market for compact fluorescent
lamps or energy efficient fluorescent tubes) in isolation and abstract from linkages to other markets or
general equilibrium effects. Consider the following simple four-equation model where (1) is the
demand curve for a specific lighting product, (2) is the stochastic process for the path of income over
time, (3) is the supply curve for the lighting product, and (4) is the stochastic process for import prices
over time, and the error terms have been suppressed for convenience.
(1) quantityt = a + b* pricet + c* incomet + d* dummyt
(2) incomet = e + f*timet
(3) pricet = g + h* world pricet + i*dummyt
(4) import pricet = j + k*timet

In these equations, quantityt is domestic market demand for the lighting product at time t, pricet is unit
price of the lighting product at time t, incomet is total domestic income at time t, timet is the year t,
world pricet is the per unit at time t, and dummyt is a shift variable that takes the value “0” in the pre-
China Green Lights program period and the value “1” in the China Green Lights program period.
Equation (1) represents the demand curve in year t. It says that market demand for the lighting
product is a linear function of product price, total domestic income and a preference variable, which
represents a shift in consumer demand towards the product of interest. This demand shift could be
driven by program advertising and promotional activities or by other factors, but in any event reflects
consumer views of relative value of the product compared to other products. Equation (2) represents
the path of income over time. It says that the path of income can be adequately represented by a
(stochastic) linear trend. Over the relatively short period of data available, this is a reasonable
assumption.
Equation (3) represents the supply curve for a lighting product in year t. It says that that lighting
product price is a linear function of the world price and a supply shift variable. The supply shift
variable represents a shift in domestic supply towards higher efficiency products (or away from lower
efficiency products) as a result of changes in producer preferences, perhaps as a result of marketing
initiatives, or the effect of a subsidy. Equation (4) is the world price which is constant over a given
year but decreases over time due to experience curve effects and increased competition due to
market transformation.
Substituting for income in (1) and substituting for price in (3) yields a simple structural equation model
with two equations in two variables, quantity and price. Solving this model for price and for quantity,
yields in turn the reduced form of this structural model as follows:
(5) quantityt = α + β*time +χ*dummyt
(6) pricet = δ + φ*time +γ*dummyt

Equation (5) represents the reduced form equation for quantity. Note that it has been rearranged so
that quantity depends on a constant plus a time trend plus a term that represents program impacts.
Equation (6) represents the reduced form equation for price, which in this model is the same as the
supply equation. Note that it has been rearranged so that price depends on a constant plus a time
trend plus a term that represents program impacts. We provide estimates of equations (5) in the next
section.
The study approach is as follows. First, information from local surveys and data sources are used to
build a database of domestic consumption for various lighting products. Second, regression
discontinuity models (sometimes referred to as interrupted time-series models) are applied to this
data to understand the impact of the China Green Lights program on quantities of lighting products
produced. In the regression models, it is assumed that all non-program factors are captured by a
time-trend term so that the coefficient on a dummy variable for the program period measures program
impact. Third, using engineering data, algorithms are used to estimate the energy savings for each
efficient technology as well as the impact of energy savings on emissions. Time-series information on
product sales is invaluable in understanding market transformation, and this is often done by using
the ordinary least squares regression model. But time-series data on sales of products is often
characterized by the first-order auto-regressive process where the errors between periods are not
independent. Applying the usual ordinary least squares evaluation framework in this context can
potentially provide misleading results. For references see Dhrymes [8], Johnston [9], Malinvaud [10]
and Theil [11].

1358
Overview of Chinese Lamp Production and Performance
Chinese production of lighting products for the period 1988 to 1998 is shown in Table 1. Production is
reported in millions of lamps per year. Lamp types include GLS or standard incandescent, halogen,
low voltage or primarily vehicle lamps, T5 to T-8 efficient fluorescent tubes, T9 to T12 standard
fluorescent tubes, compact fluorescent lamps and other lamps, which includes high intensity
discharge lamps.

Table 1: Lamp Production (million units)

GLS Halogen Low T5-T8 T9-T12 CFL Other Total
voltage tubes tubes
1988 1,303 30 50 0 156 8 214 1,761
1989 1,437 40 55 0 171 9 309 2,021
1990 1,747 50 61 4 208 10 396 2,476
1991 1,993 80 89 6 212 11 437 2,828
1992 2,264 90 131 8 278 18 551 3,340
1993 2,217 100 399 20 238 38 574 3,586
1994 2,559 189 567 14 282 52 880 4,543
1995 2,900 278 666 18 328 66 1,244 5,500
1996 3,413 300 650 50 364 120 1,498 6,395
1997 3,053 350 650 70 423 150 1,337 6,033
1998 2,624 400 650 84 336 180 1,334 5,608
Source: China Association of Lighting Industry, various years. Note that 1994 data are author’s
interpolations. Fluorescent tube shares by type, before 1995, are based on discussions with industry
officials.

Chinese lamp performance is compared to international standard performance for the mid-1990s in
Table 2. Wattage refers to the range of available lamps in watts. Efficacy refers to lamp efficiency in
terms of lumens per watt. Life refers to average hours of lamp life under normal conditions. Lamp
types included in the table are incandescent, halogen, linear fluorescent, compact fluorescent, metal
halide and high pressure sodium.

Table 2: Chinese and International Lamp Performance

China International
Wattage Efficacy Life Wattage Efficacy Life
(W) (lumens/W) (hours) (W) (lumens/W) (hours)
Incandescent 15-1000 7-18 100-1000 10-1500 5-25 100-2000
Halogen 50-2000 14-28 50-1500 10-1000 15-30 50-2000
Linear fluor 6-125 25-40 3000-5000 4-200 40-100 10000-
20000
CFL 4-28 40-65 1000-3000 5-55 50-70 5000-
20000
Metal halide 150-1500 70-80 3000-8000 30-2000 50-125 2000-
10000
HPS 35-1000 60-100 4000- 30-1000 50-150 10000-
12000 25000
Source: Min, Mills and Zhang, 1995.

Incandescent Lamps
An incandescent lamp consists of a thin, frosted glass envelope filled with argon and/or nitrogen gas.
Electricity heats a tungsten filament to about 2500°C causing the tungsten to emit visible light and
heat. The incandescent lamp is inefficient because most of the energy is used to produce heat, and it
is short lived because the tungsten evaporates and is deposited on the glass before a thin spot is
formed on the filament and breaks. China produces a wide variety of incandescent lamps, but
medium-base Type A lamps predominate. As shown above in Table 2, China in the mid-1990s
produced a narrower range of wattages for incandescent lamps than was available on the
international market. Lamp efficacy in terms of lumens per watt and length of lamp life were both
significantly worse in China than internationally.

1359
Production, export and apparent domestic consumption of incandescent lamps in China for 1995 to
1998 are shown in Table 3. Production and domestic consumption of incandescent lamps appears to
have peaked in 1996 with declines in production and domestic sales in 1997 and 1998. Exports show
a generally upward pattern, but the export share of production is lower for incandescent lamps than
for certain other products such as CFLs.

Table 3: Incandescent Lamps: Production, Exports and Domestic Use (million units)
Production Exports Apparent Consumption
1995 2900 150 2750
1996 3413 142 3271
1997 3053 98 2955
1998 2624 423 2201
Source: China Association of Light Industry, various years.

Determinants of Chinese lamp production are explored in Table 4. We use a simple econometric
model, where sales in millions of units are a function of GDP in trillions of constant 1995 yuan and of
a dummy variable for the Chinese Green Lights program (which are the years 1997 and 1998). From
a statistical perspective, both regressions fit the data well. The explanatory power of both regressions
is very high at over 85%. Auto-correlation is not a major problem with either regression, although the
maximum likelihood regression slightly improves the estimated auto-correlation.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of incandescent lamps by 520,000 units. The China Green Lights program reduces annual sales of
incandescent lamps by 857 million units. This is about one-quarter of the peak sales of incandescent
lamps of 3,053 million units in 1996. In the maximum likelihood regression (ML), a one trillion increase
in GDP increases annual sales of incandescent lamps by 530,000 units. The China Green Lights
program reduces annual sales of incandescent lamps by 861 million units. Results of the OLS and ML
regression are similar.

Table 4: Incandescent Lamps: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -104 315 0.75 -112 304 0.71
GDP 0.52 0.069 0.00 0.53 0.067 0.00
Green -857 272 0.01 -861 267 0.00
Lights
R-squared 0.87 - - 0.89 - -
D-W 1.70 0.15 - 2.19 -0.09 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated
auto-correlation is shown in parentheses beside the Durbin-Watson statistic.

Halogen Lamps
A halogen lamp uses a tungsten filament, but the envelope is made of quartz with a higher melting
point than silicon based glass, and it is relatively small. The gas inside the envelope comes from the
halogen group, allowing the filament to burn hotter thus improving lamp efficiency. At high
temperatures, the released tungsten combines with halogen and is re-deposited on the filament,
which increases the length of life of the filament. China produces a range of halogen lamps
comparable to the international market. Lamp efficacy in terms of lumens per watt and length of lamp
life were close to international standards in the mid-1990s.
Production, export and apparent domestic consumption of halogen lamps in China for 1995 to 1998
are shown in Table 5. Production steadily increased over this period. Exports show a generally
upward pattern, but the export share of production is lower for incandescent lamps than for certain
other products such as CFLs.

1360
Table 5: Halogen Lamps: Production, Exports and Domestic Use (million units)
Production Exports Apparent Consumption
1995 278 180 98
1996 300 200 100
1997 350 230 120
1998 400 265 135
Source: China Association of Light Industry, various years.

Determinants of Chinese halogen lamp production are explored in Table 6. Sales in millions of units
are a function of GDP and of a dummy variable for the Chinese Green Lights program (which are the
years 1997 and 1998). From a statistical perspective, both regressions fit the data well. The
explanatory power of both regressions is very high at over 97%. Auto-correlation is a major problem
with the OLS regression, and the ML regression substantially improves the estimated auto-correlation.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of halogen lamps by 81,000 units. The China Green Lights program increases annual sales of
halogen lamps by 16 million units. In the maximum likelihood regression (ML), a one trillion increase
in GDP increases annual sales of halogen lamps by 82,000 units. The China Green Lights program
increases annual sales of halogen lamps by 11 million units. The results of the OLS and ML
regression are very similar.

Table 6: Halogen Lamps: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -228 32 0.00 -230 38 0.00
GDP 0.081 0.0070 0.00 0.082 0.0082 0.00
Green 16 28 0.58 11 29 0.40
Lights
R-squared 0.97 - - 0.97 - -
D-W 1.38 0.31 - 1.63 0.19 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated auto-
correlation is shown in parentheses beside the Durbin-Watson statistic.

Low Voltage Lamps

Most low voltage lamps produced in China are used in motor vehicles. Most vehicle headlights use
halogen bulbs, but some headlights use high intensity discharge lamps, including mercury vapour
lamps with xenon added to reduce the start-up time. Motor vehicle lamps produced in China appear to
generally meet international standards
Production, export and apparent domestic consumption of low voltage lamps in China for 1995 to
1998 are shown in Table 7. Production and domestic consumption of low voltage lamps appears to
have peaked in 1995 with declines in production and domestic sales in 1996, 1997 and 1998. Exports
show a generally upward pattern, but the export share of production is lower for incandescent lamps
than for certain other products such as CFLs.

Table 7: Low Voltage Lamps: Production, Exports and Domestic Use (million units)
Production Exports Apparent Consumption
1995 666 80 586
1996 650 100 550
1997 650 120 530
1998 650 120 530
Source: China Association of Light Industry, various years.

Determinants of Chinese low voltage lamp production are explored in Table 8. Sales in millions of
units are a function of GDP in trillions of constant 1995 yuan and of a dummy variable for the Chinese
Green Lights program (which are the years 1997 and 1998). From a statistical perspective, both
regressions fit the data well. The explanatory power of both regressions is very high at about 95%.
Auto-correlation is not a major problem with either regression.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of low voltage lamps by 210,000 units. The China Green Lights program apparently reduces annual

1361
sales of low voltage lamps. This particular result makes little sense and is due to the rapid ramp up of
production of low voltage lamps in the period before the program. In the maximum likelihood
regression (ML), a one trillion increase in GDP increases annual sales of low voltage lamps by
220,000 units. The China Green Lights program again apparently reduces annual sales of low voltage
lamps. The results of the OLS and ML regression are very similar.

Table 8: Low Voltage Lamps: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -639 92 0.00 -656 84 0.00
GDP 0.21 0.020 0.00 0.22 0.018 0.00
Green -251 79 0.01 -268 75 0.00
Lights
R-squared 0.94 - - 0.95 - -
D-W 2.02 -0.01 - 2.11 -0.06 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated
auto-correlation is shown in parentheses beside the Durbin-Watson statistic.

Standard Fluorescent Tubes

A fluorescent lamp consists of a sealed glass tube which is coated on the inside with a phosphorous
powder. The tube contains mercury, an inert gas such as argon, and two electrodes. This mechanism
is used to excite atoms using a chemical reaction rather than by heating a tungsten element, as in an
incandescent or halogen lamp. Using a chemical rather than a thermal process increases the
efficiency of the lamp and greatly lengthens lamp life. The range of fluorescent lamps produced in
China in the mid-1990s was fairly wide, but lamp efficacy and lamp length of life were substantially
below international standards.
Production, export and apparent domestic consumption of T9-T12 fluorescent lamps in China for
1995 to 1998 are shown in Table 9. Production and domestic consumption of T9-t12 lamps appears
to have peaked in 1997 with declines in production and domestic sales in 1998.

Table 9: Standard Fluorescent Tubes (T9-T12): Production, Exports and Domestic Use (million
units)
Production Exports Apparent Consumption
1995 328 - 328
1996 364 - 364
1997 423 - 423
1998 336 - 336
Source: China Association of Light Industry, various years.

Determinants of Chinese T9-t12 lamp production are explored in Table 10. Sales in millions of units
are a function of GDP in trillions of constant 1995 yuan and of a dummy variable for the Chinese
Green Lights program (which are the years 1997 and 1998). From a statistical perspective, both
regressions fit the data well. The explanatory power of both regressions is high at about 82%. Auto-
correlation is not a major problem with either regression.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of T9-T12 lamps by 52,000 units. The China Green Lights program reduces annual sales of T9-T12
lamps by 16 million units. In the maximum likelihood regression (ML), a one trillion increase in GDP
increases annual sales of T9-T12 lamps by 55,000 units. The China Green Lights program reduces
annual sales of T9-T12 lamps by 21 million units. The results of the OLS and ML regression are very
similar.

1362
Table 10: Standard Fluorescent Tubes: Determinants of Sales (millions of units)
Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant 21 47 0.45 11 32 0.74
GDP 0.052 0.010 0.00 0.055 0.0070 0.00
Green -16 41 0.71 -21 30 0.48
Lights
R-squared 0.82 - - 0.83 - -
D-W 2.22 -0.11 - 2.34 -0.17 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated auto-
correlation is shown in parentheses beside the Durbin-Watson statistic.

Efficient Fluorescent Tubes

Efficient triphosphor tubes use significantly less energy than halophosphate tubes and suffer far less
degradation in light output levels over time. Detailed information on production of fluorescent lamps by
efficiency level is not available, so for our purposes T5-T8 lamps are viewed as efficient while T9-T12
lamps are view as standard.
Production, export and apparent domestic consumption of incandescent lamps in China for 1995 to
1998 are shown in Table 11. Production and domestic consumption of incandescent lamps appears to
have increased through 1998.

Table 11: Efficient Fluorescent Tubes (T5-T8): Production, Exports and Domestic Use (million
units)
Production Exports Apparent Consumption
1995 18 - 18
1996 50 - 50
1997 70 - 70
1998 84 - 84
Source: China Association of Light Industry, various years.

Determinants of Chinese T5 and T8 lamp production are explored in Table 12. Sales in millions of
units are a function of GDP in trillions of constant 1995 yuan and of a dummy variable for the Chinese
Green Lights program (which are the years 1997 and 1998). From a statistical perspective, both
regressions fit the data well. The explanatory power of both regressions is very high at about 93%.
Auto-correlation is not a major problem with either regression, although the ML regression
substantially improves the estimated auto-correlation.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of T5-T8 lamps by 11,000 units. The China Green Lights program increases annual sales of T5-T8
lamps by 33 million units. In the maximum likelihood regression (ML), a one trillion increase in GDP
increases annual sales of T5-T8 lamps by 10,000 units. The China Green Lights program increases
annual sales of T5-T8 lamps by 39 million units. The results of the OLS and ML regression are very
similar.

Table 12: Efficient Fluorescent Tubes: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -36 10 0.01 -32 8 0.00
GDP 0.011 0.0023 0.00 0.010 0.0018 0.00
Green 33 9 0.01 39 8 0.00
Lights
R-squared 0.93 - - 0.94 - -
D-W 2.32 -0.16 - 2.13 -0.06 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated auto-
correlation is shown in parentheses beside the Durbin-Watson statistic.

1363
Compact Fluorescent Tubes
China is currently the largest supplier of compact fluorescent lamps, which are fluorescent tubes
folded on themselves to minimize space requirements. In the mid-1990s many Chinese produced
CFLs lasted little longer than incandescent lamps, but product quality and lamp life of Chinese
manufactured CFLs have improved substantially over the past decade. Many lower end Chinese
factories have ceased production. Chinese now produces CFLs in a wide range of wattage levels and
styles.
Production, export and apparent domestic consumption of CFLs in China for 1995 to 1998 are shown
in Table 13. Production and domestic consumption of incandescent lamps have steadily increased
through 1998 and exports show an upward pattern, and the export share of production is very high
compared to other types of lamps.

Table 13: Compact Fluorescent Lamps: Production, Exports and Domestic Use (million units)
Production Exports Apparent Consumption
1995 66 39 27
1996 120 70 50
1997 150 80 70
1998 180 100 80
Source: China Association of Light Industry, various years.

Determinants of Chinese CFL production are explored in Table 14. Sales in millions of units are a
function of GDP in trillions of constant 1995 yuan and of a dummy variable for the Chinese Green
Lights program (which are the years 1997 and 1998). From a statistical perspective, both regressions
fit the data well. The explanatory power of both regressions is very high at about 96%. Auto-
correlation is not a major problem with either regression.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of CFLs lamps by 29,000 units. The China Green Lights program increases annual sales of CFLs by
45 million units. In the maximum likelihood regression (ML), a one trillion increase in GDP increases
annual sales of CFLs by 28,000 units. The China Green Lights program increases annual sales of
CFLs by 50 million units. The results of the OLS and ML regression are very similar.

Table 14: Compact Fluorescent Lamps: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -92 17 0.00 -88 15 0.00
GDP 0.029 0.0037 0.00 0.028 0.0033 0.00
Green 45 15 0.02 50 14 0.00
Lights
R-squared 0.96 - - 0.96 - -
D-W 2.04 -0.02 - 2.11 -0.05 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated auto-
correlation is shown in parentheses beside the Durbin-Watson statistic.

Other Lamps
Other lamps include low pressure sodium (widely used for street lamps in China), high pressure
sodium, mercury vapor, metal halide and a variety of specialty lamps. In the mid-1990s, a wide range
of high pressure sodium and metal halide lamps were produced. But lamp efficiency and lamp life for
these high intensity discharge lamps were at perhaps two-thirds of international levels.
Production, export and apparent domestic consumption of other lamps in China for 1995 to 1998 are
shown in Table 15. Production and domestic consumption of other lamps appears to have peaked in
1996 with declines in production and domestic sales in 1997 and 1998. Exports show a generally
upward pattern, but the export share of production is low.

1364
Table 15: Other Lamps: Production, Exports and Domestic Use (million units)
Production Exports Apparent Consumption
1995 1,244 4 1,240
1996 1,498 6 1,492
1997 1,337 8 1.329
1998 1,334 12 1,322
Source: China Association of Light Industry, various years.

Determinants of Chinese lamp production are explored in Table 16. Sales in millions of units are a
function of GDP in trillions of constant 1995 yuan and of a dummy variable for the Chinese Green
Lights program (which are the years 1997 and 1998). From a statistical perspective, both regressions
fit the data well. The explanatory power of both regressions is very high at over 96%. Auto-correlation
is a problem with then OLS regression, although the ML regression improves the estimated auto-
correlation.
In the ordinary least squares regression (OLS), a one trillion increase in GDP increases annual sales
of other lamps by 560,000 units. The China Green Lights program reduces annual sales of other
lamps by 485 million units. In the maximum likelihood regression (ML), a one trillion increase in GDP
increases annual sales of other lamps by 55,000 units. The China Green Lights program reduces
annual sales of other lamps by 477 million units. The results of the OLS and ML regression are very
similar.

Table 16: Other Lamps: Determinants of Sales (millions of units)

Ordinary Least Squares Regression Maximum Likelihood Regression
Coefficient St. Error Significance Coefficient St. Error Significance
Constant -1,453 204 0.00 -1,436 215 0.00
GDP 0.56 0.045 0.00 0.55 0.047 0.00
Green -485 177 0.03 -477 181 0.01
Lights
R-squared 0.96 - - 0.97 - -
D-W 1.62 0.19 - 2.00 -0.00 -
Note: Standard errors are shown in parentheses beside the regression coefficients and estimated auto-
correlation is shown in parentheses beside the Durbin-Watson statistic.

Impact on Energy Use and Carbon Dioxide Emissions

Energy savings are estimated in Table 17 for compact fluorescent lamps and for energy efficient
fluorescent tubes. Energy savings for the first two years of China Green Lights are defined as unit
change in load multiplied by average annual hours multiplied by twice the estimated program impact
on sales. For screw-type lamps, it is assumed that a 13 watt CFL replaces, on average, a 40 watt
GLS lamp leading to a change in load of 27 watts. Average hours of use are assumed to be 4.0 hours
per day or 1,460 hours per year. For fluorescent tubes, it is assumed that a 36 watt T8 tube replaces,
on average, a 40 watt T12 tube leading to a change in load of 4 watts. Average hours of use are
assumed to be 5 hours per day or 1,825 hours per year.
The unit savings results are shown below. Estimated savings per unit are 39.4 kWh for a CFL
replacing an incandescent lamp and 8.0 kWh per year for a T5 to T8 lamp replacing a T12 lamp.
These results are the estimated savings that would be experienced at the customers’ meters. At the
system level, savings would be about ten percent higher to allow for transmission losses and
distribution losses.
Annual energy savings from two years of China Green Lights are 4,426 GWh using the OLS
regressions and 4,954 GWh using the ML regressions. Carbon dioxide reductions are 2,753
kilotonnes using the OLS regression and 3,081 kilotonnes using the ML regressions.

1365
Table 17: Annual Energy and Carbon Dioxide Savings at the Meter and at the Generation
Busbar
Base Efficient Unit Hours Unit Line Unit Total CO2
lamp lamp demand (per meter loss busbar busbar savings
(W) (W) savings year) savings factor savings savings (GWh)
(W) (kWh) (kWh) (GWh)
Ordinary Least Squares Based Estimates
CFL/GLS 40 13 27 1,460 39.4 1.10 43.3 3,898 2,425
T8/T12 40 36 4 1,825 7.3 1.10 8.0 528 328
Total 4,426 2,753
Maximum Likelihood Based Estimates
CFL/GLS 40 13 27 1,460 39.4 1.10 43.3 4,330 2,693
T8/T12 40 36 4 1,825 7.3 1.10 8.0 624 388
Total 4,954 3,081
Source: Min, Mills and Zhang, 1995 and regressions. Emissions factor of 0.622 kilotonnes of carbon
dioxide per GWh of electricity and loss factor were provided by the China Green Lights program.

Conclusions
Lighting accounts for about 13% of the electricity used in China, and lighting is probably the largest
and most rapidly growing use for electricity in China. Over the decade from 1988 to 1988, it is
estimated that electricity used for lighting in China increased from 44 TWh per year to 152 TWh per
year. Over the past ten years, China has become the largest producer as well as the largest exporter
of lighting products. The efficiency and quality of Chinese lighting products has major impacts on
global energy use, sustainability and greenhouse gas emissions.
The China Green Lights program was established to address barriers to the development, production
and sale of high quality and efficient products in China. This paper examines the growth of lighting
products and estimates the impact of efficient lighting products on electricity use and carbon dioxide
emissions.
Estimated savings per unit are 39.4 kWh for a CFL replacing an incandescent lamp and 8.0 kWh per
year for a T5 to T8 lamp replacing a T12 lamp. Annual energy savings from two years of China Green
Lights are 4,426 GWh using the OLS regressions and 4,954 GWh using the ML regressions. Carbon
dioxide reductions are 2,753 kilotonnes using the OLS regression and 3,081 kilotonnes using the ML
regressions.

References
[1] Sinton, J.: “What Goes Up: Recent Trends in China’s Energy Consumption, Energy policy,
Vol. 26, No. 11, 2000.
[2] United Nations Development Program: Promoting Green Lights in China: Findings of a UNDP
Project, 2000.
[3] Min, G., Mills, E. and Zhang, Q.: “Energy Efficient Lighting in China: Accomplishments and
Challenges,” Energy Policy, Vol. 29, No. 10, 1998.
[4] Duke, R. and Kammen, D.: "The Economics of Market Transformation Programs," The
Energy Journal, Vol. 20, No. 4, 1999.
[5] Horowitz, M.: "Economic Indicators of Market Transformation: Energy Efficient Lighting and
EPA’s Green Lights,” The Energy Journal, Vol. 22, No. 4, 2001.
[6] Horowitz, M. and Haeri, H.: "Economic Efficiency versus Energy Efficiency: Do Model
Conservation Standards Make Good Sense?,” Energy Economics, Vol. 12, No. 2, 1990.
[7] Jaffe, A. and Stavins, R.: "Dynamic Incentives of Environmental Regulations: The Impact of
Alternative Policy Instruments on Technology Diffusion," Journal of Environmental Economics
and Environment, Vol. 29, 1995.
[8] Dhrymes, P. J. Econometrics – Statistical Foundations and Applications, Harper and Row,
New York, 1970.
[9] Johnston, J. Econometric Methods, Third Edition, McGraw-Hill, New York, 1984.
[10] Malinvaud, E. Statistical Methods of Econometrics, Second Edition, North Holland,
Amsterdam, 1970.
[11] Theil, H. Principles of Econometrics, Wiley, New York, 1971.

1366
Energy Efficiency Testing Facilities of Home Appliances in Egypt
Omneya Mostafa Kamal Sabry

Testing Dept., New & Renewable Energy Authority, Egypt

Abstract
Egypt is heavily depending on fossil fuel (oil and natural gas) to meet its growing energy demand.
This has led to limited hydrocarbon exports, one of the main sources of national income. The Energy
Strategy as one of the main components of the Egyptian Economical Development Strategy targeted
to decrease the growth rate of Energy Consumption using fossil fuel from 9% to 5.25% by the year
2017, through different Energy Efficiency Programs.
The Energy Efficiency Improvement and Greenhouse Gases Reduction Project (EEIGGR) is a main
project financed by UNDP/GEF to assist Egypt in reducing long term growth of GHG emissions from
Electrical Power Generation through a range of activities which include Standards & Labeling
Program. One of the building blocks of this program is the establishment of accredited testing facilities
with appropriate testing procedures. The EEIGGR has successfully developed the energy efficiency
standards for the three home appliances with the highest market penetration in Egypt namely:
Refrigerators, Washing Machines and Air Conditioners. Relevant Testing Facilities have been recently
built at New and Renewable Energy Authority (NREA) affiliated to Ministry of Electricity and Energy.
This paper will present the Egyptian Experience at NREA in performing Energy Consumption Tests to
Refrigerators and Washing Machines according to International and Egyptian Standards. This paper
will describe the main features of Testing Facilities, Capacity Building of Staff, Mechanism
implemented to ensure the enforcement and compliance with the standards and labels. The
Accreditation Process, Verification of Results with European Labs in addition to Impact on National
Trade Balance will also be presented.

Introduction
Egypt’s economic development outlook for the next decade calls for a challenging growth in Gross
Domestic Product (GDP) exceeding 6 % on average requiring almost similar increases in energy
supply. Since Egypt is an oil and gas producing country ,it is heavily dependent on fossil fuel to meet
the current growing energy demand. This is consequently increasing the rate of energy utilization to
cope with ambitious goals of economical growth especially with subsidized prices. This will constrain
hydrocarbon exports, one of the main sources of national income, and foreign currency. The rising
trend in energy consumption levels is likely to turn Egypt to a net oil importer soon. National studies
predict that the demand for crude oil is expected to double by 2017 but annual production will be
reduced by half. Although natural gas production is increasing, unless domestic consumption levels
and patterns are managed more effectively, Egypt’s supply of natural resources will be threatened
and so, its economic growth plans. [1]

Energy Efficiency in Egypt

o With the economic reform in Egypt in the early nineties, Energy efficiency awareness has
increased, especially with the annual growth rate of 5.5 % in energy consumption where statistics
has shown that this rate may reach 7 %, by the year 2017. Many studies have been conducted by
national and international experts which concluded the following: [2],[3]
• Any increase in economic activity by 1 %will eventually increase energy consumption by 1.3%
• Investing in Energy efficiency projects will lead to economic growth (Economic Multiplier
~ 2.32$) more than investing in production of energy (Economic Multiplier ~ 1.48$)
• With the Implementation of Energy Efficiency Projects a new market of related equipment will
arise accompanied with new skills and job opportunities (~ 25,000to 100,000) in O & M of
those equipment, in addition to job opportunities in the field of consultancy services.
• Investment in Energy efficiency in industry will decrease operating costs and consequently
will increase global competitiveness of product.

1367
 • The saving is not only as a result of reduction of energy demand but also by reducing the
investment required to meet the steadily increasing energy demand to meet economical
development objectives.
• The income to the facilities from energy saving can be reinvested in Human Resource
Development or in purchasing new equipment and facilities.
• In order to organize and coordinate efforts towards energy efficiency improvement, a
regulatory body becomes necessary.
o The Energy Strategy -one of the main components of economic and social development strategy
till 2017 - has been reviewed and formulated to include a strategy of energy efficiency
improvement. The overall objective is to decrease the annual growth rate of energy consumption
from 9 % to 5.25 % and to improve demand for oil and natural gas from 1770MTOE in 1998
to 1175 MTOE by the year 2017saving about 110 $ Billion (estimated for $ 25 barrel price).
o The current economic conditions in Egypt are creating a suitable climate for investing in energy
efficiency. Different Energy Efficiency Programs using proven technologies are being introduced
to the business community every day. These programs are usually nationally financed or
implemented in cooperation with international organizations. For example:[1]
• Cogeneration to generate electricity and steam
• Highly efficient lighting systems including local manufacturing of energy efficient lamps
• Power factor improvement using locally manufactured capacitors
• Efficient internal combustion engines, efficient combustion in boilers and improved thermal
insulation
• Waste heat recovery and reuse
• Energy Management systems in buildings

Energy Efficiency Improvement and Green House Gas Reduction Project

o The Energy Efficiency Improvement and Green House Gas Reduction Project (EEIGGR) is one of
the main energy efficiency projects implemented by Ministry of Electricity and Energy (MOEE) and
financed by UNDP/GEF, to assist Egypt in improving energy efficiency and reducing long term
growth of GHG emissions from electric power generation. One of the main activities of this project
is the standard and labeling program which has successfully developed in 2003 Egyptian Energy
Efficiency Standards to the home appliances with highest market penetration in Egypt namely:
Refrigerators, Washing Machines and Air Conditioners.[4]
o Main features of Egyptian Energy Efficiency Standards
Those standards aim at improving of energy consumption in appliances that have been selected
according to a certain criteria. These criteria included: penetration level in the Egyptian market,
average monthly energy consumption and amount of energy savings that may result from such
improvement. Refrigerators, Washing machines and Air conditioners testing requirements and
procedures have been set up for the standards which include also test conditions, measuring
equipment and max allowable values of energy consumption. Energy label that indicates monthly
energy consumption and levels of energy efficiency are now issued, where manufacturers are
committed to stick the label on the appliance to allow to the customer to select the appliance with
less energy consumption. It has been decided that these Energy levels should be reviewed every
two years for efficiency improvement. [5]
o Energy Efficiency Testing Facilities at NREA
One of the building blocks of standard and labeling program is the establishment of accredited
testing facilities to perform energy consumption testing to home appliances following the issued
standards. New & Renewable Energy Authority (NREA) as independent entity was selected to
host those testing facilities and will be responsible for conducting energy efficiency tests according
to Egyptian Standards and to the regulatory decree issued to enforce the standards and labels .
NREA affiliated to MOEE acts as the national focal point for expanding efforts to develop
renewable energy technologies to Egypt on a commercial scale together with the implementation
of related energy efficiency programs. Through its Testing and Certification Center performance
tests are performed for the evaluation of renewable energy equipment performance as well as
other equipment for energy efficiency purpose. [6]
o The UNDP has complemented the ongoing UNDP/GEF national and standards and labeling by
providing assistance to establish the energy efficiency testing facilities. NREA contracted -for a
turn key job- an Egyptian Supplier who designed the system and supplied the relevant equipment
for Refrigerators and Washing Machines according to technical specifications prepared by an

1368
 International Consultant from the Collaborative Labeling and Appliance Standards Program
(CLASP). The equipment was imported from different well known European and American
suppliers. [7]

Refrigerators Testing Facility at NREA

o This testing facility performs energy consumption tests on refrigerators with capacity up to 1100
liters and on freezers up to 850liters. Locally manufactured or imported refrigerators and freezers
can be tested according to International Standards (ISO5155, 8187 and 8561) and Egyptian
standards (ES 3794) which comply since 2005 with International standards. The test room is
fabricated from fiberglass material and is thermally insulated up to 30 º C and its size (3 * 4 * 2.85)
is sufficient for testing two Units Under Test (UUT) simultaneously.[8]
o Power supply to each UUT (with 1350 VA max. power) is controlled to provide voltage in the range
from 0 v to 270 v and frequency in the range from 45 HZ to 500 HZ with accuracy + 1% 1 .The
power supply is equipped with a power transducer with error reading 0.2± 0.04FS with maximum
error 0.8watt to measure electrical parameters supplied to the UUT.
o Test Room Preparation and Testing Conditions
• The room ambient temperature is adjustable from10 °C to 55°C within an accuracy ± 1° C. It
is controlled to maintain temperature within 25C ± 1° C with gradient < 1° C and accuracy 0.2
% of span via PID controllers used to activate heater or cooler. Temperature measuring
devices are shielded so that the indicated temperature is not affected by the operation of
condensing units.
• Humidity of test room is maintained at the standard range where it can be increased using
steam generator activated by a solid state relay to control humidity in the range from 45%
to 75% with accuracy 2% .
• Air speed inside the test room is controlled using a high efficiency distribution system. Air is
restricted to flow along the walls and floor and away from the space occupied by the UUT.
This technique is used to ensure that air circulation in the vicinity of the UUT is constant and
does not exceed 0.25m/s. Fig. (1)

Figure 1: Air circulation in refrigerator test room

In table (1) and table (2) we compare accuracies required for measurements and instruments as
specified in ISO 8187 & ES 3794 with those provided by the testing facility [9], which shows how we
exceeded those accuracies in supplied equipment

1
In this testing facility we can test locally manufactured refrigerators that can be exported to those countries that have got 220
v or 60 Hz supply

1369
Table 1: Comparison between accuracies of measurements required in standards with
accuracies of measurements provided by the testing facility
Test Room Required accuracy Required accuracy NREA
Conditions according to ISO according to Egyptian Measurement
Standard Standard Accuracy
1- Voltage + 1% + 1 Volt + 0.13 Volt
2- Frequency + 1% ‫ــــــــــــــ‬ + 0.05 HZ
3- Ambient + 0.5 °C + 1 °C + 0.24°C
Temperature
4- Humidity ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ +2%
5- Power ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.4 W

Table 2: Comparison between accuracies of instruments required in standards with

accuracies provided by the supplied equipment of the testing facility
Required accuracy Required accuracy NREA
Test Room Instruments according to ISO according to Instrument's
Standard Egyptian Standard Accuracy
1- Voltage Transducers ‫ــــــــــــــ‬ + 0.5 % of reading + 0.25 % of F.S
2- Frequency ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.1% of Span
Transducers
3- Current Transducers ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.25 of F.S

4- Energy + 1 % of F.S + 2 % of reading + 0.2 of reading +

0.04 of F.S
5- Ambient Temperature ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.25 % of Span
Controller

6- Humidity Controller ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.25 % of F.S for

Voltage
+ 0.7 % of F.S for
Current
5- Temperature Sensors + 0.3 °C + 0.5 °C + 0.1 °C
Inside The
Refrigerators
o Test Unit Preparation
• A wooden platform and separators painted dull black have been provided to hold the UUT at
30 cm height from floor and 60cm from sides, ensuring free air circulation around the UUT
according to standards.
• Temperature sensors of RTD type (4wire, 0.1° C accuracy) are connected to test loads
distributed in the freezer and cabinet according to standard, as shown in Fig.(2). Storage and
equivalent volume of the UUT are calculated according to relevant testing procedure2
It is preferred to operate the UUT in an entry room for about 6hours before introducing it inside
the test room. Thermostats or temperature control devices are adjusted before the test at the
manufacturer recommended positions. If the UUT has got an ice making unit, it must also
operate during the test period.
• Before the software starts logging data and calculating energy consumption , the UUT must
reach steady state equilibrium testing conditions indicated by:
Temperature inside the cabinet at half thermostat = +5°C,
Temperature inside the freezer = - 6°C to - 18°C depending of no of stars mentioned in the
fridge operating manual,
Compressor operates for two or three on-off cycles

2
For ES 3794: equivalent volume = storage volume of cabinet + correction factor * storage volume of freezer (correction factor
is function of ambient temperature and temperature of cabinet and freezer), for ISO 8187: equivalent volume = storage volume
of cabinet * correction factor + storage volume of freezer * corre
ction factor (correction factor is function of cabinet temperature)

1370
Figure 2: Distribution and connection of
RTD's to thermal loads in refrigerator
cabinet and freezer
o Data Acquisition System (DAS) and Software Features
The facility is supported by a DAS to collect the data from sensors .Before reaching steady state
all on line data are shown namely voltage, current frequency, temperature and humidity inside
the test room in addition to the temperature transmitted via RTD's .As soon as we set up Test
period, Standard Type for test (ISO or ES) and type of fridge (with or without defrosting) , we can
then start logging data. All above data are monitored in addition to on-off cycles of compressor
and ratio of total time of compressor operation to total test period named Run Time Ratio 3,as
well as Power Factor with logging interval that starts from 1 sec4
o Determination of UUT Energy Consumption and Class Level
At the end of the test we get total energy consumption during the test period .Energy
consumption for the whole year is calculated ( E annual = E daily * 365) as well as the class
according to both Egyptian and ISO standards 5
Table (3) compares results obtained from intercomparison laboratory tests conducted during the
Acceptance Tests of the testing facility, where a difference of 1.53 % of results shows the level of
reliability and reproducibility of the testing facility

Table 3: Intercomparison laboratory test results: NREA laboratory and SLG Pruf-Und
Zertifizierungs GmbH Testing laboratory
SLG Pruf-Und NREA Testing
NO. Item ZertifizierungsG lab.
mbH Testing lab.
1- Test duration 24 : 00 : 00 24 : 04 : 48
2- Energy consumption per day 1728.06 wh 1701.77 wh with
without defrosting defrosting
3- Annual energy consumption 630.74 Kwh/year 630.74 Kwh/year
4- Average Temperature in The Cabinet 5.1 5.11
5- Average Temperature in The Freezer -18.83 - 17.24
6- Percentage Running Time 48 % 45.9 %
7- Class C C

Washing Machines Testing Facility at NREA

This testing facility performs energy consumption tests to automatic or half automatic washing
machines with capacity up to 10kg. Locally manufactured or imported washing machines can be

3
In 8187standards this ratio excludes period of heater in operation
4
PF < 1when compressor is operating, PF = 1 indicating that it is the period where the compressor is OFF and the heater is
ON during defrosting period, PF = O in case of UUT without defrosting.
5
In ES 3794: Class = annual energy consumption / equivalent volume (KwHr/L) ,In ISO 8187: Class = (Annual Energy
Consumption/Standard Energy Consumption) *100

1371
tested according to international standard (IEC 60456) and Egyptian standard (ES 4100) which
complies since 2005with international standards.[8]
o Test Room Preparation and Testing Conditions :
• The test room is thermally insulated and has same dimensions as the refrigerators test room
and is capable of testing two UUT’s simultaneously Fig (3). Washing Machines testing facility
is equipped with two power supplies with same capabilities as in refrigerators but with max
power 3000 VA, to set electrical parameters needed to operate the UUT.

Figure 3: Washing machine testing

facility showing also temperature and
humidity sensors

•
In Washing Machines Testing Facility we don’t have any restrictions regarding air
flow around UUT. Consequently, we control ambient temperature inside the test
room using normal air conditioner to provide room temperature around 25 °C as
specified with accuracy + 0.6 º C
• Water Pressure inside washing machine is set to 240 Kpa, a pressure transmitter
detects the difference between measured and required pressure of water activating
a frequency inverter used to control the speed of a pump in the range from 0.5 to 3
HP with accuracy 0.1%.
• Temperature of water pumped to the UUT is controlled within specified values 6
6
using a water tank where detected low temperature water can be heated via a pump
and heater. For detected high temperature water, a refrigeration system with a timer
shall operate to cool the water and do the circulation of water inside the tank.
In table (4) and table (5) we compare accuracies required in standards for measurements and
instruments with those provided by the testing facility [9], which shows how we exceeded
those accuracies in supplied equipment

Table 4: Comparison between accuracies of measurements required in standards with

those provided by the testing facility.
Test Room Required accuracy Required accuracy NREA
Conditions according to IEC according to Egyptian Measurement
Standard Standard Accuracy
1- Voltage + 2 % ( 4.4 Volt) + 2 % ( 4.4 Volt) + 0.95 Volt
2- Frequency + 1% ( 0.5 HZ) + 1% ( 0.5 HZ) + 0.025 HZ
3- Ambient + 5 °C + 5 °C + 0.6°C
Temperature
4- Humidity ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ +2%

6
For ES 4100: temperature of water inside the UUT = 25 ° C, For IEC 60456 temperature of cold water inside the UUT= 15° C
and temperature of hot water inside the UUT= 60 °C.

1372
Table 5: Comparison between accuracies of instruments required in standards with those
supplied by the testing facility
Required accuracy Required accuracy NREA Instrument's
Test Room Instruments according to IEC according to Accuracy
Standard Egyptian Standard
1- Voltage Transducers ‫ــــــــــــــ‬ + 0.5 % of reading + 0.25 % of F.S
2- Frequency Transducers ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.1% of Span
3- Current Transducers ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.25 of F.S

4- Energy ‫ــــــــــــــ‬ + 2 % of reading + 0.2 of reading +

0.04 of F.S
5- Ambient Temperature ‫ــــــــــــــ‬ ‫ــــــــــــــ‬ + 0.25 % of F.S for
Controller Voltage
+ 0.7 % of F.S for
Current
6- Humidity Controller + 3 % of F.S ‫ــــــــــــــ‬ + 0.25 % of F.S for
Voltage
+ 0.7 % of F.S for
Current
5- Temperature Sensors + 0.5 K + 0.5 °C + 0.1 °C
Inside The Washing
Machine
6- Water Flow + 1 % of F.S + 2 % of F.S + 0.3 % of F.S
7- Water Pressure + 5 % of F.S ‫ــــــــــــــ‬ + 0.25 % of F.S for
Controller Voltage
+ 0.7 % of F.S for
Current
8- Mass + 0.1 % of F.S + 0.3 % of F.S + 0.01% of F.S
9- Water Temperature + 1 K of F.S ‫ــــــــــــــ‬ + 0.25 % of F.S for
Controller Voltage
+ 0.7 % of F.S for
Current

o Test Unit Preparation

The test is run on UUT without detergent where it is set for fully automatic cycle for cotton
fabric at 60 º C without pre-washing cycle. The UUT is loaded 7 by textiles that are pre-
washed at least 20 cycles and not more than 60cycles .The UUT is subject to five complete
tests for every complete washing cycle according to standards. Average annual energy
consumption (for 240cycle /year) is determined8. The class is determined as the ratio of
energy consumption to the capacity of UUT in Kwhr/Kg

In table (6) we compare results obtained from intercomparison laboratory tests conducted during the
Acceptance Tests, where a difference of 6.4 % of results shows the level of reliability and
reproducibility of the testing facility

7
Maximum capacity for full automatic and 75 % for half automatic
8
A correction factor is mentioned in the standard to calculate energy consumption for full automatic machines with no control
switch of temperature equal to 0.6 or those which having a non separated pre- washing program equal to 0.85

1373
Table 6: Intercomparison laboratory test results tests: NREA laboratory and DEA laboratory
DEA NREA Testing
NO. Item Testing lab.
Lab.
1- Test duration (Time of Program B) 02 : 07 : 12 02 : 14 : 36

2- Inlet water temperature 22 ° C 25 ° C

3- Water pressure 270 Kpa 245 Kpa

4- Water consumption ‫ــــــــــ‬ 77.968 L

5- Heater Current 7.88 A 8.05 A

6- Motor Current 1.2 A 1.17 A

7- Motor stage pressing current 2.9 A 2.48 A

8- Power of electric heater 1.68 Kw 1.779 Kw

9- Power of electric motor 0.26 Kw 0.22 Kw

10- Weight of wear 5.44 Kg 5.14 Kg

11- Temperature of washing machine water 60 ° C 50.09 ° C

12- Average Energy Consumption from Program B 1.03 Kwh 0.9842 Kwh

13- Average Energy Consumption from Program B/Kg 179.1 Wh 191.49 Wh

14- Class C D

o Data Acquisition System (DAS) and Software Features

The facility is supported by a DAS to collect the data from sensors .On the PC we can monitor
the same parameters mentioned in the refrigerators testing facility namely electrical parameters,
room temperature and humidity. Pressure and temperature of water inside the UUT are also
displayed at the set log interval that starts from 1 sec. Fig. (4)

Figure 4: Data acquisition system of

washing machine test facility

Common Features
o A Calibration Kit has been supplied to calibrate different parameters used in the system
including the following:

1374
 •
A Pressure calibrator used to calibrate: Pressure Transmitters (P-I and P-V), Pressure
Converters (P-P and I-P), Pressure switch and relief valves
• A Multifunction Calibrator to calibrate electrical parameters in the range of the supplied
equipment, pressure in the range from (- 1 to2O bar),humidity from (0 % to100 %),
Thermocouples and RTD's an also be calibrated in the range from ( -50° Cto 650° C)
o Each Testing Facility has been equipped with a manual system ( 0.05 % accuracy) supplied as
backup to the computerized system.. This system can record all parameters data for 24 hrs and
calculate daily energy consumption .This manual system has also 64analog channels and 64
alarm set points. It has also 48 digital inputs and 48alarm relay outputs in addition to an event
alarm system. Fig. (5)

Figure 5: Back up manual system

supplied to record all measured
parameters in parallel with
computerized system
• The Testing Facility has safety features including a fire alarm system, in addition to an "under
and over temperature alarm" to shut down the whole system. The software has a flashing
lighting in case of out of range parameters or in case of turn off of any equipment. Moreover,
as soon as we start logging data to calculate energy consumption human intervention is not
allowed to stop logging data.

Mechanism adopted for Enforcement of Standards and Labeling

o An agreement with Egyptian Organization for Standards and Quality EOSQ -as a regulatory
body- is being reviewed where NREA will receive Refrigerators or Washing Machines randomly
selected by EOSQ at the manufacturer site .This mechanism is expected to ensure the
enforcement of standards and labeling The quality management system established since 2004
at NREA Testing and certification center ensures effective lab management according to
ISO9001/2000 requirements.

o The reports issued from NREA will indicate the energy consumption and the class of the
appliance for the manufactured model. Those data should be mentioned on the label that will be
issued by the manufacturer prior to distribution in the stores. This label will be subject to
investigation in retail shops by another regulatory body affiliated to Ministry of Internal Trade

Accreditation Process
o The Industrial Modernization Program financed by EU has adopted accreditation of the new
testing facilities. Quality management system according to ISO 17025 requirements has been
established including quality manual and testing procedures.
o Work instructions, Inventory and log file of all testing equipment, maintenance plan and records
of troubleshooting, Calibration records are currently applied. Accreditation of the testing facility
is expected by July 2006.

1375
 o NREA is now seeking international accreditation to save high fees paid by manufacturers to test
locally manufactured appliances abroad prior to exporting

Capacity Building
Three engineers and two technicians have attended a classroom and on-site factory training to learn
operating and calibrating testing equipment and to get acquainted with Egyptian and international
standards of testing procedures. An on-the-job-training has also been conducted during the
installation phase and during commissioning to perform the complete test according to both standards
in addition to O&M procedures. They have also been trained to monitor and analyze test results and
issue reports, in addition to a hands-on training to calibrate testing parameters of the whole testing
facility using supplied calibration equipment. Accreditation requirements according to ISO 17025have
been taught to testing engineers including methods of calculation of uncertainty as one of the main
requirements of testing credibility.

Impact on National Economy

o Testing of locally manufactured appliances has shown the application of improved
technologies that have dramatically improved their energy efficiency. [2]
• For the refrigerators: use of high efficiency compressors and increased insulating
material width results in savings around 40 % of the current energy consumption.
• For the washing machines: one of the local manufacturers uses Water Jet system
instead of the normal system with results obtained showing 20 % saving of energy
consumption .Other improvements include using highly efficient compressors,
special sensors to control water consumption, and decreasing inter drum space.
These improvements can provide almost 30 % energy and 20.8 % water savings
o Moreover by applying standards and labeling to home appliances we expect a great impact
on national economy as following:
• The Egyptian market will be protected from low energy efficient products and
consequently will create a competitive market for highly efficient products.
• Bringing benefits to the consumer through less spending on electric bills
• Decrease of GHG emissions.
• Reduction of investment costs needed to build Power plants to meet growing
requirements of energy
• Improved national balance of payments due to savings of fossil fuel for exports and
petrochemical industries

References
[1] Egyptian Environmental Policy Program (EEPP) Energy Efficiency Strategy, 2000
[2] Egyptian Environmental Policy Program (EEPP) Energy Efficiency Utilization Report: What it
means to National Economy?
[3] Organization of Energy Planning Energy Efficiency Council
[4] Dr. Yassin I. Implementing Status for Energy Efficiency Improvement and Green House Gas
Reduction (EEIGGR) Project, 2nd International Conference for Scientific and Applications/ Cairo
University, December 2005
[5] Http://www.oep.gov.eg
[6] New & Renewable Energy Authority (NREA) Annual Report 2004 - 2005
[7] UNDP Office, Cairo Egypt Project Document Signed for The Establishment of Home
Appliances Testing Facilities at NREA, EGY / 03 / M22,2004
[8] UNDP Office, Cairo Egypt Tender Document Issued for the Establishment of Home Appliances
Testing Facilities at NREA, 2004
[9] Instrumentation and O&M Manuals for the Established Home Appliances Testing Facilities at
NREA

1376
Residential Gateway Standardisation
Milan Erbes

SPiDCOM Technologies

Abstract
This paper will present the standardisation work being done in ESTI, with NGN@Home and TISPAN
and work done within the CENELEC SmartHouse Project.
To be able to understand the residential gateway concept we have to look into today’s residential
access environment with different network access technologies coming to the home via the existing
twisted pair telephone network (xDSL), the coax (and fiber) cable TV networks and emerging wireless
access. Coupled with this today we also have a direct broadcast satellite networks, switched digital
video networks, PCS networks, and probably others. These are probably not all at the same place
and same time and also these are not only competing networks, but they are incompatible at various
levels of the OSI model. This would mean that they are incompatible in more than their basic physical
interface characteristics.
This incompatibility and competition among those networks that we have are in functions, price,
availability, which means that the residential customers will have to work with a complicated
connections and switching problems. So, the solution is to design a technology-based solution that
will handle this function and hide the complexity from the consumer.

Home Networking Trends

Broadband gateways are devices of the future and they have much to offer but at the moment they
are still in the laboratories and for some time to come. The innovations in telecom industry have
reshaped the way we communicate, our nature will still ask for more connectivity, more bandwidth,
more services, and more scalability and what we want is everything, not just certain elements. If we
have a high-speed Internet access at work, than why can’t we also have it at home?
The demand goes not only for a high-speed access, but also other applications such as linking
telecommuters with corporate offices, to controlling a home alarm system from the other side of the
country, turning the dishwasher ON from the office, or downloading a movie from one computer to
another within our homes.
Reliable broadband services delivered to the home will enable a variety of applications that will enrich
our quality of life with new multimedia services for voice, video, videoconferencing, interactive
gaming, high speed Internet access, telecommuting; services for white goods, metering, health care
for the ill, elderly and disabled, security, monitoring and intelligence.
While aiming for more functionality out of a high-speed connection, the challenges service providers
are facing is finding a way to uniformly serve customers with different demands. Many providers are
turning to broadband gateways as a solution that could bring together multiple technologies without
compromising the end user service. This new “digital home” is the converging point of multimedia,
computing, Internet and telecommunication industries.
The Home Residential Gateway market could be separated into three types of device:
• Home Residential Gateway - addresses entertainment (video and audio), communications
(telephony), high-speed data access, and control and monitoring (HVAC, security, lighting,
etc.) functions.
• Internet Residential Gateway – This is intended to connect multiple computers, multiple high-
speed access problems in the home.
• Set-Top Box Residential Gateway – Are the CATV digital solutions. They are new generation
boxes that came from the analogue TV set-top box with more capabilities and features.
The Home Residential Gateway has a control function between external networks and in-home
networks and devices. It serves as a traffic control and routing device and has these key functions:
• The location to terminate all external access networks to the home, with multiple residential
services being delivered over any type of access network.
• The location to terminate all home networks, such as telephone, television, computing, alarm,
telemetry, data, etc.
• To seamlessly interconnect those public and private networks

1377
The network interface units are selected to match the appropriate external network. The customer
premises interfaces are selected to match internal home networks or specific home appliances. The
processor serves to operate the back-plane as a switching point to allow highly flexible translations
and interconnections while hiding the complexity from the customer
The residential gateway enables consumers to connect electronic devices such as PCs, kitchen
appliances, audio/video equipment and security systems to their phone service and high-speed
Internet access. By connecting the electronic devices, consumers can perform tasks such as adjust
heating or air conditioning via a home computer. The home networking products came out a while
ago, but what we are missing is the home residential gateway to tie everything together. Using the
home residential gateway, users can share a single broadband connection for all PCs in their home
and share that connection, by setting up home LANs, share files and play interactive games.
The concept of Intelligent Homes and Business is intended to provide solutions for automation in
homes, buildings and related areas to assist people in their living environment. Standardized solutions
will make it easier to supply services and networking, together with numerous new applications
including remote control of home automation, alarm systems and assistance to people with special
needs. Communications’ infrastructures will form an essential part of the support to these
applications.

Home Area Networks

The Home Area Network is a short-range communications system designed for the residential
environment and is independent of the home network physical (PHY) layer medium, where this may
include, but is not limited to:
• Co-axial cable
• CAT 5 and CAT 6 cable (twisted pairs - balanced)
• Telephone wire (unbalanced)
• Power line communications
• Optical fibre
• Radio Frequency (RF) wireless
• Infra Red (IR) wireless
The home network physical layer protocol and data link protocol may include, but are not limited to:
• Bluetooth™
• Ethernet or IEEE 802.3
• HomePlug™ (Home Powerline Alliance)
• HomePNA™ (Home Phoneline Networking Alliance)
• HomeRF™ (HomeRF Working Group)
• IEEE 802.11a/802.11b
As can be seen from the above list, there are many network technologies; it may not be possible to
deliver all Home services over the network technologies listed above.

Higher Layer Protocols

TCP/ Other Transport

UDP Protocols
Transport
Internet Protocol
(IP) Network
Network Specific
Data Link
Data Link
Network Specific
Physical
Physical
Figure 1: Home Network Protocol Stack

Typically, devices within a home network communicate with each other using a peer-to-peer
architecture as opposed to the client/server model that is used in corporate networks. In peer-to-peer
networks, devices can connect to each other directly without a server as an intermediary. In
client/server architecture, all devices connect to a central server, which provides services like Internet

1378
access, applications and file sharing. Many more devices can connect in a client/server network than
in a peer-to-peer network.

Existing Communications Cabled Infrastructure “No New Wires” Networks

Phone Line Networks

This technology uses a home's internal phone wiring to connect one device to another. Most phone
line networks require that all devices be connected to the same phone line (i.e., the same pair of
wires). A phone line can carry multiple signals at different frequencies, which is why one can talk on
the phone and still use a DSL modem for Internet access. Network traffic over phone lines uses
frequencies higher than DSL or voice bands. This, de-facto industry standard for this technology, is
set by the HomePNA a Phone-line Networking Alliance offering a quality of service (QoS), which
enables real-time audio and video precedence over other data.

Power Line Networks

Since power outlets are more widespread in a home than phone jacks, a network that makes use of
electrical wiring could be more practical. Power line networks send data signals through existing AC
power lines. The challenge is that data transmission on the AC power lines is highly susceptible to
interference and noise generated by household appliances (e.g., hair dryers, dishwashers, etc.). The
HomePlug Alliance sets the industry standard for higher data rate networking over home AC wiring.
This consortium is developing standards for a second generation of power line network products with
10 Mbps transfer rates and increased reliability. The HomePlug Alliance has several focus areas:
HomePlug 1.0 + AV (in-home connectivity, including digital home and consumer electronics
applications), HomePlug BPL (to-the-home, Broadband-over-power line applications) and HomePlug
Home Automation (command-and-control applications).

Wireless Networks
IEEE 802.11 is growing in popularity and support as prices continue to fall and as consumer
awareness grows through the efforts of the Wi-Fi Alliance (an organization established to promote
interoperability between IEEE 802.11 devices and to bestow certificates of conformance to those
devices implementing a baseline subset of IEEE 802.11). It also supports other computing devices
such as PDA's. Wi-Fi IEEE 802.11 is also being adopted for wireless connectivity in public places,
college campuses and the workplace, using the same network adapters one would use in the home or
in business.
Bluetooth is one of the latest protocols that came into the wireless arena. Imaginatively named after a
Viking King, this short-range wireless protocol is not aimed solely at the home network but the
personal area network (PAN). Bluetooth is primarily meant to reduce the number of cables required to
interconnect small devices like mobile phones, PDAs, laptops, keyboards and headsets to each other.
Bluetooth network can connect to access points in pre-existing networks within a range of about 10 m
(30 ft) with the data transmitted at a maximum of 1 Mbps.

1379
Mobility Network Home Area Network (HAN) “Customer Premises” Access Core Networks
(e.g. In-Car Network)

TE TE Broadcast Audio
Existing Comms Cabled
Video N/W
Infrastructure (HAN) Broadcast Access
Network(s)
TE
IP Data N/W
Auto Area
UE TE
Network Home Area Other MM
Network N/W (IMS)
Home Area BB Access
Network Network(s)
UE RG
ETSI TISPAN
GW NGN
TE Home Area
TE Network

UE
Home Area GW Other
SCT Access NGN
Auto Area Network
Network(s)
Network PSTN /
UE ISDN
Either one (not both) of these
UE Home Area links may be provided
Network Session Control
UE Interface (Signalling)

Figure 2: NGN @ Home Interfaces

There are multiple interfaces at the boundary to the Home Area Network and within the HAN.

Access Networks

The definition of the access network(s) is outside the scope of NGN @ Home. However, the
interaction between the HAN and access network has to be managed at several levels to:
• Ensure privacy within the Home domain;
• Provisioning of end devices that may be required for specific services;
• Requests from end devices for resource allocation (QoS aspects);
• User profile management;
• Service level management;

NGN @ Home is required to support access networks of diverse technologies and capabilities.
Regardless of the type of access network technology, the NGN @ Home communications and
services are required to be available to all end users.
Whilst profiling the specific characteristics of the various access networks is within the scope of the
NGN @ Home work, the QoS and Security specifications under Release 1 will only cover the
following access network technologies:
• Switched Circuit Technology (SCT) – PSTN/ISDN/GSM
• Digital Subscriber Line (xDSL)
• Hybrid Fibre Co-ax (HFC)
• Satellite and Terrestrial Broadcast
• WLAN (used in local loop) – IEEE 802.11x

Access Network Assumptions/Requirements

The access network is defined as a collection of network entities and interfaces that provides the
underlying IP transport connectivity between the device and the NGN core entities. Examples of
“Access Network” are given below. It should be noted that this list is not an exhaustive list and other
access network may be considered.

• xDSL: this includes ADSL, SDSL and VDSL transport systems and supporting
concentration/multiplexing technologies. The TISPAN NGN will provide direct support for this
access requirement through interfaces to control resources (QoS) and coherent authentication.
• Wireless LAN networks are required to be supported by the TISPAN NGN. For Release 1 they
are regarded as just another type of access network. The TISPAN NGN interface to the wireless
LAN network will be at a fixed (not radio) point. Requirements for this support are similar to the
support for XDSL access networks.

1380
• Cable networks may be supported as another type of access network, but no interworking
requirements are placed. TISPAN NGN will not control resources or authenticate access to this
network by terminals and users.
• Fixed IP Connection (e.g. Gigabit Ethernet link to corporate network). TISPAN NGN assumes that
this interface is a direct connection to a customer owned and managed LAN or MAN. Control is
limited to admission control. Example scenarios also include FTTC (Fibre to the curb) and FTTH
(Fibre to the home).
• 3GPP PS domain. TISPAN NGN shall support the 3GPP PS domain at the same interface point
as other access networks, see Figure 2. Network attachment and associated functionality is
supported in exactly the same way as in a 3GPP network. In this sense a TISPAN_NGN supports
any 3GPP IP-CAN. TISPAN NGN does not support the CS domain as an access system. All
interfaces between the PS domain and the IMS are maintained unchanged in the TISPAN NGN.
The 3GPP PS domain is not re-documented as part of TISPAN NGN specifications.
All NGN capable access types are required to offer IP connectivity.

P rim a ry
Ho m e
D o m a in
P u b lic D o m a in

Extended
Home Domain Firew al
Vehicle l WL AN
HOT EL
Hot spots
C apable of S e c o n d a ry S h o p p in g M a ll etc
attac hing and
detac ting from
Ho m e
Hom e Netw ork D o m a in

User's
PAN

Firew al
Extended
l Wo rk p la c e D o m a in
Home Domain
Yacht M u ltip le
Ho m e
Not c apable of
attac hing to Hom e
D o m a in s User's
Netw ork PAN

Firew al
l

Figure 3: Multiple Home Domains

Home Residential Gateway is an IP based device that enables a Service Provider to offer enhanced
set of home network services with QoS, Device and Service Discovery, Security, Firewall,
Provisioning & Management. It also enables users to connect devices like PCs, kitchen appliances,
audio/video equipment and security systems to a high-speed Internet Access.
At the same time Home Residential Gateway enables users to share a broadband connection for all
PCs, by setting up home LANs, share files, printers and play interactive games and also is able to
provide solutions for automation in homes, buildings and assist people in their living environment.

Home Residential Gateway Reference Architecture

The Next Generation of in-home network should support a multiple of different home networks based
on IP but also the other non-IP based networks that should be interconnected into one overall
structure, represented in Figure 3. Connection to the Wide Area Network should be done through a
single Residential Gateway Architecture that on one side would support different access technologies
and on the home side it should connect to other home routers/switches but also legacy networks
through separate gateways.

1381
 P rovider 1 P rovider 2 P rov ider 3

Internet
L o cal Ag reg ato r

R eside ntia l G ate w ay

F ire w all, D R M , S ecu rity ....
H o u se IP b ack-b o n e n etw o rk

H om e Lo ca l G a tew ay 1
Loc a l ne tw ork3
Loc al
R oute r 1 R outer 2
Hom e
netw o rk1 Hom e N on-IP
ne tw ork 4 O pera tions &
T V /AV C ontrol
H om e N etw ork
ne tw o rk2

Figure 4: Home and Access Network Model

From the Figure 3, it is clear that there are four fundamental areas that are the key to the
NGN @ Home:
• Multiple Access Networks (using both IP and non-IP native access technologies)
• Multiple Home Area Networks (including the control domain, the transport domain, and the
applications environment)
• Multiple Services and Service Providers (multiple content streams and formats)
• Multiple End User Devices and Terminal Equipment connected within the Home Area Network

The NGN@Home Project areas:

• General Model. The Home Network Architecture targets at supporting a wide range of
services, from legacy telephony to new generation services, such as: audio, data, video
broadcast and conversational services, streaming services, interactive gaming.
• Functional Architecture Model. A distributed functional architecture with the use of a new
set of protocols to control user sessions (identification, authentication), resource allocations
and QoS, the traffic (policing, enforcement), services and applications, between various
entities within the home network, Packets filtering (depending on "IP address/port”, i.e.;
firewall functionality), Packet marking, Resource allocation and Bandwidth reservation,
Allocation and translation of IP addresses and port numbers (NAPT), Throughput limitation,
and optionally Media Ciphering/Deciphering, Media Trans-coding, Media flows topology
(conferencing, redirection of flows, etc.), User Authentication, Usage metering, IP address
allocation.
• End-to-end Quality of service
• Service platforms (including APIs)
• Network Management
• Security & DRM/CA

The NGN@Home Project functions and features:

• Easy to acquire, store, and access digital music. From anywhere in the home access to
digital music collection stored on multiple, network-enabled devices, including PCs, “virtual
jukeboxes” and portable audio players and for playback on any network-enabled playback
device in the home.
• Easy to manage, view and share digital photos. The wireless download feature transfers
all the photos to a media archive on a PC that will distributes the photos to photo frames, PC
screensavers, TV adapters, and other devices throughout the home. It even securely sends
the images across the Internet to friend or family.
• Distributed, multi-user content recording and playback. Using a universal remote, access
any of the network-enabled set-top boxes, PCs, or TVs in the home and select programs for
viewing, or for recording and later playback, utilizing available tuner resources embedded in
network-enabled TVs, dedicated PVRs, set-top boxes, and PCs.

1382
In order to deliver on digital interoperability in the home, it is required to support:
• Transparent connectivity between devices inside the AV Home Network;
• Unified framework for device discovery, configuration and control;
• Interoperable media formats and streaming protocols;
• Interoperable media management and control framework;
• Compatible authentication and authorization mechanisms for users and devices;
• Home control, communications, and more advanced entertainment scenarios.

End users can:

• Select the service provider independent of the access mechanism;
• Different end users within the same HAN can select different service providers;
• Roam between delivery networks, based upon their subscription profile with the selected
service provider.

For example: a user can register with the service provider for delivery of content when connected via
neighbours’ HAN.
The objective is to define an open network architecture enabling the provision of services to the users
in its home and in nomadic situation. The basis for a multi-service home network based on a Home
Residential Gateway (HRG) that acts as a service platform for the end user needs to be defined. The
HRG has embedded Agent that could allow a remote management of the home network by the home
network service operator.

C o n t ennt t PPr roovvidid er rss

CCo on nt et e
n t P r o v id e e
rs

A ccess A ccess
P r o v id e r A . . . P r o v id e r Z

C o n n e c t io n C o n n e c t io n
M o d u le G a te w a y M o d u le

P o lic y C o o r d in a t io n
M o d u le

H A N M o d u le

Agent . . . Agent

E n d D e v ic e E n d D e v ic e

Figure 5: Home Gateway Architecture overview

Home Residential Gateway coordinates shared access to the Internet for all of its End Devices. It has
four physical/logical modules: Connection Module, Access Module, Policy Coordination Module and
HAN Module.

A: Connection Module. Provide content to End Devices, through Access Provider(s) and through
out-of-band Broadcaster.

B: Access Provider Module. Provide Access to Internet for the Connection Module(s) on the
Gateway coming via: xDSL, HFC, PLC, xTTH, FWA, GPRS, UMTS, or any other new access
technology. He may provide local advertisement insertion.

C: Policy Coordination Module coordinates changes amongst the other modules. It handles NAT,
DHCP, IP Routing, DRM and CA Rights, Multicast optimisation, Network Address Translation, Quality
of Service & Security. It makes appropriate adjustments to the other Modules as needed.

D: HAN Module. Home Access Network (HAN) module might be separate hardware like a Hub,
bridge, switch, access point. It ties the various home networking media together. It should support
Ethernet, PLC, IEEE 1394, 802.11, HPNA, Bluetooth, USB, coaxial cable. Its role is to notify the
Policy Coordination Module of pertinent changes.

1383
 P ro g ra m m a b le
W E B In te rfa c e R e m o te S o ftw a re SNM P
M anagem ent A P I’s (H U C L ™ ,
Java, HTM L… S u p p o rt & U p g ra d e M anagem ent
H e a v y d is a b le d , ..)

V o ic e S IP , H .3 2 3 , V o ic e o v e r D S L , V o ic e o v er C a b le

M u ltim e d ia M PEG 2/ M PEG 4 over RTP , SDP, RTP, RTCP, RTSP

VPN V P N S e c u rity – P P T P , IP S e c , L 2T P , IK E , S e c u rity L ib ra ry, H a rd w a re C ryp to g ra p h y

T ra ffic P rio ritiz atio n , C la s s-B a s e d Q u eu in g , R S V P , V P N p a s s th ro u g h ,

E th e rn e t R A D IU S A u th e n tic a tio n , P a c k e t F ilte rin g
S ta c k
P PP oE DHCP DNS N AT B rid g in g R IP F ire W a ll

A c c e s s N e tw o rk HAN
B lu e
In te rfa c e xD SL C a b le xTTH PLC W i-F i IE E E 1 3 9 4 USB Hom e PNA T o o th
PLC 8 0 2 .1 1 a /b

P ossib ly S e p a ra te H W P ossib ly S e p a ra te H W

H U C L - H o m e U n ifo rm C o n tro l L a n g u a g e

Figure 6: Home Gateway Stack (example)

The Home Residential Gateway is to enable the delivery of new services to devices within the home
with the goal to interact with IP (and non-IP via specific GW’s) based home devices. This would give a
Service Provider a management, provisioning, QoS and Security to the Home Residential Gateway
together with the LAN messaging, prioritised QoS and simple remote diagnostics for all home devices
in his domain of management.

Quality of Service (QoS) Requirements and Functions

The meanings of QoS include two aspects: service performance and service differentiation [ITU-T
E.800]. The key parameters of service performance are bandwidth, delay, jitter, and packet loss. The
guarantee of service performance should be end-to-end, consistent and predictable, at a level equal
to or above a guaranteed minimum [RFC2990]. Service differentiation means providing different
performance guarantee for different service applications. And some key services should be carried
with accurate and unaffected QoS guarantee even under heavy load. Especially for voice services,
the IP network should be capable of providing a carrier-class QoS that is equivalent to the legacy
PSTN.

1. The end-to-end QoS architecture should:

• Encompass CPEs
• Be independent of access technology
• Accommodate multiple administrative domains
• Support varying services, such as real-time multimedia communications and VPN
• Support convergence of connectionless and connection-oriented networks and technology
• Support proactive (e.g., admission control) and reactive mechanisms (e.g., congestion
control) based on measurements
The Home Residential Gateway is an environment that enables home networking applications to
utilize QoS resources. These resources provide a management mechanism that prioritises data flows
to support real-time application traffic, such as VoIP, A/V streaming, and video gaming, by using
prioritised media access and queuing.

The Home Residential Gateway QoS:

• Enable home networking applications to establish prioritized data transmission among Hosts
as well as between the Hosts and the Residential Gateway using compliant messaging.
• Enable home networking applications to establish prioritized data sessions between the
Access Node and the Home RG device.
• Home Residential Gateway device supports a transparent bridging functionality for QoS
messaging from/to home compliant applications
• Ability to assign traffic priorities (differentiated media access) to specific applications

1384
 • Ability to prioritize queuing in the Home Residential Gateway device in conjunction with the
packet handling functionality.

The Home Residential Gateway QoS-enabled services:

The NGN@Home should be able to support a wide range of QoS-enabled services. To offer these
QoS services, it is necessary to define:
• Bearer service QoS Classes
• QoS Control mechanisms
• QoS Control Architecture
• QoS Control Signalling

The provision of NGN@Home QoS mechanisms should take account of different QoS control
mechanisms corresponding to different technologies and possibly different business models. The
following three scenarios have been identified:
• Proxy QoS with policy-push: The client’s terminal or Home gateway does not itself support
native QoS signalling mechanisms. It requests a specific service to the Application Manager,
which determines the QoS needs for this service (as in xDSL network).
• Client-requested QoS with policy-push: The client is able to request its QoS needs and the
terminal or the Home gateway is capable to send QoS requests over signalling and/or
management protocols for its own QoS needs, but requires prior authorisation from an
Application Manager (as in Mobile Network).
• Client-requested QoS with policy-pull: The client terminal or Home gateway is capable of
sending QoS Request over signalling and management protocols for its own QoS needs, and
does not require prior authorisation.

This approach is in conformance with PacketCable multimedia Architecture Framework (PKT-TR-MM-

ARCH-VO1-030627) as endorsed by ITU-T J.179.

VPN Service Requirements

VPN services constitute an important piece of the current service market for the service/network
providers and estimations indicate their value will continue to grow. The required VPN service
features will also continue to grow with the evolution of application scenarios and enabling
technologies and also looking at Service Providers' increasing variegated set of network and service
environments.
Some significant evolving requirements are like: simultaneous data/voice/video support, multicast
support, QoS support, enhanced security, mobility integration (nomadicity, roaming), increased
access technologies diversification, increased interworking scenarios, customer-on-demand
capabilities, multi-layer VPN services (L1/L2/L3) and related features (multi-layer resource control),
multi-provider and other complex connectivity scenarios, IPV6/IPV4 support, migration scenarios,
VPN OAM, ubiquitous user identification. A number of these requirements are common to other
services whose support is required by future NGN and NGN@Home architectures.
At the same time, some of the inherent characteristic of the VPN services, like service-transport layer
separation, virtualisation of resource, multipoint connectivity, auto-discovery capabilities, make them
particularly interesting looking at some general characteristics required by future NGN and
NGN@Home architectures.

Security
The ability to guarantee secure communications and to block unwanted traffic or access to a terminal
is beneficial for almost all types of basic service capabilities.
To support the Home user security requirements, there are security functions that reside within the
Home Residential Gateway Security Domain, that is on a per-home basis that include servers used
for key distribution, encryption, and authentication as well as some other client functions.
1. Home Security Portal (HSP). The HSP communicates with Network or Service Provider
security servers, and includes functions that provide client side participation in the
authentication, key exchange and certificate management processes. Other security functions
include management message security, participation in secure download processes, and
remote firewall management
2. Firewall (FW). The Firewall provides functionality that protects the home network from
malicious attack.

1385
 3. Key Distribution Centre (KDC) Servers. The key distribution centre (KDC) servers provide
security services to the CSP and include functions that participate in the authentication and
key exchange processes.

Basic network security issues have changed very little over the past decade. Protecting the
confidentiality of information, preventing unauthorized access and defending the network against
attacks remain primary concerns for network security specialists. Making network security significantly
more challenging is wide spread remote access by and the high number of increasingly sophisticated
attacks.

The network has been placed in a vulnerable position and will remain that way based on several key
trends.
• Changing levels of trust: An ever-widening range of network access is being granted to
different users that is making the network increasingly vulnerable
• Ubiquitous access to the Internet: The availability of the Internet has made every home, office
or business partner a potential entry point for an attack. This ubiquitous access allows
sophisticated attacks to be launched against the corporate network by deliberate attackers or
unknowingly by remote users logging onto the corporate network.
• Attack sophistication: New types of attacks that target application vulnerabilities have been
added to the long list of viruses, worms and Trojan horse attacks that need to be defended
against.

Security components describe the different security layers and their intended use:
• Virtual Private Network (VPN). Protect communications between sites and/or users with an
encrypted, authenticated communications session.
• Denial of Service (DoS). Protect against denial of service type attacks
• Network Firewall. Protect the network by controlling who and what can have access to the
network
• Application Aware Firewall. Combination of network and application level protection detects
and stop application level attacks
• Intrusion Prevention. Dedicated technology designed to protect against a wide range of
sophisticated application level attacks
• Antivirus. Protect against virus attacks at the desktop, gateway and server levels
• Personal Firewall. Protect content on personal computers and in turn, keeps corporate
networks safe.

Possible services to be supported:

• Home Residential Gateway device authentication
• Secure management messages between the WAN access link and the Home Residential
Gateway
• Secure download of configuration and software files
• Optional configuration file for security
• Remote Home Residential Gateway firewall management
• Standardized firewall configuration and reporting
• Simple parental control
• Secure Residential Gateway management messages
• Secure QoS on the WAN access link

Billing, charging and accounting

NGN In-Home should enable all possible types of billing arrangements, as well as accounting
(between providers). This includes also e-commerce arrangements. Billing, charging and accounting
will be based on the collection of information from any appropriate entities in the form of Charging
Data Records (CDRs).

Security Authentication Requirements and Copy Protection

An NGN In-Home Network should provide security from network operator and user perspective. It
should provide the possibility to establish trust relationships with other networks and with users. This
includes the capability of the network to authenticate and authorize a single user and another

1386
network. A user should have the possibility to authenticate the network. The IP multimedia
applications should be provided in a secure manner.
The NGN Home Network should allow for a user to register with multiple terminals in parallel and
multiple users to register with the same terminal. At any time the NGN Home Network should be able
to verify the identity of users and terminals.
Additionally it should be able to check the authorization of the user to use resources of the NGN
Home Network and to access services offered by it. Identities of NGN Home Network users, used e.g.
for authentication, authorization and routing, could be administered by the network operator.
A user profile, that is a collection of user related data, is provided for support of:
• Authentication
• Authorization
• Service subscription information
• Location
• Charging

The NGN Home Network should support mechanisms for the network operator to guarantee the
authenticity of a user identity presented for an incoming call to a user where the call is wholly within
that operator’s network.

NGN@Home Network Mobility Support

The ability for the user to communicate and access services irrespective of changes of the location of
user or terminal, i.e., independent of the network access point, and of changes of the terminal type.
The degree of service availability may depend on the Access Network capabilities, as well as any
service level agreements between the user's home network and the visited network.
• Terminal Mobility, where the same terminal equipment is moving or is used at different
locations. The ability of a terminal to access telecommunication services from different
locations and while in motion, and the capability of the network to identify and locate that
terminal.
• Personal Mobility, the user changes the terminal used for network access at different
locations. The ability of a user to access telecommunication services at any terminal on the
basis of a personal identifier, and the capability of the network to provide those services
delineated in the user's service profile.
• Service Mobility is applied for a specific Service, i.e. the ability of a user to use the particular
(subscribed) service irrespective of the location of the user and the terminal that is used for
that purpose.
One of the crucial requirements for NGN and NGN@Home is to provide the mobility management for
users and terminals so as to ensure the roaming across the heterogeneous networks and also the
seamless mobility for on-going sessions in the networks.

Possible user requirements for mobility:

• Users should be able to gain access from any network access point. This includes all the
access technologies (fixed, mobile radio, WLAN, WLL, etc.) and users are able to access
their services through other networks (roaming). These possibilities may be limited by
subscription and roaming arrangements among various service providers.
• Users should be able to get their services in a consistent manner, which is dependent on the
constraints they experience in their current situations (ex. terminal capabilities, bandwidth
limitations etc.). This is required for services provided by the network operator as well as
services provided by a third party.
• Users’ availability and reachability should be known to network functions and possibly, to
services and applications, including those provided by a third party.

NGN@Home Network Applications Support

In a NGN Home Network it should be possible to offer and support standardized and non-
standardized applications. The support of non-standardized services is defined by service capabilities.

The NGN Home Network should support capability negotiation of IP multimedia applications to
identify and select the available media components and resources, QoS etc. of IP multimedia
sessions. It should be possible for the capability negotiation to take place on invocation, acceptance
and during an IP multimedia session (e.g. following a change in terminal capabilities, change in media

1387
types etc.). The user, the operator or an application on behalf of them may initiate capability
negotiation.
In order to support the user's preferences for IP multimedia applications, the capability negotiation
may take into account the information in the user profile whenever applicable. This includes the
capability to route the IP multimedia session to a specific terminal, when multiple terminals share the
same NGN service subscription.
The NGN Home Network should support a rapid service creation and deployment using service
capabilities, which enables maximum flexibility for the end user devices and network & application
servers.
This would give an operator possibility to efficiently deploy IP multimedia applications without having
to wait for these applications or additional enabling technology to be standardised in the NGN
environment or somewhere else. For example download of client software from a 3rd party repository.
The NGN Home Network should support regulatory requirements especially with respect to
emergency communications and lawful interception.
The NGN Home Network should provide mechanisms that allow presenting the identity of the
connected party to the session originator, if this is not restricted by the connected party or the
network. It should be possible for the user or the network to identify an alternative destination for an
IP multimedia session or individual media of an IP multimedia session.
The sending party, receiving party or the network on their behalf, may initiate redirection to alternative
destinations. It should support more than one IP multimedia session to be run by a user in parallel
and also activation of concurrent IP multimedia applications within IP multimedia sessions.

1388
Energy Efficiency in the Refurbishment of High-Rise Residential
Buildings: Mapping Out an Integrated Policy Approach
Pedro Guertler, Winton Smith

Association for the Conservation of Energy

Abstract
In the context of pressing and frequently conflicting environmental, economic and social policy
objectives, energy efficiency investment is repeatedly found to be a cost-effective and reconcilable
component of energy policies. High-rise residential buildings are a particularly salient issue in this
regard as their poor energy efficiency is regarded as a “moderate” to “major” problem by 18 out of 27
housing ministries who responded to a Europe-wide survey [1]. Yet no previous research exists on
the Europe-wide picture of the potential for energy efficiency improvement in high-rise buildings, nor
on ways to achieve this potential.
Taking into account the present EU (25) plus Bulgaria, Romania and Turkey, this paper – based on
research funded by the International Energy Agency and the European Alliance of Companies for
Energy Efficiency in Buildings – briefly illustrates the scope for increased energy efficiency in high-rise
buildings and the benefits investment in energy efficiency in these buildings can provide, including the
cost-effectiveness and CO2 mitigation cost of this investment. Following this assessment of the
current situation and potential, and based on a map of the political, economic, social and legal
opportunities for and barriers to financing and implementing the necessary energy efficiency
investment, policy steps are identified and proposed which form the integrated policy approach
needed to exploit the energy-saving potential in this sector.

Introduction
Some 36 million European households are in high-rise residences, one in six of all households [1],
and yet many of the buildings are in urgent need of refurbishment. This paper highlights a Europe-
wide cost-effective energy saving potential of 28% from energy-efficient refurbishment of the high-rise
residential building stock. Attainment of this potential would imply a 1.5% reduction of Europe’s total
final energy demand and CO2 emissions savings of 35Mt. In practice only the less efficient buildings
need to be refurbished to realise these stock-average savings and for these buildings typical savings
in heating energy from refurbishment of between 70% and 80% are identified.
Buildings in general suffer from a variety of barriers that tend to prevent their occupants from
maintaining and refurbishing them to the levels of comfort and energy performance that would be
justified over the longer term, but collective housing is particularly susceptible to market failures. Many
occupants do not own the property while their landlords usually have little motivation to finance
improvements. Refurbishment requires collective agreement on a capital investment, which is difficult
to establish especially when some occupants expect to live in the building over the longer-term but
others only for the short-term. Furthermore, in most cases the occupants of high-rise residences are
not among the wealthier members of society and they find it difficult to raise capital for longer term
investments. It is not surprising, then, to find that this section of the building stock is the most
neglected and that there remain significant cost-effective opportunities for it to be refurbished in a way
that improves comfort, saves energy, reduces CO2 emissions and significantly improves the urban
environment.
This paper, based on a research project funded by the International Energy Agency and EuroACE
(the European Alliance of Companies for Energy Efficiency in Buildings) [2], investigates the potential
for energy savings in high-rise residential buildings in Europe – defined by the 3rd European Housing
Ministers’ Conference on Sustainable Housing, in Genval, Belgium in 2002, as multi-family buildings
with more than four storeys [3]. It then maps out an integrated policy approach for the incorporation of
energy efficiency improvements into widely needed overall refurbishment as a central element of
sustainable refurbishment. The 28 countries covered by the project were organised into eight groups,
according to socioeconomic category (‘old’ EU members (EU15), ‘new’ (EU10) and accession (AS3)
states) and climate (using three heating degree day bands) as shown in Figure 1 and Table 1.

1389
Table 1: Categorisation of countries into base regions
EU15 EU10 AS3
Warm climate (A) (B) (C)
<2700 heating degree France Cyprus Turkey
days (HDDs) Greece Malta
Italy
Portugal
Spain
Moderate climate (D) (E) (F)
2700-3700 HDDs Belgium Czech Republic Bulgaria
Ireland Hungary Romania
Luxembourg Slovakia
Netherlands Slovenia
United Kingdom
Cold climate (G) (H)
>3700 HDDs Austria Estonia
Denmark Latvia
Finland Lithuania
Germany Poland
Sweden

Figure 1: Base regions

Not all possible energy efficiency improvements were considered quantitatively. The quantitative
modelling assessment incorporated wall, roof and floor insulation, window replacement, and
improvements to the heating system – all in terms of their effect on reducing heating demand. Many
other measures, including external solar shading and the effect of insulation on reducing cooling
energy demand (potentially significant in reducing energy demand in hot, humid climates), passive
solar design, ventilation strategies, the reduction of internal heat loads and lighting play an important
role in reducing energy demand in high-rise buildings, but fell outside of the scope of the quantitative
assessment examining the cost-effectiveness and amount of energy and CO2 savings.

1390
Assessing the situation – findings for individual base buildings
Using data from a variety of European projects1 and based on commercial data from EuroACE
members, it was possible to create eight representative (of those in need of refurbishment) high-rise
buildings with construction and energy features typical for buildings in each group of countries. The
main findings from modelling the chosen measures in these individual ‘base buildings’ are shown in
Figure 2 through to Figure 5, covering a range of key indicators.

200

180

160

heating measures
140 window replacement
floor insulation

120 roof insulation

wall insulation
kWh/m2a

warm climate
100 moderate climate
cold climate

80

60

40

20

0
EU15 (A ) EU10 (B) A S3 (C) EU15 (D) EU10 (E) AS3 (F) EU15 (G) EU10 (H)

Figure 2: Reduction of heating demand – contribution of all modelled measures2

• Achievable energy savings are substantial, ranging from approximately 70% to 80% of heating
demand.

1
see [1], [4], [5], [6], [7] amongst others
2
The stacks in each column, read from top to bottom, correspond to the top-down order of the measures in the legend.

1391
 9.00

8.00

7.00

6.00

5.00
€/m2a

energy-related investment cost

retrofit investment cost
4.00

3.00

2.00

1.00

0.00
EU15 (A ) EU10 (B) A S3 (C) EU15 (D) EU10 (E) A S3 (F) EU15 (G) EU10 (H)

Figure 3: Energy-related and retrofit investment cost

• The energy-related investment cost – i.e. the additional cost of making refurbishment that
needs to take place anyway as energy efficient as is reasonably practicable – is approximately
half what it would cost to improve energy efficiency separately from general refurbishment (i.e.
‘retrofit’).
• The cost of energy efficiency investment is lowest in EU10 and AS3 countries, mainly due to
lower labour costs.

14.00

12.00

10.00
average 30 year
heating fuel
€cent/kWh

8.00 price

net benefit to
6.00
householder of
energy saved
4.00

2.00

0.00
EU15 EU10 AS3 (C) EU15 EU10 AS3 (F) EU15 EU10
(A) (B) (D) (E) (G) (H)
Figure 4: Energy prices and household benefit of energy saved3

3
Energy prices taken from [8].

1392
• Energy prices are (still) much higher in EU15 countries than in EU10 and AS3 countries.
• Taking reduced energy expenditure for households into account, there is a net benefit as a
result of investment for all base buildings; the net benefits are highest in EU15 countries.
Though net benefits in EU10 and AS3 countries are lower, as a proportion of households’
disposable income, the benefit is often greater than for EU15 households.

120

100

mitigation cost of CO2

80 emissions
warm climate
€/tCO2

moderate climate
60
cold climate

40

20

0
EU15 (A) EU10 (B) AS3 (C) EU15 (D) EU10 (E) AS3 (F) EU15 (G) EU10 (H)

Figure 5: CO2 mitigation cost

• Net CO2 mitigation costs (i.e. after taking reduced household energy expenditure into account)
are lowest in EU15 countries [2]; more importantly, from a policy-maker’s perspective, gross
CO2 mitigation costs, illustrated in Figure 5 are lowest in EU10 and AS3 countries.

Assessing the situation – findings for the high-rise stock

With respect to overall CO2 emissions there is scope for substantial reductions from the entire
European high-rise building stock; this part of the assessment, unlike the results presented so far, is
not based on the modelling of individual base buildings, but on a survey of European housing
ministries about their stock [1]. Figure 6 illustrates the annual CO2 savings possible from the high-rise
stock according to the ministries surveyed, based on their estimates of energy saving potential.

1393
 9.33
9.98
Base region A
Base region B
Base region C
Base region D
Base region E
Base region F
Base region G
Base region H

0.06

2.37

6.71
1.60

1.10 4.02

Figure 6: CO2 savings potential according to national housing ministries [MtCO2]

The highest energy saving potential is in Eastern Europe; 39% in base region E and 34% in base
region H. Europe-wide, the energy saving potential is 28%, implying a reduction of Europe’s total final
energy demand of 1.5%, and a corresponding approximate emissions reduction of 35 MtCO2. The
lower stock-wide energy saving potential compared to the potential in individual base buildings (see
Figure 2) is because each base building is assumed to be refurbishable with respect to every energy
efficiency measure considered, an assumption that holds true for many buildings, but of course not
across the whole of a country’s or region’s high-rise stock.
In addition to the financial payback and reduced CO2 emissions, the less tangible benefits of improved
energy security (in terms of avoided investment in energy generation and distribution, increased
system reliability, resource conservation and enhanced energy price stability [2]), employment gains
[2, 9] and improved residential comfort and wellbeing [2, 10] also need to be balanced against the
required energy efficiency investment cost.
Six case studies, covering the various climatic and socio-economic regions and carried out as part of
the research underlying this paper [2], highlight many practical approaches for appropriating the
benefits outlined above, and carry a number of their own findings. For energy efficiency in the
refurbishment of high-rise buildings overall, the mapping of an integrated policy approach begins
below.

Mapping opportunities for and barriers to an integrated policy approach

Having identified that there are substantial benefits associated with improving the energy efficiency of
high-rise residential buildings, in practice the realisation of the significant energy and emissions
saving potential is faced with a number of institutional, economic, legal and social barriers, but also
opportunities. A comprehensive assessment identified the following opportunities and barriers as
significant, needing to be exploited or addressed collectively in order to develop an integrated policy
approach.

Politically and institutionally,

The capacity to gain an accurate picture of the state of high-rise buildings, to administer financial
instruments and ensure best practice is applied in the refurbishment of the high-rise stock is crucial. A
number of important European projects, notably OPET Building [6], SUREURO, LOCOSOC and the
project underlying this paper can contribute to filling gaps in knowledge and know-how.
Rapid privatisation and the much higher proportion of privately owned housing in EU10 and AS3
countries [2] poses specific (but not exclusively) institutional challenges to refurbishment, requiring
new approaches and partnerships. Public-private partnership approaches to refurbishment could hold
much promise, though experience is thin on the ground.

Financially and economically,

Energy prices are a key determinant of the attractiveness of energy efficiency investment. With the
lowest European prices likely to rise more rapidly than others, the incentive to save energy should

1394
strengthen and the target groups of new and existing financial instruments to promote energy
efficiency in high-rise buildings would become more receptive to them. In this context, there is an
important opportunity in the extensive European body of knowledge surrounding the design and
implementation of effective financial instruments.
Flat-rate tariffs associated with district heating provision in EU10 and AS3 countries in particular, so
common in the high-rise stock, pose a significant barrier in that they do not create any incentive on
the part of the householder to save energy and thus undermine the effectiveness of financial
instruments. In these cases, incentivising district heating suppliers to save their customers’ energy by
providing a full energy service or third-party financing agreements may supply another means by
which to improve high-rise energy efficiency.
Financial incentives designed to link to the Energy Performance of Buildings Directive’s (EPBD)
certification requirements – and to the Energy End-use Efficiency and Energy Services Directive
(ESD) – present a powerful opportunity to strengthen the case for incorporating energy efficiency
improvement into refurbishment.
The effect of the economic cycle and interest rates on housing expenditure and competing priorities
for investment [11] – in particular for public funds – serve to highlight the fact that most investment in
high-rise buildings is needed where the least is forthcoming, mainly in EU10 and AS3 countries [1], as
illustrated in Figure 7.
3.5% 30.00%

3.0%
25.00%

2.5%
20.00%

2.0%

15.00%

1.5%

10.00%
1.0%

5.00%
0.5%

0.0% 0.00%
EU15 EU10 AS3 EU15 EU10 AS3

Figure 7: Percentage of state budget and percentage of household budget spent on housing

Legally,
The EPBD’s transposition into Member States' legislation offers a central legal opportunity to drive the
improvement of high-rise energy efficiency as part of the refurbishment cycle. The Directive stipulates
that whenever a building with a total useful floor area of over 1000m2 undergoes major renovation, its
energy performance must be upgraded to meet minimum requirements. This fits the profile of high-
rise buildings and matches the argument for integration of energy efficiency into refurbishment: the
chance must be taken to ensure the transposition of the Directive interprets it this way.
The ESD addresses a wide range of barriers, including the removal of competing incentives in the
interests of saving energy, the creation of a market for energy services and the requirement to
introduce individual metering and billing for each end-user. Potential synergies with the EPBD exist,
and the opportunities these present must be investigated further.
Widespread inadequate legislation or procedures governing the collective ownership of, and decision-
making about high-rise buildings or estates pose a significant barrier to implementing energy efficient
refurbishments [2]. Effective laws or codes of conduct are essential. Figure 8 [2, 1] shows the
inadequate coverage of rules governing occupants’ decisions about high-rise buildings in EU10 and
AS3 countries.

1395
 100.0%

90.0%

80.0%

70.0%

60.0% %age with law on condominium

%age with compulsory condominium in mixed

50.0%
ownership
%age with condominium registration
40.0%

30.0%

20.0%

10.0%

0.0%
EU15 EU10 AS3

Figure 8: Prevalence of condominium law in high-rise buildings

Socially,
Marketing and energy advice appropriate to the energy use culture and tailored to the individual to
ensure energy efficient systems are used effectively is an essential opportunity to be taken with any
refurbishment, in particular to counter the barrier of entrenched energy use practices, such as
opening windows and/or using secondary heating systems in response to the widespread problem in
high-rise buildings of over- and/or under-heating.
The potentially collective nature of living in high-rise buildings should be harnessed to get residents to
support each others’ energy-saving behaviour, especially in lieu of the requirements for individual
metering and billing.
Employing tried and tested methods of holistic stakeholder involvement with both pre-refurbishment
consultation and post-refurbishment evaluation of stakeholders’ views, helps strengthen communities,
helps eliminate potential problems before they arise and contributes to the body of good energy
efficient refurbishment experiences [2], in turn helping to improve the often negative perception of
high-rise living [12].

Moving towards an integrated policy for high-rise buildings

In recognition of the cost-effective and very substantial CO2 emissions reductions that can be
achieved, especially in EU10 and AS3 countries but also in EU15 countries with the existing pattern of
energy prices, it is proposed – on the basis of the mapping of the significant opportunities and barriers
surrounding high-rise residential buildings – that policy makers:
• Recognise the inherent market failures and barriers to energy efficiency refurbishment that
apply in the building sector as a whole, but most acutely in multi-tenanted residences, and
devise and implement policies to remedy them.
• Incorporate energy efficiency improvement as a legal requirement whenever refurbishment is
undertaken in high-rise buildings to maximise the cost-effectiveness of investment, in line with
the spirit of the Energy Performance of Buildings Directive.
• Close gaps in building or estate level condominium legislation and collective decision-making
rules to facilitate refurbishment.
• Prepare for energy market liberalisation, in particular in EU10 and AS3 countries, and ensure
that individual metering and billing replaces the existing energy consumption infrastructure, in
line with the spirit of the Energy End-Use Efficiency and Energy Services Directive.

1396
• Consider implementation of framework energy efficiency delivery mechanisms – such as the
UK Energy Efficiency Commitment or the Italian and French white certificate schemes – that
could be used, amongst other purposes, to fund energy-efficient refurbishment activities.
• Facilitate and support the creation of new European funds to accelerate sustainable, energy
efficient refurbishment – especially for EU10 and AS3 countries where it is most needed, and
because no structural funds for housing or energy demand management exist as yet.
Moreover, on a national level, consider:
o Adopting Danish-style requirements for condominium dwellers to contribute a small
monthly payment to a refurbishment fund;
o Introducing fiscal incentives for refurbishment such as tax-deductions for
refurbishments that improve the overall energy performance of the building or lower
rates of tax on the rental income of landlords that improve the energy performance of
their rental stock;
o Developing specific additional funds and obligations for energy-efficient refurbishment
in the case where high-rise residences are publicly owned.
• Link all of the above actions to the implementation of the Energy Performance of Buildings and
Energy End-use Efficiency and Energy Services Directives, to exploit as well as inform the legal
and institutional infrastructure being put into place to support the them.
Taking the opportunities and overcoming the barriers outlined in this paper will require work to
synchronise the objectives of various government departments and other authorities involved in the
delivery of sustainable housing and energy. To this end there is a need to employ consistent
methodologies across government to quantify the wider benefits of energy efficiency improvement
and to commission further research to identify the most innovative forms of financing.

References
[1] PRC Bouwcentrum International (2004). Sustainable Refurbishment of High-Rise Residential
Buildings and Restructuring of Surrounding Areas. commissioned by the Netherlands Ministry
of Housing, Spatial Planning and the Environment, Den Haag.
[2] Guertler, P and Smith W (2006). Energy Efficiency in the Refurbishment of High-rise
Residential Buildings. Association for the Conservation of Energy, London.
[3] mrw.wallonie.be/dgatlp/logement/logement_euro/Pages/Reunions/Genval/En_DefSusHou.htm
[4] Petersdorff, C et al (2004). Mitigation of CO2 Emissions from the Building Stock: Beyond the EU
Directive on the Energy Performance of Buildings. Ecofys, Cologne.
[5] Petersdorff, C et al (2004). Cost-Effective Climate Protection of the EU Building Stock. Ecofys,
Cologne.
[6] http://www.opet-network.net/opetnetworkinfo/areaofwork/buildings.html
[7] http://www.invert.at
[8] Eurostat (2005). Environment and energy statistics. European Commission, Brussels. Can be
browsed at
http://epp.eurostat.cec.eu.int/portal/page?_pageid=0,1136239,0_45571444&_dad=portal&_sch
ema=PORTAL
[9] Association for the Conservation of Energy (2000). National and Local Employment Impacts of
Energy Efficiency Investment Programmes. (SAVE contractXVII/4.1031/D/97-032). Association
for the Conservation of Energy, London.
[10] Bonnefoy, X, Braubach, M, Krapavickaite, D, Ormandy, D and Zurlyte, I (2003) Housing
conditions and self-reported health status: A study in panel block buildings in three cities of
Eastern Europe. Journal of Housing and the Built Environment Vol. 188 pp.329-352. Kluwer
Academic Publishers, Massachussets.
[11] Davis, M and Heathcote, J (2004). Housing and the Business Cycle. Finance and Economics
Discussion Series November 2004, Board of Governors of the Federal Reserve System,
Washington DC.
[12] Land, T (2002). Coping with communism’s grim legacy of high-rise buildings. Contemporary
Review June 2002. The Contemporary Review Company Ltd, Oxford.

1397
1398
A Least Cost Strategy for Increasing Energy Efficiency in
Residential Sub-Sector of Developing Countries
Seyed Mohammad SADEGH ZADEH

Faculty of Engineering, Shahed University, Iran

Abstract
From the macro-economy point of view, it is essential to give a response to the question of which
energy efficiency approach in the residential sub-sector should be given priority and with what
timetable for implementation? The objective of this paper is to propose a long term least cost plan as
a response to this question for developing countries. The Iranian residential sub-sector is selected as
a sample energy flow network to verify the proposed approach. This network contains all energy
consuming appliances and parts. Energy Flow Optimization Model is used to minimize the total
discounted cost of the proposed energy network. With due attention to the current prices of energy,
using the lowest grade evaporative coolers will be economical for the consumers in the arid regions of
the country. Whereas if the energy to be priced on the basis of the imposed costs on the national
economy, only using evaporative cooler in the highest energy grade will be selected. The shell
insulation of the building diminishes the thermal loss. It reduces the heating and cooling demands.
This will reduce the capacity of the cooling and heating equipments and required investment costs, as
an interesting result of the study. The paper looks at the introduction of compact fluorescent lamps
(CFLs) for lighting. If any limitation to be applied in using CFL lamps, then it is recommended to use
fluorescent lamps. The proposed plan results in 27%, 54% and 10% saving in energy consumption,
energy cost and investment cost, respectively.

1. Introduction
In the developing countries, rapid growth of the energy systems must be treated within a wider
context. On the one hand, it must be coordinated with the growth of other sectors of the economy in a
more explicit manner. On the other hand, it must be viewed within the framework of a longer term
evolution of the energy system. Recent experiences in some developing countries have shown that
an energy model called Energy Flow Optimization Model (EFOM) can be a useful tool for purposes of
sectoral analysis and planning within the framework of the overall energy system [1]. EFOM was
developed in mid-seventies for the needs of the overall energy systems analysis activities of the
European Commission by SYSTEMS-EUROPE (SE). It consists of a database management system
(DAMOCLES), a simulation model (SIMUL), and an optimization model involving a matrix generator
"ORESTE", a linear programming code, and a report generator "ORACLE").
Because of its wider view of the energy system the EFOM methodology deals with a more aggregated
representation of the energy system planning than what is classically used in this area. In fact EFOM
can be used in two following modes for energy system planning:
- The Single-Sectoral (SS) mode, in which it is used to provide a quick and rough plan and
define overall directions of action and policies for the energy sector.
- The Multi-Sectoral (MS) mode, in which it is used to provide boundary conditions and
consistent exogenous data for the conventional energy system planning tools.
Usually these sorts of models are a mapping of the whole energy conversion chain and possible
future options for energy supply and emission reduction into a network of energy flows. They use the
minimization of the total discounted costs as the goal function under the restriction of maximum
allowed emissions and technical conditions. EFOM has been developed for purposes of techno-
economic analysis of the overall energy system, and has been applied to different European, Asian
and American countries in this context. This paper is not intended to present the EFOM methodology
in its detail. Only a brief introduction is provided. The interested reader is referred to consult the
reference [1]. The paper will present an example application of EFOM in what was referred to as the
single sectoral mode above.
Application of EFOM for sectoral analysis and planning within the overall energy system framework
was used for the first time in Iran in the context of a 1-year project (2004-2005) [2]. Iran is one of the
world’s biggest oil producing countries with a 9% share of the world oil deposits (90 billion barrels). In
recent years, domestic energy consumption in the country has been growing rapidly and reached

1399
about 44% of the total energy production. If this trend continues, It is predicted that Iran would
become an energy importing country by 2018. Approximately 75% of Iran’s foreign currency earning
depends on the petroleum products. It is, therefore, an important issue to secure oil export through
establishing efficient energy utilization. In order to solve the problems, the government, as stated in
the 3rd five-year national development plan (2000-2004), is preparing to execute the following
countermeasures:
- Reduction of energy pricing subsidies to better reflect actual costs
- Public awareness activities on energy conservation and management
- Implementation of demonstration projects for energy efficiency
- Financial assistance to energy efficiency projects
- Enhancement of legal systems relevant to energy management

Figure 1: Residential sub-sector break down into different sub-systems

The residential sub-sector in Iran enjoys a broad potential in energy management. However it is not
so easy to achieve those objectives since it demands a wide range of measures and major initial
investments as well. The key question is: “Which group of approaches should be given priority and
what is the timetable for their implementation?” To respond to this question, the overall energy system
framework is proposed in part 2. Technical and economical specifications of the building shell have
been presented in part 3 of the paper. In part 4, the paper makes a demand estimation of useful
energy and energy services required in the residential sub-sector for a 25-year period. The 5th part
presents the results of the proposed energy efficiency plan. Finally a conclusion is provided in the last
part of the paper.

2. The overall energy system framework

A combined analysis of major final energy carriers supplied to residential sub-sector and energy
conversion technologies can provide a common framework for coordination of sectoral planning
activities. A methodology suitable for this purpose is based on representation of energy flow within
and among the parts. Starting from final energy delivered to residential sub-sector, the path of energy
flow is schematically represented as it passes through the main processes of heating, air conditioning,
co-generation, shell loses and home appliances services. This is normally represented by a
unidirectional network graph. Each link in the graph represents a "process" in a general sense (e.g.
refrigeration, air conditioning, heating, illumination, etc.). Flow of energy through each process (i.e.
each link of the graph) entails costs (investment as well as operations cost), losses, environmental
effects, manpower, equipment, etc. These are all expressed as per unit of process output or process
capacity as relevant. These per unit data are in fact describing parameters for each typical process.
Collectively, they make up the main part of the data base needed for overall energy system analysis
and planning. The "nodes" of the graph simply represent the connection among the processes, and
carry no process information. Energy flowing into a node from its upstream processes leaves it
through its downstream links; the net flow at each node is thus zero. To facilitate representation of the

1400
overall energy system the energy system graph is broken down into sub-graphs, each representing a
so-called "sub-system".
Figure 1 shows a possible breakdown of the residential energy flow network into different parts. More
details about this network can be find in reference [2] reported this study in detail. In addition to the
process data, scenario oriented data are also considered in the study. These concern mainly
scenarios related to the energy prices. One of the parameters in which influences the results of the
energy network optimization is the efficiency of energy consuming means and equipments. Since the
output of most of these equipments is usually of the type of services and can not be expressed in
terms of energy units, consequently it is not possible to determine the efficiency in regular way. To
develop the energy labelling standards for these equipments, the energy performance index over one
unit of energy services is defined. Accordingly, the equipments in different ranks are compared with
each other on the basis of this index [3, 4, 5, 6, 7, 8, 9, 10, 11].
In the lighting systems, theoretically, each Watt is equal to 680 Lumen upon which the efficiency of
the lamps can be evaluated from the relationship ηlamp = Lm W 680 . Thus, in the lamps, the real
efficiency is used instead of using the agreed concept of relative efficiency.
The time horizon of the study was chosen to be 2004-2029. The study sub-periods are (2004-2009,
2009-2014, 2014-2019, 2019-2029). It is worth mentioning that in energy modelling; logically the
investment cost is measured in comparison with the output energy capacity of an equipment or
process. It is less determined in terms of the capacity of input energy flow. On this basis, the capacity
of annual energy consumption of the equipment has been multiplied by its energy consumption
efficiency to determine the capacity of output energy. The technical and economical specification of
the modelled equipments in accordance with the data of the references [3, 4, 5, 6, 7, 8, 9, 10, 11] is
provided to this study.

3. Walls, floors, ceilings and windows data

The average quantities of the thermal loss in the present situation of the country’s buildings are: walls
35%, floors 7.5%, ceilings 7.5% and windows 50%, respectively [2]. In order to calculate the shell
investment cost, the following assumptions are considered.
The thermal loss of the walls during summer is 7% less than that of the winter [12]. In this study, the
model of thermal losses has been supposed as to be the same in all seasons of a year. The
calculations have been made on the basis of a model building having an area of 75 square meters,
area of walls: 125 square meters, the area of windows as of 10 square meters, the area of floor and
the ceiling, as 75 square meters. The cost of constructing the insulated walls has been considered 1.3
times of the non-insulated ones and the costs of insulating the already-built walls have been
considered that of 40% of the cost of the non-insulated walls. The cost of making the double-glazed
windows has been considered 1.5 times of those of single-glazed windows and the cost of doubling
the layers of the single-glazed windows has been considered 50% of the cost of those of single-
glazed windows. The cost of insulated floor and ceiling is 1.5 times of that of the type without
insulation and the cost of insulating the already-built floor and ceiling is 40% of the cost of the type
without insulation. The shell investment cost per energy flow capacity in the building in terms of
monetary unit on ton oil equivalent in a year (TOY) is calculated from the relationship
C Q. A* 8760 * 93,47 * 10 −15 in which C is the actual cost in terms of monetary unit, Q is the thermal
load of the shell in terms of kilo calorie per hour and A is the share of the heat transfer for the shell in
terms of percentage.

4. Useful energy consumption of residential sub-sector

In this part, the exogenous variables of the heating load, cooling load and the demand of home
appliances services are estimated. Useful energy is the energy form that is really demanded by the
consumer for heat, light or mechanical motion. The amount of useful energy delivered from a given
amount of final energy, that is available to the consumer, depends on the efficiency of the end-use
technology. Useful energy reaches the consumer by providing some type of energy service. Energy
services include, for example, thermal comfort, illumination, food refrigeration and cooking. While we
generally discuss energy end-use efficiency with regard to the conversion from final to useful energy,
it is actually the quantity and quality of energy services that determine whether the consumers’ needs
are met. For example, a relatively efficient air conditioner can reduce the electricity demand of an
office building, but a well-designed building could provide the same energy service of thermal comfort
with no air-conditioning at all [13].

1401
To procure the comfortable conditions, there is a demand for each person to have a 0.708 cubic
metre of fresh air in a minute with the humidity of 50% and temperature of 26 Celsius in summer and
that of 18 Celsius in winter [12, 14 and 15]. This can be provided by ventilation and infiltration of
outside air into the building that must be included as part of the heat load. The formula to be used is:

Qair = 20.12 × CMM × (Tout − Tin )

Po
(1)
PA

in which, Qair is the ventilation and infiltration heat load in terms of joule per second, CMM is the
outside air that enters or brought in through the building in terms of cubic meters per minute, Tout is
the outside temperature and Tin is the inside comfort temperature of the room in terms of Celsius, Po
is the barometer pressure of the place and PA is the standard pressure in terms of bar. The heat load
of hot water consumption is obtained from relationship (2).

Qhotwater = 5.8 × CMD × (Tneeded − Tinco min g ) (2)

in which, Qhotwater is the hot water heat load in terms of joule per second, Tneeded is the needed water
temperature and Tincoming is the incoming water temperature of the building in terms of Celsius and
CMD is the rate of needed hot water in terms of cubic meter per day. The standard average demand
of hot water for each person per day is 0.075 cubic meters daily [14]. The total heat load resulting
from the hot water and the air ventilation and infiltration have been calculated with regard to the
existing population in different regions and climate conditions of the country [16].

In order to estimate the energy services demand for home appliances during the years of the plan, the
relationship (3) is suggested:

LastGrade
Dn = ∑ No .ECapa .AUF .η .(1 + IR )
n
i i i relativei
(3)
i= A

in which, No is the number of the energy consuming equipments being used in the base year, ECapa
is the capacity of energy consuming equipments in terms of energy units in a year, AUF is the annual
utilization factor in terms of percentage, η relative is the relative efficiency of the equipment as
explained in part 2 in terms of percentage, IR is the annual increase rate of the number of equipments
in terms of percentage, n is the year under consideration in comparison with the base year, i=A, B,....,
LastGrade is the energy labelling grade of the equipment and Dn is the total demand for the service of
equipments in year n in terms of energy unit. Concerning the real definition of the efficiency for
lighting, this definition has been used instead of relative efficiency. Thus, the estimated lighting
demand not only procures the modelling needs, but also has a physical concept by itself.

5. The proposed plan

After determining the system framework and providing the necessary technical and economic
specifications and exogenous demand forecasting in previous parts of the paper, the energy flow plan
in the sub-sector is optimized with the help of Energy Flow Optimization Model (EFOM). The objective
of the optimization is to minimize the total cost of residential sub-sector while respecting some
technical and environmental constraints including energy flow, capacity and environmental pollution
bounds, exhaustible reserves and exogenously generated constraints. Interested reader may refer to
the reference [1] for more details

1402
 18
Million Ton Oil Equivalent perYear

16 Non-ducted Air Cond. Absorption Sys Evaporative Cooler

14

12

10

0
2004 2009 2014 2019 2029
Year

Figure 2: Required cooling system capacities during planning period in whole country

0.07

0.06
USD/kWh of cooling

0.05

0.04

0.03

0.02

0.01

0
0.01 0.02 0.03 0.04

USD per electrical kWh

Figure 3: Economically feasible region for CCHP cooling compared with electric compression
cooling systems

1403
 0.01

0.009
0.008
USD per kWh of hot water

0.007
0.006

0.005
0.004

0.003
0.002

0.001
0
0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09

USD per Cubic Meter of Natural Gas

Figure 4: Economically feasible region for CHP heating compared with gas fired heating
system

Continuing, the resulting plan and its impacts are reviewed for the period 2004-2029. The figures for
2004 show the present situation of the sub-sector. The least cost plan recommends using the gas-
fired heating systems to supply the building heating demand in the regions where natural gas network
is available. The central heating system has more chance among others.

80

70 Incandescent Fluorescent CFL

60

50
Percentage

40

30

20

10

0
2004 2009 2014 2019 2029

Year

Figure 5: The shares of selected lighting systems in the proposed plan

The main reason to use central heating system is its concurrency with the absorptive cooling system
that according to Figure 2, it has been broadly selected for the cooling system. In the event of the lack
of using the central heating systems, then using gas instant heater and wall-mounted gas instant
water heater have been selected to supply heating and hot water. For those areas where there is no
gas service, the best option for heating is to use the gas oil central heating system.
Fig. 3 shows the economically feasible region of supplying cooling from the combined cooling, heating
and power (CCHP) systems in comparison with other cooling alternatives in lieu of different prices of
electricity. In the above sensitivity analysis, the price of natural gas has been considered as equal to
USD 0.06 per cubic meter which is conservative assumption for the Iranian context. As it is observed,

1404
for the actual selling price of the electricity in the country, i.e. USD 0.025 per kWh, the economic
break even point of the CCHP cooling is USD 0.04 per kWh of cooling. When it is compared with
other methods of cooling supply, it proves itself as an economic method. If we consider the current
selling price of the electricity sale to the residential sub-sector, i.e. USD 0.0125 per kWh, then the
economic break even point of CCHP cooling is equal to USD 0.025 per kWh of cooling. It is clear that
this pricing regulates the behaviour of the consumers inconsistent with the national interest of the
country. The economic region of hot water and heating of the CCHP system against the gas price is
presented in Fig. 4. The economic break even point in lieu of the present price of the gas sale to the
residential sub-sector, i.e. USD 0.01 per cubic meter is USD 0.001 per kWh of heating. From the
national viewpoint, in which the actual supply cost of the gas is assumed as to be about USD 0.025
per cubic meter, the economic break even point for hot water and heating of the CCHP system is
equal to USD 0.00215 per kWh of that of thermal. It is worth mentioning that the actual cost of the
heating of these systems is estimated less than USD 0.0012 per kWh of the thermal [2].
According to the results of Fig. 5 in order to provide the residential lighting, the compact fluorescent
lamps (CFLs) have been selected. If any limitation to be applied in using CFL lamps, then it is
recommended to use fluorescent lamps. It goes without saying that in some applications, it is
inevitable to use incandescent lamp. The sensitivity analysis shows that even reducing the cost of
electricity by one third, using the CFL lamps is still of priority.
Business As Usual Energy Efficiency Plan
60

50
Milion Ton Oil Equivalent

40

30

20

10

0
2004 2009 2014 2019 2029
Year

Figure 6: Total required energy for heating and cooling

The reduction in energy consumption as a result of the proposed plan is presented in Fig. 6. The main
part of this reduction is related to the insulation of the buildings shell which reduces the loss by about
50%.
On the basis of the optimization results, with due attention to the current prices of energy sale, using
the lowest grade evaporative coolers will be economical for the consumers in the arid regions of the
country. Whereas if the energy to be priced on the basis of the imposed costs on the national
economy, only using evaporative cooler in grade A will be selected.

6. Conclusions
In this study, the energy flow needed for energy services in the residential sub-sector is optimized with
the help of Energy Flow Optimization Model to propose a long term energy efficiency plan. The main
achievements for the Iranian context as an example of developing countries are as following:
Using gas-fired absorption system is of first priority in cooling while evaporative coolers in grades A
and B are suitable in dry climate. Gas-fired system is the best option for space heating and provision

1405
of hot water. Using the central heating is a better selection due to concurrency with gas fired
absorption chillers. Gas instant heater and wall-mounted instant water heater are the following
priorities. The insulation of the shell and making double-glazed windows in the existing buildings and
those under construction, even by doubling its expenses are the priorities. Observing the highest
grades of the energy labelling standard of equipments and using compact and non-compact
fluorescent lamps for the lighting of the building have economic priority.

Acknowledgments
This research was supported and conducted in association with Energy Efficiency Office especially
the learned expert Eng. Hamid Reza Neisaz. We take this opportunity to express our sincere thanks
to all of them.

References
[1] Energy environment planning in developing countries. United Nations, RAS/92/071, New York,
1995.
[2] Energy efficiency policies and plans in building sub-sector. Energy Efficiency Office, Iranian
Ministry of Energy, E=D/7-4-200-10, 2004.
[3] Institute of Standards and Industrial Research of Iran. Specification for energy labelling of
electrical household water heaters, September, 1st edition, 2004, Downloadable at:
http://www.isiri.ir/.
[4] Institute of Standards and Industrial Research of Iran. Specification for energy consumption
and energy labelling of electrical household washing machines, 1st edition, 2004,
Downloadable at: http://www.isiri.ir/.
[5] Institute of Standards and Industrial Research of Iran. Liquid chilling packages (with water-
cooled condenser and evaporator) – Method for measuring of energy consumption and energy
labelling instructions, 1st edition, 2004, Downloadable at: http://www.isiri.ir/.
[6] Institute of Standards and Industrial Research of Iran. Refrigerant compressors – Method for
measuring of energy consumption, 1st edition, 2004, Downloadable at: http://www.isiri.ir/.
[7] Institute of Standards and Industrial Research of Iran. Specification for energy consumption
and energy labelling of electrical household refrigerators and freezers, 2nd edition, 2004,
Downloadable at: http://www.isiri.ir/.
[8] Institute of Standards and Industrial Research of Iran. Evaporative air coolers – Method for
measuring of energy consumption and energy labelling instructions, 1st edition, 2004,
Downloadable at: http://www.isiri.ir/.
[9] Institute of Standards and Industrial Research of Iran. Non-ducted air conditioners and heat
pumps – Method for measuring of energy consumption and energy labelling instructions, 1st
edition, 2004, Downloadable at: http://www.isiri.ir/.
[10] Institute of Standards and Industrial Research of Iran. Specification for energy consumption
and energy labelling of electrical lamps, 1st edition, 2004, Downloadable at: http://www.isiri.ir/.
[11] Institute of Standards and Industrial Research of Iran. Specification for energy consumption
and energy labelling of electrical household room heaters, 1st edition, 2004, Downloadable at:
http://www.isiri.ir/.
[12] Malekzadeh G. R., Kashani Hesar M. H. Handbook of air conditioning system design, Ostad
pubblishers, Mashad, Iran, 1994.
[13] Swisher J. N., Jannuzzi G. M., Redlinger R. Y. Tools and Methods for Integrated Resource
Planning, UNEP Collaborating Centre on Energy and Environment, Riso National Laboratory,
Denmark, 1997.
[14] American Society of Heating, Refrigerating and Air Conditioning Engineers, GA 30329,1997
Fundamentals, 1995 Systems, 1993 Applications.
[15] Tabatabaei S. M. Building utility calculation, Rouzbehan Publisher, Tehran, Iran, 9th edition,
2003.
[16] Statistical Center of Iran. Iran Statistical Yearbook, 2003, Downloadable at:
http://www.sci.org.ir/.

1406
Energy Saving Potential and Environmental Impacts of Televisions
Using Energy-Efficient Power Supplies
Edson Adriano Vendrusculo1, José Antenor Pomilio2, Gilberto De Martino
Jannuzzi3
1,2
School of Electrical and Computer Engineering, State University of
Campinas,
3
School of Mechanical Engineering, State University of Campinas, and
International Energy Initiative Latin American Office –IEI-LA

Abstract
This paper analyzes the standby power consumption of household televisions (TVs) and considers
the new technologies available for switching-mode power supplies (SMPS). Semiconductor
companies have introduced new, energy-efficient semiconductors, which have matured to the point of
being currently available for SMPS applications. Further, these companies claim that 25% of total
energy consumption is now consumed in the low power/sleep/standby modes.
This analysis estimates the annual reduction in energy consumption and the CO2 conservation
accruing from the reductions in natural gas power plant emissions that will result from the use of more
energy-efficient TVs. It takes into account information about production, energy consumption, lifetime,
and market share of TVs (assuming annual Brazilian sales of 20-inch TVs of 2,251,080 units in 2004
as estimated in a recent Brazilian government essay).
Results indicate that at least one television is found in 87.7% of Brazilian homes. Throughout the
country, household TVs consume 7.2 to 10.9 terawatt-hours (TWh) of electricity per year, or about
10–15% of Brazil’s residential electricity consumption. The overall consumption in 2002 was 72.7
TWh.
This paper uses the Energy Star and Ecolabel standby criteria for reference, since regulations for
establishment of standby power standards have not yet been passed by the Brazilian Congress.

I. Introduction
In 2001, the Brazilian law 10.295/2001 set the principles for the “National Energy Conservation Policy
and Rational Use of Energy”[1]. As long as this law is in place, all electric equipment commercialized
in the country will be required to comply with Brazil’s energy efficiency regulations. Energy standards
are promoted around the world through well-known ecolabeling and energy efficiency programs, such
as Energy Star in the USA, the Ecolabel in Europe, and the Top Runner in Japan. In Brazil, the Procel
labeling program has been continuously updated and revised following the worldwide trends.
However, only recently, the Brazilian Congress decided to use the labeling program to establish
standby power standards [2]. Since the Brazilian standards had not yet been established at the time
of this research, this paper uses the North American and European standards for reference purposes.
The main intent of this work is to assess the television (TV) sets available in the Brazilian market,
based on the Energy Star and Ecolabel standards, which are the standby power criteria in the USA
and Europe, shown in Table 1.
Additionally, semiconductor companies have introduced technological improvements with impacts on
energy efficiency. Some new semiconductor devices have matured to the point of being currently
available for switching-mode power supply (SMPS) applications, which can increase the energy
efficiency of TV sets. Finally, this paper summarizes the economic and environmental impacts of cost-
effective improvements in TVs, based on the reductions in energy consumption and CO2. It takes into
account information about production, energy consumption, lifetime, and market share.
In July 2005, Energy Star program celebrated the achievement of its first goal, i.e., the establishment
of a 1-watt (1W) standby limit for TVs. As seen in Table 1, beginning in March 2005, advances in
standby regulations in the European Community may yield even more strict limits on standby losses.
Electronic improvements, however, may necessitate changes in the standby consumption regulations.
In fact, the current standby definition may become inconsistent (see table note “a”) as a result of the
establishment of new operating modes, such as “sleep” and “deep sleep.” A worldwide agreement for
household appliances operating on standby mode has been discussed on numerous occasions; a low

1407
power mode (LOPOMO) designation, which is under discussion, is outlined on the standby power
home page of Lawrence Berkeley National Laboratory [6].

Table 1: Energy-Efficiency Specifications for Qualified TVs

ENERGY STAR (USA)[4]
ECOLABEL(Europe) [5]
(standby modea)
Product Phase I Phase II Phase III Effective from 1 April 2002 until 31 March 2005.
(effective (effective (effective
7/1/02) 7/1/04) 7/1/05)
< 1W (for passive standby consumptionb
≤ 1W (analog)
and < 9W (for active standby consumptionc of
TV ≤ 3W and ≤ 1W
the televisions which have an integrated digital
≤ 3W (digital) receiver/decoder)
a
Standby power is defined as the power being used when the product is connected to a power source, produces
neither sound nor picture, does not transmit nor receive program information and/or data (excluding data
transmitted to change the unit’s condition from “standby mode” to “active mode”), and is waiting to be switched to
“on” (active/play mode) by a direct or indirect signal from the consumer, e.g., with the remote control [1].
b
Passive standby – the television is connected to a power source, produces neither sound nor picture, and is
waiting to be switched into an “off”, “active standby,” or “on” mode, on receipt of a direct or indirect signal, e.g.
from the remote control [2].
c
Active standby – the TV is connected to a power source, produces neither sound nor vision, and is
exchanging/receiving data with/from an external source [2]

II. Test Criteria

The specifications that products must meet to get an energy-efficiency endorsement in different
regions depend on the local electrical distribution system and other circumstances. For Energy Star
labeling, only the consumption in the standby mode is considered, whereas the European Ecolabel
regulations also take into account the consumption in “on mode.”1.
Energy Star specifies general criteria for voltage and total harmonic distortion (THD) lower than 3%
and ambient temperature in the range of 22°C ± 4°C. The nominal voltage is a market-specific
criterion; the recommended values are listed in Table 2.

Table 2: Market-Specific Criteria for Energy Star

Market: United States Europe and Australia Japan
115 VRMS ± 3 VRMS 230 VRMS ± 10 VRMS 100 VRMS ± 5 VRMS & 200
Voltage
VRMS ± 10 VRMS
Frequency 60 Hz ± 3 Hz 50 Hz ± 3 Hz 50 Hz ± 3 Hz & 60 Hz ± 3 Hz

Ecolabel regulations use the technical standard EN 50301 [7] to measure power consumption of
appliances and equipment during normal operation (“on mode”). In addition, Working Group 9 of the
International Electrotechnical Commission (IEC) Technical Committee TC59 prepared the
international standard IEC 62301, i.e. “Household Electrical Appliances – Measurement of Standby
Power”[8].
Progressively more energy-efficient standby power devices have been offered by semiconductor
manufacturers since the Energy Star, Blue Angel, and Top Runner programs released their energy
efficiency specifications for power supplies used in consumer electronics products. To highlight these
manufacturers’ efforts, the following section discusses the energy losses in a very common topology
for SMPS.

III. Reduction of standby losses in Switching-Mode Power Supply

Semiconductor manufacturers claim that 25% of total energy is consumed in low power/sleep/standby
mode and around 75% of average total energy consumption is in active mode. Further, in active
mode, changing semiconductor efficiency from 60% to 75% can result in 15% energy savings [13].
Consequently, much analysis is being done on the losses in standby and active mode (see example
in Figure 1 for flyback topology), on driving TV sets and other audio devices [13].

1 “On mode” - the television is connected to a power source, and produces sound and vision.

1408
Figure 1: SMPS common topology: losses in standby mode

There are several key sections of the SMPS that can be optimised to minimise standby power
consumption. The losses can be categorised into two types – conduction losses and switching losses
[19]. Power switches based on MOSFET (Metal-oxide-semiconductor field effect transistor)
technology devices have a named on-state resistance, Rds(on), which represents a major area of
conduction loss. This loss can be minimised by selecting MOSFETs with lower Rds(on). Unfortunately,
these devices tend to have a larger gate capacitance, which in turn increases the switching losses.
However, depending on the output power rating, it is possible to select a MOSFET that strikes an
appropriate balance between switching and conduction losses. Power input (Pin)—see equation (1)—
helps provide a comprehensive insight into SMPS losses with regard to both aforementioned
strategies. It is deduced considering that, in a switching period, the energy drawn by the transformer
during the on-time is transferred to the output during the off-time. As a result:
1
Pin = × L p × I pk
2
× f sw (1)
2
where Lp is the transformer primary inductance, Ipk is the inductor peak current, and fsw is the normal
working switching frequency.

Two strategies, both linked to the SMPS switching frequency, have been established to reduce
losses. By employing burst-mode operation (the so-called “skip-cycle” mode) or decreasing the
switching frequency as much as possible, one may achieve a reduction in the switching losses in
standby mode. These strategies are described below.

a) Burst-mode operation or skip-cycle mode

Especially when the requirement to provide output power is minimised, it is possible to operate the
power supply in “bursts”, whereby the output power is supplied in small periodic bursts of pulse width
modulated (PWM) operation, rather than in continuous mode. This minimises both conduction and
switching losses. While this feature has been employed in high-end designs, it has only recently been
needed in power supply designs for output power under 100 W.
In this mode, switching cycles are automatically skipped when the output power demand drops below
a given level. This is usually accomplished by monitoring a feedback pin available in controllers.
When the controller enters the controlled-burst operation, the power transfer depends on the width of
the pulse bunches, as seen in Figure 2 for P1<P2<P3 [20].

1409
Figure 2: Output pulses at various power levels. Horizontal scale: 5us/div.

In an example where Lp=300 microhenries (μH), Ipk=600 miliamperes (mA), and fsw=65 kilohertz
(kHz), a theoretical power input from (1) is Pin=4W. If skip cycle or burst mode takes place with a
bunch length of 10 ms over a recurrent period of 100 milliseconds (ms), then the total power transfer
is Pin= 0.4W.

b) Decreased switching frequency

In such a strategy, the SMPS controller automatically detects a light load condition and, as a result,
decreases its oscillator frequency as well. The normal oscillation frequency is automatically resumed
when the output load builds up and exceeds a defined threshold. This function allows for the
minimization of power losses related to switching frequency, which represent the majority of losses in
a lightly loaded flyback, without giving up the advantages of a higher switching frequency at heavy
load [21]. This is accomplished by monitoring a feedback controller pin whose signal depends linearly
on the peak primary current in the transformer. If the peak primary current decreases (as a result of a
decrease of the power demanded by the load) and the voltage at the feedback pin falls below a fixed
threshold (VT1), the oscillator frequency will be set to a lower value (fSB). When the peak primary
current increases and the voltage at the feedback pin exceeds a second threshold (VT2), the oscillator
frequency is set to the normal value (fosc). This operation is graphically shown in Figure 3.

Figure 3: Standby dynamic operation.

f osc
In some SMPS controllers, the highest rate for the ratio has been set at 5.5 [21]. Thus, based
f SB
on (1), the standby losses may be similarly reduced.

1410
IV. Power Consumption Tests
In order to better represent the TV market, three leading Brazilian brands were purchased—all with
20-inch screens—and their energy consumption was measured according to the Energy Star criteria.
Standard international testing procedures specify that the voltage and frequency shall be within ±1%
of the nominal, and THD voltage shall not exceed 2% (Ecolabel) and 3% (Energy Star). Ambient
temperature shall be within 22oC ± 4oC. So, a qualified AC Source model CI 4500iL manufactured by
California Instruments provided the power, as necessary. California voltage (rms) accuracy (@25oC ±
5oC) is with 0.15%+0.3volts (V) and frequency accuracy is 0.01%+0.01Hz for the 45-100Hz range.
The CI 4500iL model allows delivering 15Amps (45Hz-1kHz) while maintaining maximum 1% of THD
voltage.
Figure 4 shows the voltage harmonic distortion in supplying a TV model HPS-2023 manufactured by
CCE. A voltage distortion of 0.08% for the high THD current of 83.45% states the power performance
specifications provided by California’s power supply. Voltage and current waveforms with regard to
these distortions are shown in Figure 5, such that supplied voltage and current are respectively
114.8V and 0.996A root-mean-square quantities, which performs with 5.85W of active power.
For true power measurements of 10W or less, the measurement instrument must have a resolution of
0.01W or less to comply with the Ecolabel regulations. True standby power measurements require the
use of a true power wattmeter. In such an exercise, care should be taken to select appropriate power
measurement equipment, since TV sets may draw current that is not sinusoidal, as shown in Figure 5.
The high crest factor of the current may cause internal peak distortion on a common wattmeter, i.e., it
may clip off the top of the current wave.
To avoid low accuracy, low resolution, and peak distortion, an oscilloscope with a current probe was
used in the tests of standby operation.

Figure 4: AC Source measurement window. Current

and voltage harmonic distortion (THD) and current
spectrum.

1411
Figure 5: Top: instantaneous active power in
TV. Bottom: voltage and harmonic current
waveforms for CCE HPS-2023 model.

Measurement Equipment
The TDS430A Tektronix digitizing oscilloscope has the following features, which enable the
measurement of energy consumption:
• 400 megahertz (MHz) maximum analog bandwidth
• 100Megasamples/second maximum digitizing rate, on each channel simultaneously
• Up to 30,000-point record length per channel
• Each channel with 8-bit resolution2. Nevertheless, in High Resolution Mode, up to 15 bits
of digitized resolution is available. This is a nominal trait of high-performance digitizing
oscilloscopes.
• Minimum voltage setting is 1mV/div.

To avoid current distortions due to a high crest factor, a TCP202 50MHz current probe containing a
Hall effect device was used. The TekProbe interface allowed the TCP202 probe to be directly
connected to the oscilloscope. The TCP202 probe has a maximum peak current of 50A, with a pulse
width ≤ 10µs (500µAseconds). As the frequency decreases, the maximum current rating increases,
limited to 15A (DC + peak AC).

V. Standby Power Consumption - Experimental Results

Estimates of the number of TVs in Brazil were taken from the CPqD Foundation’s Implementing
Model of Digital TV in Brazil Project [9], a statistical report prepared exclusively for the
Telecommunications National Agency (ANATEL) in 2002. Results indicate that at least one TV is
found in 87.7% of Brazilian homes. Nationally, residential TVs use 7.2–10.9TWh of electricity per
year, accounting for about 10–15% of Brazil’s residential electricity consumption. The overall
consumption in 2002 was 72.7TWh [10]. These figures justify research into standby power
consumption by residential TVs in Brazil.
Performing experimental tests on all TV screen sizes is expensive. As a result, the 20-inches screen
size was chosen as the focus of this study, based on its 37% market share. Models of the three most
common brands were purchased for testing; their respective features are listed in Table 3.
Although none of the brands has standby technology embedded in compliance with international
standby regulations, it is evident that these models are able to meet worldwide market requirements.
Furthermore, from Table 3, it is obvious the lack of technical specifications for power consumption
measurement as seen by the different measurements conditions used.
Experimental evaluation of Brazilian TV sets was carried out according to the aforementioned
international standby requirements. A summary of the experimental results for both rated voltages of
115 and 230V is shown in Table 4.

2
Displayed vertically with 25 digitization levels (DLs) per division. Expressed as a voltage, a DL is equal to 1/25 of a division
times the volts/division setting.

1412
Only one model currently complies with the North American energy efficiency program. Further, there
is a large discrepancy between manufacturers in the power consumption in standby and “on mode”
operation in Table 3 and Table 4.

Table 3: TVs features of three major brands (according manufacturer specifications)

Voltage Energy consumption in “on mode” Standby
Brand/model
/frequency consumption
PHILIPS Universal (90 to 46W (approximated) < 3W
/20PT3331 255)V/ 50 or 60 Measurement conditions: not mentioned Measurement
Hz conditions:
Voltage at 220V
LG / RP 20CB20A Universal (100 85W (maximum) 10W
to 240)V/ 50 or Measurement conditions: not mentioned Measurement
60 Hz conditions: not
mentioned
CCE / HPS 2023 Universal (100 54 W ± 10% Not mentioned
to 240)V / 50 or Measurement conditions:
60 Hz Voltage at 120V, “color bars” signal,
volume=15, brightness/contrast and color
= 50%

Previous work by Alan Meier (of Lawrence Berkeley National Laboratory), testing 321 TVs, has shown
that the TV active power draw (“on mode”) is closely related to screen size and manufacturer, while
TV standby power draw is related only to the manufacturer [11]. Despite the small number of TVs
tested in this study, the results of Table 4 show the same divergent standby power measurements, as
verified in Figure 6 [11].

Table 4: Standby consumption –Brazilian TV sets with 20-inch screens.

Standby Power [W]
Brand/model
115V 230V
PHILIPS /20PT3331 1,78 1,96
LG / RP 20CB20A 5,3 7,8
CCE / HPS 2023 5,85 8,1

Figure 6: Average standby power consumption

values by screen size [11]

Consequently, it may be inferred that standby power consumption depends basically on the SMPS
provided by the TV manufacturers, while the active consumption depends mainly on the vacuum tube
technology.
Despite the focus of this work on standby power, active power consumption also must be taken into
account when applying for the European Ecolabel program; this is because the Ecolabel includes
environmental3 criteria, in addition to the energy-efficiency criteria considered by Energy Star.

3
Related to the use of energy, the Ecolabel criteria aim in particular at promoting:
– the reduction of environmental damage or risks related to the use of energy (global warming, acidification, depletion of non-
renewable resources) by reducing energy consumption.
Observation: There are others criteria related to the use of natural resources and related to the use of hazardous substances.

1413
Applicants to the Ecolabel program for TVs need to consider passive standby, active standby, and “on
mode” consumption. The following section discusses the additional requirements of the Ecolabel.

VI. Ecolabel Criteria for Televisions

The main environmental impact of a television results from energy consumption during its use; this
includes both standby and “on” modes. The Ecolabel criteria set requirements for both passive and
active standby modes (refer to Table 1), which are similar to those set by other organizations [12].
Active standby is becoming more significant as televisions use this mode for receiving electronic
programme guides; this may become a standard feature in new TV sets. In addition, service providers
of digital television transmissions may require an integrated receiver/decoder (IRD) to be in the active
standby mode for significant periods of time [12].
Consumption in the “on mode” is becoming more important as improvements in picture quality
increase electricity consumption. The Ecolabel criterion set for “on mode” energy use requires that
televisions meet an energy-efficiency index (EEIon) of < 65%, which is between the proposed energy
efficiencies for the A and B ratings in the energy label for televisions. [12].

A. Ecolabel “on mode” power consumption criterion

The “on mode” energy efficiency index EEIon shall be derived from the equation:
EEI on = Pon Pon ,bc (2)
Where:
• Pon is the measured energy consumption in the “on mode”.
• Pon,bc is the base case energy consumption of the television in the on mode. This is calculated
using the following formula:
0.75 ⋅ format ⋅ scrnsize + digit ⋅ 0.33 + 0.38 ⋅ scrnarea
Pon ,bc = 16 + 16 ⋅ idd + (3)
0.825
Where:
• digit is equal to 1 if the television has digital processing for picture scanning, and 0 if it does not
have such processing.
• format is equal to 0.80 for a standard screen (4:3 aspect ratio), and 0.87 for a wide screen (16:9
aspect ratio).
• scrnsize is the screen diagonal in cm.
• scrnarea is the area of the screen in dm2, i.e., it is equal to scrnsize x scrnsize x 0.48 / 100 for a
standard screen (4:3 aspect ratio), and scrnsize x scrnsize x 0.427 / 100 for a wide screen (16:9
aspect ratio).
• idd is equal to 1 if the TV has an Integrated Digital Decoder for digital broadcast signals, otherwise
it is equal to 0.

In summary, the following assessment and verification must be ensured for applying to Ecolabel
labeling [12]:
The applicant shall provide a test report stating that the level of power consumption in passive
standby, active standby, and “on” modes has been measured using the procedures shown in EN
50301 (Methods of measurement on receivers for TV broadcast transmission). The report shall state
the measured power consumption in each mode, the calculated base-case “on mode” consumption,
and the calculated percentage of base-case consumption in the “on mode.”

B. Experimental results for “on mode” power consumption

“On mode” power consumption measurements and base case power consumption are shown in Table
5. Digital processing is not integrated in the evaluated three 20-inch screen TV models, so the
following parameters were used as a base case:
• digit=0;
• format=0.8;
• scrnsize=48 centimeters (cm);
2
• scrnarea=11.06 square decimeters (dm );
• idd=0.

In Brazil the standard NBR5258 sets to approximately 48cm the screen size of 20 inches television,
so, applying equation (3) it is found that the base case power is about Pon,bc=58.18W.

1414
As seen, measurements of standby power (Table 4) and EEIon (Table 5) show the non-conformity of
the three Brazilian televisions analyzed to Ecolabel requirements. The Philips model—EEIon=0.68
and standby power of 1.96W (230V – European specific voltage)—comes closest to complying with
the Ecolabel standard. With regard to Energy Star, the Philips standby power of 1.78W (115V North
American specific voltage) satisfies the required limit of 3W, which is effective starting July 2005.
Considering the TVs’ overall performance, with the exception of the Philips model, this paper
addresses the technological enhancements, with regard to SMPS, needed to comply with the
international standards.

Table 5: Ecolabel Energy Efficiency Index

“On mode” power Base case power Energy Efficiency Index
(Pon) [W] (Pon,bc) [W] (EEIon)
Manufacturer/
Model 115V 230V 115V 230V
# # # #
(rms) (rms) (rms) (rms)

PHILIPS/
39.6 -- 0.68 --
20PT3331
LG/
67 72 58.18 1.15 1.23
RP 20CB20A
CCE/
41 43.2 0.70 0.74
HPS 2023
#
root mean square

VII. Environmental and Economic Analysis

This analysis estimates the annual reduction in energy consumption and CO2 emissions from natural-
gas power plants that could result from the use of more-efficient TVs in Brazil. The study is based on
estimated annual sales of 20-inch screen TVs (2,251,080 units in 2004), as stated in a recent
Brazilian government essay [15]. The analysis was based on the following assumptions:
• Standby time: An average time of 6.6 hour (h) in active mode was recorded for all social groups in
Brazil [15]. Therefore, the televisions are assumed to stay in standby mode for 17.4h.
• Market Share: Brand (%) sales, as detailed in Table 6.
• Standby power consumption: The brand weighted-average power was calculated from
experimental data rather than that stated by manufacturers, which does not provide a sufficient basis
for evaluation.
Using the data from Table 6, the average power is respectively 3.45W and 4.37W, for the 115V and
230V North American and European specific voltages. A 1W mandatory standby standard would yield
an average power reduction of 2.45W (115V) and 3.37 (230V) in 2005.

Table 6: Market share

Brand Model Standby – Standby – experimental results [W] Market
as stated by 115V 230V Share
manufacturer
in the technical
manual [W]
Toshiba TV 2090AV 3.0 assumed to be assumed to be 10% 23%
3.0 higher than for
115V, i.e., 3.3
Philips 20PT3331 < 3.0 (220V) 1.78 1.96 21%
CCE HPS 2023 not declared 5.85 8.1 14%
LG RP 20CB20A 10 5.3 7.8 13%
Philco TP-2053 < 2.5 assumed to be assumed to be 10% 13%
2.5 higher than for
115V, i.e., 2.75
Panasonic not available (assumed to be <1W) 10%
Mitsubishi not available 4%
Sony not available 2%
• An increasing rate of 15% in sales was estimated to 2005 [16].

1415
In conclusion, assuming a mandatory 1W standby power requirement, one can estimate a power
reduction of 5.32MW (115V) and 7.32MW (230V) in 2005. This reduction is due to power
improvements in only 84% of Brazil’s TVs, since the last three brands (Table 6) are assumed to
already comply with the mandatory 1W standard.
Table 7 summarizes the results for energy savings, and power and CO2 reductions. The estimated
17.4h standby period was used to calculate the energy savings. As a result, the nation could save
16.25 (gigagrams of CO2) GgCO2 (115V) due to avoided emissions from natural gas-fueled
thermoelectric plants (assuming a coefficient of 0.48gCO2/Wh [17]). Similar analysis may be done to
the 230V grid.

Table 7: Environmental and economic saving.

Voltage [V] Technical/Economic data Results
Real Units sold Retail rate Time in Power Energy CO2 reduction
power in 2004 increasing standby saving saving [GgCO2]
[W] to 2005 [h] [MW] [GWh]
[%]
115 3.45 2.251.080 15% 17,4 5,32 33,85 16,25
230 4.37 2.251.080 15% 17,4 7,32 46,56 22,35

VIII. Conclusions
The recent discussion about energy-efficiency standards in Brazil’s national Congress might require
manufacturers, in the medium term, to improve the standby technology embedded in TV sets. It was
theoretically shown that at least 84% of brands currently do not comply with energy savings. As a
result, e.g., for the 115V grid, the establishment of a 1W standby standard in 2005 could save 8.53
million reais (R$) on electricity bills (for 33.85GWh in Table 7 and an electricity price of 252 R$/MWh,
including the 18% of tax on the tariff of the National Agency of Electric Energy - ANEEL). More
societal benefits could also result, for instance, avoiding 16.25Gg of CO2 emissions. Similar analysis
may be done for the 230V grid.
Currently, new SMPS controllers comply with the energy-efficiency labels worldwide. Nevertheless,
these controllers are developed for more common topologies and for limited power—for instance,
flyback converters and 150W, respectively. Therefore, more effort from researchers on studying
standby power losses should be encouraged.
Finally, assuming that 50% of all 2,251,080 units sold in 2005 are connected in the 115V grid and the
remainders are connected in the 230V grid, the average values from Table 7 are:
• Power savings = 6.32MW, energy savings = 40.2GWh, and avoided CO2 emissions = 19.3
GgCO2.
To serve as a reference, the estimated power savings are 30% of the total wind power capacity
installed in Brazil, i.e. 20.3MW [18].

IX. Acknowledgments
We gratefully acknowledge The State of Sao Paulo Research Foundation (FAPESP) for financial
support (02/08938-3) to this work.

X. References
[1] Lei 10.295/2001. Dispõe sobre a Política Nacional de Conservação e Uso Racional de Energia
e dá outras providências (Law 10.295/2001: National policy on energy conservation).
Subsecretaria de Informações do Senado Federal (Federal Senate), 17/outubro/2001. Decreto
4.059 - Regulamenta a Lei no 10.295. Subchefia para Assuntos Jurídicos da Casa Civil da
Presidência da República, 19/dezembro/2001 (in portuguese).
[2] Law project - PL-3893/2004. “Altera o art. 2º da Lei nº 10.295, de 17 de outubro de 2001,
estabelecendo limite para o consumo de eletricidade por aparelhos operando em modo de
espera (establishes standby power limit for electronics apparatus) ”. Câmara dos Deputados.
Can be downloaded at: http://www.camara.gov.br/Internet/sileg/Prop_Detalhe.asp?id=259696
(in portuguese).

1416
[3] Wiel S. and McMahon J. E. Energy-Efficiency Labels and Standards: A Guidebook for
Appliances, Equipment, and Lighting. Collaborative Labeling and Appliance Standards Program
(CLASP 2001), 205p.
[4] Energy Star – Government-backed program helping businesses and individuals protect the
environment through superior energy efficiency. Can be downloaded at:
http://www.energystar.gov/
[5] European Environmental Labeling Program - ECO-LABEL. Can be downloaded at:
http://europa.eu.int/comm/environment/ecolabel/index_en.htm.
[6] Alan Meier; “Research Recommendations to achieve energy savings for electronic equipment
operating in low power modes”, Lawrence Berkeley National Laboratory (LBNL), September 30,
2002. Can be downloaded at: http://www.standby.lbl.gov/
[7] CENELEC - European Committee for Electrotechnical Standardization. Can be downloaded at:
http://www.cenelec.org/Cenelec/Homepage.htm
[8] IEC (International Electro technical Commission). “IEC 62301 Ed 1 – Measurement of Standby
Power”, . Novembro, 2003. Can be downloaded at:
http://www.energyrating.gov.au/library/detailsiec-standbydraft2003.html.
[9] Agência Nacional de Telecomunicações – ANATEL (National Telecom Agency). “Projeto
Modelo de Implantação da TV digital no Brasil – Relatório Produto III, Análise das Condições
Brasileiras para a Introdução da Tecnologia Digital na Transmissão Terrestre de Televisão”
(evaluates Brazilian system for digital TV implementation). Can be downloaded at:
http://www.anatel.gov.br/radiodifusao/tv_digital/analise_030_2002.pdf
[10] Ministério de Minas e Energia (The Ministry of Mines and Energy). Capítulo 2: Oferta e
demanda de energia por fonte 1987/2002. Balanço Energético Nacional (energetic balance or
potency) 2003. Brasília: MME, 168p, 2003. Disponível online no site: Can be downloaded at:
http://www.mme.gov.br/paginasInternas.asp?url=../ben/.
[11] Karen B. Rosen e Alan K. Meier, “Energy Use of Televisions and Videocassette Recorders in
the U.S.”, Environmental Energy Technologies Division, Lawrence Berkeley National
Laboratory. Can be downloaded at: http://eetd.lbl.gov/ea/reports/42393/
[12] J. Poll, P Dolley, Dr N Varey; “Development of EU ecolabel criteria for televisions:A report
produced for Department for Environment, Food & Rural Affairs”, AEA Technology
Environment, January, 2002.
[13] Dhaval Dalal, “Enabling Efficient Solutions for Power Supplies“, ON Semiconductor,. June,
2004.. Can be downloaded at:
http://www.energystar.gov/ia/partners/prod_development/downloads/power_supplies/OnSemiP
resentation.pdf
[14] On Semiconductor. “PWM Current-Mode Controller for High-Power Universal Off-Line
Supplies” .Can be downloaded at: http://www.onsemi.com
[15] “Projeto Modelo de Implantação da TV digital no Brasil – Relatório Produto III, Análise das
Condições Brasileiras para a Introdução da Tecnologia Digital na Transmissão Terrestre de
Televisão”, Agência Nacional de Telecomunicações – ANATEL. http://www.anatel.gov.br/
http://www.anatel.gov.br/radiodifusao/tv_digital/analise_030_2002.pdf (in portuguese).
[16] Brazil Focus – Datamark. (Sales and marketing agency). http://www.datamark.com.br
[17] G. M. Jannuzzi, G. C. Queiroz, E. A. Vendrusculo, T. Borges, J. A. Pomilio. “A life-cycle cost
analysis (LCCA) for setting energy-efficiency standards in Brazil: The case of residential
refrigerators.”, ACEEE Summer Study on Energy Efficiency in Industry 2003 Sustainability and
Industry: Increasing Energy Efficiency and Reducing Emissions. July, 2003.
[18] CBEE - Centro Brasileiro de Energia Eólica (Wind Energy Brazilian Center). – Can be
downloaded at: http://www.eolica.com.br/energia.html
[19] Rayabhari M. Cutting Stand-by Power. IEE Power Engineer Magazine. Volume 17, Issue 2, pp.
38-40. April, 2003. Can be ordered from http://scitation.aip.org/PE (IEE Digital Library).
[20] Semiconductor Components Industries. Publication Order Number NCP1216/D. September,
2004 – Rev. 8. Can be downloaded at: http://www.onsemi.com/PowerSolutions
[21] STMicroelectronics. Primary Controller with Standby (L5991:High Performance Pwm
Controllers). Can be downloaded at: http://www.st.com/stonline

1417
1418
Designing Energy Efficiency Electric Motors (EEEM) by Using
Reliability Indicators
C. D. Pitis

Femco Mining Motors

Abstract
Millions of swimming pool motors (SPM) are running without electricity consumption being scrutinized
(EC’s efficiency grading system bands starts from 1.1 kW). Swimming pool unit is attracting the owner
attention only when it fails. In about 65 to 75 % of failures, motor replacement criterion is preferred,
decision being based on initial investment.
Approaching an EEEM design by using higher quantities and/or quality of active materials (reducing as
consequence motor losses) is a very basic method. As a result, the motor cost becomes prohibitive.
Besides lifetime expectation of 3…4 years trouble free at competitive price, the end users start asking for
energy efficient motors (EEEM) as an extra feature.
The target was to obtain a new SPM-EEEM design with minimum changes (alterations), but achieving
also an optimum balance between costs and EEEM lifetime expectancy (MTBF equivalent) in order to
ensure business sustainability. The author described one of his experiences in designing EEEM.
Multidisciplinary techniques (including statistic-probabilistic methods and thermodynamics calculations)
have been correlated to obtain a financially competitive product with improved technical performances.
Paper presents a case study of a low efficiency 0.75 kW SPM currently existent on the market (efficiency
bellow EFF 3 Class). Input data were obtained by evaluating existent SPM population reliability based on
“fault tree” method. The motor was re-designed to an EEEM – EFF 1 Class with higher life expectancy,
15…16 % energy efficiency improvement, being manufactured at a competitive price. Further energy
savings were obtained by improving power factor of the motor (reducing reactive power costs). In-house
and site tests results validated the design. Currently, the product is part of a patented series of SPMs.

1. Setting the Problem

1.1. Short presentation of application
Standard SPMs are TEFC IP55, IP41 types, fitted together with water pump in an approximate 900 x 600
x 600 mm cubicle enclosure (CE) glass enforced by resin impregnation. The application passed
homologation tests performed long time ago by South African Standards Authority. The tests have been
performed on dedicated stands, at continuous running regime, with normal (clean) water.

However, on site, the application recorded a relative high failure rate. As a result of unexpected specific
working conditions characterizing the application, upon investigations it was confirmed that the motors are
the “weak point“ of the application.
• The motors are not matching the load requirements (not always running continuously): numerous
stop/startings at very short interval are occurring.
• The water is not clean but contaminated with foreign substances as: atmospheric dust and sand,
muddy, mulch, etc.

Rotor and stator electric losses represent about 2/3 of the total losses. Critical states are the overload and
re-starting conditions when these losses are higher. During a motor re-start attempt after overload
tripping while the motor starts, the starting losses are 150 to 220% higher than continuous motor running
losses. If the overload sensor keeps tripping the motor temperature will rise beyond its standard
conditions. This situation (with many re-starts in row) is very detrimental to the motor life expectancy.
Subsequently accelerated ageing effects on winding insulation and bearings are occurring.

The condensation is appearing in specific climatic conditions, (when motor stops) due to unusual high
temperature gradients into the motor. This condensation is very detrimental to the bearings and

1419
lubricants, especially. For specific SPM types, the end user has chosen to drill draining holes at the
bottom of the motor enclosure (the motor enclosure becoming not IP55 anymore). This measure is
proving the fact that motors may run better in a close contact with environment.
Nevertheless, because of high failure rate of SPMs, trying to prevent unplanned stoppages and
warranties, the end user preferred to oversize the motors. Subsequently, SPMs are running at lower
performances than their rated values with much higher losses (125%…135 %) on active and reactive
power.

1.2. Loading an oversized motor and efficiencies.

Oversized SPMs frequently are occurring in this specific application as a consequence of various
allowances accumulated during the process of design, procurement, commissioning and imposed by
reliability (warranty) conditions. These allowances are motivated by various reasons mainly related to
SPM lifetime and guarantee period. As a result the SPMs are loaded in a range of 50%…70%, running at
lower efficiencies and power factor. Table 1 shows a typical example of a 1.1 kW SPM efficiencies when
running at various loads.

Table 1: Efficiency and power factor variations of a standard 1.1 kW SPM (hot) @ different loads
No Shaft power Input Efficiency Power factor Combined: Efficiency x p.f.
[Watts] [Watts] [%] (p.f.)
1 1350 1980 68.2 0.97
2 1100 1570 70.0 0.95 100 %
3 850 1180 67.4 0.90 91.2 %
4 580 860 59.0 0.76 67.4 %
5 290 580 50.0 0.54 40.6 %
Table source: Femco 150100069/QPS 20

At 50 % load the motor will run 16 % under rated efficiency and 20 % under rated power factor with
corresponding increases in electricity bill (combined efficiency x power factor reduction are 32.6 %, if 100
% has being considered rated values). The annually energy losses are reaching 1570 kWh/motor.
Selecting the correct size motor for an application can increase motor efficiency.

2. Reliability Analysis of SPM by Using Fault Tree Methods

2.1. Fault tree method
Since reliability is the reciprocal of failure, and failure is a random event, probabilistic measures are most
appropriate. The laws of probability theory shall apply. A simplistic example is to consider a motor having
6 (six) identical components with equal Mean Time Between Failures, MTBF = 3 years. The failure rate
function of MTBF is estimated as:
λ = 1 / MTBF (1)
Then we would expect 6 x (1/3) = 2-device failure per year, which means SPM, will fail every 6 months!
Fault tree analysis is a method of combining various component failure rate, first proposed by HA Watson
to analyze the Minuteman Launch Control System [1], [2]. This method models a system failure of
interest called Top Event (TE) i.e. SPM failure. The fault tree breaks down TE into lower-level events
(LLE).

Figure 1: Gate “OR” and gate “AND” description

Logic gates show the relationship between LLE and TE. Mathematical model is based on Boolean
algebra. The “OR” gate showed in figure 1 (a) express the idea that any of several component failures

1420
can cause output event, here, TE = SPM failure. The output event failure rate λo function of LLE failure
rates λi is:
λo = Σ λ i [1/hours] (2)
The gate “AND” showed in figure 1 (b) express the idea that both (or all) components must fail in order to
produce the output event. The output event failure rate λo function of LLE failure rates λi is:
λo = Π λi [1/hours] units (3)

2.2. Fault tree of a SPM

Regarded as a product with unique (simple) function (no reparability feature), SPM design as EEEM is
based on reliability concepts as” the aptitude of a product to fulfill correctly the specified functions in a
given period of time and in prescribed running conditions”. The failure rates estimation from processed
statistical data on MTBF is shown in Table 2. SPMs have been considered running in average 12 hours
per day, 365 days per year.

TE: SPM failure, λm = 207·10-6, MTBF ≈ 1.2 years

OR

Rotor
Running
winding
DE Bearing Environment

Starting winding NDE Bearing

Figure 2: Basic fault tree designed for a SPM.
Basic fault tree of an SPM is presented in figure 2. TE failure rate is expressed in 10-6 [1/hour]. LLE are
the bearings, windings and rotor failures, and environmental conditions, too. Any of them can produce TE
= SPM failure.
2.2. SPM evaluation based on fault tree method
SPM fault tree is the basic point in evaluation of a standard SPM behavior from reliability point of view.
According to table 2:
• SPM failure is occurring beyond warranty period (ensuring “business sustainability”!).
• Declared “weak points” of the motor are the windings and drive bearing, with environment having a
substantial contribution.
Table 2: Basic failure rates used in estimation of failure probability for standard 0.75 kW, SPM.
SPM Component MTBF λi x 10-6 Lifetime expectancy
[hours] [1/hour]
Starting winding 23490 42.5 5.3 years
Running Winding 24360 41.0 5.6 years
Drive bearing 26100 38.31 6 years
Rotor 30960 32.3 7.3 years
Environment 30450 32.8 7 years
Non-drive bearing 48550 20.6 11 years
Total for SPM 207.51 1.2 years
Table source: Femco 150100069/QPS 20

The failures of the “weak points” were found to be mostly due to thermal stresses: overheating, poor heat
transfer non co-ordination of heat transfer and cooling.

1421
As shown in figure 3, the SPM is an IP 55 type of enclosure & IC41 [3]. Because of the external fan
position, nondrive endshield (with its bearing) is the most effective in the cooling process and its weight in
heat radiation is in a range of 10%…15 % of the total heat evacuation. The casing and ribs are partially
cooled by air fillets going along. However, there is an obvious cooling gradient along the casing due to air
slight and progressive separation from the casing profile.

External fan installed

opposite to the load Drive end bearing on the
load side cannot be cooled
by the airflow
Progressive separation of
the airflow

Figure 3: Standard SPM (IP 55, TEFC) with pump. source: Femco 150100069/QPS 20
In a TEFC motor, the active materials do not take a direct cooling from airflow generated by the external
fan placed on non-drive end side of the motor. The heat dissipation is based on:
• Stator winding heat transfer by conductivity to the casing and ribs
• Casing and ribs heat transfer by conductivity and radiation (being taking away by the air flow)
• Rotor heat transfer by radiation to the winding and by conductivity to the bearings
Due to specific working conditions the rotor may influence negatively the behavior of the motor
components.

3. Directions of the Heat Flux Transferred from the Rotor

Paths of the heat radiated by the rotor in a TEFC motor are shown in figure 4. Rotor fanning effect does
not influence consistently heat dissipation. The heat flux is transmitted along 2 (two) directions.
AXIAL direction: Overheat produced into the rotor is conducted along the shaft producing bearing
journal expansion, reducing bearing clearance inducing eventually bearing failure. Rotor life span can be
drastically reduced because of bearing collapsing and rotor rubbed on the stator. A classic example is
known as “bow rotor” when, as a result of a local thermal vector, the active material is emerging out from
the rotor

1422
Figure 4: Rotor’ heat flux paths toward bearings and winding source: [4]

RADIAL direction: Rotor’ heat flux radiated on radial direction is increasing stator temperature gradient.
Hidden voids in aluminium cast material produce so called “hot spots” seen as a “circle of fire” effect on
the stator winding as shown in figure 5.The same route is applying for the heat generated by the rotor
during the overload, starts and re-starting. Thermal stress is ageing winding insulation and bearings.

Winding being unevenly

thermally stressed produced
interturns shortcircuit.
The hot spot of a thermal local
vector situated on the rotor
overheat the stator winding as
“circle of fire”.

Figure 5: Winding affected by rotor’ heat radiation

Photo source: Femco 150100069/QPS 20

4. Investigating SPM Weak Points”

4.1. Analyzing drive end-bearing failure

In a TEFC, B5, IP55 motor, the drive end bearing is situated in a worst position. Besides generating heat
by itself from the load to be taken, bearing is receiving extra heat from adjacent components.
• Heat received from the winding overhangs by radiation
• Heat received from the rotor via shaft by conduction
• Heat received from the sealing arrangement (friction of rotating shaft against the seals lips) via shaft

As shown in figure 3, the drive end bearing does not take a direct cooling from the airflow generated by
the external fan placed on non-drive end side of the motor. During load working time the heat dissipation
of the bearing is function of the drive end shield material conductibility and end shield radiating area
against motor enclosure and against pump. Any extra heat transferred by thermal conductivity via drive
end shield it takes a time delay due to thermal inertia (confirmed by measurements to be in a region of
10…15 minutes).

1423
It is well known that fractional motors perform high slip rotor features. Electric losses on the rotor (and
heat generation, too) are direct proportional to rotor slip. The critical situation (as over-heat generation)
intervenes at re-start and overload conditions.
During motor re-start in the first 10 to 15 minutes running conditions, the bearing will be overheated (and
possible collapsed) as a result of an additional thermal shock at motor re-start (superimposed on the
existent “hot” condition). This thermal shock is occurring in relative short time “Δt” and may be considered
as “quasi-adiabatic” [5] as demonstrated in the following thermo-dynamic heat transfer calculations.

Figure 6: Basic graph of the rotor’ overheat flux

transferred by conductivity via shaft.
Figure source: [5]

Consider a drive end bearing temperature rise Trise = 30°C @ ambient of To = 35°C. It result the bearing
is running at Tb = 65°C. After re-start, the rotor of mass Mr = 0.5 kg and average specific heat c = 2.2
cal/kg °C records a temperature Tr =150°C (minimum δT = 55°C/start) as “thermal shock”.
According to figure 6, heat developed into the rotor HS, is transferred via shaft area As = 78.5 10exp(-6)
m² by conduction to the “cold area” (drive bearing). Steel mean thermal conductivity is K = 14.4 cal/m sec
°C and average distance between the rotor bars and bearing is X = 0.1 m. Time duration Δt of the heat to
be transferred from the rotor to the bearing via shaft is estimated according to equation 4 [6]:

Mr x c x δT = K x As x [(Tr – Tb)/X] x Δt (4)

It needs Δt = 153 seconds for the thermal shock heat flux generated by the rotor to reach the bearing
inner ring. The bearing inner ring (diameter Do = 30 mm) dilatation due thermal shock is estimated
according to equation (5):

D = Do [ 1 + α ΔT] (5)

For ΔT = 85°C and α = 15.7·10 exp. (-6) it results an increase of 36 microns of the bearing inner race
diameter, while the maximum radial clearance for this type of bearing is only 25 microns. As a result of
inner ring dilatation beyond internal clearance allowance, the bearing will overheat in a “snow ball” effect
reducing its clearance until collapse.

4.3. Thermal shocks in the windings during re-start and overload conditions.
Consider the motor run in “HOT” thermal stabilized conditions. Extra heat generated during transients
overload or re-start conditions will be superimposed onto existent thermal equilibrium. Rotor heat
radiation will stress thermally the windings as shown in figure 4. During these transient situations a new
thermal equilibrium has to established as a result of:
• Starting currents circulating into the winding
• High currents circulation in rotor bars and short-circuit rings
-4 2
Consider copper thermal conductivity K = 91.8 cal/m sec °, slot cross section As = 1·10 m and
temperature difference between rotor Tr = 150°C and winding Tw = 120°C, (Tr – Tw = 30°C). The heat
transmission to a distance of X =0.1 m will occur in a “quasi adiabatic” thermal process with Δt = 10 to
15 seconds obviously overheating the windings insulation system. Time duration of thermal stress applied

1424
to windings will depend by the cooling thermal constant (heat has to be transferred outside of the motor
casing and radiated in the airflow stream). Windings will be overheated unevenly, until overload protection
sensor will trip.

5. New Swimming Pool Motor

5.1. Essentials of application engineering
As demonstrated, standard SPMs do not comply with five essentials of application engineering.
i. Matching the driven machine conditions.
ii. Matching the power supply conditions.
iii. Matching the environmental conditions.
iv. Matching the reliability conditions.
v. Matching the business sustainability conditions.

5.2. The targets of the new design

The new design targets were to transform existent low efficiency SPM in an EEEM without alteration of
main components.
• Stator and rotor geometrical sizes.
• Rotor and stator slots profiles
• Quantity and quality of the laminations raw material
• Quantity of aluminium used for the rotor electric circuit
• Quantity of the copper used for the windings
These targets are imposed by the above five essentials of application engineering with special references
to specific conditions of the mass production technological process and business sustainability.

5.3. Basic steps in increasing SPM reliability and efficiency.

Has been demonstrated that SPM failure rate is function of thermal ageing effect. That means the
process of heat transfer and evacuation has to be addressed before any changing of SPM
electromagnetic design.

First step is to re-direct the cooling flow directly to the declared “weak points” (winding, bearings and
rotor) in dispersing the transient overheat and “thermal shock” effects. As a result, the components
ageing effect will be suppressed.

Second step in efficiency improvement is obtained by direct cooling of active materials with consequence
of reducing motor temperature rise. This will reduce stator and rotor copper losses RI² in proportion.

Third step is obtaining by reducing friction (existent in double seal’ arrangement) and windage losses.

Fourth step is obtaining by minimizing the increase of specific losses [W/kg] of the laminations by
improving technological process (degassing, annealing, and punching). Specific losses of the lamination
packs were reduced with 25 %. Reduction of the core losses enabled an increased magnetic loading,
reducing electrical loading in proportion and subsequently RI² Joule losses.

6. Patent description [7]

New SPM (shown in figure 7), EEEM type denominated as “Evolution” (Ev) was designed and
manufactured according to conditions presented above. The motor stator core (including the winding) is
hold in position by two “spider brackets” (SB). The long screws holding the brackets have different length,
according to different core lengths enabling motor power adjustment in a range of 0.6…. 1.5 kW for the
same laminations profiles.

1425
 Spider brackets - SB

Fan & LAVAL

profile gap

Figure 7: Cross-section of new SPM, EEEM

(180° rotated) Photo source: [7]

Legend of the arrows indicating air-flow

Blue = inlet air
Sky blue = air in the winding overhang
Lime = air into the airgap
Orange = envelope of air between back iron and casing
Gold = cooling air for drive bearing
Red = Outlet air

By construction and tolerances, SB play a major role in motor functionality by ensuring:

• Mechanical alignment on radial and axial directions between stator and the rotor;
• Space for winding overhangs;
• Air flow guidance (with reduced turbulence and aerodynamic resistance);
• Suitable fitting for casing

A special profiled fan (low windage losses) is attached to the shaft, outside of nondrive spider bracket, but
inside the motor enclosure. Air is entering axially into motor enclosure. Throughout an aerodynamic
profiled gap between fan’s fins and the casing (LAVAL profile) the fan creates a suitable airflow inside the
motor enclosure, the air being accelerated without turbulence towards:
• Nondrive bearing;
• Windings overhangs;
• Motor air gap (plus windings and rotor);
• Back of the stator core (air envelope in the space between casing and stator core)
• Drive end bearing.
At the end, all stream flows are re-united on drive end side. Outlet air apertures are available at the
bottom of the motor.

7. Testing Activities & Results.

7.1. Basic comparison between standard and new Evolution (“Ev”) SPM designs.

Table 3 shows a comparison between standard and “Ev” series of 0.75 kW, 2 poles, SPMs.

1426
Table 3: Performances comparison of SPMs
SPM Items Standard New “Ev”
Rated output power [kW](ii) 0.75 0.75
Rated speed [r/m] 2787 2898
Rated current [A] 4.96 4.2
Current in main winding [A] 4.64 3.93
Current in auxiliary winding [A] 2.16 1.83
Power factor 0.910 0.952
Input power [kW] 1.095 0.925
Rated efficiency [%] 68.5 81.1
Stator copper losses [kW](iii) 0.126 0.095
Rotor losses [kW](iii) 0.070 0.031
Iron losses [kW](i) 0.100 0.039
Friction and windage [kW](iv) 0.049 0.010
Total losses [kW] 0.345 0.175
Temperature rises [°C](v)
Main winding [°C] 77.2 40
Auxiliary winding [°C] 75.6 45
Drive end bearing [°C] 55 30
Non drive bearing [°C] 30 10
Rotor [°C] 150 93
Table source: Femco 150100069/QPS 20 & [7]
Motors performances were assessed on dynamometer, real conditions as shown in figure 8.
.
SPM loaded on
dynamometer

Power analyzer
Voltage regulator
Computerized
acquisition data
system with display

Figure 8: Dynamometer tests on SPM

Source: FEMCO test bay

7.2. How were obtained such performances?

i. By changing the balance between magnetic loading and electrical loading in electromagnetic design
(increasing magnetic loading as a result of improving lamination specific losses [Watts/kg]).
ii. Unchanging geometrical dimensions of electromagnetic design but reducing electrical loading.
2
iii. Reducing thermal level of the active materials and subsequently RI losses due to reduced electric
loading and due to direct cooling of the windings and rotor.
iv. Reducing friction and windage as a result of a new cooling principle and heat coordination.
v. Reducing temperature rises (thermal stress) of the motor components slowing ageing effect and
minimizing the failure rates of these components.
The results of energy efficiency improvement validated by site validation activity are shown in figure 9.

1427
 Efficiency %
85
80
New SPM “Ev” efficiencies domain
75

70
65
Standard SPM efficiencies domain
60
Load [%]

25 % 50 % 75 % 100 % 125 %
Figure 9. SPM, EEEM efficiency domains improvement compared to standard
Figure source: Femco 150100069/QPS 20 & [7]

7.3. Matching five essentials of application

The motor new design is matching the load conditions: impeller performances are matching motor
performances, the motor running at its rated values (essential one).
Reducing electrical loading of electromagnetic design enabled a better motor behavior in different
conditions of the power supply (essential two).
The design structure complies with environmental conditions required by application (essential three).
Ageing effect was suppressed by reducing thermal stresses. According to very conservative estimations
of components failure rate presented in table 4, it results that SPM “Ev” predicted lifetime is 2.7 times
bigger than standard motors (essential four). Estimations have been done according to Montsinger and
Dakin formulas using Svante Arrhenius chemical rate equation [8]. Additional endurance tests performed
on stands simulating accelerated ageing conditions confirmed the results.

Table 4. Failure rate estimation of new “Ev”, SPM.

-6
Component of new MTBF λi x 10 Lifetime
“Ev” type [hours] [1/hour] expectancy
Starting winding 86000 11.6 20 years
Running Winding 96000 10.4 22 years
Drive bearing 77500 12.9 12 years
Rotor 75500 13.2 7.3 years
Non-drive bearing 91500 10.9 13 years
Environment 55600 17.9 11 years
Total for SPM 76.9 3.1 years
Table source: [7]

Regarding essential five, minor changes of technological process occurred with reduced expenses
(production costs have been consistently reduced).
- Tooling maintenance interval was shortened.
- Lamination processing was improved.
- Winding specifications were simplified.
- Casing, cooling system (fan cowl) and rotor costs were considerably reduced.
- Labour reduction in winding and assembly process
- Assembly processes become simpler.

1428
8. Conclusions
It was demonstrated that a multidisciplinary approach has to be considered when designing motors. This
exercise proves that statistical probabilistic methods might revolutionize the physical shape of any
product.

Proposed method ensures minimum changes and costs of EEEM manufacturing process. Energy
efficiency concept was accomplished by:
• Reducing of daily power consumption in a region of 2 (two) kWh per SPM, or 8400 kWh per
year.
• Reducing ownership costs, indirect costs related to pool water damages and logistics, by increasing
product life expectancy and reducing failure rates.
• Reducing SPMs manufacturing costs.

References
[1] Military Handbook. Reliability Prediction of Electronic Equipment, MIL-HDBK-217F, USA Dec.
1991
[2] Schweitzer, E.O. Reliability Analysis of Transmission Protection using Fault tree Methods,
Schweitzer Engineering Labs, Inc, Pullman, WA, USA, 1997, ed@selinc.com
[3] SABS IEC 600034: Rotating Electric Machines, IEC Specification, 1998, ISBN 0 – 626 –08877 - 1
[4] Pitis C.D. Novel method of improving squirrel cage induction motor performance by using Mixed
Conductivity Fabricated Rotor (MCFR), Ph.D. Thesis, North West Univ., No. 20421133, Pretoria,
March 2006, http://www.nwu.ac.za
[5] Pitis, C.D. Thermo-dynamic calculations on over-temperature protection of equipment, Vector,
April 2005, pp.30…33, vector@ee.co.za.
[6] Mitton, R.G. Heat, J.M.Dent & Sons Ltd, Aldine House, Bedford St. WC2, London, 1957, pp.
252…277.
[7] Frahm, R, Niekerk, A, Nola, N, Pitis, C.D. A Swimming Pool Electric Motor, Patent PA 138282/P,
Spoor & Fisher, Johannesburg, Jan. 2005, j.fiandeiro@spoor.com
[8] WILEC. Thermal Life Predictions on Motor Insulation, Wilec ©, Johannesburg, May 2002

1429
1430
A Tale of Two Topics: Evaluation and Incentives in Utility
Regulatory Award Mechanisms For Energy Efficiency Programs
And Measuring Indirect Non-Energy Benefits --A Poster Session
Lisa A. Skumatz, Charles Bicknell, John Gardner

Skumatz Economic Research Associates, Inc. (SERA)

Abstract and Introduction

This poster session addresses two very different and distinct topics related to energy efficiency
program evaluation:1
• The issue of the use of well-designed financial incentives to encourage implementation of
effective energy efficiency programs by utilities is examined. In this section, the poster reviews
four different designs of award mechanisms that have been used in the US, and discusses
advantages and disadvantages of each. Then, the apparent performance of each award
mechanism is examined based on the findings from an extensive review and verification
performed on the activities, outcomes, and rewards for four award periods and four utilities in
California. Information on award mechanisms from other states, and proposals for modifications
to generate improved performance are also presented.
• The complexities of deriving estimates of the indirect non-energy effects from energy efficiency
measures and programs – and feasible approaches – are presented. In this section, we provide
background and methods for estimating non-energy benefits (NEBs) in a credible and defensible
manner. The pros and cons of alternative measurement approaches, and issues related to bias,
implementation ease, and applicability to different program types are presented. Results
applying these advanced measurement approaches to several US programs – both residential
and commercial – are provided for illustration and discussion.
The poster presents charts, graphs, and results to compare the performance and impacts of
alternative methods of addressing both these two topics.

Lessons in Incentivizing Energy Efficiency: Review of US Energy Efficiency

Award Mechanisms for Utilities
Introduction
The authors conducted a review of alternate versions of award mechanisms that have been proposed
and/or used to incentivizing utilities for energy efficiency programs and investments around the nation.
The work tracked changes in mechanisms over time, examined the impacts of these modifications on
activities and outcomes, and compared the results to provide “lessons learned” in the design and
implementation of incentives. Advantages and disadvantages of the various award mechanisms were
assessed from two perspectives – from the perspectives of
• providing a performance incentive and
• verifying achievement.

There are pros and cons in the ability of the award mechanism to encourage effective progress in the
delivery of energy efficiency services and programs by the utilities (first perspective). Secondly, there
are advantages and disadvantages in terms of how readily and unambiguously the activities can be
verified (second perspective). Types of award mechanisms reviewed by the authors included:
• expenditure-based,
• accomplishment-based,
• milestone-based, and
• retention-based.

The authors found some of these mechanisms to be more straightforward and effective than others.
Results showed that the award mechanisms varied in their suitability for various program types and
1
Combining these two separate topics into one poster session was requested by EEDAL. As such, each discussion is
truncated. For more information on either topic contact www.serainc.com.

1431
their abilities to reflect progress in different types of program goals; however, some common themes
were identified. The authors examined the strengths and weaknesses of different award mechanisms
and developed recommendations on specifications.
Overall, the question is how metrics can best be designed to carry forward the original intent related to
comprehensive, effective, and efficient service delivery of programs; lead to well and directly carried
out program and service activities delivering this service; and support reflection of those efforts and
intentions through reasonable documentation. The results have implications for states and regulatory
agencies looking to refine their energy efficiency incentives and design the “next generation” of
mechanisms to incorporate “best practices” from national experience.

Background on the Project

The Annual Earnings Assessment Proceeding (AEAP) provides the basis for California Public Utilities
Commission (CPUC) review of the level of success achieved by the four California IOUs (independent
owned utilities) in implementing a variety of energy efficiency (EE) programs. The AEAP reviewed
was a consolidation of the IOUs’ AEAP applications filed in 2000, 2001, 2002 and 2003 for
shareholder incentive earnings for Program Years (PY) 1999-2001 and for PY 2002
accomplishments. The AEAP review process is the mechanism that determines how the CPUC
distributes IOU shareholder earnings to each of the four utilities (Pacific Gas and Electric (PG&E),
Southern California Edison (SCE), San Diego Gas and Electric (SDG&E), and Southern California
Gas (SCG) based on their activities and accomplishments. The review process included an
assessment of the achievement of energy efficiency program related milestone goals by the IOUs.
2
Skumatz Economic Research Associates (SERA) was hired to conduct an independent third-party
assessment of energy efficiency program-related milestone goals and to identify the shareholder
earnings claims that could be validated and thus, should be paid to the utilities. Three sets of PY
assessments conducted as part of this project:
• PY 1999-2000, which was based on cost reimbursement for program delivery, but also provided
an additional incentive based on superior or acceptable achievement of program-related
milestones;
• PY 2001, which provided incentive awards partially based on achievement of energy savings
targets, and partially based on market transformation and program implementation targets as in
the previous years;
• PY 2002, which included no shareholder earnings. Instead, reimbursement of a portion of
program costs was at risk if reasonable efforts toward meeting program accomplishment goals
were not made.

The steps that were conducted for this work included:

• Reviewed the utilities’ AEAP filings and developed requests for documents to establish milestone
goals, award values, and verification requirements;
• Developed a milestone claims inventory and checked both that the claims were consistent with
Commission-approved milestones and that individual claims summed to totals claimed by the
utilities;
• Selected a sample of milestones for detailed review and analysis;
• Submitted claims verification requests to the individual utilities related to the selected milestone
claims; and
• Assessed milestone award claims to identify which portions of award claims could be supported
based on a review of the documents submitted, and which portions of dollar award claims may
possibly be at risk.

Across the years for which milestone incentives were available, utilities pursued a variety of different
milestone types, each with a unique set of measurement metrics and award mechanisms. In order to
facilitate the milestone verification process, the SERA team categorized milestones into four major
groups:

2
The project team for this assignment included Skumatz Economic Research Associates (SERA) and its subcontractors
Summit Blue Consulting (SBC), blueConsulting Company, and GEP (Global Energy Partners). SERA reviewed claims from
three utilities and Summit Blue reviewed one. Thanks to Leah Fuchs (SERA), Ingrid King (SERA), and Stuart Schare (SBC) for
suggestions helpful to the development of this paper. The client assignment consisted of a review of documentation and
support for award claims. Additional work to examine and compare a variety of award mechanisms was not was part of the
client assignment, but was undertaken and funded independently by the authors.

1432
• Energy Savings milestones, related to the achievement of specific savings targets for kW, kWh,
and therms (for PY01 only, including possible bonus awards).
• Progress Indicator-based milestones, related to indicating market progress toward
transformation of the market (used in PY02)
• Expenditure-based milestones, based on the utilities spending most or all of the approved
program budgets (e.g., Aggressive Implementation or Performance Adders)
• Miscellaneous milestones, which include all those classified as Administrative, Base, Activity, or
Market Effects milestones. Some were similar to Progress indicators, but were less directly
“market progress” related.

The list of award types starts with the indicators most directly related to some of the end-goals or
desired outcomes of programs; however, they are harder to measure and validate. Expenditure and
miscellaneous mechanisms are easier to measure and validate, but reflect indicators less directly
linked to outcomes and more geared to activities. Energy savings milestones perhaps most directly
reflect one of the key outcome indicators associated with the programs. Progress indicator
milestones explicitly recognize outcomes broader than energy savings, including market
transformation changes related to supply chain effects and market progress. For the California
project, across all the utilities and years, the SERA team checked 30% of the utilities’ milestone award
claims worth 50% of the total claimed dollars for the period, providing a wide sample of award types,
program types, and dollar awards at risk. The work took many months, and the records requested
filled several dozen full-sized storage boxes.

Pros and Cons of Award Mechanisms

Inspired by the project work, the authors then reviewed mechanisms from the project and other states,
and undertook work to analyze the strengths and weaknesses of alternative award mechanisms.
Advantages and disadvantages of the various award mechanisms are assessed from two
perspectives – from the perspectives of:
• providing a performance incentive and
• verifying achievement.

There are pros and cons in the ability of the award mechanism to encourage effective progress in the
delivery of energy efficiency services and programs by the utilities (first perspective). Secondly, there
are advantages and disadvantages in terms of how readily and unambiguously the activities can be
verified (second perspective). Overall, the question is whether the metrics carry forward the original
intent related to comprehensive, effective, and efficient service delivery of programs; whether it leads
to well and directly carried out program and service activities delivering this service; and whether
those efforts and intentions can be well reflected and demonstrated with reasonable documentation.
Each award mechanism has advantages and disadvantages; none is perfect. Those mechanisms
directly requiring delivery of energy savings may be harder to demonstrate and verify. Those
addressing “activities” may be easy to verify but may suffer from difficulties in that activities may not
be effective at capturing energy savings. Further, the ease of verification is affected directly by the
quality and comprehensiveness of records maintained and analyses conducted by the utilities.
Incentives need to address the cost of this data collection, or the utilities may find it is not worth
pursuing particular activities or goals if the cost of recordkeeping is too high. Although utilities may
wish to be reimbursed on the basis of activities (reflected through costs incurred), a balance may
need to be achieved between demonstrating efforts and effects. Advantages and disadvantages of
each award mechanism are summarized in the tables to be presented in the Poster session.

Summary And Recommendations

The authors examined the strengths and weaknesses of the different award mechanisms from the
perspectives of encouraging good performance and the ability to verify performance. The review
included an analysis of achievements under a variety of award mechanism designs used in California
over a period of years. Based on this analysis, verification issues could be addressed from both a
conceptual side and a practical side, and detailed results presented at the Poster session. As
discussed, the ability to verify is affected by several factors: the specificity of the intended goal or
achievement, the utility’s assessment of the cost to implement and maintain an appropriate tracking
methodology and the quality of tracking mechanism they implement as a result; and the work involved
in the actual verification. As a result, the utility’s incentives for pursuing particular goals are affected
by the incentive or award, balance by the cost to achieve and monitor. Therefore, incentives need to
be designed to take all these issues into account. Overall, the results of the review indicate that each

1433
type of award mechanism was more suited to meeting different types of program goals. However,
common themes related to what works included:
• For savings based awards: verification of saving levels is difficult; therefore, clear, approved, or
commonly accepted savings and lifetimes for buildings and measures need to be agreed upon at
the beginning of a program
• For progress / accomplishment based awards: the relevant indicators can be derived from
program theory and logic; however, best efforts being utilized toward an end was not an easily
independently verified award mechanism.
• For expenditure based awards: the acceptance of committed funds in a program year led to
verification difficulties and left questions related to the degree of follow-through in completing
committed projects.
• For miscellaneous milestone based awards: again, they can be derived from program theory/logic
and clear dollar amounts for attaining well defined and easily documented activity or other
milestones proved highly effective; however, electronic databases as a verification requirement
left room for uncertainty.

In constructing award mechanisms, outcomes are very important; however it is also important to
acknowledge that efforts are what incur costs. Given this, and the fact that there are inherent risks
that even good-faith efforts will not always result in the expected levels of outcomes, the best basic
concept may be expenditure-based awards requiring that either energy savings or progress
3
milestones be established and achieved in order to be eligible for the award. For example, such an
award might be structured as follows:
• If 2 GWh of savings are achieved (or 5,000 meters installed), the utility receives an award of an
additional 30 percent on the first $1 million in expenditures, and an additional 5 percent on the
remaining expenditures.
• If 3 GWh of savings are achieved (or 10,000 meters installed), the utility receives an award of an
additional 40 percent on the first $1 million, 25 percent on the second $1 million, and 5 percent on
the remaining expenditures.

This general type of structure may constitute a candidate for the next generation of awards by
providing an incentive for greater achievement by having the award increase with greater
accomplishments. It may also encourage efficiency by having the magnitude of the incentive
decrease with each increment of financing required to achieve the specified result.

Finally, in addition to assessing the structure of award mechanisms, the review also led the
researchers to provide suggestions about the actors that should be involved in designing
mechanisms. There may be issues that arise from the involvement of the utilities in the initial
establishment of the mechanisms and details of the awards (i.e. energy savings per CFL, and
threshold levels). It may be that these should either be written purely by the government agency
responsible, or by the agency in conjunction with an independent contractor that has no conflict of
interest with any potential recipient of the awards – with review and consent by the utilities. Review by
the utilities can provide valuable input on: whether the ultimate metrics are cost-effective, potential
monitoring costs or monitoring (M&V) options, adequacy of incentives, risks related to timeliness of
awards, and other factors.

The authors are continuing to review award mechanisms from other states to identify strategies that
provide the most effective balance of incentives and verifiability for awards for good performance in
delivering effective energy efficiency services. This work helps provide a basis for consistent and
effective design of award mechanisms in the future.

References
[1] Skumatz, Lisa A., Charles Bicknell, Leah Fuchs, David Bell, and Stuart Schare. 2005.
“Review of AEAP Milestone Incentive Awards, Program Years 1999-2002”, Prepared for the
California Public Utilities Commission, San Francisco, CA, October 2004.
[2] Skumatz, Lisa A., and Charles Bicknell, “Comparing Award Mechanisms - What Works?”,
Proceedings of the IEPEC Evaluation Conference, 2005.

3
Presumably derived at least partly from program logic models.

1434
Non-Energy Benefits: Measurement and Applications in Energy Efficiency
Evaluation
Measurement of NEBs
NEBs are becoming more recognized as relevant omitted program effects in evaluation. NEBs include
a variety of program impacts– positive and negative – that are beyond energy savings and result from
the program.4 Starting with work in the mid-1990s, the literature began to sort these benefits into three
“perspectives” (See Reference [6]):
• Utility / Agency NEBs: Net benefits accruing to the utilities or program-sponsoring agency, including
fewer billing-related calls and other follow-ups, lower bad debt from unpaid bills, lower T&D losses,
and other benefits, which result in lower revenue requirements for the agency, and are appropriately
valued at the agency’s marginal cost and discount rates.
• Societal NEBs: Net benefits beyond those accruing to the utilities / agencies or directly to
participants, including economic multipliers or job creation benefits, reduced environmental impacts
from emissions, and other benefits valued at societal costs and discount rates.
• Participant NEBs: Positive and negative impacts that are realized and recognized by program
participants. For residential buildings these translate to greater comfort, lower noise, and a variety
of other benefits for residents. For commercial applications, they may include comfort, fewer tenant
complaints, better ability to sell or lease the property, productivity, fewer lost days at work, and many
others. These effects are measured using valuation methods appropriate to the residents or
owners.

Many of the net effects in the first two categories can usually be measured using programmatic, utility,
5
and secondary data. However, measuring the effects realized by the third category – participants – is
more complex. The benefits to participants derive from several main “drivers” – specifically “net”
impacts from:
• Payment and collection-related effects,
• Education and knowledge of energy use, building, and equipment,
• Changes in Building stock / building value,
• Health-related changes,
• Direct and indirect changes from equipment service (including comfort, O&M, services, etc.)
• Changes in other utility bills (e.g. water bills, etc.), and
• Other changes.

Well-researched measurement work on NEBs, based on detailed literature research were pioneered in
6
the late 1990s, and followed up with detailed research over the last ten years. Granted, NEBs are,
almost by definition, Hard to Measure (HTM); however, not measuring the effects means that decisions
about programs are likely to be suboptimal because they ignore the effects. Running scenario analysis
around ranges or order of magnitude values would be preferable to excluding the impacts altogether.
Thus, approximate estimate provide value; the improving sophistication of measurement methods
implies that these approximations are getting better and better.

To provide credible estimates of the participant NEBs actually attributable to the program, the results
must be “net” and represent only the effects due to the program or measure – above and beyond what
would have occurred without the program. The authors have addressed these concerns in important
ways, presented in the Poster. The authors have also researched, tested, and applied a number of
measurement methods for estimating NEBs. These include:
• Willingness to pay (WTP) / Willingness to accept (WTA) / contingent valuation (CV),
• Comparative and scaling approaches,
• Direct measurement and regression approaches,
• Discrete choice, ranking, and logit methods,
4
The literature historically calls these effects “non-energy benefits” even though they may be negative in the “net”. There have
been several suggestions to call them non-energy effects or non-energy impacts (See Reference [3]), using the traditional term
better respects the literature, and there is nothing lost by calling them net-NEBs or NEBs, and the literature remains more
robust.
5
A model for estimating all three categories of NEBs has been developed (“NEB-It ©SERA) and results from programs using
this model will be demonstrated at the poster session.
6
Measurement methods have been discussed in detail in previous papers including in References [5] and [7]. Choice models
have also been applied in several projects, including projects in this paper, with strong results. Results for this paper were
gathered via phone, in-person, fax, and web approaches.

1435
• Other revealed and stated preference approaches, and
• Other approaches.

Each method has advantages and disadvantages, and an assessment of the advantages,
disadvantages, and illustrative outcomes are presented as part of the poster session.
The authors have conducted detailed work on NEBs for scores of residential, commercial, and other
programs (low income, renewables, real-time pricing, commissioning, etc.). The findings have
demonstrated that the total of NEBs from the three perspectives can easily exceed the value of the
direct energy savings from the program. Although NEBs from the utility point of view tend to be small,
extensive work by the authors on the societal benefits and the participant benefits show estimated
NEBs to range from 25% to 400% of the value of program savings, depending on the program design,
measures included, region of the country, and other variables. NEBs are a significant and
measurable omitted program effect.

Application / Uses of NEBs

Typical process evaluations examine program design, delivery, and progress in an array of program
progress indicators. They gather scalable responses on program awareness, understanding,
strengths, weaknesses, barriers, and provide feedback on decision-making factors and a variety of
other program feedback. While typical process evaluations can provide input to many “researchable
questions”, the answers relate mostly to “more” vs. “less” of an effect – and fall critically short of
providing quantitative information that would help inform their importance or identify which, if any, next
steps are warranted (and cost-effective).
The work in non-energy benefits (NEBs) has progressed to the point where it can fill this gap and go
beyond typical marketing, targeting, and benefit-cost applications. A well-designed analysis of NEBs
provides key quantitative information on program strengths, weaknesses, and decision-making factors
that are not typically outcomes of process evaluations. State of the art practices in NEBs measure net
NEBs in the sense that they:
• measure both positive and negative NEBs,
• net of the impacts that would have been realized from the choice of new standard efficieny
equipment, and
• measure the impacts above and beyond what would have been experienced without the program
(net of free riders).

Measured well, the positive impacts provide dollar values of the value of the program’s benefits. More
importantly, the negative impacts provide dollar estimates of cost of remaining “barriers” participants
and non-participants associate with the equipment that are not addressed by the program. Recent
research by the authors illustrates the point, providing quantitative estimates of the dollar value of
remaining maintenance, labor, and other barriers for several programs. These estimates provide
dollar estimates of the unaddressed barriers, and thus provide estimates of the investment that might
be needed to address those barriers. Program planners can then make intelligent decisions whether
to address the barriers, given the investment requirements indicated. This is far more practical and
valuable than standard process evaluation feedback that notes a barrier score has fallen from,
hypothetically, 3.3 to 3.2 on a 5 point scale. Finally, NEB analysis can show the degree to which
program information is getting to key program actors – and the degree to which the understanding of
measure features is leading to underinvestment in efficiency.

Drilling Down in NEBs: Applications and Implications

Analysis of NEBs has wide applications beyond the simple “valuation” of the NEBs. Examining the
perceptions of NEBs that are positive and negative, and those that are most valuable, provide
information important to program evaluation, decision-making, marketing, and other applications.
• Marketing and Targeting: Highly valued NEBs are likely easier to “sell” than energy efficiency, and
more importantly, they are likely to appeal to owners or decision-makers. Tailoring the program
message to the high scoring NEBs for the audience of interest is potentially more fruitful than
continuing to push energy efficiency on efficiency or bill savings grounds. Furthermore, the
research on NEBs identifies those participants that gain the highest NEBs, which can provide
information for targeting the program.
• Benefit-Cost: The NEB values provide information for the benefit/cost analysis and payback
computations from participant point of view, and may be useful as inputs for scenario analysis
around regulatory tests as well. Some states are looking into scenario analysis around program
benefit costs analysis.

1436
• Barriers Analysis: Negative benefits are indications of program barriers that remain – either
perceived or real (or both) depending on which actors report the negative NEB. If non-participants
or vendors report a negative NEB but the participants do not, then the program may benefit by
providing greater education or data on that factor. The program would likely obtain more
applicants, and the vendors may be able to make a stronger case for the energy efficient
equipment. If, however, the barrier represents a real cost – if participants or others (A&E,
contractors) notice the problem as well – the NEB results provide an estimate of the cost of the
rebate, refund, warranty buy-down or other interventions that may help participants become
indifferent to the barrier – and spur participation and adoption of new measures. Tracking these
negative values over time provides useful information feedback to let program staff check whether
the program is decreasing these barriers over time.7 The dollar value provides information on the
level of investment that may be needed to overcome the barrier.
• “Disconnects”: The authors believe a robust evaluation of the NEBs gathers information from
multiple actors involved in the program, as well as non-participants. These results allow an
examination of differences in positive and negative perceptions about NEBs as well as differences
in associated values. Using this approach, our work has been able to identify situations in which
architects / engineers / contractors assign more “negatives” to NEBs than do owners – leading to
underinvestment in energy efficiency. The implication is that bids and construction may be including
less energy efficiency than owners might be willing to “buy”. Additional education, incentives, or
other program interventions targeted at those with more skepticism may aid the program; feedback
on the owner perspective may also help.

The work shows that “hard to measure” NEBs can be measured in all three perspectives – utility,
societal, and participant. The authors have applied this work to more than 50 residential, low income,
commercial / industrial, and specialty programs. Examples of these applications are presented in the
Poster session.

References
[3] Bicknell, Charles and Lisa A. Skumatz, Ph.D.. “Non-Energy Benefits in the Commercial
Sector: Results from Hundreds of Buildings.” Proceedings of the 2004 ACEEE Summer
Study, Asilomar, CA. August 2004.
[4] Skumatz, Lisa A.. “Non-Energy Benefits (NEBs) – A Comprehensive Analysis And Modeling
Of NEBs For Commercial & Residential Programs.” Florida: Proceedings of the AESP
Conference. December 2001.
[5] Skumatz, Lisa A.,” Comparing Participant Valuation Results Using Three Advanced Survey
Measurement Techniques – New Non-Energy Benefits (NEB) Computations of Participant
Value”, Asilomar, California: Proceedings of the 2002 ACEEE Conference. August 2002.
[6] Skumatz, Lisa A., and Chris Ann Dickerson. "Recognizing All Program Benefits: Estimating
the Non-Energy Benefits of PG&E’s Venture Partner’s Pilot Program (VPP).” Chicago, Illinois:
Proceedings of the 1997 Energy Evaluation Conference., 1997.
[7] Skumatz, Lisa A. and John Gardner. Differences in the Valuation of Non-Energy Benefits
According to Measurement Methodology: Causes and Consequences, NESP Conference
Proceedings, San Diego, California, January 2006.

7
This feedback is potentially more useful than tracking barrier “scores”, which provide less information on the importance of the
barrier before or after.

1437
1438
Energy+ Pumps – Technology Procurement for Very Energy
Efficient Circulation Pumps
Claus Barthel, Stefan Thomas

Wuppertal Institute for Climate, Environment, Energy

Abstract
Circulators in single or double family homes and flats cause about 2–3 % of the overall electricity
consumption of the EU. A new technology of pumps with electronically commutated (EC) motor
pumps is available now; it is one possible way to achieve a reduction in circulator annual electricity
use by 60 % or more.
If these new very energy-efficient pump technologies became the new technology standard for
circulators, they would save about 1 % of current EU electricity consumption, that is 25 TWh/year, and
reduce CO2 emissions by 10 million tons per year.
This paper gives an overview about the potentials, describes the technical background of the new
technology and presents a new “Intelligent Energy - Europe” project that addresses these potentials.
The project’s objective is a market transformation towards new very energy-efficient pump
technologies – Energy+ pumps – for circulators in heating systems, both stand alone and integrated in
boilers. The most important market barrier is a high initial price due to low production numbers.
Therefore, only few manufacturers have so far introduced the new pump technology to the market for
single or double family homes and flats; among these, the two most important ones on the European
market have recently launched it.
The short-term objective of the project is, therefore, to bring more products to the market from all
major manufacturers, to contribute to their market success, and to reduce their prices through mass
production. A conservative estimate is that the buyers aggregated by the project might purchase
10,000 Energy+ pumps during the project; this would save 2.5 million kWh per year. If at the end of
the project, a market share of 5 % has been achieved, it will mean annual savings of 100 GWh/year
from the products sold in that year alone.
In order to achieve its short-term objective, the project will adapt and apply the technology
procurement methodology as it was very successfully tested in the European Energy+ project on
energy-efficient cold appliances.
Large buyers, mainly (social) housing companies will be aggregated, to activate the pump and boiler
manufacturers. Sales and training materials, and sizing spreadsheet software for installation
contractors will be developed and applied. A competition both for energy-efficient products and
marketing campaigns will be organised and the information on the Energy+ pumps will be
disseminated widely through website, newsletter, media, and fairs.

Introduction
In the EU-15, the electricity consumption by circulators for heating purposes in households amounted
to about 41 TWh per year in 1998 [1]. This is caused by 87 Million circulators, most of them with a
power input below 250 W. But private households often do not even know that they have a circulator
in their heating system, much less do they know that the circulator is responsible for 5 to 10 % of their
electricity bill. As long as their rooms get warm, they don’t care about this and trust their installer that
he or she installed a good system. The consumer does not know the electricity costs he or she has to
spend for running the circulator and if he/she knew, he/she also might not care because the absolute
amount of money of about, say, 60 Euros per year is not so high that the consumer would try to spend
much effort to minimise this.
But for society the energy used by circulators is equal to between 2 and 3 % of the overall electricity
consumption and causes CO2 emissions of more than 20 million tons per year, so efforts to minimise
this would be worthwhile both for economic and for climate-change mitigation reasons.

1439
Potentials
A typical circulator used in European heating systems has a power input of 80 to 100 W. Several
studies show that this is far oversized (e.g. [2], [3]). Installation contractors tend to install a big pump
so as to receive no complaints from their customers; the contractors do not have to pay the electricity
bill. Normally, a smaller pump would be sufficient in a heating system. An additional issue is the
hydraulic balance: A correct hydraulic balance secures the same heat supply to all radiators. If the
heat supply is uneven, a stronger pump will be necessary to compensate this. So, if the hydraulic
balance were correct in a heating system, which is not the case under normal circumstances, even a
smaller pump could be installed. For the above mentioned example, a circulator with 35 W will be
sufficient [2].
For a number years, a new technology of pumps with electronically commutated (EC) motor pumps
has been available. By this high-efficiency circulator, a reduction in circulator annual electricity use by
60 % or more is achievable. Fig. 1 shows a comparison of the possible savings compared to a typical
situation.
By which technical measures has it become feasible to increase circulator energy efficiency? First, the
motor efficiency of “conventional” asynchronous motors of the size used in small-scale circulators is
around 50 %, whereas for EC motors, e.g., permanent magnet motors, this is around 60-80 %. In
case the high-efficiency motor is, as the second measure, combined with an improved impeller design
enabled by the high revolving speeds of the high-efficiency motor, the hydraulic efficiency can also be
raised from on average 35 % to 60 %. Combining these two measures achieves a total circulator
efficiency of approximately 40 %, compared to an average efficiency of 5 % to 24 % for circulators
with asynchronous motors.
100

90

80

70
Power input [W]

60

50

40

30

20

10

typical circulator, circulator, new

oversized well-adjusted high-efficiency
circulator

Figure 1: Comparison of the input power of different circulators. Source: [2]

This new high-efficiency circulator, that was first developed by the Swiss manufacturer Biral in 2000
and brought to the market soon afterwards (see Fig. 2), would save about 1 % of current EU electricity
consumption, that is 25 TWh/year, and reduce CO2 emissions by at least 10 million tons per year. For
the first few years, this new pump was not very successful on the market. It was caught in a vicious
cycle of high initial prices and low production numbers. Therefore, Biral remained for a number of
years the only manufacturer to have introduced the new pump technology to the market for single or
double family homes and flats. Only in the autumn of 2005, the two European pump market leaders,
Grundfos and Wilo, entered the market with a similar pump. With an end-consumer price of about 300
Euros, the price is still high but much more affordable now.

1440
 Zur Anzeige wird der QuickTime™
Dekompressor „Foto - JPEG“
benötigt.

Figure 2: Swiss manufacturer Biral’s EC motor circulator for single family homes, and
Grundfos´and Wilo’s actual product

The Europump label

Synchronised to the launching of the high-efficiency pumps in the market, Europump, the “European
Association of Pump Manufacturers”, proposed a pump energy label similar to the European energy
label for appliances. This label may act as a marketing instrument for the installation contractors and
promotes the high-efficiency pumps, because only these pumps can achieve the A-label. But this
label is voluntary and only new pumps are labeled. In addition it only covers stand-alone circulators,
built-in circulators are excluded.

A < 0,40 high efficiency drive

B 0,4 < EEI < 0,6variable speed drive

C 0,6 < EEI < 0,8unregulated
D 0,8 < EEI < 1,0unregulated
E 1,0 < EEI < 1,2
F 1,2 < EEI < 1,4
G 1,4 < EEI

Figure 3: The Europump Label Source: [4]

EEI = Energy Efficiency Index, calculated according to measurement standards and formulas described in [4]

The Energy+ Pumps project

Though the conditions for the introduction of the new technology in the market are given now, there
exist a lot of barriers. Current market barriers include a high initial price (300 Euros) due to low
production numbers, low customer interest caused by the fact that not the final customer but
installation contractors or even boiler manufacturers are usually choosing the pump, and vendors
selling to the final customer on product price only.
This is a notable difference to the market for medium-sized circulators (200 to 400 Watts of input
power, for office and other larger buildings), for which major manufacturers (Grundfos, Wilo, Biral;
KSB may follow) all have a range of EC motor pumps on the market. The reason is that this is a

1441
market of institutional buyers that specify the pumps themselves and are used to economic
calculations, hence easier to convince than the single homeowner.

Objectives of the project

This project will therefore mainly target the market for the small-scale circulators, to foster the market
entry and uptake of these.
The short-term objective of the project is, therefore, to bring even more products to the market (at
the end of the project we expect at least 10 pumps for space heating and sanitary hot water by at
least 3 manufacturers, and at least 10 condensing or low energy boilers with the new pump built in),
to support their rapid break-through in the market and thereby to reduce their prices through mass
production. According to one manufacturer, an annual production volume of 30,000 to 40,000 pumps
would be needed to start mass production and bring down prices.
We expect that the project will contribute to bringing small-scale Energy+ circulators to the market for
single or double homes and flats from all major manufacturers, and to reducing the price premium
over conventional electronically controlled circulators for that market to less than 50 % or 60 Euros. A
conservative estimate is that the buyers aggregated by the project might purchase 10.000 Energy+
pumps during the project; this would save 2.5 million kWh per year. If at the end of the project, a
market share of 5 % has been achieved, it will mean annual savings of 100 GWh/year from the
products sold in that year alone.

How to achieve the goals?

This will be achieved through the methods of technology procurement that have been successfully
tested in the European Energy+ project for energy-efficient refrigerators and freezers, co-funded by
the SAVE programme (cf. [5] and [6]).
The project will, therefore, bring together institutional buyers such as housing companies,
intermediate buyers such as boiler manufacturers, installation contractors and their associations, and
pump manufacturers to build mutual trust:
• confidence for the pump manufacturers in the existence of a market demand large enough to
start mass production,
• confidence for installers and boiler manufacturers that it will be feasible and easy to convince
final buyers of the benefits of the new pumps,
• and buyer awareness of the benefits of the new technology and confidence that a supply at
appropriate prices will be feasible.
The methodology of technology procurement therefore implies an active co-operation with the
following target groups.
• On the demand side of circulators: particularly large institutional buyers to sign procurement
declarations, but potentially all building owners as purchasers of the new kind of pumps, or
boilers that have these pumps built in;
• On the supply side of circulators: pump and boiler manufacturers, as well as installation
contractors and their associations;
• For support of the dissemination of information: further national, regional, and local energy
agencies than those already in the consortium, environmental and consumer NGOs, energy
companies, associations of building owners and market participants, and other organisations
interested in the subject.

Potential large institutional buyers will be approached and invited to sign a procurement
declaration. This will either concern an indicative number of pumps, e.g., for replacement, or of boilers
that have the new pumps integrated. Final customers buying such pumps in larger quantities could be
big housing companies. Apart from bigger multifamily buildings where the bigger pumps are needed,
they may also have blocks of smaller multifamily or even row houses, or they may have heating
systems by apartment (with wall-mounted boilers). In the bigger multifamily houses, such small pumps
could be used in sanitary hot water circulation systems. Although the small circulators are at the focus
of this project, the procurement declarations could also include medium-sized circulators (200 to 400
Watts of input power, for office and other larger buildings) used in bigger apartment blocks, in order to
achieve synergies.

The pump manufacturers will be invited to develop and launch to the market pumps fulfilling the
criteria developed by the project, and to present them to the Energy+ team for inclusion in lists of
qualifying products. They are also expected to actively market their products once they are on the
market.

1442
Boiler manufacturers will be contacted, too. They are in an intermediate position as buyers of the
pumps, but manufacturers of products that are sold to the final customers. Hence, they must see a
potential market for boilers with the efficient pumps integrated, in order to become buyers for the
pumps. The project will build their trust in such an initial market.
Finally, further energy agencies, environmental and consumer NGOs, energy companies,
associations, and other organisations interested in the subject will be invited to support the project
with publicity. Governments and energy companies might also support the market uptake of the new
pumps with financial incentive programmes. This will indirectly target the whole market of all
building owners in order to achieve a wider market break-through for the new energy-efficient
circulators.
To further support this process and increase sales volumes, installation contractors must be
integrated in the project. They too must be convinced that there is a market for the new pumps and
that it is worth the effort to actively sell them to their customers. For this purpose, they must receive
easy-to-use materials to convince their customers and be trained about selling the efficient pumps
based on life-cycle costs. In order to achieve synergies, this will cover both the small and medium-
sized circulators.
The project will actively co-operate with professional training agencies, institutes and organisations,
associations of installation contractors, and pump and boiler manufacturers to organise the training
seminars.
None of the primary market agents (buyers, manufacturers, installers) are Members of the
Consortium. The reason is that the Co-ordinators of a technology procurement at European and
national level must be independent from those directly involved in the target market.
All of these market agents will furthermore be actively involved in the project from the beginning.
Their interest in the project and its objectives will be stimulated during the first work package with
direct information and interviews. They will then be invited to sign procurement declarations, support
declarations, or to file products to the lists. The participants and products will be featured on the
website and the other publicity organised by the project. As has been described above, the different
target groups – manufacturers, supporters, associations of installers – will also be actively involved in
the dissemination of project results.
The project will also enable a field-test of the appropriateness of the method for classification of
circulators proposed by the European manufacturer organisation Europump. The consortium will also
actively seek to co-operate with Member State governments and energy suppliers about potential
financial incentive programmes and awareness campaigns. E.g., the new pumps could be targets for
energy efficiency programmes in the framework of the Italian White Certificate system, or the
obligations for Flemish grid companies, or the Energy Efficiency Commitment in the UK.
Furthermore, in the context of the EU Directive on the overall energy performance of buildings,
reducing the electricity consumption for the circulator will make it easier to achieve the primary energy
performance targets for new and refurbished buildings. A close link is also intended to other planned
EU projects on related subjects.
The ultimate long-term objective is to transform the market so that the new technology will become
the standard technology, at prices close to those of today’s pumps. This seems feasible, since the
new technology enables smaller pumps and hence requires less material, and the extra costs of the
electronic controls are constantly decreasing.

The project consortium

Many members of the consortium for this project (7 out of 10) have been partners to the recent
Energy+ projects on household cooling appliances ([5], [6]).

Figure 4: The Energy+ logo

The consortium was rounded off with SEVEn to represent the Czech Republic, Escan to represent
Spain, and DENA to organise the award and share with the Co-ordinator the national work in
Germany.

1443
All in all, the team consists of:
• seven national energy agencies, who have a lot of experience in working with market actors
on the dissemination of energy-efficient technology, and of which four were already partners
to the Energy+ cold appliances project; one of the agencies (ADEME) co-operates with the
subcontractor SOWATT, which was assisting the agencies leading the last Energy+ project;
• two research institutions that were partners in the Energy+ cold appliances project and have
worked on applied energy efficiency research for at least 10 years;
• and a company that has a lot of experiences in energy efficiency markets as well.
• Furthermore, the Swiss company ARENA will co-operate with this project as an external
partner, acting as the national procurement Co-ordinator for Switzerland.
The team also represents all regions of the EU-25, and 9 countries with a total of around 120 million
households.
The following table presents an overview of the roles of the partners in the project.

Table 1: Partners of the Energy+ project

Partner Role in the project
Wuppertal Institute Project Co-ordinator;
shares national procurement co-ordination for Germany with
DENA
A.E.A national procurement Co-ordinator for Austria
eERG national procurement Co-ordinator for Italy
VITO national procurement Co-ordinator for Belgium
CRES national procurement Co-ordinator for Greece
Escan National procurement Co-ordinator for Spain
SEVEn national procurement Co-ordinator for the Czech Republic
MOTIVA National procurement Co-ordinator for Finland
DENA shares national procurement co-ordination for Germany with the
Wuppertal Institute
ADEME National procurement Co-ordinator for France
SOWATT Subcontractor of ADEME
ARENA External national procurement Co-ordinator for Switzerland

Work Programme
The first project phase will be a market and feasibility study (work package 2), for which 8 months are
planned. The market and feasibility study needs to fine-tune the methodology of co-operative
procurement to be used during the project and to define the energy+ product criteria. With interviews
and market research, the market structure and the barriers for the market introduction and diffusion of
the new circulator technology will be analysed. Furthermore, the willingness of the market players to
co-operate in the technology procurement will be assessed. Based on this research, the methodology
developed in the Energy+ project on cold appliances will be refined and adapted to the pump market.
After these preparations, the second project phase will implement the technology procurement for 28
months until the end of the project. This phase includes the work packages 3 (implementation of the
technology procurement with buyers, manufacturers, and installation contractors as well as
supporters, development of lists of qualifying products, interested buyers, and supporters eager to
promote the products, and development and application of sales and training materials and a training
course for installers), 4 (award, given to the most energy-efficient products and the best promotion
campaigns; also including an independent test of products, likely to include the award winners), and 5
(dissemination through website with product and participant database, newsletters, brochure, and a
wide media coverage organised by the partners as well as the supporters).
During the last 6 months of the project and in parallel to the second phase, an evaluation of the
project and its impacts will be carried out (work package 6). It will provide the basis for the final report.
Overarching these three phases are the work packages 1 (project management) and 7 (common
dissemination activities upon request of the European Commission).
The following flow chart provides an overview of the structure of the Energy+ pumps project and the
interaction of its work packages.

1444
 WP2: Market and Feasibility Study
WP7: Common Dissemination Activities
WP1: Project Management

WP3: Implementation WP4: Award

WP5: Dissemination

WP6: Evaluation and Report

Figure 5: Flowchart of the energy+ project

Conclusions
Since the Energy+ Pumps project has only begun in January of 2006 and is at the time of the EEDAL
conference at the stage of the feasibility study, it is to early to draw conclusions on the success of the
project. We therefore wish to give some indications on why we expect it will not fail.
There are only two critical steps where the project could fail: if either no manufacturers would list their
products, or no institutional buyers would declare their interest in purchasing the new, energy-efficient
pumps and boilers.
However, based on the experiences with the Energy+ project on cold appliances and on preparatory
talks with market players for circulators, we are confident that the technology and market procurement
approach of the Energy+ project will also work for the circulators.
For cold appliances, the Energy+ project started with two products from one manufacturer on the first
list and ended with almost 900 models from 21 manufacturers. This is unlikely for circulators since
there is only a handful of manufacturers and a smaller range of product types, but leads us to the
expectation that there will be interest from manufacturers for this project too. This is shown by letters
of support from three manufacturers and the fact that two major manufacturers have recently
launched small EC motor circulators.
On the buyer side, the Energy+ project for cold appliances attracted around 20 institutional buyers
owning more than 1 million dwellings, and dozens of retailers with more than 15,000 retail outlets
across Europe. Installation contractors and boiler manufacturers are playing a similar intermediary
role for the circulator pumps, as are the retailers for the cold appliances. We will, therefore, involve
these two types of market actors from the very beginning.
The success of Energy+ among housing companies also shows that it will be feasible to overcome
the problem of split incentives between landlords and tenants. This will certainly be easier in a market
with a surplus of apartments for rent, and/or a building certification in place, where the energy costs
are important for the ability to rent out apartments.

References
[1] Bidstrup N. Promotion of Energy Efficiency in Circulation Pumps, especially in domestic
Heating Systems. EU SAVE II Project, 2001, can be downloaded at:
www.eci.ox.ac.uk/lowercf/eusave_circulation.html
[2] Nipkow J. Klein-Umwälzpumpe mit hohem Wirkungsgrad. Bundesamt für Energiewirtschaft,
1994

1445
[3] Wolff D., Jagnow K. OPTIMUS –Optimal Energie nutzen, Technische Optimierung und
Energieeinsparung. Deutsche Bundesstiftung Umwelt 2005
[4] Europump Industry commitment to improve the energy performance of Stand-Alone Circulators,
2005
[5] STEM energy+ Aggregated purchase of energy efficient refrigerator-freezers ar European level.
EU SAVE project 2001
[6] SenterNovem 2E+ Procurement on very efficient white goods. EU SAVE project 2005

1446
Energy Consumption and Efficiency Potentials of Lifts
Jürg Nipkow; Max Schalcher

Swiss agency for efficient energy use S.A.F.E.

HTW Chur University of Applied Sciences

Abstract
Lifts can account for a significant proportion of energy consumption in buildings. In a research project
that was concluded in 2005, the energy consumption of 33 lifts from a variety of manufacturers was
measured. The most important finding concerned the surprisingly high stand-by consumption, which
accounted for between 25 and 80 percent of the total consumption. Average efficiency rates were
favourably high at around 60 percent. One notable finding was that modern hydraulic lifts can be just
as efficient as traction lifts thanks to the use of counterweights or energy storage. Overall efficiency
shall be achieved by reducing stand-by consumption and using energy-efficiency concepts and
criteria for architects, designers, planners and customers.

Current situation, objectives

The electricity consumption of lifts is dealt with in the new SIA standard 380/4 (“Electricity in
buildings”) [1], and is also an integral part of the declaration of overall energy consumption by building
systems (energy certificate for buildings). A research project that was concluded at the end of 2005 [2]
set out to extend the somewhat limited knowledge about the electricity consumption of lifts by carrying
out measurements. The broad-based project team was supported by the lift industry in that Schindler
Aufzüge AG provided technical assistance while seven other manufacturers provided the necessary
local support for the measurements.

Project partners
• Swiss Agency for Energy Efficiency (S.A.F.E.), project management
• Swiss Federal Office of Energy, SwissEnergy programme
• Swiss Society of Engineers and Architects (SIA)
• City of Zurich civil engineering office, and energy-efficiency fund of the City of Zurich
electricity works (ewz)
• Office for Environmental Protection and Energy (AUE), Basel-Stadt
• Schindler Aufzüge AG
• Other companies affiliated to the Association of Swiss Lift Manufacturers (VSA)
• S.A.L.T. (Swiss Alpine Laboratories for Testing of Energy Efficiency], Chur), implementation
of measurements

The main objectives of the project were to determine the level and structure of electricity consumption
by lifts in Switzerland and to identify efficiency measures and ways of implementing them in various
situations. For measurement purposes, typical lift systems in use in Switzerland were classified by
type, size, building type and lift properties (Table 1, Figure 1).

Table 1:No. of measured lifts by year of manufacture and building category

Year of manufacture Prior to 1980 1980 - 1997 After 1998 total
Residential buildings 1 1 12 14
Hospitals/clinics - 2 1 3
Shops - - 1 1
Office blocks 1 2 5 8
Car parks 1 - 3 4
Industrial buildings - 1 2 3

1447
Figure 1: Measured lifts (technology matrix). Circled figure = project no. of lifts

Key to technology matrix:

Traction lifts:
A Worm gearing with AC motor for fine adjustment (2 speeds)
B1 Gearing with AC motor, voltage control
B2 Gearing motor, with frequency converter
C Gearless drive, permanent-magnet motor with frequency converter
MR 1:1 = Lift with machine room, cabin suspension 1:1 (direct, centrical)
MR / RS = Lift with machine room, direct eccentrical suspension (“backpack”)
MR 2:1 / UF = Lift with machine room, suspension via rollers beneath cabin / 2:1 indirect
suspension (lower block)
MRL / RS = Lift without machine room, with eccentrical suspension
MRL / UF = Lift without machine room, suspension via rollers beneath cabin
Hydraulic lifts:
D/E Hydraulic valve control
Direct centrical = Central hydraulic hoist beneath cabin
Indirect RS = Hydraulic hoist beside cabin, suspension indirect via roller on hoist
Indirect UF = Hydraulic hoist beside cabin suspension, indirect via roller on hoist, suspension
2:1 via rollers beneath cabin

1448
Figure 2: Diagram showing construction of modern lifts. Source: Schindler Aufzüge AG

Measurements
To measure the energy consumption while in operation, each lift was put through a single travel cycle
(ascent and descent) while empty, and its stand-by consumption was measured. In this way the
minimum and maximum loads were recorded. Due to the extremely wide range of loads, the demands
placed on the measuring equipment were very high: it was necessary to work with mobile recording
devices that were easy to install, covered a range from a few watts in stand-by mode up to 30 kW
three-phase, including negative feed for recuperation, and guaranteed a high degree of accuracy. On
top of this, the relevant safety requirements also had to be met. The readings per travel cycle were
recorded three times per second in order to ensure that peak levels due to acceleration were included
in the measurements.
Travel up - down (empty)

4500

4000 P_1_
P_2_
3500 P_3_
P___
3000
Power [W]

2500

2000

1500

1000

500

0
0
2.93
5.87
8.8
11.7
14.7
17.6
20.5
23.5
26.4
29.3
32.3
35.2
38.1
41.1
44
46.9
49.9
52.8
55.7
58.7
61.6
64.5
67.5
70.4
73.3
76.3
79.2
82.2

time (seconds)

Figure 3: Readings for travel cycles of a traction lift (when empty) with
counterweight.
P (strong) = total load, P_1/2/3 (thin/dotted) = 3 phases.

1449
The graph in Fig. 3 depicts the typical readings for travel cycles of a traction lift (when empty) with
counterweight. The segments before and after the almost constant hoist segments show highs and
lows (due to acceleration and braking), and minor loads due to door operation are also visible.
In order to determine the energy consumption for standard usage (representing typical use) from the
highs and lows recorded during ascents and descents over the full hoist height, the calculation
method defined for SIA standard 380/4 [1] was applied. This method describes load factors and the
proportion of corresponding travel cycles by drive technology, as well as a hoist height factor, so that
the energy required for moving the lift can be calculated using the maximum hoist height, motor
output and travel speed, as well as the number of travel cycles; stand-by consumption can be
calculated on the basis of the corresponding power consumption and 8,760 hours of operation.

Stand-by and travel-cycle energy consumption

The number of travel cycles is the decisive factor for determining the energy consumption of a lift in
motion. For all lifts in the study (with the exception of 5 brand-new ones), this figure was obtained by
evaluating the travel cycle meter, and was used for calculating the energy consumption (Fig. 4). The
figures thus obtained were used in the calculation for standard usage by type of building in
accordance with SIA standard 380/4: 60,000 for residential buildings, 200,000 for office buildings. The
results can vary enormously, even for the same types of buildings.
No. of travel cycles of measured lifts by type of building/purpose

700,000

600,000
Travel cycles per
500,000 annum

400,000

300,000

200,000

100,000

0,000
Commercial building

Car parking

Car parking

Car parking

Car parking

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential / surgery
Shopping centre

Hospital (visitors’ lift)

Hospital (visitors’ lift)

Office

Office

Office

Office

Office

Office

Office

Office

Hospital (lift for beds)

Vehicle lift

Figure 4: Travel cycles of measured lifts by type of building

To compare stand-by and travel-cycle consumption (Figs. 5 and 6), the annual travel-cycle energy
consumption (EF,a) was calculated as follows, in accordance with the method defined in SIA standard
380/4:

Z F ∗ k1 ∗ k 2 ∗ hmax ∗ Pm
E F ,a =
v ∗ 3600
EF,a Energy requirements for moving cabin (travel cycles) in kWh per annum
ZF Number of travel cycles per annum
k1 Average load factor (technology factor):
rope traction 0.35 (with recuperation 0.21), hydraulic w/o counterweight 0.3
k2 Hoist height factor, average/maximum hoist height = 1 if 2-storey, otherwise 0.5
hmax Maximum hoist height, between lowest and highest stop, in metres
Pm Motor output (as a rule, nominal output as per rating plate), in kilowatts
v Speed in metres per second.
Formula “1/(speed * 3600)” indicates the travel time in hours (simplified!)

1450
 Energy consumption shown by travel cycle
kWh/a
* Cable traction with recuperation converter
25,000

Travel cycle consumption

20,000
Standby consumption * * *

15,000

10,000

5,000
*

0,000
11

16
26
33
37
40
42
49
55
62
77
100
100
100
107
116
119
157
165
196
211
279
300
300
313
330
340
353
513
671
688
Travel cycles (1,000 p.a.)

Figure 5: Stand-by and travel-cycle energy consumption of lifts by number of travel cycles;
travel-cycle consumption calculated using the above formula.
Lower travel-cycle consumption despite a higher number of travel cycles may be attributable
to nominal load, motor output and hoist height. No basic conclusions can be made.

Proportion of standby to overall energy consumption, by type of building

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
Vehicle lift

Office

Office

Office

Office

Office

Office

Office

Office
Car parking

Shopping centre

Car parking

Car parking

Car parking
Residential / surgery

Residential building

Residential building

Residential building
Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Residential building

Hospital (lift for beds)

Hospital (visitors’ lift)

Hospital (visitors’ lift)

Commercial building

Figure 6: Proportion of stand-by to overall energy consumption, by type of building, sequence

as in fig. 5, except 3 lifts where not applicable. Figures below 30% indicate systems with very
high numbers of travel cycles.

Efficiency of lift drives

By using the method for calculating drive energy consumption per travel cycle as defined in SIA 380/4
[1, 3] it was possible to measure the efficiency of the drives of the lifts included in the study on the
basis of fundamental physical data (hoist height, cabin weight, speed). For this purpose, the “load
cycle” data with maximum motor output were used. In the case of traction lifts, this concerns the
descent when empty (due to the counterweight); in the case of hydraulic lifts without counterweight or
air accumulator, this concerns the ascent with nominal load. The figures thus obtained indicate an
uncertainty factor, since the exact cabin weights were not known. The calculated average degree of
efficiency was 60 percent, which was in line with expectations. The efficiency rates for hydraulic lifts
with counterweight were similar to those for traction lifts. Measurements concerning lifts with
recuperation (which require significantly higher investments) were of particular interest. Degree of

1451
recuperation refers to the ratio of recuperated energy during ascent divided by the required energy
during ascent and descent. For the 5 lifts with recuperation, the readings varied from a disappointing
9 percent up to a satisfactory 47 percent. Here too, careful optimisation appears to be essential in
order to utilise the technical potential.

Projection and composition of consumption

The analysis of the electricity consumption per travel cycle and year for typical traction lifts resulting
from the measurement campaign (Table 2) allows for a comparison of energy requirements. But as
the measurements indicated, significant variations may occur in some systems for a variety of
reasons.
Table 2: Energy consumption of typical traction lifts (type C as per Fig. 1)
Type of Capacity Speed No. of Wh per No. of kWh p.a., % in
building/purpose stops cycle travel including stand-by
cycles p.a. stand-by mode
Small apartment 630 kg 1 m/s 6 4 40,000 950 83%
building:
Office 1,000 kg 1.5 m/s 8 13 200,000 4,350 40%
block/medium-
sized apartment
block:
Hospital, large 2,000 kg 2 m/s 12 19 700,000 17,700 25%
office block:
The projection in Table 3 shows that 58 percent of the energy consumption of lifts is attributable to
stand-by mode. This figure would be even higher if similar systems that are not covered by the Lifts
Ordinance (e.g. inclined haulage, stairlifts – which have relatively low utilisation frequencies) were
included in the projection.

Table 3: Projection of energy consumption of lifts in Switzerland (type of building/purpose

shares estimated). The total of 280 GWh p.a. is equivalent to 0.5 percent of the country’s
electricity consumption.
Type of No. of % Typical lift Projection as per SIA 380/4
building/purpose lifts
Capacity Energy % of total
Storeys

energy
Travel cycles p.a.
Hoist height

In motion

In motion
Stand-by

Stand-by
Speed
Motor

Total

Total
Stand-by

m m/s kW W
GWh

GWh

GWh

Residential
dwellings 97,500 65 6 14 40,000 1 6 90 77 16 93 28 6 34
Hospitals 1,500 1 12 30.8 700,000 2 25 500 7 19 26 2 7 9
Clinics 13,500 9 8 19.6 300,000 1.6 10 200 24 24 48 8 9 17
Shops 6,000 4 3 5.6 200,000 1.6 20 150 8 4 12 3 1 4
Offices 18,000 12 8 19.6 200,000 1.5 21 200 31 48 79 11 17 28

Car parks 6,000 4 4 8.4 60,000 1.6 18 100 5 2 7 2 1 3

Industrial buildings
(goods lifts) 7,500 5 4 8.4 40,000 0.8 30 150 10 4 14 4 1 5
Total 150,000 100 162 117 279 58 42 100

1452
Measures to reduce stand-by consumption
Stand-by consumption incorporates a broad variety of components. From the point of view of energy
efficiency there are two major factors that can cause unnecessarily high stand-by consumption:
permanently running cabin lights and door locking devices, which require constant power. With the
present-day status of technology, the stand-by consumption of a lift in an apartment block ranges
from 40 to around 100 watts, but in view of the negative factors referred to above, this figure may also
be considerably higher.
Composition of standby consumption

Control device

Floor display

Operating console at each floor

Operating console inside cabin

Cabin door lighting

Frequency converter

Composition of standby consumption

Control device

Floor display

Operating console at each floor

Operating console inside cabin

Cabin door lighting

Frequency converter

Figure 7: Shares of the different components to stand-by

consumption – example with switch-off of cabin lighting facility.
Source: U. Lindegger

Potential ways in which stand-by consumption can be reduced include:

• Switching off the frequency converter control device and other control functions when the lift is
not in motion. During off-peak periods, a stand-by mode requiring a lower power supply would
be feasible, e.g. similar to the sleep mode used in electronic devices, would be feasible.
Under certain circumstances this would lead to slightly longer waiting (wake-up) times.
• Use of more efficient power supply units (switched units, toroidal transformers)
• Use of efficient display options (e.g. LEDs)
Despite the use of automatic switching devices, lighting in lifts still contributes to overall energy
consumption to a substantial degree if inefficient solutions such as halogen filament lamps are used.
In this context, lift suppliers could take efficiency into account as a criterion in addition to design.

Developments in the area of drive technology

Hydraulics versus rope traction

Our measurements and studies have shown that hydraulic lifts, which up to now have been generally
regarded as inefficient compared to traction lifts, are no less efficient if the newest technologies are
used. Any progress is slow here because the investment costs are slightly higher, but the advantages

1453
to be gained thanks to lower requirements concerning motor capacity on the other side can help in
reduceing costs. A variety of advanced technologies are available on the market or are currently
under investigation:
• Closed instead of open loop control of valves
• Counterweight (possible with “hydraulic indirect”)
• Energy storage (instead of counterweight)

Recuperation converters

A perfect lift drive would feed exactly the same amount of energy back into the grid during ascent as it
requires for its descent (min./max. load travel cycle). In this case, the ratio of recuperated energy to
required energy would be 1:1. However, real lifts also require energy for accelerating, braking,
stopping, overcoming friction and for motor losses, and if at all this energy can hardly be recuperated.
Thus the degree of recuperation (ratio between recuperated energy fed back during minimum load
travel divided by the energy required during complete ascent and descent cycle) is unlikely to exceed
50 percent, and in the case of smaller lifts the limit is closer to 30 percent (see “Efficiency of lift drives”
section). In terms of both energy efficiency and economic viability it therefore primarily makes sense
to use recuperation converters in large lifts with a high degree of utilisation.

Optimisation of counterweight

According to data provided by the lift industry, the average occupancy rate of lifts is only 20 percent of
the nominal load, whereas the figure for counterweights is 40 to 50 percent. Optimisation in terms of
smaller loads would result in a more favourable balance with corresponding savings in energy
required for travel cycles.

New technologies

Matrix converters do not have an intermediate DC circuit and thus have the potential for reducing
losses. However, there are still a few technical problems associated with their use in lifts, and in the
next few years we will find out whether it will be possible to develop a suitable solution. Linear motors
would in fact be suitable as lift drives in view of the advantage of fewer moving parts and more precise
positioning, but a number of obstacles still have to be overcome, including strong lateral force
(friction!) and the technical complexity associated with the greater length of such motors.

Energy-conscious planning and dimensioning

Lift systems are going to be installed in buildings to an increasing extent, including in those with a low
number of storeys. As more and more lifts are put into use, additional attention should be paid to their
energy efficiency. In order to find suitable solutions from the point of view of economic viability and
energy efficiency, it is important to incorporate the following aspects during the planning stage:

For planning purposes, the required transport capacity should be specified on the basis of the
following criteria:
• Type of building (residential, office, etc.)
• Occupancy of building (no. of people per floor)
• Location of lifts (thoroughfares, location of rooms, storage areas, etc.)
• Utilisation patterns

The number and size of lifts should be determined on the basis of anticipated demand for transport
capacity. The following general rules apply with respect to the specified cabin size:
• In residential buildings with maximum 5 storeys, normally 1 lift is required with a capacity of
630 kilograms and interior cabin dimensions of 1.1 x 1.4 metres (in order to accommodate
wheelchairs).
• In apartment blocks with more than 7 storeys, at least 1 lift is required with a capacity of 1,000
kilograms and cabin dimensions of 1.1 x 2.1 metres (to allow for transport of furniture,
stretchers, etc.). For higher capacities it is important to consider whether to install a bigger
and faster lift or a second one.
• For all other buildings, detailed analyses of transport demand and patterns have to be carried
out.

1454
Accelerating and braking cause a substantial share of the travel-cycle energy consumption,
increasing with lift speed (cf. "Recuperation converters" above). The choice of lift speed should
therefore be made on the basis of the following rules while taking energy consumption and travel time
into account:

Travel time for entire hoist height:

• For residential buildings 25 to 35 seconds
• For office blocks, hotels and hospitals 20 to 32 seconds

These criteria result in the following typical theoretical speeds:

• Residential building, 4 storeys, hoist height 9 metres 0.26 to 0.36 metres per second
• Office block, 10 storeys, hoist height 27 metres 0.84 to 1.35 metres per second

Since time for acceleration and braking also has to be taken into account, slightly higher speeds are
required in practice. Thus for residential buildings with up to 6 storeys, the standard minimum speed
of 0.63 metres per second is sufficient. With respect to transport capacity, the possibility should be
considered that, under certain circumstances, choosing a slightly faster lift could eliminate the need
for a second one. In office blocks, speeds of over 1 metre per second are only truly required in
buildings with more than 8 storeys.

Choice of an energy-efficient lift system

The following construction properties have an influence on the energy efficiency of lifts:
• System architecture: Suspension (centrical, eccentrical, etc.). Centrical suspension and low-
friction guide elements reduce (friction) losses.
• Drive: Adjustable speed motors accelerate with lower losses than is the case with pole
changing motors that have been widely used in the past. By comparison with conventional
worm gear mechanisms, drive losses can be reduced by using gearless or planet gear drives.
A travel cycle with slower acceleration is more efficient but takes slightly longer.
• Control mechanism: Control mechanisms with collective operation save travel energy versus
taxi operation (i.e. without stops in between). Adjustments can be made according to time of
day. And of course, attention has to be paid to stand-by consumption!

Conclusions
There are high saving potentials in lift systems. Two major paths lead to higher overall efficiency:
Lowering stand-by consumption, that should be addressed by the lift industry and demanded by the
buyers' side . On the other hand, in an integral planning process of architect, planner, orderer and lift
supplier the system should be optimized regarding design and technology to satisfy comfort, cost and
energy requirements.
References

[1] Electricity in buildings (2006), SIA standard 380/4, Swiss Society of Engineers and Architects
(SIA), Zurich
[2] Nipkow, J., Elektrizitätsverbrauch und Einspar-Potenziale bei Aufzügen, Forschungsprojekt-
Schlussbericht, Bundesamt für Energie 2005. Download from: www.electricity-research.ch /
Electricity consumption and efficiency potentials of lifts. Research project, final report. Swiss
Federal Office of Energy, 2005
[3] M. Lenzlinger, Report on SIA standard 380/4: Energetische Vorgänge während der Fahrt eines
Aufzugs, 2005/2006, Swiss Society of Engineers and Architects (SIA), Zurich /
Energy-related processes during the travel cycle of a lift.

1455
1456
The Best Choice: Comparison of Alternatives for Residential Water
Heating in Brazil
Marcelo Caetano Simas, Fundacao Getulio Vargas, Raymundo Moniz de
Aragao Neto

IIEC – International Institute for Energy Conservation

Abstract
In Brazil, the typical residential appliance used for bath water heating is the electric shower. The
equipment is cheap (models can cost less than US 10.00), easy to install but mean a significant
electricity load in the house, and a problem to the utility: in residential areas, the peak load can be
doubled in a few minutes if all residents decide to go to bath at the same time. For larger residences,
an interesting alternative is to use an electric heater with storage tank.
In some locations, gas is available and the resident can decide what the best alternative is: electric
shower or a gas heater, with our without water storage. Although the gas heater is expensive in
Brazil, it represents reduced costs during its lifetime. A new source is expanding its penetration in
Brazil: natural gas. In cities as Rio de Janeiro and Sao Paulo, natural gas pipelines are replacing old
manufactured gases, and offering a cleaner alternative – but slightly more expensive too.
The residential consumer in those areas now has four alternatives: electric showers, solar heating,
LPG or natural gas. The paper will compare the costs and benefits of different technologies,
considering financial indicators. The operational costs will be considered, and possibly the best
alternative will depend on family size and habits.
As additional results, the paper can demonstrate to utilities companies how they can approach
consumers, showing what specificities should be highlighted in order to avoid the migration to another
energy source or attract the consumer.
The paper will consider a specific location where the three sources live together, as in Rio de Janeiro.

The water heating market in Brazil

There are few statistics of installed equipments for residential water heating in Brazil. According to
Sociedade do Sol [1], 36 million residences use electric showers for water heating. IBGE [2] presents
the evolution of total equipments sold, from 2001 to 2003.

Table 1: Number of water heaters sold in Brazil,

units per year
Year Electric Gas
showers heaters
2001 13,456,772 332,740
2002 13,546,392 29,999
2003 23,172,749 31,081
Source: IBGE – Industrial Survey

The numbers show the penetration of electric showers, due to several reasons:
1. Cost. A standard electric shower costs nearly US$ 18.00, but are cheaper models costing
less than US$ 10.00.
2. Electric showers are easy to install, requiring only an electrical connection to work (besides
water piping).
3. Electric showers can be replaced and repaired. The component to fail is the resistance that
can be replaced at low cost (less than US$ 5.00 per unit).
4. Lack of regulatory restrictions. Although some equipments have power greater than 5
kilowatts, the utility cannot avoid their installation in residences.
Electric showers are cheap only from consumer standpoint. According to Ferreira [3], in 2001 each
additional kilowatt costs US$ 663.00 – more than 100 times the cost for a single shower with 3 kW.
Equipments with storage tanks and stand-by resistances can have total power of more than 9
kilowatt, corresponding then to almost US$ 6,000 in investments for generation, transmission and
distribution.

1457
The utilization of gas for water heating until the eighties was limited to specific locations such as Sao
1
Paulo and Rio de Janeiro, where manufactured gas pipelines are available since the nineteenth
century. The expansion of oil exploitation brought together natural gas, offered in new places in the
last ten years. Consumers in other places now can decide what the best choice is.
Considering that, from a technical and safety standpoint all alternatives are similar; the financial
comparison of these available alternatives becomes the main issue. If considering only acquisition
costs, electric showers would remain incomparable; but a deeper evaluation brings new questions to
be answered.
The evaluation will consider a typical residence in the State of Rio de Janeiro, where all technologies
are well known and a network of suppliers and installers is available.

Considered costs
Energy
2
For the present evaluation, the following unitary energy costs will be considered. All include taxes
and other charges to residential consumers, according to Brazilian legislation valid in April 2006.
1. Electricity: US$ 0.278 / kWh.
2. Natural gas: US$ 1.399 / cubic meter.
3. LPG: US$ 2.646 / kilogram.

Purchase of appliances
3
The appliance costs are estimated based upon a survey with local suppliers , and are shown bellow.
1. Electric shower: US$ 20.00
2. Gas heater (both natural gas and LPG): US$ 195.00
3. Solar heater: US$ 340.00

Installation
Installation costs were estimated according to Lorenzetti [4], and are presented bellow:
1. Electric shower: US$ 9.00
2. Gas and solar heaters: US$ 68.00

Operation
For estimating operation costs, the following parameters were considered:
1. Family with 4 people, each one taking 2 baths / day.
2. Average duration: 8 minutes per bath.
3. Hot water flow: 3 litres/minute (electric shower), 2 litres/minute (gas and solar heaters mix hot
with cold water).
The following table presents additional assumptions, for each technology.

Table 2: Operational data of different water heaters

Specification Electric Gas heater Gas heater Solar heater
Shower (natural gas) (LPG)
Electric power 4400 W 3000 W
Consumption (per 4.4 kWh 0.72 m3 0.70 kg 3.0 kWh
hour of utilization)
Utilization 32
(hours/month)
Consumption (per 140.8 kWh 23.0 m3 22.4 kg 96.0 kWh
month)
Monthly cost (US$) 39.14 32.18 59.27 26.69
Sources: Electric shower and solar heater [4].
Gas heaters (150 litres model) [5].

1
Manufactured gas was initially obtained from coal.
2
For all purposes, a reference Exchange rate of 1US$ = 2.2 R$ (Brazilian currency) is considered.
3
For electric shower and gas heater, prices were obtained in www.bondfaro.com.br. For solar heater, the source was
www.shopping.clickgratis.com.br.

1458
Maintenance costs and useful life

For considered technologies, the following maintenance costs and useful life are considered.

Table 3: Maintenance costs of the 4 alternative technologies during their life-time and the
length of useful life
Item Electric Gas heater Gas heater Solar heater
Shower (natural gas) (LPG)
Maintenance actions Replace Cleaning and adjusting
resistance
Maintenance costs 5.00 35.00
(US$/yr)
Useful life (yr) 4 10

Cash flow analysis

Correspondent cash flows were created to each technology, considering all costs and:
1. Tariff readjustments of 8.0%/year, applicable to electricity, natural gas and LPG.
2. Inflation of 5% / year, applicable to appliance, installation and maintenance costs.
3. Discount rate of 15.75%, equivalent to basic interest rate in Brazil (April 2006).
4. Period of evaluation equal to 10 years.

The calculations are given below in tables 4-7, with a summary of the results and the total costs of
using different technologies in table 8.

Table 4: Electric shower cash flow analysis.

Year Appliance Installation Operation Maintenance Total cost Total cost
cost cost cost cost (nominal) (discounted)
0 20.00 9.00 469.68 5.00 503.68 503.68
1 507.25 5.25 512.50 442.77
2 547.83 5.51 553.35 413.01
3 591.66 5.79 597.45 385.25
4 24.31 10.94 638.99 6.08 680.32 378.99
5 690.11 6.38 696.50 335.21
6 745.32 6.70 752.02 312.68
7 804.95 7.04 811.98 291.68
8 29.55 13.30 869.34 7.39 919.58 285.38
9 938.89 7.76 946.65 253.80
10 1,014.00 8.14 1,022.15 236.76
Total 73.86 33.24 7,818.05 71.03 7,996.18 3,839.20

1459
Table 5: Gas heater (natural gas) cash flow analysis.
Year Appliance Installation Operation Maintenance Total cost Total cost
cost cost cost cost (nominal) (discounted)
0 195.00 68.00 386.16 35.00 684.16 684.16
1 417.05 36.75 453.80 392.05
2 450.42 38.59 489.00 364.98
3 486.45 40.52 526.97 339.80
4 525.37 42.54 567.91 316.37
5 567.40 44.67 612.07 294.57
6 612.79 46.90 659.69 274.29
7 661.81 49.25 711.06 255.42
8 714.76 51.71 766.47 237.86
9 771.94 54.30 826.23 221.52
10 317.63 110.76 833.69 57.01 1,319.10 305.54
Total 512.63 178.76 6,427.82 497.24 7,616.46 3,686.57

Table 6: Gas heater (LPG) cash flow analysis.

Year Appliance Installation Operation Maintenance Total cost Total cost
cost cost cost cost (nominal) (discounted)
0 195.00 68.00 711.24 35.00 1,009.24 1,009.24
1 768.14 36.75 804.89 695.37
2 829.59 38.59 868.18 647.99
3 895.96 40.52 936.47 603.86
4 967.63 42.54 1,010.18 562.75
5 1,045.04 44.67 1,089.71 524.45
6 1,128.65 46.90 1,175.55 488.78
7 1,218.94 49.25 1,268.19 455.55
8 1,316.46 51.71 1,368.17 424.59
9 1,421.77 54.30 1,476.07 395.75
10 317.63 110.76 1,535.51 57.01 2,020.92 468.10
Total 512.63 178.76 11,838.94 497.24 13,027.57 6,276.43
Table 7: Solar heater cash flow analysis.
Year Appliance Installation Operation Maintenance Total cost Total cost
cost cost cost cost (nominal) (discounted)
0 340.00 68.00 315.48 35.00 758.48 758.48
1 340.72 36.75 377.47 326.11
2 367.98 38.59 406.56 303.45
3 397.41 40.52 437.93 282.39
4 429.21 42.54 471.75 262.80
5 463.54 44.67 508.21 244.59
6 500.63 46.90 547.53 227.66
7 540.68 49.25 589.93 211.91
8 583.93 51.71 635.64 197.26
9 630.65 54.30 684.94 183.64
10 553.82 110.76 681.10 57.01 1,402.70 324.90
Total 893.82 178.76 5,251.32 497.24 6,821.14 3,323.19

1460
Table 8: Summary
Total
Appliance Installation Operation Maintenance Total cost
Technology cost
cost cost cost cost (discounted)
(nominal)
Electric
shower 73.86 33.24 7,818.05 71.03 7,996.18 3,839.20
Gas heater
(natural gas) 512.63 178.76 6,427.82 497.24 7,616.46 3,686.57
Gas heater
(LPG) 512.63 178.76 11,838.94 497.24 13,027.57 6,276.43
Solar heater
893.82 178.76 5,251.32 497.24 6,821.14 3,323.19

The results for discounted cash flow show that the solar heater is the best choice, from a financial
perspective. It’s necessary to reinforce that results are obtained for a 10 years period, and the
difference for electric shower results are not so relevant (close to 15%). Another constraint is that
solar heating can be used specially for houses; when a building is considered, even with the
possibility of using a roof area for installing solar collectors, the installation costs are increased and
the numbers would change significantly.

Conclusions
Several technologies can be used for residential water heating in Brazil. The calculations presented in
this paper show that a solar heating system is the best alternative, when technically possible and for a
10 years period of analysis.
Utilities and other stakeholders could consider the assumptions presented here to stimulate the
market to adopt more efficient technologies. The use of natural gas for heating results in competitive
costs, but to demonstrate them is necessary. If the client is inclined to consider only short-term costs,
the utilization of electric showers will remain dominant.
Energy planners should take these results as reference, in order to create mechanisms to avoid the
utilization of electric showers in Brazil. If a small amount migrates from grid and generation expansion
to reduce initial costs for natural gas heaters, for example, consumers would be stimulated to adopt
this technology, reducing peak demand significantly and postponing investments only for meeting the
needs of residential water heating.

References
[1] www.sociedadedosol.org.br
[2] IBGE. Pesquisa Industrial (2001, 2002, 2003).URL: www.ibge.gov.br.
[3] Ferreira, O. O sistema eletrico brasileiro. ECEN – Economia e Energia, vol 32, 2001.
[4] www.lorenzetti.com.br

1461
1462
Barriers and Drivers to Energy Efficiency –
A New Taxonomical Approach
B. Sudhakara Reddy

Indira Gandhi Institute of Development Research

Abstract
This paper develops a new systematic classification and explanation of barriers and drivers to energy
efficiency. Using an ‘actor oriented approach’, the paper tries to identify (i) the drivers and barriers
that affect the success or failure of energy efficiency investments and (ii) the institutions that are
responsible for the emergence of these barriers and drivers. This taxonomy aims to synthesise ideas
from three broad perspectives, viz., micro (project), meso (organization), and macro (state, market,
civil society). The paper develops a systematic framework by looking at the issues from the
1
perspective of different actors. This not only aids the understanding of barriers and drivers; it also
provides scope for appropriate policy interventions. This focus will help policy-makers evaluate to
what extent future interventions may be warranted and how one can judge the success of particular
interventions.

Introduction
Economic development has traditionally been linked to energy adequacy. Countries pursuing
economic growth are expected to resort to increasing levels of energy use. Achieving these levels of
energy production and utilisation through present technologies is not only difficult and expensive, but
also environmentally unsustainable. Various studies indicate that increased energy efficiency can
bridge the gap between growing demand and reduced energy supply without adversely affecting the
quality of service (Golove and Eto 1996; S. Reddy 2003) However, as the past experience has
shown, this may not happen, unless the issues that hinder the penetration of efficient technologies
are addressed (A.K.N. Reddy 1991; Hollander and Schneider 1996; Sorrell 2000). There is a gap
2
between the theoretical opportunities for cost-effective energy efficiency investments and the levels
that can be achieved practically. The origins of the gap seem to lie in the set of barriers which may be
divided into categories such as financial, legal, organizational, or informational. These barriers
prevent investments in energy efficient technologies. It is also certain that there are drivers that help
increase investments. The barriers hinder the penetration of energy efficient technologies, even
though these technologies have been shown to be economically cost-effective. If policies to
encourage investments in improved energy efficiency are to be successful, understanding the nature
of these barriers and drivers is essential. These policies must succeed in the context of liberalising
energy markets, falling energy prices, and increasing the development of a broad-based energy
service industry.
Lighting is an important household energy service. This is because lighting usually involves the use of
commercial energy and often not many alternatives exist. Nearly 0.3 billion people in India — more
than the world’s population in Edison’s time — still have no access to electricity. The majority of
people who lack direct access are mostly from rural areas. Even those households which us electric
lighting, the level of illumination are far lower than with modern electric lighting. The result is a
substantial amount of primary energy use with little service received in return. The aim of the present
paper is thus to examine the nature of barriers and drivers for efficient electric lighting. It also
analyses the circumstances in which they arise, their relative importance in different contexts, and the
manner in which different actors intervene to overcome these barriers. The paper reviews current
perspectives on barriers and drivers, classifies them according to their influencing patterns, and

1 The actors include: the consumer/investors, utilities, government agencies (ministries, state agencies, parliamentary
commissions, and intergovernmental commissions), financial institutions, regulatory bodies, local authorities, research and
development organisations, equipment manufacturers, market institutions, energy consultants, NGOs, energy service
companies, the International organizations (e.g., Intergovernmental Panel on Climate Change), etc.
2
Economic cost-effectiveness depends on both: where the actor-concern boundary is placed (the individual or the society as
whole), the time perspective (static or long term dynamics included) and the system regarded (ceteris paribus – all other things
equal - all things concerned are changed). Here, the perspective is individual-static-ceteris paribus.

1463
provides supporting evidence for their prevalence. Finally, the paper tries to evaluate the
effectiveness of different institutions for improving energy efficiency. The debate on barriers and
drivers is contentious and is characterised by disagreement over basic theoretical and conceptual
principles. Hence, the primary objective of this work is to develop a new systematic theoretical
framework.

The energy efficiency debate

There has been a long running debate over the issue of energy efficiency (EE) between energy
economists and energy analysts. One issue concerns the rebound effect. Although definitions vary, this
effect describes the following linkage: the efficient use of energy leads to an increase in the use of
energy. This may partly offset the savings in energy use achieved by the EE improvement. The
rebound effect is rooted in neoclassical economic theory. The extent of the rebound effect depends on
the price elasticity of demand. Therefore, the assumption of rational decision making is the precondition
for an explanation of this effect.
Henry Saunders (1992), argues that ‘energy efficiency gains can increase energy consumption by two
means: by making energy appear effectively cheaper than other inputs; and by increasing economic
growth, which pulls up energy use.’ The debate grew more intense in the 1990s, spurred by global
warming concerns. The argument for EE, however, is independent of environmental concerns. The
market failure that distorts energy use is under pricing of energy by regulators and deregulation is
preferable to EE into which utilities were forced to reduce energy prices. There are also other market
failures, such as split incentives (landlord-tenant), etc (Reddy, 1991). The reality is that while EE is
technologically feasible, it will require a significant change in our collective approach to making it work.
What this means is that, while technologies are already available, the problem lies in their application.
Another aspect to which contributions to the EE debate can be associated is what one describe as a
matter of governmental intervention. Gunn (1997) who investigated the paradigms of EE stated that it is
important to recognize that the primary debate is over the optimal level of governmental intervention in
energy markets rather than over the optimal level of EE. There are many forms of government
intervention such as Subsidies and taxes; Special purpose loans; Facilitation (information systems;
well-structured markets; approved suppliers); Guarantees for specific risks, or offering insurance; and
Arranging objective non-partisan product information (e.g energy labelling)
The justification and degree of governmental intervention is a matter of debate in the international
literature. Barriers which are attributed to market failures make governmental intervention necessary
and justifiable and a large body of international literature relates barriers to market failures. However,
others claim that only few market failures can be defined as such. Haugland, Bergesen, and Roland
(2000) argue that most barriers merely reflect ‘unaccounted (transaction) costs or simply result from the
consumer’s liberty to choose freely his/her convenience and service levels and willingness to accept a
higher energy bill for their personal taste or lifestyle.’ Therefore, governmental intervention might be
questionable.
A large body of international literature puts effort in an empirical approach to this question with different
objects of investigations and findings (Helm 2002, Sloman and Sutcliffe 2000, Vine et al 2003,
Banerjee and Solomon 2003) Governmental stimulation of the implementation of new technology by
promoting associated research and development was found to be counterproductive. Although it leads
to technological progress it may hinder corporate investments in new technology . Firms may favour to
wait for the next generation of technological developments. In the case of a restructuring electricity
market pro-interventionists ask whether the market alone is able to overcome EE barriers. It is argued
that governmental support in promoting EE and load management can be advantageous. Further, an
evaluation of US energy labelling programs led Banerjee and Solomon (2003) to conclude that,
‘government support is the most critical factor for the success of a labelling program.’ Altogether, the
challenge of reconciling government and free market contributions with regard to the energy market and
EE remains.
With regard to EE potential, a distinction has to be made between: (i) the economc potential: achievable
by removing market failures; ii) the technological potential: achievable by the additional removal of ‘non-
3
market market barriers ’; and (iii) the hypothetical potential: achievable through the additional
4
elimination of market failures in fuel and electricity markets (Jaffe-Stavins (1994) . This framework is
summarised in Figure 1.
3
Technical barrier is the one where the new technology might be found wanting or become rapidly outdated.
4
Market failure is a distinct notion for such problems that can not be solved (in full) by use of market instruments, such as public
goods, externalities, etc. Such failures cannot be eliminated though there are some “Pareto-sanctioned” measures that redeem

1464
Figure 1: Characterization of energy efficiency potential

On the right hand side of Figure 1, various potentials for EE are represented. The market potential is
the efficiency improvement that can be expected to be realized for a projected year under a given set
of conditions (e.g., energy prices, consumer preferences and energy policies). The market potential
reflects barriers and market imperfections that keep efficiency potential from being fully realized.
The economic potential is the energy saving that would result if during each year of the period in
question, all replacements, retrofits and new investments were shifted to the most energy-efficient
technologies that are still cost-effective at given energy market prices. The economic potential implies
a well-functioning market, with competition between investments in energy supply and demand. It also
assumes that the barriers to such competition have been corrected by energy policies. It is assumed
that as a result of such policies, all users have easy access to reliable information about the cost-
effectiveness and technical performance of existing and emerging options for energy efficiency.
The technical potential represents achievable energy savings under theoretical considerations of
thermodynamics, where final energy consumption is kept constant, and energy losses can be
minimized through process substitution, heat and material re-use, and avoiding heat loss. This can be
considered as hypothetical potential and represents achievable energy savings that result from
implementing the most energy-efficient technology available at a given time, regardless of cost
considerations and reinvestment cycle.
On the right hand side we examine the social optimality. The narrow social optimum in the market for
energy efficient technologies represents the rate of energy efficiency uptake that would be observed if
all barriers that were deemed to be irrational on a cost-effective basis were eliminated, i.e., if people
adopted all measures that could leave them economically better off given the current pricing

them. The market failure here means “market imperfections” where market pricing, ownership etc. can be applied successfully
(at least in part).

1465
environment. In this situation, one has to get energy prices right implying that narrow social optimum
should include cost effective removal of market failures for energy. The true social optimum would
include additional efficiency diffusion that would likely to be seen by considering environmental
externalities.

Actors in the field of energy efficiency

There is a growing recognition that the usual investment decisions such as payback period, rate of
return on investment, net present value, etc, do not account adequately for actual business behaviour.
Technical solutions above are not likely to succeed unless there is an interface between various
actors in the field of EE. In conventional energy programmes, there are only two main participants,
that is, the suppliers and the consumers. In the EE system, there are many actors such as equipment
manufacturer, financial institutions, NGOs, etc, who play important roles. If the EE programmes are to
be widely accepted, all these parties should work together. Each actor has relations to other actors.
The actor experiences constraints and stimuli; has abilities and weaknesses; and holds rights,
responsibilities and obligations. All these lead to series of decisions. All the actors, and their
decisions, describe the whole socio-technical structure and the processes that occur.
While discussing the drivers and barriers, it is important to consider the role of actors such as the
national government, regional and local authorities, supranational bodies such as international
development agencies, the United Nations and its specialised bodies, the World Bank Group,
international and national professional and trade bodies, etc. One could try to indicate which ‘actor’
has the power to create/reduce/remove certain barriers and on which actors the particular barrier has
an influence. One could also try to indicate whether the barrier can be modified in the short term (eg
through a subsidy), in the medium term (eg through new legislation), or only in the long term (eg
through improving general education), or probably never (eg in the case of cultural/religious barriers).
One could also look at the mechanisms as means to overcome the usual constraints and pave the
way for smooth functioning of programmes. This, in turn, encourages the removal of barriers and
positively affects investments. Such an ‘actor-oriented’ approach would give clearer insights into
barrier analysis. Through such an approach we can find out the role of actors. For example, if any
barrier is named, we can see who created it, and who, therefore, is (in principle) able to remove it.
The actors can be divided into the three sub-groups – micro, meso and macro.

Figure 2 illustrates actors in the field of EE in general, from ‘macro’ and ‘micro’ perspectives

Figure 2: Actors in energy efficiency

At the micro level: To design better programmes, they are the persons/consumers to address
(information, training, support by specialists, etc).
At the meso level: Organisations such as utilities, energy development agencies, service companies,
etc. Through new incentives, organizational reform, and other changes, barriers can be reduced or
removed.

At the macro level: Relate to the ‘higher-level’ institutions (state/market/civil society) that determine
the setting under which the lower levels have to operate.

1466
The actions needed to address barriers are different for each. Through this approach, we not only
need to look at the barriers themselves, but also at the institutions/situations that create the barriers.
Each actor would then have two roles: (i) to carry out a project at his own level, the actor has to work
within existing ‘external constraints’ given by ‘higher-level’ institutions, and (ii) establish conditions
(either barriers or drivers) for other actors at a ‘lower level’.
Each actor in the field of EE has a specific function, but even without being an active performer,
everybody can be positively influenced by EE thanks to its character of positive externality.
Everybody has a responsibility: governments, industries, business associations, donors, and
international institutions; they all should recognize their respective roles. The EE process can be
completed only by cooperation, by a common achievement. If the government wants to promote EE,
it needs to see how the whole system can be modified which can lead to more EE-directed decisions.
Table 1 demonstrates main functions of the actors in the field of EE. Each of them has a specific
role(s) in the process of the implementation of EE. They are mutually connected:
Table 1: Functions of the actors in energy efficiency lighting systems
Actors Functions
Establish legal and institutional frameworks;
Governments Integrate EE in decision-making in all sectors,
Support administrative efforts to enhance EE;

Collect and propagate information about activities, experiences,

programs and projects;
EE agencies Develop and implement EE programmes;
Put programmes together to ensure a larger market response.
Disseminate the achieved results;
Business and industry Appeal to members to apply EE;
enterprises Negotiate with involved partners in terms of achieving EE
targets, eg with government, foreign organizations, etc.
Improve energy services – supply, transmission and distribution;
Energy Supply agencies Propagate EE lighting technologies;
(Utilities) Provide incentives to those who use EETs and disincentives to
those who do not .
Provide a whole range of lighting devices to the consumers;
Cooperate on developing and later promoting EE lighting
Equipment
technologies;
manufacturers
Make consumers understand they should not consider only
purchasing but life-cycle costs.
Help to finance EE programmes;
Financial Institutions Disseminate information about EETs;
Cooperate with other agencies to implement common
programmes;
Publicise good examples;
Non-governmental Network to make use of the latest experiences in research both
organizations in technology and applications.
Develop supporting instruments for monitoring and evaluation;
Support mutual interest by adapting routines and instruments;
International
organizations Serve as a forum to disseminate results;
Act as a clearing-house to establish collaborative actions.

Source: based on OECD, 1999

Barriers to energy efficiency

Micro barriers
These can be referred to as the obstacles that are unique to a particular project. Here the barriers
could be in terms of project design. A poorly designed project can make insufficient use of synergies

1467
or drivers, or take too little account of barriers. The person to address is the project designer
(information, training, support by specialists, etc). Examples include:
A lighting programme, which only focuses on upgrading the lighting unit, is likely to be less profitable
than the one which, at the same time, upgrades leakage in distribution systems, creates incentives
for energy savings, such as metering in households, and gradually raises lighting prices to recover
costs.
A medium size or large project, consisting of households from a group of villages or a town, is usually
more profitable than dispersed and one-off small projects due to lower transaction costs and
economies of scale.
A project that consults the representatives of benefited target groups (e.g., consumers, equipment
manufacturers, financial institutions) is usually more feasible than the one that is imposed from
above.
By changing the features of a project – for example, by modifying incentives for energy savings,
replacing the technology, increasing the project size, or creating legitimacy through consultation – the
financial viability and feasibility could be improved. Also, changes in project design can reduce the
internal barriers to profitability and feasibility.

Meso barriers
These relate to the organizations affiliated with the project. These barriers can be common to a wide
variety of projects and can be tackled with efficient organizational design, human resource, as well as
time management. Examples include:
The implementing agency may be understaffed, bureaucratic, or lack proper incentives for promoting
energy efficient lighting systems;
The project target groups (eg rural households) may be small, inexperienced and
undercapitalised;
The consumers may lack experience in a particular lighting technology
The implementation authority may be unaware about the details of a geographic area where the
project has to be implemented;
The government authorities may put forth rules and procedures that can raise the cost of the
project and/or reduce the feasibility of implementation.

Macro barriers
The macro barriers can be divided into three categories: state, market, and civil society related. Since
these barriers are not project or organization-specific, they cannot be altered by changing project or
organizational design. For project sponsors and financiers, macro barriers are externally driven and
are difficult to influence (unless they have the power of influencing policies, market, or culture). In
some cases, projects include policy components, which can affect macro variables (eg electricity
tariffs, laws about who will keep financial savings from energy efficiency projects, subsidies, etc). It is
usually easier for project sponsors and investors to change the project characteristics than it is to
influence government policies such as electricity tariffs and subsidies. Therefore, many projects do
not even attempt to change macro variables and instead focus on overcoming or neutralizing the
adverse effects of macro barriers through increased financial subsidies – or, more rarely, through
innovative project and organizational design. The benefits of tackling macro barriers are usually much
greater than focusing merely on micro and meso level barriers, and the sustainability of results over
time is much greater as well.

Barriers relating to the state are those that can be traced to the behaviour (action or inaction) of
governments or state-run organizations (e.g public utilities). Barriers relating to the market are those
that can be traced to the behaviour of individuals, private firms and financial institutions, which reflect
the prevailing market structure. And finally, barriers relating to civil society can be traced to the
behaviour of NGOs, academic institutions and other civil society organizations (CSOs). While the
distinction between state, market and civil society barriers is useful as a means of classification, in
practice, there are linkages between them. For example, markets react to policy changes and vice
versa; policy is affected by the lobbying of firms; and NGOs and other civil society organizations
operate within a political and economic context. Efforts to remove or reduce macro barriers need to
pay attention to these relationships in order to be effective.
The barriers can be further classified into internal (barriers due to flaws in the project or the
organization) and external (policy, market, and civil society barriers). Internal barriers are easier to
overcome, because they require only changes in the project or the organizations involved in the
project whereas external barriers require policy changes, measures to affect the workings of markets,

1468
or measures to influence civil society or the culture of a country. In designing a complete model, there
are numerous variables that could potentially be relevant: there are variables in the three categories
of project, organization, and policy design, and there are variables in the causal pathway (the
categories of drivers and barriers).

Removing, reducing, and avoiding the barriers

It is important to explore the relationship between the independent variables (the stimuli mechanisms)
and the linking variables (the drivers and barriers). This relationship can be described as a process
of stimulating the drivers and overcoming the barriers to private investment. Overcoming the barriers
would reduce the overall financing need for energy efficiency and promote sustainable development.
It is important to recognize that the process of 'overcoming' barriers is not a single process, but, in
fact, three separate processes.

Type 1: Removing a barrier (or risk) means getting rid of a barrier altogether, so that all present and
future projects no longer face that barrier. For example, if a government repeals a law that obstructs
energy efficiency, the change of legislation will affect all present and future projects. Unless the law is
reintroduced later, the barrier is removed altogether. In terms of policy objectives, this is the most
desirable form of overcoming barriers.

Type 2: Reducing a barrier (or risk) means that the barrier remains in place, but that its deterrent
effect is diminished. For example, if a government increases electricity tariffs but not sufficiently to
cover long-run marginal costs, the barrier of distorted electricity pricing is reduced but not removed.

Type 3: Avoiding a barrier (or risk) means that the barrier can be overcome or avoided during a
particular programme, while remaining in place for others.

While Type 1 and Type 2 are actions addressed primarily to the actor who has created the barrier,
Type 3, is addressed primarily to the actor who wants to ‘work around’ the barrier (a different
person/institution).

Of all the approaches to overcome barriers, removing obstacles (Type 1) may be the most expensive
and difficult. Yet, it is probably the most cost-effective, because in this way, barriers disappear for all
projects in an economy – present as well as future. In practice, however, the most common approach
seems to be the least effective, namely, the avoiding of barriers (Type 3). It seems that many
agencies promoting energy efficiency merely ‘lift’ projects over the same hurdles time and again.
From a public policy perspective, this is not a desirable approach, as it represents a waste of public
funds.
A case in point is the Global Environment Facility (GEF). The Operational Strategy of GEF states that
the removal of barriers to energy efficiency and renewable energy are central to the mission of the
organization (GEF 1996, 33). The GEF supports projects (and ‘programmes’) that tackle institutional
and structural shortcomings, and in this way modify the barriers and drivers. Two of the ten
operational programmes at GEF are specifically aimed at removing barriers to energy efficiency
5
(Operational Programme #5 ) and renewable energy (Operational Programme #6) (GEF, 1996).
However, the incremental cost principle, based on which GEF distributes its funds, contradicts this
objective because it usually results merely in ‘lifting’ the projects over barriers, rather than reducing or
removing the barriers that create the incremental cost problem in the first place. As a result, the
barriers remain in place, ‘waiting’ to obstruct the next project.
To illustrate this important point, consider an energy efficiency lighting programme, which addresses
two main barriers: high initial cost of efficient light bulbs and lack of consumer awareness. Given
enough financial resources and successful project implementation, both barriers can be overcome.
This, however, does not guarantee that those barriers will permanently disappear. It may happen that
the consumers, who got used to highly subsidized prices, make it difficult for the manufacturer to sell
the bulbs at commercial prices once the project is over. This is because the consumers tend to forget
the benefits of efficient light bulbs after a while, or because they are simply unwilling to accept
increased prices. In order to prevent this, incentives must be given to the project developers to
ensure the durability of the barrier removal measures so that the impact stretches beyond the

5
The United Nations Framework Convention on Climate Change (UNFCCC) seeks to stabilize atmospheric greenhouse gas
concentrations at levels that would prevent dangerous anthropogenic interference with global climate. The Operational Strategy
of the GEF puts initial emphasis, among others, on the removal of barriers to energy conservation and energy efficiency.

1469
individual programme. The sustainability of barrier removal should be one of the main criteria by
which development agencies allocate funding to programmes. This should be made mandatory in all
the business plans submitted as part of applications for energy efficiency funding.
While barrier removal (Type 1) is quite rare, many projects achieve barrier reduction (Type 2). It can
be hypothesized that all successful energy efficiency projects contribute to the reduction of costs or
other barriers, thus accelerating the process of commercialisation. Types 1 and 2 refer to possible
actions by an authority, which is responsible for the barrier/driver, or has the power to modify it. The
primary target for barrier removal and reduction should be the government, but other institutions may
also be able to influence the process in direct and indirect ways. For example, rather than lobbying a
reluctant government, it may in some cases be more effective to work with private sector or civil
society organizations, if they have an influence reducing or removing barriers.
Type 3 refers to the ways in which the implementing agency and the programme manager can ‘get
round’ the problem. Although this is usually the least cost option in the short term, avoiding a barrier
is a short cut that should not be taken, as it does not improve the process of commercializing energy
efficiency. Programme developers and financiers are likely to choose the cheapest and easiest
method of overcoming the barriers – which is usually to avoid them, leaving the task of reducing and
removing barriers for other agencies. However, even for private programme developers and
financiers, this makes only short-term sense, as they would themselves benefit from barrier removal
in the long term.

In the light of this analysis, one can arrive at two approaches to overcome the barriers to private
investment. The first one has the primary objective of removing, or at least reducing barriers, which
can be referred to as a targeted barrier removal effort. The second approach has the primary
objective of maximizing the project’s profit. Examples for targeted barrier removal efforts include:
Policy initiatives to remove direct and indirect subsidies for fossil fuels;
Initiatives to provide energy efficiency information (eg through mass media, Internet) to create
awareness;
Initiatives to train programme developers, financiers and government officials, and to provide them
with the means and incentives to change the structure of barriers and drivers.
If successfully implemented, these activities are likely to yield high economic and environmental
benefits per unit of consumer expenditure on energy.

Analysis of drivers
Along with the barriers one should understand the motivation and forces that lead consumers to
adopt energy-efficient measures. Information directed towards understanding consumer’s decision-
making behaviour and preferences as well as the behaviour of other stakeholders would give a better
understanding of the drivers that push energy efficiency measures. A few examples of drivers are
given below:
Awareness: It is clear from the above discussion that there are a wide variety of players that can
contribute to barrier removal measures, who can also stimulate the drivers and thereby help the
penetration of energy efficient technologies. A case in point is the strong competition between
technology manufacturers that results in aggressive advertising campaigns. The advertising
campaign in this example is the measure (stimulant), and the high level of awareness of energy
efficient technologies, thus created, is the driver.
Decrease in technology price levels: A high level of awareness is usually not sufficient to attract
private investment and guarantee market success. The general understanding of market
mechanisms dictates that price of a technology is an important factor in its speedy penetration.
Hence, one can assume that educational/promotional activities are important, even though, there
should be other considerations as well. Along with advertisement campaigns, the competition should
lead to a decrease in the cost of the technology. Such reductions in prices can safely be assumed to
lead to an increase in the sales of the technology.
Increase in energy prices: Cost savings in energy bills through reduced use of energy is one of the
reasons for the decision to buy energy efficient equipment. A look at the electricity prices in
developing countries over the past few years indicates that nominal electricity prices increased
manifold during the 1990s. In real terms, the price increase may be less but is still significant to affect
purchasing and investment decisions. Increased energy prices place a higher burden on consumers.
If there is a continuous and predictable increase in its price, consumers are more likely to be
motivated to adopt energy efficient equipment to conserve electricity and heat.

1470
Technology appeal: While analysing drivers, one factor that may be of worth considering is the
‘smartness’ of the technology. If the energy-efficient equipment gives an impression that it looks
‘modern’, ‘appealing’, and ‘fashionable’, there is a higher probability of consumers purchasing the
technology. These non-economic motivations, in general, dominate the decisions primarily of high-
income groups, for whom, technological appeal is a major driving factor.
Non-energy benefits: Non-energy benefits are important drivers of energy efficiency. They accrue
at the national level, e.g. via improved competitiveness, energy security, job creation, From a
consumer perspective, it is often the non-energy benefits that motivate decisions to adopt energy
efficient measures. The benefits to the consumer through these measures include (i) improved indoor
environment, comfort, health, safety, and productivity; (ii) reduced noise; (iii) labour and time savings;
(iv) improved process control; (v) increased reliability, amenity or convenience; and (vi) direct and
indirect economic benefits from downsizing or elimination of equipment.
Environmental Regulations: Environmental regulations, if properly designed, can serve as a driver
for investments in energy efficiency. In the absence of environmental regulations, the societal costs
of electricity generation in the form of air emissions, water use and other environmental impacts are
not borne by the energy producer or by the consumer. Consequently, these actors do not see the
true societal costs of their production and consumption decisions. Environmental regulations can
force producers and consumers to internalize these environmental costs into the price of their energy
goods and services in the form of increased environmental compliance costs. These increased
environmental costs can send a price signal for increased investments in energy efficiency by making
efficiency investments comparatively more attractive financially. Not all environmental regulations are
created equally, however. If an environmental regulation simply mandates that an industry install a
particular pollution control device, then the industry’s response will be to seek ways to minimize its
cost of compliance with the pollution control requirement and the price signal for efficiency
investments will be muted. Once the pollution control device is installed, there will be little incentive
to improve the efficiency of the overall production process. If, on the other hand, the environmental
regulation uses market mechanisms to reward industry for reducing emissions through, for example,
tradable permits, then the industry would have the incentive to improve the efficiency of and
continuously improve its manufacturing process and potentially turn the environmental regulations
into a source of profitability. A more efficient manufacturing process would naturally follow.

Conclusions
This paper attempts to study the barriers and drivers that influence investments in energy efficiency
using an actor-oriented approach. It starts with the development of a new taxonomy of
barriers/drivers classifying them in terms of profitability and feasibility of private investments in energy
efficiency. The barriers are classified into three broad categories, viz., micro, meso and macro. In
practice, these barriers are of the following types: perceptual-behavioural, financial-economic,
institutional-structural and market oriented. Such classification is expected to help devise the
response measures to remove, reduce, or avoid the barriers. The paper is also aimed at
understanding which drivers contribute to the successful diffusion of energy efficiency measures. This
would facilitate development of appropriate support mechanisms at financial, policy, institutional,
regulation, and information levels.
Further, using this taxonomy, the paper develops a theoretical framework which proposes a
methodology to analyze the causal relationship between barriers/drivers and the appropriate
response measures. This work brings out clearly the need for a different set of response measures,
depending on which group a barrier belongs to. At the policy level, some barriers can hardly be
influenced by an energy efficiency project team, and whoever encounters them has to accept them.
But if the ‘project’ is of a wider scope, let us say, a programme of institutional development financed
by international donors, that programme may be able to modify some of the barriers. Hence, it is
important to try to assess which barriers are more or less ‘unchangeable’, and which may be worth
tackling by such programmes. This would greatly help both the multilateral and government agencies
in devising their strategies in terms of support to future barrier removal programs. This analysis has
profound implications for barrier taxonomy, which, in turn, helps design energy efficiency projects.
The paper underlines the significance of the identification and classification of real barriers, which is a
precondition for the successful diffusion of energy efficient technologies.

Acknowledgement
The author would like to thank the anonymous referees who had provided excellent reviews on
an earlier draft of this paper.

1471
References
Banerjee, Abhijit, and Barry D. Solomon, 2003. ‘Eco-labeling for energy efficiency and sustainability: a
meta-evaluation of US programs’. Energy Policy, vol. 31, no. 2: 109-123.
Edward Vine, Drury Crawley, and Paul Centolella (Eds.). 1991, Energy Efficiency and the
Environment: Forging the Link.. American Council for an Energy-Efficient Economy,
Washington, D.C., 1991.
Ferguson, Eric F., 2002, personal communication (e-mail).
GEF (Global Environment Facility), 1996. ‘Operational Strategy of the Global Environment Facility’.
Washington, DC: Global Environment Facility.
Gunn, C 1997. Energy efficiency vs economic efficiency? New Zealand electric sector reform in the
context of national energy policy objectives. Energy policy. 25(2): 241-257.
Guy, Simon, and Elizabeth Shove, 2000. ‘A Sociology of Energy, Buildings and the Environment.
Constructing knowledge, designing practice’. London, UK .
Haugland, T, Bergesen, HO, and Roland, K 2000. ‘Energy structures and environmental futures in
Europe’. Oxford University Press, Oxford.
Helm, Dieter, 2002. ‘Energy policy: security of supply, sustainability and competition’. Energy Policy,
vol. 30, no. 3: 174-184.
Hollander, J. M. and Schneider, T. R., 1996. Energy-Efficiency: Issues for the Decade. Energy vol. 21,
no. 4: 273-87.
OECD, 1997, The role of technology in closing the efficiency gap: A global challenge, A Hagler Bailly
study, USA.
Preston, John, and Theodosios Palaskas, 28 September 2000, oral communication (thesis
defense).
Reddy, Amulya K. N., 1991. ‘Barriers to Improvements in Energy Efficiency’. Energy Policy, vol.
19, no. 10: 953-61.
Reddy, B.Sudhakara, 2003, Overcoming the energy efficiency gap in India's residential sector, Energy
Policy, Volume 31, No 11.
Saunders, Harry, 1992, The Khazzoom-Brookes Postulate and Neoclassical Growth, Energy Journal,
Vol.13, No4, pp131-148.
Sanders, Maria, 1997. Green Base Conversion Project (415) 968-8798. Center For Economic
Conversion, Mountain View, CA 94041-1344.
Sloman, John, and Mark Sutcliffe, 2000. ‘When markets fail’. In: Vivek Suneja (ed), Understanding
Business: Markets. London, UK and New York, NY: Routledge in Association with The Open
University, 147-166.
Sorrell S, Scleich J, Scott S, O'Malley E, Trace F, Boede U, Ostertag K, Radgen P, 2000,
Reducing barriers to energy efficiency in private and public organisations, Report to the
European Commission, in the framework of the Non-Nuclear Energy Programme,
JOULE III. Brighton.
Stavins Jaffe, 1994, The energy efficiency gap, Energy Policy 22, 804-810
Thorleif Haugland, Helge Ole Bergesen, Kjell Roland: Energy supply structures and environmental
futures in Europe. Oxford University Press, 2000.
USAID, 1991, Opportunities for improving end-use electricity efficiency in India, Electricity
Efficiency in India, A Report of the Office of Energy, Washington D.C.
Vine, Edward, Jan Hamrin, Nick Eyre, David Crossely, Michelle Maloney and Greg Watt, 2003. ‘Public
policy analysis of energy efficiency and load management in changing electricity businesses’.
Energy Policy, vol. 31, no.5: 405-430.
Weber, Lukas, 1997. ‘Some Reflections on Barriers to the Efficient Use of Energy’. Energy
Policy, vol. 25, no. 10: 833-35.

1472
Load Demand Pricing - Case Studies in Residential Buildings
Jurek Pyrko

Department of Energy Sciences, Lund University

Abstract
Since the liberalisation of the Swedish electricity market in 1996, the competition between utilities has
increased, and the generation capacity has gradually been adjusted to suit the demand.
Consequently, the earlier excessive electricity production capacity has been reduced. However, if the
gap between the generation capacity and demand will be too narrow, this may result in notable power
shortages in the electricity market. In order to achieve lower load demand, to avoid load peaks and to
reduce electricity cost, a Swedish electrical utility - Skånska Energi Nät AB (SENAB), is planning to
include a load demand component in its electricity tariff to make customers more aware of their
energy consumption pattern and (possible) load demand problems. This study investigates the impact
of the new tariff from the viewpoint of the utility as well as its customers, compared to the existing
tariff. The project was carried out by the Efficient Energy Use in Buildings Research Group at the
Department of Energy Sciences, Lund University.
The results of the investigation show that if a load demand component were to be introduced into
SENAB’s network tariff, primarily customers with a 16-ampere fuse would incur higher network
charges whereas customers with a higher fuse level would incur lower charges. With the existing
network tariff, customers with high fuse levels pay relatively high standing charges in relation to their
exploitation of the grid and as such they are subsidising customers with lower fuse levels. The study
also shows that it is important that the new load demand pricing strategy (tariff) is communicated to
customers in a comprehensive manner, so that they understand it and furthermore realise that they
can save money by changing their energy consumption patterns without lowering their standard of
living or comfort.

Introduction
Sweden has a relatively high electricity consumption per-capita, about 17 000 kWh per inhabitant
annually, more than twice as high as the European Union average. In the year 2005, Sweden was in
fourth place in the world, in terms of electricity consumption, after Norway, Iceland and Canada. The
high electricity consumption in Sweden is due to electricity-intensive industries and the high demand
for space heating caused by the cold climate. Over the past thirty years, the electricity consumption in
Sweden has increased at the rate of almost 3 % annually [1].
The Swedish electricity market was reformed in 1996 and then again in 1999 for household users. As
a result of the electricity market reforms, consumers may now choose their electricity supplier and all
trading must be competitive. However, the grid operator can not be chosen by the consumer, and is
still regulated. A corporation that pursues network operations may not pursue trading in or generation
of electricity. Therefore, there must be a clear distinction between generation of and trading in
electricity and network operations.
Electricity consumption varies between different hours of the day, between days of the week and
between seasons of the year. The highest power demand occurs only during a few hours when the
outdoor temperature drops. In recent years, the power demand has reached new peak levels but due
to predominantly economic and political reasons the load reserves have dwindled. The reliability of
supply criteria that determined the required peak load generation capacity before the market reform
was abandoned in conjunction with the liberalisation. The problem of load capacity has become more
and more obvious during the last years. According to the law (in force until March 2008) the Swedish
national grid operator is obliged to ensure reliability of electricity supply by purchasing reserve
capacity.
One possible solution to the load problem may be the introduction of a new pricing strategy with a
load demand component, which means that consumers pay for load demand instead of electricity
consumption only. In this way, the customers would be more aware of their energy consumption
pattern and may be incited to lower the load demand, which could help the utility to avoid high load
peaks.

1473
The objective of this study was to investigate how such a tariff would affect one of the Swedish
electricity utilities and its more than 16 000 electricity customers.

Electricity price at user level

The total electricity price charged to the Swedish customers consists today typically of three parts:
electricity fee, network fee and taxes.

Figure 1: Residential electricity price structure [4].

The only part of the electricity bill that the customers themselves are able to influence is the electricity
fee. All customers have the opportunity to switch their electricity supplier or renegotiate their existing
contract, and, in this way, get a lower price.
The second part of the total electricity price, the network fee, is paid to the network owner in the area.
The network owner provides the physical transmission of electricity from the generation plants to the
end-user. Customers cannot choose their network provider so the network fee must be reasonable
and non-discriminatory. Network tariffs are supervised and published by the Swedish Energy Agency.
The third part of the electricity charge is taxes. In Sweden, like in all the other Nordic countries, the
consumption of electricity is taxed. Swedish customers have to pay two different types of taxes, an
energy tax and a value added tax (VAT). The energy tax for domestic customers depends on the
region. Industries pay no taxes at all at user level. The VAT is applied to the total price of electricity,
including the energy tax.
About 40 % of the total electricity price to a domestic customer is the price of electrical energy, 20 %
is the share of the network tariff and taxes account for 40 % [2].
Residential electricity customers can often receive two bills: one from the electricity supplier and
another one from the electricity grid company in the area. Both bills divide the fees into variable
(depending on the amount of electricity used) and standing subscription fees (see Figure 1). The
variable fee on the network bill is the charge for transmission and network services. The fixed part is
based on the main fuse used in the household and includes also governmental charges (as green
certificates etc) [3].

Previous experience from load demand tariffs

The main purpose of implementing a load demand component into electricity pricing is to draw the
customer’s attention to load demand (kW) rather than energy demand (kWh). In this way, customers
will hopefully become more conscious of their energy consumption pattern and possible load demand
problems.
As of January 1st 2001, Sollentuna Energi became the first Swedish energy utility to have
incorporated a load component into their grid tariff. Their experience is therefore of great interest
when other utilities are investigating the possibility of implementing load based electricity pricing
strategies.

1474
Sollentuna Energi’s load charge depends on the customer’s average load value of three daily 1-hour
load peaks during one month. This means that through achieving more even electricity use pattern,
customers can lower their network bill. The utility introduced the new tariff in a broad campaign
explaining “load demand” terms and giving many advices about different ways to lower load demand
in residential buildings, with and without electrical heating.
Sollentuna Energi’s new tariff showed that customers living in flats with a 16 ampere fuse level had
paid, with the old tariff, a lower grid fee than other customers. Some customers in flats had a
surprisingly high load demand and relatively large electricity use. Generally speaking, customers
living in flats with a 16 ampere fuse level incurred small increase in their grid fee while customers with
higher fuse levels (25 – 63 ampere) got a significant price reduction [5]. According to the evaluation
made by the utility itself it was possible to lower load demand about 5 % thanks to this new load
based tariff.
The experience from Sollentuna Energi also shows the importance of customers’ understanding the
difference between “power/load” and “energy” terms. In a study made on 1020 of Sollentuna Energi’s
customers in October 2002 [6], 78 % preferred the old tariff (where customers only paid for their
electricity consumption) to the new one. Some argued that it was bothersome to have one more thing
to think about concerning the electricity bills. Others argued that the new tariff created higher and
unfair electricity costs.

Case study - Skånska Energi Nät AB

Skånska Energi AB (SEAB) is an electric utility that operates in the southern-most county of Sweden,
Scania, supplying electricity to about 17 000 customers. The vast majority of these customers, about
75%, are residential customers, but there are also schools, agricultural properties and industrial
companies in the customer base [7]. SEAB is divided into a retail company - Skånska Energi Marknad
AB (SEMAB) and a grid company which owns the grid in the area - Skånska Energi Nät AB (SENAB).
SENAB is buying electricity from the high voltage grid owner within this area - E.On. The contract
states the highest hourly load demand, so called subscribed load, which was at the time of this
investigation 78 MW. If this level is exceeded, the utility pays fine per each kW, depending on the
terms of the contract with E.On. Over the past 5 years, the subscribed load capacity has been
exceeded twice (by 2 MW) - once on the morning of January 21st, 2004 and once on New Years Eve,
2001. The morning peak on January 21st, 2004 cost the company about half a million SEK (54 000
EUR). In order to avoid penalty charges from the supplier and to reduce load demand, and in the long
term decrease the subscribed load level, SENAB is interested in incorporating a load component into
the grid tariff. In 1998, SEAB invested in an advanced remote metering/billing system, CustCom. This
system, which is based on 1-hour measurements for all customers, makes it possible for the utility to
introduce such a tariff.
A specialised Internet module makes it possible for SEAB’s customers to enter a website and to
monitor their electricity use statistics (in kWh/h) whenever they wish, which may help them to verify
their network bill and to give more attention to their electricity use and load peaks.

Load demand tariff simulations

With a view to analyse how a grid tariff with a load demand charge could affect the utility and its
customers, a new pricing strategy (tariff) was constructed and price simulations, with varying load
tariff component values, were carried out [9].
The simulations were conducted as cost comparisons between the cost that the customers would
have with the new load demand tariff and the cost that they have currently, with the existing tariff.
The structure of the load demand tariff can be seen in Equation (1).
Φ = Pav a + s (1)
Pav [kW] denotes the average value of the customer’s three highest hourly load peaks from three
separate days during each month. a [SEK/kW] is a constant load price that takes two different values
- one from April to October and another from November to March. s [SEK] is the fuse level fee of the
network tariff (standing charge). Taxes and governmental fees are excluded from the analysed
pricing.
The structure of the existing tariff can be seen in Equation (2).
Φ = 0,149 E + S (2)

1475
E [kWh] is the electricity consumption during one month. 0,149 [SEK/kWh] is the energy unit price of
the network and S [SEK] is the standing charge of the network tariff. Taxes and governmental fees
are not included.
The price simulations were run for all of SENAB’s customers with fuse levels between 16A and 200A.
Customers were divided into groups depending on their fuse level. Customers with a 16-ampere fuse
were separated into tree subgroups: customers living in flats 16L, electric heated houses 16V and
houses with other heat source 16A.
In all four simulations, the condition that SENAB’s total revenue would be close to zero, seen over the
whole year, was applied. Component a was adjusted in order to achieve this.
In order to get a distinct difference between low and high demand periods, the component a in the
load demand tariff was almost doubled during the high demand period November - March, compared
to the low demand period April - October.

Simulation results
In the first price simulation the following premises were given: (1 SEK = 0.11 EUR)
s = S/2
a = 73 SEK/kW November-March
a = 35.5 SEK/kW April-October.
Figure 2 shows the difference in SENAB’s income (load demand tariff – existing tariff) for each fuse
group. Figure 3 shows the average cost increase for customers in each fuse group, when using the
new load tariff compared to the existing tariff.

1200000
1000000
800000
600000
400000
SEK

s=S/2
200000
0
-200000 16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-400000
-600000

Figure 2: Difference in SENAB’s income for each fuse level group (load tariff – existing tariff).
a = 73 SEK/kW November-March, a = 35.5 SEK/kW April-October, s = S/2.

2000
1000
0
-1000 16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-2000
SEK/year

-3000 s=S/2
-4000
-5000
-6000
-7000
-8000

Figure 3: The average cost increase for customers in each fuse group with the new load tariff
compared to the existing tariff.
a = 73 SEK/kW November-March, a = 35.5 SEK/kW April-October, s = S/2. (1 SEK = 0.11 EUR)

Negative values in Figure 3 imply that the average customer would be charged less with the new load
tariff. The results show that customers with low fuse levels would generally be charged more,
whereas customers with higher fuse levels would be charged less.
The second price simulation was preformed for s = S/3, a = 80 SEK/kW November-March, and a =
39.5 SEK/kW April-October. The findings from the second simulation were similar to that of the first

1476
one. 16L, 16A and 16V customers would incur higher charges with the load tariff, whereas the other
groups would be charged less (see Figure 4 and 5).

1200000
1000000
800000
600000
SEK

400000 s=S/3
200000
0
-200000 16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-400000

Figure 4: Difference in SENAB’s income for each fuse level group (load tariff – existing tariff).
a = 80 SEK/kW November-March, a = 39,5 SEK/kW April-October, s = S/3. (1 SEK = 0.11 EUR)

2000
1000
0
-1000 16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-2000
SEK/year

-3000 s=S/3
-4000
-5000
-6000
-7000
-8000

Figure 5: The average cost increase for customers in each fuse group with the new load tariff
compared to the existing tariff.
a = 80 SEK/kW November-March, a = 39,5 SEK/kW April-October, s = S/3. (1 SEK = 0.11 EUR)

In order to compare how a tariff based only on a load demand component would turn out, s was set to
zero (s = 0) in the third simulation. a = 95 SEK/kW November-March, a = 46.6 SEK/kW April-October.
In this case, 16A customers would be charged less with the load tariff and 20A-group would be
charged more, thus achieving the opposite result to the previous two cases. The other fuse groups
however were still following the trend achieved in the first two simulations (higher charges for 16L and
16V and lower charges for the others groups). The results can be seen in Figure 6 and 7.

1000000

800000

600000

400000
SEK

s=0
200000

0
16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-200000

-400000

Figure 6: Difference in SENAB’s income for each fuse level group (load tariff – existing tariff).
a = 95 SEK/kW November-March, a = 46,6 SEK/kW April-October, s = 0. (1 SEK = 0.11 EUR)

1477
 2000

0
16L 16A 16V 20A 25A 35A 50A 63A 80A 100A 125A 160A 200A
-2000
SEK/year

s=0
-4000

-6000

-8000

Figure 7: The average cost increase for customers in each fuse group with the new load tariff
compared to the existing tariff.
a = 95 SEK/kW November-March, a = 46.6 SEK/kW April-October, s = 0. (1 SEK = 0.11 EUR)

In the fourth and final simulation, the aim was for SENAB’s total revenue change, for each fuse level
group, to be as close to zero as possible. In this case, s was the component that was adjusted. a was
given the value of 70 SEK/kW from November to March and 35 SEK/kW during April-October. Table 1
shows the existing fuse fee and predicted fee for the new load tariff, if the goal was the one
mentioned above. Customers with higher fuse levels would in general incur a higher fuse fee
compared to customers with low fuse level. This means that with the existing tariff, customers with a
higher fuse level pay a relatively high standing charge in relation to their load demand. It is worth
noting that 125A customers would get a higher standing charge with the load tariff. This confirms the
conclusion that with the existing tariff, higher fuse level customers are subsidising customers with
lower fuse levels.

Table 1: Comparison between existing tariff’s and load tariff’s fuse fee
Fuse level Existing tariff’s fuse fee Load tariff’s fuse fee Ratio: load tariff /
SEK/year SEK/year existing tariff
(Ampere) (without VAT) (without VAT) (%)
16L 696 50 7,2 %
16A 1462 606 41,5 %
16V 1800 966 53,7 %
20A 2238 1333 59,6 %
25A 2792 1820 65,2 %
35A 3861 2500 64,8 %
50A 5438 3804 70 %
63A 6758 5162 76,4 %
80A 8568 7415 86,5 %
100A 10700 8567 80,1 %
125A 13344 14570 109,2%
160A 17072 15670 91,8 %
200A 21007 18000 85,7 %

Conclusions and recommendations

Conclusions from this study and recommendations that can be relevant for energy utilities when
planning load based pricing, have been gathered under some selected headings:

Existing tariff with load component

The main purpose of including load demand components into the network tariff is to achieve a lower
load demand and avoid load peaks. The analysis has shown that:
• Load based tariff adjusts pricing between fuse groups,
• Totally, load based tariff together with remote meter reading is profitable for utilities,
• The difference between “energy” and “power” must be explained in a comprehensive manner,

1478
 • To reach tariff’s goals, it is very important that customers understand the structure of load
tariff and its aim,
• Customers have to understand that they can save money by changing their energy
consumption patterns without the deterioration of comfort or standard of living,
• According to the utility’s own evaluation, it was possible to lower load demand about 5 %
thanks to the new load based tariff.

Tariff simulations
The results of this study show that:
• If a load demand component were to be introduced into SENAB’s network tariff, primarily
customers with a 16-ampere fuse would incur higher network charges compared to customers
with higher fuse levels, who would be charged less.
• With the existing network tariff, based on electricity use, customers with high fuse levels pay
today relatively high standing charges in relation to their exploitation of the grid.
• Several households would lower their fuse level (and the costs),
• It is not clear what would the introduction of load based tariff mean for total load demand level
within the simulated area.

Some important issues when introducing load based tariff

Electricity pricing should reflect real marginal costs of electricity production and the utilities’ costs.
Load based price could achieve higher price elasticity and thus limit the needs for expensive peak
load production. Many utilities have already invested in modern Automatic Meter Reading systems
(AMR) which facilitate implementation of load based tariffs. Customers are in such a case both an
exposed target and a vital potential - in many situations they really want to “help” society, and even
“their” utility, to avoid problems and shortages. Therefore, promotion of a new tariff with load based
price signal requires a solid and carefully prepared information campaign. It is of great importance for
the result that the purpose of such a tariff is clearly introduced to the customers from the very
beginning. The difference between “load demand” and “energy use” is not easy to understand and
keep after for the majority of customers. They need help to gain a better insight into how their
electricity costs will depend on their habits and usage of appliances and installations at home.

Load tariff structure

Load demand tariff should, as any tariff, be simple end easy to understand. The structure and price
levels are of decisive importance when trying to influence and change the patterns of energy use. The
tariff can always be adjusted afterwards but a comprehensive knowledge about consequences for
both customers and utility will help to avoid unnecessary sources of irritation and complaints.
Construction of a new tariff should start with an analysis of load characteristics for a grid company in
question - load curves for different customer groups, load factors and superposition factors as well as
load aggregation on selected levels in the grid should be investigated.
It is also essential to update the customer register regarding heating system, load guards etc. The
new tariff should be tested on some limited groups of customers.
A conceivable solution for a utility, when implementing a new load demand tariff, could also be to offer
its customers installation of diverse electronic devices (displays, load guards, soft heating systems)
helping them to “keep an eye” on load demand. Together with the new tariff, these investments
should be paid back in a relatively short time, helping at the same time to lower load demand in the
grid- a win-win solution for both partners.

Customer feed-back
Several investigations and studies have indicated that a continuous feed-back to energy customers is
of great significance while different energy related measures and changes are in progress. Possibility
to compare the results “before” and “after” or “own” with “others” can intensify and establish more long
lasting behavioural changes. Introduction of load demand tariff should therefore be supported by
continuous customer focused information. Market segmentation could give a hint how different
customer groups should be reached and influenced, depending on their energy related behaviour,
lifestyle, information sources and frame of reference.

1479
Extra values
Introduction of load demand tariff needs, or is made possible by, a remote meter system (AMR) with
hourly readings. This means that this new tariff should be seen as a part of a development of products
and services connected to the AMR system. A number of applications can for example improve
customer service and save needs of administration. Extra value-added services related to billing,
energy statistics, monitoring, energy guidance, grid optimisation etc, can create new possibilities and
values for the company and its customers.

Acknowledgements
This work was financed by the ELAN programme - a joint research programme on electricity utilisation
and behaviour in a deregulated market. The ELAN programme is financed by the Swedish Energy
Agency, the Swedish research council Formas and Swedish utilities through Elforsk.
The simulations described in this paper were performed by MSc Mattias von Knorring. The author
would also like to express his gratitude to the staff at the electric utility Skånska Energi AB who made
this research project possible.

References
[1] Swedish National Energy Administration. The Electricity Market 2003. ET13:2003.
[2] Swedenergy’. Http:// www.svenskenergi.se, 2004.
[3] The Swedish Consumer Electricity Advice Bureau. Http:// www. elradgivningsbyran.se, 2004.
[4] North, G. Residential Electricity Use and Control, Technical aspects. Report LUTMDN/TMVK--
7051--SE, Department of Heat and Power Engineering, Lund University, Lund, Sweden.
[5] Ahnland R. Effektavgift kan mota effektbrist. ERA1-2, 2002. (In Swedish)
[6] Fernström P. and Mackhé Å. Allmänhetens syn på energiavgifter. Temo-survey. Sollentuna
Energi, 2002, T-nr:23460. (In Swedish)
[7] Skånska Energi’s Web Site www.skanska-energi.se, 2004-04-28.
[8] Pyrko J. Direkt och indirekt laststyrning i samspel. (Interplay between load demand pricing and
direct load control). Lund University. Report LUTMDN/TMHP--05/3017--SE. (In Swedish)
[9] Knorring, von M. Analys av en ny prissättning med effektkomponenet för elkunder. (Analysis of
a new electricity pricing strategy incorporating a load demand component). Lund University.
LUTMDN/TMHP--04/5040--SE. (In Swedish)

1480
Innovative Thermal Energy Storage Systems for Residential Use
Andreas Hauer

Bavarian Center for Applied Energy Research, ZAE Bayern

Abstract
Innovative thermal energy storage (TES) systems are beneficial for residential use, especially in
HVAC applications. Thermal Energy from solar collectors or waste heat from industrial processes are
examples for energy sources, which can not be utilized effectively without TES. A huge potential of
energy sources substituting fossil fuels can only be exploited by the integration of TES systems. They
enable a greater and more efficient use of fluctuating energy sources by matching the energy supply
with the demand. This finally leads to a substantial energy conservation and reduction of CO2
emissions.
The growing peak demand of today’s energy consumption, essentially caused by air conditioning,
leads more often to black-outs all over the world. Such a problem - the shifting of a peak demand for
only a few hours or minutes - can be solved by TES. In this context TES can be the best solution not
only from the technical point of view, but also for economical reasons.
Different technologies can be used to store thermal energy. Advantages and disadvantages of these
technologies have to be taken into account in order to find the most suitable application for each TES
system. Demonstration projects are the best way to present the possibilities and proof the reliability of
innovative TES systems. Two examples carried out by the Bavarian Center for Applied Energy
Research are described in this paper.

Introduction
There is a huge potential for the application of energy storage systems. The fact that energy storage
systems are not as widely used as they could, is due to several reasons, in particular because some
of these systems are not yet economically competitive with the combustion of fossil fuels and their
long term reliability is not yet proven. There are still some regulatory and market barriers which have
to be overcome. Therefore, further attempts are being made to resolve these issues.
The IEA Implementing Agreement on Energy Conservation through Energy Storage (http://www.iea-
eces.org) was established in 1978 with the objective to facilitate international cooperation on
research, development and demonstration (RD&D) of new, innovative energy storage technologies.
Initially, attention was primarily focused on energy storage technologies improving the energy
efficiency of energy supply.
Over the last few years, the emphasis of the co-operative RD&D efforts has shifted towards storage
technologies that improve the manageability of energy systems or facilitate the integration of
renewable energy sources. In the future more application oriented topics like thermal energy storage
for cooling and industrial processes or mobile thermal storage systems for the utilization of waste heat
will be discussed. And since many storage systems are already developed products and have entered
the market or are close to it, the topic of market deployment will become more important within the
ECES.
The Bavarian Center for Applied Energy Research, ZAE Bayern, is active in different TES
technologies. In this paper two examples of demonstration projects of innovative thermal energy
storage systems, which were performed by the ZAE Bayern within the framework of the IEA ECES,
are presented.
The first project developed a combination of a water and a borehole storage. The advantage of such a
system is the minimization of heat losses from the water tank to the surrounding boreholes. The
second project is an open Sorption storage using Zeolite for heating and air conditioning of a building
in Munich. Both projects are the first of their kind world wide.

Thermal Energy Storage Technologies

Thermal energy can be stored in different ways given by the thermodynamics of the storage process.
If a storage medium is heated up or cooled down the storage is called sensible. If a phase change of
the medium is included in the temperature change, the thermal energy storage - TES - is called latent.

1481
These latent thermal storages can provide higher storage capacities and a constant discharging
temperature. One example is ice storage for cooling. If the thermal energy input during the charging is
initializing a reversible chemical reaction, the storage can achieve even higher capacities and is able
to deliver thermal energy at different discharging temperatures. The most developed reaction for
thermal energy storage is the adsorption of water vapour on micro porous materials.

Sensible TES
The use of hot water tanks is one of the best known thermal energy storage technologies. The hot
water tank serves the purpose of energy saving when e.g. applied to a solar tap water system and an
energy supply system with cogeneration. The major aim of an electrically heated hot water tank in a
tap water system is to shave the peak in electricity demand and consequently improve the efficiency
of electricity supply. A state-of-the-art research project as part of the energy storage programme has
resulted in the conclusion that the water tank storage technology has become mature and reliable.
The storage efficiency can be further improved by ensuring optimum stratification in the tank.

Figure 1: Aquifer thermal energy storage system [1]

The most frequently used storage technology which makes use of the underground is Aquifer Thermal
Energy Storage. This technology uses a natural underground layer as a storage medium for the
temporary storage of heat or cold (see figure 1). The transfer of thermal energy is realized by
extracting groundwater from the layer and by re-injecting it at the modified temperature level at a
separate location nearby.
Most applications are about the storage of winter cold to be used for the cooling of large office
buildings and industrial processes. It can easily be explained that aquifer cold storage is gaining more
and more interest: Savings on electricity bills for chillers are approx. 75 %, and in many cases, the
payback time for additional investments is shorter than five years. A major condition for the application
of this technology is the availability of a suitable geologic formation.
Other technologies for underground thermal energy storage are borehole storage, cavern storage and
pit storage. Which of these technologies is selected, strongly depends on the local geologic
conditions.
With borehole storage, vertical heat exchangers are inserted into the underground, which ensure the
transfer of thermal energy towards and from the ground (clay, sand, rock, etc.). Many projects are
about the storage of solar heat in summer for space heating of houses or offices. Ground heat
exchangers are also frequently used in combination with heat pumps, where the ground heat
exchanger extracts low-temperature heat from the soil.
With cavern storage and pit storage, large underground water reservoirs are created in the subsoil to
serve as thermal energy storage systems. These storage technologies are technically feasible, but the
actual application is still limited because of the high level of investment.

Latent TES and Chemical Reactions for TES

Sensible heat energy storage has the advantage of being relatively cheap but the energy density is
low and there is a variable discharging temperature. To overcome those disadvantages phase change
materials (PCM’s) could be used for thermal energy storage. The change of phase could be a
solid/liquid or a liquid/gas process. Melting processes have energy densities in the order of 100
3 3
kWh/m , e.g. ice, compared to 25 kWh/m for sensible heat storage.

1482
Figure 2: PCM storage system for utilization of night cold [2]

The incorporation of micro-encapsulated PCM materials such as paraffin wax into the gypsum walls or
plaster increases considerably the thermal mass and capacity of lightweight buildings. By night the
PCM in the microcapsules cools and solidifies. During the day the cool walls, reducing the daily
temperature swing by several degrees, and thereby avoiding the need for electric chillers or, at a
minimum, reducing the cooling requirements. Another application of active cooling systems is macro-
encapsulated salts that melt at an appropriate temperature. The PCM is stored in a building’s air vent
duct and the cold air is delivered via large-area ceiling and floor a/v systems. In figure 2 such a
system, called CoolDeck, in an office building in Stevenage in the UK is shown [3].
Higher energy densities can be reached by the utilization of chemical reactions for thermal energy
3
storage. Energy densities in the order of 300 kWh/m are possible. Thermochemical reactions like
adsorption (the adhesion of a substance to the surface of another solid or liquid) of water vapor to
Silicagel or Zeolites (micro-porous crystalline alumo-silicates) can be used to generate heat and cold
as well as to regulate humidity. Of special importance in hot, humid climates or confined spaces
where humidity levels are high, these open sorption systems use lithium chloride to cool water and
Zeolites to absorb ambient humidity.

Examples of TES Applications

Sensible TES for Solar Applications
The solar system in Attenkirchen, a village near Munich / Germany, is an example for a sensible TES.
The system includes a solar collector array, a seasonal TES, two compression heat pumps and a
distribution network to deliver the heat to the buildings [4].
If higher solar fractions, above 30 %, shall be reached for solar assisted heating systems, seasonal
storage of thermal energy is unavoidable in European climate. For the storage of large amounts of
thermal energy in the temperature range from 0 – 90 °C underground TES systems are for technical
and economical reasons the best solution. Two categories can be distinguished:
• The storage medium water is stored in non pressurized volume like rock caverns or concrete
containers.
• The storage medium is the ground, earth or rock, itself. Boreholes are necessary to charge
and discharge the ground.
The storage capacity is determined by the difference between the highest and the lowest operation
temperature in both cases.
The storage medium can be used as heat transport medium at the same time in the case of water
storages. This leads to an easy heat distribution and high thermal power of the storage, which is
important for the utilization of solar energy. Furthermore water has a comparable high specific heat, is
inexpensive and ecologically safe. The high cost for the concrete container is a disadvantage of such
a system.

1483
Borehole storages have an additional heat transfer between the storage medium ground and the heat
transport fluid. Additionally a slow velocity of the ground water is important. This type of sensible TES
is only applicable under certain boundary conditions and needs in most cases a water storage for
power buffering. Advantages of borehole storages are the low cost of construction and the possibility
to enlarge the system easily.
Parking
Thermal Insulation Heat Distribution

Water Tank TES

Borehole TES

Concrete Wall without

Insulation

Figure 3: Hybrid TES system including water tank and boreholes

In the project Attenkirchen a combination of a water and a borehole storage were developed (see
figure 3). A central cylindrical water container is surrounded by a number of boreholes. The central
water tank is used as a buffer and short term storage, while the borehole ring represents the long
term TES. There is no thermal insulation of the water tank, except the upper surface. The thermal
losses of the water storage can be collected by the surrounding borehole storages. This system
combines the operational advantages of a water storage with the economical advantages of a
borehole storage. This hybrid system allows the matching of a growing demand by simply adding
more boreholes.
At the end of the charging period, in late summer, the temperature within the water storage rises,
which leads to a decrease in the possible solar thermal input. The maximum temperature is crucial for
the reachable storage capacity. The highest temperature for non pressurized seasonal TES is about
95 °C.

Figure 4: Construction of the hybrid TES system in Attenkirchen

Figure 4 shows the construction of the hybrid storage system in Attenkirchen. It takes about 3 years
for the system to reach steady operation conditions. In the summer of 2003 and 2004 mainly the
water tank was charged in order to have enough energy for the winter heating. The heating load was
small in the first winter. Due to the partial completion of the residential buildings in the area, the
boreholes were not used in the beginning. Ground temperatures of about 45 °C were measure at the
end of the charging phase. During the heating period this temperature was dropped to 9 °C [5].
Chemical Reactions for TES: Sorption Storage Systems for Heating and Air Conditioning
An example for TES by chemical reactions is the open sorption storage using Zeolite as storage
medium. Figure 5 shows the schematically the charging and discharging process. In an open sorption
system an air stream transports heat and water vapour in and out of a packed bed of Zeolite pellets.

1484
Temperature and humidity of the air stream were influenced by the sorption process. The input and
output of heat and water vapour is shown in figure 5 [6].
Desorption / Adsorption /
Charging Discharging

air + heat of air + heat of

water condensation water evaporation

Zeolite

heat of heat of
air desorption air
adsorption

Figure 5: Open adsorption storage system

During desorption, which is the charging process, the air is heated up by the heat of desorption. This
heat is loosing the bound water in the microporous Zeolite and evaporates it. The cool air saturated
with water vapor is leaving the packed bed. During adsorption, the discharging process, the air is
transporting water vapor into the packed bed, where it will be adsorbed. The haet of adsorption is
released and heats up the air. The air leaves the packed bed hot and very dry.
For heating applications the heat of adsorption during adsorption and in some cases the heat of
condensation during desorption can be utilized. Heat for evaporation during adsorption has to be
available at a low temperature level. For air conditioning, which is based on the dehumidification of
the air (“desiccant cooling”), cooling can be delivered only during adsorption.
Thermal energy can be stored by sperating the desorption and adsorption step. Desorpbed Zeolite
stays charged until the adsorption process starts.
An open sorption storage was built up in Munich Germany by the ZAE Bayern. It is used for the
heating of a school building by utilizing thermal energy at 130 °C from the district heating net for
charging and the condensate of the return line for evaporation during adsorption (see figure 5). Air
conditioning is provided during Adsorption / discharging by the humidification of the dried air. Before
the cooling of the air by humidification the dried air is precooled by humidification of the exhaust air.

Figure 6: Open sorption storage system

The sorption storage in Munich provides peak shaving fort he district heat. Figure 6 shows the storage
modules in the background. Charging takes place during night or in times of low demand. During the
day and in peak hours the school building is heated by the stored energy. 700 kg of Zeolite can match

1485
the heating demand of one day at -16 °C outside temperature. In the summer time a jazz club, which
is located in the cellar close to the sorption storage, is air conditioned by the system. In this
application district heat is converted into cooling energy.
In the heating mode 92 % of the stored thermal energy could be delivered to the building. The
capacity of the sorption storage reaches 124 kWh/m³, which is about 3 times the capacity of a hot
water storage, at a charging temperature of 130 °C. If higher charging temperatures up to 300 °C are
available, the capacity could reach 250 kWh/m³ [7].
About 85 % of the stored heat could be converted into cooling energy for air conditioning by the
system. This result was accomplished at a charging temperature of 80 °C. The capacity of the
sorption storage for this application was still high at about 100 kWh/m³. At higher charging
temperatures the ratio of usable cooling energy to charged district heat would decrease, while the
storage capacity would increase.
If such a TES can be economically interesting, depends on the price for heat and cold delivered by
the system. This depends strongly on the number of charging and discharging cycles per time, which
was clearly increased by the double application – heating and air conditioning. 150 cycles for heating
and 100 cycles for air conditioning were assumed for the sorption storage application in Munich. This
leads to a pay back time for the system of about 6 to 7 years. Therefore this technology can compete
with conventional systems in the near future [8].

Conclusions and Outlook

The innovative TES systems in HVAC applications described above proofed that they can store
thermal energy efficiently and that they are not too far from conventional system concerning their
economics. Especially the use of sorption storage system allows beside the high storage capacity a
transformation from heat to cold and dehumidification [8].
Within the IEA ECES the topic of TES for cooling was emphasized over the last years. A new Annex
“Sustainable Cooling with Thermal Energy Storage” was formed last year (http://www.iea-
eces.org/annexes/annexes_ongoing.html). Within IEA ECES previous Annexes 7, 8, 10, 13 and 14
have looked at various aspects of cooling with TES alternatives. The results of these Annexes have
lead to an increase in awareness followed by initiation of TES activities. There is a need for the new
annex to provide new combinations of TES for different energy systems in different climates and
spread implementation of TES systems.
Another interesting field of application for innovative TES could be domestic appliances. Here short
term storage could avoid very efficiently power peaks in the future. For example the use of PCM
storages in cooling appliances or the application of sorption storages for drying processes look very
promising. Certainly this should be a field of future R&D activities.

References
[1] IF Technology b.v., Frombergstraat 1, P.O. Box 605, 6800 EA Arnhem, The Netherlands, e-
mail: office@IF-tech.nl .
[2] Climator AB, Kylarv. 2, SE-54134 Skovde, Sweden, Telephone: +46 (0)500 48 23 50, Fax: +46
(0)500 42 84 99, web: http://www.climator.com/index.html .
[3] Final Report of Annex 17 “Phase Change Materials and Chemical Reactions for Thermal
Energy Storage” of the ECES Implementing Agreement, can be ordered from: fredrik.
setterwall@comhem.se .
[4] M. Reuß. Solare Nahwärmeversorgung Attenkirchen, Proceedings of the 7. International
Symposium for Solar Energy SOLAR 2004 (8.-12.09.2004) in Gleisdorf, Austria, S. 111 – 120.
[5] M. Reuß, W. Schölkopf, Solar District Heating with Seasonal Storage in Attenkirchen,
Proceedings of the ISES International Solar World Congress in Gothenburg, Sweden, 14.-
19.06.2003, published on CD Gothenburg 2003.
[6] A. Hauer, Thermal Energy Storage with Zeolite for Heating and Cooling Applications,
Proceedings of the 3rd Workshop of Annex 17 ECES IA / IEA, Tokyo, Japan, 1.-2. Oktober
2002.
[7] A. Hauer, Thermal Energy Storage with Zeolite for Heating and Cooling Applications,
Proceedings of the 2nd International Heat Powered Cycles Conference, Paris, France, 5.-7.
September 2001.
[8] A. Hauer, Thermal Energy Storage with Zeolite for Heating and Cooling, Proceedings of the 7th
International Sorption Heat Pump Conference ISHPC ´02, Shanghai, China, 24.-27. September
2002.

1486
The Evolution of CLASP: A Status Report on a UN Sustainable
Development Partnership Devoted to Energy Efficiency Standards
and Labels
Christine Egan, Stephen Wiel

CLASP

Abstract
CLASP (Collaborative Labeling and Appliance Standards Program) was formed in 1999 as a
partnership devoted to advancing the extent and quality of energy efficiency standards and labels
(S&L) in developing countries. Since then it has evolved into a globally oriented, globally governed
non-profit corporation. In 2001, in preparation for the Johannesburg World Summit on Sustainable
Development, CLASP registered with the UN Commission on Sustainable Development as a
Sustainable Development Partnership. Recently, CLASP has provided major assistance to S&L
programs in China and India; played a key role in the development of a UNDP-GEF global initiative in
regional S&L; partnered with APEC in creating a website for global information on S&L; incorporated
as a 501(c)(3) non-profit organization; formalized collaborative relationships with sponsoring partners,
country partners, implementing partners, and affiliates; and created the S&L Energy Trust Fund. This
paper reports on the recent and continuing evolution of CLASP.

Background
CLASP maintains that an energy-efficiency S&L program for appliances is the most efficient and cost-
effective greenhouse gas mitigation practice. Once implemented, this practice effectively results in
the reduction of greenhouse gas emissions while promoting economic growth and the development
objectives that accompany that growth.
A total of 34 percent of global energy consumption results from human activities in buildings. Energy
consumption in buildings accounts for: (1) about 25 to 30 percent of all energy-related CO2
emissions; (2) 19 to 22 percent of all anthropogenic CO2 emissions; and (3) 10 to 12 percent of
mankind’s net contribution to climate change from all greenhouse gases. Furthermore, while energy
use in buildings is growing by just under 1 percent per year on average worldwide, in many
developing countries this growth rate is considerably higher—reaching nearly 9 percent in the
commercial sectors of many Pacific and Asian nations.
Building energy consumption involves the use of appliances, equipment, and lighting. Energy use in
buildings can be cost-effectively reduced by 15 to 25 percent over the next decade or two by the
accelerated adoption of more energy-efficient appliances, equipment, and lighting. Achieving this goal
would enhance economic development (especially in developing countries), reduce consumer energy
bills, and similarly reduce CO2 and local pollutant emissions. By promoting energy efficiency
standards and labels (S&L) for appliances, equipment, and lighting products, the Collaborative
Labeling and Appliance Standards Program (CLASP) is working toward this goal.
There are seven core stages in the S&L process including: 1) Considering a program; 2) Developing a
testing capability; 3) Designing and implementing a labeling program; 4) Analyzing and setting
standards; 5) Designing and implementing a communication campaign; 6) Ensuring program integrity;
and 7) Evaluating the program. These are described in detail in the CLASP S&L Guidebook [1], each
being a separate chapter. By the end of 2005, there were 62 countries engaged in this process that
implemented at least one energy-efficiency standard or label.

CLASP History
Since 1999, the Collaborative Labeling and Appliance Standards Program (CLASP) has been helping
energy efficiency standards and labels (S&L) policymakers and practitioners foster socio-economic
development, alleviate poverty, improve the environment, and stimulate global trade. It is an
outgrowth of an initiative begun in 1996 at Lawrence Berkeley National Laboratory to help developing
countries pursue energy-efficient S&L policies. Shortly afterward, the US Agency for International

1487
Development embraced the effort and funded Berkeley Lab, the Alliance to Save Energy, and the
International Institute for Energy Conservation to pursue this initiative. They formed CLASP.
For the five years 2000 through 2004, CLASP, operating as a partnership of its three founders with
increasing collaboration with additional partners and with $9.4 million from 12 different donors,
provided assistance for the development and implementation. CLASP has conducted work in
Argentina, Bahrain, Brazil, Colombia, Chile, China, the Dominican Republic, Ecuador, Egypt, Ghana,
India, Mexico, Nepal, Poland, South Africa, Sri Lanka, Thailand, Tunisia, and Uruguay. It has
supported regional S&L projects in 30 additional countries.
CLASP registered as a Sustainable Development Partnership with the UN Commission on
Sustainable Development (CSD) in 2002. Since registering with CSD, CLASP has partnered with
three other CSD Sustainable Development Partnerships – Efficient Energy for Sustainable
Development (EESD), Renewable Energy and Energy Efficiency Partnership (REEEP), and
Promoting and Energy-efficient Public Sector (PEPS). CLASP has also partnered with Asia-Pacific
Economic Cooperation (APEC) by co-sponsoring the joint APEC-CLASP Energy Standards
Information System (ESIS) website, a database with information on over 1700 standards.
In 2005, CLASP completed its originally intended transformation into an independent global
organization open to all willing and able participants. It is now governed by 12 directors from 8
countries in four continents. Upon its incorporation, CLASP established a formal collaboration of
Sponsoring Partners who fund CLASP activities, Country Partners who are the recipients of CLASP
services, Implementing Partners who provide CLASP services, and interested stakeholders as
Affiliates.

Highlights of CLASP’s Accomplishments to Date

Overall Impacts
CLASP measures its success in terms of saved energy and reduced tonnes of carbon dioxide
emitted. Saving energy reduces energy intensity. CLASP has assisted with the implementation of 21
new minimum energy performance standards, energy efficiency endorsement labels, and energy
information labels that will save 90 terawatt hours (TWh) of electricity and 86 megatonnes of CO2
(MtCO2) annually by 2014– avoiding 140 new large 1000-megawatt power plants. Most of this
savings is contributing to reducing energy intensity in China.

China
CLASP, with Lawrence Berkeley National Laboratory (LBNL) as its primary Implementing Partner, has
helped China implement a robust energy-efficiency standards and labeling program (S&L) that
includes minimum standards, voluntary energy labeling, and a residential energy consumption survey.
China's S&L program has transformed several product markets, improved the nation's economic
efficiency, and accelerated the pace of China's GHG mitigation
CLASP has assisted China in implementing S&L programs in the sector of the most rapid energy
consumption growth in China's economy. China has now implemented 11 minimum energy
performance standards (MEPS) for 9 products and endorsement labels for 11 products, including
refrigerators, air conditioners, televisions, printers, computers, monitors, fax machines, copiers,
DVD/VCD players, external power supplies, gas water heaters, and set-top boxes (under
development). These measures are estimated to save 85 TWh annually by their 10th year of
implementation. By 2020, China's S&L program is estimated to save 11% of its residential energy
use, reduce CO2 emissions by 34 million tons of carbon annually, and avoid the need for $20 billion
investment in power plant construction.
As China’s capacity for S&L implementation has grown, the nature of CLASP’s support has shifted
from technical training and capacity building for the domestic program to assistance in extending
market transformation effects internationally through harmonization of efficiency specifications. Most
notably, in 2005, China, Australia, and the US adopted a harmonized set of efficiency specifications
for external power supplies, based on a single testing standard. Current efforts support both the
application of China’s S&L programs into new market transformation programs—such as government
procurement—domestically as well as the expansion of China’s outreach internationally in additional
harmonization efforts.
The essence of CLASP's work in China is technology transfer, transferring to China the last 20 years
of experience and toolkits that have been developed around the world to support S&L programs. The
success relies heavily on cooperation with a wide range of organizations and groups and training of
our Chinese counterparts. LBNL has provided 196 person-weeks of training for 90 officials from five
agencies, split roughly evenly between training at LBNL and training inside China.

1488
S&L has become a prominent element in China's increasing emphasis on more sustainable energy
development and its recently announced energy intensity goals.

Regional S&L Projects

In order to help foster the alignment or harmonization of S&L among nations, CLASP has been
providing technical support to seven regional activities with this objective involving 42 nations.
CLASP has taken an active role in S&L programs within the Asia-Pacific Economic Cooperation
(APEC), the South Asia Regional Initiative for Energy Cooperation and Development (SARI), the Asia
and South East Asia Network (ASEAN), and the North American Energy Working Group (NAEWG),
and is helping UNDP-GEF develop several regional S&L projects. These efforts address the adoption
of the same test procedures, mutual recognition of test results, common comparative energy label
content, harmonized endorsement energy labels, and harmonized minimum energy performance
standards (MEPS).

Besides participating in numerous seminars, workshops, meetings, and teleconferences, and

providing technical information on S&L, CLASP has participated in the following major ways:
• APEC [21 Pacific-rim countries] – In 2004, APEC partnered with CLASP to expand its web-
based Energy Standards Information System (ESIS) beyond the APEC economies. The joint
ESIS-CLASP database has information on over 1700 standards. Visit <www.apec-esis.org>
and <www.clasponline.org>. CLASP has been participating in all of the activities of APEC's
Expert Group on Energy Efficiency and Conservation (EGEE&C) since 2002.
• ASEAN [Brunei Darussalam, Cambodia, Indonesia, Laos, Malaysia, Myanmar, Philippines,
Singapore, Thailand, Vietnam] – ASEAN engaged CLASP to help conduct an appraisal of the
ASEAN ballast market, conduct comparative “round-robin” testing for magnetic ballasts in six
ASEAN countries, and develop a common testing procedure for magnetic ballasts – all in
pursuit of an endorsement label for energy efficient products that would be available for
voluntary use by any ASEAN member country starting with magnetic ballasts. CLASP has
been working with ASEAN's Energy Efficiency and Conservation Sub-sector Network (EE&C-
SSN) and it’s ASEAN Centre for Energy (ACE) since 2001.
• SARI [Bangladesh, Bhutan, India, Maldives, Nepal, Sri Lanka, Afghanistan, Pakistan] –
USAID's South Asia Regional Initiative for Energy Cooperation and Development
(SARI/Energy) program, among other energy efficiency programs, engaged CLASP in efforts
to harmonize standards developed by each country in the region in order to isolate the region
from low quality and inefficient appliances and improve the energy efficiency of manufactured
appliances in the region. Focusing first on harmonizing refrigerator standards, CLASP
prepared white papers on testing facilities and protocols, and led workshops in Sri Lanka with
key regional technical expects to discuss the regional implications of the refrigerator
standards already developed by India & Sri Lanka. This historic meeting led to the formation
of an informal regional technical group to pursue future regional standards.
• NAEWG [Canada, Mexico, United States] – The 3-country North American Energy Working
Group's (NAEWG's) Energy Efficiency Expert Group has relied solely on CLASP to provide
technical support since its formation in 2001. In 2002, CLASP assembled detailed
comparisons of the three countries’ test procedures, to identify areas for potential
harmonization, showing that of the 46 energy-using products for which at least one of the
three countries had energy efficiency regulations, three products had nearly identical test
procedures in the three countries and ten other products had different test procedures but
near-term potential for harmonization. Since then, CLASP has met regularly with the Expert
Group conducting several other activities, all related to S&L. Recently NAEWG has become
the energy arm of the broader trilateral Security and Prosperity Partnership of North America
(SPP).
• UNDP-GEF – In 2002, CLASP began working with UNDP-GEF to develop a series of regional
projects to foster regional collaboration in S&L. In 2004 UNDP-GEF brought an S&L
international expert to its staff full-time in order to further develop and coordinate this effort.
As of the spring of 2006, UNDP-GEF's regional S&L initiative has projects just getting
underway in Central America (Costa Rica, El Salvador, Nicaragua, Panama), in the Andean
region of South America (Bolivia, Colombia, Ecuador, Peru, Venezuela), and in Southern
Europe (Bulgaria, Croatia, Romania, Turkey), with more such projects under development.
Besides playing a key role in the development of the initiative, CLASP is providing several
international experts to each of the three projects. For example, to the Southern Europe
project CLASP is providing: 1) a policy, legal & institutional expert; 2) a market studies expert;

1489
 3) a stakeholder assessment and awareness raising expert; and 4) a verification &
enforcement capacities expert.

S&L Tools
CLASP has developed four tools for policy makers and practitioners of S&L to facilitate their
development and implementation of their S&L programs:
ƒ CLASP maintains and keeps current a website making available comprehensive information
about all aspects of S&L and its proponents.
ƒ In 2005, CLASP published a second edition of Energy-Efficiency Labels and Standards: A
Guidebook for Appliances, Equipment, and Lighting, designed to help train the many
thousands of people worldwide needed to advance the quantity and quality of S&L practice.
ƒ In 2005 CLASP developed a survey instrument to help practitioners of S&L collect the home
and business energy use necessary to design effective S&L programs.
ƒ CLASP has developed, applied, and is continuously improving PAMS (Policy Analysis
Modeling System), a policy calculator that simplifies and decreases the cost of analysis to set
MEPS levels and calculate potential savings/benefit to the economy.

CLASP Today
CLASP promotes the world's best practices in energy efficiency S&L for residential, commercial and
industrial equipment and lighting. Four features allow CLASP to assist governments in saving energy
and reducing greenhouse gas emissions more cost-effectively and reliably than any other
organization in the world:
1. CLASP is the only organization in the world devoted solely to addressing the world's most
effective policy for fostering economic development and reducing greenhouse gas emissions –
energy efficiency standards and labels.
2. CLASP is one of only two UN-sanctioned Sustainable Development Partnerships devoted solely
to energy efficiency.
3. CLASP embraces all organizations and individuals who have the ability and interest to foster
CLASP's mission.
4. CLASP assigns the world's most qualified individual to each task independent of nationality or
affiliation.

CLASP’s Future
CLASP recently announced the establishment of a Standards and Labeling (S&L) Trust Fund. CLASP
hopes the fund will accelerate the adoption of S&L globally as well as enhance the efficacy of the
standards and labels that are adopted for the betterment of the planet. In the countries to which funds
are applied, the fund will:
ƒ Reduce national utility bills and thus enhance socio-economic development, improve economic
efficiency and alleviate poverty;
ƒ Reduce global greenhouse gas emissions as well as local pollutants;
ƒ Stimulate global trade and enhance local competitiveness; and
ƒ Expand the delivery of S&L technical assistance based upon international best practice.
CLASP has successfully attracted funding for big projects in what are generally considered to be key
countries including: China, India, and Brazil. With the Trust Fund, CLASP can also bridge gaps in the
basic S&L infrastructure, develop new country relationships, and contribute vital support to small
nations. There are many steps involved in building S&L infrastructure that sometimes don't fit the
standard criteria used by sponsors—work is not country-specific, it has no measurable energy or
carbon savings, its payoff is too far in the future (sometimes over-one-year is too long). Yet this work
often provides the critical foundation upon which quality S&L programs are built.
CLASP has secured a central place in the S&L world and, in that position must balance varied and
numerous opportunities. CLASP seeks to secure $1,000,000 in 2006 to allow it and its partners to
pursue short-term and foundational investments that otherwise would be foregone. These funds would
be applied to projects that satisfy the following strict criteria:
ƒ Would result in substantial long-term impact in advancing S&L;
ƒ Would enable or stimulate major effort by others in the long-term;
ƒ Would establish key collaborations; and
ƒ Would be a lost opportunity without CLASP funding.

1490
Trust Fund projects will deliver the world's best practices provided by the most highly competent
project teams assembled from among the world's most accomplished S&L experts. Sponsors will
serve as Trustees, receive reports on the use of the funds and its impact, and participate in annual
meetings conducted in order to focus on lessons learned and refine the Trust Fund’s annual program
strategy. Trust fund sponsors will receive special recognition in CLASP’s outreach materials.

References
[1] Wiel, Stephen, and James E. McMahon. Energy-Efficiency Labels and Standards: A
Guidebook for Appliances, Equipment and Lighting; 2nd Edition. CLASP, Washington, D.C.,
February 2005. Can be downloaded at www.clasponline.org under "The S&L Guidebook".

1491
1492
European Commission

EUR 22317 EN – DG Joint Research Centre, Institute Environment and Sustainability

Title: Energy Efficiency in Domestic Appliances and Lighting – Proceedings of the 4th International
Conference - EEDAL'06, Volume 3
Authors: Paolo Bertoldi, Bogdan Atanasiu
Luxembourg: Office for Official Publications of the European Communities
2006 – 534 pp. – 21 x 29,7 cm
EUR - Scientific and Technical Research series; ISSN 1018-5593
ISBN 92-79-02752-2

Abstract
This book contains the Proceedings of the 4th International Conference on Energy Efficiency in Domestic
Appliances and Lighting, London (UK), 21-23 June 2006. The EEDAL’06 conference has been very
successful in attracting an international audience, representing a wide variety of stakeholders involved in
policy implementation and development, research and programme implementation, manufacturing and
promotion of energy efficient residential appliances and lighting. The international community of
stakeholders dealing with residential appliances and lighting gathered to discuss the progress achieved in
technologies and policies, and the strategies to be implemented to further this progress.
EEDAL'06 has provided a unique forum to discuss and debate the latest developments in energy and
environmental impact of residential appliances and installed equipment, and lighting. The presentations
were made by the leading experts coming from all continents. The presentations covered policies and
programmes adopted and planned in several geographical areas and countries, as well as the technical
and commercial advances in the dissemination and penetration of energy efficient residential appliances
and lighting.
 LB- NC- 22317- EN- C
The mission of the JRC is to provide customer-driven scientific and technical support for the
conception, development, implementation and monitoring of EU policies. As a service of the European
Commission, the JRC functions as a reference centre of science and technology for the Union. Close
to the policy-making process, it serves the common interest of the Member States, while being
independent of special interests, whether private or national.