Tesis - Rellenos Sanitarios - 02
Tesis - Rellenos Sanitarios - 02
Since the mathematical models can use different parameters, yielding different
estimates, the project activity efficiency must be evaluated according to each
model/parameters set, which can be independently done year after year, and then
compared with the planned project efficiency.
As the carbon credits generated by CDM projects will assist other countries to
meet their reduction targets required by legally binding international agreement,
every caution is taken during the process to ensure the integrity of the mechanism.
A project activity can only use a methodology approved by the CDM Executive
Board, scientifically supported by a Methodologies Panel, with detailed conditions
and steps to be followed, even maintenance frequency for measuring instruments.
A Designated Operational Entity (DOE) acting on behalf of the Executive Board
should validate the compliance with the methodology. During the project life,
another DOE will be required to verify and approve monitoring reports that will be
issued throughout the years.
Calculations of expected methane generation as contained in the Project Design
Document (PDD) have been redone and checked using the same model. From the
PUC-Rio - Certificação Digital Nº 1213364/CA
3. THE LANDFILLS
The landfills studied here had their potential for biogas evaluated at their inception
and which later on had their generation models and their parameters re-evaluated.
This is done upon renewal of their first 7-year crediting period, according to the
CDM procedures, when it is necessary to update the baseline and monitoring
methodology. Only four CDM projects in Brazil fell under this condition in May
2014.
EXPANDED ABSTRACT
3.2 Bandeirantes Landfill Gas to Energy Project (BLFGE) – Ref 0164
The landfill received waste from 1979 to 2007. The first crediting period ranges
from 2004 to 2010. Monitoring reports cover the period up to 31 August 2012.
In the first PDD, the IPCC (1997) model is used, with emission maximum since
the year of disposal; in the second PDD (March 2012), the CDM Tool model is
used, and there is an update of the waste amounts disposed of in the past seven
years.
Extraction efficiencies – for each model – were calculated after the monitored
methane extraction amounts were recovered from the final monitoring reports,
which cover uninterrupted periods. Whenever necessary, the beginning years or the
final years of these periods were completed based on daily averages, so as to permit
full year comparison. In such way, 73 days were added in 2007 for Adrianópolis;
for Bandeirantes, 122 days were added in 2012; for ESTRE’s-Paulínia, 256 days
were added in 2006, and 92 days in 2012; for Caieiras, 31 days were added in 2007,
and 92 days in 2012.
In order to allow comparison among the models used in different opportunities
for each landfill, care was taken with:
All the calculations have been redone in order to check the PDDs, following the
indicated parameters. In the first PDDs there was no worksheet available, except
for ESTRE’s Paulínia; for the second ones, worksheets were available in the
CDM site. To arrive at the results presented in PDDs, some corrections in the
EXPANDED ABSTRACT
parameters were necessary as indicated by the reverse calculations. In one case,
a double discount related to the oxidation factor was found.
In the first PDDs, no oxidation factor was used. As this can be added simply in
the end, this factor is used here in all models for final comparison.
Only in the CDM Tool model there is an uncertainty factor. Prior to its version
06.0.0 of 25 November 2011, this factor was 0.9, and then changing to 0.75. This
uncertainty factor did not exist in the first PDDs using IPCC (1997) or IPCC
(2000) models. For comparison with the others, the CDM Tool model is
presented here with three uncertainty factors: 1, 0.9 and 0.75.
As the second PDDs give actual values for the waste disposal amounts in the
seven previous years, the updated values are used here for all models.
The values for k and L0 used for the landfills in the first PDDs are shown in Table
3, along with the corresponding values suggested by IPCC (2006) for bulk waste,
which depend on the parameters reveled in the second PDDs.
In the second PDDs, parameters kj, DOCj are IPCC (2006) default values,
considering that Caieiras has a temperate-wet condition and the others have a
tropical-wet condition. Waste composition for each landfill is shown in Table 4. All
have their own waste composition, except Caieiras that adopted IPCC (2006)
default values for South America.
In Table 5 there is a summary of the information from PDDs and monitoring
PUC-Rio - Certificação Digital Nº 1213364/CA
Table 3. Decay constant and methane potential used in the first PDDs
PDD-1
-1
Landfill k (year ) L0 (m3 CH4)
Model
Used IPCC 2006 Used IPCC 2006
Adrianópolis 0.1 0.17 117.14 58.37 IPCC 1997
Bandeirantes 0.102 0.17 84.10 73.01 IPCC 1997
ESTRE’s Paulínia 0.09 0.17 140.00 68.71 IPCC 1997
Caieiras 0.08 0.09 115.78 75.43 IPCC 2000
EXPANDED ABSTRACT
Table 5. General information of the landfill sites studied (from PDDs and
monitoring reports)
Disposed Projected
Number of Recovered
wastes recovery Monitored
Landfill Start monitoring methane
up to 2012 efficiency period
reports (t)
(t) in PDD-1
Adrianópolis 2003 7,780,010 85% 6 6 years 26,292
Bandeirantes 1979 37,226,873 80% 18 9 years 232,075
ESTRE’s Paulínia 2000 13,512,504 80% 14 7 years 123,460
Caieiras 2002 20,234,148 80%* 7 6 years 180,365
* In 2007, the recovery efficiency was estimated in PDD as 70%.
With the annual amounts of recovered methane and the estimation by each key
model, the annual efficiencies were calculated for each year of the monitored
periods. Maximum, minimum and average annual values are shown in Table 6 for
each model. The first two are the model used in PDD-1 – IPCC (1997) model or
IPCC (2000) model with factor A – and its correction as discussed here, the IPCC
(2000) model with factor A’ and delay. The other five models are options using data
from PDD-2: the IPCC (2006) model with bulk option; the CDM Tool model and
its correction, considering the uncertainty factor as 1, and the corrected CDM Tool
model with uncertainty factor as 0.9 and 0.75.
PUC-Rio - Certificação Digital Nº 1213364/CA
The results in Table 6 are showing methane overestimation in the first PDDs –
column A, which presents the lower figures, not only because of the model used, as
discussed before, but also due to a higher value for L0 in relation to IPCC (2006)
defaults, as shown in Table 3. Or due to the assertion of Themelis and Ulloa (2007)
that a methane generation of about 50 m3 CH4 / t MSW is conservative. Note that
if oxidation factor of 10% were not used for this comparison, the efficiencies would
be even lower. The correction of the first model gives lower estimations and higher
efficiencies, as per column B. If used IPCC (2006) bulk option – column C, the
efficiencies would be even better except for Bandeirantes.
In the second PDDs, the fraction approach is used. The actual CDM Tool but
without uncertainty factor – column D – already shows better efficiencies in relation
to the first PDDs – column A – or its correction – column B, except in Caieiras.
Column D can also be compared to the IPCC (2006) model in the bulk option –
column C, where there are better results in two cases – Bandeirantes and ESTRE’s
Paulínia, but worse in Caireiras. The specific case of Caieiras will be analyzed
further ahead.
If column D model is changed for its correction – column E – the improvement
in relation to the columns A and B are true for all landfills. However, the
comparison of column E with column D shows an efficiency reduction in
Bandeirantes landfill. It should be noted that the variations in efficiency in the
corrected CDM Tool model depend on the waste disposal time series, as said before.
In this sense, the waste disposal at Bandeirantes landfill ending in 2007 explains
the decreased efficiencies when CDM Tool model is corrected.
When the uncertainty factor is taken into account, as prescribed in the CDM
Tool, methane estimations decrease, and efficiencies increase. This is clear by
observing columns E, F, and G It seems that φ=0.75 leads to an excessive discount.
Adrianópolis has the lowers results, even lower if compared to the projected
recovery efficiency of 85%. This value will be reconsidered to be 40% in the second
EXPANDED ABSTRACT
PDD, which is consistent with the assertion that a zone in operation that has no top
cover and is connected to a biogas recovery system has a collection efficiency of
35% (Scharff and Jacobs, 2006). However, maybe the main reason for a low
efficiency is that, according to PDD-2, an extraction system was not in place for all
areas receiving wastes.
IPCC 2006
CDM Tool
CDM Tool
CDM Tool
CDM Tool
(φ=0.75)
corrected
corrected
corrected
corrected
(φ=0.9)
PDD-1
PDD-1
Indicator
(φ=1)
(φ=1)
bulk
Landfill
A B C D E F G
Max 21% 27% 32% 30% 34% 38% 45%
Adrianópolis Min 12% 14% 18% 19% 21% 23% 28%
Avg 16% 20% 25% 24% 28% 31% 37%
Max 47% 46% 43% 64% 52% 58% 69%
Bandeirantes Min 25% 23% 28% 51% 41% 45% 54%
PUC-Rio - Certificação Digital Nº 1213364/CA
It also worth noting that for Adrianópolis in Table 6, it was used a methane
fraction in biogas of 39.74%, according to PDD-2 and based in past measurements.
As the usual value is about 50% (Themelis and Ulloa, 2007; IPCC, 2006), this could
be an indication that the negative pressure for extraction could be excessive leading
to dilution of the biogas with outside air, lowering the CH4 ratio to CO2 (Bingermer
and Crutzen, 1987). Scharff and Jacobs (2006) also indicate that the extraction
system can lower the methane concentration and reduce it from the default value of
55%. Moreover, IPCC (2006) guidelines note that the fraction of methane in
generated landfill gas should not be confused with measured methane in the gas
emitted from SWDS, because CO2 is absorbed in seepage water, which would lead
to a higher methane fraction. If 50% were used instead of 40%, the efficiencies for
Adrianópolis would be even lower in inverted proportion.
On the other hand, Caieiras has very high results. As it can be observed, default
waste fractions from IPCC (2006) were adopted for it, differently from
Bandeirantes, a neighbor landfill, both serving the São Paulo municipality. The
food fraction – the fastest degrading one – is 44.9% in the first case and 60.6% in
the second. Another difference is that Caieiras was considered to be in a temperate-
wet condition but Bandeirantes is said to be in a tropical-wet condition: they are in
a threshold situation that has consequences in the decay constants. If the same
Bandeirantes’ parameters were used for Caieiras, its maximum efficiency –
EXPANDED ABSTRACT
considering the CDM Tool with corrections but without uncertainty factor – drops
from 106% to 64%, which seems more reasonable.
Table 7. Differences from the used models to the ones without the intrinsic errors
First crediting period Second crediting period
Landfill
PDD-1 PDD-2
Adrianópolis 30% 3%
Bandeirantes 1% -13%
ESTRE’s Paulínia 23% -6%
Caieiras 28% 7%
5. CONCLUSIONS
When the Kyoto Protocol entered into force in 2005, or even earlier, as Brazilian
entrepreneurs became aware of its imminence, it became a driving force to
transform uncontrolled SWDS into landfills associated with the capture and
destruction of methane. This took place in Brazil as well as in the developing world.
EXPANDED ABSTRACT
Having the incentive of obtaining carbon credits, entrepreneurs made investments
in projects that could solve environmental problems; and, when the credits were
sold, the economic equation would be completed.
In the recent years, in addition to the drop in carbon-credit prices, the results
demonstrated in this thesis show that the CDM projects face difficulties in
predicting methane generation. Although partially solved with improved models,
there are still uncertainties in these models, such as in establishing values for the
decay constants that depend on local climatic conditions; and, there are also
uncertainties on the composition and amount of waste annually disposed of. In this
regard, additional research is necessary in order to look further into the local reality.
This thesis also recommends the CDM Executive Board to correct its
methodological tool to calculate emissions from solid waste disposal sites, as its
use yields a significant overestimation in the forecasted amount of methane
generation in new landfills and an underestimation in landfills closing their
operation, as demonstrated in this study.
In view of a Brazilian law, which enforces the end of uncontrolled solid waste
disposal sites, the use of improved models to predict methane generation is very
important in order to allow a correct understanding of how this greenhouse gas can
be used as a clean energy source in lieu of being irresponsibly released into the
atmosphere.
PUC-Rio - Certificação Digital Nº 1213364/CA
REFERENCES
EXPANDED ABSTRACT
____. Personal communication, received at <mauro.meirelles.mct@gmail.com>
on 16/10/2014.
Scharff, H. and Jacobs, J. Applying guidance for methane emission estimation
for landfills. Waste Management 26, 2006: 417-429. 13 p.
Themelis, N. J., Ulloa, P. A. Methane generation in landfills. Earth Engineering
Center and Department of Earth and Environmental Engineering, Columbia
University, New York, NY 10027, USA. Renewable Energy 32, 2007: 1243-
1257. 15p.
UNFCCC. COP Decision 17/CP.8 – Annex, paragraph 11, 2002a. Link:
http://unfccc.int/resource/docs/cop8/07a02.pdf#page=2
____. COP Decision 18/CP.8 – Annex, paragraphs 9 and 10, 2002b. Decision’s
link: http://unfccc.int/resource/docs/cop8/07a02.pdf#page=13; annex’s link:
http://unfccc.int/resource/docs/cop8/08.pdf
____. Methodological Tool – Emissions from solid waste disposal sites, 2012.
Link: http://cdm.unfccc.int/methodologies/PAmethodologies/tools/am-tool-04-
v6.0.1.pdf., version 06.0.1
____. COP Decision 24/CP.19 – Annex I, paragraphs 3, 9, and 10, 2013. Link:
http://unfccc.int/resource/docs/2013/cop19/eng/10a03.pdf#page=2
US-EPA. Landfill Gas Emissions Model (LandGEM) Version 3.02 User´s Guide,
United States, 2005a. EPA-600/R-05/047.
____. First-Order Kinetic Gas Generation Model Parameters for Wet
PUC-Rio - Certificação Digital Nº 1213364/CA
Keywords
Landfills; biogas generation; models; CDM; Clean Development Mechanism.
EXPANDED ABSTRACT
Sumário
Resumo 6
Expanded Abstract 7
1. Introdução 37
1.1 O problema 37
1.2 Objetivos (final e intermediários) 41
1.3 Relevância do estudo 41
1.4 Delimitação do estudo 43
1.5 Literatura consultada 44
1.5 Organização desta dissertação 46
PUC-Rio - Certificação Digital Nº 1213364/CA
5. Resultados e discussão 96
PUC-Rio - Certificação Digital Nº 1213364/CA
1.1 O problema
13
Conforme informação do site da Convenção do Clima, acessado em 23/11/2014, no
endereço http://cdm.unfccc.int/Projects/projsearch.html.
39
país dividia-se entre 76% em lixões a céu aberto, 22% em aterros controlados e
sanitários e 2% em outras destinações. Não havia até então maiores preocupações
com o metano que emanava para atmosfera, a não ser manter uma chama em alguns
drenos no chão dos aterros para queimar o gás e evitar riscos de explosão pelo seu
acúmulo, já que o metano não é tóxico, é inodoro, embora comburente e
potencialmente explosivo. É preciso que se diga que os lixões têm geração de biogás
menor do que os aterros, já que não possuem a mesma condição anaeróbica –
ausência de oxigênio – que caracteriza os últimos.
Com o MDL, abriu-se a oportunidade para, ao lado de se melhorar as
condições sanitárias da disposição dos resíduos sólidos urbanos, através da
construção de aterros sanitários em substituição a lixões, captar e oxidar o gás de
efeito estufa a ser emitido em maior quantidade – o que geraria créditos de carbono
a serem transformados em recursos financeiros que viabilizariam os
empreendimentos. Tal oxidação poderia ser a simples queima eficiente do metano,
PUC-Rio - Certificação Digital Nº 1213364/CA
quanto seu uso para a geração de energia elétrica, com redução adicional de gases
de efeito estufa pela redução do uso de termelétricas do sistema interligado
nacional. A oportunidade foi aproveitada pelos empresários no Brasil, o primeiro a
registrar um projeto MDL no mundo14, justamente desse tipo, o NovaGerar, Projeto
de Energia a partir de Gases de Aterro Sanitário, situado em Nova Iguaçu, no Estado
do Rio de Janeiro (será visto no item 4.2.1).
Registre-se que existem outras soluções em que se evita a geração do metano
nos aterros sanitários e que já podem ser projetos MDL: evitando-se a disposição
de resíduos orgânicos no aterro, como sua queima de resíduos para energia ou sua
compostagem para adubo, ou fazendo-se uma aeração passiva ou ativa do aterro.
Soluções para a oxidação do metano por conta de uma camada especial de
recobrimento do aterro, com uso de bactérias metanotróficas, ainda são objeto de
estudos (Bahr et al., 2006; Teixeira et al., 2007), mas ainda não têm uma
metodologia aprovada pelo Conselho Executivo do MDL.
Ao se pensar em fazer um projeto MDL de redução de emissões de aterro
sanitário é importante prever a quantidade de gás a ser gerada durante a vida útil
desse aterro. Uma das dificuldades para se fazer um projeto MDL é demonstrar que
ele só seria feito a partir da perspectiva dos ganhos a serem gerados pelos seus
14
Ver a ordem dos projetos MDL em http://cdm.unfccc.int/Projects/projsearch.html, com os
campos de pesquisa em branco.
40
(ICLEI, 2009), o que se confirmava também pela análise dos projetos MDL
brasileiros, onde se explicitava que, antes do MDL, não havia interesse em investir
recursos para a queima do gás metano (von Randow et al., 2009). Embora seja
necessário para os projetos de aterro sanitário sob MDL, saber com antecedência
quanto um aterro vai gerar de emissões é importante também nos países
industrializados, para quantificação do potencial energético do efluente gasoso ou
simplesmente para a contabilização em seus inventários nacionais de gases de efeito
estufa.
Em resumo, a questão que se quer aqui discutir, já que os aterros sanitários
atuais preveem a captação e o uso do biogás, é: