2020 IEEE 17th International Conference on Smart Communities: Improving Quality of Life Using ICT, IoT and AI (HONET)

Dynamic Line Rating (DLR) by Weather-Based

Calculation for Power Grid Optimization in Tenaga
Nasional Berhad (TNB)
2020 IEEE 17th International Conference on Smart Communities: Improving Quality of Life Using ICT, IoT and AI (HONET) | 978-0-7381-0527-7/20/$31.00 ©2020 IEEE | DOI: 10.1109/HONET50430.2020.9322810

Muhamad Shahmi Muhamad Shokri Farah Adilah Mohd. Kasran Muhammad Hanif Abdul Aziz
Utility Automation Unit Utility Automation Unit Utility Automation Unit
TNB Research Sdn. Bhd. TNB Research Sdn. Bhd. TNB Research Sdn. Bhd.
Selangor, MALAYSIA Selangor, MALAYSIA Selangor, MALAYSIA
shahmi.shokri@tnb.com.my farahadilah.kasran@tnb.com.my mhanif.aziz@tnb.com.my

Normi Salwana Miswan Mohd Noorfazly Noran Azlan Abd Rahim

Utility Automation Unit Utility Automation Unit Utility AutomationUnit
TNB Research Sdn. Bhd. TNB Research Sdn. Bhd. TNB Research Sdn. Bhd.
Selangor, MALAYSIA Selangor, MALAYSIA Selangor, MALAYSIA
salwana.miswan@tnb.com.my noorfazly.noran@tnb.com.my azlan.rahim@tnb.com.my

Abstract—Dynamic line rating (DLR) is a practical solution • Wind speed = 0.5 m/s
for line optimization in power transmission by dynamically
increase and decrease line rating. Instead of limiting the line • Wind direction = 45o
rating to the fixed value in Static Line Rating (SLR) based on With the conservative values assigned to the ambient
worst conditions, DLR uses the line monitoring sensor and
parameters as above and line sag being limited by the ground
weather monitoring system to provide the actual line operating
condition. TNB, as the power utility operator in Malaysia,
clearance, overhead line rating is then limited to the conductor
always aims to find the best solution to improve overhead line temperature of up to 75°C. There is an opportunity to utilize
performance. To date, two DLR sensors have been installed at the spare capacity (estimated about up to 30% more) of
selected line spans and the DLR data for these lines is accessible the transmission line if the actual operating conditions (i.e.
via a web portal of the system. However, the current DLR data line thermal rating) are carefully monitored with the
is limited only to the line span with the installed sensors. deployment of DLR system in the grid system operation.
Therefore, this paper attempts to provide the DLR data based
The deployment of DLR system uses a sensor installed at
on weather parameters that will be useful for line rating
the transmission line to measure the line temperature, line sag,
indication without line sensor installation. A detailed
comparison of DLR data is analyzed between the weather-based ground temperature, and tilt/angle for wind direction. Besides
calculation and the existing system at the line with sensor and it that, it also requires the weather data from the outsource
shows a consistent pattern between both DLR data. This work weather provider or in-situ weather monitoring system for
will then serve as a base for future research in the weather-based more accurate data [3]. All of the data is then applied to
calculation to obtain DLR value for TNB grid optimization. calculate the line thermal rating to indicate the line
performance. In order to do so, several DLR sensors and
Keywords—Overhead transmission line, Dynamic Line weather monitoring systems need to be installed for getting
Rating, DLR weather-based calculation accurate data for line rating calculation especially for the
longer transmission line.
I. INTRODUCTION
Currently, many technologies in DLR system can be found
The current approach in power utility, to ensure the in the market such as Lindsey, USi Power Donut [4, 5],
optimum capacity carried by the transmission lines is using Ampacimon Dynamic Rating (ADR) [6], Span Sentry, Easy
Static Line Rating (SLR). By fixing the current capacity of the Monitoring Overhead (EMO), and Overhead Transmission
transmission line based on the conductor characteristic will Line Monitoring. Different technologies offer different
limit the line performance where it can actually carry more methods to measure the line rating and different software to
current capacity with detailed regard on the ambient provide DLR information to clients. For instance, some DLR
conditions (i.e. air temperature, solar radiation, and wind technologies (i.e. Span Sentry, Easy Monitoring Overhead
speed and its direction) and the conductor temperature [1, 2]. (EMO), and Overhead Transmission Line Monitoring) come
The recent system, so-called Dynamic Line Rating offers a with meteorological sensors to measure the ambient
great solution to improve the current capacity of transmission parameters such as wind speed and direction, ambient
lines while maintaining its performance. Instead of carrying a temperature, and solar radiation. While SMARTLINE System
fixed current capacity in transmission line, the DLR system by Lindsey uses the outsource weather monitoring to provide
dynamically increase and decrease the line capacity based on the ambient parameters, so-called DarkSky. Not only
real-time ambient monitoring. Currently, the calculation on providing the real-time DLR data, SMARTLINE System
overhead line ampacity rating is based on the following offers several hours-ahead DLR forecast data that helps the
parameters and values: customer in line operation optimisation. For that reason, TNB
• Ambient temperature = 40oC has chosen the SMARTLINE system technology by Lindsey
to deploy the DLR system in the grid network. To date, two
• Solar Radiation = 850 W/m2 TNB line-spans have installed the Lindsey sensors after a

978-0-7381-0527-7/20/$31.00 ©2020 IEEE

113

Authorized licensed use limited to: California State University Fresno. Downloaded on June 19,2021 at 23:19:07 UTC from IEEE Xplore. Restrictions apply.
thorough study on the span critical parameters such as the line-
span, low ground clearance, and the human activity under the
line. After successful completion of the pilot project of sensor
installation at the selected span, TNB Grid can now access the
DLR data through the SMARTLINE web portal.
However, to maintain the grid performance and to have
better accuracy of DLR data, more transmission lines sensors
(Lindsey sensor) need to be installed at the strategic location.
Therefore, it is necessary to do a deep research on the DLR
calculation based on the available parameters without using
the sensor data. This work will be useful for TNB Grid in the Fig. 2. Lindsey transmission line monitoring sensor; bottom view of the
future for optimizing the current capacity. Hence, this present sensor (left), and sensor features (right)
work aims to develop an algorithm to calculate DLR, by
taking the existing DLR data extracted by Lindsey sensors as Under steady-state, the conductor temperature, Tc is
a benchmark. defined based on the thermal heat balance of overhead lines
(2) and (3):
II. BRIEF EXPLANATION ON SMARTLINE SYSTEM
As TNB currently uses SMARLINE system to provide the Heat loss = Heat gain (2)
line rating, the Lindsey sensors data (i.e. line temperature,
line sag or ground clearance, and load) at the selected TNB qc + qr = qs + qj (3)
line-span is transmitted to SMARTLINE cloud server via
satellite communication [7]. Meanwhile, this system uses the where qc and qr is the cooling due to convection and due to
outsource weather service provider which is Darksky to radiation of surrounding. While qs is the solar radiation
calculate the current rating (IDLR) and provide the information heating and qj is Joule heating due to current flows in the
(i.e. calculated DLR and weather data) to TNB through its conductor [8-11].
web portal. The system overview is presented in Fig. 1 and Besides, the line-monitoring sensor (Lindsey) also
the final equation is expressed as in (1). measures the line clearance to ground/line sag using LiDAR
sensor that built-in with the Lindsey sensor as depicted in Fig.
IDLR(t) = TC,IEEE 738(t) + IDELTA(t) (1) 2. It is important to identify the relationship between the line
sag and the conductor temperature, in order to calculate the
In detail, this system compares the conductor temperature dynamic rating for the line [12] [13].
measured by the Lindsey sensor from actual load with the line III. DATA SELECTION
temperature using IEEE standard, TC,IEEE 738 to get the final
ampacity adjustment factor, IDELTA. Based on IEEE standard Instead of using transmission line monitoring sensor
738 [8], the conductor temperatures depends on: (Lindsey sensor) and weather data to monitor the DLR data,
this project aims to calculate the line rating for a specific
• Material properties of the conductor conductor type using the weather parameters. Thus, it will
• Conductor diameter determine the DLR value without the monitoring data from
installed DLR sensor at transmission line. As this is a sequent
• Conductor surface condition project from the SMARTLINE DLR deployment, no weather
station has been installed nearby the DLR sensor. Therefore,
• Weather conditions (solar heating, wind speed, and it is sufficient to use the same weather data source as the
direction, and air temperature) current system, which is Dark Sky, to calculate the line rating
• Electrical current in the conductor for TNB Grid. However, it is important to perform a thorough
study on the data accuracy, based on the data comparison
between different weather providers before performing the
calculation.

TABLE I. SIX WEATHER SERVICE PROVIDERS

No. Weather provider Website link
1 AccuWeather https://www.accuweather.com

2 The Weather https://weather.com

3 Ventusky https://www.ventusky.com

4 Weather Underground https://www.wunderground.com

5 Dark Sky https://darksky.net

6 Weather Forecast https://www.weather-forecast.com

Fig. 1. The overview of SMARTLINE system from the measuring line
data using Lindsey sensor and weather data to the calculated DLR data,
which is accessible via web portal.

114

Authorized licensed use limited to: California State University Fresno. Downloaded on June 19,2021 at 23:19:07 UTC from IEEE Xplore. Restrictions apply.
A. Weather Data
There are several accessible weather data providers but it
needs a detailed study on the data comparison between
different service providers to obtain better accuracy on
weather data. Six different weather services have been
selected as in Table 1. A series of observations have been
conducted to compare and to record the difference between
each of the selected weather services. All of the observation
on weather parameters has been done to a fixed location.
1) Ambient Temperature, oC
A one-week data observation has been conducted to
compare the ambient temperature between different weather
providers. However, only one-day temperature data is shown
in Fig. 3 due to the repetitive trend. This figure shows an
almost similar temperature trend between the weather
providers but it is clear that temperature data from Data 1 is
quite higher especially during daytime than other weather
providers. Meanwhile, Dark Sky, which is used in conjunction Fig. 4. Davis portable weather station; the overall view of Davis set up at
the office area for local weather data monitoring (left), and the clear view
with SMARTLINE system, provides almost continuous
of the weather sensor (right)
weather data with similar accuracy among the other weather
providers. The surrounding objects such as a tower, building, or even
To verify the accuracy of the weather service Dark Sky, a a tree will highly influence the measuring data of the weather
portable weather station as shown in Fig. 4 is used to capture station. For the height factor, the wind speed and the direction
the weather data, which is then set as a local reference for are recorded from ground level up to several meters of the
further comparison. The assessment is conducted at the TNB device itself. It means that the weather station is limited to
research office and the brand of the portable weather station is monitor the wind parameters within local area coverage only.
so-called Davis. Meanwhile, for weather forecast services, the wind speed and
This assessment is conducted for a month starting from direction are monitored from the satellite image. Cloud
March to April 2020, but in this report, only one-day data coverage and movement are applied to predict the wind speed
observation on 26 March 202 is discussed in presenting the and direction. It is indeed that for the wind parameters, the
overall finding. In detail, the polling rate data is every 30- weather forecast service is an excellent choice for the job.
minutes starting from 12.30 am to 11.30 pm each day. As the
result in Fig. 5, it depicts that the difference between both data
(Dark Sky and Davis) is slightly small, with the margin of
error is found to be around ±2°C.
2) Wind Speed, km/h, and Wind Angle
For wind speed comparison as presented in Fig. 6, it is
found that some weather providers (Data 1 and Data 4) show
a large difference, especially during the daytime. While Dark
Sky data records a consistent trend as other weather providers.
However, for local comparison on wind speed data, there
is a large difference between Dark Sky data and Davis as
depicted in Fig. 7 and it is due to different geo and ecosystem
of both data. The portable weather station (Davis) is able to Fig. 5. The comparison on Dark Sky ambient temperature data (orange
star-line) and local weather monitoring data (blue dotted line)
measure wind speed only at a specific given location it
depends on the equipment height and the surrounding
conditions.

Fig. 6. The comparison of wind speed data between six different weather
providers. Dark Sky is presented in grey line.
Fig. 3. The comparison of ambient temperature data between six different
weather providers. Dark Sky data is presented in grey line.

115

Authorized licensed use limited to: California State University Fresno. Downloaded on June 19,2021 at 23:19:07 UTC from IEEE Xplore. Restrictions apply.
 = . . . − (6)

= . . . (7)

where ε is the emissivity of conductor surface, Tc is the

conductor temperature, and Ta is the ambient temperature.
While for heat loss due to convection (qc), the thermal air
conductivity, Nusselt number, and conductor temperature to
ambient temperature-dependent are defined as λf, Nu, and θ
respectively [1, 8, 9, 14]. In general, the calculation on qc is
determined from the ambient conditions and using the
Fig. 7. The comparison of wind speed data between Dark Sky weather specific conductor type in TNB Grid, which is ACSR.
provider (orange star-line), and local weather monitoring data (blue dotted
line) With all the variables defined in the thermal balance of
conductor (4)–(7), it is possible to calculate the current flows,
B. Conductor Characteristics I, which also represents the line rating, since other parameters
In TNB grid, the transmission line involves for DLR are obtainable. MATLAB software is used to develop the
installation is 275kV Pantai – Ayer as presented in Fig. 8. The algorithm based on a numerical method for solving the
line spans are selected based on the analysis data of line differential equation. This method is used to estimate the line
loading and generation profile in the area. The line type is rating value, I from the heat balance equation (2).
double circuit line and the conductor type is double conductors
of ACSR – Zebra. The measured parameters using the sensor V. RESULTS AND DISCUSSIONS
at the line span will indicate the line rating and thus, it will
The DLR data comparison between SMARTLINE DLR
then use a DLR comparison between the weather-based
data with installed Sensor 1 (as referred to Fig. 8) and DLR
calculation (the present work) and SMARTLINE data.
weather-based calculation is analysed. The result is shown in
IV. DLR DATA BY WEATHER-BASED CALCULATION Fig. 9. Overall, this figure depicts that the trend of weather-
based DLR follows the DLR SMARTLINE data, where the
The weather-based calculation is developed based on the increasing and decreasing trends are consistent with
energy balance at the transmission line, which is expressed in SMARTLINE system and it is higher than SLR value.
(2) and (3). From these equations, the Joule heating, qj is However, there is a clear difference especially for the spikes
described as in (4). of certain points (as highlighted in Fig. 9). This behaviour can
be related to the gap parameter in the weather-based
qj = Rac(NcI)2 (4) calculation as SMARTLINE system considers the ground
clearance and actual line temperature in the DLR calculation
where Rac corresponds to AC resistance of the line at a [7]. Moreover, higher DLR range means lower ambient
specific temperature, Nc is the number of conductors and I temperature during the period. Thus, it is crucial to monitor
defines the current flows in the conductor. Then, the heat the actual ambient parameters (i.e. wind speed and its
gained from Sun, qs can be further elaborated in (5). direction, temperature, and solar radiation) to get high DLR
accuracy as demonstrated in SMARTLINE system.
qs = αs.D.S (5) To analyse the difference between the DLR data from
SMARTLINE system and weather-based calculation, an error
where αs is absorptivity of conductor surface, D is the outer measurement technique is done, so-called mean average error
diameter of the line, and S is global solar radiation. For the (MAE). Based on the MAE analysis, it yields that the
cooling variables due to radiated heat loss, qr, and convective difference is 7.97A. Then, the data comparison is further
heat loss, qc, the equations are determined as (6) and (7). analysed using normal distribution analysis on SMARTLINE
DLR, and weather-based calculation as in Fig. 10. SLR is also
plotted in the same figure as the reference.

Fig. 9. DLR data comparison between weather-based calculation (in red

line), SMARTLINE system (black dashed line), and SLR (blue dashed line)
Fig. 8. Two installed DLR sensor at 275kV Pantai – Ayer transmission line

116

Authorized licensed use limited to: California State University Fresno. Downloaded on June 19,2021 at 23:19:07 UTC from IEEE Xplore. Restrictions apply.
 REFERENCES
[1] Andrea MICHIORRI, et al., “Dynamic Line Ratings Deployment on
the Orkney Smart Grid”, in 21st International Conference on
Electricity Distribution, CIRED, Frankfurt., 2011.
[2] Kosec, G., M. Maksić, and V. Djurica, “Dynamic thermal rating of
power lines – Model and measurements in rainy condition,”
International Journal of Electrical Power & Energy Systems, vol. 91,
pp. 222-229, 2017.
[3] Kim, S.D. and M.M. Morcos, “An Application of Dynamic Thermal
Line Rating Control System to Up-Rate the Ampacity of Overhead
Transmission Lines,” IEEE Transactions on Power Delivery, vol. 28,
no. 2, pp. 1231-1232, 2013.
Fig. 10. The normal distribution of calculated DLR (red line), [4] N.D. Sadanandan and A.H. Eltom, ‘Power donut system laboratory test
and data analysis,” IEEE Proceedings on Southeastcon, New Orleans,
SMARTLINE DLR (black line), and SLR (blue line) as reference
LA, USA, 1990.
This finding indicates that DLR from weather-based [5] Arne BERGSTRÖM, a. Urban AXELSSON, and V. NEIMANE,
calculation is higher most of the time compared to SLR. In “Dynamic capacity rating for wind cooled overhead lines,” 22nd
International Conference and Exhibition on Electricity Distribution
addition, the normal distribution trends between both (CIRED 2013), Stockholm, Sweden, 2013.
SMARTLINE DLR and weather-based calculation are
[6] Schell, P., et al., “Dynamic prediction of energy delivery capacity of
consistent except for the maximum DLR value. The maximum power networks: Unlocking the value of real-time measurements,”
DLR (weather-based) value is 2217A, which is 783A greater 2012 IEEE PES Innovative Smart Grid Technologies (ISGT), pp. 1-6,
than SLR and 70A lower than SMARTLINE DLR. These 2012.
present findings indicate that with the difference with the [7] Robert Aaserude, Walnut Creek, and R. Bliss, “Systems and Methods
SMARTLINE DLR data, the DLR value from the weather- for Calculating Power Transmission Line Capacity,” I. Dynamic
based calculation is reliable since the maximum rating is Engineers, Dynamic Engineers, Inc. US, 2016.
smaller than the existing system rating. It means that the [8] Society, I.P.a.E., “IEEE Standard for Calculating the Current-
Temperature Relationship of Bare Overhead Conductors,” IEEE Std
calculated DLR can indicate the current rating without causing 738™-2012., New York, USA, 2012.
damaging effect on the line performance. [9] CIGRE, “Guide for Thermal Rating Calculations of Overhead Lines,
in Considerations for High Temperatures and Changing Weather and
VI. CONCLUSION AND RECOMMENDATION Load Data,” CIGRE. 2014.
This project is undertaken to develop a DLR calculation [10] Morozovska, K. and P. Hilber, “Study of the Monitoring Systems for
based on weather parameters and to evaluate its accuracy by Dynamic Line Rating,” Energy Procedia, vol. 105, pp. 2557-2562,
2017.
comparing the calculated DLR data with the DLR of existing
system. The findings suggest that in general, the DLR value [11] Saied, M.M., “Assessing the dynamic rating of overhead transmission
lines,” European Transactions on Electrical Power, vol. 17, no. 5, pp.
from the weather-based calculation is reliable to use by 526-536, 2007.
showing the consistent data trend with the existing system [12] Karimi, S., P. Musilek, and A.M. Knight, “Dynamic thermal rating of
(SMARTLINE). At the same time, the calculated DLR value transmission lines: A review,” Renewable and Sustainable Energy
can help in increasing the line rating of 10 – 50% from SLR Reviews, vol. 91, pp. 600-612, 2018.
without causing damage to the line performance based on the [13] Black, C.R. and W.A. Chisholm, “Key Considerations for the Selection
normal distribution trend. of Dynamic Thermal Line Rating Systems,” IEEE Transactions on
Power Delivery, vol. 30, no. 5, pp. 2154-2162, 2015.
It is indeed that there is a clear difference between [14] Greenwood, D.M., et al., “A Comparison of Real-Time Thermal Rating
calculated DLR data and SMARTLINE DLR data, but this Systems in the U.S. and the U.K,” IEEE Transactions on Power
present work is a good initiative for future algorithm Delivery, vol. 29, no. 4, pp. 1849-185, 2014.
development in indicating the DLR value for line optimisation [15] D. J. Spoor, a. and J.P. Roberts, “Development and Experimental
use in TNB Grid without installing the line sensor. Hence, Validation of a Weather-Based Dynamic Line Rating System,” 2011
IEEE PES Innovative Smart Grid Technologies, Perth, WA, Australia,
further research needs to be conducted especially in finding 2011.
the best alternative to monitor the weather parameters without [16] Bhattarai, B.P., et al., “Improvement of Transmission Line Ampacity
dependency on a single weather service provider as been Utilization by Weather-Based Dynamic Line Rating,” IEEE
highlighted by previous research [15, 16]. This will assist to Transactions on Power Delivery, vol. 33, no. 4, pp. 1853-1863, 2018.
establish a high DLR data accuracy as other established DLR
systems.

117

Authorized licensed use limited to: California State University Fresno. Downloaded on June 19,2021 at 23:19:07 UTC from IEEE Xplore. Restrictions apply.