Load Factor and Unserved

Energy Webinar

October 25, 2019

Barbara O’Neill, NREL, US DOE

James Elsworth, NREL, US DOE
Todd Levin, ANL, US DOE
Travis Douville, PNNL, US DOE
Juliet Homer, PNNL, US DOE
Agenda

1. Fundamentals, macroeconomics, primary drivers

2. Measures to influence peak and future trends
3. Unserved energy
4. International comparisons and general observations
-Country case studies: India, Afghanistan, Sri Lanka,
Bangladesh, Nepal, Denmark
5. Load Forecasting in Pakistan

2
Fundamentals
𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 (𝑀𝑀𝑀𝑀𝑀)
Definition of load factor: 𝐿𝐿. 𝐹𝐹. (%) =
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜

Northwest Power Pool USA Load Factor

(actual 2015-2018, forecast 2019-2029)
2018 Example of Northwest Power Pool 74.0%
in USA 72.0%
(parts of states Washington, Oregon, Idaho,
Montana, Nevada, Utah, Wyoming) 70.0%

68.0%

66.0%

288,002,000 (𝑀𝑀𝑀𝑀𝑀) 64.0%

66.8 % = 62.0%
49,235 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
60.0%

3
Macro Trends Affecting Load Factor
Traditional variables used to predict electricity consumption
- Gross domestic product
- Weather factors (heating degree days, cooling degree
days, seasonal differences)
- Tariffs, lagged
- Individual customer sector growth
- Population growth
- Employment or other labor statistic
- Household size

4
Macro Trends Affecting Load Factor

Historically electricity intensity (MWh/GDP) increased as an

economy developed
Third world  Second world  First world
Agricultural  Industrial  Service economy
economy economy

Factors that may dampen the traditional relationship

- Distributed generation (as inhabitants get wealthier, more likely
to purchase rooftop solar and possibly storage)
- Energy efficient technologies are more ubiquitous
- Smart technology making customers more price responsive 5
Primary Drivers of Peak Load
Residential
- Air conditioning, heating,
television (Sri Lanka peaks
during cricket matches)
Commercial
- Business hours
Industrial
- Seasonal production
Source: Peak Electricity Demand Prediction Model for Sri Lanka

- Single, double, or triple shift Power System (G.V.Buddhika et al. De Silva and Lalith A.
Samaliarachchi)

work days
6
Measures to Influence Peak
California Modeled Load for Increased PV Penetration
- Industrial
demand
response
- Time Of Use
pricing
- Smart meters
- Energy
efficiency
- Electric vehicle
policy
Source: Denholm, P., Overgeneration from Solar Energy in California – A
Field Guide to the Duck Chart, https://www.nrel.gov/docs/fy16osti/65453.pdf 7
 Future Trends
Electrification
• Transport sector, i.e., electric vehicles
• Hot water heaters (natural gas, electric, on-demand)
• Heating (natural gas furnace, electric baseboard heating,
directed space heating)
• Increased plug load, yet more efficient appliances
(e.g., computers, phones, electronic devices)

Other considerations
• Individual market players both generate and consume power (e.g., blockchain
technology)
• Storage technology use increases
• Players intentionally maximize the economics of doing so
• Net load daily and seasonal peak that the utility sees will change.
8
Unserved Energy Categories Not straightforward

𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 (𝑀𝑀𝑀𝑀𝑀)
𝐿𝐿. 𝐹𝐹. (%) =
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
1. Not connected to grid

Source: Hodge, Bri-Mathias, Future Use of Big Data in the 9

Smart Grid, https://www.nrel.gov/docs/fy16osti/66326.pdf
Unserved Energy Categories Not straightforward

𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 (𝑀𝑀𝑀𝑀𝑀)
𝐿𝐿. 𝐹𝐹. (%) =
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
1. Not connected to grid
2. Self-generation (captive demand)

Source: Hodge, Bri-Mathias, Future Use of Big Data in the 10

Smart Grid, https://www.nrel.gov/docs/fy16osti/66326.pdf
Unserved Energy Categories Not straightforward

𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 (𝑀𝑀𝑀𝑀𝑀)
𝐿𝐿. 𝐹𝐹. (%) =
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
1. Not connected to grid
2. Self-generation (captive demand)
3. Load shed

Source: Hodge, Bri-Mathias, Future Use of Big Data in the 11

Smart Grid, https://www.nrel.gov/docs/fy16osti/66326.pdf
Unserved Energy Categories Not straightforward

𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸𝐸 (𝑀𝑀𝑀𝑀𝑀)
𝐿𝐿. 𝐹𝐹. (%) =
𝑃𝑃𝑃𝑃𝑃𝑃𝑃𝑃 𝑀𝑀𝑀𝑀 𝑥𝑥 8760 ℎ𝑜𝑜𝑜𝑜𝑜𝑜𝑜𝑜
1. Not connected to grid
2. Self-generation (captive demand)
3. Load shed
4. Emergent demand:
increased demand due
to increased grid
reliability/availability

Source: Hodge, Bri-Mathias, Future Use of Big Data in the 12

Smart Grid, https://www.nrel.gov/docs/fy16osti/66326.pdf
Comparison against current load factors in other countries
Average Cold Average Hot
Country Load Factor
Season Temp (°C) Season Temp (°C)
Pakistan (Islamabad) 61.0%* 13 29
India (New Delhi) 71.5% 17 39
Afghanistan (Kabul) 60.0% 1 24
Nepal (Kathmandu) 56.9% 12 23
Sri Lanka (Colombo) 60.8% 26 28
Germany (Berlin) 86.2% 1 17
Turkey (Istanbul) 70.7% 7 22
France (Berlin) 62.2% 4 18
Denmark (Copenhagen) 61.4% 1 16
Italy (Rome) 62.1% 9 23

*2019 NTDC Regression modeling 13

https://en.climate-data.org
Country Case Studies

- Difficult to compare load factors across countries

- Apples and oranges for energy parameter
- A. gross generation
- B. net generation after house power
- C. net energy after imports and exports
- D. energy sales after transmission and distribution losses
- E. energy sales after non-technical losses

14
 Example of Varying Energy Parameter
Puerto Rico Generation and Peak Load Load Factor
1
35000

0.9
30000
0.8

25000 0.7

0.6
20000

0.5
15000
0.4

10000 0.3

0.2
5000
0.1
0
2007 2009 2011 2013 2015 2017 0
2007 2009 2011 2013 2015 2017
Sales Gross Generation
Net Generation Peak demand for 8760 h Load Factor - Sales Load Factor - Gross Gen Load Factor - Net

15

Source data from Puerto Rico Integrated Resource Plan 2018-19 by Siemens PTI for Puerto Rico Electric Power Authority
Country Case Studies
Load is forecast using two different methods
I. Load factor as a result
A. Use regression (or other method) to forecast energy
B. Use regression (or other method) to forecast peak
C. Load factor time series is a calculated result

II. Load factor forecast as a driver

A. Use regression (or other method) to forecast energy
B. Explicitly forecast load factor
i. Hold constant or
ii. Show improvement, i.e., increase or higher capital use efficiency
C. Peak is a calculated result

• However, in reality load factor usually changes over time depending on drivers
discussed previously e.g., air conditioning use, tariff structure, economic
development curve 16
Specific Country Examples

17
 Case Study 1 of India
National Electricity Plan based on 2011, projected to 2037
• Used a Partial End Use Method (PEUM) for Energy & Demand
• “bottom-up” approach focusing on end uses
• Based on past demand + projected policies, development,
energy efficiency
• Done on a state-by-state basis
• Included stakeholders from across the power sector
• Considerations
• Category-wise energy and load, T&D losses, population, #
consumers, # irrigation pumps & capacity, electrification, railway
expansion,
• Policies – reduction of T&D losses, EE & DSM programs, Power
for All Initiative, Dedicated freight corridor, make in India
• Rooftop Solar Prog. – 100 GW solar by 2022, 40 GW rooftop
• EVs – 6 mill by 2020, charging schedule won’t increase demand
18

Source: National Electricity Plan (Volume 1), Generation, Government of India Ministry of Power, Central Electricity Authority
Case Study 1 of India
3000000
Forecasted Energy Consumption and Peak 100%
Demand Consumption in India
90%
2500000
80%

70%
2000000
60%
GWH

1500000 50%

40%
1000000
30%
Peak Load for 8760 hours (GWh)
20%
500000 Energy use forecast (GWh)
Load factor (right axis) 10%

0 0%
2011 2013 2015 2017 2019 2021 2023 2025

19
Sources: India CEA National Electricity Plan, https://community.data.gov.in/gross-electricity-generation-in-india-from-2010-11-to-2016-17/
Case Study 2 of Afghanistan
Power Sector Master Plan, analysis period 2011-2032
• 3 consumer types – residential, commercial/industrial, public
sector
• 34 provinces each with a demand forecast (23 with no or low
elect.)
• Base, Low, and High forecasts
• Simplified econometric approach, based on estimates and
assessments from similar countries (economically) considers:
• Income development – available income (res), GDP (comm/ind & public)
• Electricity price development
Can combine these to calculate elasticity of electricity demand
• Other considerations: reductions in T&D losses, no DSM or EE
• Future development of load calculated from load factor
20
Source: Islamic Republic of Afghanistan: Power Sector Master Plan, https://www.adb.org/sites/default/files/project-document/76570/43497-012-afg-tacr.pdf
Case Study 2 of Afghanistan

• Assumptions for load factor forecasting

21
 Case Study 2 of Afghanistan

Load Factor
2012 2015 2020 2025 2030
55% 56% 58% 59% 60%
22
Source: Islamic Republic of Afghanistan: Power Sector Master Plan, https://www.adb.org/sites/default/files/project-document/76570/43497-012-afg-tacr.pdf
 Case Study 3 of Sri Lanka
Load forecast considerations
• Historical energy and peak consumption growth rates
• Historical load factor
• Decreasing trend in losses
• Future impact of national energy policies
• Future developments in transport, urban growth, tourism, industry,
agriculture, manufacturing
• Excluded special days – holidays, etc.

Methodology
• Forecasted electricity consumption
• Forecasted load factor
• Calculated forecasted demand

23

Source: Ceylon Electricity Board Long Term Generation Expansion Plan, 2018
 Sri Lanka System Load Factor

24

Source: Ceylon Electricity Board Long Term Generation Expansion Plan, 2018
Sri Lanka Forecast Scenarios

25
Source: Ceylon Electricity Board Long Term Generation Expansion Plan, 2018
 Sri Lanka Forecast of Flattening Daily Curve

26
Source: Ceylon Electricity Board Long Term Generation Expansion Plan, 2018
Case Study 4 of Bangladesh
- Peak demand forecast using
- GDP elasticity method
- Sectorial analysis method
- Consistency confirmed
- Integrated both projections to hybrid approach
- Included theoretical estimation of load curves considering
historic black-outs and energy not served
- Composite daily load curve estimated from summer and winter
actual values
- Exogenous forecast of peak demand reduction relative to GDP
- Daily load curve estimated using assumed load factor

Source: Power Sector Master Plan 2016, supported by Japan International Cooperation Agency and Tokyo Electric Power Company for Power
27
Division of Ministry of Power, Energy, and Natural Resources of the Government of the People’s Republic of Bangladesh
28
Case Study 6 of Nepal

29
 Case Study 6 of Nepal
Nepal Base Case and Optimistic Case Electricity Forecasts
70000 80%

60000 70%

60%
50000

50%
40000
40%
30000
30%

20000
20%

10000 10%

0 0%
2013 2015 2017 2019 2021 2023 2025 2027 2029 2031 2033

Optimistic Case System Peak Load for 8760 h (GWh) Base Case System Peak Load for 8760 h (GWh)
Optimistic Case Total Energy Demand (GWh) Base Case Total Energy Demand (GWh)
Optimistic Case Load Factor (%) Base Case Load Factor
30

Source data: Nepal Electricity Authority Load Forecast Report

 Case Study 7 of Denmark

Demand forecasting
• Biggest factor:
increase in data
center energy
consumption (65%
of demand increase,
16.7% of total
energy by 2030)
• EVs and heat pumps
are also significant
31
Source: Ostergaard, Andersen, and Kwon, “Energy Systems Scenario Modelling and Long Term Forecasting of
Hourly Electricity Demand”. International Journal of Sustainable Energy Planning and Management, Vol. 07 2015 99-116
Case Study 7 of Denmark

2012 2020 2030

Demand (GWh) 30922 31783 33512
Peak Load (GW) 5.72 5.8 6.04
Load Factor (%) 62% 63% 63%
Source: Ostergaard, Andersen, and Kwon, “Energy Systems Scenario Modelling
and Long Term Forecasting of 32
Hourly Electricity Demand”. International Journal of Sustainable Energy Planning
and Management, Vol. 07 2015 99-116
Load Forecasting in Pakistan

33
 Load Forecasting in Pakistan

• NTDC provides the primary load forecast that is used

within the Government of Pakistan
• A full report outlining the regression methodology and results were
published in 2014
• This process was updated in 2019
• Data from the 2019 analysis are utilized in NTDC’s Indicative
Generation Capacity Expansion Plan (IGCEP)
• The methodology is briefly reviewed in the IGCEP

Indicative Generation Capacity Expansion Plan 2018-40, NTDC February 2019. 34

https://www.ntdc.com.pk/ntdc/public/uploads/downloads/NTDC.PSP.IGCEP.pdf
NTDC 2019 Load Forecasting
• Ordinary Least Squares regression model based on historical data
Sectors Independent variables
• Domestic • Annual GDP (by sector)
• Commercial • Electricity tariff (by sector)
• Industrial • Number of consumers
• Agriculture • Population trend
• Lag of dependent variables
• Consumer Price Index
• Dummy variables
Y T = YT-1 * (1+GR of G)b
*(1+GR of R)C
* (1+GR of L) d
35
NTDC 2019 Load Forecasting
• Regression accounts for historical load management
• Projections are for actual demand
• Regardless of whether or not that demand can be served by the power system
• The total annual electricity demand (GWh) is projected for each sector
of the economy
• Through regression analysis
• National generation requirements are then determined by considering
projected T&D losses
• Method for projecting future changes in T&D losses is not outlined in detail
• Peak demand (MW) is then obtained by applying a projected load
factor
• Method for forecasting load factor is not outlined in detail
• Load factor depends on many system interactions and can be very challenging to
forecast

36
 NTDC 2014 Regression Study
• Assumed reduction in load factor from 69.1% to 68.1%
• PEPCO only
• Assumed reduction in T&D losses
• 265kV+: Decrease from 2.92% (2012) to 2.52% (2016 and beyond)
• 132 kV: Decrease from 2.10% (2012) to 2.06% (2016 and beyond)
• Distribution: Decrease from 14.52% (2012) to 10.0% (2021 and beyond)
• Considered 3 GDP growth scenarios
• Low: 3.56% (IMF value for 2012-13)
• Normal: 5% (based on last 30 years)
• High: 6.5% (goal)
• Results are presented for six subsets of country demand
1. Sectoral sales for PEPCO and KESC
2. Load forecast for PEPCO excluding export to KESC
3. Load forecast for KESC
4. Load forecast for PEPCO + KESC + Self-generation
5. Load forecast for PEPCO + KESC + Self-generation with Demand Side Management
37
“Electricity Demand Forecast Based on Multiple Regression Analysis: Period 2014 to 2037”, NTDC, March 2014.
 NTDC 2014 Regression Results
PEPCO excluding exports to KESC – Normal Growth Projection

Base Year Energy Sale (GWh) Energy Generated (GWh) Load Factor (%) Peak Demand (MW)
2012-13 (Recorded) 70,508 90,893 82.8 12,527
2012-13 (Computed) 89,535 116,611 69.1 19,265
2013-14 94,222 120,469 68.1 20,195
2014-15 99,537 126,296 68.1 21,171

2015-16 105,418 132,747 68.1 22,253

2016-17 111,741 139,655 68.1 23,411
2017-18 118,829 147,591 68.1 24,741
2018-19 126,651 156,335 68.1 26,207
2019-20 135,005 165,624 68.1 27,764
2020-21 144,283 175,928 68.1 29,492
2021-22 154,497 187,241 68.1 31,388
2022-23 165,568 200,658 68.1 33,637

Recorded Demand represents the load that was actually served in the base year
Computed Demand represents the load that would have been served without load shedding
Auxiliary Consumption is electricity consumed by IPPs
• Electricity demand is projected to grow faster than GDP
• Increasing from 5% to 9% per year by 2036, vs. assumed constant 5% per year GDP growth
• Load factor projected to remain constant at 68.1%
• Projected load factor in KESC is greater (~80-90%) 38
• The peak generation shortfall (computed – recorded) in 2012-13 was roughly 6750 MW
Source: “Electricity Demand Forecast Based on Multiple Regression Analysis: Period 2014 to 2037”, NTDC, March 2014.
 NTDC 2014 Regression Study

39

*PEPCO Only
 NTDC 2019 Regression Results
Low Normal High
Fiscal Year Gen (GWh) Peak (MW) Gen (GWh) Peak (MW) Gen (GWh) Peak (MW) Load Factor
2017-18* 120,791 26,741 120,791 26,741 120,791 26,741 0.516
2018-19 144,665 27,072 145,674 27,261 147,188 27,545 0.610
2019-20 151,062 27,814 152,914 28,155 155,718 28,671 0.620
2020-21 158,842 28,782 161,841 29,325 166,429 30,157 0.630
2021-22 166,267 30,127 170,645 30,921 177,416 32,147 0.630
2022-23 173,178 30,889 179,142 31,953 188,476 33,618 0.640
2023-24 181,051 32,294 188,914 33,696 201,374 35,919 0.640
2024-25 188,749 33,640 198,744 35,422 214,788 38,281 0.641
2025-26 193,948 34,062 206,155 36,206 226,011 39,693 0.650
2026-27 202,763 35,610 217,664 38,227 242,228 42,541 0.650
2027-28 211,718 37,183 229,603 40,324 259,492 45,573 0.650
2028-29 220,940 38,802 242,104 42,519 277,960 48,816 0.650
2029-30 231,142 40,594 255,989 44,958 298,670 52,453 0.650
2030-31 241,889 42,481 270,792 47,557 321,130 56,398 0.650
2031-32 253,101 44,451 286,441 50,306 345,315 60,645 0.650
2032-33 265,289 46,591 303,554 53,311 372,070 65,344 0.650
2033-34 278,069 48,835 321,710 56,500 400,940 70,414 0.650
2034-35 291,403 51,177 340,888 59,868 431,973 75,865 0.650
2035-36 305,685 53,686 361,590 63,504 465,916 81,826 0.650
2036-37 320,652 56,314 383,529 67,357 502,479 88,247 0.650
2037-38 336,293 59,061 406,719 71,429 541,779 95,149 0.650
2038-39 352,917 61980 431,584 75,796 584,509 102,653 0.650
2039-40 370,348 65,042 457,939 80,425 630,529 110,736 0.650
• 2017-18 data are recorded, 2018-19 onward are computed
• Recorded load factor in 2017 was 51.6%
• Load factor projected to grow from 61% to 65% 40
• Data provided in the IGCEP Report Table 5-2
 NTDC 2019 Regression Study

41

*Data provided in the IGCEP Report Table 5-2

 World Bank*: Review and Contributions
• World Bank reviewed a number of past internal NTDC forecasts
• Compared forecasts to actual realized demand in future years
• Goal to determine if they are suitable and prioritize methodological
improvements
• NTDC conducts regular load forecasting driven by regression analysis
• Long Term (20+ years)
• Top-down macroeconomic regression analysis
• Base years: 10/11, 12/13, 15/16 and 16/17

*Funding provided by the Energy Sector Management Assistance Program (ESMAP) 42

Source: http://pubdocs.worldbank.org/en/871161554646154248/pdf/VRE-Study-Stakeholder-Workshop-Tasks-1-2-Slides-04Apr2019.pdf
 Regression Analysis

• Overall very accurate, within +/- 5%

43
Source: http://pubdocs.worldbank.org/en/871161554646154248/pdf/VRE-Study-Stakeholder-Workshop-Tasks-1-2-Slides-04Apr2019.pdf
 World Bank Findings and Recommendations

“The NTDC forecasts provide very accurate results to

forecast future electricity demand for the NTDC system on a
national level.”

44
Source: http://pubdocs.worldbank.org/en/871161554646154248/pdf/VRE-Study-Stakeholder-Workshop-Tasks-1-2-Slides-04Apr2019.pdf
 2014 vs 2019 Peak Demand Projections

• Projected demand growth declined between the 2014 and 2019 regression studies 45
• Lahmeyer revised the 2019 NTDC regression forecast moderately upwards
 JICA: 2016 Least-cost Expansion Plan

• Energy intensities are estimated based on past sectoral consumption

• Consumption per unit GDP (or per capita for residential)
• Base case: Assumes 5% GDP growth - 5.8% electricity demand growth rate
• High case: Assumes 6.5% GDP growth - 7.0% electricity demand growth
• Project constant future load factors for three regions
46
• North (57.5%) South (61.0%) and Karachi (62.0%)
Source: “Project for Least Cost Power Generation and Transmission Expansion Plan”, JICA and IIEP, 2016.
University Demand Projections
At least four academic studies have projected electricity demand in
Pakistan, three using LEAP and one using TIMES

1. Gul and Qureshi 2012 (LEAP)

2. Perwez et al. 2015 (LEAP)
3. Ishaque 2017 (LEAP)
4. Valasai et al. 2017 (TIMES)

• Typically directly project total demand

• Calculate peak demand based on exogenous load factor

47
Comparison of University Peak Projections

48
• Comparable to 2019 NTDC high projections
Comparison of University Energy Projections

49
• Comparable to 2019 NTDC high projections
Comparison of all Projections

50
Comparison of all Projections

51
Conclusions
• Most projections are for total demand, based on macro-economic indicators
• GDP growth
• Population growth
• Electricity tariffs
• Often converted to peak demand by applying a load factor
• Load factor is hard to predict with a top-down economic model
• Need to be careful with definitions when making comparisons
• Recorded demand vs. Computed Demand
• Sales vs. Generation (T&D losses)
• NTDC methodologies have been externally validated
• Demand projections are still always limited by macroeconomic projections
• No single demand projection is ever perfectly correct
• All projections are based on assumptions
• It is important to draw upon a number of studies
• Help to understand how different factors may influence future demand

52
 Contact:
Questions?
[email protected]
[email protected]
[email protected]
[email protected]

Photo by Dennis Schroeder / NREL

53
Citations
“Electricity Demand Forecast Based on Multiple Regression Analysis: Period 2014 to 2037”, National Transmission and Despatch Company, 2014.
Gul, M., Qureshi, W.A., 2012. Modeling diversified electricity generation scenarios for Pakistan, in: 2012 IEEE Power and Energy Society General
Meeting. Presented at the 2012 IEEE Power and Energy Society General Meeting, pp. 1–7. https://doi.org/10.1109/PESGM.2012.6344821
“Indicative Generation Capacity Expansion Plan 2018-40, National Transmission and Dispatch Company (NTDC) 2019.
https://www.ntdc.com.pk/ntdc/public/uploads/downloads/NTDC.PSP.IGCEP.pdf
Ishaque, H., 2017. Is it wise to compromise renewable energy future for the sake of expediency? An analysis of Pakistan’s long-term electricity
generation pathways. Energy Strategy Reviews 17, 6–18. https://doi.org/10.1016/j.esr.2017.05.002
Peak Electricity Demand Prediction Model for Sri Lanka Power System (G.V.Buddhika De Silva and Lalith A. Samaliarachchi)
Perwez, U., Sohail, A., Hassan, S.F., Zia, U., 2015. The long-term forecast of Pakistan’s electricity supply and demand: An application of long range
energy alternatives planning. Energy 93, 2423–2435. https://doi.org/10.1016/j.energy.2015.10.103
“Project for Least Cost Power Generation and Transmission Expansion Plan”, Japan International Cooperation Agency (JICA) and International
Institute of Electric Power, Ltd. (IIEP), 2016.
Valasai, G.D., Mirjat, N.H., Uqaili, M.A., Memon, H.U.R., Samoo, S.R., Harijan, K., 2017. Decarbonization of Electricity Sector of Pakistan—An
Application of Times Energy Model. Journal of Clean Energy Technologies 5, 507–511. https://doi.org/10.18178/JOCET.2017.5.6.425
“Variable Renewable Energy Integration and Planning Task 2: Demand & Generation Forecast Analysis”, Lahmeyer International, Report Final
DRAFT, 2019.

NREL/PR-7A40-75311

This work was authored, in part, by the National Renewable Energy Laboratory (NREL), operated by Alliance for Sustainable Energy,
LLC, for the U.S. Department of Energy (DOE) under Contract No. DE-AC36-08GO28308. Funding provided by the United States Agency
for International Development (USAID) under Contract No. IAG-17-2050. The views expressed in this report do not necessarily represent
the views of the DOE or the U.S. Government, or any agency thereof, including USAID. The U.S. Government retains and the publisher,
by accepting the article for publication, acknowledges that the U.S. Government retains a nonexclusive, paid-up, irrevocable, worldwide 54
license to publish or reproduce the published form of this work, or allow others to do so, for U.S. Government purposes.