long‐termdemandforecasting term demand forecasting power ... · bihar mp up 0 500 per capita cons...
TRANSCRIPT
1
Long‐term Demand Forecasting
11th Capacity Building Programme for Officers of Electricity Regulatory Commissions
Long term Demand Forecasting
&
Power Procurement Planning
for Distribution Utilities
1
for Distribution Utilities
(Uttar Pradesh) Dr. Anoop Singh
Dept. of IME, IIT Kanpur1
Need for Demand Forecasting &Power Procurement Planning (PPP)
• Long‐term Forecast – Planning for capacity addition/power procurementg p y /p p– Upgrade transmission facilities
• Medium‐term Forecast– Planning for power procurement– Design tariff structure– Demand‐side management– Time of use pricing
2
p g
• Short‐term Forecast– Merit order dispatch– Minimising deviation from schedule – Decision making for short‐term power procurement– Optimising use of renewable energy
2
2
Objectives*
• Demand projection for the power requirement in Uttar Pradesh
h l bl d d d f• Assess the available, proposed and expected power procurement fromconventional and renewable energy sources
• Optimize the power procurement to meet future peak load & energyrequirement
• Develop a power procurement scenario with a mix of long‐term, medium‐t PPA d h t t t
3
term PPA and short‐term power procurement
* ‐ The study was conducted on behalf of UPPCL so as to enable the utilities to meet their obligations underPower for All programme.
3
Trend • Study the past growth pattern
Methodology – Four Stages
1. Projection of electrical energy demand1. Projection of electrical energy demand
• Study category‐wise connected load, electricity consumption and growth pattern
End Use method
• Forecast considering economic change
Econometric Models
AnalysisStudy the past growth pattern energy demand
2. Load profile analysis & projection
3. GAMS based optimisation
4
•Inference from historical load profile
•Account for projected solar addition and DSM
•Account for demand profile influenced by supply
2. Load profile analysis & projectionmodel
4. Choosing power procurement strategy
4
3
• Develop optimisation model considering
3. GAMS based optimisation model
Methodology (contd…)
1. Existing & candidate plants’ quantum and prices
2. Projected solar capacity addition and DSM activities
3. With and without short‐term procurement
4. Generator & contract specific constraints
• Defining different power procurement scenarios
5
• Estimation of social cost and utility cost in different economic and power procurement scenarios
• Identify optimal power procurement strategy
4. Choosing power procurement strategy
5
Projection of Electrical Energy Demand
6 6
4
90%
100%
Electricity Demand Pattern Across States
1200
Per Capita Electricity ConsumptionCategory‐wise Electricity Consumption (2014‐15)
50%
10%
8%
28%
10% 11% 5%12%
7%
11%
7%
3%
3%
11%
15%5% 3% 3%
8%
6%
9%4%
16%
37% 38%
27%25%
18%
25%
23%
28%21%
3%
18% 22%
30%33%
36%
25%38%
16%18%
30%
40%
50%
60%
70%
80%
90%
672717 734
779819
884 914957
1010
346 372 387 412450 450 472 502400
600
800
1000
kWh
7
41%
27%
50%
17% 21% 19%24% 23% 20%
28%39%5%
11% 7%
0%
10%
20%
Domestic Commercial Industrial (LV & MV)Industrial (HV) Agriculture So: CEA (Gen. Review 2016)
341 346 372
0
200
India Uttar Pradesh
So: CEA (Gen. Review 2006‐077 to 2014‐15) 7
Electricity Consumption vs Economic Activity
ChhattisgarhDelhi
GujaratHaryana
Karnataka MaharashtraRajasthan
TN
1000
1500
2000
umption
kWh)
Bihar
MPRajasthan
UP
0
500
0 20000 40000 60000 80000 100000 120000
Per Cap
ita Consu
of Electricity (k
Per Capita SGDP (Rs. at 2004 prices)
Elasticity ‐ 0.78
2011‐12450
500
ty
h)
Uttar Pradesh
8
2006‐072007‐08
2008‐092009‐10
2010‐11
2012‐13
300
350
400
450
14000 14500 15000 15500 16000 16500 17000 17500 18000 18500 19000
Per Cap
ita Electricit
Consumption (kWh
Per Capita SGDP (Rs. at 2004 prices)
Elasticity ‐1.16
8
5
Power Sector Scenario in Uttar Pradesh
41.30 %
70 %
%30 00
35.00
40.00
45.00Category‐wise Consumption (in %)
14000160001800020000
nds
Number of Consumers CAGR ‐ 8.55%
7.40 %
22.7
18.30 %
0.00
5.00
10.00
15.00
20.00
25.00
30.00
2010‐11 2011‐12 2012‐13 2013‐14 2014‐15 2015‐16
Domestic Commercial Industrial Agriculture
02000400060008000
100001200014000
2010‐11 2011‐12 2012‐13 2013‐14 2014‐15 2015‐16 2016‐17
Thousan
Source –Data from UPPCL
4500050000
ds
Total Connected Load (kW) CAGR ‐8.04%
70000
Uttar Pradesh Electrical Energy Consumption(in GWh) Utility Only
9
05000
1000015000200002500030000350004000045000
2010‐11 2011‐12 2012‐13 2013‐14 2014‐15 2015‐16 2016‐17
Thousand
Source Data from UPPCL
30000
35000
40000
45000
50000
55000
60000
65000
Source ‐ CEA, General review 2016
CAGR ‐ 8.16%
9
Trend Analysis
• Useful for preliminary estimate for forecast
Forecasting Methods
• Uses time factor, assuming pattern of demand in the past will continue into future
• Does not explain underlying factors– Demographics– Economic factors Consumer behaviour and price sensitivity
1 0
– Consumer behaviour and price sensitivity– Government initiative, tech. development and RE induction
• Easy to use and understand
10
6
Forecasting Methods (Contd…)
End‐use method
d h d l l– Focuses on various end‐uses in the residential, commercial, agriculture and industrial sectors
– Aggregate energy demand is summed over different end‐uses in a sectors
– Effective method when there is lack of adequate past data
1 1
– This method requires a high level of detail on each of the end‐uses
– Does not take in account demographics and socio‐economic factors
11
Forecasting methods (Contd…)
Econometric Analysis
• Utilises information from historical data along with factors that• Utilises information from historical data, along with factors that influence electricity demand
– Demographic factors (urbanisation, number of households & size, connected load etc.)
– Economic factors (GDP/per capita income, industrial growth etc.)
C ti b h i b t
1 2
– Consumption behaviour by consumer category
– Price sensitivity
• Estimate econometric relationship and use it to predict future demand
12
7
Econometric Model
1 3 13
Data & Data Sources • CEA ‐General Review (2003‐04 to 2014‐15)
i. Category‐wise Connected Load, No. of Consumers and Consumption all states
ii. Per capita electricity consumption
iii. Number of pump‐sets energized, Mid year population
• CSO, MOSPI
i. State Gross Domestic Product at constant price (base year 2011)
• PFC Reports
i. Weighted average price of electricity (base year 2011)
• Tariff Orders
i. Power procurement cost
1 4
p
• UPSLDC
i. Unrestricted demand, rostering, emergency rostering, electricity generation etc.
• UPPCL
i. U.P. number of consumers, connected load and consumption category‐wise
ii. PPA Information's and rate of electricity
iii. CS3 & CS4 reports 14
8
Forecasting Scenarios
• High Growth Scenario
• Realistic Growth Scenario
• Medium Growth Scenario
• Low Growth Scenario
1 5 15
Projected Per Capita Electricity Consumption in U.P
10111000
1200
953
854
721
400
600
800
1000
kWh
High Growth
Realistic Growth
Medi mGro th
1 6
0
200
200
3‐04
200
4‐05
200
5‐06
200
6‐07
200
7‐08
200
8‐09
200
9‐10
201
0‐11
201
1‐12
201
2‐13
201
3‐14
201
4‐15
201
5‐16
201
6‐17
201
7‐18
201
8‐19
201
9‐20
202
0‐21
202
1‐22
202
2‐23
202
3‐24
202
4‐25
202
5‐26
202
6‐27
Medium Growth
Low Growth
16
9
Projected Electrical Energy at Bus‐bar for Utility Only
244238
300000
244238
230398206808
175223
50000
100000
150000
200000
250000
GWh
High Growth
Realistic Growth
1 7
0
50000
200
3‐04
200
4‐05
200
5‐06
200
6‐07
200
7‐08
200
8‐09
200
9‐10
201
0‐11
201
1‐12
201
2‐13
201
3‐14
201
4‐15
201
5‐16
201
6‐17
201
7‐18
201
8‐19
201
9‐20
202
0‐21
202
1‐22
202
2‐23
202
3‐24
202
4‐25
202
5‐26
202
6‐27
Medium Growth
Low Growth
17
Results Comparison With Other Reports
Note: For utilities only
Compassion Projected Energy ( 19th EPS vs Estimated Value) GWh
FYCEA Econometric model results (IIT Kanpur)
19 EPS Realistic High Medium Low* Without Captive Generation
Note: Energy sold
19 EPS Realistic High Medium Low
2016‐17 108070 114512 114512 114512 1145122021‐22 150797 163562 166115 153757 142298
2026‐27 195323 227838 244238 206808 175223
Projected Total sales (In MU)FY PFA Econometric Model ∆ %
1 8
Note: Energy sold * Without Captive and losses2016‐17 83789 92882 11%
2017‐18 95131 101267 6%2018‐19 103173 110511 7%2019‐20 116385 120706 4%2020‐21 126046 130958 4%2021‐22 136700 141753 4%
18
10
Load Profile Analysis & Projection
1 9 19
Low Duration Curve
2 0 Source ‐ Night Report ,UPPCL 20
11
Uttar Pradesh – Load profile
13500
15500
17500
MW
Monthly Avg. Load ProfileJan,17
Feb,17
Mar,17
14500
15500
16500
17500 June month avg. load profile
June, 2013
June, 2014
MW
9500
11500
13500
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
M
Apr,17
May,17
Jun,17
2 1 Source ‐ Night Report ,UPPCL
9500
10500
11500
12500
13500
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
June, 2014
June, 2015
June, 2016
June, 2017
M
21
Projected Load Profile – High Growth Scenario
Annual Load Profile (High Growth Scenario)
26206
30058
20000
25000
30000
MW
2026-27
2025-26
2024-25
2023-24
2022-23
2021-22
2020-21
2 2
10000
15000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
2019-20
2018-19
2017-18
2016-17
2015-16
Hour22
12
Projected Load Profile – Realistic Growth Scenario
2833730000 Annual Load Profile (Realistic Growth Scenario)
2026 27
24704
16000
18000
20000
22000
24000
26000
28000
MW
2026-27
2025-26
2024-25
2023-24
2022-23
2021-22
2020-21
2019 20
2 3
10000
12000
14000
16000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
2019-20
2018-19
2017-18
2016-17
2015-16
Hour23
Projected Load Profile – Medium Growth Scenario
28000.0 Annual Load Profile (Medium Growth Scenario)
2026 27
16000.0
18000.0
20000.0
22000.0
24000.0
26000.0
MW
2026-27
2025-26
2024-25
2023-24
2022-23
2021-22
2020-21
2 4
10000.0
12000.0
14000.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
2019-20
2018-19
2017-18
2016-17
2015-16Hour
24
13
Projected Load Profile – Low Growth Scenario
24000
Annual Load Profile (Low Growth Scenario)
2026 27
16000
18000
20000
22000
MW
2026-27
2025-26
2024-25
2023-24
2022-23
2021-22
2020 21
2 5
10000
12000
14000
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
2020-21
2019-20
2018-19
2017-18
2016-17
25
GAMS Based Optimisation Model
2 6 26
14
Input Variables
Minimise social cost of power supply
2 7 27
GAMS Simulation for Different Power Procurement Scenario
Demand Load Profile High, Realistic, Medium & Low Growth Scenario
With or Without Short‐term Contracts
Solar & DSM ‐Realistic Targets
Fi d P iti Fl ti P iti
Solar & DSM –Policy Targets
For candidate
2 8
Fixed Position
Fixed All PlantsFixed 3 Plants Float Panki @ (2023,24,25)
Floating Position
Float All PlantsFloat 3 Plants & cancel 1 (H,
J, O, P)
Similar Scenarios as for Realistic
Targets
Plants only
28
15
GAMS Output
• Demand & Generation Chart
PLF f Pl t• PLF of Plants
• Optimal position for candidate plants
• Utility Cost & Social Cost
2 9 29
Demand & Gen. Curve for Realistic Growth Scenario (Fix All Without Considering STPP)
Load Shed
30
Existing Generators
Candidate Generators
10
15
20
25
rocured in thousand M
W
3 0
Solar
Existing Generators
0
5
1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
Power Pr
Solar Existing Generators Candidate Generators Load Shed Demand
30
16
Demand & Gen. Curve for Medium Growth Scenario(Fix All Without Considering STPP)
Load Shed25
30
ExistingGenerators
Candidate Generators
10
15
20
25
Power Procured in thousand M
W
3 1
Solar
Existing Generators
0
5
1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19 1 7 13 19
2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027
P
Solar Existing Generators Candidate Generators Load Shed Demand
31
Choosing Power Procurement Strategy
3 2 32
17
Conclusions & Key Observations
• Most optimal strategy ‘All Float’ case
• Certain existing plants/PPAs have low PLFg p /
• Progress on DSM and Solar capacity addition should be reviewed for medium‐term power procurement
• Planned and effective measures should be in place for metering, meter reading, billing and collection
• Periodical Power Procurement Analysis (at least every three years)
3 3
• Extension of Time of Day (ToD) tariff for all large consumers
• fair and transparent competitive bidding
• Adopt a state‐level UMPP model
33
Disruptive Changes in Future
Open Access
f l Rooftop Solar
Retail competition
Metro & Electric Traction
Electric vehicles
Smart Grid
3 4
Smart Grid
Storage
Franchisee (with exist clause for power procurement?)
34