Flexibility costs under high variable renewable energy generation: - - PowerPoint PPT Presentation

Flexibility costs under high variable renewable energy generation: the Chilean case

.

• S. Mocarquer, R. Quinteros, S. Binato, M. V. F. Pereira

2018 IEEE GENERAL MEETING Portland, August 2018

1

Overview

• Objectives of the Analysis
• Chilean Market Overview
• Methodology
• Results
• Conclusions

2

Objectives

• Quantify the effects of massive VRE1 insertion in the operation
• f the electric system with focus on ‘Flexibility Costs’
• Provide valuable inputs for the regulatory discussion in Chile
• The study was commissioned by the Chilean Generators

Association (AG)

• Independent analysis based on public information

3

Note: (1) VRE: Variable Renewable Energy (Solar, Wind)

Wind 5% Solar 5% Biomass 3% Coal 40% Natural Gas 16% Diesel 1% Small Hydro 3% Hydro Dam 13% Hydro RoR 14%

Generation

74,222 GWh/y

Chilean Market Overview

4

Wind 6% Solar 8% Biomass 2% Coal 21% Natural Gas 19% Diesel 13% Small Hydro 2% Hydro Dam 15% Hydro RoR 12%

Capacity

22,343 MW

Source: Annual Energetic Report CEN, January 2018. Installed Capacity SEN, Anuario CNE 2017. GDP, The World Bank.

Key information 2017 Renewable generation 15% (excluding large hydro) Peak demand 10,363 MW Energy Sales 49% regulated, 51% un-regulated Transmission lines 32,100 km GDP per capita (PPP) US$ 24,085

Consulting Team

• Consulting firm founded in 2013 by

executives from the electricity sector, in Santiago - Chile, to support investors and stakeholders in decision-making in the energy sector.

• Wide range of services taking

advantage of extensive experience and high degree of specialization;

– Market and regulation analysis – Business strategy – Due diligence for transactions – Business development

www.morayenergy.com

• Provider of analytic tools and

consultancy (economic, regulatory and financial studies) in electricity and natural gas since 1987, based in Rio de Janeiro – Brazil.

• Team of 54 specialists (17 PhDs, 31

MSc) in engineering, optimization, energy, statistics, finance, regulation, IT and environmental analysis.

• In more than 70 countries on all

continents. www.psr-inc.com

Methodology of the study

• The aim was to estimate the flexibility costs associated with

different VRE (solar-wind) expansion scenarios

Assumptions and scenarios definition Flexibility costs identification Flexibility costs functions Flexibility costs estimation Detailed hourly simulation of the system operation Generation, reserve and transmission expansion Detailed electric studies Simulation verification with the detailed electrical network Flexibility costs System Modeling AC Verification

Modeling Tools: PSR Core planning system

7

ePSR: Information management environment

OptGen/OptFolio

Integrated generation / interconnection / reserve expansion + financial analysis

SDDP/NCP/CORAL

Probabilistic hourly simulation of system operation and G&T supply reliability evaluation

NetPlan

Detailed probabilistic transmission and VAr planning

OptFlow

Optimal Power Flow linearized active power w/ losses and full AC model

HERA

Inventory of hydro basins and other renewable resources

Time Series Lab

Multiscale stochastic scenarios (inflow, renewable, demand)

PSR Cloud

Modeling Tools: PSR Core planning system

7

ePSR: Information management environment

OptGen/OptFolio

Integrated generation / interconnection / reserve expansion + financial analysis

SDDP/NCP/CORAL

Probabilistic hourly simulation of system operation and G&T supply reliability evaluation

NetPlan

Detailed probabilistic transmission and VAr planning

OptFlow

Optimal Power Flow linearized active power w/ losses and full AC model

HERA

Inventory of hydro basins and other renewable resources

Time Series Lab

Multiscale stochastic scenarios (inflow, renewable, demand)

PSR Cloud

Scenario Definition

• VRE insertion level was driven by investment costs and demand scenario

8

Scenario coding: DXCY – Plan with demand scenario X and investment costs Y – X: A (high demand), M (average), B (low) – Y: A (high price); M (average); B (low)

81 senarios

Flexibility

• Flexibility → ability of the

system to efficiently respond to supply and demand imbalances

• Massive insertion of VRE →

greater challenges in system

• peration require system

flexibility

9 1500 3000 4500 6000 7500 9000

Solar FV Eólica

CAGR 80+%@5 years (2012-2017)

Solar PV Wind

Wind and Solar Generation [GWh] Sources of Flexibility

• Generation technologies ← focus of the study
• Demand response
• Storage technologies
• Interconnections

Flexibility Cost Components

• The following flexibility costs were evaluated:
• Cost functions1 were applied to output variables obtained

from the simulations

• Ex-post analysis to assess unrecovered costs under current

regulation

10

Type of Cost Components Function Direct Start Up Costs Fuel and emission costs 𝑔 #Start Cycles Indirect Start Up Costs Capex and maintenance 𝑔 #Start Cycles Ramp Up/Down Cost Capex and maintenance 𝑔 #Ramp Cycles Efficiency Cost Fuel and emissions 𝑔 𝐸𝑗𝑡𝑞𝑏𝑢𝑑ℎ Opportunity Costs Lost variable margin 𝑔 𝐸𝑗𝑡𝑞𝑏𝑢𝑑ℎ 𝑏𝑜𝑒 𝑇𝑞𝑝𝑢 𝑄𝑠𝑗𝑑𝑓

Note: (1) The functions related to the start up and ramp up/down costs have been estimated using approximations based on international sources (Power Plant Cycling Costs, NREL, 2012)

Criteria for operational reserve

• Challenge to incorporate VRE effect to current methodology in

use in Chile ➢R = ƒ(D, G, VRE)

11 Demand (forcast and variation error) Generation (grater unit contingency) Non-supply energy (NSE) cost Optimal annual reserve

Current Metodology in Chile 𝑺𝒃𝒐𝒐𝒗𝒃𝒎 = 𝒏𝒋𝒐(𝑫𝑷𝒒𝒇𝒔 + 𝑮𝒃𝒋𝒎𝒗𝒔𝒇)

VRE reserve Generation (grater unit contingency) Dynamic probabilistic reserve

Implemented Modeling 𝑺𝒊 = 𝑬𝒇𝒏𝒃𝒐𝒆 + 𝒏𝒃𝒚(𝑺𝑯

𝒊, 𝑺𝑾𝑺𝑭 𝒊

)

Demand (forcast and variation error)

VRE Reserve (𝑆𝑊𝑆𝐹)

• Required reserve to account for the uncertainty associated

with the forecast error of VRE generation from the simulated series: 𝑆VRE

∗

= 𝜇 × 𝐹 𝑆 + 1 − 𝜇 × 𝐷𝑊𝑏𝑆90%(𝑆)

• With this type of risk criterion, 𝜇=0.8 represents a reasonable

compromise between reliability and cost

12

Step 1 Forecast error by hour and by series 𝜀ℎ

𝑡

Step 2 Difference of forecast error between consecutive hours for each series Δℎ

𝑡

Step 3 Optimal VRE reserve calculation for hour h, 𝑆𝑊𝑆𝐹

∗

under CVaR criteria.

Wind – Solar Complementarity

• The complementarity of the wind and solar generation is

captured in the optimized generation expansion considering hourly profiles

13 0% 20% 40% 60% 80% 100% 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 Wind Solar

Main Results: Generation Expansion

• Wind and solar technologies grow between 9.000 and 16.000 MW by

2030 (investment potential between US$ 8.000 and 18.000+ millions)

• Reserve expansion is identified in the North of Chile (200 – 1.000 MW)

14

2 4 6 8 10 12 14 16 18

VRE Expansion 2018-2030 (GW)

Solares Eólicas Solar PV Wind

Main Results: Generation by Technology

(Median Hydrology - DMCM)

• VRE generation share of 42% by 2030
• Including hydro, renewables account for 75% of energy

generation by 2030

15 Thermal 25% Other Renewables 4% Solar PV 30% Wind 12% Hydro 29%

2030

Thermal 31% Other Renewables 4% Solar PV 20% Wind 11% Hydro 34%

2025

Thermal 35% Other Renewables 5% Solar PV 8% Wind 13% Hydro 39%

2021

Main Results: Generation by Technology

(Dry Hydrology – DMCM)

• Thermoelectricity is still relevant by 2030 under dry hydrology

(33% share)

• No decommissioning of coal fired power plants was

considered

16 Thermal 48% Other Renewables 5% Solar PV 8% Wind 13% Hydro 26%

2021

Thermal 41% Other Renewables 5% Solar PV 20% Wind 11% Hydro 23%

2025

Thermal 33% Other Renewables 4% Solar PV 30% Wind 12% Hydro 21%

2030

Main Results: Transmission Expansion

• Relevant capacity expansion

is needed at 500 kV level by 2025

• Expansion plan proposed by

the Government includes a longer 500 kV HVDC line between Kimal and Polpaico (US$1,8 billion)

17

Kimal Cardones Polpaico Cautín

• P. Montt

Charrua

500 kV 220 kV 1.500 MW 500 kV 220 kV 1.500 MW (01/2019) 4.900 MW (01/2025) 224 MW 270 MW (01/2019) 750 MW (06/2022) 500 kV 220 kV 4.200 MW 900 MW 500 kV 220 kV 2.100 MW 2.600 MW (12/2018) 3.750 MW (05/2021) 5.350 MW (01/2025) 500 kV 220 kV 172 MW 580 MW (07/2021)

Main Results: Daily Dispatch – 2021

• The reservoirs and thermoelectricity will provide flexibility in

an increasing manner

– Hydro Dam: Daily storage (solar hours) – Coal: Ramping/minimum operation – CCGT: Cycling

18

Note: Median Hydrology - DMCM

Main Results: Daily Dispatch – 2025

• The reservoirs and thermoelectricity will provide flexibility in

an increasing manner

– Hydro Dam: Daily storage (solar hours) – Coal: Ramping/minimum operation – CCGT: Cycling

19

Note: Median Hydrology - DMCM

Main Results: Daily Dispatch – 2030

• The reservoirs and thermoelectricity will provide flexibility in

an increasing manner

– Hydro Dam: Daily storage (solar hours) – Coal: Ramping/minimum operation – CCGT: Cycling

20

Note: Median Hydrology - DMCM

Main Results: Marginal Cost

• Fluctuation of intraday marginal cost increases over time leading to

potential collapse during solar hours by 2030

• The minimum cost expansion requires long-term signals (contracts)

– Pure short-term marginal cost signals (in solar hours) may be insufficient to trigger investment

21

1 34 67 100 133 166 199 232 265 298 331 364 397 430 463 496 529 562 595 628 661 694 727

Hourly Marginal Costs (2021)

1 34 67 100 133 166 199 232 265 298 331 364 397 430 463 496 529 562 595 628 661 694 727

Hourly Marginal Costs (2025)

1 34 67 100 133 166 199 232 265 298 331 364 397 430 463 496 529 562 595 628 661 694 727

Hourly Marginal Costs (2030)

Main Results: Reserve Requirements

• Increasing reserve requirements in certain periods of the day
• Dynamic probabilistic reserve determination criteria will be

key to address high VRE insertion levels

22

2021 2025 2030 500 1000 1500 1 3 5 7 9 11 13 15 17 19 21 23 MW

Minimum Reserve SING (DACB)

2021 2025 2030 500 1000 1500 1 3 5 7 9 11 13 15 17 19 21 23 MW

Minimum Reserve SIC (DACB)

• Residual generation1: the Chilean ‘Duck Curve’

Main Results: Flexibility – 2021

23

Note: (1) Discounting VRE generation. Median Hydrology - DMCM

Winter – 2021

ΔP = 1.000 MW/h

MW/min Between Hours MW/min Between Hours DACB 15,7 18-19

• 10,1

20-21 DMCM 15,8 18-19

• 9,9

20-21 DBCA 15,9 18-19

• 9,7

20-21 Case Max Increase Max Distribution

Summer – 2021

ΔP = 1.000 MW/h

MW/min Between Hours MW/min Between Hours DACB 19,8 18-19

• 18,2

21-22 DMCM 19,3 18-19

• 17,8

21-22 DBCA 18,8 18-19

• 17,4

21-22 Case Max Increase Max Distribution

• Residual generation1: the Chilean ‘Duck Curve’

Main Results: Flexibility – 2025

24

Note: (1) Discounting VRE generation. Median Hydrology - DMCM

Winter – 2025

ΔP = 4.000 MW/h

Summer – 2025

ΔP = 4.000 MW/h

MW/min Between Hours MW/min Between Hours DACB 38,7 20-21

• 61,2

7-8 DMCM 29,0 19-20

• 46,3

7-8 DBCA 23,9 18-19

• 35,4

7-8 Case Max Increase Max Distribution MW/min Between Hours MW/min Between Hours DACB 63,2 18-19

• 36,0

8-9 DMCM 49,6 18-19

• 29,6

8-9 DBCA 38,7 18-19

• 22,4

8-9 Case Max Increase Max Distribution

• Residual generation1: the Chilean ‘Duck Curve’

Main Results: Flexibility – 2030

25

Note: (1) Discounting VRE generation. Median Hydrology - DMCM

Winter – 2030

ΔP = 6.000 MW/h

Summer – 2030

ΔP = 6.000 MW/h

MW/min Between Hours MW/min Between Hours DACB 94,0 18-19

• 66,3

8-9 DMCM 84,9 18-19

• 60,8

8-9 DBCA 68,7 18-19

• 47,9

8-9 Case Max Increase Max Distribution MW/min Between Hours MW/min Between Hours DACB 63,7 20-21

• 92,6

7-8 DMCM 53,0 20-21

• 81,5

7-8 DBCA 39,1 20-21

• 62,3

7-8 Case Max Increase Max Distribution

Main Results: Thermoelectricty Cycling

26 2021 2025 2030

CT A

Note: Results corresponds to April, DMCM scenario.

Coal Plant: CCGT:

Dispatch (MW) Dispatch (MW) Dispatch (MW) Dispatch (MW) Dispatch (MW) Dispatch (MW)

Main Results: Flexibility Costs

• Thermoelectric generators will incur in increasing flexibility costs

– US$150 to 350 millions per year in 2030 (mostly driven by start up/down cycles)

27

50 100 150 200 250 300 350 400 2021 2025 2030 MUS$

Flexibility Costs (DACB)

Opportunity Cost Less Efficiency Cost Ramp up/down cost Indirect Starting Cost Direct Starting Cost

Conclusions

• Unitary operation costs

will be reduced between 11% and 20% by 2030

• Flexibility costs could

reach up to US $ 350 millions by 2030

28 2021 2025 2030

Unitary Operation Costs

Operational (Variable) Cost CO2 Emissions Tax Cost Flexibility Cost Cost Savings since 2021

Flexibility costs increase 3+ times compared to 2021 Operational and emission costs reduced 30+% compared to 2021 Costs savings between 11% and 20%

• Flexibility costs must be addressed in the regulation so that the

potential VRE expansion can be achieved in an efficient manner

Further reading

• Full report and additional presentations can be downloaded

from our website:

29

http://www.morayenergy.com/

Flexibility costs under high variable renewable energy generation: the Chilean case

.

• S. Mocarquer, R. Quinteros, S. Binato, M. V. F. Pereira

2018 IEEE GENERAL MEETING Portland, August 2018

30

Disruptive factors

• The analysis presented is subject to the following uncertainty

factors:

– Changes in the CO2 tax level and treatment – Corporate decarbonization policies – Effect of climate change on hydrology – Greater competitiveness of storage systems – International interconnections development (electricity and gas)

31