Local Market Power Mitigation Enhancements. Conference Call March 25, 2011

Transcription

1 Local Market Power Mitigation Enhancements Conference Call March 25, 2011

2 ISO Stakeholder Initiative Process We Are Here Page 2

3 Agenda for today s meeting Estimated Topic Presenter Time 10:00 10:10 Introduction Chris Kirsten 10:10 10:20 Background Cynthia Hinman 10:20 10:40 Straw Proposal Cynthia Hinman 10:40 11:15 Examples Cynthia Hinman / Lin Xu 11:15 11:50 Inline Competitive Path Jeff McDonald Assessment 11:50 12:00 Next Steps Chris Kirsten Page 3

4 Goal of the LMPM straw proposal Meet the FERC requirement to use bid-in demand Incorporate design elements to reflect the implementation of: Convergence bidding Proxy demand resource Improve accuracy of mitigation in real-time market Provide the opportunity to incorporate inline competitive path designation Page 4

5 From issue paper to straw proposal Input from stakeholders MSC Meeting Stakeholder comments Other ISOs' mitigation processes The DMM approach or Option B approach Page 5

6 Straw Proposal Locational Marginal Price (LMP) Decomposition process overview 1. Perform all constraints (AC) run 2. Decompose the LMP for each location a. Energy component b. Loss component c. Congestion cost component i. Competitive constraint component ii. Non-competitive constraint component 3. If the non-competitive component of a physical supply ppy bid > 0, it potentially has market power and requires mitigation. Page 6

7 Straw Proposal LMP Decomposition process overview (continued) 4. A physical supply bid will be mitigated to the higher of: a. Its default energy bid (DEB) b. Its competitive LMP (analogous to the LMP produced in today s competitive constraints run) if that is lower than its unmitigated bid 5. Virtual supply bids are not mitigated 6. Run IFM with mitigated bids Page 7

8 Straw Proposal LMP Decomposition details For location i: LMP i = LMP i EC + LMP i LC + LMP i CC + LMP i NC Where: EC = the energy component LC = the loss component CC = the competitive constraints congestion component NC = the non-competitive constraints congestion component Page 8

9 Straw Proposal LMP Decomposition details (continued) For location i: Congestion Component LMP i = LMP i EC + LMP i LC + LMP i CC + LMP i NC Where: EC = the energy component LC = the loss component CC = the competitive constraints congestion component NC = the non-competitive constraints congestion component Page 9

10 Straw Proposal LMP Decomposition details (continued) For location i: Competitive LMP LMP i = LMP i EC + LMP i LC + LMP i CC + LMP i NC Where: EC = the energy component LC = the loss component CC = the competitive constraints congestion component NC = the non-competitive constraints congestion component Page 10

11 2 Bus Example System side Local side S 100 MW Non competitive G0: 600MW at $40/MWh G1: 100MW at $ 20/MWh G2: 200MW at $100/MWh G3: 200MW at $130/MWh G4: 100MW at $150/MWh VS1: 200MW at $110/MWh VD1: 200MW at $140/MWh PD1: 300MW at $160/MWh L Page 11

12 2 Bus Example (Continued) Page 12

13 2 Bus Example - LMPM under current approach Supply CC run (MW) AC run (MW) CC LMP AC LMP DEB LMPM Mit. Bid G $40 $40 $30 N $40 G $40 $110 $10 N $20 G $40 $110 $30 Y $40 G3 0 0 $40 $110 $60 N $130 G4 0 0 $40 $110 $70 N $150 VS $40 $110 N/A N $110 Terminology: CC run Competitive constraints run LMP Locational marginal price LMPM Local market power mitigation AC run All constraints run DEB Default energy bid Mit. Bid Mitigated bid Page 13

14 2 Bus Example - LMPM under current approach Dispatched up in AC run Supply CC run (MW) AC run (MW) CC LMP AC LMP DEB LMPM Mit. Bid G $40 $40 $30 N $40 G $40 $110 $10 N $20 G $40 $110 $30 Y $40 G3 0 0 $40 $110 $60 N $130 G4 0 0 $40 $110 $70 N $150 VS $40 $110 N/A N $110 Terminology: CC run Competitive constraints run LMP Locational marginal price LMPM Local market power mitigation AC run All constraints run DEB Default energy bid Mit. Bid Mitigated bid Page 14

15 2 Bus Example - LMPM under current approach Dispatched up in AC run Supply CC run (MW) AC run (MW) CC LMP AC LMP DEB LMPM Mit. Bid G $40 $40 $30 N $40 G $40 $110 $10 N $20 G $40 $110 $30 Y $40 G3 0 0 $40 $110 $60 N $130 G4 0 0 $40 $110 $70 N $150 VS $40 $110 N/A N $110 Terminology: CC run Competitive constraints run LMP Locational marginal price LMPM Local market power mitigation Bypasses mitigation due to VS1 crowd out AC run All constraints run DEB Default energy bid Mit. Bid Mitigated bid Page 15

16 2 Bus Example (Continued) Bypasses mitigation due to VS1 crowd out Page 16

17 2 Bus Example - LMPM straw proposal System side as reference bus Supply Sched. (MW) LMP EC CC LC NC UB DEB LMPM Mit. Bid G0 100 $40 $40 $0 $0 $0 $40 $30 N $40 G1 100 $110 $40 $0 $0 $70 $20 $10 Y $20 G2 200 $110 $40 $0 $0 $70 $100 $30 Y $40 G3 0 $110 $40 $0 $0 $70 $130 $60 Y $60 G4 0 $110 $40 $0 $0 $70 $150 $70 Y $70 VS1 100 $110 $40 $0 $0 $70 $110 N/A N $110 Terminology: Sched. Schedule CC Competitive constraint congestion Component NC Non-competitive constraint congestion component EC Energy component LC Loss component UB Unmitigated bid Page 17

18 2 Bus Example - LMPM straw proposal System side as reference bus Supply Sched. (MW) LMP EC CC LC NC UB DEB LMPM Mit. Bid G0 100 $40 $40 $0 $0 $0 $40 $30 N $40 G1 100 $110 $40 $0 $0 $70 $20 $10 Y $20 G2 200 $110 $40 $0 $0 $70 $100 $30 Y $40 G3 0 $110 $40 $0 $0 $70 $130 $60 Y $60 G4 0 $110 $40 $0 $0 $70 $150 $70 Y $70 VS1 100 $110 $40 $0 $0 $70 $110 N/A N $110 Terminology: Sched. Schedule CC Competitive constraint congestion Component NC Non-competitive constraint congestion component EC Energy component LC Loss component UB Unmitigated bid NC > 0 = potential market power Page 18

19 3 Bus Example PD1 400 MW G1 600 MW at $400 L MW G3 600MW at $50 S G4 G5 G2 600 MW at $200 L 2 Page 19

20 3 Bus Example PD1 400 MW G1 600 MW at $400 L MW G3 600MW at $50 S G4 G5 G2 600 MW at $200 L 2 System side of the constraint Page 20

21 3 Bus Example Local constrained areas PD1 400 MW G1 600 MW at $400 L MW G3 600MW at $50 S G4 G5 G2 600 MW at $200 L 2 Page 21

22 3 Bus Example - LMPM under current approach Supply CC run (MW) AC run (MW) Bid CC LMP AC LMP DEB LMPM Mit. Bid G1 0 0 $400 $50 $350 $50 N $400 G $200 $50 $200 $80 Y $80 G $50 $50 $50 $45 N $50 Terminology: CC run Competitive constraints run AC run All constraints run LMP Locational marginal price DEB Default energy bid LMPM Local market power mitigation Mit. Bid Mitigated bid Page 22

23 3 Bus Example - LMPM under current approach Supply CC run (MW) AC run (MW) Bid CC LMP AC LMP DEB LMPM Mit. Bid G1 0 0 $400 $50 $350 $50 N $400 G $200 $50 $200 $80 Y $80 G $50 $50 $50 $45 N $50 Dispatched up in AC Terminology: run $200 X 2-$50= $350 1 extra MW of load at CC run Competitive constraints run AC run All constraints run L1 will be served by LMP Locational marginal price DEB Default energy bid dispatching G2 up by LMPM Local market power mitigation Mit. Bid Mitigated bid 2 MW and G1 down by 1 MW Page 23

24 3 Bus Example - LMPM straw proposal Supply Sched (MW) LMP EC CC LC NC UB DEB LMPM Mit. Bid G1 0 $350 $50 $0 $0 $300 $400 $50 Y $50 G2 200 $200 $50 $0 $0 $150 $200 $80 Y $80 G3 200 $50 $50 $0 $0 $0 $50 $45 N $60 Terminology: Sched. Schedule CC Competitive constraint congestion Component NC Non-competitive constraint congestion component EC Energy component LC Loss component UB Unmitigated bid NC > 0 = potential market power Page 24

25 Straw Proposal RMR Condition 1 & Condition 2 Mitigation Concern - Use of bid-in demand and virtual bids could cause over or under commitment t of RMR resources Proposed solution Condition 1 units - market bids will be utilized in the AC run and RMR proxy bids will be used in place of DEBs. The same LMP decomposition will be used to determine when RMR proxy bids replace market bids. Condition 2 units ISO operators will manually dispatch these resources if needed and RMR proxy bids will be utilized in the market. For 2011, there is only 1 RMR contracted resource Page 25

26 Proposal for in-line competitive path assessment Stakeholder Call on March 25, 2011 Jeff McDonald, Ph.D. Department of Market Monitoring

27 Topics covered Proposed approach compared to current approach. In-line testing for competitiveness of constraints using the residual supply index Calculating the residual supply index for a constraint Identifying potentially binding constraints Other issues Page 27

28 Proposed approach compared to current approach Current approach Perform assessment four times a year (manual) Use range of system conditions: low, medium, and high conditions for hydro and load. Use anticipated generation and transmission availability can t account for changes / current conditions. Proposed approach Perform assessment in-line with market software daily for day ahead and every 15 min. for real time. Uses current system conditions. Uses current generation and transmission avail. Page 28

29 In-line testing for competitiveness of constraints using the residual supply index Competitiveness test is applied by the market software in a pre-market run. Use the residual supply index (RSI) to determine if there is competitive (effective) supply of counter-flow. Remove effective supply of up to three largest net suppliers for each potentially binding constraint. If RSI 1 then constraint is competitive, otherwise it is uncompetitive. Pass tested competitive / uncompetitive designations to LMPM process for potential mitigation. Default designation (if not tested) is competitive. Page 29

30 Calculating the residual supply index for a constraint The residual supply index for a constraint is the ratio of the effective supply of counter-flow without selected suppliers to the demand for counter-flow. Effective supply: Only remove effective supply of net suppliers. Accounts for current availability and dispatch level. Reflects current effectiveness from changes in network conditions. Demand for counter-flow: Measured using dispatch of effective resources from premarket run. Formulas for calculating the RSI are found on pages 3 and 4 of fthe DMM paper. Page 30

31 Identifying potentially binding constraints Local market power is created by congestion in the market run. The path competitiveness test t and the mitigation test t are applied before the actual market run can t perfectly predict congestion. Can use the critical constraint list from pre-market runs to expand list from binding to potentially binding. Trade-off: Adding non-binding constraints reduces under- identification and increases over-identification. For real time, using set of constraints with flow 97.5% or 95% of limit provides good accuracy. Note: constraint must fail test to be uncompetitive. See Table 3 from DMM paper for statistics. Page 31

32 Identifying potentially binding constraints real time market timing Currently Use congestion in HASP for mitigation in RTD. Propose Use critical constraint list from RTPD A/S run for mitigation in RTD. Page 32

33 Other issues Information identifying net suppliers and portfolio changes due to tolling agreements into market process. Default values for path designations if market software issues prevent testing. ti Method to calculate the mitigation trigger (price decomposition test) for non-competitive constraints that are not binding. Page 33

34 Next Steps April 1 Comments Due May 2 Draft Final Proposal May 9 Conference Call May 23 Comments Due June 29, 30 Board of Governors Page 34

35 Please send comments to Questions? Contact Cynthia Hinman Dynamic Competitive Path Assessment Comments to