1.
Please provide your feedback and perspectives on the value and need for Fast Start Pricing within ISO markets, including perspectives on the Fast-Start Pricing analysis provided in Working Group Session #16.
The Public Advocates Office at the California Public Utilities Commission (Cal Advocates) is the state-appointed independent ratepayer advocate at the California Public Utilities Commission (CPUC). Our goal is to ensure that California ratepayers have affordable, safe, and reliable utility services while advancing the state’s environmental goals. Our efforts to protect ratepayers include energy, water, and communications regulation advocacy.
The California Independent System Operator (CAISO) should not implement Fast Start Pricing (FSP)
The CAISO is re-evaluating FSP given the existence of a diverse resources fleet in a broad regional market.[1] The CAISO’s recent analysis identified increases to market costs if FSP is adopted,[2] consistent with the findings of other grid operators.[3] The CAISO has stated that FSP may provide benefits; namely, to enable certain fast-start resources (FSRs) to inform energy market price formation if those FSRs are dispatched on the margin but fail to meet the entire need of marginal load.[4] Today, FSRs that do not inform the LMP when dispatched for marginal load receive bid cost recovery (BCR) payments to be made whole with respect to their economic bid.[5] FSP’s ability to decrease BCR payments has not been studied in the analysis, though the CAISO discussed that BCR costs would not be significantly offset by FSP.[6]
Given that FSP does not appear to benefit grid reliability and would increase costs to ratepayers, Cal Advocates continues to oppose the adoption of FSP in the CAISO real-time market and its inclusion in the scope of the Price Formations Enhancement (PFE) initiative.[7] The CAISO should instead maintain the status quo of using BCR payments to ensure dispatched generation is made whole when appropriate, as this approach is relatively cheaper to FSP and provides the same result of dispatching energy when needed.
However, Cal Advocates provides feedback and several recommendations on the April Analysis if the CAISO continues to develop an FSP mechanism.
- FSP costs and net benefits must be clarified and further developed.
The April Analysis determined that FSP may result in incremental costs of up to $380,000 per month on an average annual basis.[8] The previous December Analysis found that incremental costs could be as high as $13.5 million per month on an average annual basis.[9] Powerex and the Public Power Council (PPC) estimated FSP costs of $1.3 billion per year.[10] Cal Advocates requests that the CAISO clarify the differences leading to the disparity in estimated costs for its own analyses.
Additionally, the net costs of FSP should be better understood. The CAISO has not yet determined the magnitude of potential reductions in BCR costs due to adoption of an FSP. The Department of Market Monitoring (DMM) determined that potential FSRs receive $18-$32 of BCR payments today; however, DMM was unable to state what amount would be mitigated by an FSP structure in the real-time market.[11] DMM also noted that reductions in BCR payments “may be offset” by increased BCR payments to other units caused by FSP.[12] DMM’s concern that BCR payments may in fact increase due to FSP is consistent with PJM’s experience with FSP. In PJM, the adoption of FSP led to additional uplift costs as well as distorted price incentives.[13] As the CAISO narrows the analytical scope of the FSP track, the CAISO should provide additional information showing the potential impact of FSP on BCR payment reduction as well as potential BCR payment increases.
- CAISO should narrow the analytical scope to focus on the lowest-impact fast start pricing (FSP) options by discontinuing analysis of the constant amortization methodology and discontinuing analysis of sensitivities involving startup times over 30 minutes and minimum up times over 60 minutes; and consider analyzing sensitivities involving resources with startup times less than or equal to 15 minutes.
Cal Advocates appreciates the twelve sensitivity scenarios and attendant data exploration and analysis presented at the April 8, 2024 Working Group.[14] The overview was helpful for understanding potential permutations and some of the tradeoffs involved with the options. In particular, the analysis demonstrates that the constant amortization methodology is likely to yield the highest cost impacts—roughly double that of the other two amortization approaches.[15] Focusing the analysis on fewer sensitivities would allow stakeholders to examine the options in greater depth and understand the detailed tradeoffs between FSP mechanisms.
If the CAISO proceeds with FSP development, the CAISO should focus the analytical scope in three ways. First, CAISO staff should discontinue analysis of the sensitivities utilizing the constant amortization methodology, which would eliminate four sensitivities. Based on the CAISO’s analysis, the constant amortization methodology would result in the:
- Highest average price increases;[16]
- Highest hourly price spikes in 2022-2023;[17]
- Highest intra-CAISO monthly price impacts;[18]
- Most pronounced skew in the distribution of price changes in CAISO and in the system-wide market arrangement,[19] and;
- Largest increases in costs of supply with FSP for CAISO and in the system-wide market arrangement.[20]
Conversely, the other two amortization methodologies (average and adjusted) consistently lead to smaller cost increases. However, between the average and adjusted methodologies, one is not always less costly than the other. Thus, the CAISO should focus its analysis on the average and adjusted amortization methodologies to examine the methodologies in comparative detail and identify the best outcome for ratepayers.
Second, CAISO staff should discontinue analysis of sensitivities involving startup times (STUT) over 30 minutes and Minimum Up Times[21] (MUT) over 60 minutes, which would eliminate an additional four sensitivities. CAISO analysis indicates that utilizing STUT ≤ 60 minutes is more costly than a STUT ≤ 30 (although the differences are less pronounced than the difference between using constant amortization methodology and the other options).[22] Likewise, the CAISO noted that “[a]ll other [independent system operator (ISO)] FSP models use definitions of 60 minutes or less for start-up and minimum run time”[23] while no other ISO uses an MUT greater than one hour.[24] Narrowing the scope of analysis to resources with STUT that are between 0 and 30 minutes and MUT that are between 0 and 60 minutes would still include 19% of gas resource megawatts in the Western Energy Imbalance Market (WEIM) footprint.[25]
Third, if CAISO staff decide to analyze additional sensitivities, Cal Advocates requests analysis of a STUT ≤ 15 minutes option. One slide of the CAISO’s presentation identified the subset of gas resources which would fit a STUT ≤ 15 minutes and MUT ≤ 60 minutes definition, suggesting that analysis of a STUT ≤ 15 minutes option is viable within the current scope.[26] The Southwest Power Pool provides precedent for utilizing an even narrower definition for its FSP resources, at STUT ≤ 10 minutes.[27] Cal Advocates requests this analysis to understand whether utilizing an FSP definition of STUT ≤ 15 minutes and MUT ≤ 60 minutes would reduce impacts to ratepayer costs.
- The CAISO should clarify if the market dimension is a policy question.
The CAISO’s April Analysis presentation identified three dimensions for analyzing FSP, including market configuration. The slides note that market configuration “is only for analysis set up to determine if price setting is at each [Balancing Authority Area] individually or at a single system wide area.”[28] The comparison of Balancing Authority Area (BAA)-level pricing against system-wide pricing is certainly useful for inferring the differential impacts of FSP on individual BAAs, including the CAISO. However, the discussion during the Working Group was unclear as to whether this dimension is intended as a policy question for stakeholders. The system-wide pricing approach consistently yielded smaller cost increases compared to when pricing was determined within the CAISO only, across all amortization methodologies and STUT configurations. This is an intuitive result when considering that the system-wide pricing approach would increase competition among the WEIM-wide set of FSP resources and the system-wide price would be set by the lowest-priced marginal FSP resource available across a larger portfolio. Insofar as the market configuration dimension may be up for consideration, Cal Advocates urges the CAISO to utilize the system-wide pricing approach because this approach is likely to lead to lower ratepayer costs.
- FSP analysis should consider the risk of gaming.
The CAISO’s outside expert, Michael Cadwalader, presented on several topics during the Working Group. Of note, Cadwalader highlighted the potential for gaming FSP depending on the configuration selected, and stakeholder discussion highlighted possible avenues for gaming.[29] As the CAISO narrows the FSP options considered by stakeholders, CAISO staff should also commit their efforts and stakeholder discussion time to “minimize the susceptibility to gaming.”[30]
[1] In 2016, the CAISO argued that FSP would be “unlikely to improve price formation in the CAISO’s market.” Price Formation Enhancements: Stakeholder Workshop, June 9, 2022 at 9. Available at: https://www.caiso.com/InitiativeDocuments/Presentation-PriceFormationEnhancementsWorkshop-June9-2022.pdf.
[2] Price Formation Enhancements: Analysis on Fast Start Pricing, April 8, 2024 (April Analysis) at 10 and 74. Available at: https://www.caiso.com/InitiativeDocuments/Presentation-Price-Formation-Enhancements-Apr8-2024.pdf.
[3] The Midcontinent Independent System Operator (MISO) found that FSP increased prices between $0.50 and $1.45/megawatt-hour (Potomac Economics, 2022 State of the Market Report for the MISO Electricity Markets, June 15, 2023 at 41. Available at: https://www.potomaceconomics.com/wp-content/uploads/2023/06/2022-MISO-SOM_Report_Body-Final.pdf.). The Pennsylvania-New Jersey-Maryland Interconnection (PJM) found that FSP increased prices as well as increased reliance on uplift payments (Monitoring Analytics, LLC, State of the Market Report for PJM: Volume 1, 2023, March 14, 2024 at 21 and 30-32. Available at: https://www.monitoringanalytics.com/reports/PJM_State_of_the_Market/2023/2023-som-pjm-vol1.pdf.).
[4] Price Formation Enhancements: Working Group Session #6, October 12, 2023 at 11-17. Available at: https://www.caiso.com/InitiativeDocuments/Presentation-Price-Formation-Enhancements-Oct12-2023.pdf.
[5] Price Formation Enhancements: Working Group Session #9, December 11, 2023 (December Analysis) at 17. Available at: https://www.caiso.com/InitiativeDocuments/Presentation-Price-Formation-Enhancements-Dec12-2023.pdf.
[6] Price formation Enhancements Working Group Recording, April 8, 2024 at 2:08:40-2:10:20. Available at: https://www.youtube.com/watch?v=4ULeI0pAo20&t=7714s.
[7] Cal Advocates previously recommended that the CAISO remove FSP from the scope of the PFE initiative, citing that the CAISO’s findings to not adopt FSP in 2016 remain relevant today in addition to unnecessary increased costs to ratepayers. Cal Advocates, Comments on Price Formation Enhancements Working Group 6, November 3, 2023. Available at: https://stakeholdercenter.caiso.com/Comments/AllComments/6705ab14-132f-4a74-a0db-4ad6ba98b9b5#org-20bc9248-0b32-4d23-b075-70bbd48cf411.
[8] Based on incremental real-time energy supply relative to day-ahead schedules. April Analysis at 74.
[9] Based on total demand and an FSP-informed clearing price. December Analysis at 42.
[10] Powerex and PPC, The Importance of Fast-Start Pricing in Market Design, June 2022 at 2. Available at: https://www.caiso.com/InitiativeDocuments/Powerex-and-Public-Power-Council-Report-Importance-of-Fast-Start-Pricing-in-Market-Design.pdf.
[11] DMM, Comments on Price Formation Enhancements October 12, 2023 Working Group, November 2, 2023 (DMM October WG Comments) at 2. Available at: https://stakeholdercenter.caiso.com/Common/DownloadFile/f821f8da-a367-4da5-b833-74915608c5cb.
[12] DMM October WG Comments at 2.
[13] “The differences between the actual LMP and the fast start LMP will distort the incentive for market participants to behave competitively and to follow PJM's dispatch instructions. PJM is paying new forms of uplift in an attempt to counter the distorted incentives inherent in fast start pricing.” Monitoring Analytics, LLC, State of the Market Report for PJM: Volume 1, 2021, March 10, 2022 at 30. Available at: https://www.monitoringanalytics.com/reports/PJM_State_of_the_Market/2021/2021-som-pjm-vol1.pdf.
[14] April Analysis at 69.
[15] April Analysis at 9.
[16] April Analysis at 71
[17] Assuming a 60-minute startup time. April Analysis at 72-73
[18] Including the highest differences relative to if prices were calculated using system-wide dispatch. April Analysis at 75-79.
[19] More skew indicates more price increases due to FSP. April Analysis at 80-86
[20] April Analysis at 87-98.
[21] Also referred to as “minimum run time.”
[22] For example, the incremental costs per month for using STUT ≤ 60 minutes are $25,000 to $29,000 higher than using STUT ≤ 30 minutes across all three amortization methodologies. April Analysis at 74.
[23] April Analysis at 9.
[24] April Analysis 17.
[25] April Analysis at 50.
[26] April Analysis at 51.
[27] April Analysis at 9 and 17.
[28] April Analysis at 43.
[29] April Analysis at 108.
[30] April Analysis at 108.