Commissioner Clements Dissenting in part regarding PJM Interconnection, L.L.C.

Commissioner Allison Clements Statements
March 9, 2021
Docket Nos. EL19-58-005

As is the case with several other proceedings during these early days of my Commission tenure, my involvement in this proceeding originates near the end of a longer, comprehensive consideration of the issues – here, the design of several aspects of PJM’s reserve market.  As a result of this late involvement, I now consider the justness and reasonableness of only the aspects of PJM’s reserve market design that remain under consideration as a result of requests for rehearing in this docket.

As relates to these remaining issues, I must dissent in part on today’s order because I disagree that it is clear that PJM’s application of a 10% adder to combustion turbine resources (CTs) and use of historical reserve prices are assumptions that will yield just and reasonable capacity rates.  I recognize that these technical details are challenging and do not always lend themselves to easy answers.  But our acceptance must nonetheless be supported by record evidence, and here I find that evidence lacking.

Approval of the application of a 10% adder to CT offers and use of historical reserve prices are assumptions with the potential for material real-world consequences.  These assumptions feed into the net cost of new entry (Net CONE) value that affects all critical parameters of the capacity market.  An inflated Net CONE value can lead to artificially high capacity quantities and prices, results that can harm consumers. 

As an initial matter, I agree with the Commission’s decision in its May 2020 order to find that the reserve market changes adopted in this proceeding render the preexisting energy and ancillary services (E&AS) offset in PJM’s capacity market unjust and unreasonable.  PJM’s energy, ancillary services and capacity markets are meant to work in tandem to compensate suppliers for providing electric service to PJM customers.  The capacity market is designed to be responsive to expectations of E&AS revenues in the delivery year three years hence.  When those expectations are thrown out of whack, such as with a major market reform like the reserve changes in this proceeding, a correction may be warranted to ensure capacity market rates are not artificially high.  The shift to a forward-looking E&AS offset was such a correction, and it was warranted.

The calculation of the E&AS offset is complex and involves numerous assumptions.  I acknowledge there may be multiple just and reasonable methods for calculating it.  The Commission can offer some deference to PJM’s expertise so long as its choices result in just and reasonable rates as supported by record evidence. On rehearing here, however, parties again point to a paucity of record evidence supporting the 10% adder PJM applies to offers from CTs in its new dispatch model.  

In support of its compliance proposal, PJM argues that CTs may face additional fuel procurement costs as a result of being dispatched in real-time, and that such risks may not be limited to peak hours.  But, as Public Interest and Customer Organizations (PICOs) point out, PJM’s new dispatch model assesses all hours of the year, such that PJM’s assumptions about resource offers should be reasonable across all of those hours.  Yet real-world evidence from the Independent Market Monitor for PJM (Market Monitor) and PICOs indicates that CT offers are frequently below their cost-based level, including 40-50% of the time in 2019-2020.[1]   PICOs also point to a PJM simulation showing that if CTs included a 10% adder in their offers during all hours, their run hours would decline by 29% and their revenues would decrease by 30% compared to offering without the adder.[2]  These arguments suggest that PJM’s assumption that CTs will include the 10% adder during all hours of the year is contrary to the financial interests of the owners of those resources and likely is not a reasonable assumption.

PJM also argues the 10% adder is appropriate because it is applied to the reference resource CT in constructing the capacity market demand curve.  But, as PICOs argue, PJM has changed the model used to calculate the E&AS offset from one that looks only at peak hours to a new model that looks at dispatch in all hours.[3]  An adder that was arguably reasonable when examining only peak hours, during which fuel procurement costs may be higher, is not necessarily reasonable when examining all hours of the year, particularly in light of the information cited above that consistently offering at 10% above cost would be contrary to a CT owner’s financial interest.

I also find support lacking for PJM’s proposal to use historical reserve prices as a stand-in for projected reserve prices in its forward-looking E&AS offset.  I recognize the practical challenges in seeking to project revenues when no futures prices for a product exist.  Nonetheless, PJM’s new operating reserve demand curves clearly and purposefully assign a positive value to reserves in excess of the minimum reserve requirement, most of which had a zero valuation under the previous curve.  One can reasonably expect the reserve market price to be positive in far more hours of the year once the new curves are implemented in 2022.  It also appears a strong possibility, then, that reserve revenues for many resources will be greater under these changes.  This intuitive conclusion is supported by Brattle Group data, cited by PICOs, suggesting that, as a general rule, when energy prices rise, reserves prices rise at an even faster rate.  So again, PJM’s assumption appears to underestimate E&AS revenues.

I find protestors’ arguments and evidence on these points compelling, and PJM offers little in rebuttal.  I therefore find insufficient evidence in the record to assuage my concern that the 10% adder and use of historical reserve prices will lead to an understated assumption of E&AS revenues.

These questionable assumptions would be less concerning if their impact on capacity prices was not so great.  The E&AS offset is a major determinant of Net CONE, and Net CONE affects every aspect of the capacity auction, including the demand curve and market power mitigation.  PICOs assert that PJM’s 10% adder assumption translates to a but-for increase in Net CONE of roughly $30/MW-day, which is nearly 12% of the recently posted RTO-wide Net CONE of $260.50/MW-day for the 2022/2023 Base Residual Auction.[4]  These estimates translate to significant additional capacity costs to customers.

I appreciate the context in which I dissent.  PJM and market participants have expressed a commercial and operational urgency to running the delayed 2022/2023 capacity auction as soon as practicable.  But that factor does not lessen our responsibility to ensure that the capacity rates that emerge from that auction are just and reasonable.  The parameters I highlight here are just two among many in the capacity market, but they have significant impacts on the rates ultimately borne by consumers.  PJM will soon be filing with the Commission updates to its capacity market parameters as part of its quadrennial review.  The contents of that filing will include a broader set of considerations about capacity market design, including, as Chairman Glick’s concurrence notes, the choice of reference resource.  I look forward to reviewing that filing carefully when it comes before us.

For these reasons, I respectfully dissent in part.