The Reliability Backstop Procurement is not dead. It is in a rewrite sprint — and the calendar is doing the talking. On Thursday, October 22, PJM holds a special Members Committee meeting, scheduled for 9 a.m. to noon Eastern, to take stakeholder feedback on a revised filing it has told federal regulators it intends to submit by October 29. Separately, FERC has ordered PJM to file tariff revisions by November 11 addressing the large-load capacity rules the Commission rejected on October 9. And if neither effort satisfies the Commission, a paper hearing in Docket No. EL26-108-000 — with initial briefs reportedly due in mid-November — stands ready to let FERC impose its own version.
Three dates, two dockets, one underlying question: who pays for the capacity the data-center boom is demanding, and by what mechanism. FERC’s objections were not to the procurement’s existence but to its plumbing — specifically, to whether the costs would land on the right parties. The Commission found three defects in 196 FERC ¶ 61,245, all variations on the same theme: cost allocation, transmission-owner exit rules, and collateral requirements for load-serving entities. The next three weeks are when PJM has to prove it can build those pieces to the Commission’s specifications. As Chair Laura Swett’s concurrence put it, the Commission would not be forced into accepting what she characterized as a deeply flawed, eleventh-hour procurement with billion-dollar implications for consumers. The rewrites are where that judgment gets tested.
The three clocks
Start with the calendar, because the sequencing determines the leverage. The backstop’s bid window was supposed to open September 30; FERC’s order instead suspended the procurement for five months with an effective date of February 28, 2027, subject to refund and a paper hearing. But the order deliberately left a faster door open: FERC encouraged PJM to refile a revised Section 205 proposal curing the three defects, and indicated that a filing by October 29 could prompt the Commission to hold the paper-hearing schedule in abeyance while it considers the new proposal. PJM spokesman Jeffrey Shields confirmed to Utility Dive that PJM intends to meet that October 29 target and will hold the special Members Committee meeting on October 22 to consult members on the substance of the filing. A special Transmission Owners Agreement–Administrative Committee meeting is also planned.
The second clock belongs to a different docket. On October 9, in Docket No. ER26-3515, FERC issued a unanimous order on PJM’s separate interim plan for large loads: it accepted PJM’s definition of a Large Load — cumulative peak load of 50 MW or greater at a single electrical site or affiliated facilities within a one-mile radius — and the Large Load Registry, effective October 12. It rejected nearly everything else: the curtail-first interim resource adequacy service (IRAS) and the bring-your-own-new-capacity (BYONC) mechanism, which FERC said PJM had not justified in its proposal to lock in resource accreditation over the life of a contract. FERC ordered PJM to file tariff revisions by November 11, saying most of the revisions would take effect June 1. The vote was unanimous; Commissioner David Rosner concurred.
The third clock is the one PJM is racing against. If the October 29 refiling does not cure the defects, the paper hearing proceeds — initial briefs from PJM are reportedly due November 13 — and FERC may ultimately impose its own backstop design through the Section 206 proceeding in EL26-108. The effective date of February 28, 2027 remains the outside bound. Every week the rewrite slips is a week the 6,831 MW hole — the shortfall from the July 2028/29 Base Residual Auction — sits unaddressed, and the December 2029/30 auction creeps closer on the existing, flawed rules.
Fix one: cost allocation, the load-growth question
The first and hardest repair is cost allocation, and FERC’s guidance was unusually specific. PJM had proposed allocating backstop costs on a snapshot of current load; the Commission found that method inconsistent with cost causation because it does not capture the forecast growth from large new users — precisely the loads driving the shortfall. FERC’s alternative: allocate using updated load forecasts that reasonably capture all forecasted load growth, and set the initial procurement target with the most current forecasts available. It also rejected PJM’s proposal to let qualifying cooperatives and municipal utilities opt out, finding the carve-out could be unduly discriminatory against other load-serving entities with data centers in their footprints. On October 7, the Department of Energy filed a statement of position in the backstop docket — its first such filing at FERC in at least five years, according to Utility Dive — arguing that “PJM should implement the cost allocation and other reforms identified by the Commission to ensure that the costs of serving new data centers or other large electric energy loads are not unfairly shifted to PJM’s existing ratepayers.” The administration has separately gathered commitments to a “Ratepayer Protection Pledge” under which large electric loads fund the generation and infrastructure they require; seven PJM utilities — American Electric Power, AES, CenterPoint Energy, Dominion Energy, Exelon, FirstEnergy, and PPL — have signed.
The principle being assembled is beneficiary-pays: new large loads should arrive with their own supply or pay for it directly, rather than socializing costs through the capacity market. PJM’s own design already gestures at this — the bilateral matchmaking track lets large loads contract privately and shrinks the central auction’s target accordingly, and starting with the 2029/30 delivery year PJM plans to exclude new large loads from the demand forecasts used in general capacity auctions. The October 29 refiling is where that principle has to become arithmetic: a forecast-based allocation that actually lands costs on the growing loads. Commissioner David Rosner’s concurrence pointed in the same direction, discussing assignment of costs to the zones experiencing data-center growth.
Fix two: the exit rules
The second repair is narrower but revealing. PJM had proposed that if a transmission owner leaves the RTO before backstop costs are fully recouped, the remaining costs would be assigned to the exiting transmission owner. FERC said PJM had not shown those costs would land on the entities causing them, and directed that such costs instead be assigned to the load-serving entities within the transmission zone.
The principle is cost causation again, but applied to corporate form: costs follow the load that caused them, not the company that happens to exit. In practice this matters because the backstop envisions commitments of up to 15 delivery years — long enough that transmission ownership could change hands within the cost-recovery period. The refiling must write a rule that survives that scenario without reopening the allocation fight. The special Transmission Owners Agreement–Administrative Committee meeting suggests PJM knows this one is legally intricate even if it is analytically straightforward.
Fix three: collateral, and the $2 billion example
The third repair is about financial plumbing that almost no one outside the market watches but everyone inside it feels. FERC found PJM’s collateral requirements for power suppliers appropriate, but the proposed requirements for load-serving entities potentially flawed — failing, in the Commission’s words, to strike a reasonable balance “between protecting PJM’s markets and its members against the risk of default and ensuring that market participants have adequate flexibility to manage their collateral requirements.”
The concrete illustration came from Northern Virginia Electric Cooperative, which said it would have faced roughly $2 billion in collateral postings under the proposal. A cooperative serving one of the fastest-growing load pockets in the country — the Northern Virginia data-center cluster, the largest concentration of data-center load in the United States — facing a $2 billion posting requirement is the kind of number that turns a technical parameter into a Commission-level objection. Set collateral too low and a load-serving-entity default cascades onto everyone else; set it too high and you price the fastest-growing participants out of the market they are being asked to help fix. Expect this to be one of the most negotiated sections of the October 22 discussion.
The second docket: what survived October 9
The November 11 deadline deserves its own explanation, because the October 9 order is a consequential ruling in its own right. FERC accepted the Large Load definition and the registry — effective October 12, meaning PJM now has a formal, tariff-backed inventory of loads of 50 MW and above — while rejecting the curtail-first IRAS framework and the BYONC mechanism. It also severed and rejected, without prejudice, PJM’s definition of a “new large load” (end-use load going into service after June 1, or load increased to 50 MW), since that definition was needed to describe the subset of loads subject to curtailment under the rejected IRAS framework. The tariff revisions due November 11 must stay on the right side of that line — a registry and forecast exclusions, yes; curtail-first retail service, no.
PJM had wanted the revisions approved by October 12 to prepare for the December 2029/30 Base Residual Auction. FERC instead said most of the revisions would take effect June 1. The registry itself, though, is live now, and it feeds directly into the backstop fix: better visibility into large-load project progress was something commissioners explicitly asked for, and a registry that tracks which projects are real — versus speculative announcements — is what makes forecast-based cost allocation something more than an aspiration.
What FERC’s own version would look like
The Section 206 proceeding in EL26-108 is the stick behind the encouragement. If PJM’s refiling does not cure the defects, FERC can impose its own design — and the September 29 order plus the concurrences amount to a fairly detailed specification: forecast-based cost allocation capturing large-load growth, no carve-outs that shift costs to neighbors with data centers in their footprints, exit costs assigned to load-serving entities in the zone, balanced collateral for load-serving entities. Commissioner Lindsay See’s concurrence put the standard in kitchen-table terms: existing customers should not be left paying costs attributable to new demand.
The bilateral track matters more than it appears. Matchmaking between large loads and new-supply developers continues independently of the frozen auction, and every qualifying bilateral deal shrinks the eventual central target. If large loads contract privately at scale, the backstop that eventually opens could be materially smaller than 6,831 MW — the beneficiary-pays principle working as designed, with FERC’s allocation formula dividing a smaller pie. If bilateral deals stall, the full weight falls back on the central procurement and the unresolved allocation fight. PJM has called the bilateral route the preferred path for new capacity to enter the market.
What it means for December and beyond
The December 2029/30 Base Residual Auction remains on schedule, and it will run under whatever rules are in place — which is precisely why the November 11 registry revisions and the forecast-exclusion mechanism matter beyond the backstop. Starting with that delivery year, PJM plans to exclude new large loads from the demand forecasts used in the general auction, so existing consumers do not bear higher capacity costs caused by large new loads that have not contracted for their own supply. The July auction’s lesson — two consecutive auctions at the $325/MW-day price cap delivering barely 525 MW of new supply against a 6,831 MW shortfall — is what made both dockets necessary.
For developers, the freeze extends the uncertainty about the single question that matters: what price and contract length can a new project count on. The backstop’s 15-year commitments at up to $555/MW-day were designed to solve the financing problem that one-year capacity commitments at a capped $325 cannot. For consumers, the proceedings change nothing about bills already set — capacity costs for already-cleared delivery years remain in place. What is being decided in the next three weeks is the shape of the future bill: which loads it lands on, by what formula, and with what collateral backstopping it.
What to watch
Four things will determine how this resolves. The first is the October 22 Members Committee meeting itself: it will show whether PJM’s stakeholders are converging on the allocation formula or relitigating it — watch especially for pushback from cooperatives and municipal utilities on the opt-out question, and from data-center interests on how aggressively the forecast method captures their growth. The second is the October 29 filing: whether it lands, whether it cures all three defects or only some, and whether FERC pauses the paper-hearing schedule in response — that pause, or its absence, is the cleanest real-time signal of whether the Commission considers the sprint successful.
The third is the November 11 tariff revisions in the large-load docket: they will reveal how PJM interprets FERC’s jurisdictional boundary after the October 9 rejection, and whether the registry starts producing the project-level visibility commissioners asked for. The fourth is the December 2029/30 auction under the forecast-exclusion rules: it will show whether the regular market’s machinery can begin closing a gap that two capped auctions could not, or whether the backstop — rewritten or imposed — remains the only bridge. The procurement is not dead. But the version that opens, if it opens, will look substantially different from the one FERC froze — and the next three weeks are when that difference gets written.