The PJM 2026/2027 Base Residual Auction cleared at USD 329.17/MW-day, the auction's price ceiling, while leaving the 13-state region 6,831 MW short of its target, the second consecutive year of a capacity shortfall [S1][S3].
Capacity charges flowing into supply contracts have run at roughly nine times their pre-2024 level since June 2025, and contracted forward positions keep that multiplier locked in through at least May 2029 [S1]. The pipeline for new generation is small relative to the gap, with the auction adding 2,669 MW of new and uprated unforced capacity against a multi-thousand-MW deficit [S5].
Auction results: cap binding, 6.8 GW shortfall, 14.4% reserve margin
The 2026/2027 auction cleared 138,318 MW of unforced capacity inside the auction and 149,182 MW including self-supply commitments from utilities, against a target of 156,013 MW, a 6,831 MW gap almost identical to the 6,516 MW shortfall in the prior auction [S1]. The resulting reserve margin is 14.4%, well below PJM's 20% target and thin enough that a sustained heat wave or cold snap would leave little headroom [S1].
For the 2026/2027 and 2027/2028 auctions, PJM agreed with FERC to a price floor of USD 64,693/MW-year and a cap of USD 120,147/MW-year (equivalent to about USD 177.24/MW-day and USD 329.17/MW-day on a per-day basis), with the 2026/2027 auction printing at the cap [S4]. Demand-response ELCC ratings were also adjusted, dropping from 77% to roughly 69%, which reduces the accredited capacity each enrolled MW contributes but does not undermine the economics for flexible loads [S4].
Price path from 2022 to 2027: USD 28.92 to USD 329.17 per MW-day
Capacity prices in most PJM zones moved from USD 28.92/MW-day in the 2024/2025 delivery year to USD 269.92/MW-day for 2025/2026, then to USD 329.17/MW-day in all zones for 2026/2027, a factor of roughly 11 over two years [S1][S3].
The Independent Market Monitor estimated that data centers drove 63% of the increase in the 2025/2026 auction, translating to about USD 9.3 billion in higher capacity costs recovered from customers across the PJM footprint in a single year [S3]. Load forecasts in the Dominion Zone illustrate the scale shift: the 2022 forecast projected about 5,700 MW of growth by 2037, while the 2025 forecast projects more than 20,000 MW of data-center load alone by 2037 in the same zone [S3].
Who actually pays: a flat per-MW-day charge scaled by peak demand
The capacity charge follows a customer's contribution to peak demand rather than its total energy use, so an industrial plant running one shift a day ends up paying two to three times more per unit of power than a 24/7 data center, even though the data center drove the shortage [S1]. Residential customers feel it as a flat monthly adder: in Washington D.C., Pepco residential bills rose by an average of USD 21/month starting June 2025, with more increases already locked in through the next several delivery years [S3].
For commercial and industrial buyers, the practical levers are demand response, behind-the-meter generation, and self-supply contracts. Demand response participation in PJM's Emergency Load Response Program is paid at the clearing price, so the same cap-binding environment that hurts flat-rate customers is a direct revenue stream for participants with flexible load [S4]. Industrial buyers evaluating on-site generation should also weigh that the 2,669 MW of new and uprated capacity added in the 2026/2027 auction is a fraction of the 6,831 MW gap, and that the next major procurement, the Reliability Backstop, will set the multi-year contract template for new plants [S1][S5].
New-build economics: a few billion in subsidies could offset tens of billions in capacity costs
Analysts at Ascend Analytics argue that capacity markets will become politically unsustainable once prices stay at the cost of new entry, and that the bill can be contained only by moving part of new-build recovery off the market-clearing price [S2]. Their illustrative case for a market the size of PJM: a direct subsidy to new unit entry of a few billion dollars could reduce total capacity market costs by roughly USD 20 billion, by allowing the clearing price to be set by low-cost existing units rather than by new entrants [S2].
The mechanisms under discussion are direct subsidies or contracts for new plants, a return to cost-of-service regulation for utilities, or expanded behind-the-meter solutions for large buyers. Each changes how the cost is collected, not how much generation is needed, and PJM's own analysis indicates that lifting the cap on the 2026/2027 auction would have raised the clearing price about 70% while attracting almost no new supply [S1][S2]. For the next delivery year, capacity buyers can monitor PJM's Reliability Backstop awards, FERC dockets on cost allocation for large new loads, and the 2027/2028 BRA scheduled to clear under the same floor and cap, as the three signals that will determine whether the 6.8 GW gap narrows or compounds into the 2028/2029 delivery year.
Trackable signals for the next 90 days: PJM's Reliability Backstop multi-year contract awards for new generation, the 2027/2028 BRA floor/cap parameters, and any FERC order on whether hyperscale data-center load is carved out of the socialized capacity charge. Capacity market prices remain the headline number, but the policy fight over who pays the new-build subsidy is where the next USD 20 billion swing will land.
For component-level specifications, see powder new material, construction machinery and equipment, and lamps and light fittings.
This topic is covered further in Calculate Air Cylinder Force: Bore, Pressure, and Rod Area in Practice.