PJM's emergency capacity fix delegates the hard problem to states as queue conversion risk mounts
PJM's board proposed a limited emergency auction to cover a 6.8 GW capacity shortfall, but the durable supply problem lands on states and bilateral channels with a poor track record.
PJM's board moved on Wednesday (2026-08-05) to plug a 6.8 GW shortfall in its last capacity auction through a limited emergency procurement, according to Canary Media. Rather than restructuring how large loads connect to the grid or waiting for its scheduled centralized backstop, the board delegated the longer-term supply problem to individual states and bilateral contracts.8
The scale of what was acknowledged deserves attention. The 6.8 GW gap sits in a grid serving 67 million people across 13 states and Washington, D.C. Data centers alone drove $6.3 billion of the $16.4 billion in charges from PJM's most recent capacity auction, representing 38% of total system costs attributable to a single demand category, according to Monitoring Analytics president Joseph Bowring, as reported by Utility Dive.7,4
PJM spot power at the Western Hub was marked at $62.49/MWh on Wednesday (2026-08-05). That contained level sits against a board decision that effectively acknowledged the grid cannot currently match supply to the demand trajectory it faces. The bearish consensus on PJM real-time prices rests partly on the volume of generation visible in the queue. The queue numbers are large. But cleared-queue and built are two different things.8
Only 13% of the capacity that submitted interconnection requests from 2000 to 2019 had reached commercial operations by the end of 2024, with 77% withdrawn, per IEA data cited by QZ. Over 55 GW of new generation has cleared PJM's interconnection queue and is ready to build, with another 220 GW in the latest review cycle, per Electric Power Supply Association figures. When PJM asked developers whether they were willing to contract directly with large loads, over 130 GW came forward. That figure sounds large until weighed against the 77% historical withdrawal rate for interconnection requests.1,5
The bilateral channel now bears considerable weight in the board's response, but its delivery record is thin. Canary Media reported that bilateral contracts between individual data centers and generation developers represent a potential major channel — yet no contracted volume has been publicly reported to match the 130 GW of developer interest. State-level delegation introduces political and regulatory friction that has historically slowed coordinated supply response.6,8
Demand is not holding still while the institutional machinery works through those delays. PJM saw electricity demand surge to roughly 163 gigawatts on Thursday (2026-07-02) during a heat dome that pushed heat indices past 110 degrees from Washington to New York, falling just short of the grid's 2006 all-time peak of 165,563 MW — even after PJM had forecast load could exceed 166,000 MW that day, OilPrice.com reported. Global electricity demand from data centers grew 17% in 2025, with AI-focused consumption up 50%, according to the IEA. Data centers account for roughly half of U.S. incremental demand growth, the IEA's global energy assessment found.4,1
Transmission infrastructure compounds the supply-side drag. Regional grid operators including PJM requested extensions on a federal FERC deadline to upgrade transmission capacity, a delay with direct implications for how quickly new supply can reach the grid, Datacenterdynamics.com reported. Grid Strategies projects the U.S. data center market will need between 65 GW and 90 GW of additional capacity by 2029 — a horizon close enough to matter for anyone pricing medium-term power contracts in PJM territory.2
PJM had already pulled its backstop reliability auction forward to September, moving it up from 2027. "Waiting until 2027 to execute the centralized procurement, considering the expedited need, is not prudent," the grid operator said, as reported by E&E News. September will test how much supply developers are prepared to commit at current economics.3
The bearish case leans on queue volume and backstop timing as evidence the gap will be managed. The supply pipeline does look deep on paper. But 77% historical withdrawal rates suggest the paper figure overstates what will actually get built, state-level delegation slows any coordinated response, and the emergency procurement covers only the immediate 6.8 GW hole rather than the structural growth in demand.8,51
Thin clearing volumes at September's backstop auction, or a wide gap between what developers commit and what PJM actually needs, would validate the concern. Watch also whether any state-level bilateral deals close before year-end; developer interest at 130 GW remains unmatched by reported contracted volume, and that gap is where the supply story either holds or breaks.3,6