PJM’s last two capacity auctions each cleared roughly 7 GW short of the grid operator’s reliability target, and the board has now directed staff to file a Reliability Backstop Procurement plan with FERC that would run an emergency auction from September 30 to October 21 at a price cap of $555/MW-day, up from the prior cap of $325/MW-day in the base market. The same initiative also asks for more dispatch control over large customers: loads that have not secured sufficient capacity can be ordered to reduce consumption before an operating emergency hits. Taken together, the two measures are PJM’s official concession that its forward capacity market no longer does the reliability work by itself – and that the giant load growth of the 2030s will force the grid into emergency procurement and demand-side management as a structural policy, not a one-off patch.
Why two consecutive auctions fell about 7 GW short
The PJM capacity construct is a forward, product-driven market: resources bid capacity for a delivery year, the auction sets a single clearing price, and loads pay for that commitment through transmission and capacity charges. When the auction clears below the requirement, parts of frontier capacity layers were not retained, and the shortfall falls directly on the reliability margin, which the operator increasingly relies on in peak hours. The first 7 GW miss was already in the constrained system; the second is not just a repetition but a new marker – the board is now replacing the traditional adjust-and-retry cycle with a separate, time-boxed backstop that acts in reserve slot.
Two structural forces have compressed both sides of the bid stack. A full 15 GW of baseline resources has retired in PJM’s footprint since 2022, pulled by economics, unit anniversaries and plant life decisions more than by fuel prices. At the same time, large load – data centers, electrical industrial, and announcements of the order of new 70 GW by 2038 – is stepping into the projection, while much of the physical interconnection and construction of that new capacity is still scrolling through the queue. When the large-bulb wave slows, the auction has less retained supply to bid at the confining cap; capacity costs start rising more for the same MW than any real market expectation.
In that new equilibrium, the old cap of $325/MW-day simply leaves no room for the last tranche of resources – some peakers, using their marginal cost of staying open, and some front-end storage projects with a price that would cover recovery only at a higher ceiling. The $555/MW-day emergency cap is not so much a new pricing rule as a new market: at that level, resources that were bid off in the base auction can re-enter and be paid without breaking the whole clearing mechanism. The short window – four weeks to final terms – is also very deliberate, producing what the design already calls a backstop, not a wholesale auction cycle.
The economics of a $555/MW-day topop: what it expands for ratepayers and developers
If the emergency auction were to actually clear the entire 7 GW shortfall at the new cap, the annual capacity bill for that piece alone would be on the order of $420 million more than the same volume would have cost at the old $325 cap – roughly half a billion dollars per year, bus. Put differently, at a 7-GW replacement programme, backstop capacity will cost about $200 per kW-year at max price, roughly before the auction has historically payback at intermittent values. The result of this steep step is exactly the signal that a resource in the gap starts to look like real insurance: at those numbers, peaking capacity, existing oil less flexible plants and longer-duration storage all can be a significantly better business.
For a large customer, the second instrument is equally important. The “primary field” that insists on large loads (equal to a contracted capacity amount) effectively converts a part of the load curve into an option: if the electricity retailer has not bought the MW, the operator can push curtailment early – before, say, an extreme weather event. That fits the wider subsidy-by-another-name trend. PJM is responding to a supply-side context that is now common to every corps in the eastern U.S.: fossil closing schedules met rising data-center demand peaks have made interruptibility an economically tenable first-order option. It is effectively a rule that forces large demand to become the last line of defense when the market price is not looking.
Look at the consequences in the real budget. If 2026-27 delivery year follows the same pattern, a 100-MW industrial facility paying a backstop clearing price could face a capacity cost on the order of $25 million/yr more than in the post-2020 era – and the load-curtailment armature forces it to pay for that capacity or formally go to the resilience curve. For it, backstop pricing is already functionable as a revenue floor; for rate maker rate design, the gap between $325 and $555 is nearly the total issue in policy-making.
Who the plan hits first
- Utility and distribution planners must now add a second-ee auction into short-term resource plans: every monthly target will – the capacity shortfall can reach beyond the backstop window and the tariff rider update earlier than in legacy.
- Peak capacity, gas and storage developers finally have a credible revenue signal: a dispatchable asset that can offer real backstop MW against the base 7 GW gap is a clear card in the coming do‑the‑math push, especially in Northwest zones before winter.
- Data center and industrial load owners must now buy or clear pivots, because the “sufficient capacity” test means the price of substitution is high – either anticipate direct pay capacity at FERC-driven caps, or plan the control of the physical load and the operator dispatch.
- Policy analysts and state regulators should read this as a tests of FERC’s willingness in a new price form: the agency decision on the Reliability Backstop will inevitably be the precedent used by the next load-driven bidding in other regions.
What to watch next
- FERC order on the PJM filing – whether the reliability backstop is approved in a that runs, on the Sep 30 window, and if there are more conditions (fuel guarantees, completion deadlines) added, changes the whole trajectory.
- The quantitative auction exit from September 30 – and the process of the 7 GW that actually underwrites and the clearing price relative to the cap: whether <¨ shows $555 is a true cap or the real contract.
- The definition of “sufficient capacity” in the large-load provisions – statement order: where the threshold is set, how many GW of new data centers jump above it, and whether curtailment is actsiumized along with who carries the value.
- Next full base auction – if an new annual sign-of-success has actually repaired the market, the next auction gap will shrink; if it remains at 7 GW, the backstop becomes a permanent gate for the year and the 70-GW wind by 2038.
Bottom line
Securing the grid must now clear the price – the base auction is no longer the last word, but the emergency backstop at $555 with shoulder-load curtailment is the cheapest area that says reliability is a customer’s preference and their people at the same time. For PJM’s 70 GW decades, this is the defining order: either a $550 per MW-day set of capacity is yours for the next year, or the lights are for event h. Those are now the only two real paths in large-load governance.
Read the full report at Energy Central.
Note: facts and figures attributed above to reflect that outlet's original reporting. Broader context, cross-sector connections, and forward-looking scenarios reflect independent analysis by our editorial team.
About this article: Drafted by Energy Ai with AI-assisted research and writing based on public reporting, then reviewed under our editorial process before publication.
Leave a Reply