By Tom Rutigliano
PJM’s resource adequacy woes have taken on an air of inevitability. PJM warned of impending shortfalls three years ago, and since then we’ve all been watching it like a slow-motion train crash. Consumer demand for power feels unstoppable, but supply is frozen, unable to respond. Interconnection queue reform, record high prices and special purpose fast tracks have all so far been unable to deliver the new capacity the region needs.
Now PJM, with prompting from 13 governors, is trying another solution: the “Reliability Backstop Auction.” While details still are being negotiated, this boils down to throwing money at new power plants. But it’s not clear that money is the problem right now — after all, tech companies have famously deep pockets. (See White House and PJM Governors Call for Backstop Capacity Auction.)
The backstop auction itself isn’t a bad idea. It recognizes the commonsense fact that financing new power plants is different than covering the operating costs of existing ones. If done right, this offers new supply attractive terms without raising prices for everyone else.
That fixes the worst-of-both-worlds problem we have now: Rates are going up by billions, but that money is going to windfalls for existing power plants instead of investment in new ones. If PJM can design a backstop auction that doesn’t put costs or risk on the public and is open to all technologies on fair terms, they might have a winning formula on their hands.
But even the best auction won’t work unless new power plants can get built, which is still a huge problem. If the many other problems blocking new supply aren’t solved, this new auction could fail. (See Government-proposed ‘Backstop’ Auction to Test PJM Stakeholder Process.)
Disappointing Queue Reform
For many years PJM’s interconnection queue was the problem. But following reforms, PJM’s queue began to approve projects in 2025. What should have been great news turned into disappointment as developers discovered their projects will have to wait four years or more for transmission upgrades.
PJM’s next batch of projects isn’t doing much better — almost a third of the capacity in that group dropped out when they got their first estimate of interconnection costs. The hand-picked “shovel ready” projects in the RRI fast track dropped out at about the same rate.
Behind the queue is a deeper problem: The transmission system isn’t ready to accept tens of gigawatts of new generation. If PJM doesn’t deal with this, there won’t be enough new supply for the Reliability Backstop Auction. This isn’t a problem PJM can build its way out of. At best, the transmission we need can’t be completed for many years, and transmission projects often run into serious obstacles that delay or cancel them.
Instead, PJM needs to embrace solutions that work around the limits of the electrical grid we have. This is a major paradigm shift for Valley Forge; for as long as PJM has been around, it has insisted on “full deliverability,” which means that every generator in PJM must be able to supply power to the entire region, and not be bottled up in a small area.
This made sense in an era of diffuse demand growth, but what we’re seeing now is generators built to supply individual data centers in specific locations. If PJM recognizes this, and the generators and data centers accept the risk that comes with not being fully integrated into the larger grid, we could unlock gigawatts of faster supply.
Expand the Co-location Model
This already has begun. FERC recently approved an SPP proposal along similar lines, and, spurred by a recent FERC order, PJM has come up with some innovative solutions for generators located next to data centers. For the most part, these are variations on the idea of partial and/or as-available service rather than “full service all the time.”
This is a great start. PJM now needs to expand this model and open opportunities for new supply sited where the transmission system can deliver its power to particular customers. Imagine a storage development a few counties away from a data center that charges when there’s spare grid capacity and discharges to offset any overloads the data center would place on the interstate grid. If carefully sited, an arrangement like that brings the storage online years earlier by avoiding transmission upgrades.
Old approaches are just not up to the task of powering the data center boom. We don’t have and can’t build an electrical grid that ships power from distant fossil plants. Now is the time for bold innovation around carefully sited projects, especially energy storage, that can quickly work with and around the limits of the current grid. Without changes in how PJM interconnects new supply, no auction will be successful, no matter how well designed.
Tom Rutigliano is senior advocate for climate and energy at the Natural Resources Defense Council.



