The report also leaves out the hundreds of gigawatts of projects from independent energy developers seeking interconnection with the regional transmission organizations (RTOs) and independent system operators (ISOs) that manage transmission grids and energy markets covering about two-thirds of the U.S. population. Most of those projects won’t overcome the permitting, interconnection, and financing challenges to reach completion. But RTOs and ISOs acknowledge at least a portion of these queues in their grid reliability assessments — and the DOE is “just ignoring them in its analysis,” Pierpont said.
The DOE told Canary Media that its use of NERC data “was intentional and grounded in reliability planning principles. By focusing on only the most mature and certain projects, those with signed interconnection agreements, executed power purchase agreements, or inclusion in integrated resource plans, the DOE aimed to model a conservative yet realistic baseline. This approach is consistent with how NERC and planning authorities assess near-term reliability risks.”
The DOE’s report pairs this undercounting with an overcount of power plant retirements, Ricks said. The agency finds that 104 GW of power plants will close by 2030, a figure that combines “confirmed” retirements of about 50 GW with a slightly larger amount of “announced” retirements from NERC’s December 2024 report, he said.
The department’s figure for confirmed retirements appears to align with current data from the U.S. Energy Information Administration, DOE’s statistics arm, which is tracking 52 GW of capacity scheduled to retire by 2030, according to a Thursday note from Clearview Energy Partners.
But its inclusion of announced retirements fails to account for the fact that many power plant closure plans are being “walked back and delayed in response to load growth projections,” said Greg Wannier, senior attorney at the Sierra Club. Sierra Club opposes delays in closing polluting power plants, but effective grid reliability planning ought to take them into account, he said.
As for how much more electricity U.S. grids will need, DOE’s load growth forecast of 101 GW by 2030 is on the high end of expectations, Pierpont said, ”if not outside of the bounds” of reason. The report forecasts 51 GW of increased electricity demand from factories, electric vehicles, and broader economic growth, and another 50 GW from data centers.
“What [the DOE is] effectively doing is painting this picture of high load growth and lots of resource retirements, but assuming away the ability of utilities to build the resources they’re planning,” Pierpont said, as well as “the ability of markets to incent new resources to be built beyond those under construction or shovel-ready.”
The politics behind the analysis
To be clear, Ricks said, the criticism is “not to say there aren’t real reliability risks on the horizon.”
Utilities and grid operators are struggling to overcome the grid backlogs preventing new generation from being connected. Climate change-intensified summer heat waves and winter cold snaps are increasing the risk of electricity demand exceeding supply. And the megabill signed into law by President Trump on July 4 will make it much more expensive to keep building the solar farms, batteries, and wind turbines that now provide most of the U.S.’s new power.
But given the flaws in DOE’s methodology, it does appear that “this report seems designed from the ground up to justify keeping coal plants open with emergency orders,” Ricks said.
The DOE spokesperson told Canary Media that the report is meant to “complement, not override, the more granular, region-specific planning processes that incorporate a broader range of resources. Future updates may explore alternative scenarios with broader resource assumptions.”
However, the DOE faces a problem in using a study of grid conditions in 2030 to justify such emergency orders, Sierra Club’s Wannier said. The agency’s Section 202(c) authority is strictly limited to solving immediate and unexpected grid emergencies, not to deal with forecasted grid challenges years in the future. State regulators, environmental groups, and consumer advocates have already challenged DOE’s use of the law to keep the J.H. Campbell and Eddystone plants open.
Section 202(c) also allows the DOE to order plants to stay open for only 90 days at a time, Wannier said. The DOE’s roles in long-term grid planning, which are codified in Section 215 of the Federal Power Act, are strictly limited, ”and for good reason,” he said. “The federal government shouldn’t be stepping on the states and implementing a command-and-control, Soviet-style system.”
But Alison Silverstein, an energy analyst and former adviser to the Public Utility Commission of Texas and the Federal Energy Regulatory Commission, noted that aging fossil-fueled power plants are less reliable amid extreme weather events responsible for the most damaging grid outages.
“The plants that are most likely to close are the least likely to show up when we need them the most,” she said. “They’re very old, they require a great deal of maintenance, and they are most likely to break when we need them.”
The oldest fossil-fueled plants also tend to be the most expensive to operate, Wannier said. “We’re talking about keeping a potential fleet of hugely expensive oil, coal, gas plants around for five years, when there are untold alternatives available,” he said. “If you need new generation by 2030, build new generation by 2030. Don’t keep old generation around.”
