FERC told PJM Interconnection its data-center power plan was "deeply flawed" and sent it back to the drawing board, with a new proposal now expected by the end of February, according to Bloomberg. PJM responded by suspending its planned capacity auction for meeting surging data-center demand entirely, per a separate Bloomberg report. Analysts at TD Cowen didn't mince words: the regulator "punted" on a critical policy overhaul.
Here's the civilizational cost of that punt. Every month PJM spends relitigating its own tariff design is a month the largest grid in the country isn't building the interconnection pathways that AI compute, manufacturing, and electrification all need simultaneously. This isn't an isolated PJM problem — it's systemic. FERC issued show-cause orders in June to six major grid operators, including MISO, SPP, CAISO, ISO-NE, and NYISO, demanding they justify or reform how they integrate large loads, with FERC Chair Willie Phillips framing the stakes bluntly: America's ability to lead in advanced industries "hinges on the speed with which large energy loads can safely and efficiently plug into the grid," according to EnergyPulse.
I've written before about PJM's interconnection queue as the real bottleneck behind every gas and nuclear headline this year. What's new is the sequencing failure: the regulator meant to speed things up just slowed its biggest grid down by five months at minimum. That's the paradox of reform-by-committee — you can order six grid operators to fix large-load rules, but if the fix itself requires another round of filings, comment periods, and resubmissions, you've added friction in the name of removing it.
The abundance case doesn't need PJM to be perfect. It needs PJM to be fast. Right now it's neither.