Behind the Meter: Why Big Tech Is Buying Nuclear Power Plants
With utility interconnect queues stalled for up to 7 years, gigawatt-scale AI clusters cannot wait for regional grid upgrades. The hyperscaler workaround? Park the server halls directly behind the reactor meter to pull 24/7 baseload generation straight from the turbine.
1. The Interconnection Wall
Hyperscalers are sitting on hundreds of billions in AI CapEx, but they have hit a physical wall. The traditional process of building a data center and requesting a grid interconnection from a regional utility now takes anywhere from 5 to 7 years.
For companies racing to train next-generation models, waiting half a decade for a regional grid operator to upgrade transmission lines is a non-starter. They need gigawatt-scale power immediately, and the public grid simply cannot deliver it.
2. The Behind-the-Meter Bypass
Enter the nuclear pivot. Deals like Amazon's acquisition at the Susquehanna Steam Electric Station and Microsoft's contract to restart Three Mile Island aren't just about clean energy PR. They are about physics and speed.
By co-locating data centers "behind the meter"—physically plugging the server halls directly into the nuclear plant's switchyard before the power hits the public grid—hyperscalers completely bypass the regional transmission queues.
3. The FERC Blowback
This strategy has immediately triggered an existential turf war. Public utilities are actively petitioning the Federal Energy Regulatory Commission (FERC) to block these deals, arguing that Big Tech is siphoning off existing baseload generation, leaving everyday ratepayers vulnerable to grid instability and price hikes.
The battle lines are drawn: will the government allow private tech monopolies to corner sovereign nuclear assets, or will they force AI to wait in line with the rest of the public grid?