Big Tech's AI Boom Is Supercharging Old Nuclear Plants
Your electricity bill and Google's AI could both depend on this nuclear upgrade.
Nuclear operators are racing to squeeze more power out of existing reactors. Constellation announced plans to spend more than $4.3 billion on uprates at 11 reactors, enough to power roughly 700,000 US homes, and Google agreed to buy electricity from those plants under a 20-year deal. The Department of Energy also said it will commit over $4 billion to help finance upgrades at Vistra plants across Pennsylvania and Ohio, adding 433 megawatts, with Meta already signed on to buy power.
The context matters, though. The US nuclear fleet has roughly 98 gigawatts of capacity today, and the total potential from uprates is somewhere between three and eight gigawatts. PJM, the grid operator covering much of the East Coast, expects about 30 gigawatts of additional peak demand from AI by 2030—so all the possible nationwide uprates would cover just a fraction of that. Making the most of existing reactors is a relatively easy win, but if we want more nuclear power to meet demand, eventually that will require actually building new reactors.
- Nuclear plants are being upgraded to produce more power for AI data centers.
- Google and Meta are buying the extra electricity, making the projects financially viable.
- Even all upgrades won't meet the expected 30 gigawatt demand increase from AI by 2030.
Why It Matters
More clean nuclear power could keep electricity reliable and prices stable as AI demand surges.