Google to Pay for Georgia Nuclear Reactor Upgrades — SkimNews

Get the Energy newsletter
Daily energy & climate — solar, EVs, oil, the policy fights and tech bets shaping the transition. Free.
- Georgia Public Service Commission approved a deal letting Google and Georgia Power upgrade four older nuclear reactors at Plant Vogtle and Plant Hatch, with work starting in 2028 to add 96 megawatts of capacity to the grid.
- Google will pay a subscription fee on top of its electricity bill to fund the upgrades, earning credit toward its goal of running entirely on carbon-free energy by 2030, with the deal projected to deliver approximately $900 million in customer benefits over the units' lifespan.
- Georgia Power framed the subscription model as a way large-load customers can help finance clean energy that benefits all ratepayers, with SVP of strategic growth Aaron Mitchell calling it an example of how "large-load growth can benefit all electric customers."
- Google separately announced a partnership with Constellation Energy to fund nuclear upgrades at reactors in Illinois, New Jersey, and Pennsylvania, adding 890 megawatts of capacity to the PJM grid, the country's largest grid operator.
- Amazon has announced a parallel deal to add capacity to a Constellation nuclear plant in Maryland, underscoring how multiple hyperscale data center operators are turning to existing nuclear sites.
- Nuclear upgrades sidestep the multi-decade timeline for new plants — roughly five years instead, per the Corporate Energy Buyers Association — and avoid the supply-chain delays now hitting natural gas turbines that utilities are racing to install for data center demand.
Why it matters: Large tech customers are now directly financing nuclear upgrades that previously fell to ratepayers, with Google committing to a 96 MW subscription model in Georgia (worth ~$900M in projected benefits) and an additional 890 MW through Constellation across three more states — a concrete alternative path for utilities facing multi-year waits for new gas turbines while hyperscale data center load keeps climbing.
Ask SkimNews



