Georgia Power adds 1,137 MW of solar through other people's balance sheets

Georgia Power has approval for 1,137 megawatts of solar power purchase agreements, Renewables Now reported on Sept. 11, with no counterparties, no prices, and no body identified as granting the approval. The omission has become the house pattern.

In August this publication covered the 3.2GW grid slot OpenAI is underwriting, a build in which the customer pays full freight and the utility constructs against the promise. The same month, Harbert's Georgia cogeneration acquisition paired a Georgia Power contract with thermal sales to the Frito-Lay plant next door. Both structures put someone other than the utility on the construction side of the meter, and the solar approval reported roughly two weeks later looks like the same trade at larger volume.

A power purchase agreement is output the utility buys rather than capacity it owns, so the capital cost of the 1,137 megawatts lands on developers' balance sheets while Georgia Power books the capacity as procurement. That is an efficient way to add supply and a slow way to build an asset base, and it suggests where the utility thinks its leverage sits. In a territory where new load arrives in gigawatt increments, the scarce asset is the interconnection right, and the cheapest way to monetize one is to let a third party finance the panels inside it. The OpenAI arrangement is decent evidence that Georgia Power's territory holds more load than it holds cheap ways to serve it.

As this publication has argued, a milestone without a price is not capital allocation: an approval covering 1,137 megawatts is a procurement target with a date attached. Steel in the ground is a different milestone, and the market has spent two years treating completion as a financing event rather than proof of infrastructure—PPA approvals deserve the same skepticism until strike prices and counterparties are on the page.

The commercial question is who signs, and nothing in the reporting yet distinguishes the two possibilities. If the counterparties turn out to be one or two very large offtakers, Georgia Power will have replicated its hyperscaler structure across the solar book, with a single credit standing behind a portfolio of contracts; if they are a spread of independent developers selling on standard terms, the approval is ordinary procurement and the story is volume rather than structure. For a utility that has already shown it will let a customer underwrite its generation, the first executed contracts will settle whether the 1,137 megawatts are a supply addition whose economics Georgia Power controls, or a queue position it has learned to rent out.