GA Power: Your Bill, Outages, and What the Data Says About Service

The Megawatt Mirage

Georgia Power is facing a crucial decision from the Georgia Public Service Commission (PSC) on December 19th: whether to approve a massive 10,000-megawatt expansion to support the influx of data centers. The problem? PSC staff analysts are raising serious red flags, suggesting that this buildout is largely speculative and could leave residential customers on the hook for substantial cost overruns. We're talking potential bill increases of $20 or more per month.

The core issue boils down to risk. Georgia is indeed a hot market for data centers, those energy-guzzling warehouses powering everything from AI to streaming cat videos. Georgia Power argues this justifies the unprecedented expansion. They point to 11,000 megawatts of "committed" electricity service from large customers. But the PSC staff isn't buying it, and for good reason: they claim only a fraction of this is backed by signed contracts under the new "large load" billing framework.

What happens if these data centers don't materialize as projected? We're left with stranded assets – power plants and grid upgrades that no one needs, but everyone pays for. This isn't just theoretical. The PSC staff notes that data center operators have a habit of canceling contracts or relocating to other states, and that Georgia Power's forecasts have historically overestimated data center demand.

To put this in perspective, each reactor at Plant Vogtle, the state's nuclear power plant, generates roughly 1,100 megawatts. We're talking about adding the equivalent of nine more Vogtle reactors. That's a massive bet on a sector known for its volatility.

The Shifting Sands of Demand

The numbers tell a story of shrinking commitments, not surging demand. According to a Georgia Power filing from late November 2025, their pipeline of large load economic development projects shrank by a net 6 GW (gigawatts) from Q2 to Q3 of 2025, landing at 50.9 GW. New projects only added 6.8 GW, while a whopping 14.3 GW exited the pipeline. Georgia Power’s large load pipeline shrinks by 6 GW

Here's where it gets interesting, and frankly, a little concerning. While the overall pipeline shrinks, Georgia Power highlights an increase in "commitments" from 26 to 28 customers, representing 2.2 GW of new load. They also state that five additional customers have broken ground, increasing long-term load from 3,721 MW to 7,313 MW.

But what does "commitment" really mean? Is it a legally binding contract with penalties for backing out, or a handshake agreement and a vague promise? The PSC staff clearly suspects the latter. Their testimony points out that the majority of the proposed generation isn't backed by executed contracts under the new large load framework. Only about 1,900 MW is supported by such contracts. The rest is speculative.

I've looked at hundreds of these utility filings, and the language here is crucial. "Commitment" is a marketing term, not a legal one. The discrepancy between the shrinking pipeline and the touted "commitments" suggests a potential overreliance on projections that haven't materialized into solid deals. Since the 2023 IRP Update, 33 data center projects with 11,332 MW of announced load have been removed from the pipeline, representing roughly 55% of all project removals. It's unclear whether Georgia Power's model accounts for this level of project removal.

The Customer's Burden

The PSC staff recommends approving only about a third of the proposed buildout (3,100 MW) and approving another 4,200 MW with conditions. They argue against adding the remaining 2,400 MW entirely, deeming those projects the most expensive.

Georgia Power, of course, disagrees. They argue that there are options to mitigate risk, such as adding new contracts or delaying construction. They also point to new billing terms designed to protect non-data center customers. However, the PSC staff isn't convinced. They note that most existing contracts were signed before the new billing rules took effect, meaning there's no guarantee these costs won't be passed on to residential customers.

And here's the kicker: the PSC staff alleges that Georgia Power stands to profit "tremendously" from this expansion, regardless of whether the data centers actually materialize. The proposed additions would nearly double the company's "rate base," a key factor in determining its earnings. In essence, Georgia Power gets paid to build, even if what they build isn't needed.

The question becomes: who bears the risk? Is it Georgia Power, or is it the average homeowner struggling to pay their monthly bills? The PSC's decision on December 19th will determine the answer.

So, What's the Real Story?

This whole thing smells like a classic case of regulatory capture. Georgia Power is incentivized to build, regardless of actual demand, because it directly boosts their profits. The PSC is supposed to be a check on this behavior, but the revolving door between the utility and the commission raises serious questions about their independence. Until we see ironclad, legally binding contracts – not just "commitments" – backing this expansion, customers should be very, very worried.