Georgia Power: rising bills and the shrinking capacity outlook
Georgia Power's Data Center Dream: A Numbers Game or a Customer Nightmare?
Georgia Power wants to build big. Really big. We’re talking about an ambitious plan to tack on an additional 10,000 megawatts (MW) to its power generation fleet over the next five years. For perspective, that’s roughly nine times the output of a single nuclear reactor at Plant Vogtle, a facility whose cooling towers still cast long shadows over Burke County. The utility’s pitch? This massive expansion is essential to fuel Georgia’s booming data center industry, which, they argue, demands electricity on a scale comparable to entire cities. It sounds like progress, a testament to Georgia’s digital future, right? But peel back the corporate gloss, and the numbers tell a far more complicated story.
The Georgia Public Service Commission (PSC) staff, the folks whose job it is to scrutinize these proposals, aren’t buying the whole narrative. Their analysts have delivered a stark, almost chilling, warning: existing residential customers could see their monthly bills jump by $20 or more if the PSC greenlights Georgia Power’s full data center buildout. Robert Trokey, the PSC’s electric section director, along with outside consultants, didn't mince words in their testimony. They contend that only a fraction of what Georgia Power is asking to build—about a third, or 3,100 MW, to be exact—is actually backed by signed contracts. The rest, they conclude, is "speculative and exposes customers to the risk of stranded costs if the anticipated load does not materialize." This isn't just a minor disagreement; it's a fundamental clash over who bears the risk in Georgia's digital gold rush.
The Slippery Slope of "Commitments"
Here’s where the numbers get really interesting, and frankly, a bit unsettling. Georgia Power maintains that the demand is anything but speculative, pointing to a "portfolio" of large customers who have "committed" to receiving electricity service, totaling 11,000 MW. They insist their load forecast is robust, based on these commitments and "organic growth." But let's apply a little analytical rigor here. Recent filings reveal that Georgia Power’s pipeline of large load economic development projects shrank by a net 6 gigawatts (GW) from Q2 to Q3 2025. To be more precise, a staggering 14.3 GW of projects exited the pipeline in that quarter. Meanwhile, only 6.8 GW of new projects entered. This isn't a minor fluctuation; it’s a significant contraction. As reported by Utility Dive, Georgia Power’s large load pipeline shrinks by 6 GW.
How can the utility claim a growing number of "commitments" when its overall pipeline is actively shrinking, with projects bailing out at an alarming rate? This is the part of the report that I find genuinely puzzling. It feels a bit like trying to fill a leaky bucket by only counting the water you pour in, ignoring the torrent gushing out the bottom. My analysis suggests a critical methodological critique is in order: what constitutes a "commitment" in Georgia Power's model, and how precisely does it account for the high churn rate—the fact that thirty-three data center projects representing 11,332 MW of announced load have been removed from the pipeline since 2023? That’s roughly 55% of all project removals, and 65% of all announced load removed. If 24% of data center projects that entered the pipeline have subsequently been removed, with an average of five exiting each quarter, how can the underlying forecast be considered sound? It suggests Georgia Power might be building a stadium for a team that keeps losing players to free agency before the season even starts.
The PSC staff's testimony is explicit: data centers are "primarily underperforming expectations due to a mixture of lower materialization rates, project cancellations, and delays." This has already happened in Georgia, and the risk of overbuilding for electrons that have no place to go is high. If Georgia Power builds these plants, and the data centers either scale back or vanish, existing customers are stuck paying for the infrastructure, whether it’s used or not. It’s like being asked to chip in for your neighbor’s swimming pool, only for them to fill it with dirt and grow a vegetable garden, while you're still on the hook for the filtration system.
The political backdrop here is impossible to ignore. This staff warning drops on the heels of a historic election where two Republican incumbent PSC commissioners, Fitz Johnson and Tim Echols, lost their seats. Customer frustration over rising utility bills was the central issue, a sentiment I’ve tracked in online forums as a clear, quantifiable pattern of discontent. And guess what? Those two outgoing commissioners get to cast their votes on this critical case on December 19th, less than two weeks before their Democratic replacements take office. The air in the PSC hearing room on that day will likely be thick with the unspoken weight of political legacy and customer anxiety. Georgia Power claims there are "options" if demand falters, like delaying construction or finding new contracts. But the PSC staff argues that the existing "guardrails" designed to protect non-data center customers aren't enough, especially since most contracts were inked before new billing rules took effect. They state, unequivocally, "there is no guarantee those costs will not be passed on to existing customers." And who stands to gain? Georgia Power, of course, which stands to profit "tremendously" from the expansion, potentially nearly doubling its "rate base"—a key determinant of its earnings.
The Customer's Uneasy Future
The PSC staff recommends approving only a third of the proposed buildout, with conditions on another 4,200 MW, and outright rejecting the remaining 2,400 MW—the most expensive projects. This isn’t a small adjustment; it’s a fundamental re-evaluation of the utility’s ambition. While customers likely won't feel any direct bill increases from this data center expansion before 2029 (the PSC voted to hold base rates steady until then), other cost decisions are looming next year, like storm repair and fuel costs. The cumulative effect, coupled with this potential data center gamble, paints an uneasy picture for the average Georgian trying to keep their lights on. The question isn't just if Georgia Power will build, but who will ultimately pay for the speculative bets on the digital frontier.