Unbuilt data centers are raising Ohio power bills

Ohio has become the testing ground for a power market idea that most ratepayers never think about: a data center does not have to exist to raise an electricity bill. Canary Media reports that Ohio is one of the top states for data center developers, with roughly 100 facilities already running and nearly 150 more in development, many of which may never be built—but their projected electricity needs are already driving up what Ohioans pay.

The mechanism is the forecast itself: utilities plan around the load they expect to serve, and a pipeline full of data center applications supplies an expectation large enough to justify spending, so the forecast arrives as a charge before a single server is ordered and stays in rates if the project disappears. The developer can walk away, but the preparation that forecast triggered does not dissolve at the same speed—and that lag is where the ratepayer's exposure lives.

The interconnection queue has become the new lease, the asset developers line up to hold in a world where power is the constraint, and Ohio is the other half of that trade. The queue is being treated in rate setting as a guarantee of future demand when it should be treated as a stack of options: some options get exercised, but many, by the outlet's telling, will not, and the bill sent to Ohioans does not discriminate between those outcomes.

The fix is not to abandon forecasting but to price the probability: load with a signed tenant and power agreement is real load and deserves capacity, while load with only an application is merchant risk that belongs on the developer's balance sheet, not in the regulated rate base. Regulators should demand that distinction and make utilities expose how many megawatts in each request come from anchored projects versus speculative ones, because the alternative is to keep treating each application as a certain customer and let the ratepayer make up the difference when the project disappears.

Until then, infrastructure investors get something close to an option written on the Ohio rate base: upside if the data center arrives, no cost if it does not. The developer who cancels loses an application; the ratepayer who pays the bill does not get a refund. Ohio's experience is a useful warning that in the power-constrained data center economy, the first bill can precede the first load.