30,000 Megawatts: Who Pays When the Data Centers Move In

Thirty thousand megawatts is an abstract number until you translate it. It is more than a dozen large power plants' worth of demand, and AEP Ohio has cited more than that much in requests to connect to its grid. Most of it is data centers: the enormous, power-hungry buildings that run the cloud and, lately, artificial intelligence.

That demand does not arrive for free. Serving it means new lines, new substations, new capacity, and all of that costs money to build. The only real question is who pays. That question has a boring, decisive answer, and the answer is a tariff.

What a tariff actually decides

A tariff is a pricing rule a utility files and regulators approve. It sounds like paperwork. In this case it is the difference between a giant new customer paying for the grid it demands and that cost being smeared across every household in the territory.

Here is the trap the rule is trying to avoid. A data center announces it wants to connect. The utility builds capacity to serve it. Then the project shrinks, or delays, or never fully materializes, but the infrastructure is already built and someone has to pay for it. If the rules are loose, that someone is you, on your monthly bill, subsidizing capacity that a corporation asked for and then didn't use.

The 2025 rule, and the fight over it

In 2025, PUCO approved a tariff meant to close that trap. It requires large loads, users over 25 megawatts, to pay for at least 85% of the capacity they subscribe to, locked into long-term contracts. In plain terms: if you tell the grid you need this much power, you commit to paying for most of it whether or not you end up using it, and you commit for years, not months.

That is a real attempt to keep the buildout's cost off households. It puts the risk of an overbuilt grid on the company that asked for the build, which is where it belongs.

Whether it holds is a live fight. Big customers do not like being told to pay for capacity they might not use, and they have lawyers and lobbyists to say so. The tariff can be challenged, softened, carved with exceptions, or undermined at the next rate case. A rule that protects ratepayers on paper only protects them if it survives contact with the people it costs money.

This is the same question HB 6 got wrong

Ohio has been here before, and recently. The HB 6 scandal was, underneath the bribery, a fight over exactly this: whether ratepayers would be forced to subsidize generation that couldn't pay its own way. A utility wanted the public to carry the cost of plants the market had rejected, and it bought a law to make that happen.

The data-center tariff is the same structural question with the corruption stripped out. Who pays to keep the lights on for a favored set of large customers, the customers themselves or the households around them? The difference this time is that the decision is being made in the open, through a regulatory process, rather than in the dark through a bribe. That is progress. It is not a guarantee.

Why it runs through the ballot

The commission writing and defending this tariff is PUCO, and PUCO's commissioners are appointed by the governor. That is the whole reason a regulatory rule about megawatts belongs in an election story.

A commission inclined to protect ratepayers writes and enforces a tariff like the 2025 one, and holds the line when the big customers push back. A commission inclined to accommodate industry finds reasons to loosen it. Both are staffed by the same appointment power, held by whoever wins the governor's race. You do not vote on the tariff. You vote for the person who picks the people who decide the tariff.

The data centers are coming either way. The grid gets built either way. The only thing genuinely undecided is whose bill absorbs the cost, and that is being settled right now, in a regulatory docket most Ohioans will never read, by officials an election puts in place. When your electric bill moves in the next few years, this is the fight it will have come from.