The minimum bill in Ohio and Virginia
On July 9, 2025, the Public Utilities Commission of Ohio approved a new tariff for data centers served by AEP Ohio, the utility covering Columbus and much of central Ohio. The commission adopted a settlement supported by the utility, commission staff and consumer and industrial groups. New data centers with demand above 25 megawatts would pay for at least 85 percent of the capacity they contract for, whether they use it or not, under agreements that can run for up to twelve years, with an exit fee if they leave early. Existing facilities were left under the old rules. The same order required AEP Ohio to lift a pause on new data center service agreements that it had imposed in central Ohio in 2023. The Data Center Coalition, which represents large operators, opposed the design.
Why would a regulator want a customer to pay for power it does not use? The answer lies in how utilities recover their costs. When a utility builds a substation or a transmission line to serve a new customer, it recovers that investment over decades through rates. If the customer shrinks its plans or leaves after a few years, the cost does not disappear. It is spread across everyone else. Economists call this stranded cost risk, and it has become the central worry in data center rate cases.
Data centers make the risk unusually sharp for three reasons. They are large relative to the systems they join, so a single cancellation can move a utility's forecast. Their developers often file requests in several utilities at once and build only where power comes first, so some of the demand a utility sees is speculative. And the technology driving demand is young enough that nobody can say with confidence what a facility will need in ten years. A minimum bill is the regulator's way of making the customer carry the risk it creates.
Virginia arrived at a similar place by a different route. On November 25, 2025, the State Corporation Commission, Virginia's utility regulator, issued its final order in Dominion Energy Virginia's biennial rate review. It created a new rate class, called GS-5, for customers with demand of 25 megawatts or more, taking effect on January 1, 2027. Those customers must pay a minimum of 85 percent of their contracted transmission and distribution demand and 60 percent of their contracted generation demand, under contracts of fourteen years, with collateral requirements and advance notice before reducing load. The commission also approved a base rate increase for Dominion that was smaller than the utility had requested.
The two decisions share a structure worth naming: a size threshold that separates large new loads from ordinary customers, a minimum demand charge expressed as a share of contracted capacity, a long contract term and protection against early exit. Faraday's read is that the structure is spreading: Indiana Michigan Power's large-load settlement, approved by the Indiana Utility Regulatory Commission in February 2025, pairs a size threshold with minimum payments on contracted capacity and long contract terms. The numbers differ; the logic is consistent.
It is fair to ask who wins under these rules. Households gain some protection from paying for infrastructure built for customers who later vanish. Utilities gain confidence to build. Data center developers lose some flexibility and must commit capital earlier and more firmly. Whether the designs strike the right balance is exactly what the parties argued about, and reasonable people disagreed. In Ohio, consumer and industrial groups signed the settlement while the Data Center Coalition argued that it unfairly singled out one industry and would push investment toward states with friendlier terms.
Faraday's read is that minimum bills sort serious projects from speculative ones more effectively than any queue rule. A developer who signs a fourteen-year commitment to pay for 85 percent of its contracted capacity has told the utility something that no press release can. For someone reading a jurisdiction, the existence of a large-load tariff is therefore not simply a cost. It is evidence that the utility has a way to say yes.
The practical point for an investor is to read the tariff before the incentive package. Tax breaks are easy to announce and easy to compare. The terms on which a utility will actually serve a large load, and what happens if the load changes its mind, often matter more to the economics of a site over its life.