FirstEnergy’s Ohio Data Center Tariff Sets an 85 Percent Minimum-Demand Floor and Writes Rules for Behind-the-Meter Self-Supply
FirstEnergy’s Ohio utilities filed a proposed large-load tariff on June 15 that bills data center customers on the higher of 85 percent of their contracted capacity or their actual usage. The filing does something most rate cases do not: it includes explicit provisions governing what happens when a customer supplies part of its own load with behind-the-meter generation.
Both features matter beyond the data center class they target.
The floor. Under a conventional demand charge, a customer pays for the highest interval of consumption in a billing period, so cutting that peak cuts the bill. A minimum-billing-demand floor severs part of that link. The tariff sets contracted capacity by agreement between the customer and the utility, then charges no less than 85 percent of it, or actual usage if that runs higher. A customer that trims actual demand below the floor keeps paying the floor.
For a facility running a flat load near its contract level, the floor changes little. For any asset whose value comes from shaving peaks below that level, the floor is a ceiling on savings.
Self-supply provisions. The tariff permits a customer to offset its contracted capacity with on-site generation, and attaches explicit conditions to how that self-supply is allowed to count. The reduction in billed capacity is not automatic: it is governed by the tariff’s behind-the-meter provisions rather than left to the customer’s discretion. This is a utility writing, in tariff language, the terms under which behind-the-meter supply may be applied against a customer’s obligation. The generation is allowed. The credit against billed capacity is conditioned.
The pattern. The filing is not an isolated Ohio experiment. Its take-or-pay structure mirrors tariffs already moving through other states, according to EnergyChoiceMatters: Oregon, where Portland General Electric’s Schedule 96 isolates large-load costs; North Carolina, where Duke Energy has proposed a data center tariff; and Montana, where NorthWestern Energy has filed comparable terms. The specific tool varies by state, but the design intent is consistent. Force the largest new loads to pay for the capacity they reserve, whether or not they use it, and stop them from shifting that cost onto other ratepayers.
The standard reading of these tariffs is that they protect smaller ratepayers from hyperscaler-driven grid buildout. That reading holds. The 85 percent floor exists because a data center can sign up for hundreds of megawatts of firm capacity, prompt the utility to build transmission and distribution to serve it, then ramp slower than promised and leave the fixed cost stranded on everyone else. The floor makes the reservation binding.
The effect on storage economics. The prevailing argument for behind-the-meter batteries in commercial buildings is that demand charges keep rising and a battery that shaves the monthly peak captures the difference. Minimum-demand floors interrupt that logic for the customers they cover. If a share of the bill is fixed at 85 percent of contracted capacity, peak-shaving below the floor produces no demand-charge savings on that share. The battery still arbitrages energy prices and can still provide backup, but the demand-charge case, the one that usually underwrites the investment, weakens in proportion to how high the floor sits.
That effect is narrow today. The tariff applies to large-load data center customers. It does not reach the office towers, hospitals, cold-storage warehouses, and multi-tenant retail that make up the commercial demand-charge market. Those customers sit well below the large-load thresholds these tariffs target, and their lumpy, peaky load profiles are exactly what peak-shaving is built for.
The precedent. Utilities are learning to draft tariff language that does two things at once. It caps the value any behind-the-meter asset can extract, through the floor, and it governs how self-supply is allowed to count, through the behind-the-meter provisions. Both are being tested on hyperscale data centers, the load class with the least political sympathy and the deepest pockets. The open question is whether the same drafting migrates down-market.
No filing yet extends an 85 percent floor into a general commercial class, and the economics that justify it for a several-hundred-megawatt data center do not obviously transfer to a two-megawatt building. But rate design propagates by template. The FirstEnergy filing and the parallel tariffs in Oregon, North Carolina, and Montana each started as an answer to a specific problem, and each becomes reference language for the next commission that faces a similar one.
For now, the mid-market commercial customer is untouched, and rising demand charges elsewhere continue to favor on-site storage. The signal in the FirstEnergy filing is not a present threat to that case. It is a demonstration that a utility, when it chooses to, can write a tariff that decides in advance how much a battery behind the meter is allowed to save.
Sources