Duke Energy Asks North Carolina Regulators to Make Data Centers Pay for 75 Percent of the Power They Reserve

Duke Energy wants any customer drawing 50 megawatts or more in North Carolina to pay for at least 75 percent of its maximum potential energy use, whether or not that power is ever consumed, under minimum contracts running 10 to 15 years. The proposed large-load tariff sits alongside the utility’s roughly 15 percent overall rate increase request, and the North Carolina Utilities Commission opens hearings on the rate case July 7.

The state’s consumer advocates want the terms tightened. They asked for an 85 percent minimum take, 20-year contracts, and a lower 25-megawatt threshold that would sweep in more customers. The distance between those two proposals, 75 versus 85, is not a rounding difference. It is the share of a data center’s reserved capacity that other ratepayers underwrite if the facility runs below its contracted peak or never fully materializes.

The mechanism. A minimum-take clause, sometimes called take-or-pay, forces a large customer to pay for a fixed floor of demand regardless of actual load. The logic is straightforward. Data centers ask utilities to commit large sums of capital against load forecasts that the customer alone controls, and a minimum-take contract converts a speculative promise into a contractual liability the developer cannot easily walk away from.

Who absorbs the forecast risk. Consider a data center that contracts for 100 megawatts. If it takes 75, a 75 percent floor means the utility collects on the full reservation and no cost shifts. If that same center takes only 60, the 15-megawatt gap between actual use and the contractual floor is covered by the customer. Below the floor, the arithmetic reverses. A center that reserves capacity, prompts the utility to build for it, and then underdelivers leaves stranded investment. The higher the minimum take, the more of that stranded cost the data center bears rather than the general body of ratepayers, which in North Carolina means residential customers and the commercial and industrial accounts that pay demand charges.

The 10-point spread between the two proposals maps directly onto that exposure. Duke’s 75 percent floor protects other customers against moderate underuse. The advocates’ 85 percent floor protects them against a deeper shortfall, and their proposed 20-year term protects them against a data center that performs for a decade and then relocates for cheaper power once its contract lapses. Every point the commission shaves off the floor is residual exposure it assigns to everyone else on the system.

The wider field. North Carolina is not the first state to confront this question. Regulators and utilities across the country have moved to wall off data center costs from the general rate base, with dozens of large-load tariffs now filed or in force. Duke’s version sits toward the middle of that range rather than at the strict end. The threshold is high at 50 megawatts, the term is shorter than advocates want, and the take is set at three-quarters rather than the higher shares the toughest proposals seek. The specific mix of threshold, term, and take is what distinguishes one utility’s approach from another.

The precedent. Southeast commissions tend to follow one another, and the outcome in Raleigh will carry beyond it. A 75 percent floor blessed by the North Carolina commission becomes a reference point for the next large-load schedule filed in the region, and a number other Southeastern utilities can cite when they draft their own. A commission that pushes the figure to 85 sets a different precedent. Either way, the decision is likely to inform how neighboring states treat the same class of customer.

The effect on commercial customers. For the commercial and industrial accounts already facing Duke’s proposed 15 percent increase, the tariff is not an abstraction. Rate cases allocate the cost of new generation and transmission across customer classes, and any data center cost that a minimum-take clause fails to capture flows into that allocation. A weaker floor does not raise a specific commercial line item. It raises the pool of shared infrastructure cost that class allocation then divides, and demand-heavy commercial accounts carry a large share of that pool. That is the channel through which a percentage buried in a tariff schedule reaches a monthly bill.

What the commission decides. The question before the commission is narrow. Not whether data centers should pay their own way, which nearly every party now accepts, but the precise percentage and term at which “their own way” is defined. The hearings that open July 7 will turn on figures that read like contract boilerplate: a take percentage, a contract length, a megawatt threshold. Those three numbers determine how much of the largest load-growth wave in a generation the rest of North Carolina’s customers help finance.

The broader pattern is a shift in how utilities price speculative demand. For most of the past century, a large new customer was an unambiguous benefit to the system, spreading fixed costs across more kilowatt-hours. Data center load inverts that assumption, because the capital required to serve it can strand if the forecast proves wrong, and the forecast belongs to a customer with no obligation to be right. Minimum-take contracts are the instrument regulators are reaching for to close that gap. North Carolina’s proceeding will test how tightly one commission is willing to draw the line.

Sources