North Carolina just made the customer contract part of the permit

North Carolina's utility commission has denied Duke Energy's application for a 255-megawatt combustion turbine at the Smith Energy Complex on a ground that deserves close reading: the record does not establish which customers the plant is for. The proposed unit, hydrogen-capable and F-class, was originally estimated at 240 MW and revised to 255 MW after Duke Energy Progress, the Duke subsidiary serving North Carolina, concluded negotiations with a turbine vendor the order does not name.

The commission was specific about what the record lacked: it "does not adequately establish the specific extent to which the Proposed Facility is intended to serve the anticipated growth in data center customer demand," the order states, and therefore "does not demonstrate how [Duke] intends to adhere to its commitments under the Ratepayer Protection Pledge with regard to the Proposed Facility." The pledge — a voluntary promise by utilities and hyperscalers not to shift data center costs onto other customers — is "not a substitute for the statutory CPCN standard," the commission allowed, though Duke's agreement to its provisions is "relevant to the Commission's consideration of ratepayer impact and determination of whether the Proposed Facility is in the public interest."

A voluntary pledge with no statutory force has now been treated as relevant to a ratepayer-impact finding. It lands against a company whose scale makes 255 MW look small: in February, Duke executives told investors the company's $103-billion capital spending plan was the largest on file at any regulated U.S. utility, framed around the data center boom. One turbine certificate against a program that size is immaterial; the test applied to it is not, because a utility that must name the load a generator serves before it can build the generator inherits an exhibit list it has never had to produce.

The commission's own public staff agreed the load is coming, reversing an initial opposition to recommend approval after a June 30 errata cited Duke's "aggressive signing of new load" and found a capacity need, while still testifying the unit was not an ideal solution. That leaves a position worth sitting with: demand is arriving in North Carolina, the plant would help serve it, and the filing failed anyway.

The gap the order describes is a contract gap: the record does not show which customers the load serves, or how their costs stay off other ratepayers' bills, and the commission read that absence as fatal. It also questions the load forecast underneath Duke's capital plan, which assumes data center demand converts into contracted, cost-allocated customers on a schedule the utility can build against; on this record, one proposed turbine could not show that conversion in writing.

Smith has been waiting a long time already: the complex went into commercial operation in 2001, houses five F-class simple-cycle units and two F-class combined-cycle units, and was configured for a sixth simple-cycle machine that was never built. One witness in the proceeding estimated a new generator could be in service by Jan. 1, 2030, and that date is the denial's real cost — a refiling means a new docket, new testimony and new staff review running against a construction calendar while the load arrives on its own schedule.

As this publication noted when the order came down, North Carolina's normally deferential panel has converted a capacity decision into a scheduling problem and attached a price to time. Grid permission is the underwriting asset: queue positions and interconnection rights get priced before electrons do. North Carolina pushes that argument one layer up the capital stack: permission now runs through the contract, and a utility can hold the site, the turbine slot and the air permit and still lack the right to build, because the certificate turns on whether it can produce agreements that identify who pays for the output.

Duke's answer is that it has "demonstrated that the Smith [combustion turbine] is part of a least-cost path to maintain reliable and affordable service for customers as energy demand continues to grow across North Carolina," and that it remains committed to working constructively with regulators and stakeholders "to identify the best path forward." Neither statement supplies the exhibit the commission asked for, pointing to a refiling whose appendix, more than its testimony, will decide the case.

The narrow reasoning holds: the record genuinely did not show who pays, and the commission was within its statute to say so. The broader doctrine carries a risk worth naming, because a standard that requires a named customer for each new generator hands a utility's construction schedule to its largest customers' procurement calendars; applied to the next baseload filing, it would let one hyperscaler's real estate timing set when 255 MW of capacity the rest of North Carolina is paying for actually gets energized. Duke's refiling will show which reading the panel meant, and the in-service date one witness put on the record — Jan. 1, 2030 — is the clock it will be judged against.