A majority of the North Carolina Utilities Commission pointed to concerns about timing and whether data centers would bear enough of the cost in taking the rare step of denying Duke Energy a key regulatory approval for a new gas plant.
Duke was asking the commission for a certificate of public convenience and necessity to build a new combustion turbine at its Sherwood H. Smith Jr. Energy Complex near Hamlet, a Richmond County town near the South Carolina border. The proposed plant would also be close to Amazon's under-construction $10 billion data center, and questions about how much of its power would be used for data center projects played a major role in commissioners' skepticism.
The commissioners found that Duke failed to prove that the new turbine is necessary in part because Utilities Commission staff has not yet vetted Duke's modeled load growth — of which about 70% comes from data centers — as part of its ongoing 2025 Carbon Plan and Integrated Resource Plan case.
Beyond that, commissioners found that Duke hadn't discussed how much of the power from the anticipated $584 million turbine would be used to power data centers and, to the extent it would be, how those customers would pay for the project. Duke needed to better demonstrate, commissioners found, how it would comply with the White House Ratepayer Protection Pledge the utility signed onto earlier this year.
"The application for a CPCN is premature and the record is consequently insufficient for the Commission to determine whether the proposed facility satisfies all the applicable requirements for issuance of a CPCN," commissioners wrote in their order.
Commissioners voting to deny the certificate included Bill Brawley, Tommy Tucker and Donald van der Vaart - all Republican appointees. Commissioner Floyd McKissick, a Democrat,
voted to grant it and wrote a dissenting opinion. Commissioner John Gadja did not vote.
It is exceptionally rare for the Utilities Commission to deny Duke Energy a certificate of public convenience and necessity.
"I think that unprecedented wouldn’t be too strong a word to use here. You know, CPCNs have generally been sort of a formality in recent years, and I think it just shows how skeptical the Commission is of putting these costs on ratepayers," Will Scott, the Environmental Defense Fund's North Carolina policy director, said in an interview.
In a concurring opinion, Tucker wrote specifically to the concerns about how much of the turbine's electricity would go to data centers. While witnesses said the turbine wasn't being built for a specific project, he wrote, it was not lost on him that it is very close to the proposed Amazon facility and that the Utilities Commission had approved a new transmission line to that facility.
"(Duke) should provide additional evidence regarding the extent to which the Proposed Facility is needed to provide service to data centers, and how much those data centers will contribute to costs," Tucker wrote.
Duke Energy's stance
Duke's Smith Complex already features a pair of combined cycle gas plants that can generate 1.08 gigawatts of power, as well as five combustion turbines that can generate up to an additional 820 megawatts.
Combined cycle plants are typically used to generate baseload electricity, or the power that is needed from day to day, while combustion turbines are often referred to as "peakers" because they are used when electricity demand peaks.
For example, Duke modeled the proposed turbine's use in 10 years with typical weather. It found that the turbine would be used more than 10% of the time in four of those years.
Craig Wilson, a Duke Energy spokesman, wrote that the company is reviewing the commission's decision and considering its next steps.
"We believe we have demonstrated that the Smith CT is part of a least-cost path to maintain reliable and affordable service for customers as energy demand continues to grow across North Carolina," Wilson wrote.
During evidentiary hearings, Duke officials told the commission that because Smith was configured for a sixth turbine, it could be built and start generating power relatively quickly. They were hoping to have the new turbine operating by 2030.
Duke witnesses also warned commissioners that not granting the certificate might not only delay the in-service date of the turbine, but also potentially result in higher construction costs.
McKissick's dissenting opinion called the decision to deny the certificate "a serious error of judgment," focusing in part on how much it would cost to build the turbine.
Failing to move forward with the project now would likely result in higher costs to ratepayers if Duke seeks approval in the future, McKissick wrote. He noted that in rebuttal testimony, a Duke witness said the utility had received two bids that were much lower than the initial estimates.
"Ratepayers should not be required to bear the costs of delay resulting from the denial of a facility which the record plainly demonstrates is needed," McKissick wrote.
McKissick also wrote that there is no other plan that can add firm power to Duke's energy mix like the proposed turbine.
Are there other options?
Public Staff, the state agency that represents the interest of ratepayers in utility cases, also recommended that the Utilities Commission approve the new turbine despite what it called a "staggering" $584 million price tag.
During the evidentiary hearing, Dustin Metz, a Public Staff engineer, said he would recommend moving forward with the project despite that price tag because of concerns Duke's ability to reliably provide power to a growing state in coming years.
"Because Duke hasn't put forward an alternative that actually can be implemented in a timeline to mitigate that reliability concern that we're seeing, we are now, for lack of a better word, stuck with this resource," Metz said.
But commissioners disagreed, pointing to 2025's Senate Bill 266, which among other things tweaked the state's utilities laws to say the commission can grant a certificate of public convenience and necessity only after comparing the cost of a proposed facility to other resources like grid modernization, storage, energy efficiency and demand-side management, among others.
In his concurring opinion, van der Vaart wrote that Duke had failed to make those comparisons. While some kind of dispatchable resources may be necessary by 2030, van der Vaart wrote, the record did not prove that the turbine was the least-expensive way of meeting that need as required under SB 266.
"The record should allow the Commission to evaluate the alternatives, their costs, their reliability attributes, their limitations, and the reasons they cannot establish or maintain a more cost-effective and reliable system. An assertion that the Proposed Facility is the only resource that can be completed in time does not by itself substitute for that analysis," van der Vaart wrote.
Commissioners left open the possibility that Duke can better address the concerns they raised and re-file the application.
