NRC Permits Four X-energy Reactors at Dow’s Texas Site | Nuclear Now #78
The NRC’s construction permit for Dow and X-energy’s four-reactor project is real progress, but it is not yet a nuclear project. It is a regulatory authorization to build one. That distinction matters because the industry has spent years confusing permission, partnerships, and PowerPoint capacity with concrete, steel, qualified fuel, and revenue.
The Long Mott project at Dow’s Seadrift, Texas, site would deploy four Xe-100 high-temperature gas reactors totaling 320 megawatts, with electricity and process steam supplied to an industrial complex. This is a more credible use case than the familiar promise to sell generic grid power. Dow has an existing industrial load, a reason to value firm heat, and a site where replacing fossil process energy could produce more value than selling electrons into a wholesale market.
But the permit covers construction only. X-energy and Dow still need an operating license before loading fuel or generating power. They also need to finance the build, complete detailed engineering, manufacture pressure vessels and reactor components, establish a dependable supply of high-assay low-enriched uranium fuel, and prove that the plant can be delivered at something close to its commercial assumptions. The NRC decision removes a major institutional obstacle. It does not remove the execution risk.
That is why this milestone deserves neither the anti-nuclear dismissal that all advanced reactors are imaginary nor the investor shorthand that an NRC permit equals commercialization. The project has crossed a meaningful line, and the next line is much harder: building four identical reactors in sequence, on schedule, with a customer that can use their heat. The nuclear revival needs precisely this kind of project. It also needs the industry to stop pretending that regulatory progress is the same thing as economic proof.
The economic case for Long Mott is not primarily “SMRs are cheaper.” That claim has not been demonstrated. The case is that a reactor providing both power and high-temperature process steam can compete against a different incumbent than a standalone electricity generator does.
A 320 megawatt plant operating at a 90 percent capacity factor would produce roughly 2.5 terawatt-hours annually. That is useful, but the electricity alone is unlikely to justify an advanced-reactor price premium. US nuclear construction history is filled with projects where the financing cost, schedule risk, and first-of-a-kind engineering overwhelmed the theoretical advantages of smaller units. The NuScale cancellation in Idaho remains the warning: its six-module project was abandoned after projected costs rose sharply, despite substantial federal support.
X-energy’s design tries to change the calculation in three ways. First, the Xe-100 is a high-temperature gas reactor, so its output includes industrial steam rather than only grid electricity. Second, four identical units are intended to create repetition within one site. Third, the project is attached to a large industrial customer, reducing dependence on volatile wholesale power prices. Those are sensible design choices. None has yet been validated at commercial scale.
The fuel question is especially important. X-energy’s reactor depends on TRISO fuel, and the United States has had to rebuild domestic capacity for the specialized fuel needed by advanced reactors. Centrus has made progress on high-assay low-enriched uranium production, but fuel availability remains a supply-chain constraint rather than a solved commodity market. A reactor vendor can have a certified design and still lack the industrial ecosystem required to deliver multiple units.
The comparison with renewables also needs precision. Solar and wind frequently produce cheaper marginal electricity, but they do not provide Dow with continuous industrial heat without storage, overbuilding, firming generation, or a major redesign of the customer’s energy system. A nuclear project should not be judged as though its only product is an undifferentiated megawatt-hour. Conversely, the ability to provide firm heat does not excuse poor capital discipline. If Long Mott requires a premium large enough to make Dow’s products uncompetitive, the industrial use case collapses.
The commercial test is therefore not the NRC permit. It is the eventual price of delivered steam and electricity, including financing, contingency, fuel, operations, and the cost of delay. The project needs to show that nuclear can be valuable enough to an industrial host to pay for reliability and decarbonization, while still being buildable enough for investors to finance the next four units.
The clearest beneficiary is X-energy, because the permit gives the company something far more valuable than another memorandum of understanding: a regulator-reviewed project with a named industrial customer and a defined site. That improves credibility with utilities, suppliers, and governments evaluating whether to place orders for the Xe-100.
It also strengthens the case for industrial nuclear deployment in the United States. Dow’s Seadrift operation is not a remote greenfield power project. It is a large existing manufacturing site, which means the nuclear plant can be evaluated as part of an integrated energy system. If the economics work, chemical producers, refineries, steelmakers, cement companies, and data-center developers have a more relevant template than a reactor built solely to sell into the grid.
The project also raises the stakes for competitors. TerraPower, Holtec, NuScale, and other advanced-reactor developers have spent years accumulating federal support, site agreements, and licensing activity. Those are necessary assets, but X-energy now has one of the strongest near-term tests of whether its commercial model can survive contact with construction. The company’s advantage is not that the Xe-100 is automatically cheaper. It is that the project has a customer whose energy problem is specific and expensive.
New York’s decision to qualify ten nuclear developers for potential partnerships shows how quickly states are moving from broad support to vendor selection. But qualification lists are not orders. Neither are proposed projects in Indiana, where X-energy, Amazon, and AEP are studying a 12-unit, 960 megawatt deployment at a retired coal site. Those efforts benefit from Long Mott if the Texas project demonstrates repeatability. They are exposed if Long Mott becomes another first-of-a-kind cost escalation.
The next signal to watch is not another announcement. It is the engineering and financing sequence: an operating-license application, a credible construction schedule, binding supply contracts, a final investment decision, and visible site work. Each step separates a buildable fleet from a company selling future optionality.
Long Mott is the kind of project the nuclear revival needs because it links an actual industrial customer to an actual reactor design and an actual regulator. It is also exactly the kind of project that can damage the revival if the industry treats a construction permit as a victory lap. The winners will be companies that turn regulatory permission into repeatable delivery. The losers will be those that keep raising capital on the assumption that the hard part is getting permission to begin.