Approval is easy. Fuel is the bottleneck.
Kemmerer, March 2026
TerraPower has secured the federal green light that most advanced-nuclear projects spend years chasing, and the implications go far beyond a single construction site in Wyoming. The U.S. Nuclear Regulatory Commission has issued a construction permit for the company’s Natrium reactor, a milestone being described across multiple outlets as the first U.S. approval for a new commercial reactor in nearly a decade and the first permit for a commercial-scale non–light-water reactor in more than forty years. For Bill Gates, TerraPower’s founder and primary backer, it is a validation moment. For the U.S. energy system, it is a live experiment in whether “next-generation nuclear” can move from PowerPoint to concrete without collapsing under cost, supply chain constraints, or political fatigue.
The Natrium concept is not a conventional nuclear build with a new paint job. It is a sodium-cooled reactor design paired with an integrated energy storage system intended to smooth output and meet peak demand without forcing the reactor itself to ramp aggressively. The headline numbers attached to the project capture the ambition: roughly 345 megawatts of electric output with the ability to temporarily boost to about 500 megawatts using stored thermal energy. That hybrid logic matters because it is aimed at a modern grid problem, not a 1970s one. Grids are becoming more volatile as wind and solar expand, as electrification pushes demand upward, and as data centers increase baseline load. A reactor that can behave more like a dispatchable asset, rather than a constant-output plant, is attempting to compete in the language of the current market.
Yet the permit is not the finish line. It is the start of the part that fails most often: execution. Building an advanced reactor is less about engineering novelty than about engineering discipline under scrutiny. The NRC permit authorizes construction of the plant as designed, but TerraPower still needs an operating license later, and it still must demonstrate the real-world manufacturability of components, the stability of supply chains, and the ability to control schedule creep. This is where advanced nuclear projects historically lose credibility. They either drift into indefinite “development,” or they become so expensive that the technology’s theoretical advantages stop mattering to utilities and taxpayers.
The most immediate structural risk is fuel. Natrium is designed to use HALEU, high-assay low-enriched uranium, enriched to levels below weapons grade but higher than the fuel used in most existing U.S. reactors. In the current geopolitical environment, that is not a detail. It is the central constraint. HALEU supply has been heavily linked to Russian production capacity, which creates a strategic dependency that becomes politically unacceptable precisely at the moment Washington is trying to rebuild domestic nuclear capability. TerraPower’s project therefore sits inside a paradox: the U.S. wants advanced reactors to strengthen energy security, but some advanced reactor designs depend on a fuel ecosystem that is not yet secure. This is why the “hardest problem” is not the reactor vessel. It is the fuel pipeline.
Washington has started treating that pipeline as an industrial priority, including large public funding signals aimed at accelerating domestic HALEU capacity. That policy direction matters because it reveals the true shape of the bet being made. TerraPower is not only building a plant. The United States is trying to rebuild a nuclear supply chain layer that has atrophied, including enrichment, conversion, and fabrication for fuels that the next generation of reactors expects. If the fuel chain does not mature on schedule, the project can become a success on paper and a delay in practice, which is the most damaging outcome for public trust.
There is also a cost-and-credibility debate that will not go away. Advocates argue that smaller, factory-influenced designs and modular construction approaches can reduce risk and shorten build times compared with classic gigawatt-scale projects. Critics argue that smaller reactors may struggle to achieve economies of scale and may end up expensive per unit of electricity, especially if the supply chain is bespoke and the workforce must be trained from scratch. Both positions can be true at different phases. The first-of-a-kind unit is rarely cheap. The argument is whether it can become cheaper on the second and third builds, and whether the market will still be patient by then.
Waste and long-term governance remain the quiet pressure points. Advanced designs often claim better fuel utilization or reduced waste profiles, but they do not erase the political reality that the U.S. still lacks a settled, durable solution for high-level waste disposal. Every new nuclear project inherits that unresolved national issue, and every delay in national waste policy becomes a reputational weight on nuclear expansion, even when the plant itself operates safely.
The larger context is that the Natrium approval arrives during a renewed global push to treat nuclear as a serious decarbonization and reliability tool, especially as electricity demand rises from electrification and computational infrastructure. TerraPower’s permit is therefore not merely a company milestone. It is a governance signal: the U.S. regulator is willing to authorize advanced designs again, and the state is willing to spend to make the enabling fuel ecosystem real. But the permit does not guarantee that the project will be delivered on time, on budget, or at a price the market can accept.
In the end, the story is not “Bill Gates revolutionizes nuclear.” The story is that the United States has reopened the door to commercial advanced reactors, and now must prove it can walk through it. TerraPower’s Natrium plant will be judged less by the elegance of its concept than by whether it can be built, fueled, operated, and replicated without turning ambition into another decade-long delay.
Phoenix24: clarity in the grey zone. / Phoenix24: claridad en la zona gris.