TerraPower Wins First Advanced Reactor Permit in 50 Years, Marking U.S. Nuclear Comeback Attempt
Bill Gates-backed sodium reactor gets NRC clearance in Wyoming, setting up race with China on next-generation nuclear and test of whether new designs can break cost spiral.
The U.S. Nuclear Regulatory Commission on March 4 approved construction of TerraPower’s 345 MW Natrium reactor in Kemmerer, Wyoming—the first commercial non-light-water reactor permit issued in more than 40 years and a test case for whether advanced nuclear can escape the industry’s chronic delays and cost overruns. The project, backed by TerraPower founder Bill Gates and $2 billion in Department of Energy funding, combines a sodium-cooled fast reactor with molten salt thermal storage capable of ramping output to 500 MW for over five hours.
The approval arrives as China operates a 600 MW sodium fast reactor that began low-power testing in 2023 and plans a 1,200 MW successor by the early 2030s. While TerraPower’s reactor uses similar coolant technology, the U.S. project distinguishes itself through integrated energy storage—a grid-balancing feature China’s CFR-600 lacks. According to Cowboy State Daily, the NRC completed its safety review in 18 months, nine months ahead of the original 27-month schedule.
Regulatory Sprint vs. Economic Reality
The accelerated timeline reflects pressure from the 2024 ADVANCE Act, which mandated faster advanced reactor reviews. The Breakthrough Institute notes the permit validates TerraPower’s strategy of separating the nuclear island from the energy storage system, exempting roughly two-thirds of the plant from NRC jurisdiction. That regulatory gambit reduced licensing burden but does not resolve the harder question: can Natrium break even?
NucNet reports the total project cost at approximately $4 billion, split equally between DOE and TerraPower. At 345 MW base capacity, that implies roughly $11,600 per kilowatt—well above the $6,600/kW average for conventional nuclear cited by the International Energy Agency and significantly higher than utility-scale solar ($1,500-$2,000/kW) or onshore wind ($1,300-$1,700/kW). TerraPower claims its system will produce electricity at half the cost of traditional nuclear, but provided no public data backing that assertion. The company did not respond to questions on levelized cost projections.
| Technology | Cost Range |
|---|---|
| Natrium (Kemmerer) | ~$11,600 |
| Conventional Nuclear | $6,600 |
| Utility Solar | $1,500-$2,000 |
| Onshore Wind | $1,300-$1,700 |
Passive Safety, Active Concerns
Sodium coolant operates at atmospheric pressure and boils at 883°C—far above the reactor’s 662°F operating temperature—eliminating loss-of-coolant accident risk that plagues water reactors. TerraPower states the design relies on gravity-driven passive cooling using air convection rather than powered pumps. The pool-type vessel has no penetrations below the lid, removing leak pathways present in conventional designs.
But sodium reacts violently with water and ignites in air. Edwin Lyman, director of nuclear power safety at the Union of Concerned Scientists, told reporters in December that “this type of reactor has major safety flaws compared to conventional nuclear reactors.” Sodium fires damaged Japan’s Monju fast reactor multiple times before its 2016 closure after generating almost no power in 22 years. The U.S. abandoned its own Clinch River Breeder Reactor in 1983 after costs tripled. Fort St. Vrain, the last U.S. non-light-water commercial reactor, operated for 15 years before shutting down in 1989 due to persistent technical problems.
The Fuel Supply Chokepoint
Natrium requires high-assay low-enriched uranium (HALEU) fuel enriched to 5-20% U-235, versus the 3-5% used in conventional reactors. Until 2022, Russia’s TVEL was the only commercial supplier; the U.S. lacks domestic capacity at scale. Neutron Bytes reports TerraPower signed a term sheet with ASP Isotopes in 2024 for a South African enrichment facility, though no production timeline was disclosed. Each Natrium core requires 15-20 metric tonnes of HALEU; annual refueling needs 3.6 tonnes.
The DOE’s 2025 budget includes funding for HALEU production at Centrus Energy’s Ohio facility, but output remains below commercial demand. China sources its CFR-600 fuel from Russia without supply constraints. According to KCRG, TerraPower acknowledges fuel availability as a critical path item but provided no contingency plan if domestic production lags.
The Kemmerer site sits adjacent to PacifiCorp’s retiring Naughton coal plant. TerraPower broke ground on non-nuclear infrastructure in June 2024, employing several hundred workers on site preparation, roads, and administrative buildings. The reactor will supply power to PacifiCorp’s grid, which also serves Utah, Idaho, and Oregon. Kemmerer, population 2,500, was historically a coal mining town; the nuclear plant represents the largest economic development project in Lincoln County history.
Meta’s $2 Billion Bet and the Serial Build Question
In January 2026, Meta signed an agreement with TerraPower to fund up to eight Natrium units for data center power, with initial delivery targeted for 2032. Neutron Bytes reports each dual-reactor site would provide 690 MW baseload with 1 GW dispatchable capacity. If realized, the deal would provide the serial production volume needed to drive down costs through manufacturing learning curves.
But every advanced reactor developer promises nth-of-a-kind cost reductions that rarely materialize. NuScale, the only U.S. small modular reactor to receive design certification, saw its flagship Utah project collapse in 2023 when projected costs doubled to $9.3 billion. The International Energy Agency estimates first-of-a-kind small modular reactors at $10,000/kW in Europe, declining to rough cost parity with conventional nuclear only by 2050 under optimistic scenarios. A 2023 ScienceDirect techno-economic analysis found molten salt SMRs at $3,985/kW overnight capital cost—still 50% above conventional nuclear once financing and construction duration are factored.
What to Watch
TerraPower plans to pour first nuclear concrete in 2027 and submit its operating license application in 2028. The project’s credibility hinges on three execution tests: maintaining the $4 billion budget through fuel loading, securing reliable HALEU supply chains by 2030, and demonstrating load-following capability with molten salt storage under grid conditions. Cost overruns exceeding 25% would push levelized costs above $140/MWh—uncompetitive with firmed renewables even before factoring transmission.
China’s CFR-600 reached criticality without the U.S. regulatory overhead, but operates without the energy storage system that TerraPower positions as its grid-integration advantage. If Beijing’s planned CFR-1000 comes online before 2032 at lower cost per megawatt, it will establish Chinese sodium reactor technology as the global reference design. The Kemmerer permit is a regulatory milestone, not an economic one. Whether it leads to serial deployment or joins the list of one-off demonstration projects depends on performance no model can predict.