Urenco USA announced this week it will expand America's only commercial uranium enrichment facility by nearly 50%, a multi-billion-dollar bet on nuclear fuel demand that won't materialize for years. The company will install 2.1 million separative work units of new capacity at its New Mexico plant, with initial cascades coming online in 2032 and the buildout running through 2036.
This is the move nobody's pricing yet. Urenco's locking in capacity today to supply a reactor fleet that's still renderings and construction permits, not baseload gigawatts. The Department of Energy celebrated its first advanced reactor reaching criticality on June 4, and X-energy just signed Fluor to support four 80-MW small modular reactors at Dow's Seadrift facility in Texas—but commercial deployment at scale is a 2030s story, not a near-term one.
Wood Mackenzie notes that 62 reactors are under construction globally, but the prohibitively long timelines for large-scale builds cannot address near-term demand. Vogtle Unit 3 took 14 years from groundbreaking to commercial operation, per the same source—an execution risk Urenco's now financing six years ahead of first fuel deliveries. The expanded capacity will produce low-enriched uranium and serve as feedstock for high-assay low enriched uranium, the fuel advanced reactors planned for the 2030s will require.
Power demand is forecast to grow 21% between now and 2030, largely driven by AI data centers, and Google, Oracle, Meta, Microsoft, and Amazon are investing tens of billions in small modular reactors to power their facilities. The thesis is sound—AI load plus decarbonization mandates equals nuclear renaissance. But thesis and cash flow are separated by a chasm of permitting, construction timelines, and first-of-a-kind execution risk that Urenco's now bridging with shareholder capital, not customer contracts.
The real signal here is confidence in policy durability, not demand visibility. President Trump signed four executive orders to strengthen U.S. nuclear leadership and called for construction of 10 new reactors and reform of the Nuclear Regulatory Commission to accelerate deployment. Europe reversed course in 2025—Belgium abandoned its nuclear phase-out, Italy lifted its ban, Germany recognized nuclear as green. Urenco's betting the regulatory tailwind stays at their back through 2036, a long time to hold a multi-billion position in a sector where Three Mile Island killed momentum for 40 years.
The company's not alone in the forward bet. DOE awarded $19 million in February for used fuel recycling R&D, then called for private partners to design and operate fuel recycling and fabrication facilities. DOE finalized three HALEU fuel allocation contracts in 2025 to support Trump's goal of reactor criticality by July 4, 2026. The entire supply chain's being capitalized ahead of the build cycle, on the assumption execution catches up to enthusiasm.
Watch three things that determine whether Urenco's early bet pays or becomes stranded capacity:
- Advanced reactor FIDs in 2027-28: X-energy's four Seadrift SMRs and new deployments with PPL and Talen in Pennsylvania need to move from MOUs to construction contracts. If they stall, Urenco's 2032 cascades come online into a demand void.
- NRC licensing timelines: X-energy's Seadrift project became the first commercial reactor to receive environmental clearance via Environmental Assessment instead of the years-long Environmental Impact Statement process, with NRC completing it ahead of schedule. Scale that across 20 projects or watch the enrichment glut build.
- HALEU production ramp: The NRC approved X-energy's TRISO fuel fabrication license in February, the first new fuel license in nearly 50 years. Enrichment capacity means nothing if fuel fab can't convert it, and that's still a bottleneck with one licensed plant.
Urenco's invested over $5 billion in the facility since 2006 and is the only company to have licensed, built, operated, and expanded a commercial enrichment plant in the United States. That track record buys credibility. But betting billions on 2030s reactor demand in mid-2026 is either the shrewdest supply-chain play of the energy transition or the most expensive example of building the track before anyone ordered the train.