Power Needs People (and Fuel)

We talk about “next-gen nuclear” like it’s a switch you flip. But the chokepoint isn’t a shiny new reactor — it’s the unglamorous fuel supply chain behind it. If we want reliable, zero-carbon power at data-center scale, industrial-heat scale, or just to keep the lights on when the wind is still, we have to rebuild the U.S. nuclear fuel ecosystem from the ground up.

Nuclear power quietly supplies about 19% of U.S. electricity, a 24/7 backbone that doesn’t depend on the weather. The next chapter isn’t just “more reactors,” it’s better fuel and smaller plants. Most advanced designs use HALEU — high-assay low-enriched uranium, enriched to under 20% U-235, still low-enriched (not weapons-grade) — to pack more energy into a smaller core, run longer between refuelings, and, with robust fuels like Tri-structural Isotropic fuel (TRISO), add extra safety margins by design. Pair that with small modular reactors (SMRs), factory-built units that can be added in blocks, and you get firm, zero-carbon power sized for modern loads, including AI/data centers that are straining grids faster than new wires and plants can be built. In short: fuel and people, not hype, will decide whether SMRs scale fast enough to keep our lights — and our servers — on.

Start with this reality: most advanced reactors need HALEU. Until recently, the only commercial source at scale was Russia. Congress finally banned Russian uranium imports last year, with limited waivers for reliability, and put real federal money on the table to re-shore the fuel cycle. That’s the right call strategically, but it also means we’re now living with a supply gap we have to close . . . fast.

There is progress. In June, Centrus Energy hit a milestone no U.S. company has reached in decades: 900 kilograms of HALEU produced and delivered to the Department of Energy (DOE) from a single demonstration cascade in Ohio. DOE has extended production through mid-2026, with options to scale further as funding allows. That’s a big step, but measured against the multi-ton annual needs of a commercial fleet, it’s still proof-of-concept, not an industrial base. 

Meanwhile, Urenco USA — the country’s only operating commercial enricher, just secured U.S. Nuclear Regulatory Commission (NRC) authorization to enrich up to 10% U-235, a level that can support some advanced reactor designs and bolster resilience for today’s fleet. Urenco is also adding capacity at its New Mexico plant. These are tangible moves toward a domestic supply chain we control. 

Why the urgency? Because the projects we champion on paper run on timelines that collide with fuel reality. TerraPower’s Natrium demonstration, for example, pushed to 2030 in part due to HALEU constraints. That’s not a technology problem; it’s a supply chain problem. We won’t deploy what we can’t fuel. 

So what does it take to fix this? Three tracks: capacity, conversion/fabrication, and people, moving together.

1. Capacity — Enrichment and Front-End Conversion: Centrus has proven the machines; Urenco has regulatory headroom and expansion underway. Now the task is to move from pilot-scale kilograms to market-scale metric tons. That means multi-year off-take contracts (not one-year pilots), creditworthy counterparties, and public-private finance that treats HALEU like strategic infrastructure, not a science project. DOE’s recent HALEU allocation and fuel-line pilot selections are a start; industry needs to meet that with binding demand.
2. Fabrication and Form Factors: For advanced designs - TRISO, metallic fuels, novel cladding - the bottleneck will shift quickly from enrichment to fabrication. The U.S. needs qualified lines, licensed Quality Assurance (QA), and a regulatory cadence that certifies fuel forms on timelines that match plant construction, not years after. Think “design-to-license” toolchains that pull NRC, vendors, and utilities into the same calendar.
3. Workforce, Testing, and Quality Assurance: Nuclear’s constraint isn’t just centrifuges — it’s craft labor, welders, chemists, and QA professionals who can build, run, and audit a first-of-a-kind fuel enterprise. If we can stand up a data-center workforce in months, we can stand up a nuclear fuel workforce in years — if we fund apprenticeships, certify programs with real employers at the table, and keep the pipeline filled with projects, not press releases.

There’s also a market design problem we should be honest about: fuel vendors won’t build capacity without bankable demand signals. Utilities and developers won’t sign those signals without clarity on policy risk.

This is where federal tools can de-risk the first movers without socializing everything:

• Long-dated, take-or-pay fuel framework for first tranches of HALEU (think five–ten years) tied to qualified projects and milestones.
• Cost-share with teeth — federal dollars contingent on private capital and delivery, not open-ended subsidies.
• Portfolio procurement — aggregate multiple advanced reactors’ needs so fabricators can scale on one purchase order instead of ten.
• Waiver guardrails on the Russian ban — use them sparingly, transparently, and only to avoid near-term reliability events, not to undercut domestic build-out. 

To the skeptics: this isn’t about picking winners. America, and the world for that matter, needs an all-the-above approach to energy, not just to meet burgeoning energy demand but also to provide cleaner alternatives. This is the backbone of the Administration’s energy policy. For nuclear, it’s about building the commons every reactor design requires — conversion, enrichment, fabrication, and people — and letting projects compete on cost and performance above that floor. If we do that, the best ideas will scale. If we don’t, everyone will stall for the same boring reason: no fuel.

There’s a national-security layer here, too. Before the ban, roughly a quarter of U.S. enrichment relied on Russia. Re-shoring that capability is as much about resilience as it is about climate or cost. Adversaries shouldn’t sit at the choke point of our baseload power, our medical isotopes, or our defense fuels. The bipartisan law recognized that; now we have to deliver on it. 

What should success look like by the late 2020s?

• Multiple domestic sources of HALEU and LEU+ with licensed, running capacity.
• Qualified fabrication lines for the major fuel forms in the advanced pipeline.
• A stable demand stack—utilities, data-center power buyers, industrials—signing long-dated fuel contracts that justify private investment.
• A skilled workforce with career pathways that keep talent in the sector.
• Measured, transparent waivers only when reliability is at stake, sunset as capacity comes online. 

We can keep arguing about reactor brands, or we can fix the part of the system that all of them need. In energy, physics and logistics always win. Right now, the logistics say: no fuel, no future. Let’s build the fuel.