The fuel is missing before the reactors even exist. By end of June 2026, Centrus Energy was to reach 1.8 MTU of HALEU, generally produced as UF6 before its deconversion and manufacture into fuel required by most advanced reactor designs—less than 10% of demand estimates associated with possible deployments of advanced reactors. Advanced SMRs face an incomplete HALEU chain with limited commercial capacity, rather than one that doesn’t exist.
The Essentials
- HALEU (uranium enriched between 5% and 20%) is the fuel required by the majority of advanced SMR designs undergoing certification.
- Centrus was to reach 1.8 MTU of HALEU as UF6 by end of June 2026, less than 10% of demand estimates compiled for possible deployments of advanced reactors (GAO-26-107385).
- Industrial players hesitate to invest in enrichment plants without guaranteed long-term purchase contracts.
- The construction and certification timeline for a HALEU plant spans several years, compressing the available window for investments.
- Breaking the deadlock requires a risk-sharing mechanism between states and industry, guaranteed purchase-type contracts, or public reserve capacities.
HALEU: Ordinary Fuel Turned Strategic Matter
Regulatory constraint also weighs on certification timelines for the facilities themselves. A cascade of centrifuges intended for HALEU must satisfy safety and non-proliferation requirements distinct from those applicable to conventional enrichment, which extends authorization procedures independently of operators’ technical capacities. This additional regulatory layer does not invalidate the industrial model, but it compresses the available window to achieve significant production before the first reactors come online. Administrative delay mechanically adds to construction delay, without either being able to absorb the other.
Ordinary uranium in conventional power plants is enriched to 3 or 5% in uranium-235. HALEU is enriched between 5% and 20%. This difference in proportion makes it possible to design smaller, denser reactors and potentially less costly to build—the industrial wager of almost all advanced SMR designs, from TerraPower to concepts from EDF or Rolls-Royce.
The problem lies in the history of nuclear proliferation. Facilities capable of enriching uranium to these levels have been subject to very strict regulations since the 1970s. The only commercial plant capable of producing HALEU in industrial quantities in the United States is Centrus Energy’s facility in Piketon, Ohio. It was to reach 1.8 MTU of HALEU as UF6 by end of June 2026, after restarting a centrifuge cascade financed in part by the U.S. Department of Energy.
The GAO reports external projections of HALEU demand. The ratio speaks for itself: current production covers less than one-tenth of projected needs.
The Investor Awaits the Buyer, the Buyer Awaits the Reactor
The loop is all the more difficult to break because risk is not symmetric between the parties. An investor who commits without a contract bears certain losses if the market lags; a buyer who defers an order only suffers a delay. This asymmetry incentivizes waiting on the capital side, when it is precisely capital that must mobilize first for the chain to start. Public or pooled intervention can help reduce initial risk when private commitments are insufficient.
A high-precision enrichment centrifuge is expensive to build and more so to certify. In January 2026, the DOE awarded Centrus a $900 million task order to expand HALEU production in Piketon. Long-term purchase contracts can facilitate private investment, but they were not a prerequisite for DOE task orders.
HALEU needs and allocations are tied to licensing milestones and project schedules. Long-term fuel commitments remain underdeveloped before reactor deployment. The NRC issued the Natrium construction permit on March 4, 2026. The first Natrium plant is projected for commercial operation in 2030, subject to project schedule. Result: long-term fuel commitments remain underdeveloped before reactor deployment.
This loop is classic in capital-intensive industries. It resembles, in relative terms, the problem of electric vehicle charging stations in the early years of EVs: drivers would not buy electric cars for lack of charging stations, and operators would not install stations for lack of drivers. Electromobility needed temporary public guarantees to break this circle, in the form of purchase subsidies and infrastructure tenders, before the market took over. The HALEU chain poses the same problem, in an industry where timelines are measured in years.
European Production and Global Demand
Dependence on HALEU touches a particular geography. Russia was the main commercial supplier, but China also has large-scale production infrastructure. After Russia’s invasion of Ukraine in 2022, the United States, United Kingdom, Canada, Japan, and the European Union undertook efforts to reduce this dependence, but the industrial effects of these decisions are slow to materialize.
In Europe, Urenco, the British-German-Dutch consortium, is developing HALEU capacity in the United Kingdom targeting up to 10 tonnes per year from 2031 onwards. Its capacity remains limited and oriented toward the conventional commercial market. Orano, in France, is in a similar position. HALEU expansion programs have been announced, but their adequacy to future needs depends on demand scenarios.
The geopolitical picture is thus as follows: HALEU demand is global and rising, non-Russian Western supply is limited, but China has large-scale HALEU infrastructure, and the few Western players capable of producing are awaiting a market signal that is slow to arrive. The Carnegie Endowment analysis from October 2026 emphasizes that this bottleneck is now better documented than in the early SMR announcements, which does not make it easier to resolve.
The Window Closes Between 2027 and 2029
Schedule is the hardest constraint. Building a commercial-scale HALEU enrichment plant takes several years, once the investment decision is made and permits are obtained. For significant capacity to be available before the first SMR deployments expected, the investment decision must be made in the years ahead.
This schedule tightens the available investment window. An absence of commitments before this period would increase the risk of delays for deployments in the 2030s. GAO-26-107385 notes that the U.S. government has become aware of this schedule but that available instruments lack economic documentation and programmatic coordination.
This type of blockage is documented in other capital-intensive infrastructure sectors. Industrial automation creates skill shortages analogous to this one: the machine arrives but technicians capable of operating it are missing, because training takes time and no one trains without guaranteed job prospects. The dynamic is identical here, except that timelines are twice as long and the sums at stake are considerably higher.
Decisions to Make Today for 2032-2035
Building a non-Russian HALEU enrichment chain before the first SMR deployments would require a public financing mechanism compensating for the absence of prior purchase contracts.
Two trajectories are emerging. In the first, governments that have made advanced nuclear energy a pillar of their climate policy—the United States, Canada, United Kingdom, France, Japan—coordinate investments and actions to strengthen a non-Russian nuclear fuel supply chain. This can take the form of long-term contracts with Centrus or Urenco, public strategic capacities paid for by states in advance of the market, or joint tenders among SMR buyers. The comparison with guaranteed purchase contracts put in place for renewable energy in the 2000s is relevant: these mechanisms enabled financing of solar panel and wind turbine factories before the market was mature. They cost taxpayers dearly at first; they made the transition competitive later.
In the second trajectory, governments continue financing research and reactor certification without tackling the fuel problem. HALEU availability would then risk remaining limited and uncertain, dependent on existing production capacities and their possible expansion. In case of shortage, non-Russian options would remain limited, but Rosatom would not be the sole potential source, which nonetheless poses a risk for the energy sovereignty objectives declared since 2022. The IAEA observes commercial dependence on Russian HALEU and diversification efforts of supply sources.
Signals to watch in the coming eighteen months are identifiable. The U.S. Department of Energy’s budget for 2027 will or will not contain a substantial HALEU reserve contract line. Urenco will or will not announce an expansion program. The first SMR operators, TerraPower, X-energy, British candidates under the Great British Nuclear selection program, will or will not pass fuel letters of intent. Each of these signals is worth more, as an indicator of trajectory, than any political statement of intent.
The question of regulating upstream technologies with heavy infrastructure is precisely the one being posed here, in reverse: the risk is not regulating too early, but investing too late.
The Instruments Exist; Political Decision Is Missing
Nothing in this equation is technically insoluble. The physics of centrifuge enrichment is understood. Centrifuges exist. Centrus has proven that a HALEU cascade can operate under industrial conditions in Piketon. Urenco has the expertise.
What is missing is a financing mechanism that uncouples the investment decision from the certainty that contracts will follow.
Precedents exist in other critical industries. Messenger RNA vaccines benefited from massive government prepurchases in the United States and Europe before clinical trials were even complete; this prepurchase financed production capacity. Advanced semiconductors are the subject of direct subsidies and public purchase guarantees in the United States (CHIPS Act), Europe, and Japan. HALEU could benefit from a comparable mechanism: a temporary public guarantee that makes private investment viable, without nationalization of production.
The cost of this guarantee would be modest compared to the stakes. Such an investment represents a limited share of amounts dedicated to the energy transition. What is blocking, therefore, is not cost but the priority given to SMR fuel relative to other energy transition technologies.
Governments that have made advanced nuclear energy a public commitment must decide whether they are prepared to secure access to HALEU fuel through public guarantee mechanisms before reactors are in service. Public guarantees of access to HALEU are an important factor in the SMR deployment timeline, among several industrial and regulatory constraints.
Sources
- GAO-26-107385, U.S. Government Accountability Office, “Nuclear Energy: DOE Should Take Additional Steps to Address HALEU Availability”, September 2026
- World Nuclear Association, report on SMR fuel needs, August 2026 (no link: document available on world-nuclear.org)
- Carnegie Endowment for International Peace, analysis on SMR fuel cycle constraints, October 2026 (no link)
- Discovery Alert, review of the nuclear fuel cycle, October 2026 (no link)
- IAEA, 2026 modeling on fuel cycle constraints (no link: document available on iaea.org)
