GAO Puts $140 Billion Price Tag on U.S. Enrichment Build-Out

A newly released Government Accountability Office report, dated 20 August 2026, places a $140 billion cost estimate on the domestic uranium enrichment infrastructure the United States must construct to sustain its weapons programmes, naval reactors, and tritium production through 2105. The GAO concluded that lawmakers have not received the itemised cost information required to exercise meaningful oversight of that effort.

The supply cliff is concrete. Existing unobligated enriched uranium stocks, material free from foreign peaceful-use agreements and therefore eligible for national security missions, cover tritium production through the early 2040s and naval propulsion through the 2050s. After those thresholds, new purpose-built domestic enrichment infrastructure must be operational or mission continuity fails.

The report, GAO-26-107385, answers three questions the market has been circling for years: which companies sit in the critical path, what scale of federal spending is now confirmed, and why the transparency gap the GAO flagged creates funding risk for every contractor in the programme.

A $140 billion supply cliff hiding in plain sight

The countdown starts in 2027. That is when the National Nuclear Security Administration’s (NNSA) campaign to downblend surplus highly enriched uranium (HEU) into low-enriched uranium (LEU) reaches completion. The LEU produced from that downblending campaign supplies a pair of Tennessee Valley Authority reactors dedicated to tritium production, which in turn sustains the nuclear weapons stockpile. Once that inventory is drawn down, the supply runs out in the early 2040s.

HEU reserved for naval nuclear propulsion, the fuel that powers aircraft carriers and submarines, remains adequate through the 2050s. HEU set aside for research reactors and medical isotope production runs through 2040.

The countdown starts in 2027, but the structural drivers behind it extend well beyond NNSA planning cycles; the uranium supply deficit affecting both commercial and national security fuel chains traces back to decades of underinvestment in domestic mining, conversion, and enrichment capacity.

After those dates, the inventory is gone. NNSA planning documents project national security enrichment demand will reach approximately 1.7 million separative work units (SWU) per year by the 2050s, a measure of the industrial effort required to enrich uranium to usable concentrations.

GAO-26-107385 documents the NNSA’s enrichment mission across three distinct programme categories, each with separate inventory horizons, demand projections, and infrastructure requirements that the GAO found Congress had not received in a unified, long-term cost framework.

Of that total, roughly 1.5 million SWU per year must be sourced from newly constructed domestic facilities capable of producing unobligated output, meaning uranium produced entirely without foreign technology, components, or feedstock that would activate international peaceful-use commitments.

The word “unobligated” is the constraint that makes existing U.S. commercial enrichment capacity largely irrelevant to these missions. A centrifuge plant can be physically capable of producing weapons-grade material and still be legally disqualified if any component in its supply chain traces back to a foreign agreement. That single distinction is what drives the entire $140 billion programme.

The 2040s and 2050s sound distant. They are not. Enrichment plants of this scale take over a decade to design, license, build, and commission. The procurement window for the contracts that will determine who builds this capacity is already open.

Two contractors, two technology paths, and one fatal disqualification at Piketon

The most obvious candidate to fill the gap is already enriching uranium. Centrus Energy began producing high-assay low-enriched uranium (HALEU), uranium enriched between 5% and 20% U-235, at the American Centrifuge Plant in Piketon, Ohio, in late 2023 under a Department of Energy contract.

There is one problem. The centrifuges currently operating at Piketon include components sourced from overseas suppliers, which means the facility cannot produce unobligated uranium for national security purposes. Centrus’s operational track record does not automatically translate into national security contracts until that constraint is resolved.

The NNSA strategy at Piketon has two phases:

• Near-term bridge: Install one or two centrifuge cascades built entirely from U.S.-sourced components, creating an unobligated LEU stream to partially cover demand after the early 2040s.
• Long-term scale-up: Expand Piketon’s capacity, potentially to approach the 1.5 million SWU per year national security requirement from the 2050s onward, using large centrifuge technology at full production scale.

The resolution of the obligated-components constraint is the gate event. Until Centrus demonstrates an all-American cascade producing unobligated material, Piketon’s existing infrastructure is a proof of concept, not a national security asset.

BWX Technologies and the small centrifuge alternative

The federal strategy does not rest on a single contractor. BWX Technologies (BWXT) is developing a parallel enrichment path using small centrifuge technology through its DUECE pilot programme at the Nuclear Fuel Services site in Erwin, Tennessee, with related centrifuge manufacturing in Oak Ridge, Tennessee.

BWXT was awarded an indefinite-delivery/indefinite-quantity (IDIQ) contract valued at approximately $1.5 billion in September 2025. The pilot’s output sequence is deliberate:

• First phase: produce limited quantities of unobligated LEU.
• Second phase: once domestic LEU feedstock is available, potentially produce unobligated HEU suitable for naval reactor fuel.

The two-contractor, two-technology structure reduces the risk of a single point of failure. It also creates competitive pressure that could shape how NNSA awards the larger production contracts still ahead.

What Congress does not know about the $140 billion programme

The GAO’s sharpest finding is not about centrifuges or uranium chemistry. It is about money and who has been told how it will be spent.

The three major national security enrichment programmes carry a combined projected price tag of roughly $140 billion through 2105, yet Congress has not received granular, long-term cost breakdowns that would allow it to track spending across each effort individually.

The GAO issued five recommendations. Three carry the most weight for how this programme gets funded and tracked:

1. NNSA should submit to Congress multi-decade cost projections that span the full life of each of the three national security enrichment programmes.
2. NNSA should provide sufficiently detailed cost estimates for each of the three programmes individually, enabling Congress to exercise meaningful line-item oversight.
3. DOE should complete an independent economic assessment examining whether the scale and structure of current federal interventions are adequate to bring domestic enrichment capacity to the level national security missions demand.

This is not a new concern. The GAO raised similar warnings in 2018, finding that NNSA’s enrichment mission and cost estimates were not adequately defined for planning purposes. Eight years later, the same deficit persists. NNSA and DOE have outlined planned actions in response to all five 2026 recommendations, but planned actions are not completed actions.

Without detailed cost architecture reaching Congress, the appropriations process that funds these contracts becomes unpredictable. For Centrus, BWXT, and their supply chains, that introduces funding-continuity risk across contracts that are designed to run for decades. The five recommendations, if implemented, would force the kind of public disclosure that turns an internal NNSA planning exercise into a congressionally tracked acquisition programme, one where milestones, cost overruns, and contract awards become part of the public record.

The FY 2027 energy budget allocations set the near-term appropriations context within which NNSA must fund its response to the GAO recommendations, making the budget’s nuclear investment line items a leading indicator of how quickly the agency can act on planned cost disclosure commitments.

DoD microreactors and the demand variable that could accelerate everything

The inventory adequacy dates in the GAO report assume current demand projections hold. There is at least one identified variable that could pull the timeline forward.

The Department of Defense is actively developing mobile and expeditionary microreactor concepts designed to provide power in remote or contested environments. These systems would require HALEU fuel, and because they serve defence applications, that HALEU must be unobligated, drawn from the same constrained domestic enrichment pool allocated for weapons and naval programmes.

Three demand vectors now compete for unobligated enrichment capacity:

• Tritium production: confirmed demand, inventory adequate through the early 2040s.
• Naval propulsion: confirmed demand, inventory adequate through the 2050s.
• DoD microreactors: contingent demand, timeline and scale unconfirmed but explicitly identified by the GAO as an unclassified risk factor.

The GAO also noted that some demand factors affecting future unobligated uranium consumption are classified and not publicly disclosed. The 1.7 million SWU per year figure is a floor, not a ceiling.

Why the supply window may be narrower than NNSA’s current timeline suggests

If DoD microreactor programmes move to production scale, procurement of unobligated enrichment capacity could shift forward on a compressed schedule, pulling contract awards and capital deployment earlier than current NNSA planning assumes. NNSA itself acknowledges that domestic fuel cycle constraints may hinder its objectives for increasing enriched uranium output, reinforcing that the 1.5 million SWU per year target is an ambition subject to execution risk, not a guaranteed outcome.

For investors treating the 2040s and 2050s supply deadlines as fixed, the microreactor variable introduces a scenario where urgency arrives earlier and at higher volume than published planning documents reflect.

What the GAO report changes for the domestic enrichment sector

Before this report, domestic uranium enrichment was an internal NNSA planning matter, discussed in classified briefings and procurement documents that rarely reached public markets. After it, the programme has a named price tag ($140 billion), a defined timeline (through 2105), and five formal recommendations pushing toward greater congressional transparency.

The programme horizon extends through 2105, approximately eight decades from the report date, a duration of confirmed federal demand rarely seen in government procurement.

That structural shift matters. Congressional visibility means appropriations battles become more public, contract awards more formally documented, and programme milestones more trackable. Three focal points now define the sector outlook:

The $2.7 billion in domestic enrichment orders placed ahead of the GAO report illustrates how federal procurement strategy has already begun translating long-range planning documents into contracted commercial obligations, years before the inventory adequacy thresholds arrive.

• NNSA recommendation responses: Whether planned actions on the five GAO recommendations translate into actual cost disclosures and the comparative economic analysis DOE has been asked to conduct.
• Piketon obligated-components resolution: Whether Centrus can demonstrate an all-domestic cascade producing unobligated LEU, converting its existing infrastructure advantage into national security contract eligibility.
• BWXT DUECE scale-up trajectory: Whether the $1.5 billion pilot at Erwin produces unobligated LEU on schedule and demonstrates the small centrifuge technology’s viability for naval HEU production.

The confirmed multigenerational federal demand signal extends beyond these two contractors. Uranium miners, converters, and equipment suppliers whose material and technology can qualify as unobligated feedstock now have the clearest government demand backstop the domestic nuclear fuel sector has received in decades.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. These statements regarding future programme costs, timelines, and contract outcomes are based on the GAO’s published findings and NNSA planning documents. They are subject to change based on appropriations decisions, classified demand variables, and programme execution risk.