DOE Spent Fuel Liabilities Hit $44.5 Billion as GAO Presses for Cost Transparency
Nuclear fuel supply programs are advancing, but cost overruns at legacy projects threaten to overshadow new subsidies and private capital.
A 2024 financial audit found that the Department of Energy has accumulated liabilities of $37.6 to $44.5 billion from its failure to manage spent nuclear fuel over the long term, a figure now under scrutiny in a new government watchdog assessment of federal nuclear fuel supply efforts. The number lands as Washington pushes billions in new loan authority toward domestic uranium enrichment and conversion, yet offers no clear answer on who ultimately pays for the waste.4,8
The federal government owns the spent fuel liability, but disposal and reprocessing cost trajectories are climbing in ways that could crowd out the supply-side investments the loan program is meant to stimulate. The URA uranium ETF slipped 0.25% to $48.14 on Wednesday (2026-08-26), but the policy arithmetic runs deeper than spot prices.8,4
The transparency problem has precedent. At Savannah River, the MOX fuel fabrication project saw cost estimates climb from about $4.9 billion to roughly $17 billion before its projected completion date slipped to 2048, leading to the project's termination in 2018. Every new nuclear fuel initiative the government backs with loan guarantees inherits that history.2
Japan offers a live case in the same dynamic. JNFL's Rokkasho spent fuel reprocessing plant has seen its final completion cost rise to ¥15.98 trillion, up ¥360 billion from the previous year's estimate and the highest yet recorded. Construction costs alone rose ¥180 billion to ¥3.92 trillion, and the MOX fuel plant at the same site reached ¥2.68 trillion. Reprocessing spent fuel, which is roughly 95-97% reusable uranium and plutonium, reduces high-level radioactive waste volumes. The price tag for doing so keeps climbing anyway, and Tokyo has no cheaper alternative on the table.3
Back in the United States, the fuel supply picture is tightening even as reactor demand holds. U.S. nuclear plants loaded 40.9 million pounds of uranium into reactor cores in 2025, a 15% drop from the 48.1 million pounds installed in 2024. Yet 93% of the uranium used in U.S. reactors in 2026 was imported, with total fuel loading reaching about 41 million pounds.7,8
Domestic production is rising but starts from a thin base. Import dependence is structural, not cyclical. Most global supply chains run through Russia, with raw uranium supplies largely cornered by Russia and China.4
Private capital is moving into the gap. Oklo, the largest of the advanced reactor startups, reported $2.5 billion in cash and marketable securities as of Q1 2026, supported in part by a $1.2 billion At-the-Market offering. SHINE has raised $240 million in equity funding, bringing total investor backing to $1 billion. But the bottleneck is not at the reactor design stage.2
ASP Isotopes announced on July 21 (2026-07-21) a research agreement with Texas A&M Engineering Experiment Station to develop scalable domestic uranium conversion technology, targeting what the company describes as a critical U.S. nuclear fuel supply chain gap. Conversion capacity is where U.S. redundancy is thinnest.5
India faces the same structural bind from a different angle. The country currently operates about 7,900 MW of capacity across 24 nuclear plants and has 17 reactors totaling 13,100 MW either under construction or in pre-project activities. NPCIL's own estimates show PHWRs alone could require roughly 5,400 tonnes of U3O8 per annum for about 25 GWe of capacity, while UCIL currently supplies only around 30% of that need. Persistent import reliance is built into the expansion plan from the start.6
The United States, meanwhile, holds more than 50 tons of plutonium left behind by nuclear weapons programs. Nuclear startups are in advanced negotiations to acquire some of that Cold War stockpile, but the Savannah River experience shows the gap between negotiating and delivering is precisely where cost overruns accumulate.1,2
The GAO assessment ties these threads together: federal loan programs can accelerate enrichment and conversion capacity, but without cost transparency on the back end, accumulated liabilities could erode the benefits. A DOE carrying $44.5 billion in unresolved spent fuel obligations may find its new loan authority stretched thinner than projected once waste bills come due. The first concrete signal arrives when the department names specific companies receiving conditional loan commitments and whether their disclosed cost projections cover the full spent fuel cycle.4,5