Commercial fusion and US nuclear warheads depend on the same scarce, decaying isotope, and Washington has only a few years to rebuild the supply chain for both.

The United States began funding fusion research in 1951 through Project Sherwood, a classified Atomic Energy Commission initiative. Then, a December 2022 ignition at Lawrence Livermore National Laboratory changed the trajectory of fusion. It ushered in a sizable group of modern private fusion companies, which have raised over $14 billion in capital. Commonwealth Fusion Systems has raised $4 billion as it finishes its SPARC demonstration machine and prepares to build ARC, a 400-megawatt plant in Virginia meant to connect to the grid in the early 2030s. Similarly, the Department of Energy (DOE) has a milestone program that is funding pilot-plant designs with eight companies.

While considerable attention and capital are focused on containment machines and plasma physics, the specific fuel needed is getting less scrutiny. Most private ventures plan to burn a mixture of deuterium and tritium. While deuterium is relatively easily extracted from seawater, no significant commercial production industry exists for tritium. Already, the Bulletin of the Atomic Scientists has tracked the emerging risks of this technology, focusing on proliferation risks and the entanglement of fusion with the broader drawbacks of deuterium-tritium fuel cycles. The world is now racing to build fusion machines that depend on a fuel for which no commercial market exists, while every modern nuclear arsenal depends on the same isotope for warheads. The tritium production industrial supply bottleneck connects these two sectors.

Why Fusion and Nuclear Weapons Compete for Tritium

The National Nuclear Security Administration (NNSA) sustains the nuclear-weapons stockpile, while the US Air Force and Navy operate the delivery systems. Critically, tritium acts as a boost gas to improve warhead yields. With a radioactive half-life of 12.33 years, 5.5 percent of any tritium inventory decays into helium-3 each year, meaning more than 40 percent disappears over a decade. Thus, the DOE treats tritium components as limited-life assemblies that require scheduled replacements for operational warheads.

The civilian tritium market is barely a market since supply comes almost entirely from heavy-water power reactors, which generate tritium as a byproduct. Ontario Power Generation’s Darlington facility in Canada, the world’s largest commercial extraction site, collects around 1.5 kilograms a year. South Korea operates a smaller recovery facility at Wolsong, extracting under one kilogram annually. Both sites exist to remove tritium for worker safety and environmental compliance. Total accessible global civilian stocks cluster between 20 and 30 kilograms, priced near $33,000 per gram. Outside of fusion labs, buyers purchase minuscule amounts of tritium primarily for “powering the glow” of emergency exit signs and watch dials, niche scientific research, and micro-batteries for long-duration space and sensor applications. 

While fusion developers expect their future commercial plants to breed their own fuel by surrounding the burning plasma with lithium blankets, a reactor must first be fueled with externally supplied tritium before it can breed on its own. For example, a plant producing 1 gigawatt of fusion power, close to the scale of what ARC is designed for, will need several kilograms of tritium on day one, with published estimates ranging from under 1 to about 18 kilograms, with an annual consumption of approximately 55 kilograms. The ITER project in France, which is slated to come online in the late 2030s, has already staked a substantial claim on that small stockpile and will consume most of the tritium in Canadian reserves, leaving commercial startups potentially dry.

Because tritium decays constantly, stockpiling is not a viable strategy for the commercial sector or the military. Production capacity is what must be “stockpiled.”

How the US Nuclear Weapons Program Makes Tritium

The tritium supply chain for a nuclear warhead begins with lithium-6 sealed in lithium-aluminate pellets inside tritium-producing burnable absorber rods. These rods travel to the Tennessee Valley Authority’s Watts Bar nuclear plant, a commercial facility that has irradiated defense rods since 2003.

After absorbing neutrons in Tennessee, the assemblies are shipped to the Department of Energy’s Savannah River Site in South Carolina. There, inside the Tritium Extraction Facility, the rods are heated under vacuum to release the gas, which is purified and loaded into warhead reservoirs. Savannah River also unloads returned reservoirs from serviced military warheads, purifying residual material back into the pipeline.

To keep pace with warhead modernization, utility regulators raised the ceiling to permit up to 2,496 rods per reactor unitunder a 2024 amended record of decision. In 2026, the National Nuclear Security Administration announced that Savannah River completed 13 extraction cycles in nine months. These technical gains satisfy military requirements, but they provide no fuel to commercial fusion ventures.

Under international nonproliferation agreements, Watts Bar can irradiate defense rods only when fueled by low-enriched uranium free of foreign peaceful-use obligations. The United States shuttered its last commercial enrichment plant utilizing domestic technology in 2013. Government audits confirm that domestic stockpiles of unobligated low-enriched uranium will support defense tritium production only into the early 2040s. While replacement centrifuge facilities are planned (through commercial contracts) at Piketon, Ohio, any disruptions to their tight schedule could cause reactors to be short of needed fuel.

Every gram harvested at Watts Bar is designated for military use. Selling tritium meant for weapon systems to commercial startups would undermine the principle of keeping civilian and military nuclear activities separate. This boundary blurs into nonproliferation norms, which require strict separation between national security stockpiles and commercial markets. 

Why Lithium-6 Is Tritium’s Upstream Choke Point

Making more tritium leads directly to another isotope the United States stopped producing more than 60 years ago. Lithium-6 constitutes just 7.5 percent of natural lithium. The United States ended domestic lithium enrichment in 1963, when the Y-12 National Security Complex closed its production lines after deciding Cold War reserves were adequate. Since then, the US nuclear arsenal has survived by recovering lithium from retired warheads. While a new Lithium Processing Facility at Y-12 will modernize recycling in the 2030s, Washington has built no production-scale capability to enrich new lithium isotopes from raw feedstocks.

Commercial fusion will eventually require enriched lithium-6 at scale, because most proposed breeding blankets require lithium enriched far above natural levels to produce replacement tritium. Today, Russia and China are the only commercial exporters of separated lithium isotopes, selling lithium-7 for civilian reactor coolant chemistry as a direct byproduct of their military enrichment programs.

Two allied projects will add modestly to the civilian tritium pool. Romania and South Korea are constructing Europe’s first tritium removal and a commercial recovery facility at Cernavoda, still several years from operation, which will capture roughly 130 grams per reactor annually. In Britain, the UK Atomic Energy Authority and Eni are building the H3AT tritium loop at Culham to demonstrate closed-loop fuel recycling at pilot scale by 2030. These projects expand scientific know-how, but they cannot close the multi-kilogram startup deficit facing commercial fusion fleets. 

Fuel-chain node	Vulnerability
Lithium-6 feedstock	No US lithium enrichment since 1963; entire national supply depends on finite lithium recovered from retired warheads, leaving no margin for civilian fusion scale-up
Reactor irradiation	Single-point failure risk; Watts Bar Units 1 and 2 are co-located at a single Tennessee Valley Authority site, creating a major vulnerability
Unobligated low-enriched uranium*	Impending supply cliff; current domestic inventory of unobligated uranium can only support defense tritium production into the early 2040s. Replacement domestic enrichment capabilities will produce meaningful volumes sometime between 2030 and 2035
Extraction, recycling, and reservoir loading	Monopolized infrastructure; all warhead reservoir loading operations have been consolidated to the Savannah River Site, leaving the entire US nuclear arsenal vulnerable to a single industrial disruption
Civilian tritium supply	Approximately 25 kilograms worldwide; decaying 5.5 percent a year and tied to aging heavy-water reactor fleets

The Energy Department has a fusion roadmap and admits that no integrated breeding blanket has ever been constructed and validated under power plant conditions. Testing facilities like UNITY-3, being developed at Oak Ridge National Laboratory, are only beginning to address continuous extraction and accounting. 

How Washington Can Rebuild the Tritium Supply Chain

Commercial fusion ventures may not be consuming kilograms of tritium from the market, but the National Nuclear Security Administration is fulfilling its near-term nuclear stockpile targets. Policymakers in Washington have only a few years left to intervene before the demands of these two programs start to falter. 

Four things could help:

First, the National Security Council should direct an interagency process, led by the Department of Energy and the National Nuclear Security Administration, to build an integrated, 20-year projection of tritium needs. This strategic assessment must reconcile classified stockpile modernization targets with commercial fusion milestones, startup fuel inventories, and the retirement schedules of Canadian and South Korean heavy-water reactors. While weapons data must remain classified, this analysis should help the government proactively identify where civilian demand will strain identical materials, facilities, and specialized workforces.

Second, lithium isotope separation must be recognized as critical national defense and energy infrastructure. Relying indefinitely on the dismantling of Cold War weapons for lithium-6 is not viable. The Department of Energy and the Pentagon should utilize Defense Production Act Title III authorities to finance and scale modern lithium-6 separation technologies, including advanced chemical-exchange and laser-based separation methods. 

Third, the civilian breeding fuel cycle should be stress-tested. Alternative civilian production sources should be supported by the government, such as the accelerator-driven molten-salt project at Los Alamos National Laboratory, which transmutes commercial nuclear waste into tritium for private fusion ventures. Developing parallel production channels alongside allied facilities like the British H3AT tritium loop can help decouple commercial fusion development from nuclear stockpile dependencies.

Finally, Congress must resolve the unobligated uranium deficit feeding the supply chain for nuclear weapons. Under international peaceful-use agreements, the Tennessee Valley Authority’s Watts Bar reactors can only irradiate tritium-producing rods if fueled by domestic, unobligated low-enriched uranium. Unfortunately, the National Nuclear Security Administration projects its unobligated low-enriched uranium inventory will be exhausted by 2041. Keeping the Piketon and BWXT enrichment plants funded and on schedule is the precondition for sustaining weapons tritium production, and Congress should treat any slip as a national security problem rather than a procurement delay.

Tritium is yet another lesson of what happens when strategic materials are taken for granted for advanced technologies. Governments usually prepare for scarcity by accumulating stockpiles, but that does not work for a decaying isotope. The United States is now set on two nuclear paths, and both depend on the functioning of a specialized industrial ecosystem that cannot scale. Getting ahead of the tritium gap means building the capacity to replenish supply before defense and civilian demand strains the fragile chain with no capacity to surge. 

About the Authors: Macdonald Amoah, Morgan Bazilian, and Jahara Matisek

Macdonald Amoah is an independent researcher with interests across critical mineral supply chains, advanced manufacturing gaps, the defense industrial base, and the geopolitical risks of the mining sector.

Morgan D. Bazilian is the director of the Payne Institute and professor at the Colorado School of Mines, with over 30 years of experience in global energy and minerals security policy and investment. A former World Bank lead energy specialist and senior diplomat at the UN, he has held roles at NLR and in the Irish government.

Dr. Jahara Matisek is a senior fellow at the Payne Institute for Public Policy and a visiting scholar at Northwestern University. He has published over 200 articles on defense industrial base issues, strategy, and warfare.

Disclaimer: The views expressed are those of the authors and do not reflect the official position of the US Air Force, Department of Defense, or the US Government.

The post How America’s Tritium Shortage Threatens Fusion Power and Its Nuclear Arsenal appeared first on The National Interest.