Join War on the Rocks and gain access to content trusted by policymakers, military leaders, and strategic thinkers worldwide.
Most accounts of the Manhattan Project tell a familiar story: American scientists relentlessly racing to design and build the fission bombs that would ultimately fall on Japan. A neglected chapter of the saga was the obsessive advocacy of a handful of physicists who favored a far more potent weapon design. The “Super” of their imagination would derive its explosive energy not simply from fission, or the splitting of atomic nuclei, but also from the fusion of deuterium and tritium, two isotopes of hydrogen. Although wartime expediency demanded the more straightforward fission path to a weapon, such hydrogen bombs would soon become the mainstays of the world’s nuclear arsenals and remain so to this day.
The need for plutonium or highly enriched uranium to build a nuclear weapon is widely understood. But to achieve a modern, two-stage thermonuclear warhead of the sort fielded by the United States and every other advanced nuclear power, deuterium and tritium are no less indispensable. Deuterium is abundant in the Earth’s oceans and presents no barrier to the nuclear club. Tritium, on the other hand, is among the rarest substances known to man and must be produced in a nuclear reactor. If one could condense all the commercial tritium in the world into a single mass, it would amount to no more than 25 kilograms and fit comfortably in a duffel bag.
Every active warhead in America’s nuclear stockpile contains tritium, which has a half-life of just over 12 years. This requires the reservoirs that hold the gas to be regularly replenished and, in turn, necessitates the continuous production of fresh tritium to supplement the small amount that can be recycled from existing weapons. Tritium is the seemingly inconsequential item upon which everything else depends. Without it, there can be no modern nuclear weapons, and thus no nuclear deterrence, the cornerstone of American defense.
My colleagues and I at the National Nuclear Security Administration are responsible for manufacturing the thousands of components and materials that comprise each of America’s nuclear warheads. Because the lead time for producing many of these ingredients is measured in years, we often must develop them — or at least the wherewithal to do so — long before a formal requirement is issued. It remains to be seen whether the nuclear build-ups of Russia, China, and North Korea will require a reciprocal expansion of America’s atomic arsenal. But prudence demands that such a decision not be constrained by want of a particular component or material. For this and other reasons, the United States is now increasing its capacity to produce tritium, a process that provides a fascinating glimpse into the scientific ingenuity of the U.S. nuclear weapons enterprise.
For Want of a Nail…
Tritium was discovered in 1934 by a team of physicists led by the pioneering scientist Ernest Rutherford. Unlike the nucleus of the hydrogen-1 isotope, which contains one proton and zero neutrons, or hydrogen-2, which contains one of each, the hydrogen-3 nucleus holds one proton and two neutrons, hence the derivation of “tritium” from the Ancient Greek trítos, or “third.” Owing to this property, physicists recognized its potential in an atomic bomb even before consensus had been reached on the feasibility of such a device, much less before one was built. In 1942, J. Robert Oppenheimer convened a gathering of luminaries in the field of physics to explore the theoretical basis of a weapon design. During discussion of a bomb driven by the fusion of deuterium with deuterium, a young physicist named Emil Konopinski suggested adding tritium to boost the yield of the device. Years would pass, however, before his insight shook the world — literally and figuratively.
Tritium production for nuclear weapons began in earnest in the United States in late 1953, when R reactor at the Savannah River Site in South Carolina went critical. The complex was part of a constellation of facilities that President Harry Truman ordered built in 1950 following his decision to develop hydrogen bombs in response to the Soviet nuclear threat. Throughout the Cold War, almost all of the stockpile’s tritium was bred in five Savannah River reactors operated specifically to produce weapons materials. But by the late 1980s, antagonism with the Soviet Union had begun to recede, and scrutiny of nuclear facilities following the Chernobyl disaster cast an unforgiving spotlight on the safety of the site’s reactors. By 1988 the last of the units had ceased operations, leaving the United States without a dedicated tritium production facility.
From Cold War Reactors to Civilian Power
What followed was a 15-year interregnum before the Department of Energy adopted a new approach to making tritium, during which stockpile needs were met by harvesting tritium gas from weapons being dismantled after the Cold War. The new paradigm was the most affordable of the options studied. Rather than building an expensive sole-purpose facility, a unique arrangement was made with the Tennessee Valley Authority, a federally owned electric utility corporation, to produce tritium by piggybacking on the operation of the Watts Bar Nuclear Plant. Situated between Chattanooga and Knoxville, the civilian plant’s two light water reactors supply electricity to 1.3 million American homes. But with a slight modification to the units’ configuration, and without hitting ratepayers’ pocketbooks, the plant would also provide for the entire nation’s defense.
The protagonist of this storyline is the tritium-producing burnable absorber rod. Roughly 13-foot-long tubes of lithium pellets, the rods are inserted into the fuel assemblies of the Watts Bar reactors. The “absorber” in the name refers to their role in capturing excess neutrons, which must be carefully controlled in the reactor core to maintain a smooth and continuous nuclear reaction. A neutron “poison” like boron or gadolinium is usually used for this purpose, but swapping them for lithium serves a complementary function, and the irradiation process breeds tritium to boot.
The rods begin their lives at the Pacific Northwest National Laboratory. As the design agency for the rods, the lab performs research and engineering to optimize their manufacture and defines experiments on constituent materials to understand their performance inside a reactor. Experimental capsules are irradiated in a reactor at Idaho National Laboratory, and the ever-evolving materials that comprise the rods are then sourced from private companies all over the country.
Once fabricated, the rods are inserted into the Watts Bar reactors, where they marinate for an 18-month operating cycle. Finally, a commercial vendor ships the irradiated rods to the Savannah River Site, where tritium is extracted and purified for a range of national security purposes. Chief among these is its use in “gas transfer systems,” the tiny metal bottles that store tritium gas inside America’s nuclear warheads.
Engineering More from the Same Core
The first tritium harvest from Watts Bar in 2007 yielded a mere 223 grams. To meet stockpile needs, in 2015 the Department of Energy set a target to produce a combined 2,800 grams of tritium in each rolling 18-month cycle of the two reactors by 2025. The goal was later raised to 3,300 grams by 2027, which engineering improvements allowed to be reached two years early. By the spring of 2025, engineers calculated that the next tritium yield would be higher still, a veritable bumper crop of roughly 3,400 grams. (The exact figure will not be known with precision until the tritium is extracted from the rods and will then be classified.) How they managed this feat is a case study in innovative nuclear engineering that has enormous implications for the future weapons stockpile.
Tritium output is a function of a soup of variables: the number of tritium-producing burnable absorber rods inserted into the reactors, their tolerances, the finicky process of extracting tritium, and so on. The fastest and most fruitful means of boosting tritium production is simply to stuff more rods into a reactor, which of course is not as simple as it sounds. How they are positioned in the cores matters a great deal. Moreover, to accommodate additional rods, the enrichment level of the uranium and the ratio between the rods and the fuel assemblies must be adjusted. The reactors are there to produce commercial energy, after all, so the outcome must be that the Tennessee Valley Authority gets all the electricity it needs, the stockpile gets its tritium, and everyone is happy.
Naturally, there is more to the story. The Watts Bar units were already operating near the maximum allowable number of rods. Upon careful review, the Nuclear Regulatory Commission agreed to increase the quantity that can be inserted in each unit. Then there were changes to the rods themselves. More lithium going in means more tritium coming out, but only a certain amount of tritium can build up inside each rod before it bursts or cracks. Years of operating experience have shown that each rod can withstand a higher tritium build-up than originally thought, allowing more lithium to be added and contributing to the recent banner yield. The designers are now flushed with optimism that the output of the two reactors’ 18-month cycles can reach as high as 5,800 grams.
Fuel, Sovereignty, and the Nonproliferation Constraint
To increase the tritium supply, however, the ability to juice the production process is not the only factor. The reactor fuel used to make tritium must fall into a special category having nothing to do with the fuel’s physical characteristics. As one of only five countries permitted to possess nuclear weapons under the Treaty on the Non-Proliferation of Nuclear Weapons, the United States relies exclusively on domestically enriched uranium for tritium production. Likewise, the United States assures foreign suppliers that their fuel exports will be used entirely for peaceful purposes. In doing so, U.S. policy avoids incentivizing other countries to develop capabilities that could be used for weapons programs, encouraging them to honor their own treaty obligation not to assist any state in developing nuclear weapons. Further, a supply chain free from foreign obligations gives us the flexibility to meet deterrence requirements without overseas interference. The U.S. stockpile’s thirst for tritium has therefore motivated an energetic effort to secure enough permissible nuclear fuel.
Happily, the Department of Energy has identified enough unobligated fuel to power the Watts Bar reactors through the early 2040s. Much of this material is highly enriched uranium “scrap” that is being downblended to a lower enrichment level for use in the units. Meanwhile, the department is reconstituting the technology to enrich uranium domestically, which the U.S. government discontinued in 1992. The resulting fuel will provide low-enriched uranium for tritium production as well as highly enriched uranium for the Navy’s nuclear fleet.
A Growing Appetite
Deterrence is materially fragile. Without steady tritium production, America’s stockpile of nuclear weapons will atrophy regardless of policy intent. This also means that capacity is leverage against America’s rivals. With Washington facing simultaneous nuclear competition with Russia and China, U.S. leaders will want options: expand, hedge, upload, adjust the nuclear posture, and so on. These options only exist if the industrial base is already capable of scaling. And critically, sovereignty matters. Because tritium production depends on unobligated, domestically enriched uranium, supply chain independence becomes a strategic variable, not merely a procurement issue. Therefore, fuel policy is deterrence policy.
In the years ahead, a variety of civil and military applications of tritium will inevitably drive demand higher. Use of the isotope in the private sector, now just a few hundred grams per year worldwide, is limited by its exorbitant expense. At a commercial price of $30,000 per gram, an ounce of tritium would fetch around $850,000, compared to $5,200 for an ounce of gold. So, it is restricted to niche uses stemming from its luminescent properties, such as firearm sights, the dials of high-end watches, and night illumination of traffic signs. But its potential is immense.
Perhaps tritium’s most promising economic utility is not in expensive Swiss wristwatches but in the possibility of limitless low-carbon energy. Commercial-scale nuclear fusion reactors would devour vastly more tritium than is now available, requiring leaps in breeding technology. Although formidable technical and economic obstacles must be overcome before this dream can be realized, researchers across the nuclear enterprise are aggressively advancing tritium science.
Savannah River National Laboratory, for example, is improving the systems and processes for tritium production, as well as studying efficiencies in tritium use to enable fusion energy plants. From purely commercial uses to capabilities being investigated for homeland defense, the capacity to produce tritium will one day be central to some of the nation’s most ambitious economic and security projects. For now, though, tritium’s predominant use will remain its most iconic one: as a fusion fuel in nuclear weapons.
Whether Russia’s and China’s nuclear build-ups must be met with an expansion of the U.S. nuclear force is an open question. But signs point to the possibility at least that this response may be necessary. The Strategic Posture Commission concluded in 2023 that our current nuclear modernization campaign is “necessary, but not sufficient” to meet the unprecedented challenge of deterring two nuclear near-peers simultaneously. As the nation’s leaders assess the future requirements of deterrence, the nuclear enterprise would do them a profound disservice by limiting their decision space. Building the means to produce strategic materials for an expanded stockpile will maximize America’s policy options in a perilous new nuclear age. But it will do more than that, underscoring to our adversaries that this strategic competition is one we do not intend to lose.
Audrey Beldio serves as the principal assistant deputy administrator in the National Nuclear Security Administration’s Office of Defense Programs, where she is responsible for modernizing the production of strategic materials and components for the U.S. nuclear weapons stockpile.
Image: National Nuclear Security Administration
