For seventy years, the United States has been accumulating a problem. It was a problem it once had a plan to solve, but it gave up on that plan almost fifty years ago. During the Cold War, the US intentionally produced weapons-grade plutonium in quantities that were in excess of military needs — about a hundred tonnes of weapons-grade plutonium (>90% Pu-239), much of it machined into pits and warheads. In the 1960s, production began to taper, reactors were shut down, and the very last intentional plutonium production in the US ended in 1987. The Cold War ended and a succession of arms control treaties moved more and more material out of the weapons and into surplus, which was placed in storage. Those materials have been behind heavy security in facilities in South Carolina, Texas, and New Mexico, waiting for someone to decide what to do with it.

Surplus highly-enriched uranium was relatively straightforward to turn to productive civilian use: simply “down-blend” it back the enrichment levels needed for commercial reactors. It was a sin against thermodynamics, but it could be done, and it was done.

Plutonium was different. There is no “natural” form of plutonium with which to downblend it. Fresh civilian nuclear fuel is not designed to include any plutonium at all. Thus there was no immediate path forward with this surplus material. The previous answer was to attempt something akin to downblending and to dilute it with inert material and bury it at the Waste Isolation Pilot Plant in southeastern New Mexico. That program was running years behind schedule and billions of dollars over budget when the Trump administration scrapped it in May 2025 and redirected the material toward a different purpose: fuel for advanced nuclear reactors. The DOE’s Surplus Plutonium Utilization Program then solicited applications from companies that could actually use it, and in late May 2026 selected five for advanced negotiations: Oklo, Exodys Energy, SHINE Technologies, Standard Nuclear, and Flibe Energy.

Flibe Energy is the company I founded and run. So when George Harvey at CleanTechnica published a piece this week titled “Utilize Bomb-Grade Plutonium For Energy? Really?” — anchored by a photograph of a softball and the observation that the Fat Man bomb core was roughly that size — I want to respond to it directly and explain what we are actually proposing to do, and why it makes sense.

What the program is and is not

Harvey’s article, and the Senator Markey letter it reprints at length, frame this as the Trump administration handing weapons-grade plutonium to five startups with minimal oversight. That framing is wrong in ways that matter.

The plutonium will not leave federal custody until the receiving companies have satisfied DOE’s Office of Nuclear Energy, the Nuclear Regulatory Commission, and the International Atomic Energy Agency safeguards requirements that govern all special nuclear material. These are not new requirements invented for this program. They are the same requirements that have governed every gram of special nuclear material in the United States since the Atomic Energy Act. The negotiations currently underway between DOE and the five selected companies are precisely about establishing those conditions in detail. Nothing has been transferred yet.

More fundamentally: the softball photograph, while viscerally effective, obscures the actual risk calculus. The question is not “should weapons-grade plutonium exist?” It already exists, in quantity, with a substantial surplus designated for disposition. The question is what disposition path is safer and more sensible — burying it in a form that requires perpetual monitoring for the next quarter-million years (ten half-lives of 24,000-yr Pu-239), or destroying it through fission in a reactor, converting it irreversibly into fission products that mostly decay in days and weeks rather than millennia. Fission products cannot be reassembled into weapons material. Buried plutonium, in principle, can be recovered. Disposition through peaceful and productive use is a permanent solution, and should be exactly what we do.

Why a molten-salt reactor is the right technology for this

Harvey notes that no commercial reactor has ever run entirely on weapons-grade plutonium and suggests that building one would require starting from scratch. For solid-fueled reactor designs, that is a legitimate concern. Fabricating mixed-oxide fuel from plutonium dioxide and uranium dioxide represents challenging engineering. It can be done and it has been done, but it’s difficult.

In a liquid-fueled reactor, the process is different. The plutonium doesn’t get fabricated into anything. It gets dissolved. Plutonium metal or oxide is converted to plutonium trifluoride — a stable, solid fluoride salt — and that salt goes into solution in the lithium-fluoride-based carrier salt that fills the reactor core. There are no pellets, no cladding, no fuel assemblies. The plutonium becomes part of a fluid, managed by the reactor’s chemistry control systems, circulating through the core where it undergoes permanent fission and generates useful thermal energy.

Think of it this way: instead of pressing a ceramic powder into a pellet and sealing it inside a metal tube, you’re dissolving a fluoride salt into a larger fluoride salt mixture. The liquid already contains the rest of the reactor fuel. When a neutron hits a plutonium nucleus in that solution, fission happens exactly as it would in any other reactor. Energy is released, heating the salt. The plutonium gets consumed. What’s left is fission products dissolved in the salt, which the reactor’s online processing systems can remove continuously.

This is not a hypothetical. The Oak Ridge Molten-Salt Reactor Experiment ran from 1965 to 1969 on uranium-233 and uranium-235 dissolved in FLiBe salt, accumulating roughly 15,000 hours of operation. In the last few months of operation in 1969, plutonium fuel was added to the reactor, and the tests were very successful. The chemistry is understood. The materials are understood. What Flibe Energy is proposing to add is the plutonium trifluoride dissolution step, which is straightforward fluoride chemistry, and the integration of that step with our electrochemical processing work — for which DOE awarded us funding earlier this year.

The thorium transition

If we try to consume plutonium in the presence of uranium, whether in solid or liquid form, we’ll just make more plutonium. But if we consume it in the presence of thorium, we won’t.

Thorium-232 is fertile, not fissile. It cannot sustain a chain reaction by itself. To breed U-233 from thorium, you need an initial source of neutrons — a “driver” fuel that provides the neutrons to begin the conversion of Th-232 into Pa-233 and then into U-233. Once the reactor reaches equilibrium, the U-233 it has bred sustains itself and continues breeding. But you need something to start the process.

Historically the assumption was that highly-enriched uranium, or U-233 from another operating LFTR, would serve as this driver. Plutonium, whether from surplus weapons-grade stockpiles, or recovered from existing spent nuclear fuel, is an excellent alternative. Its fission cross-section in the thermal-neutron spectrum is high — it fissions readily when struck by slow neutrons, which is the environment a graphite-moderated reactor provides. It dissolves cleanly as the trifluoride into the carrier salt. And as it fissions, the neutrons it releases can generate new fuel from the thorium also dissolved in the salt, gradually allowing the startup of more and more dedicated thorium reactors.

In other words: the plutonium goes in, generates electricity, leads to the generation of new fuel, and gets consumed in the process. At the end of that transition, you have reactors running on thorium, and you have destroyed plutonium permanently through fission. I am hard-pressed to think of a more complete solution to the disposition problem.

On Senator Markey’s concerns

Senator Markey’s letter raises two distinct concerns, and they deserve separate treatment.

The proliferation concern I have addressed above in part, but I want to add something that I think gets lost in this entire debate: proliferation, by definition, cannot apply to the United States.

Proliferation means the spread of nuclear weapons capability to nations that do not already possess it. The United States already possesses nuclear weapons. It already possesses surplus plutonium — that is the entire premise of this program. The very existence of the Surplus Plutonium Utilization Program is proof that the United States is not attempting to harvest weapons-grade material from civilian reactors. It is trying to get rid of the weapons-grade material it already has. You cannot proliferate something you already have in abundance and are actively trying to consume.

The legitimate proliferation concern in nuclear energy policy has always been about other nations — about Iran using a civilian enrichment program as cover to approach weapons capability, about the spread of reprocessing technology to states that might extract plutonium from spent fuel for weapons use. U.S. policy for five decades has been precisely to discourage that spread. Nothing in the Surplus Plutonium Utilization Program changes that policy or exports that capability. The plutonium stays in the United States, in federally supervised facilities, and gets fissioned into electricity. As a weapons state, the U.S. is not obligated to place its facilities under International Atomic Energy Agency safeguards the way non-weapons states are — but it has long maintained a voluntary offer to do so, and it may well choose to place reactors in this program under safeguards. That decision hasn’t been made yet. Either way, the material never leaves the most heavily regulated custody chain on earth. Calling that a proliferation risk is a category error. The word doesn’t mean what Senator Markey is using it to mean in this context.

The real answer to that answer is that the concern is about setting a precedent — that if the U.S. normalizes plutonium in civilian reactors, other nations will use that as justification to do the same with less oversight. That is a more serious argument and deserves a serious answer. But it has already been overtaken by reality, and has been for forty years.

France has been using mixed-oxide (MOX) fuel in its commercial reactor fleet since the 1980s. Japan has pursued a civilian plutonium fuel cycle for decades. Germany used MOX before it shut down its reactors. The United Kingdom reprocessed spent fuel at Sellafield for civilian use. The notion that the United States must set a good example by refusing to use plutonium as reactor fuel was Jimmy Carter’s policy, announced in 1977 — and it was promptly ignored by nearly every other nuclear-power-generating country in the world. It did not stop France. It did not stop Japan. It did not stop Russia, which reprocesses spent fuel as a matter of national policy. Carter’s restraint accomplished exactly one thing: it left the United States without a civilian plutonium fuel cycle while the rest of the world built one anyway.

And here is what forty-five years of that policy have actually produced: world plutonium inventories continue to rise. Every operating reactor that uses uranium fuel produces plutonium as a byproduct. That plutonium accumulates in spent fuel assemblies, sitting in cooling ponds and dry casks around the world, with no agreed disposition path and no end in sight. The global separated plutonium inventory — material already extracted from spent fuel and awaiting use — stands at hundreds of tonnes. Getting angry at the United States for wanting to consume its own surplus plutonium in a supervised federal program, while the world’s total inventory climbs without a sink, is not a serious nonproliferation policy. It is a gesture.

The only technology with the genuine potential to reduce world plutonium levels — not stabilize them, not manage them, but actually reduce them — is a molten-salt reactor jacketed by thorium, which consumes plutonium and produces new fuel. Wishing the plutonium away will not make it happen. Getting mad at the people trying to fission it accomplishes nothing. The plutonium exists, and it will continue to exist, in growing quantities, until someone builds the reactors capable of consuming it. That is what Flibe Energy is trying to do.

On solar, wind, and what the grid actually needs

Harvey closes by asserting that solar, wind, and batteries are already cheap, already available, and can be deployed fast enough to solve the energy problem — and that nuclear therefore exists primarily to move large quantities of cash around. This is the laziest argument in the clean energy discourse, and it is flatly contradicted by the data that grid operators publish every year.

Start with demand. NERC’s 2025 Long-Term Reliability Assessment, published in January 2026, projects that summer peak demand across North America will grow by 224 gigawatts over the next decade. That is 69% more growth than NERC projected just one year earlier. The revision is so large because the data center buildout for artificial intelligence is happening faster than anyone modeled. Thirteen of 23 North American assessment areas currently face elevated or high resource adequacy risk. MISO — the grid operator serving the upper Midwest — is rated at high risk of energy shortfalls beginning as soon as this year. PJM, which serves 65 million people from Illinois to New Jersey, faces rising demand with inadequate replacement for retiring thermal capacity. These are not projections from nuclear industry lobbyists. They are the assessments of the organization Congress created to ensure the lights stay on.

Now consider what solar, wind, and batteries actually do and don’t do. Solar produces power roughly six to eight hours a day when the sun is out and the panels are clean. Wind produces power when the wind blows, which is correlated across large regions during the same weather systems that drive peak demand in summer heat waves and winter cold snaps. Battery storage at utility scale currently delivers four to eight hours of discharge — enough to shift the solar peak a few hours into the evening, which is genuinely useful. It is not enough to carry a grid through a week-long winter anticyclone when wind output collapses across an entire region and solar angles are low. Uri, the February 2021 Texas winter storm, lasted six days. Batteries rated at four hours of storage would have been empty by the morning of day two.

The math on what it would actually cost to back a wind-and-solar grid with enough storage to handle multi-day low-resource events is not complicated, but it is never discussed in pieces like Harvey’s. You would need storage measured in weeks of average demand, not hours. The U.S. currently has roughly 50 gigawatts of installed battery storage capacity. Total annual electricity consumption in the United States is approximately 4,000 terawatt-hours. A week’s worth of that is about 77 terawatt-hours. Current installed battery capacity is around 200 gigawatt-hours. We are three orders of magnitude short of what would be needed to back a primarily variable generation grid through a serious multi-day weather event. That gap does not close in ten years. It does not close in twenty. The physics and the economics do not permit it at any cost that a functioning society would accept.

What Harvey calls “cheap” is highly questionable. The levelized cost of electricity (LCOE) figures that renewable advocates cite measure the cost of generating a kilowatt-hour when the resource is available. They do not measure the cost of delivering a reliable kilowatt-hour when the customer needs it. Those are different numbers, and the gap between them grows as variable penetration increases, because each additional increment of variable capacity requires more backup, more transmission, and more grid services to maintain reliability. Germany has spent twenty years and hundreds of billions of euros building out wind and solar, and its residential electricity prices are among the highest in the developed world. California ratepayers pay some of the highest electricity prices in the United States despite — or rather partly because of — the state’s aggressive renewable buildout.

Nuclear energy is the only proven, scalable, carbon-free technology that generates firm, dispatchable power at high energy density without fuel supply chains vulnerable to weather. The argument that it exists primarily to move cash around is an argument that can only be made by someone who has decided in advance that the conclusion is true and is working backward. The grid operators issuing reliability warnings are not making that argument. The industrial customers building factories that need 24/7 power are not making that argument. The data center operators signing long-term nuclear power purchase agreements are not making that argument. The only people making that argument are writers who do not have to answer for what happens when the grid falls short.

A final word on the softball

The Fat Man core weighed about 6.2 kilograms. Weapons-grade plutonium has a critical mass of roughly 10 kilograms in an unreflected sphere, and significantly less with a beryllium reflector and implosion lens system. These are real numbers that demand real safeguards.

But a surplus plutonium stockpile sitting in federal storage is not made safer by leaving it there indefinitely. It is made safer by destroying it productively — converting it, through fission, into energy and fission products that cannot be reassembled into weapons. That is what Flibe Energy is proposing to help the United States do. I hope that this explanation helps you see why this is the right path forward for this material.