A thing that actually happened · figures unverified this session and labeled · the objections kept on the page

Megatons to Megawatts

TL;DRFor twenty years, uranium from dismantled Soviet warheads was downblended into ordinary reactor fuel and burned in American power stations. A large share of US electricity came out of the arsenal that had been aimed at it, and nobody put a sticker on the bill.

Swords into plowshares, actually executed, at industrial scale, for two decades — and it is the least-known good news of the last fifty years.

What happened

Beginning in 1993 and running until 2013, under an agreement between the United States and Russia, Soviet warheads were dismantled and their highly enriched uranium was downblended — diluted with ordinary uranium down to the low enrichment a civilian power reactor burns. The United States bought the resulting fuel. The programme was called Megatons to Megawatts.

Roughly 500 metric tons of weapons-grade uranium, reported as something on the order of twenty thousand warheads’ worth.

Supplying, at its height, about half of all American nuclear fuel.

Which worked out to roughly one in ten American lightbulbs, for about twenty years.

Sit with the last one. For two decades, something like a tenth of this country’s electricity came out of warheads that had been pointed at it. People ran their refrigerators, their dialysis machines and their school gymnasiums on the Soviet arsenal, and there was no sticker on the bill, no press campaign, no line on the statement. It simply happened, quietly, at industrial scale, while everyone argued about something else.

The single greatest disarmament story of the century, and it arrived disguised as a utility invoice.

Why it stopped

The agreement ran its term and ended in 2013, and it was not renewed. Russia concluded it would rather sell enrichment services than sell off the material itself — a one-time sale of stock versus an ongoing sale of capability. Whatever cooperation made the arrangement possible in the first place is plainly not on the table now.

And there is a genuinely bitter irony buried in it. The flood of cheap downblended fuel undercut the uranium market badly enough to damage domestic mining and enrichment. Part of the reason the United States has an enrichment capacity problem today is that for twenty years it did not need one. The peace dividend ate the supply chain that would have replaced it.

The plutonium half, which failed the way everything else in nuclear fails

Uranium was the easy one. Plutonium is harder, and the American attempt to do the same trick with it is a case study in the real constraint.

The MOX facility at Savannah River was built to turn tens of tons of surplus weapons plutonium into mixed-oxide reactor fuel. It was cancelled after billions had been spent and the schedule had collapsed. The fallback is “dilute and dispose”: blend the material with inert filler and bury it. Not fuel. Not power. Just gone, at a cost, forever.

The physics worked. The accounting didn’t.

Which is the same sentence that explains Vogtle, V.C. Summer, and every cancelled small reactor. Nuclear in America does not usually fail on science or on protest. It fails on a spreadsheet, and it fails there so reliably that the spreadsheet is the actual subject.

The strongest objections

It was a commercial deal, not a moral one. Russia was paid, in hard currency, at a moment when it badly needed hard currency. Calling it swords into plowshares flatters everyone involved; calling it a well-structured purchase is more accurate.

It processed surplus, not deployed weapons. Downblending material already declared excess is not the same as reducing an active arsenal. The warheads that mattered strategically stayed where they were.

It did real economic damage. The cheap fuel hollowed out domestic enrichment, and that bill is being paid now with interest.

And it does not scale to the present question. Anyone pointing at this as the answer to data-centre demand should notice it ended over a decade ago, the geopolitical conditions are gone, and the remaining weapons-usable stockpile is a fraction of what was sold.

The ruling

All four granted. It was commercial, it was surplus, it hurt an industry, and it is not a plan for 2026. This page is not proposing it as an energy policy, because it is not available as one.

What it is, is proof of concept at a scale nobody expected to see. The standard objection to disarmament is that the material has to go somewhere and no one will pay to handle it. For twenty years, the material went into the grid and somebody paid for it happily, because it was the cheapest fuel on the market. The thing everyone said was impractical turned out to be the practical option, and the practical option turned out to be so undramatic that the public never learned it was happening.

So when someone says the arsenal is a sunk asset that can only sit there, the honest answer is: a fifth of it already got burned for electricity and your lights were on the whole time.

The invoice ran backward

The argument next door is that if deterrence worked, the rest of the budget is a subsidy — you already paid for the peace, and everything since is a separate charge.

This is the one stretch where the ledger reversed. The arsenal was paid for once, in the largest public expenditure of the century. Then, for twenty years, a piece of it paid something back, in kilowatt-hours, to the households that had funded it.

For twenty years, the invoice ran backward. It can be done. It was done. Nobody noticed.

Where the house stands, and how to check it. The figures above are unverified this session — the machine that drafted this could not reach a source, so 500 tonnes, the share of American nuclear fuel, the warhead equivalent and the MOX tonnage are all given from memory and hedged in the text on purpose. Check them at the National Nuclear Security Administration, the Department of Energy, the operator records of the USEC/Centrus programme, and, for the MOX cancellation, the GAO reports on its cost. The argument survives being wrong about the exact numbers: whether it was a tenth of American electricity or a twentieth, the fact that a large share of a country's power came out of the weapons of its former adversary for two decades is the point, and it is robust. If a figure here is off, send the correction and it goes up louder than the original. Opinion: the ruling, the framing, and the reading of what it proves. Held to the one rule: no lying. ;