The US Hasn't Finished a Nuclear Plant on Time in 30 Years. Japan Just Bet $40 Billion It Can.

Japan just committed $40 billion to build small modular nuclear reactors in the United States. The announcement came from the White House on March 20, during a summit between President Trump and Japanese Prime Minister Sanae Takaichi. There is one problem that nobody in the press conference mentioned: America has not completed a nuclear plant on time or on budget in three decades.

The deal is the centrepiece of a $73 billion US-Japan energy package that also includes $17 billion for natural gas plants in Pennsylvania and $16 billion in Texas. GE Vernova and Hitachi will build BWRX-300 small modular reactors at sites in Tennessee and Alabama. The Tennessee Valley Authority already has a construction permit application under NRC review at its Clinch River site near Oak Ridge, with a decision expected by December 2026.

On paper, the numbers are clean and the timeline is brisk. In practice, every data point from American nuclear construction history argues against optimism.

The Vogtle Problem

The last nuclear reactors built from scratch in the United States — Vogtle Units 3 and 4 in Georgia — were supposed to cost $14 billion and start producing power in 2016 and 2017. They ended up costing $31 billion and entering service in 2023 and 2024. Seven years late. $17 billion over budget. The cost overruns drove Westinghouse Electric, one of the oldest names in American industrial history, into bankruptcy.

Vogtle was supposed to prove that America could still build nuclear. Instead, it proved something else: that the skills, supply chains, and regulatory muscle memory for large nuclear construction had atrophied so badly that even a well-funded project backed by federal loan guarantees couldn't avoid disaster.

SMR proponents argue that small modular reactors solve this problem by design. Factory-built components. Standardised modules. Faster assembly. The BWRX-300 is a 300-megawatt boiling water reactor — roughly a quarter the size of a conventional plant — designed for serial production rather than bespoke construction.

The theory is sound. The evidence is thin.

Zero Operating, One Cancelled

Not a single small modular reactor operates commercially in the United States today. The most advanced American SMR project — NuScale's Carbon Free Power Project in Idaho — was cancelled in November 2023 after costs ballooned and the utility consortium couldn't sign up enough subscribers. NuScale had the only SMR design certified by the NRC. It still couldn't close the economics.

The Union of Concerned Scientists' Edwin Lyman was blunt at the time: "SMRs cannot overcome the fundamental principle of economies of scale." He warned that if NuScale — the simplest, most conventional SMR design — couldn't make the numbers work, the dozens of more exotic reactor types being promoted (sodium-cooled, molten-salt, gas-cooled) faced even steeper odds.

Against this backdrop, the $40 billion figure is less a cost estimate and more a statement of intent. The White House fact sheet said "up to $40 billion" — a phrase that in nuclear construction history has never meant "possibly less."

The Country That Already Did It

While the US announces plans, China finishes buildings. The Linglong One (ACP100), built by China National Nuclear Corporation at the Changjiang site in Hainan Province, is expected to begin commercial operation in the first half of 2026. It will be the world's first commercial land-based SMR.

China broke ground in 2021. It is now completing steam tests and approaching fuel loading. Five years from permit to power. The US, by comparison, has been in pre-application discussions with the NRC since the 2010s. Site preparation at Clinch River "could begin as soon as 2026" — the year China's reactor starts generating electricity.

This isn't just a construction race. It's a regulatory race. China's nuclear safety regulator operates on a fundamentally different timeline. The NRC's review of TVA's construction permit alone is expected to take 18 months. China built an entire reactor in less time than the US takes to review one permit.

Why Japan Is Betting Now

The timing is not coincidental. Three weeks into the Iran war, Japan is facing its worst energy crisis since Fukushima. The Strait of Hormuz — through which 74% of Japan's crude oil transits — is effectively closed. On March 16, Japan began releasing 80 million barrels from its strategic reserves, the largest drawdown since the reserve system was created in 1978.

Japan gets 95% of its oil from the Middle East. That dependency, tolerable in peacetime, becomes existential when a single waterway shuts. Japan's bilateral deal with Iran for passage rights — announced March 21 — bought time. The $40 billion nuclear investment is the structural answer: reduce the dependency itself.

Japan restarted its own largest nuclear plant, Kashiwazaki-Kariwa Unit 6, on February 9 — just days before the war began. The restart was already planned, but the Hormuz closure turned a policy choice into an emergency measure. Japan's 3 weeks of LNG reserves and shrinking oil stockpile make the case for nuclear more forcefully than any climate target ever did.

This is the pattern the Hormuz crisis is accelerating worldwide. Vietnam fast-tracked its switch to ethanol-blended gasoline on March 20, explicitly citing the Iran war. Sri Lanka cut to a four-day work week to conserve fuel. Bangladesh closed universities. Every oil-dependent nation is now asking the same question: what happens when the strait closes again?

The Real Question

The $40 billion bet is not really about whether SMRs work. The BWRX-300 design is technically sound — a simplified boiling water reactor that eliminates much of the complexity (and cost) of traditional plants. Ontario Power Generation is building the first one at Darlington, Canada, with construction well advanced.

The question is whether the United States can build anything nuclear at the quoted price, on the quoted timeline, without the project becoming another Vogtle. The American Bar Association published an analysis in March 2026 noting that SMRs are "poised to make progress toward commercial operation" — the kind of careful language that acknowledges how many previous promises went unfulfilled.

Japan knows this history. Hitachi knows it intimately — the company builds nuclear reactors. The $40 billion isn't blind optimism. It's a calculated hedge: even if the project runs over by 50%, even if the first reactor doesn't operate until 2032 instead of 2029, the strategic value of diversifying away from Hormuz-dependent oil makes the investment rational.

The irony is that it took a war to make the economics work. Not because nuclear got cheaper — it didn't. Because oil got dangerous.

What Happens Next

The NRC is expected to complete its Clinch River permit review by December 2026. If approved, site preparation begins immediately. The first BWRX-300 could be operational by the late 2020s. Alabama sites will follow.

Meanwhile, China's Linglong One will start feeding power to the Hainan grid within months. The EU just allocated €222 million to fusion research for 2026-2027. The global nuclear buildout is accelerating on every continent simultaneously — driven not by climate ambition but by the sudden, violent reminder that oil dependency is a strategic liability.

The US and Japan are betting $40 billion that small modular reactors are the answer. History says the budget will double and the timeline will stretch. The Hormuz crisis says they have to try anyway.

The strait closes. The reactors don't build themselves. And the country that solves this first wins a lot more than an energy contract.