For most of the 21st century, the story of US nuclear power was one of slow, quiet decline. With no new plants constructed, a string of early retirements, and a fleet coasting on operating licenses extended into their sixth and seventh decades, it was clear that nuclear was not a priority in US energy policy. The nuclear page of this site still carries that history with 23 decommissioned plants that sit alongside the aging 54 active plants. However, for the first time in a generation, the course of nuclear power has begun to shift, and the map now shows categories that did not exist in any picture of US power a decade ago.
The 54 active US nuclear plants represent about 97 gigawatts of capacity, but their real significance is in how relentlessly they run. Reactors operate at capacity factors above 90 percent which is far higher than any other source. They generate roughly a fifth of the nation’s electricity and about half of its carbon-free electricity from a tiny number of sites. That concentration cuts both ways. It means a single large station produces an enormous, steady block of clean power; it also means the closure of just one plant is national news, because there is no quick way to replace that much firm, zero-carbon output. It is also why the fleet’s economics became precarious in the 2010s. In wholesale markets flooded with cheap shale gas and subsidized renewables, several profitable-to-build but expensive-to-run reactors could no longer cover their costs. The result was that owners closed plants which created the substantial number of retirements that make up about a third of the past, present, and future plants. Keeping the remaining fleet open then became a bipartisan priority, backed by state zero-emission-credit programs and federal support, on the logic that losing carbon-free megawatt-hours and replacing them with gas was a step backward.
The most surprising entries on the nuclear map are the plants marked “restart pending.” A few years ago, restarting a closed reactor was essentially unheard of. Once a plant began decommissioning, it was permanently removed from service, but now that assumption has broken. Driven by surging electricity demand, primarily caused by data centers and artificial-intelligence computing, owners have moved to bring shuttered reactors back online. Operators of massive data centers want round-the-clock, carbon-free power. They are more than willing to sign long contracts and shell out funds to bring these decommissioned plants back up to code. Restarts are attractive because the hardest, most expensive work is already done: the site, the reactor, the transmission connection and the trained workforce mostly exist. Recommissioning is cheaper and faster than building new, even after years of inspections and upgrades. The appearance of this category on the map is arguably the clearest single sign of how sharply sentiment around nuclear has turned.
The plants marked “planned or under construction” capture the other half of the revival. The completion of two new large reactors in the Southeast, the first newly built US units in decades, proved the country can still construct conventional, large-scale nuclear. With the project running years late and billions over budget it also showed the country just how hard it is to build from the ground up. Those overruns are the backdrop to everything happening now, because they explain why much of the industry’s hope has migrated to a fundamentally different way of building reactors.
Small modular reactors (SMRs) aim to replace giant, bespoke construction projects with smaller units built in factories and assembled on site, deployed in series rather than one at a time. The promise is that standardization and repetition tame the cost and schedule risks that have plagued large builds. This is the same logic that made aircraft, ships, and cars cheaper once they were mass-produced. A range of designs is advancing, from light-water SMRs to high-temperature gas, molten-salt and sodium-cooled fast reactors. Several of these new reactor designs are backed by federal cost-sharing and site selections, with the first advanced-reactor construction permits already issued. However, none of the current progress towards SMR actualization guarantees success. SMRs still must prove they can be built on time and on budget at commercial scale. Furthermore, the first units will be the most expensive, and they will require massive upfront capital to complete the research and development required to produce a reliable, fully-tested reactor. The shift towards nuclear is visible; with a fleet that only shrank for twenty years, it now has a credible path to growth. The unique technological developments happening in our time have created a queue of new customers, from utilities to technology companies, who are watching closely.
A reactor is only as reliable as its fuel supply, which is why this site devotes an entire page to the uranium supply chain. The page includes: the mines, conversion and enrichment plants, fabrication facilities, and spent-fuel stores that feed and follow the reactors. US power reactors consume more uranium than any country on Earth, on the order of 50 million pounds a year, yet most enrichment and conversion capacity is located abroad. Domestic mining is a shadow of its former self, and the US currently imports roughly 95% of the uranium yellowcake that is required to fuel its nuclear reactors. The uranium supply-chain map makes the supply vulnerability starkly visible; it displays a remarkably thin domestic fuel base on which a vast, dependable nuclear reactor fleet relies. The new reactor designs sharpen the problem. Many advanced reactors need high-assay low-enriched uranium HALEU, enriched to higher levels than today’s fuel. Up until now, HALEU uranium has been produced almost entirely abroad. Standing up domestic HALEU and conventional enrichment capacity is now treated as a strategic priority, the less-visible companion to the reactor revival on display. The reactors get the headlines, but the US's uranium supply chain will dictate whether this new nuclear renaissance is successful or fails to be fully realized.