Small modular reactor

Small modular reactors were born twice. First they were compact reactors hidden inside naval hulls. Later they reappeared as civilian power plants for grids too small, remote, or financially wary to swallow a conventional gigawatt station. That second birth became undeniable at Pevek in Russia, where the floating Akademik Lomonosov sent electricity ashore in 2019 and entered commercial operation in 2020. What changed was not fission physics. What changed was the niche: places that wanted nuclear heat and power without betting everything on one enormous build.

The adjacent possible sat on top of `nuclear-reactor` science and `nuclear-powered-surface-ship` engineering. The reactor had already proved that controlled fission could run for years. Naval programs and the `nuclear-submarine` had already proved that compact reactors could survive isolation, motion, tight footprints, and long refuelling intervals. What the civilian sector lacked was a way to translate that compactness into bankable infrastructure. Traditional `nuclear-power-plant` projects had been optimized for scale. Small modular reactors emerged when utilities and governments started caring less about the cheapest marginal megawatt on paper and more about staged capital, remote deployment, passive safety, and factory fabrication.

That is `knowledge-accumulation`. Decades of reactor physics, fuel management, containment engineering, digital controls, and marine nuclear practice had to accumulate before a sub-300-megawatt unit looked like an energy product rather than a shrunken experiment. The IAEA's definition captures the threshold: an SMR is a reactor of up to 300 megawatts electric, small enough to be built as modules and assembled in repeatable ways. Size alone was never the point. The point was moving nuclear construction away from one-off monuments and toward something closer to manufactured infrastructure.

Why did the idea return now rather than in 1970 or 1990? Because the twentieth century had selected hard for bigness. Utilities, regulators, financiers, and supply chains had all been trained by `path-dependence` to think in very large reactors. Bigger plants promised lower unit costs, and the whole industry arranged itself around that promise. Then the weaknesses of that model became harder to ignore: long build times, giant financing risk, weak fit for remote grids, and poor economics for customers that did not need a full-size station. Small modular reactors became attractive when the problem shifted from maximizing reactor size to minimizing project fragility.

That new demand performed `niche-construction`. Remote Arctic towns, mining sites, district-heating networks, military bases, desalination schemes, and smaller national grids created spaces where a giant plant was a bad fit even if nuclear power itself made sense. Russia reached the market first because Pevek needed both electricity and heat in a place where fuel logistics were punishing and a floating plant solved siting problems. The United States attacked the same niche through licensing. NuScale Power won U.S. standard design approval in 2020 and design certification in 2023, turning SMRs from a recurring policy speech into a licensable reactor class. Canada pushed the concept toward utility deployment when Ontario secured a construction licence in 2025 for General Electric's BWRX-300 at Darlington.

That spread across Russia, the United States, and Canada shows `convergent-evolution`. Different countries with different politics and reactor lineages kept arriving at the same answer: if large nuclear plants are too financially brittle, shrink the unit, standardize the design, and try to recover economics through repetition instead of sheer size. The forms vary. Russia's first move was floating heat-and-power service. NuScale's pitch centered on modular light-water units for civilian grids. Ontario's Darlington project turned the idea into a utility-scale construction program. The pressures, though, were recognizably the same.

Commercial reality has been harsher than the pitch deck. Small does not automatically mean cheap. NuScale's Carbon Free Power Project in Idaho was terminated in 2023 when expected subscription and expected cost no longer matched. That failure mattered because it exposed the central tension in the whole field: modularity is only an economic advantage after factories, regulators, and buyers all repeat the same design often enough for learning curves to bite. Until then, an SMR can be small in hardware while still behaving like a custom nuclear megaproject in finance.

Even so, the invention has changed the direction of nuclear strategy. It has given policymakers, utilities, and industrial users a new way to ask the question. Instead of debating whether every grid needs another giant station, they can ask where compact nuclear heat or power fits better than gas, diesel, or long transmission lines. That is a narrower claim than the grand promises surrounding atomic energy in the 1950s, but it is also a more durable one.

No company has yet achieved the kind of manufacturing lock-in that early advocates imagined, but the race is plainly commercial now. NuScale Power showed one regulatory path. General Electric showed another route through an existing utility and supply chain. If the SMR era becomes real at scale, it will not be because someone invented a tiny reactor from scratch. It will be because the nuclear industry finally learned how to sell repetition instead of singularity.