The Modular Atom: Small Reactors and the Coming Nuclear Renaissance

01 The Scale Problem That Broke Nuclear

Nuclear power's central paradox is that its economics worsen with scale. Conventional light-water reactors grew larger over decades — from 600 megawatts in the 1970s to 1,600 megawatts today — on the assumption that bigger plants produce cheaper electricity through economies of scale. Instead, the industry discovered diseconomies of scale: each new plant was a bespoke mega-project, built on-site over a decade or more, with costs that escalated unpredictably. The Vogtle plant in Georgia, the first new US nuclear construction in thirty years, came in years late and billions over budget.

Small modular reactors — defined as reactors with a rated electrical power of less than 300 megawatts — start from the opposite assumption. If bigger is worse, then smaller and standardized might be better. The idea is to build reactors in factories as modular units, ship them to sites, and assemble them like appliances. The factory captures economies of scale through production volume rather than unit size. The site becomes an assembly point rather than a construction site.

02 What Makes a Reactor Modular

Modularity in the SMR context means more than just small size. A true SMR is designed for factory fabrication, transport by truck or rail, and on-site installation with minimal field construction. The reactor vessel, internals, and primary cooling system are built as a single shippable module. Some designs — the so-called micro-reactors — are small enough to be transported on a single truck and installed in days rather than years.

The modular approach also enables incremental capacity addition. Instead of building one 1,200-megawatt plant that takes a decade, a utility could install four 300-megawatt modules over four years, with the first module generating revenue while subsequent modules are still being installed. This changes the financial profile dramatically: the cost of capital is incurred over a shorter period, and revenue begins flowing years earlier. For an industry where financing costs can exceed the construction costs themselves, this is not a marginal improvement — it is a structural change.

03 Passive Safety: No Operators Required

Traditional reactors rely on active safety systems — electrically powered pumps, valves, and control systems that require operator action and external power to shut down the reactor in an emergency. The failure of these systems at Fukushima demonstrated the vulnerability of this approach. SMR designs overwhelmingly incorporate passive safety: gravity, natural circulation, and convection cool the reactor without any external power or operator intervention.

In a passive design, if the reactor loses power, control rods drop by gravity, and emergency cooling water flows by natural convection. The reactor shuts itself down and cools itself for days without any human action. This is not a new concept — it was validated in Generation III+ reactor designs like the AP1000 — but SMRs make it the default rather than an upgrade. The smaller size makes passive cooling more tractable because the decay heat that must be removed is proportionally smaller.

04 The Economic Case Under Scrutiny

The SMR economic argument rests on a proposition that has not yet been tested at scale: that factory production of nuclear reactors will achieve the cost reductions seen in other manufactured goods. Aircraft, automobiles, and electronics all became cheaper through standardized production. But nuclear reactors are not consumer goods — they are safety-critical systems subject to regulatory oversight, and the history of nuclear construction suggests that standardization has been difficult to achieve even within a single country.

Critics point to the NuScale project, which was the first SMR design to receive US Nuclear Regulatory Commission certification. The planned Utah Associated Municipal Power Systems project was cancelled in late 2023 after costs rose from an estimated $3 billion to over $9 billion for a 462-megawatt (six-module) plant. This was a sobering signal: if the most advanced SMR project in the United States could not control costs, the economic premise requires reassessment. Proponents argue that the first-of-a-kind costs are expected and that subsequent projects will benefit from learning effects — but the nuclear industry has a long history of learning effects that never materialized.

05 Advanced Reactor Technologies

Not all SMRs are scaled-down light-water reactors. A significant subset uses advanced reactor technologies that differ fundamentally from the pressurized water reactors that dominate the current fleet. Sodium-cooled fast reactors, high-temperature gas reactors, molten salt reactors, and lead-cooled reactors are all in development at small and medium scales. These designs operate at different temperatures, use different coolants, and in some cases can consume nuclear waste as fuel.

TerraPower's Natrium reactor, backed by Bill Gates, combines a sodium-cooled fast reactor with a molten salt energy storage system that allows the plant to ramp output up and down — a capability that traditional nuclear lacks and that could make SMRs complementary to intermittent renewables. Oklo's Aurora micro-reactor is designed for autonomous operation with years between refueling. These are not incremental improvements on existing technology; they are different reactor architectures that have been studied for decades but never commercialized. Whether SMRs are the vehicle that finally brings them to market remains to be seen.

06 The Regulatory Bottleneck

Nuclear regulation was designed for large, site-specific reactors, and the licensing process reflects that assumption. A traditional reactor licensee submits a combined construction and operating license application that can run to thousands of pages and take years to review. SMRs challenge this framework because the safety case, design parameters, and operational profile are fundamentally different. A factory-built module certified once should not require each installation to repeat the full certification process — but the regulatory infrastructure for design certification without site-specific review is still being built.

The US Nuclear Regulatory Commission certified the NuScale design in early 2023, the first SMR certification — a process that took over five years and cost NuScale approximately $500 million. The NRC has since been working on a framework for advanced reactor licensing, but the pace of regulatory reform has not kept up with the pace of design innovation. In the absence of a streamlined pathway, each new SMR design faces a multi-year, multi-hundred-million-dollar licensing effort before a single module can be installed.

07 The China Factor

While Western SMR programs navigate regulatory and economic hurdles, China is building. The Linglong One (ACP100) SMR on Hainan Island became the first land-based SMR to begin construction, and China's state nuclear corporation has announced plans for dozens of additional units. Russia's Akademik Lomonosov, a floating nuclear power plant with two 35-megawatt reactors, has been operational since 2020, supplying power to remote Arctic communities. These projects demonstrate that SMRs can be built — the question is whether they can be built economically in market economies where cost overruns have consequences.

The geopolitical dimension is significant. If SMRs become a viable export product, the countries that master their production first will have a strategic advantage in the global clean energy market. The US, France, South Korea, Russia, and China are all positioning themselves as SMR exporters, and the competition for international orders is already underway. Nuclear energy is not just a climate technology — it is a domain of industrial and geopolitical competition, and the SMR race is its latest chapter.