The nuclear resurgence in the United States and the West more broadly has focused heavily on innovation in small modular reactors. The advanced reactor field is increasingly crowded with organizations announcing ambitious timelines and aggressive commercialization plans, many led by capable engineers pursuing meaningful innovation.
Innovation, however, should not be mistaken for commercial readiness. Nuclear engineering in the U.S. has never been limited by a lack of creative reactor concepts — its greatest challenges have come in moving from a promising design to a licensed, commercially operating power plant. China, on the other hand, is building fleets of larger reactors and is likely to overtake the U.S. by 2030.
While the impetus to develop AI capabilities and the ensuing electricity demand have provided enough incentive for large U.S. companies to back some innovative SMR and microreactor designs, the transition requires a demanding sequence of engineering, experimentation, materials qualification, and supply chain maturation that cannot be compressed simply because demand has grown. That distinction between innovation and readiness is the lens for evaluating the progress underway across today’s advanced nuclear industry.
The diversity of advanced reactor technologies under development today is unlike anything seen since the earliest decades of civilian nuclear power. Developers are pursuing sodium-cooled fast reactors, molten salt reactors, high-temperature gas reactors, lead-cooled fast reactors, integral pressurized water reactors, heat-pipe microreactors, and numerous hybrid concepts, each targeting some combination of safety, economics, flexibility, or manufacturing efficiency.
That diversity is a strength: A small, transportable reactor for a remote community faces different engineering challenges than a 300-megawatt electrical plant serving a hyperscale data center, and no single concept will satisfy every application.
Government programs, including the Energy Department’s Advanced Reactor Demonstration Program, the National Reactor Innovation Center, the War Department, and international counterparts, have accelerated this work by funding experimental facilities and demonstration projects. Several developers, including those pursuing designs under DOE’s Advanced Reactor Demonstration Program, have achieved baseline or “criticality” demonstrations under streamlined DOE authorization.
While these are genuine engineering achievements that confirm reactor physics, they do not yet imply a pathway for deploying reactors quickly and economically. For nearly every advanced reactor developer, the journey from experimental success to commercial deployment remains substantial.
The history of commercial nuclear power has demonstrated this lesson repeatedly. Vogtle Units 3 and 4, the first new AP1000 reactors built in the U.S., used a modular design that the Nuclear Regulatory Commission had already certified, yet ran behind schedule and above the original budget. NuScale’s design, similarly, became the first SMR to receive NRC design certification, a major regulatory and technical achievement, yet its flagship Carbon Free Power Project was canceled in 2023 when a customer could no longer be secured.
In both cases, the technology worked. Execution became unrealistic when confronted with manufacturing readiness and commercial arrangements. First-of-a-kind reactors absorb design, construction, and licensing costs and will not be the most accurate predictors of the cost or time required to build a fleet of reactors. Building that fleet requires a stable supply chain for the new materials these reactors require, as well as a commercial supply of the advanced fuels for which they are designed. Long-term competitiveness depends less on the economics of the first unit than on the ability to build many nearly identical units.
In developing new reactors, history shows a recurring pattern. Early commercial nuclear programs generally underestimated the effort required after conceptual design was complete, assuming licensing, manufacturing, and construction could proceed in parallel without surprises. Even a successfully prototyped reactor does not imply a commercially deployable reactor.
TerraPower and X-energy are nearing construction with offtake agreements and capital support from large corporations. Many other companies have excellent reactor concepts but no binding offtake agreements or financial arrangements. Far fewer have assembled the engineering, manufacturing, and licensing capabilities needed to deliver commercial reactors at scale. As the industry matures, execution is becoming as decisive a differentiator as the underlying technology.
France’s nuclear buildout in the 1970s and 1980s shows how standardizing around a few pressurized-water reactor designs can reduce construction costs and complexity. SMRs are expected to offer similar benefits through factory-built modular components and less on-site construction.
This, however, does not apply to the first few SMRs. Early SMR costs will depend heavily on design complexity and constructability. When complex designs encounter engineering challenges and nonexistent supply chains, costs can rise sharply. According to the Office of Scientific and Technical Information, SMRs were still projected to require more on-site labor per megawatt than large reactors.
Advanced computational tools and analytical methods can complement the testing process but cannot replace it. They can simulate reactor behavior under many hypothetical conditions, but their predictions still require validation through experiments.
Kairos Power, which is developing a fluoride salt-cooled high-temperature reactor, has spent the past few years testing non-nuclear units rather than moving directly to a commercial-scale plant. Many innovative reactor designs will likely face significant challenges in building new supply chains and specialized industrial knowledge.
Kadmos Energy, founded by a group of nuclear scientists and engineers from Idaho National Laboratory, is developing an SMR that relies on components and materials supported by existing supply chains. According to Dr. Youssef Ballout, the founding CEO of Kadmos Energy, the company’s reactor design has an identified list of certified suppliers to support construction of its first-of-a-kind SMR.
The supply chain needed to support the manufacture of SMR fleets cannot be rebuilt overnight and must begin developing long before a design reaches final licensing. Fuel supply faces an equivalent challenge across enrichment, conversion, fabrication, transportation, and eventual disposal, all of which must function before customers can order a fleet of SMRs or microreactors.
Vogtle’s experience is a useful case study: Module fabrication problems and on-site rework, rather than the reactor design itself, accounted for much of the project’s schedule slippage, while later units benefited measurably from lessons learned on the earlier ones.
AMERICAN NUCLEAR POWER TAKES AN IMPORTANT STEP
The next few years will determine whether the current wave of advanced nuclear investment produces a sustainable industry or another generation of promising prototypes. Innovation remains essential, but innovation alone has never built a reactor fleet. The organizations that shape the future of nuclear energy will not necessarily be those with the most innovative reactor designs. They will be the ones that transform innovation into repeatable deployment.
In the end, the defining question for the SMR industry is not whether the technology can work. It is whether the industry can build, license, manufacture, finance, and operate these reactors at scale.
Rashmi Singh has been chief lending officer of a San Francisco-based commercial bank, with a certificate in clean energy systems from the Massachusetts Institute of Technology and a master’s in energy law from Texas A&M University.