Technology companies like Amazon, Google, and Microsoft are heavily investing in Small Modular Reactors (SMRs) as part of their efforts to meet growing energy demands while transitioning to carbon-free power sources. However, despite these significant investments, recent developments highlight that SMR technology remains largely unproven and faces significant challenges. These challenges include cost overruns, regulatory hurdles, and questions around the commercial viability of SMRs, casting doubt on their ability to deliver the promised benefits on time and within budget.
Amazon has taken a notable step by signing agreements to support the development of SMRs, including a partnership with Energy Northwest to develop up to four reactors in Washington. These reactors are expected to provide up to 960 MW of power to Amazon’s data centers, aligning with the company’s net-zero carbon goals for 2040. While Amazon has positioned SMRs as a key part of its clean energy strategy, the technology remains untested on a commercial scale, particularly in the Western world. Amazon’s investment follows similar moves by Google and Microsoft, but these initiatives come at a time when SMR technology is facing serious obstacles.
Google recently signed a landmark deal with Kairos Power to develop multiple SMRs, aiming to bring the first reactor online by 2030. This would add up to 500 MW of carbon-free energy to the U.S. grid, supporting Google’s ambitious 24/7 clean energy goals. While Google is among the first to sign a corporate agreement to purchase nuclear energy from SMRs, the global track record for SMRs is not encouraging. France’s nuclear giant EdF recently scrapped its plans to develop an internal SMR design, citing engineering challenges and cost overruns. EdF had been working on the "Nuward" SMR for four years, but the design was abandoned due to escalating costs and concerns from potential customers over high energy prices. Similar concerns have plagued other SMR projects, including in the United States.
Microsoft’s recent commitment to nuclear energy through a power purchase agreement with Constellation Energy to restart the Three Mile Island Unit 1 reactor reflects its reliance on nuclear power to meet sustainability goals. The plant will provide 835 MW of carbon-free energy to power Microsoft’s data centers, but like Amazon and Google, Microsoft’s nuclear ambitions rest on the success of SMR technologies that have yet to prove commercially viable. The failure of NuScale Power’s SMR project in the U.S., which faced massive cost overruns and was abandoned after customer pushback, illustrates the significant hurdles facing these next-generation reactors.
The broader nuclear industry is facing similar issues. EdF’s large-scale nuclear projects have been marked by extreme delays and cost overruns, most notably at the Hinkley C project in the UK and the Flamanville project in France. Hinkley C, originally slated for completion by 2017, is now delayed until 2031, with costs skyrocketing from £9 billion to an estimated £48 billion. Flamanville, announced in 2004 with a budget of €3 billion, is still not operational, with its costs now exceeding €13 billion. These setbacks raise concerns about whether SMRs, with their smaller scale but similar technological complexities, can avoid the same pitfalls.
SMRs have been promoted as a solution to provide flexible, round-the-clock carbon-free energy, particularly for energy-intensive industries like data centers. However, despite their potential, the technology faces mounting challenges. A report by The Breakthrough Institute outlined five key issues facing the industry: high interest rates and commodity prices, constrained supply chains, a regulatory regime that stifles innovation, high system costs, and difficulties in securing nuclear fuel. These challenges, coupled with recent project failures, suggest that the commercialization of advanced nuclear technologies like SMRs is far from assured.
Even as tech giants invest in SMRs, public skepticism remains. In Australia, opposition leader Peter Dutton has faced criticism for promoting SMRs as part of his energy policy. Reports from the Australian Energy Market Operator and the CSIRO have estimated that the cost of nuclear energy, including SMRs, would be more than double that of wind, solar, and storage. These assessments suggest that SMRs may not be the silver bullet they are often portrayed to be. Public polling in Australia has shown that while nuclear energy has support, renewables like wind and solar are seen as more desirable and better for the environment.
The challenges facing SMRs are compounded by the fact that no SMR has yet been licensed or built in the Western world. NuScale, once considered a frontrunner, saw its plans fall apart after cost estimates soared from $58/MWh to $89/MWh, leading customers to back out. Similar issues have affected other advanced nuclear projects, with X-energy also delaying its commercialization timeline due to a lack of fuel. These setbacks have led to growing doubts about whether SMRs can deliver on their promises of affordable, scalable, and safe nuclear energy.