An obituary for small modular reactors

CAREM SMR in Argentina

The latest cost estimate for the CAREM small modular reactor under construction in Argentina is 22 times greater than the initial cost estimate.

Small modular reactors are being heavily promoted but the economics don't stack up.

The costs of SMRs, in part because of the size and in part because of the security that's associated with any nuclear plant, are prohibitive.

The nuclear industry is heavily promoting the idea of building small modular reactors (SMRs), with near-zero prospects for new large power reactors in many countries. These reactors would have a capacity of under 300 megawatts (MW), whereas large reactors typically have a capacity of 1,000 MW.

Construction at reactor sites would be replaced with standardised factory production of reactor components then installation at the reactor site, thereby driving down costs and improving quality control.

The emphasis in this article is on the questionable economics of SMRs, but a couple of striking features of the SMR universe should be mentioned (for details see the latest issue of Nuclear Monitor).

Fossil fuels and militarism

First, the enthusiasm for SMRs has little to do with climate-friendly environmentalism. About half of the SMRs under construction (Russia's floating power plant, Russia's RITM-200 icebreaker ships, and China's ACPR50S demonstration reactor) are designed to facilitate access to fossil fuel resources in the Arctic, the South China Sea and elsewhere.

Another example comes from Canada, where one application of SMRs under consideration is providing power and heat for the extraction of hydrocarbons from oil sands.

A second striking feature of the SMR universe is that it is deeply interconnected with militarism:

  • Argentina's experience and expertise with small reactors derives from its historic weapons program, and its interest in SMRs is interconnected with its interest in small reactors for naval propulsion.
  • China's interest in SMRs extends beyond fossil fuel mining and includes powering the construction and operation of artificial islands in its attempt to secure claim to a vast area of the South China Sea.
  • Saudi Arabia's interest in SMRs is likely connected to its interest in developing nuclear weapons or a latent weapons capability.
  • A subsidiary of Holtec International has actively sought a military role, inviting the US National Nuclear Security Administration to consider the feasibility of using a proposed SMR to produce tritium, used to boost the explosive yield of nuclear weapons.
  • Proposals are under consideration in the US to build SMRs at military bases and perhaps even to use them to power forward operating bases.
  • In the UK, Rolls-Royce is promoting SMRs on the grounds that "a civil nuclear UK SMR programme would relieve the Ministry of Defence of the burden of developing and retaining skills and capability".

Independent economic assessments

SMRs will almost certainly be more expensive than large reactors (more precisely, construction costs will be lower but the electricity produced by SMRs will be more expensive).

They will inevitably suffer diseconomies of scale: a 250 MW SMR will generate 25 percent as much power as a 1,000 MW reactor, but it will require more than 25 percent of the material inputs and staffing, and a number of other costs including waste management and decommissioning will be proportionally higher.

It's highly unlikely that potential savings arising from standardised factory production will make up for those diseconomies of scale.

William Von Hoene, senior vice president at Exelon, has expressed scepticism about SMRs: "Right now, the costs on the SMRs, in part because of the size and in part because of the security that's associated with any nuclear plant, are prohibitive," he said last year.

"It's possible that that would evolve over time, and we're involved in looking at that technology. Right now they're prohibitively expensive."

Every independent economic assessment finds that electricity from SMRs will be more expensive than that from large reactors.

A study by WSP / Parsons Brinckerhoff, commissioned by the 2015/16 South Australian Nuclear Fuel Cycle Royal Commission, estimated costs of A$180‒184/MWh (US$127‒130) for large pressurised water reactors and boiling water reactors, compared to A$198‒225 (US$140‒159) for SMRs.

A 2015 report by the International Energy Agency and the OECD Nuclear Energy Agency predicts that electricity costs from SMRs will typically be 50−100 percent higher than for current large reactors, although it holds out some hope that large volume factory production of SMRs could help reduce costs.

A report by the consultancy firm Atkins for the UK Department for Business, Energy and Industrial Strategy found that electricity from the first SMR in the UK would be 30 percent more expensive than power from large reactors, because of diseconomies of scale and the costs of deploying first-of-a-kind technology.

An article by four current and former researchers from Carnegie Mellon University's Department of Engineering and Public Policy, published in 2018 in the Proceedings of the National Academy of Science, considered options for the development of an SMR market in the US.

They concluded that it would not be viable unless the industry received "several hundred billion dollars of direct and indirect subsidies" over the next several decades.

No market

SMR enthusiasts envisage a large SMR market emerging in the coming years. A frequently cited 2014 report by the UK National Nuclear Laboratory estimates 65‒85 gigawatts (GW) of installed SMR capacity by 2035, valued at £250‒400 billion.

But in truth there is no market for SMRs.

Thomas Overton, associate editor of POWER magazine, wrote in 2014: "At the graveyard wherein resides the "nuclear renaissance" of the 2000s, a new occupant appears to be moving in: the small modular reactor (SMR)...

"Over the past year, the SMR industry has been bumping up against an uncomfortable and not-entirely-unpredictable problem: It appears that no one actually wants to buy one."

Let's briefly return to the National Nuclear Laboratory's estimate of 65‒85 GW of installed SMR capacity by 2035. It is implausible and stands in contrast to the OECD Nuclear Energy Agency's estimate of <1 GW to 21 GW of SMR capacity by 2035. But even if the 65‒85 GW figure proved to be accurate, it would pale in comparison to renewable energy sources.

As of the of end of 2017, global renewable energy capacity was 2,195 GW including 178 GW of new capacity added in 2017. On current trends, even in the wildest dreams of SMR enthusiasts, SMR capacity would be roughly 50 times less than renewable capacity by 2035.

SMRs under construction

SMR projects won't be immune from the major cost overruns that have crippled large reactor projects (such as the AP1000 projects in the US that bankrupted Westinghouse). Indeed cost overruns have already become the norm for SMR projects.

Estimated construction costs for Russia's floating nuclear power plant (with two 35-MW ice-breaker-type reactors) have increased more than four-fold and now equate to over US$10 billion / GW (US$740 million / 70 MW).

A 2016 OECD Nuclear Energy Agency report said that electricity produced by the Russian floating plant is expected to cost about US$200 per megawatt-hour (MWh), with the high cost due to large staffing requirements, high fuel costs, and resources required to maintain the barge and coastal infrastructure.

The CAREM (Central Argentina de Elementos Modulares) SMR under construction in Argentina illustrates the gap between SMR rhetoric and reality. Cost estimates have ballooned. In 2004, when the CAREM reactor was in the planning stage, Argentina's Bariloche Atomic Center estimated an overnight cost of US$1 billion / GW for an integrated 300 MW plant.

When construction began in 2014, the estimated cost of the CAREM reactor was US$17.8 billion / GW (US$446 million for a 25-MW reactor). By April 2017, the cost estimate had increased to US$21.9 billion / GW (US$700 million with the capacity uprated from 25 MW to 32 MW).

The CAREM project is years behind schedule and costs will likely increase further. In 2014, first fuel loading was expected in 2017 but completion is now anticipated in November 2021.

Little credible information is available on the cost of China's demonstration high-temperature gas-cooled reactor (HTGR). If the 210 MW demonstration reactor is completed and successfully operated, China reportedly plans to upscale the design to 655 MW.

According to the World Nuclear Association, China's Institute of Nuclear and New Energy Technology at Tsinghua University expects the cost of a 655 MW HTGR to be 15-20 percent more than the cost of a conventional 600 MW PWR.

A 2016 report said that the estimated construction cost of China's demonstration HTGR is about twice the initial cost estimates, with increases due to higher material and component costs, increases in labour costs, and increased costs associated with project delays. The World Nuclear Association states that the cost of the demonstration HTGR is US$6,000/kW.

NuScale Power's creative accounting

Cost estimates for planned SMRs are implausible. US company NuScale Power is targeting a cost of just US$65/MWh for its first plant. But a study by WSP / Parsons Brinckerhoff, commissioned by the South Australian Nuclear Fuel Cycle Royal Commission, estimated a cost of US$159/MWh based on the US NuScale SMR design. That's 2.4 times higher than NuScale's estimate.

A 2018 Lazard report estimates costs of US$112‒189/MWh for electricity from large nuclear plants. NuScale's claim that its electricity will be 2‒3 times cheaper than large nuclear is implausible. And even if NuScale achieved costs of US$65/MWh, that would still be well above Lazard's figures for wind power (US$29‒56) and utility-scale solar (US$36‒46).

Likewise, NuScale's construction cost estimate of US$4.2 billion / GW is implausible. The latest estimate for the AP1000 reactors under construction in Georgia is US$12.7 billion / GW.

NuScale wants us to believe that it will build SMRs at less than one-third of that cost, even though every independent assessment concludes that SMRs will be more expensive to build (per GW) than large reactors.

No-one wants to pay for SMRS

No company, utility, consortium or national government is seriously considering building the massive supply chain that is at the very essence of the concept of SMRs ‒ mass, modular factory construction. Yet without that supply chain, SMRs will be expensive curiosities.

In early 2019, Kevin Anderson, North American Project Director for Nuclear Energy Insider, said that there "is unprecedented growth in companies proposing design alternatives for the future of nuclear, but precious little progress in terms of market-ready solutions."

Anderson argued that it is time to convince investors that the SMR sector is ready for scale-up financing but that it will not be easy: "Even for those sympathetic, the collapse of projects such as V.C Summer does little to convince financiers that this sector is mature and competent enough to deliver investable projects on time and at cost."

A 2018 US Department of Energy report states that to make a "meaningful" impact, about US$10 billion of government subsidies would be needed to deploy 6 GW of SMR capacity by 2035. But there's no indication or likelihood that the US government will subsidise the industry to that extent.

To date, the US government has offered US$452 million to support private-sector SMR projects, of which US$111 million was wasted on the mPower project that was abandoned in 2017.

The collapse of the mPower project was one of a growing number of setbacks for the industry in the US. Transatomic Power gave up on its molten salt reactor R&D last year.

Westinghouse sharply reduced its investment in SMRs after failing to secure US government funding. MidAmerican Energy gave up on its plans for SMRs in Iowa after failing to secure legislation that would force rate-payers to part-pay construction costs. 

The MidAmerican story has a happy ending: the company has invested over US$10 billion in renewables in Iowa and is now working towards its vision "to generate renewable energy equal to 100 percent of its customers' usage on an annual basis."

Canada and the UK

Canadian Nuclear Laboratories has set the goal of siting a new demonstration SMR at its Chalk River site by 2026. But serious discussions about paying for a demonstration SMR ‒ let alone a fleet of SMRs ‒ have not yet begun.

The Canadian SMR Roadmap website simply states: "Appropriate risk sharing among governments, power utilities and industry will be necessary for SMR demonstration and deployment in Canada."

Companies seeking to pursue SMR projects in the UK are seeking several billion pounds from the government to build demonstration plants. But nothing like that amount of money has been made available.

In 2018, the UK government agreed to provide £56 million towards the development and licensing of advanced modular reactor designs and £32 million towards advanced manufacturing research.

An industry insider told the Guardian in 2017: "It's a pretty half-hearted, incredibly British, not-quite-good-enough approach. Another industry source questioned the credibility of SMR developers: "Almost none of them have got more than a back of a fag packet design drawn with a felt tip."

State-run SMR programs

State-run SMR programs ‒ such as those in Argentina, China, Russia, and South Korea ‒ might have a better chance of steady, significant funding, but to date the investments in SMRs have been minuscule compared to investments in other energy programs.

And again, wherever you look there's nothing to justify the high hopes (and hype) of SMR enthusiasts. South Korea, for example, won't build any of its domestically-designed SMART SMRs in South Korea ("this is not practical or economic" according to the World Nuclear Association).

South Korea's plan to export SMART technology to Saudi Arabia is problematic and may in any case be in trouble.

China and Argentina hope to develop a large export market for their high-temperature gas-cooled reactors and small pressurised water reactors, respectively, but so far all they can point to are partially-built demonstration reactors that have been subject to significant cost overruns and delays.

All of the above can be read as an obituary for SMRs. The likelihood that they will establish anything more than a small, niche market is vanishingly small.

This Author

Dr Jim Green is the lead author of a Nuclear Monitor report on small modular reactors, and national nuclear campaigner with Friends of the Earth Australia.

Help us keep The Ecologist working for the planet

The Ecologist website is a free service, published by The Resurgence Trust, a UK-based educational charity. We work hard - with a small budget and tiny editorial team - to bring you the wide-ranging, independent journalism we know you value and enjoy, but we need your help. Please make a donation to support The Ecologist platform. Thank you!

Donate to us here