Cost Escalation and Delays for Small Modular Reactors Suggest Caution about Nuclear Power Renaissance

David Kemp and Peter Van Doren

Is nuclear power back? Recent private and public investment in new nuclear technologies has created expectations that the nuclear industry will finally be able to surmount its construction cost problem. But news about costs at some of the most promising nuclear projects suggests caution.

The Washington Post reported that two leading companies, NuScale Power and the Bill Gates founded TerraPower are facing difficulties. Both companies are developing Small Modular Reactors (SMRs), which, as the name implies, use smaller reactor designs and off‐​site modular construction to avoid the high construction costs of conventional reactors. NuScale’s first power plant, the Carbon Free Power Project (CFPP) at Idaho National Labs, recently increased its initial estimated electricity cost from 5.8 cents per kWh to 8.9 cents per kWh and TerraPower’s first plant, in Wyoming, pushed back its planned opening by two years because of difficulties sourcing fuel from Russia after the war with Ukraine.

These projects’ difficulties resemble the problems seen by other recent US nuclear projects at V.C. Summer in South Carolina and Vogtle in Georgia. Both experienced ballooning costs, and V.C. Summer was canceled in 2017 while total Vogtle construction costs are projected to be nearly $34 billion and the start of operation continues to be delayed.

We wrote a paper last year that explored the economics of nuclear power and included a specific focus on SMRs and NuScale. Overall, our findings were pessimistic. Based on recent construction costs in the US and Europe and comparing the lifetime costs of a new nuclear power plant to a new natural gas plant, we found that the benefit of no carbon emissions from nuclear power was not large enough to offset its high construction costs. Only in an overly optimistic scenario, where nuclear construction costs were 55 percent lower and there was a high projected natural gas price, did the costs of building a nuclear power plant become worthwhile if we assumed a reasonably sized carbon tax. We concluded:

Unless nuclear construction costs can be substantially reduced, the carbon tax levels recommended by many analysts are too low to induce private investors to build nuclear rather than natural gas generation units. Even at significantly reduced nuclear construction costs, nuclear would still require a high natural gas price to compete with a new natural gas power plant. The history of nuclear construction in the United States does not support hopes that nuclear’s high cost will be reduced.

The economic promise of SMRs, however, is that their smaller, modular design and factory construction can bring down construction costs through standardization and replication that will allow firms to reduce costs as more reactors are built. Our analysis found the prospects for SMR cost reductions are largely reliant on a high rate of manufacturing (e.g., 5–10 reactors per year) and the factories and supply chains necessary for this mass production are not close to being in place.

The increase of the NuScale target price from 5.8 to 8.9 cents per kWh doesn’t actually change this calculus. It just means that the NuScale SMR has a higher starting point from which its costs need to be reduced for the benefits of mass production to pay off. But it does reaffirm our conclusion that high nuclear construction costs currently outweigh nuclear’s low‐​carbon benefit.

In a press release and talking points, NuScale says the new cost was “primarily influenced by external impacts, not by the project’s development.” They note both inflationary increases for the costs of commodities, including steel, copper wiring, and electrical equipment, and increased interest rates. This has boosted their projected construction costs, including financing costs, 75 percent from $5.3 billion to $9.3 billion.

Based on our estimated costs per kWh for a new natural gas plant (adjusted from 2020 dollars to 2022 dollars using the GDP deflator from the Bureau of Economic Analysis), we can use the same calculations outlined in our paper to assess the carbon tax necessary for the NuScale CFPP to be cost competitive when carbon emissions are considered. NuScale’s updated cost of 8.9 cents per kWh compared to our estimated costs for natural gas of 4.2–6.2 cents per kWh (depending on the projected natural gas price) suggests that there would need to be a carbon tax of around $54–92 per metric ton of CO₂ to make up the difference in costs. However, NuScale’s target price includes a 3 cent per kWh production subsidy from the Inflation Reduction Act (along with other subsidies, which for the sake of simplicity we ignore here). Clean energy subsidies are an implicit carbon price and when the additional 3 cents are included (bringing the cost of the CFPP up to 11.9 cents per kWh), the total carbon tax needed for cost parity is $113–150 dollars.

This is outside the range of carbon taxes typically recommended by experts. For example, the US Federal Government’s Interagency Working Group’s most recent estimate (adjusted to 2022 dollars) is a carbon tax of $19–93 per metric ton of CO₂ in 2025. Thus, at the most recent target price the CFPP is not worth the combined cost of electricity to ratepayers and cost of the production subsidy to taxpayers.

NuScale might be able to reduce costs as they build more reactors. But considering that the CFPP has yet to break ground and is still in the process of getting construction and operating approval from the Nuclear Regulatory Commission, the recent cost increase should temper some of the optimistic predictions that NuScale has found a way to avoid nuclear power’s historical cost issues. As we said in our paper,

Of course, because of the possibility of cost surprises, the success of the CFPP cannot be assessed until much later in its development and construction. With over 20 years of development time at this point, the NuScale SMR still faces some of the most difficult obstacles for nuclear projects: construction management and regulatory instability during construction. As the most advanced SMR design, and with the issues at Vogtle and V.C. Summer, hopes for the future of the US nuclear industry rest largely on NuScale’s ability to avoid these historical drivers of cost escalation.

inflation
commodities
interest rates
inflationary