Deep Decarbonization Models Miss the Mark on Advanced Nuclear Energy

WASHINGTON, D.C. In the past three years, nine U.S. states enacted legislation to fully decarbonize the electricity sector by 2050 or sooner.[1] Three of those state laws require 85-100 percent reductions in economy-wide greenhouse gas emissions. At least 29 U.S. electric and gas utilities have pledged to reduce CO2 or GHG emissions by similar amounts.[2] Earlier this year, President Biden proposed a target of 100 percent clean electricity by 2035. Economic models that simulate decarbonization are indispensable to policymakers and utilities considering pathways to zero carbon emissions.[3] Those modelling tools inadequately characterize advanced nuclear technologies, a source of firm low-carbon power pivotal to deep decarbonization.

We reviewed 15 high-profile deep decarbonization modelling studies published since 2016 and assessed the characterization of advanced nuclear capabilities.[4] All such studies model decarbonization of 80 percent or more in the contiguous United States or in specific regions or states within the United States. Ten studies model decarbonization of the electric power sector and five model economy-wide decarbonization. 13 studies employ least-cost capacity expansion models, one study employs a “computable general equilibrium” (CGE) model, and one employs a scenario-based model without optimization.

We conclude that most models fall far short of characterizing the full range of advanced nuclear capabilities. Many advanced nuclear designs are capable of ramping and load-following. Certain designs include shut-off of smaller reactors within larger plants, thermal heat storage with molten salt, on-site battery storage, or hydrogen or other co-product production. Many advanced nuclear designs refuel less frequently and can operate at higher capacity factors than conventional designs. There are numerous advanced nuclear designs, with power ratings from 1.5 MWe to more than 1 GWe, that can be used in a variety of applications including district heating, co-generation, industrial heating, hydrogen production, desalination, and repowering idled plants.[5]

Just four studies model non-zero ramp rates for nuclear power plants and the remainder either assume a fixed power production profile or do not indicate otherwise. The four studies that model non-zero ramp rates assume maximum changes in reactor output of 0.04-0.42 percent of rated power per minute, compared to demonstrated ramp rates of 3-5 percent of rated power in modern commercial reactors in France, Germany, and other locations (Figure 1).[6] Some advanced nuclear designs may have even greater ramping and load-following capabilities than modern commercial reactors.

Seven studies assume average new nuclear capital costs of more than $6,000 / kW and only a single study assumes a capital cost lower than the average anticipated for commercial advanced nuclear designs (Figure 2).[7] Just four studies model modest annual reductions in the capital cost of new nuclear in the range of $17-81 / kW / year, short of expected capital cost reduction rates for advanced nuclear designs.[8]

A single study models the use of nuclear electric energy in industrial heating and two studies model the use of nuclear energy in hydrogen production. The remainder of the studies do not model the use of nuclear energy in district heating, co-generation, industrial heating, hydrogen production, desalination, or repowering. None of the studies model multiple nuclear designs despite a proliferation of designs.[9] Three studies permit power production from existing nuclear facilities but do not allow for the construction of new nuclear powerplants.

Most studies assume operational costs lower than those anticipated for commercial advanced nuclear designs (Figure 3). The studies assume total operational costs (fuel and non-fuel variable O&M) of $9-17 / kWh, compared to an anticipated range of $15-23 / kWh for commercial advanced reactors.[10] Nonetheless, none of the studies model intended reductions in variable or fixed O&M costs.[11] Annual operational costs for conventional and advanced nuclear designs are generally much lower than annual investment-related capital costs.

Figure 1. Nuclear Power Ramping Rates in Deep Decarbonization Modelling

Figure 2. Nuclear Power Capital Costs in Deep Decarbonization Modelling

Figure 3. Nuclear Power Variable O&M Costs in Deep Decarbonization Modelling
Many of the studies we reviewed rely on decarbonization models that directly inform energy policy, planning, and investment decisions. For example, the NREL ReEDS model was the primary analytical tool in the U.S. Department of Energy’s Sunshot, Wind Vision, Hydropower Vision, and GeoVision studies, each of which makes specific policy, regulatory, and investment recommendations related to renewable energy.[12] The E3 “PATHWAYS” model was used recently to support the California Energy Commission’s deep decarbonization study and Xcel Energy’s corporate climate strategy.[13] The EER EnergyPATHWAYS and RIO models were used to support the Massachusetts Office of Energy and Environmental Affairs’ decarbonization roadmap and Portland General Electric’s decarbonization study.[14]

A 2018 literature review concludes that deep decarbonization modeling scenarios that fairly characterize firm low-carbon generation resources—such as advanced nuclear energy—are more likely to yield lower overall decarbonization costs.[15] Our review reveals the exclusion or mischaracterization of advanced nuclear energy in 15 of the most prominent deep decarbonization studies published in recent years. It is incumbent upon the decarbonization research and modeling communities to fairly characterize advanced nuclear energy, and all such resources likely to contribute to decarbonization, given the widespread uses of such models in public policy and planning decisions.