CS37: Nuclear Power and Firm Capacity in Energy Transition

Under carbon neutrality targets, power systems with high penetration of variable renewable energy (VRE) face simultaneous economic and reliability constraints. This study develops a multi-regional long-term power system planning model covering 31 regions and eight planning periods, and incorporates six representative days with hourly operational resolution to quantify system costs and structural changes during the carbon-neutral transition. Based on a systematic cost decomposition framework, we analyze marginal carbon abatement costs and their evolution, characterize regional heterogeneity in generation costs, and estimate system cost increments under different levels of VRE utilization. The results show that excessively stringent curtailment constraints or reliability requirements significantly increase both total system costs and levelized electricity costs, and that a clear trade-off exists between renewable energy utilization and system reliability. Our findings provide insights into the cost-effective design of carbon-neutral power systems and underscore the need to balance renewable deployment with system reliability and economic feasibility.

Overview

How do import tariffs on nuclear construction inputs and fuel-cycle services affect the economic viability of new nuclear power projects, and how do these effects compare to changes in the cost of capital in capital-intensive energy systems? 


Methodology

A discounted cash flow model is applied to a portfolio of representative reactor designs, including both small modular and large light-water reactors. Results are aggregated into a technology-neutral average reactor for greenfield deployment, with coal-to-nuclear redevelopment used for comparison. Tariffs are modelled as cost multipliers applied to import-exposed direct construction costs and fuel costs within operating expenditures. A structured marginal analysis is conducted to compare the incremental effects of tariff changes and changes in the weighted average cost of capital on net present value (NPV) and levelized cost of electricity (LCOE).

Findings

Under a baseline case of 25% construction tariff (55% imported share) and 10% fuel tariff (70% imported share), overnight capital cost increases by 5.15% and operating costs by 1.31%. At 10% cost of capital, NPV declines from $168.4 million to $123.7 million (-36%), while LCOE rises from $61.73/MWh to $63.84/MWh (+3%). 


Scenario(10% WACC)NPV($)LCOE($/MWh)0% Tariff $ 168,389,276 $ 61.7325% Construction Tariff $ 123,670,285 $ 63.84+ 1% WACC $ 91,571,550 $ 65.04



Comparative marginal analysis shows that a one percent increase in the cost of capital reduces NPV by over $76 million and increases LCOE by $3.31/MWh – nearly twice the effect of a 25% construction tariff. Tariff effects are therefore amplified through capital intensity, compressing the margin of economic viability rather than proportionally increasing operating costs.

Implications

Trade policy and investment conditions jointly shape the deployment prospects of capital-intensive nuclear technologies. In higher cost-of-capital environments, moderate tariffs can materially narrow the economic margin for new projects, with implications for long-term decarbonization and supply chain resilience.

Many countries are currently weighing nuclear power for their low-carbon energy strategies, while others - including Switzerland - assess its possible role amid ongoing public debate. Parallelly, international energy system modelling studies evaluate the role of nuclear energy within broader transition pathways. Nevertheless, although some studies examined the influence of modelling choices on nuclear scenario outcomes, they focus solely on the power sector or on constructing a complete fuel cycle. We provide best-practice guidance for frameworks seeking to integrate such scenarios, answering the following research question(s): Which methodologies better capture the systemic impact of nuclear energy? What are the main technical (flexibility, operational, etc.), cost (investment, O&M, decommissioning, etc.), and policy (climate/renewable target, etc.) assumptions for nuclear energy? How can nuclear contribute to import reduction or relieve the need for storage and flexibility in the energy system? We focus on a Swiss case study, drawing on the Swiss TIMES Energy System Model (STEM). STEM is a bottom-up, technology-rich, whole-energy-system model with full sector coupling and hourly temporal resolution, making it well-suited for assessing the impact of nuclear energy, including its potential compatibility with renewables and potential operational constraints. From a methodological perspective, we model new nuclear plants accounting for sector-coupling implications and long-term horizons (2050+) to understand their influence on deployment outcomes. Moreover, we collect and compare key assumptions regarding nuclear energy technologies across international modelling frameworks, particularly those of the IAEA and IEA ETSAP. Among these are costs, operating and flexibility parameters, outputs (shares) to different markets, and policy settings. First results indicate that, assuming a 9,500 EUR/kW CAPEX and a construction time of 10 years, nuclear plants entering the Swiss energy system from 2050 generate a 27 BEUR systemic benefit by reducing electricity, energy carrier imports, and enabling deeper demand electrification.

Opening a nuclear power plant has historically been a precarious endeavor ridden with cost overruns and prone to cancellation; historically, the chance of completion has been around 50% in the U.S. Nonetheless, the U.S. is currently working to expand its nuclear capacity in a manner reminiscent of its first nuclear buildout during the mid-1960s and 70s (Harris & Wynne 1991). Alongside the well-documented factors behind cost overruns (Sovacool et al. 2014a, 2014b), characterizing the cancellation decision can help regulators better navigate the buildout process and thus lessen the potential cost-recovery burden on ratepayers.


We use data on electricity demand, reactor characteristics (e.g., size, type, location), interest rates, utility bond ratings, coal and natural gas prices, cost and time overruns, and policy shocks. Our first contribution and primary empirical specification is a survival model that casts plant cancellation as the end of a plant's survival spell. We capture dynamics in the cancellation decision by allowing our variables to vary during each plant's time-to-cancellation (with the exception of reactor characteristics). Our second contribution is the construction of a nuclear policy index, which we use as a continuous time-varying measure of policy shocks instead of binary indicators for events such as Three Mile Island and Chernobyl. This index is derived from the number of newspaper articles discussing matters related to the U.S. nuclear industry, thereby enabling us to simultaneously measure regulatory and public sentiment towards nuclear power.