Evaluation of post‐Fukushima Japanese electricity strategies: A stochastic simulation model

Japan has not settled on a coherent long‐term electricity strategy for its post‐Fukushima reality. In this analysis, a stochastic electricity sector simulation model is formulated and used to evaluate alternative strategies prescribing different generation mixes for 2050. The generation mix evolves deterministically, demand is endogenous, and social cost of carbon is randomly sampled. The model treats fossil fuel prices and advanced technology cost dynamics stochastically. Monte Carlo simulation produces a distribution of net present social cost for each strategy, and the alternatives are assessed based on the median (expected cost) and standard deviation (risk). Aggressively expanding renewables is a promising strategy; it has the lowest expected cost and moderate risk. Whether this is preferred to substantial nuclear generation is sensitive to parameter assumptions, but a general finding is robust: severely limiting nuclear generation and substituting liquefied natural gas (LNG) and other fossil fuels is costly and risky. Principal component analysis confirms the vulnerability of such strategies to LNG price and social cost of carbon uncertainty. Japan's dominant short‐term response to Fukushima has been a dramatic decrease in nuclear generation coupled with increased use of fossil fuels, particularly LNG. This study suggests that continuing this approach would constitute a poor long‐term Japanese electricity strategy. Copyright © 2014 John Wiley & Sons, Ltd.