Valuing fuel diversification in power generation capacity planning

Deterministic capacity planning problems in electricity systems can be solved by comparing technology specific long-term and short-term marginal costs. In an uncertain market environment, Mean-Variance Portfolio (MVP) theory provides a consistent framework to balance risk and return in power generation portfolios. Focusing on fuel price risks, MVP theory can be adopted to determine the welfare efficient system generation technology mix.Existing literature on MVP applications in electricity generation markets uses predominantly numerical methods to characterize portfolio risks. In contrast, this article presents a novel analytical approach combining conceptual elements of classical capacity planning models and MVP theory to derive the efficient portfolio structure consisting of an arbitrary number of plant technologies given uncertain fuel prices. For this purpose, we provide a static optimization model which allows to fully capture fuel price risks in a mean variance portfolio framework. The analytically derived optimality conditions contribute to a better understanding of the optimal investment policy and its risk characteristics compared to existing numerical methods. Furthermore, we demonstrate an application of the proposed framework and provide results for the German electricity market which has been hardly treated in MVP literature on electricity markets.This article provides easily interpretable analytical optimality conditions for efficient generation portfolios from a societal point of view and therewith contributes to a better understanding of MVP in electricity applications.