Moderately extreme weather contributions to power supply inadequacy: Identification using rapid loss-of-load estimation

Abstract

Power systems with high shares of variable renewable energy (VRE) are increasingly vulnerable to extreme weather events. While existing studies typically identify extremes based on meteorological thresholds, such as sustained low wind, low solar radiation, or extreme temperatures, these individual-variable extremes do not always cause the most severe power shortages. In contrast, moderately extreme events, namely compound weather conditions that are not individually severe in any single meteorological variable but jointly create unfavorable electricity supply-demand imbalances, can pose greater risks. To address this gap, this study aims to develop a consequence-based framework that directly identifies weather events causing the most severe power inadequacy risks, rather than relying solely on meteorological definitions of extremes. First, multi-decadal time series of wind, solar, and electricity demand are generated under various future capacity mixes with high VRE penetration. Then, a computationally efficient loss-of-load estimation method is proposed based on algebraic computations rather than mathematical optimization to identify weather events most likely to cause severe power shortfalls. Finally, power shortage risks are evaluated using power system economic dispatch simulations and compared across different types of extreme weather. Simulation results show that the proposed method can estimate loss-of-load with high accuracy and at a speed hundreds of times faster than dispatch optimization models. The case study reveals that identified events often involve moderately low VRE output and moderately high demand occurring simultaneously, resulting in severe shortages. At equal occurrence frequencies, these identified events pose several times the risk to power supply adequacy compared to individual-variable extremes and should be prioritized in power system planning.