Power System Risk Analysis Under Increasing Levels of Variable Energy Resources

Abstract

As the portfolio of energy resources serving the electric power system transitions from a low to high percentage of stochastic energy resources, grid planners and operators must reevaluate the conventional methods used to characterize resource adequacy and the ability of power system generation sources to meet the modern electricity demand. This thesis reviews the historical and state-of-the-art methods used to evaluate resource adequacy, and presents a case study examining the reliability characteristics of a bulk power system with increasing wind and solar penetration (commonly referred to as variable energy resources). The scope of reliability in this thesis centers on probabilistic analysis of system resource adequacy. It provides examples of the pitfalls of misrepresenting dependencies in power generation from variable resources and proposes future work for further examination and extension of these concepts. A case study using historical weather patterns and correlated wind and solar generation in the WECC region of the U.S. is used to compare different reliability metric analysis methods. Results show that neglecting the generation uncertainty dependence model of variable energy resources significantly underestimates the risk of electric grid operation served primarily by variable energy resources; this highlights a major disconnect with the net load treatment of stochastic generation used prominently by system operators today. The accuracy of probabilistic models that represent system generation are of increasing importance in systems with high amounts of variable energy generation.