Uncertainty quantification in power system reliability using a Bayesian framework

Abstract

Long-term generation investment (LTGI) models have been widely used as a decision-making tool of design of energy policy. Adequate LTGI models with detailed modelling of operations are often computationally intensive. Uncertainty involved in these models poses a great challenge to the uncertainty quantification in power system reliability. This paper presents a Bayesian framework for addressing this challenge systematically. The use of Bayesian techniques enables an efficient model calibration and quantitative study on the robustness of different market designs. In the case study on the future UK power system, the robustness index estimated by the calibrated model is obtained through uncertainty analysis of loss-of-load expectation.