Quantification of aleatory and epistemic uncertainty in bulk power system reliability evaluation

Uncertainty can be generally classified as aleatory uncertainty, related to random behaviour, and epistemic uncertainty, related to a lack of information. This paper studies the uncertainty in bulk power system reliability assessment. It analyses the source of uncertainty in components outage model and refers it to the congruous uncertianty forms. This leads to the characterization of outage model uncertainty by a mixed aleatory-epistemic model. In consequence, the paper suggests two methods for representing and quantifying the effect of mixed aleatory-epistemic uncertainty on system reliability indices. These methods are second order probability and evidence theory. The paper validates the feasibility of the two methods by conducting case studies on the IEEE-RTS system using the non-sequential Monte Carlo simulation approach. The mixed aleatory-epistemic model is shown to provide more detailed estimate of uncertainty than the typical aleatory model.