Bivariate statistics of lightning density and guaranteed quality of service in distribution network using copulas

Abstract

To properly set the guaranteed quality of service for different distribution system operators (DSOs), the regulator has to determine the influence of extreme events on the annual reliability indices and predict the reliability of these events. In this paper, the spatial correlation of lightning storms and annual reliability indices for different DSOs are determined. The correlation is calculated using the global annual reliability and lightning data aggregated from individual DSOs. Using Gaussian and t copula, the joint distribution of individual reliability indices (SAIFI and SAIDI) and lightning density for different DSOs and geographical regions has been determined. After the distribution fitting, it is possible to calculate the risk of not attaining the guaranteed quality of service for every individual DSO on condition that the lightning activity does not surpass the required level. The methodology is tested by using 7 year's period real operational data from 27 Serbian distribution utilities.