Evaluating the approximate probability distributions of load point reliability indices in power distribution networks

The reliability assessment of power distribution networks is concerned with the evaluation of a set of five basic indices for each load point of interest. These indices are random variables because they are functions of the component performance parameters and the characteristics of system operating procedures which are random variables by nature. Computational modelling techniques have been published to evaluate the mean values of the indices, applying both the total and partial loss-of-continuity criteria. It is, however, important to obtain more information about the variation of the indices around their means. The purpose of the paper is to describe the computational techniques that have been developed to evaluate the approximate probability distributions associated with the load-point reliability indices in a meshed distribution system with any given configuration. These techniques incorporate all realistic failure and restoration modes that are known to exist in power distribution networks. The results of an analysis on a 150/20 kV Hellenic distribution system are also presented in the paper