New probabilistic approach for identification event severity index due to short circuit fault

In this paper, it is introduced a new approach to identify bus voltage severity profile due to short circuit fault at a certain point in a distribution power system. Short circuit causes voltage decrement for duration of time related to opening time of a relay. Data containing 2 variables, depth and duration of voltage sag due to short circuit faults on every buses, are generated. Subsequently, these data from all of buses will be clustered using K-means Clustering. Clustering data will produce center clusters and cluster membership. To be able to perceive voltage sag severity, center clusters will be converted to Event Severity Index which explains severity of a voltage sag event based on CBEMA-ITI Curve. Data of a certain bus which undergoes voltage sag events will be classified based on its cluster membership or center cluster. Thus, it will be obtained frequency of events that are classified into particular clusters; how many events that is classified into particular clusters. In order to observe data well, it is better to present it making use of histograms.