Automatic fault location on distribution network using voltage sags measurements

This paper presents a method that combines a power system analysis program, a database search method and a pattern recognition technique to identify a faulty section in a typical 11kV distribution network automatically. The method uses a three-phase unbalanced load flow and fault analysis program to analytically generate a voltage sag database for faults on all nodes in the distribution network. Database search and pattern recognition algorithms were developed to identify the faulty section when an actual fault occurs anywhere on the distribution network. The proposed method was evaluated and tested on the distribution network using EMTDC, a power system simulator.