Fault area estimation via intelligent processing of fault-induced transients

This paper presents a new method for estimating the fault area in electric power systems. The method is based on processing fault-induced electromagnetic transients through pattern matching. In this process, distinctive features of the recorded fault event are matched with the features of a large number of representative faults. The method originates from the concept of vulnerability contours which are used to assess the likelihood of voltage sags effecting a given network region. A limited number of sparsely located voltage waveform recording devices is used. Intelligent computing techniques are used to identify the most likely fault area. The method is aimed at complementing an earlier-developed technique for estimating fault location based on phasor voltage measurements. Good performance of the method is demonstrated through several examples using ATP simulations.