FAULT LOCATION TECHNIQUES IN DISTRIBUTION SYSTEMS

Abstract

Between generating stations and customers, there are large transmission and distribution networks. Faults (either temporary or permanent) are common in these networks. Temporary faults are self cleared so the continuity of power supply is not affected by them but permanent faults cause interruption in electric power continuity. In this deregulated environment the quality of power has an important issue for the customers. By locating the permanent fault, rapid restoration of supply is possible, which further improves the reliability indices (SAIFI, SAIDI etc.) Also an early detection and location of temporary fault helps the utility to take timely remedial. A lot of research work has been carried out in the field of fault location, in transmission as well as in distribution system. These fault location methods can be grouped into three categories- traveling wave based methods, artificial intelligence based methods and apparent impedance based methods. Apparent impedance based methods are simpler and widely used for locating the faults in distribution systems. They require local voltage and current measurements as input data and then use different formulations for fault location, based on fault type. Generally in fault location problems it is assumed that fault has been detected and classified accurately. However in real situations both are unknown. In this work wavelet transform is used for fault detection. Then a combined wavelet and ANN based approach is used for fault type identification and finally location of fault is pinpointed using apparent impedance based methods. This technique has then been tested extensively on an 11 bus real distribution system ALIPL from Electrical Energy Distribution State Company of Rio Grande do Sul (CEEE-D), Southern Brazil.