Fault Location Detection Using Impedance and Impulse Injection on 20 kV Distribution System Wonokromo Base on Geographic Information System

Abstract

The distribution system is one of the most important part in the power system electric supply. Long-distance topology made the distribution system is difficult to restore when the fault happened, because information about the fault location mostly counting on the maintenance crew or operator that patrol in the field, so it spent a lot of times for restoration and means economic losses not only on the supply side but also on the customer side. Unfortunately, fault cannot be avoided, even we have already done preventive and use the best protection device, it must have happened anytime and anywhere. This paper described a comparative analysis between impedance base method and impulse injection method to find fault location on overhead 20 kV distribution system, Wonokromo, Surabaya, East Java. Wonokromo distribution system capacity is about 60 MVA that used to supply PDAM Surya Sembada Kota Surabaya for IPAM Ngagel 1 and IPAM Ngagel 2 with a load capacity of 3,645 kVA and 1,500 kVA with a distance of about 738.9 m. The distribution system has been modeled using Simulink, MATLAB r2016a that validated using ETAP 12.6. The simulation is done by varying the fault type (LG, LLG, and LLLG), fault resistance (Rf : 0.001, 10, 50, and 100 ohm), then the location of the fault ( 25 %, 50%, 75%, and 100% of distance). From the simulation result, we get information that the impedance base method is better than the impulse injection method when the detail parameter like line branch, load, sequence impedance, and fault resistance is completed, but for the system that have no detail data, the impulse injection is better. The accuracy of the impedance method is about 0.25%, while the injection method accuracy is about 1.7%. The simulation result can be visualized using Geographic Information System (GIS) to track and check the real location base on the calculation result.