Modelling and Simulation of Fault Distance Locator for Underground Cable Detection

Abstract

Underground cables have been widely employed for power distribution networks for hundreds of years. Meanwhile, the preponderance of underground faults is detected by excavating and approach known the entire length of armoured cable. If a visual check is still not practicable, the entire cable length is replaced. To address the issue of faults in underground cables, this project will use wavelet analysis to detect fault locations and an artificial neural network (ANN) to evaluate fault performance. The objective of this method is utilized to detect the point and distance location of faults from the source end of an underground cable through transmission line. The underground cable system is to designed with MATLAB Simulink based on the IEEE bus system, which has 13 buses and 16 lines. There are three types of faults: single phase to earth fault (LG), double phase to earth fault (LLG), and three phase to earth fault (LLLG). The simulation results show that the voltage and current wavelet delaying in onset at 1.37ms and 1.66ms with distance (D) fault is 2.886 km from bus 4. The voltage and current wavelets of LLG is delayed in onset at 1.36ms with a distance (D) fault of 16.223 km from bus 9. Meanwhile, the voltage and current delaying in onset 1.35ms is observed for LLLG with a distance (D) of 2.499 km from bus 8.