Detection of high impedance faults in medium voltage distribution networks

Abstract

High-impedance faults (HIFs) on distribution system still present the most persistent and challenging problem to protection engineers due to the fact that HIFs do not produce enough fault current, thus not detectable by conventional overcurrent protection devices. Previous research shows that it is possible to distinguish HIFs from other similar waveforms such as nonlinear load currents by analyzing the harmonic contents. A method for HIF detection based on the harmonic analysis of current waveforms is presented here. A harmonic detection program is implemented in MATLAB using both Interpolation Windowed Fast Fourier Transform and All-phase Fast Fourier Transform algorithms. Various simulation results and real-world data analysis show that this harmonic detection program can accurately, reliably and quickly determine the harmonic contents (including frequency, amplitude and phase angle of each harmonic) in an arbitrary signal without knowing its mathematical expression. Simulation results also show that this harmonic detection program could be used for feature extraction and pattern recognition for HIF detection in the future.