Feasibility of magnetic signature-based detection of low and high impedance faults in low-voltage distribution networks

Abstract

This article investigates the feasibility of using noncontact magnetic sensor measurements as the basis for the detection of low and high impedance faults in 0.44 kV distribution networks. It is compared with fault detection using current measurement. A harmonic source free power system was simulated in DIgSilent, in which single line to ground faults was staged. Magnetic field intensity was calculated from simulated phase currents and then the potential detection of faults using total harmonic distortion (THD) and Discrete Wavelet Transform (DWT) was assessed. The use of magnetic sensor measurements as the basis for the detection of low and high impedance faults in 0.44 kV distribution networks was shown to be feasible. As with current-based measurements, wavelet analysis was found to be a better method for HIF faults. However, detection schemes proposed for current-based measurements will need to be modified to suit magnetic field measurements as the shape of the magnetic field waveform differs from the current waveform.