Analysis and detection of volt-ampere characteristics for high impedance faults in distribution systems

High impedance faults (HIFs) often occur in effectively grounded medium voltage (MV) distribution systems, while their fault currents are too low to be tripped by zero-sequence over-current protection, but the existing arc flash is easy to produce fire hazard and personal injury. Existing HIF detection (HIFD) algorithms, which are mainly based on arc thermal equation and frequency domain analysis based on Fourier transform, however, have a relatively poorer performance because of strong noises accompanied by arc flash. Aiming to solve this problem, an accurate nonlinear resistance model, based on the solid dielectric electrical break-down theory, was proposed to simulate the fault point. After analyzing the Voltage-Current characteristics of linear electrical components and nonlinear arc resistance, a nonlinear resistance identification based HIFD algorithm was proposed with fault characteristic extraction by least square fitting. Extensive field data and simulation results testified the higher accuracy of the proposed fault model over traditional models, as well as high sensitivity and reliability of the proposed algorithm. © 2014 Chinese Society for Electrical Engineering.