Analysis of online capacitor switching transient voltages on experimental MV cables

Underground residential distribution (URD) power cables are aged due to electrical, thermal, mechanical and environmental stress during their service. For utilities, the online monitoring and diagnostics of the dielectric conditions of the cables help in managing their cable asset. One of the cable online diagnostic methods is based on analysis of the capacitor switching surge transient voltages on cable. Switching transient voltages measured at both ends of cable carry significant information on the overall health condition of the cable insulation. An experimental MV power cable grid using field-aged and new cables was setup. Analysis of the switching transient voltages on the experimental grid using wavelet and principal component analysis will be presented. Significant-feature extraction and neural network based field-aged and new cable classification results will be shown.