Underground Power Cable Fault Detection and Localization Based on the Spectrum of Propagation Functions

Abstract

Multiple potential defects in cables cannot be detected and localized due to the low detection sensitivity of the time-domain reflectometry (TDR) method, as well as the cross-term interference problem of the time- and frequency-domain reflectometry method. A new method for fault detection in underground power cables is proposed based on the spectrum of propagation function. First, the propagation characteristics of various frequency signals in cables are analyzed to identify the causes of local irregularities in the spectrum of the propagation function (SPF) resulting from defects. Furthermore, the locations of defects are successfully identified using the integral transform algorithm. As broadband impedance spectroscopy (BIS) cannot accurately measure the input impedance of a 163 m cable for defect detection, the spectrum of propagation function is more suitable for long-distance cables. Finally, cables of 50, 66, and 163 m lengths are examined for mechanical damage and thermal aging defects. The fault localization error is within 0.21%, providing a technical reference for fault detection of underground cables.