Identification, Location, and Remediation of Incipient Fault and Failure Conditions Using Waveform Monitoring and Automated Analysis

Abstract

Faults and failures are inevitable on power systems. Many power system faults are singular, one-off events – for example a danger tree which falls and tears down a conductor, or a squirrel which spans the bushings of a single-phase transformer and falls to the ground. Over decades, protective engineers have developed strategies for dealing with these events and minimizing their effect on customers. Some faults, however, including some faults which are cleared by system protection and self-heal, are not one-off events. These faults are caused by certain environmental conditions which remain present even after the fault is cleared, and as a result have a chance to recur in the future, should similar conditions manifest. For example, wind might cause vegetation to push two conductors together, but only when it blows from a certain direction at a certain wind speed. In this scenario, protection might operate correctly for each fault event, but the series of faults is indicative of an incipient fault, because the latent condition persists on the system, and the possibility of future faults remains.Research at Texas A&M University has documented numerous cases of latent, incipient fault conditions. Drawing from a database of over 1,500 circuit-years of operational experience on real-world systems, researchers have developed a series of technologies which automatically identify faults with similar characteristics and present actionable information to utility personnel. In many cases, this enables utilities to find and remediate incipient fault conditions before they escalate to final failure. This paper presents illustrative case-studies from actual field examples.