Electric distribution utilities generally operate circuits in a reactive mode, responding and making repairs after outages occur. They perform periodic maintenance on certain equipment, such as capacitor banks, but most apparatus (e.g., connectors, insulators, service transformers) are numerous, long-lived, and geographically dispersed, making inspection resource-intensive. It would be preferable to make repairs proactively before outages occur, but utilities lack information that enables this. Recent “smart grid” technologies restore service more rapidly after outages occur yet remain reactive.
Working closely with the Electric Power Research Institute (EPRI), Texas A&M University researchers collected and analyzed an extensive library of high-fidelity current and voltage waveform data from more than 70 in-service circuits. They discovered waveform signatures caused by nascent failures of line apparatus and in the process validated the notion that apparatus often deteriorate over time before failing. As a result, researchers demonstrated numerous cases where detecting incipient failures enabled utilities to avoid outages. In some cases, utilities also were able to schedule corrective actions during normal working hours and in favorable weather, rather than responding to outages in adverse conditions (e.g., storms, nights), thereby improving efficiency and crew safety.
Researchers have developed algorithms to characterize circuit health and events based on high-fidelity data digitized by substation-installed devices. On-line algorithms deliver real-time information to improve awareness of circuit conditions, thereby enabling improved reliability and operations. This enables a move away from reactive operations and toward condition-based approaches.
Multiple electric distribution utilities have participated in multi-year trials of the technology. Their experience includes detection of incipient failures, improved response to vague customer problems (e.g. flickering lights, lights out), and prevention of faults. This paper discusses the technology and relies on case studies illustrative of events on real circuits and of how personnel can use improved system awareness to better manage assets and improve operations.