Toward Unattended 24/7 Monitoring and Localization of Leaks of Electrical Insulating Gases and Their Decomposition Products

An electrical power grid consists of diverse types of gas-insulated equipment (GIE) that experiences different environmental stress depending on geographical locations and climate conditions. Over the operational lifetime, gradual aging of GIE can lead to partial discharges, decomposition of electrical insulating gas media, and gas leaks. Prognostics and elimination of leaks of electrical insulating gas media is an important unmet need in making the grid safer. Here, we demonstrate an approach for unattended monitoring and localization of leaks of electrical insulating gases and their decomposition products from GIE assets using spatially distributed sensor node systems. These sensors are based on multivariable design principles and utilize a multiband dielectric excitation scheme. In initial laboratory tests, the Limits of Detection (LODs) at part-per-billion levels of gas concentrations were achieved in detection of model decomposition products of g3. Further, different leak volatiles were differentiated indoors and outdoors under dynamic test conditions. Localization of leaks of different volatiles was also achieved on a model outdoor industrial asset. The detector sensitivity and low noise levels allowed detection of model leaks as far as 52 meters in indoor settings and 43 meters outdoors under ambient wind conditions. We are working toward field validation of developed sensor systems to deliver reliable alarms 1 – 6 months earlier than conventional alarms, thus providing solutions for more cost-effective monitoring and management of GIE assets of utilities.