Increasing Grid and Asset Resilience: An Automatic and Unsupervised Approach to Partial Discharge Monitoring in Cables

Increasing grid and electrical asset resilience, while reducing safety risks for people and environment, can be achieved through effective condition monitoring of grid components. This can be done collecting diagnostic properties related to the most harmful aging mechanisms, and activating automatic alerts when those properties go beyond pre-fixed conditions. Partial discharge is probably the most important diagnostic property to assess heath condition and reliability of electrical insulation,. Inception and growth of partial discharges can cause, indeed, premature and unexpected grid component failure. An innovative partial discharge monitoring algorithm able to record, separate, recognize, identify and trend partial discharge signals, rejecting noise, is presented in this paper. The main purpose is to automatize the whole testing and alerting process, getting rid of the need to interpret and exploit each recorded partial discharge pattern. An application of this innovative technique on a defective polymeric cable, illustrated here, shows that the automatic algorithm is effective in rejecting noise and focusing on partial discharge time trend, thus providing a tool for alerting and condition maintenance approach that does not need any expert support.