Analysis of partial discharge features as prognostic indicators of electrical treeing

The aim of this paper is to identify promising indicators for prognosis of electrical treeing. Both phase-resolved partial discharge analysis (PRPDA) and pulse sequence analysis (PSA) are utilized. The prognostic properties of the features are evaluated in terms of monotonicity, prognosability, and trendability. The investigation reveals that PSA has a higher prognostic suitability index than PRPDA. An exponential fit is applied to the feature with the highest suitability index, in order to demonstrate its use for prognostic modeling and prediction of time until breakdown.