Physical modelling and mathematical description of PD processes with application to PRPD patterns

Abstract

A general framework for the physical description of partial discharge (PD) processes is presented that holds for different types of PD defects. A PD process is a special stochastic process consisting of point-like discharges and drift/recombination phases between the discharges where only the fast point-like discharges can be measured as PD signals. It is determined by few basic physical parameters. Mathematically, a PD process can be formulated either in a point process framework or in a stochastic process framework. Both approaches are presented. The link between the process and observable quantities like phase resolved PD (PRPD) patterns is derived. With the new approach a much easier interpretation and analysis of PD patterns is possible. For all of the three discharge types (internal, surface, corona) examples are presented that demonstrate the applicability of the model.