Application of statistical summary numbers to rotating machine PD Data for nonexpert identification of multiple aging mechanisms

In the past, statistical summary numbers such as skew and kurtosis have been applied to PD data to help establish a diagnostic framework for identifying the type, nature or location of a PD source as well as assessing the insulation health. This paper discusses a framework for applying the mode, mean, standard deviation, skew and kurtosis (advanced summary numbers or ASN for short) to on-line partial discharge (PD) data for the identification of individual PD mechanisms out of a group of multiple discharge mechanisms acting on a set of stator windings. The study shows how simple trending of these advanced summary numbers can help establish relative increases in the individual discharge mechanisms and consequently establish which PD sources constitute the predominant increase in partial discharge activity as evaluated via a normal NQN or Qm trend. Finally, a practical application on the application of ASN will be given.