Abstract
On-line partial discharge (PD) testing has been used for decades to help maintenance personnel detect rotating machine stator winding insulation problems at an early stage and thus optimally plan any required maintenance. Specifically, the test can find loose, overheated, and contaminated windings, well before these problems lead to failure. The test has also found use in determining the effectiveness of any repair work.
The widespread application of on-line PD testing has enabled the accumulation over the past 20 years of a database with, currently, approximately 700,000 test results. This proposed contribution presents the basics of on-line data collection, including the implementation of noise rejection algorithms and possible methods in improvements in interpreting test results. One case study describing application of on-line partial discharge testing on a hydro generator will be presented.
