Distinguishing between specific deterioration phenomena in stator windings and cross coupled PD

On-line partial discharge testing has been used over 50 years to assess the condition of motor and generator stator winding insulation systems. In the past decade, development of advanced software enhancing PD interpretation by graphical means has helped identify specific insulation deterioration phenomena acting on stator windings. Thus, in many cases the root cause for insulation deterioration may be identified and proper preventive maintenance may be applied to the insulation system to help prevent insulation failure. By investigation of typical PD data acquired on large rotating machines during normal operational conditions, this paper demonstrates how specific PD sources can be identified using advanced PD interpretation software and classic electrical field theory. Using recent data extracted from a large PD database and processed in a newly developed software program, the paper evolves around initial PD pattern recognition using pulse phase analysis leading to detection and identification of specific discharge sources and their cross-coupled components. In addition, the paper discusses practical limits in PD interpretation techniques.