Statistical analysis of partial discharge data

Recently, considerable efforts were deployed at Hydro-Quebec to retrieve all partial discharge (PD) files recorded with the PDA-H instrument over the last 25 years and which were scattered all over in plant computers. These initial raw data files have been centralized on a corporate computer, where all new measurements are now systematically saved. The statistical analysis of the first 10000 files has served as a framework for establishing five levels of discharges severity. This paper presents the results of this statistical analysis which proposes a two dimensional classification of the discharge severity. It also attempts to compare the statistical distribution of the maximum discharge apparent amplitudes with what can be found in the industry. Finally, a description is given to show how the evolution of partial discharges over the years is used in a more general diagnostic process

[1]  P. Morshuis,et al.  Partial discharges at DC voltage: their mechanism, detection and analysis , 2005, IEEE Transactions on Dielectrics and Electrical Insulation.

[2]  G.C. Stone,et al.  Effect of manufacturer, winding age and insulation type on stator winding partial discharge levels , 2004, IEEE Electrical Insulation Magazine.

[3]  C. Hudon,et al.  Partial discharge signal interpretation for generator diagnostics , 2005, IEEE Transactions on Dielectrics and Electrical Insulation.