The relationship between partial discharge activity and hydrogen diffusion in epoxy resin and epoxy/mica composites

Ontario Hydro has used partial discharge (PD) testing to help assess the condition of generator stator insulation on a routine basis for over 30 years. However, in hydrogen-cooled turbine generators, PD activity is suppressed by the 300-500 kPa of hydrogen. For example, tests on a 24 kV, 542 MW generator showed that a PD magnitude of 240 pC occurred when the winding was pressurized, whereas the discharge magnitude was 2100 pC when the winding had been in atmospheric air for a period of time. Thus, it is more difficult to detect the PD pulses in the presence of electrical interference in operating generators pressurized with hydrogen. Under such conditions, the PD test is less sensitive to dangerous conditions such as loose wedges and delamination. To investigate the feasibility of reducing hydrogen pressure to increase temporarily PD activity, a series of small model experiments have been performed. The objective was to determine the variation in PD after an external pressure reduction. The controlled variables are hydrogen pressure and temperature. Two sample types were used. Theoretical and experimental diffusion rates of hydrogen through polymer and polymer/mica composites were also determined