Black-out test vs. UV camera for corona inspection of HV motor stator endwindings

As part of a factory acceptance test program for high voltage motors, end users sometimes specify a black-out test. This is a traditional offline inspection where the stator is placed in complete darkness, each phase is energized to 115% of rated line-to-neutral voltage, and both ends of the stator are observed to determine the presence, location and severity of endwinding surface partial discharges (PD). The setup for the test may be complex and risky due to the need for observers to be standing in darkness close to the energized parts of the stator. The test results are qualitative and strongly depend on the observer's eyesight and individual perception. A safer and more accurate alternative is to use an ultraviolet (UV) camera or viewer. The observed PD activity may be observed in ambient lighting, recorded, and quantified through simultaneous offline PD measurements. This paper describes the two inspection techniques, and presents experimental validation of the UV corona camera inspection method as a suitable replacement for a black-out test. Sample 13.8-kV coils were wound in a fixture simulating their relationship in a stator winding, and subjected to high-potential tests while observed under black-out conditions and with a UV corona camera in ambient lighting. The stator windings from two high voltage compressor motors were inspected using the same camera. Recorded images of the observed discharges and measured PD activity in the sample coils and stator winding were used to compare the evaluation by each test method.