Edinburgh Napier University

  Partial Discharge detection is crucial to diagnose the condition of the insulation system of High-voltage equipment. Despite the existence of many different methods to monitor the presence of PD, ultra-high frequency methods are frequently used since galvanic contact between the antenna and the equipment under test is not required during the measurement. However, the measurement of electromagnetic emissions using antennas may be influenced not only by noise from outside sources but also by reflections or changes in the signals caused by obstacles in the measurement environment. This work presents the results of a series of experimental measurements that led to characterizing the spectral behavior of the partial discharge signals captured with a Vivaldi antenna in the presence of two different configurations of obstacles located in the vicinity of the partial discharge emission source. From these results, it was possible to establish that the spectral content of the partial discharge signals in ultra-high frequency changes for certain bands or frequency intervals depending on the shape of the obstacles located in the measurement environment. Likewise, it was evidenced that the changes in the signals can modify the characteristic parameters based on the spectral behavior of the signals, generating variations in the results obtained in the processes of separation and identification of sources.