Monitoring of Partial Discharges on Cable Terminations : An experimental approach to evaluate non-conventional online PD measurement techniques

Sammanfattning: The transmission of electric power over long distances has historically posed challenges. However, the advent of high-voltage engineering has not only addressed distance barriers and power losses in electricity transmission and distribution, but it has also significantly improved the efficiency and transmission capacity of power grids. However, the use of high-voltage techniques has presented new challenges in identifying suitable insulator materials capable of withstanding high electrical stresses associated with elevated voltages. A significant issue arising from these extreme electrical stresses is the occurrence of partial discharges (PDs). Those are electrical sparks or pulses in the magnitude order of pico- or nano-coulombs (pC or nC) emitted from high voltage conductors due to the presence of impurities, contaminants or defects in their insulation system. PDs pose a serious threat to the insulation material due to their aggressive nature in breaking down weak points or links inside the insulation system leading to short-circuit and system failure. Means of offline or conventional testing of power assets against partial discharges has proven to be extremely precise but it is often an unavailable option since it requires a total power shutdown. In this paper, the performance of existing online PD detection techniques is tested and evaluated in terms of performance against conventional PD monitoring methods. Five non-intrusive detectors including an infrared camera (IR-camera), an ultrasonic sensor , a temperature and relative humidity sensor (TRH-sensor), high-frequency current transformer (HFCT sensor) and transient earth voltage antenna (TEV-antenna) were tested in a laboratory experiment for detection of PDs emitted by artificially-created defects inside a medium-voltage cable termination. The results showed varying sensitivity levels among the sensors, with the HFCT sensor demonstrating the highest sensitivity to all types and magnitudes of PDs. The IR-camera and ultrasonic sensor also showed potential, while the TRH-sensor exhibited poor sensitivity. The TEV-antenna had limited reliability. The findings of this study are that the HFCT system proved to be highly reliable for online PD monitoring, followed by the IR-camera and ultrasonic sensor, while the TEV-antenna and TRH-sensor showed lower reliability. In future work, further research on testing the HFCT system on-site can be conducted along with performing longer thermographic detection tests using the IR-camera to further investigate their potential in online PD detection.