Partial discharges at DC voltage - measurement and pattern recognition

Knowledge of the current status of the components of an electrical power supply is essential for power grids to operate safely and economically. This means that the partial discharge diagnosis is of great importance for AC grids. It also plays an important role in connection with the increase in HVDC technology; the PD diagnosis is important for the condition assessment of the involved components. The AC PD measurement, the measurement procedure, and especially the interpretation of results cannot fully be adapted to DC voltage. In this paper, the previous findings of investigations about the PD behavior of some typical defects under different artificial environmental conditions and different measuring procedures are summarized. To analyze the numerous measured PD sequences, a software tool called “PDDC Analyzer” was developed. With this tool, well-known DC PD graphs can be shown and new methods for creating a DC PD pattern for defect identification and risk assessment can be investigated. The quality of the improved NoDi* Pattern is shown and the measurement procedures are discussed. At the end, new methods and aspects for PD identification at DC are introduced and a short overview of further research is given.