Simple Calibration Method for Improving the Comparability of PD Measurement Results at HVDC Conditions

Abstract

Partial discharge (PD) measurement at DC voltage becomes more and more important due to the increased application of HVDC components in electrical power systems. Numerous studies of PD behaviour on various type of insulation at DC have been currently performed to find out appropriate test methods and procedures as well as the physical behaviour of insulation at DC stress. To compare different results of PD measurement one of the necessary conditions is given by the performance of a calibration procedure according to the relevant standards. However, the current measurement practice shows, that despite of using the same calibration procedure in various investigations different results were obtained at apparently the same test conditions. Therefore, it is proposed an additional calibration method for checking the validity of measurement results obtained in separate test HV laboratories. That method was proposed already in the past, but could be applied also nowadays with some modifications even for PD measurement at DC conditions. It is based on the well-known PD behaviour of a spark gap with inhomogeneous electric field at negative HVDC polarity. Under certain conditions the measurable PD have a stabile behaviour in magnitude and time. That could be used, at the same time, to estimate the charge value of pulse discharge generated by the PD activity of the spark gap at DC voltage stress. This method could be applied as a kind of calibration under real HV condition. The charge magnitude can be estimated easily by using a common oscilloscope and by calculation of the current-time integral value. The obtained value should be compared with the charge magnitude measured by a commonly calibrated PD measuring system. If the test conditions are chosen in the right way, the deviation of both values should be less than 10%. If possible, that test can be performed also directly with a test object in parallel connection to the spark gap under HVDC conditions. The paper deals with the description of necessary test conditions and design parameter of the spark gap. The described method enables a simple check of a PD test circuit including the PD measuring system in addition to the standardized calibration procedure.