Evaluation of Uncertainty in Partial Discharge Measurement: A Case Study

Abstract

ISO/IEC 17025:2017 recommends all accredited laboratories under this standard, performing testing and calibration, shall identify and evaluate the contributions to the uncertainty of their measurements. This paper represents the results of the 'measurement uncertainty' assessment carried out on a partial discharge (PD) test set in our laboratory. The major sources of uncertainty include sampling error, calibrator inaccuracy, the resolution limit of display panels, temperature and humidity factors, etc. The uncertainty for each factor was calculated by the principles outlined in NIST-1297:1994 and JCGM-100:2008. The test PD levels were 1, 5, 10, 20, 50 and 100 pC. The combined uncertainty was calculated based on all uncertainty contributions, and the expanded uncertainties were expressed with 68%, 95% and 99% confidence level. Apparently, the measurement uncertainty was highest for the lowest test PD which decreased with the increase in PD level. The uncertainty was around 18% at 1 pC PD whereas it dropped to 2.84% at 100 pC PD. Within the whole test range of 5 to 100 pC, the maximum uncertainty was equal to 5.81%.