Multispectral detection of partial discharge in SF6 gas with silicon photomultiplier-based sensor array

The upgrading demand on the fault prevention necessitates highly reliable and precise approaches for monitoring partial discharge (PD) which is the essential potential hazards existing in SF6-insulated power system under electric field. In this paper, a micro SiPM-based multispectral PD sensor (SMPDS) is first developed for in-situ optical PD monitoring and the corresponding strategy for multispectral PD diagnosis is also proposed. Experiments indicates that SMPDS can efficiently response to various types of PD including point corona, gas spark and surface creepage with significant dissimilarities in the three spectral bands. The triangle clustering approach based on SMPDS stochastic detection can distinguish the different natures of PD with a satisfactory discrimination even without conventional phase-resolved statistics. This proposed solution provides a way for optical PD in-situ monitoring, circumventing the necessities of anti-interference measure and complex diagnosis algorithm.