Research on Characteristics of Typical Partial Discharge Pattern in GIS Based on SiPM
Zihan Xu,Lijun Zheng,Ruixiang Liu,Chuan Peng,Zhijie Zhang,Pengyue Gao,Liang,Zhipeng Lei
DOI: https://doi.org/10.1109/icd59037.2024.10613210
2024-01-01
Abstract:Inevitable defects often occur in the process of GIS manufacturing, transportation and assembly, which will affect the operation of the power grid. In this paper, an array optical sensor is designed based on a Silicon photomultiplier tube (SiPM) to detect partial discharge in GIS, and a measurement platform and defect model are built to simulate partial discharge in GIS according to the conditions required by the optical measurement method. Corona discharge, creepage discharge, internal discharge and discharge between metallic parts on floating potential are measured by SiPM array, and the influence of sensor distances of 10 cm, 15 cm and 20 cm on the sensitivity is investigated. According to the obtained data, the phase-resolved partial discharge (PRPD) spectra of four discharge types are obtained. Then, the PD feature spectra, such as phase-maximum discharge quantity, phase-average discharge quantity, and discharge frequency-discharge quantity, are analyzed, from which statistical characteristics, including skewness, kurtosis, discharge factor, phase asymmetry, and correlation coefficient, are extracted. The results indicate a negative correlation between the sensor's sensitivity and the distance between discharge sources. Different discharge types can be effectively classified by comparing phase-discharge quantity spectra and statistical characteristics.