Development of the on-line iPDM system for the PD diagnosis for 25.8kV C-GIS

As the recent trend of the monitoring and diagnosis technique for high-voltage GIS, many power industrial companies are researching and developing on-line preventive and diagnosis system of PD (partial discharge) for GIS. In succession of high-voltage class, the demand in transmission and distribution class is increasing and the scope of its application is spreading. Thus, we developed on-line iPDM (intelligent partial discharge monitoring) system for 25.8 kV cubicle-GIS (C-GIS). We are briefly introduced the configuration of on-line iPDM system using UHF methods for effectively diagnosing the insulation condition of C-GIS. For the internal UHF sensor, we designed the novel model with high sensitivity and checked experiment results as antenna performance factor. Comparing of ERA method and UHF method, the relationship between pC-dBm is described based on experimental results. The measured data is converted to frequency domain data and we can search the frequency with peak value except external noise. And then input data of ANN (artificial neural network) model is made from normalized zero span data. Finally, our expert system can judge the risk level from PD magnitude and its cause. In this paper, we will describe the development results and the actual experience of the on-line iPDM system for C-GIS in KOREA.