Temporary Monitoring of Stator Winding Insulation Using an Advanced PD System

In this paper, the condition assessment of the stator insulation of a 12 MVA, 10.5 kV synchronous generator with the use of an advanced partial discharge (PD) system is presented. The system is especially designed for temporary monitoring of partial discharges in various medium and high voltage equipment. A built-in computer enables long-term data collection and archival. Users can access this computer from any remote location to set up monitoring sessions and to view the collected PD data with the system's web interface. The system is configured for automatic data evaluation. An advanced synchronous multi-channel evaluation technique is implemented in the monitoring software to separate the PD signals from noise. The evaluation of the PD defect's type is automatically performed by the pattern classification system. It uses knowledge-based analysis to provide a deterministic decision for clear defect cases. For unclear cases, where the PD signal can be assigned to more than one PD defect class, an additional pattern recognition procedure is applied, which gives a probabilistic decision. The functionality of the system has been successfully confirmed in the case study described in the paper.