Oil-impregnated paper is usually the preferred choice of insulation in power transformers. The working life of any transformer depends on its insulation condition. The paper conductivity, moisture content and dissipation factor of solid insulation are a few crucial parameters used to determine the insulation condition. The reliable estimation of such parameters for diagnosis of solid insulation requires a complete profile of polarization current which is time-consuming. Hence, utilities always prefer non-invasive and rapid techniques to predict insulation conditions and minimize shutdown time. The proposed technique is based on a short-duration polarization current data. The present manuscript suggested a method that uses the initial decay rate of recorded polarization current data to predict various performance parameters related to paper insulation. In order to show the effectiveness of the proposed method, the analysis is performed on the data measured from various in-situ transformers.
Read full abstract