Partial Discharge and Conduction Current Signature Analysis for the Assessment of Transformer Oil Behavior under Accelerated Stress Condition

Abstract

2 Abstract: Performance and serviceable life of oil filled transformers are primarily governed by the behavior of insulating oil. Such transformers, during operation, may experience several stress evolving operating conditions. Deterioration in the performance of oil filled transformers are directly associated with the degradation of oil during the course of operation with different loading conditions. Further, the oil characteristics undergo appreciable variations since the chemical compositions in the oil are subjected to variations. Partial Discharges (PD) in insulating oil is one of the proven indices which exhibit the changing characteristics of oil. PD activity in oil insulating medium is influenced by several factors that are invariably directly correlated to the physics and chemistry of the insulant. The proposed article focuses on assessing the PD activity in different transformer oil insulations under different applied voltages. Further, the investigations also confined towards the correlation of PD phenomenon with Conduction Current (CC) through insulating oil. Various categories of iso-paraffin base and naphthenic base transformer oil have been taken into consideration which include new untreated oil, thermally degraded oil, oil degraded with transient signals and mixture of oil with solid dielectric materials. Detailed analysis based on experimental studies reveals that the dynamics of PD activity in degraded oil display unique discharge signatures when compared with new untreated oil. Also, it is evident that the PD activity is enhanced with the oil which contains a mixture of solid dielectric materials and with thermally degraded oil. Further, simultaneous measurement of conduction current and partial discharge activity in oil depicts that the dynamics of oil degradation is well observable in conduction current wave and correlation of both signatures yields better results interms of understanding the changes in oil behavior with time.