Determination of Health Index of Insulating Oil of In-Service Transformer & Reactors Based on IFT Predicted by Multigene Symbolic Regression

This article aims at developing a multigene symbolic regression model of interfacial tension (IFT) for determining the health index of insulating oil. The oil parameters considered for modeling of IFT are break down voltage, resistivity, acidity, tan delta and water. The unique study undertaken in this article attempts to develop a multigene symbolic regression model for generating and evaluating the relationship between IFT and other oil parameters. Data utilized in this study involves actual oil testing parameters of in-service transformers and reactors of age more than 15 years. A sensitivity analysis has been carried out to investigate the relative contribution of input variables to the model output. The result of sensitivity analysis has further been used to determine the weight of parameters for calculating the health index of insulating oil. Finally, the health index of 400 in-service transformer and reactors are evaluated and analyzed. This research will help in undertaking decisions for the replacement of oil based on health index, thereby increasing the life span of transformer and reactors. The health index so obtained will act as a reference for planning necessary maintenance activities of the equipment and thereby saving unprecedented failure of the oil-filled transformer and reactors.