A novel mathematical approach for modeling the dielectric response surface of the transformer oil to predict the optimized functions of transformers in high-power applications

Abstract

Due to increase in the number of transformers, the need for transformer dielectric coolant (TDC) is gradually rising. Transformers are continuously under stress, which shortens the TDC's shelf life. In this experimental investigation, the antioxidants (Axes), butylated hydroxytoluene, and citric acid, as well as the emulsifier lecithin (Ln), were blended with TDC to increase the dielectric performance of the TDC. Using the Box–Behnken and response surface methodology, the ideal quantity of the Ax and Ln is determined. In this experimental design, dielectric breakdown voltage (DBV), kinematic viscosity (KVIS), fire point (FRPT), and electrical conductivity (ELC) are the four dependent response variables, while the Ln, PriAx and PreAx are the three independent process variables. Based on the experimental design, the optimal values for process and response variables are examined using the response and contour plots. The optimal quantity of the process variables is used for blending the TDC with Ax and Ln. After optimizing the process conditions, Ln = 0.295 g, PriAx = 0.377 g, PreAx = 0.248 g, the experimental response was DBV = 66.092 kV, KVIS = 15.465 cSt, FRPT = 189.734° C, ELC = 0.411 µS / m. These actual values deviate slightly from the expected response values, DBV = 65.69 kV, KVIS = 15.49 cSt, FRPT = 189.75° C, ELC = 0.404 µS / m. This demonstrates how well the mathematical model influences the dielectric response of the TDC. They also show that treating TDC with optimal amounts of Ln and Ax improves dielectric response and durability. Lecithin helps emulsify immiscible liquid mixtures by lowering surface tension. Lecithin's innate antioxidant properties provide protection against product degradation, wherever it is used.