Development of a Hybridized Model for Predicting the Life Span of Power Transformers

Abstract

Power transformers are important equipment of the electrical switchyard whose failure leads to long hours of outage. In this research, a two-stage hybridized model for determining the lifespan of power transformers is presented by using the furan content to determine the Degree of Polymerisation (DP) of transformer. For a 'virgin' transformer, the furan content was about 0.01ppm while a transformer with about 10ppm was within its end of useful life. 2-Furaldehyde (2FAL) content values of 0.01ppm and 10ppm correspond to DP values of approximately 1200 and 250, respectively. These parameters were used in developing a DP model using Jacobi and Gauss Seidel numerical analysis iterative techniques. The techniques were implemented in Matrix Laboratory 8.2 (R2013b) environment. The second stage involved the hybridisation of the developed DP model with another rate constant model adopted from Arrhenius. This stage was also implemented in Matrix Laboratory 8.2 (R2013b) environment. The life span of the transformer was determined by adding the service age at any point in time to the remaining lifetime at that point. A GUI of the hybridised model was developed using SIMULINK blocks. The developed model yielded a DP range of 247 ≤ DP ≤ 1184. Factors such as the hotspot temperature, activation energy and pre-exponential factor were useful for the determination of lifespan.