Power transformers are one of the important components in an electric power system that is expected to operate optimally. Along with its use, some factors cause the depreciation of the transformer’s life, such as loading factors and environmental temperature. This study aims to predict loading based on historical loading data using linear regression to determine the depreciation and estimated life of the transformer. The research was conducted on three transformers in locations with different environmental temperatures. The test results showed that the linear regression model had a high R-Squared validation value of 0.949 for Transformer#1, 0.948 for Transformer#2, and 0.945 for Transformer#3. Then the MAPE error value obtained for the three transformers was low, namely 2.99%, 1.76%, and 4.69%. The prediction results are used to calculate the estimated remaining life of the transformer, which shows that Transformer#1 is expected to have a lifespan of <1 year by 2029, Transformer #2 in 2026, and Transformer#3 in 2025. The results showed that the load and ambient temperature significantly affect the transformer's life loss.
Copyrights © 2026