Remaining Useful Life Prediction of 20 kV Distribution Insulators Using Linear Regression

Insulators are essential components in 20 kV electrical distribution systems, serving as both mechanical supports and electrical isolators for HV conductors. Continuous exposure to environmental factors such as humidity, pollution, and ultraviolet radiation accelerates the aging process of insulating materials, causing gradual degradation in electrical performance. This study examines the degradation characteristics of three insulator types polymer, glass, and porcelain based on two diagnostic parameters: insulation resistance (IR) and leakage current. The objective is to estimate the remaining useful life (RUL) of each type under real field operating conditions in the UP3 Tanjungpinang distribution area. A descriptive quantitative method was applied using 90 consecutive days of IR and leakage current measurements obtained from a Megaohm meter and Clamp Meter. Linear regression analysis was used to model degradation behavior and determine the time when IR falls below 20 MΩ or leakage current exceeds 1.5 mA, indicating the critical aging threshold. The regression analysis yielded high coefficients of determination ranging from R2 = 0.9999 – 1.0000, indicating excellent model accuracy and a nearly perfect linear relationship between aging time and degradation rate. The results show that glass insulators have the longest estimated RUL of approximately 24.4 years, followed by porcelain at 19.8 years, and polymer at 9.7 years. These findings confirm that IR and leakage current are reliable diagnostic indicators for assessing insulation health and support the application of predictive maintenance strategies to enhance the reliability, efficiency, and safety of 20 kV distribution systems.