Perpustakaan Digital ITB

Transmission equipment such as substations are critical assets that require proper management to ensure optimal performance and longevity. Gas Insulated Substations (GIS), which are more compact and durable than Air Insulated Substations (AIS), still require effective asset management—especially for major maintenance like overhauls, typically performed every 20–25 years. Recently, condition- or risk-based maintenance has become more favorable than time-based approaches, requiring diagnostic indicators that are processed into a Health Index (HI). In this study, the HI is developed using Multi-Criteria Decision Analysis methods, including AHP, TOPSIS, and their combination. Due to incomplete data, imputation methods are applied: KNN is used to estimate the number of mechanical operations based on age, while Miss Forest imputes subsystem-level data using both age and operation count. Results show NRMSE values below 20%. After scoring, the AHP-TOPSIS method demonstrated greater sensitivity in distinguishing degraded GIS conditions compared to AHP or TOPSIS alone, making it the preferred method for overhaul prioritization. To address scoring bias from imputed data, a Data Completeness Index (DCI) is introduced for advanced analysis. Ultimately, overhaul prioritization for GIS is determined by combining HI and DCI scores.