Penentuan Prioritas Pemeliharaan Trafo Distribusi dengan Integrasi Metode Delphi dan AHP

Abstract

Distribution transformer maintenance plays a critical role in ensuring the continuity and reliability of PLN’s electricity supply. Currently, PLN employs a Condition-Based Maintenance (CBM) approach, relying on a Health Index (HI) derived from Tier 1 and Tier 2 inspection outcomes. However, in practice, determining maintenance priorities remains largely dependent on the intuition and technical experience of decision-makers, leading to potential inconsistencies in decision-making. Additionally, the absence of a structured weighting mechanism among HI parameters and unclear relevance of indicators used further hampers valid and reliable decision-making aligned with current operational conditions. This research aims to propose a structured and adaptive Health Index evaluation method through the integration of the Delphi and Analytical Hierarchy Process (AHP) methods. The Delphi method was employed to identify the most relevant parameters based on expert consensus, while the AHP method provided logical and consistent weighting among these parameters. Although AHP does not entirely eliminate subjectivity, as it still depends on expert judgments, it offers a systematic and transparent way to manage conflicting criteria effectively. The Delphi application yielded seven sub-criteria parameters grouped into three main criteria: oil leakage, transformer body temperature, load percentage, grounding value, physical condition, transformer age, and neutral current percentage. Based on AHP analysis, the oil leakage parameter was identified as having the highest weighting (0.23), indicating its significant impact on transformer performance. Applying this methodology to 253 distribution transformers resulted in prioritizing maintenance into three categories: Priority I (4 transformers), Priority II (54 transformers), and Priority III (195 transformers). The findings and validation from this study demonstrate that a structured, weighting-based approach significantly enhances the accuracy and transparency of decision-making processes in asset maintenance management.