Development of indicator based on oil analysis for estimating risk level of power transformers

Abstract

Risk assessment for power transformers has long been an effective tool to analyze transformer conditions in order to prevent damage on transformers, extend transformer's life time and reduce maintenance costs. Various parameters can be evaluated, and decision can be made manually or with the help of mathematical model. A new mathematical risk-assessment model for power transformers is being developed for power system in Latvia, based on specific maintenance features and 4 indicators: dissolved gas analysis; oil analysis, electrical measurements and operation history. This paper specifies oil analysis as one of the indicators of regular risk assessment of power transformers. Development of this indicator is presented based on traditional binary evaluation algorithm combined with fuzzy logic, and as a case study results of oil parameters from different transformers are analyzed and discussed.