Collation of Mamdani-Type and Sugeno-Type Fuzzy Inference System for Condition Monitoring of Power Transformer

The condition of the solid insulation paper within transformers can be directly determined from the degree of polymerization (DP) measured from samples. Since this is difficult for in service units, many studies have been performed in regard with indirectly measuring the degradation of paper insulation by chemical markers for several decades. The 2-FAL concentration is being used decades ago. In this contribution, condition monitoring history and experience with transformer fleets are reported for a service aging of 75 years. 37 oil-filled transformers (146.7 kV to 157 kV) of a Canadian utility were periodically monitored. As part of the maintenance of the apparatus, the analysis of the furanic compounds is carried out. These are obtained by high performance liquid chromatography in a laboratory, with a regularity that varies according to the age of the transformer. This article deals with a transformer aging evaluation tool built from fuzzy logic. A fuzzy inference system is implemented by taking as input the values of 2-FAL and CO2/CO ratio. The tool helps directly providing the paper condition.