Dissolved gases forecast to enhance oil-immersed transformer fault diagnosis with grey prediction-clustering analysis

Abstract

: A method is proposed for dissolved gases forecast and fault diagnosis in oil-immersed transformers using grey prediction–clustering analysis. Incipient faults can produce hydrocarbon molecules and carbon oxides due to the thermal decomposition of mineral oil, cellulose and other solid insulation. Dissolved gas analysis is employed to detect and monitor abnormal conditions in oil-immersed power transformers. However, the procedure takes a long time to decompose overall key gases and monitor conditions. The grey prediction GM(1, 2) model uses the variant information of hydrogen to forecast the further trends of both combustible and non-combustible gases. Grey clustering analysis is applied to diagnose internal faults including thermal faults, electrical faults and faults involving cellulose degradation. Numerical tests with field gas records were conducted to show the effectiveness of the proposed model, and are easy to implement with the help of portable devices.