Natural esters, such as rape seed oil and soybean oil, are increasingly being applied as insulating oils for transformers in view of carbon neutrality and fire protection. Their chemical compositions vary from those of mineral oils, therefore, the type and amount of gases generated under the faults, such as overheating and discharge, could differ from each other. In this study, we discuss the differences in gas generation characteristics between mineral oils and natural esters with the help of the random forest method, which is one of the machine learning techniques.
NETL
