Condition prediction for oil-immersed cellulose insulation in field transformer using fitting fingerprint database

Abstract

The results of frequency domain spectroscopy (FDS) will be affected by moisture which often leads to an unreliable condition prediction of oil-immersed cellulose insulation in transformers. In view of this, a solution is proposed in this study by adopting a fitting fingerprint database (FFD) technique, which can be utilized for condition prediction of aging and moisture content. In the current work, the relevant fingerprints that could characterize the insulation aging condition and moisture content are extracted from the FDS curves and the DC conductivity of the transformer oil. The FFD is then constructed by fitting fingerprints. The accuracy of the reported FFD technique is verified from tests on oil-immersed pressboard prepared in the lab and a transformer from the field. In this respect, the present contribution highlights FFD might be used as a powerful tool for condition prediction of oil-immersed cellulose insulation in transformers.