Identification of the minimum detection of transformer bushing failure based on Frequency Response Analysis (FRA)

Abstract

Frequency Response Analysis (FRA) is the most reliable technique for detecting a mechanical failure within electrical power transformers. Associated literature reports the ability of the FRA technique to identify non-mechanical faults such as transformer bushing failures. However, the interpretation of the FRA signature is still a challenge because of its reliance on graphical analysis, which requires an expert person for the interpretation process. Although the impact of transformer failure on the signature of the FRA technique has been intensively investigated, no standard code has been widely adopted. Moreover, the FRA technique is frequently ineligible in detecting minor levels of power transformer fault that decreases the transformer's ability to withstand any further electrical, thermal and mechanical stresses. In order to apply the FRA technique properly, the minimum detection of transformer faults should be estimated. This paper aims to identify the impact of the minimum level of transformer bushing failure on the FRA signature that can be visually detected. In this regard, a 3-D geometric of a 3-phase power transformer is simulated (based on the Finite Element Analysis (FEA) technique) to imitate its real operation during both healthy and faulty conditions. The impact FRA signatures (based on FEA) for various bushing failure levels was obtained and then compared to the healthy signature, and variations are detected if they exist. Results show that the transformer bushing failure has a minimum level that can be visually detected using the FRA technique.