A simple method to detect internal and external short-circuit faults, classify and locate different internal faults in transformers

Abstract

In new power systems, transformer monitoring and on-time detection of internal faults in power transformers are of great importance because if these faults are not identified on-time, the lifetime of the transformer might be reduced, the transformer might be completely disconnected, and network reliability might decrease. In this paper, a new method is proposed to detect internal and external short-circuit faults. It also allows us to classify and locate different internal short-circuit faults offline in power transformers. The proposed method is based on creating a locus curve of voltage–current difference ( $$ \Delta V - I $$ ) for all three phases of the transformer. This method can be implemented simply without requiring special equipment and only through monitoring terminal quantities of the transformer. In order to investigate the behavior of transformers in the presence of internal short-circuit fault, the finite element method is used. By investigating variations of locus diagrams concerning to the normal state, fault type and fault location are determined. The proposed method is applied to a distribution transformer white Yzn5 connection successfully. Simulation results verify the capability of the proposed method in transformer monitoring.