An online short-circuit impedance estimation approach to power transformer fault detection

Abstract

In addition to offline testing in transformers, advanced diagnostic techniques are experiencing significant advancements in on-line monitoring to ensure the continuity of electrical service and to minimize the costs associated with equipment repair and replacement. This paper presents a strategy for incipient fault detection in power transformers through online measurement of winding impedances. The strategy is based on calculating the series impedance for each phase, which includes the transformer and the connected equipment, by measuring the electrical variables of the transformer. Finally, a single fault indicator is obtained that provides information on the fault severity and the phase affected. The strategy is tested by finite element method simulations for different load conditions and fault severities. The strategy is validated with laboratory results on a 10 kVA three-phase transformer. The results obtained show the effectiveness of the proposed strategy in detecting faults in power transformers.