Detection of inter-turn fault in transformers at incipient level

Transformers have emerged as an integrated part of a power system. Any fault in the transformer can cause a severe outage, which therefore necessitates continuous monitoring and diagnostics of its operation. The renewed thrust in smart power system networks along with the development of advanced methods in the monitoring and diagnostics has resulted in a major impetus for research in the related domain. Among the detection of various faults in the transformer (or in any electrical machine in general), detection of winding inter-turn fault is critical since its effect is not easily detectable at lower magnitudes in the signatures of terminal voltages and currents. Several techniques have been reported in the literature for detecting the turn-fault ranging from the classical differential current protection scheme to the modern online Dissolved Gas Analysis (DGA). Even in the presence of these techniques, the Customer has to commission a transformer relay for voltage-current monitoring. This paper proposes a new simple algorithm which can be used for incipient detection of inter-turn fault in transformers. The algorithm does not require any special hardware for the fault detection; it uses the same voltage-current data monitored by the relay which does not include any additional cost. This, easy to code, algorithm is analyzed by developing 2D non-linear Finite Element (FE) model for 10 MVA transformer, and tested by creating faults on 5 kVA lab prototype transformer.