Transformer Inter-Turn Faults Detection by Dynamic State Estimation Method
Rui Fan,A. P. Sakis Meliopoulos,Liangyi Sun,Zhenyu Tan,Yu Liu
DOI: https://doi.org/10.1109/naps.2016.7747999
2016-01-01
Abstract:Transformers are protected by relays to ensure the overall reliability of the generation, transmission, and distribution electric systems. Among all kinds of transformer faults, the inter-turn faults are considered to be the most difficult faults to be detected. Inter-turn faults are extremely harmful to power transformers. When only few turns are shorted at the windings, the currents in the shorted loop can be very high while the transformer terminal currents change little. It is very difficult for the commercial relays or legacy protection methods to detect the existence of inter-turn faults. In this paper, the dynamic state estimation (DSE) based protection scheme is used to detect inter-turn faults. This method monitors the transformer terminal measurements, and it feeds the measurement data and transformer model to a dynamic state estimator. A chi-square test is performed to evaluate the probability (confidence-level) that the transformer model matches measurement data. Any mismatch between the model and measurement indicates something wrong inside the transformer, and protection action should be taken. Numerical simulations are conducted on a 115/10kV three-phase transformer with a 1% inter-turn fault. Three protection methods are compared, namely the percentage-differential scheme, negative-sequence differential scheme and the proposed DSE-based scheme. Analysis shows that the DSE-based protection is more sensitive and reliable than the other two methods, and in general provides faster protection speed.