In this study, a hyperbolic S-transform-based method is proposed to discriminate most important transient fault currents of transformers. First, the proposed method discriminates external faults from other disturbances. Then, the S-transform is applied to differential currents. An index is suggested by using absolute deviations of the S-matrix of differential currents to discriminate internal incipient faults in addition to inrush currents and internal faults. The relay issues an alarm signal in the case of the incipient fault but restrains during the magnetising inrush current. If an internal fault is recognised, the relay will issue a trip signal. To study the robustness of the suggested method, a program is developed in the MATLAB environment. The inputs of this program are differential current signals, derived from a system modelled by EMTP software. In order to simulate the internal incipient fault along with internal turn to turn and turn to earth faults, the transformer is modelled as 8 × 8 RL matrices, derived by using subroutine BCTRAN in EMTP. Also, differential currents are contaminated by noise and it is shown that the suggested method is not affected by noise and it can discriminate incipient faults, inrush currents, internal faults and external faults.

No abstract

No abstract

In this paper, the application of the hyperbolic S-transform is proposed for detecting internal incipient faults of transformers during the impulse test. This test is performed on the under consideration transformer and during this test, the input current and applied voltage are recorded. The S-transform is applied to the recorded input current. An index is proposed and computed considering absolute deviations of the S-matrix. Faulty and non-faulty windings are discriminated by using the suggested index. The data obtained from field tests of a 66 kV/25MVA interleaved transformer winding and computer simulations, are recorded. Then, these data are fed to the proposed method, and faulty and normal conditions are discriminated, accordingly. Also, a wavelet-based method is implemented in MATLAB environment and the proposed method is compared with the wavelet-based method. According to simulations and experimental results, the proposed method is superior to wavelet transform-based method.

No abstract