Experimental Data-Based Transient-Stationary Current Model for Inter-Turn Fault Diagnostics in a Transformer

Inrush current has been well-studied in transformers using different methods to diagnose internal faults. In this paper, a new methodology is proposed using experimental data for obtaining a time-model based on inrush currents. The main contribution of this paper consists of providing a model that accurately reproduces the inrush and steady state currents of single-phase transformers, offering the unique ability to analyze several severity levels of turn to turn faults. The obtained model considers eleven cases of inter-turn fault (2, 3, 4, 5, 10, 15, 20, 25, 30, 35, and 40 short-circuited turns) and allows the study of indices to detect inter-turn faults. The proposed model is based on two functions: a steady-state function based on Fourier series and a residual transient function based on a sum of Gaussian functions. The model allows the study of harmonic content because of its direct time-frequency representation. The model is developed using experimental signals measured from a single-phase transformer, which is validated by means of error criteria that assesses experimental and modeled signals. Its capability to reproduce non modeled signals with different inter-turn fault cases is also evaluated. The modeling of these signals can be a useful tool for simulation-based applications where a reliable waveform reproduction is needed. On the other hand, indices such as total harmonic distortion, energy spectral density and second harmonic ratio are obtained from the proposed model to study its behavior, depending on the severity of the inter-turn faults. Results demonstrate its usefulness to characterize the level of fault severity. (C) 2017 Elsevier B.V. All rights reserved.