Online Transformer Internal Fault Detection Based on Instantaneous Voltage and Current Measurements Considering Impact of Harmonics

This paper investigates the performance of a recently proposed online transformer internal fault detection technique and examines the impact of harmonics through detailed nonlinear simulation of a transformer using 3-D finite-element modelling. The proposed online technique is based on considering the correlation between the instantaneous input and output voltage difference (ΔV) and the input current of a particular phase as a fingerprint of the transformer that could be measured every cycle to identify any incipient mechanical deformation within power transformers. To precisely emulate real transformer operation under various winding mechanical deformations, a detailed 3-D finite-element model is developed. Detailed simulations with (non)sinusoidal excitation are performed and analyzed to demonstrate the unique impact of each fault on the ΔV-I locus. Impacts on harmonic order, magnitude, and phase angle are also investigated. Furthermore, practical measurements have been performed to validate the effect of a winding short-circuit fault on the proposed ΔV-I locus without and with the impact of system harmonics.