Fault Identification for Transformer Axial Winding Displacement Using Nanosecond IFRA and SFRA Experiments

Nanosecond IFRA has the potential to realize online detection of power transformer winding deformation and displacement. But this method is not mature even in offline condition, and the reliability, accuracy and repeat ability of which are in doubt. To verify the reliability and accuracy of nanosecond IFRA a comparison experiment is conducted between it and a mature method which is applied widely all over the world, SFRA. In this experiment, three levels of axial displacements are experimentally simulated on a dry-type distribution transformer in the lab and both a nanosecond IFRA system and a commercial SFRA analyser are applied to obtain the frequency responses at each fault level separately. To verify the repeatability of Nanosecond IFRA, two Nanosecond IFRA measurements conducted in a time interval of 30 days with all the test condition remaining the same are compared. The results are analysed by visual inspection and quantitative comparison. (paper)