Design and Development of Digital Protection for Power Transformer

Abstract

newline Power Transformer is one of the essential components in electrical power network and hence its protection is equally important. Existing methodologies to protect the transformer from disturbances provide fine protection to transformer. However, these existing methodologies may mal-operate during other external disturbances. Hence, further research can be made in terms of more accuracy, speedy operation and obviously lower protection cost with zero error in detection of faults. This work presents various methodologies for detection and estimation of severity of over-fluxing condition, inrush conditions, external fault with and without CT saturation and cross country fault against the internal fault based on various types of classifier and adaptive digital techniques. Algorithms are prepared to perform step by step procedure for filtration of internal fault condition from various anomalies. The proposed algorithms are successfully validated on PSCAD software and after collection of data, MATLAB software is used for further validation. Further, to prove the competency of the proposed schemes, hardware experiment is performed on 50kVA, 440/220 V transformer. After analyzing variety of practical, it is noted that the presented methods provides promising classification accuracy within a short and acceptable time period. This PhD Thesis would be usefulfor protection engineers and researchers who are engage in the field of protection of transformer. newline