Time–Frequency Decomposition Based Protection Scheme for Power Transformer— Simulation and Experimental Validation

A new protection scheme for power transformers using time–frequency decomposition is presented. On detecting an abnormal condition based on the average energy of both sides of the transformer derived from the significant intrinsic mode function (IMF) of each phase current, the sum of the mean absolute value (SMA) of the normalized significant IMF of each phase current on both sides of the power transformer is calculated. The calculated value of SMA acts as a discriminating factor between an internal fault and an abnormal condition. The performance of the proposed technique is verified through simulation (by modeling an existing Indian power system network in PSCAD/EMTDC software) as well as a laboratory prototype along with real field data (COMTRADE file from a substation relay) by generating/gathering various cases of internal faults and abnormal operating conditions, like inrush current, overexcitation, external faults, and saturation of CT during heavy through faults. The results indicate that the proposed technique is not only capable of detecting all internal faults but also remains inoperative during various external events, including noise, errors in CTs, and disturbances that occur during measurements. Moreover, relay response time is improved compared to the prevailing methods, which confirms the superiority of the proposed scheme.