An Exponential Smoothing Based Power Swing Detection Technique for Distance Protection

Power system network is more susceptible to grid disturbances due to its frequent operation in stressed conditions to meet increase in energy demand. This may lead to power swings due to predominant changes in power flow measurements. If such disturbances are not distinguished from fault, then the distance relays may picks up and trip the healthy transmission lines. To mitigate this, it is essential to introduce an algorithm in numerical relay to block its operation during swing. If a fault occurs in a line which is under influence of swing, the relay must sense and trip the concerned breaker instantaneously to disconnect faulty line from the network. This paper introduces an exponential smoothing prediction based power swing detection technique to estimate impedance seen by the distance relay during power swing. The actual and estimated impedance values are examined to segregate a fault from swing in the transmission system. The performance of the model is evaluated statistically by measuring forecasting accuracy and model validation. The proposed method is tested for a 400 KV two-area power system network and compared with available methods. The method can initiate the blocking correctly during power swing and able to identify faults during power swing.