A Wide-Area Scheme for Power System Fault Location Incorporating Bad Data Detection

Phasor measurements have found many applications in modern power systems. Wide-area fault location utilizing the bus-impedance matrix and available measurements is a recent achievement in the context of smart transmission grid. This paper aims at developing a wide-area fault-location scheme capable of detecting and identifying erroneous measurements. As a result, while multiple available measurements contribute to fault location in order to increase accuracy, the reliability of fault location is not compromised as the proposed scheme detects, identifies, and removes erroneous measurements. The transfer impedances between the fault location and the phasor measurements, which can be remote from the faulted line, are utilized to formulate fault location as a nonlinear fitting problem. Having estimated the fault location and total fault current, we employ statistical hypotheses testing to detect, identify, and remove bad data. In the case of unsynchronized measurements, the fitting problem is modified to include only the magnitude of measurements. The proposed method considers the distributed parameter model of the transmission lines and, therefore, the obtained results are also valid for long transmission lines. Electromagnetic transient simulations for the WSCC 9-bus and a 41-bus subnetwork of the IEEE 118-bus test system reveal successful fault location by synchronized as well as unsynchronized measurements.