Maximally secure mitigation of protection misoperations in power systems

Protection misoperations have been recognized as a main culprit in causing cascading failures in modern power systems. This paper introduces the processes of misoperations and their mitigation into a scalable stochastic discrete-state model. In this framework, a mitigation strategy is developed to minimize the risk of cascading failures. The risk is quantified by a set of security indices that formally incorporate the uncertain knowledge of the continuous-state of rotor angles/speed deviations and the discrete-state of equipment faults. The security indices serve as mitigation criteria for selection of protection control actions. Real-time security evaluation poses a significant challenge because of the fraction-of-a-second response time required of protection functions. Two problems addressing this challenge are formulated and solved through a novel application of the phasor measurement units (PMUs)?their input samples in time-domain, rather than their output samples in phasor-domain, are used as redundant feedback variables. These time-stamped input waveform samples are processed to isolate transmission line faults in the presence protection false trips, and track the generators´ rotor angles and speed deviations without invoking the swing dynamics. The computational issues and results are explained through a WSCC 3-generator 9-bus test system.