Dynamic State Estimation Assisted Out-of-Step Detection for Generators Using Angular Difference

Abstract

Out-of-step detection methods for generators depend on impedance measurements that are mapped from the angular difference between the generator source and the Thevenin equivalent of the interconnected system. Though implemented successfully with decades of operating experience, the mapping is not the most accurate due to a simplified model. This paper proposes an improved out-of-step detection method for generators by means of the most direct indicator—angular difference. The generator's rotor angle is estimated with a particle filter-based dynamic state estimator and the angular separation is then calculated by combining the raw local phasor measurements with this estimate. The method is also enhanced with the real-time stability analysis that is able to predict marginally unstable swings. The approach is illustrated via dynamic simulations on the New England 10-generator, 39-bus system. While consistent with the traditional schemes for out-of-step detection, the proposed method is advantageous in that it does not make any simplifying assumptions in its formulation, and provides early detection for unstable swings resulting in reduced stresses on the breaker and generator. The proposed approach shows great potential to supervise the conventional relay with availability of synchrophasor measurements and computational power.