A method for predictive out-of-step tripping based on synchrophasors

This paper presents a method for predictive out-of-step (OOS) tripping based on synchrophasors. The OOS condition is predicted based on information of the ongoing power oscillation, where the input signal can be either the phase angle between two nodes in the power grid or the current magnitude of a line. The proposed algorithm first detects power swing by comparing the oscillation magnitude with a set threshold. Once a power swing is detected, a damping coefficient associated to the power oscillation is estimated to assess its stability. The oscillation is predicted unstable and eventually will lead to OOS condition when both the damping coefficient and the input signal are larger than set thresholds. The tuning of these thresholds is relatively simple and straightforward. Results from simulations in the Kundur system and the Nordic 44 model demonstrate that the proposed algorithm is able to correctly predict coming OOS conditions from the analysis of power swings in the power grid.