Coherency identification for large electric power systems

An important problem associated with contingency and reliability studies for large-scale electric power systems is the use of dynamic equivalent models to represent the behavior of aggregable parts of the system. Fundamental to the aggregation process associated with any coherency-based approach is an ability to identify groups of coherent generators or generators that swing together in response to a contingency. Coherency identification is thus a key step in the reduction process. In this paper, an analytical process is examined which permits identification of coherent groups based upon the closed-form solution of a linearized power system model and the use of the Cayley-Hamilton theorem. The result is an algebraic characterization of coherency. Computationally practical coherency indices are developed which do not require significant computer storage. A physical interpretation of this characterization is discussed which suggests that coherency depends not only of the electrical distances between generators but also on generator inertia constants. The method of coherency identification developed in this paper is evaluated on a detailed model of a 39-bus test system.