A method to compute flexible static equivalents for very large power networks

This paper presents an effective methodology for power system network reduction. The simulation of very large systems when the focus is restricted to a relatively small portion of it is cumbersome and unnecessarily inefficient. The methodology used is based on the Ward approach. The concept of buffer zone in which external components in a neighborhood of the internal area are retained is also implemented. Automatic and manual means of defining components in this zone are provided. Accuracy tests based on power flow contingency analysis are included, in which post-contingency voltages and power flows for the original and reduced networks are compared. An academic system and two large real networks are used to test the performance of the proposed method.