GENERATION EXPANSION PLANNING USING BENDERS' DECOMPOSITION AND GENERALIZED NETWORKS

This paper presents an optimization model and its application to a generation expansion planning problem. The proposed model has a generalized network structure and is exploited effectively by Benders' decomposition algorithm, where a master problem generates trial expansion plans and a set of subproblems compute production cost and system reliability for the trial plan. The applicability of our decomposition algorithm is demonstrated in the case study of Korea's generation expansion planning. The results demonstrate that the model is a practical and flexible tool in solving realistic long-range generation planning problems.