Multi-area decentralized optimal VAr planning using the Dantzig-Wolfe decomposition principle

Abstract

This paper presents a new methodology for solving the optimal VAr planning problem in multi-area electric power systems, using the Dantzig-Wolfe decomposition. The original multi-area problem is decomposed into subproblems (one for each area) and a master problem (coordinator). The solution of the VAr planning problem in each area is based on the application of successive linear programming, and the coordination scheme is based on the reactive power marginal costs in the border bus. The aim of the model is to provide coordinated mechanisms to carry out the VAr planning studies maximizing autonomy and confidentiality for each area, assuring global economy to the whole system. Using the mathematical model and computational implementation of the proposed methodology, numerical results are presented for two interconnected systems, each of them composed of three equal subsystems formed by IEEE30 and IEEE118 test systems. © 2011 IEEE.