Hydrothermal Scheduling Using Benders Decomposition: Accelerating Techniques

A new decomposition method is presented that includes the network through ac modeling within the hydrothermal scheduling optimization process including the losses. In short-term hydrothermal scheduling, the transmission network is typically modeled with dc power flow techniques. Such modeling, however, can lead to impractical solutions when it is verified with ac power flow. Another proposal considers in thermal systems the ac network modeling but not the optimization of losses. The approach presented here addresses issues such as congestion management and control of service quality that often arise in large and weakly meshed networks - the typical pattern of power systems in Latin America. Generalized Benders decomposition and traditional, well-known optimization techniques are used to solve this problem. The master problem stage defines the generation levels by regarding the inter-temporal constraints, whereas the subproblem stage determines both the active and the reactive economical dispatches for each time interval of the load curve. It meets the electrical constraints through a modified ac optimal power flow (OPF). Another important contribution is the inclusion of accelerating techniques aimed at reducing the number of iterations and CPU time. The methodology was proven in a real system and test systems. Results are discussed in this paper.