Optimization models for the economic design of wind power systems

Abstract This paper presents optimization models for the economic design of large scale wind power systems. Two basic models are introduced: (1) Systems without storage—all power generated is fed directly into the existing network, and (2) systems with storage—the systems are operated as part of base load or peak load capacity. The objective of the models is to maximize the total net value of power generated under assumed operating rules and general conditions regarding wind speed and demand variations. The model for wind power systems without storage is developed to determine the total capacity of wind turbines that vary in the values of design parameters such as rotor diameter, tower height, and rated power. The model for systems with storage uses an analytical storage model as a basis of representing storage requirements. The model is designed for the case of little serial correlation in successively measured wind speeds. It is then modified for the case of a low degree of correlation. Separable programming is used as a solution technique in both models and limited computational results, based on available cost estimates and Oklahoma wind and demand data, are presented to illustrate the use of the models. In the first model, separable programming will yield a globally optimal solution for certain types of cost functions used. However, in the second model, the problem structure is such that a global optimum cannot be guaranteed.