WINDPOWER 2007 Conference and Exhibition, Los Angeles, CA 1 OPTIMIZATION ALGORITHMS FOR OFFSHORE WIND FARM MICROSITING C.N. Elkinton*, J.F. Manwell, J.G. McGowan Renewable Energy Research Laboratory Dept. of Mechanical and Industrial Engineering University of Massachusetts 160 Governors Dr., Amherst, MA 01003 * celkinto@ecs.umass.edu ABSTRACT In the United States, offshore wind energy is poised to facilitate substantial growth in wind energy production. Unlike most onshore projects, this growth has the potential to occur in close proximity to large load centers (New York, Boston, Houston, for example). In order for offshore wind to be able to compete with other energy generating technologies, however, further reductions in the cost of energy are required. Making optimal use of current technology is one simple approach to this problem. As part of a larger project focused on offshore wind farm analysis and optimization, this research examines the use of optimization algorithms for wind farm micrositing. The paper starts with a discussion of five different types of optimization algorithms: gradient search, heuristic, pattern search, simulated annealing, and evolutionary algorithms. The relevance of each algorithm to wind turbine micrositing is evaluated by considering two separate objectives: minimization of the levelized production cost and maximization of the energy production. Two algorithms, the genetic and greedy heuristic, are further evaluated for the specific case of offshore wind farm design through the use of design simulations. In these simulations, a full set of site conditions are considered, including as water depth, soil conditions, wind climate, and distance from shore. In addition, comparisons are made with previous studies in the literature. Finally, these algorithms are employed to optimize the layout of a potential, real-world offshore wind farm near Hull, Massachusetts. This process, results, and lessons learned are discussed. 1. OVERVIEW The problem of optimizing wind farm layouts falls into the class of problems called combinatorial optimization. Several papers and books have been written about this class of problem, including multiple papers related to wind farm layout. All of these papers have been concerned with onshore wind farm design, but provide a good starting place and sources of comparison for this investigation of offshore wind farm layouts [1, 2, 3, 4]. The objective of this study is to investigate the use of optimization algorithms for offshore wind farm micrositing. Five optimization algorithms are considered and the applicability of each to wind farm micrositing is evaluated. The two most promising algorithms are evaluated further. These two algorithms have been incorporated into the offshore wind farm layout optimization (OWFLO) software [5] and are used to lay out farms. Comparisons are made with optimization results from previous studies in the literature.