Designing Strategies for Optimal Spatial Distribution of Wind Power

One of the common problems with the production of wind energy is the increased uncertainty surrounding the delivered output. Wind power producers adopt various strategies to reduce production risk, one of the most popular being the dispersion of wind farms across distant locations. The purpose of this paper is to examine the effectiveness of different methodologies for deriving optimal capacity allocation schemes. We collect a large dataset of wind-speed measurements for various locations across the Netherlands and use copula techniques to capture seasonal variations and cross-site dependencies in wind profiles. We exploit the information provided by statistical models to determine optimal resource allocation plans assuming a variety of operational and investment objectives. Experimental results designate the importance of accurately modelling site dependencies when deciding how to optimally allocate the overall wind capacity. Carefully chosen subgroups of sites can result in a more reliable energy supply as opposed to other allocation schemes that ignore spatial correlations