Stochastic modeling of future wind generation scenarios

This work uses historical data from existing wind generation facilities as a sample of a larger future population of wind generation to create realistic long term time series of future wind generation scenarios. Models are created that reasonably capture both the deterministic and random characteristics of wind generation that are locally unique. Observed correlations in outputs between existing wind generators are used to create scenarios of future wind generation that can estimate the impact of greater diversity of wind generation on overall wind generation variability. Publically available historical data for ten wind generation facilities in Alberta are used to illustrate the methods proposed.