Abstract

This paper proposes a methodology for optimal allocation of a set of wind farms in order to mitigate its daily power variability, using daily mean standard deviation of the wind generation as metric. A version of the Mean Variance Portfolio Theory was developed in order to distribute a certain amount of rated capacity over several wind farms, minimizing global wind energy variability and achieving fixed mean power levels. This methodology was applied to allocate wind generation in the National Electric System (SEN, in Spanish) by 2030. From a Markowitz curve resulting from an optimization process, four wind power distribution scenarios were selected and compared with a base scenario which was built without using the proposed methodology. Different indicators were used to compare the performance of each scenario. The results show that it’s possible to obtain expansion plans with the same rated capacity, higher energy production and lower variability than the base scenario. Finally, a Unit Commitment model applied to the SEN was used to evaluate the feasibility of the resulting scenarios.