Abstract

Minimisation of variability of energy delivered from a group of wind plants into the power system using portfolio theory approach was studied. One of the assumptions of that theory is Gaussian distribution of the sample, which is not satisfied in case of wind generation. Therefore, optimisation of a “portfolio” of plants with different goal functions was studied. It was supposed that a decision on distribution of a fixed amount of generation capacity to be installed among a set of geographical locations with known wind statistics is to be made with minimised variability of generation as a goal. In that way the statistical cancellation of variability would be used in the best possible manner. This article is a brief report on results of such an investigation. An example of nine locations in Croatia was used. These locations’ wind statistics are known from historic generation data.