An optimised methodology to predict the wind farms production

Renewable energy sources (RES), are widely recognized as an effective solution to face with the rapid depletion of oil resources and an interesting energy option to ensure the energy supplies security and the meeting of the targets imposed by the international regulatory frameworks devoted to contrast the global warming. Wind energy source is undoubtedly one of the primary energy source able to provide a significant contribution to the fulfillment of the load curve. Nonetheless, wind energy source is one of the most challenging to predict due to the (i) high level of uncertainty and randomness of the physical phenomenon, (ii) strong dependence on the site topography, (iii) high non-linearity of the involved processes. The massive penetration of wind farms in the existing electrical power systems hence sensibly affects the related secure operation. To provide a solution at the above issue, the present paper aims at the development of an advanced methodology for estimating the wind farms producibility and for characterizing locally the predictions of European Center for Medium-Range Weather Forecasts (ECMWF) model or equivalently the ones of the Consortium for Small-scale Modeling (COSMO). The ingredients which makes the developed methodology optimal are high resolution digital terrain models, proper lateral boundary conditions provided by COSMO-I2 model, optimized wind generation curves derived by the application of statistical identification techniques on wind speed-power. The research activities are included in a research project with the partnership of the Italian TSO, TERNA, and the Italian Center for Aereospatial Research, CIRA.