Wind power forecasting models for very short-term operation of power systems

The integration of intermittent and volatile wind power into the electric grid poses different challenges to grid operators in the planning and operation of electric power systems. In particular, in case of system-wide oversupply or local transmission constraints, the grid operator could reduce or restrict energy production from renewable generation plants for some periods, lasting minutes to hours depending on the meteorological condition and corresponding loading of the power system. In this context, this paper presents a model based on an artificial neural network approach for wind power nowcasting based on real-time measurements data exchanged between the wind energy producers and the Italian transmission system operator. The developed model can be a valuable aid for the system operator and can be integrated into future tools designed to support grid operators for the real-time management of the wind generators during the curtailments, for having greater control of the wind parks when returning to service. A real case example is used to show the usefulness and effectiveness of the developed methodology.