Validation of European wind generation time series simulation: Importance of wakes, micro-scale adjustments and stochastic simulations

Share of wind generation has been increasing in the energy mix of European energy system and is expected to increase even further. In order to plan such a future energy system where wind energy becomes a prominent part of the energy mix, energy system studies require accurate time-series simulation of wind generation. In this paper we introduce and validate European scale wind generation time series simulation models and suggest the most suitable model for both onshore and offshore wind generation simulation. The model chain consists of adjustment of ERA5 reanalysis wind speeds with different microscale wind datasets and stochastic fluctuations, generation of plant-level power curves with detailed wake modeling and simulation of stochastic unavailability. The models are validated and compared using hourly-resolution measured wind generation data on country- and plant-level. The results indicate that the model where reanalysis wind speeds are adjusted based on Global Wind Atlas ver. 2 microscale data and wake modeling is carried out with appropriate turbulence intensity provides the most accurate results. Stochastic fluctuation and unavailability modeling are suggested when detailed modeling of generation ramps and distribution of high generation values are of importance. • A novel plant-level generic power curve model, which considers detailed wake losses and turbulence intensity. • Comparison of multiple microscale datasets for wind-speed adjustments for continental-scale wind generation simulation. • Inclusion of stochastic models for unavailability and wind speed fluctuations to better represent measured generation data. • Validation with country- and plant-level onshore and offshore wind generation measured data covering 13 countries and 14 wind power plants.