loading page

Sensitivity of Multiscale Large Eddy Simulations for Wind Power Production in Complex Terrain
  • +1
  • Giorgia De Moliner,
  • Paolo Giani,
  • Giovanni Lonati,
  • Paola Crippa
Giorgia De Moliner
Politecnico di Milano Dipartimento di Ingegneria Civile e Ambientale
Author Profile
Paolo Giani
University of Notre Dame College of Engineering Department of Civil & Environmental Engineering & Earth Sciences
Author Profile
Giovanni Lonati
Politecnico di Milano Dipartimento di Ingegneria Civile e Ambientale
Author Profile
Paola Crippa
University of Notre Dame College of Engineering Department of Civil & Environmental Engineering & Earth Sciences

Corresponding Author:[email protected]

Author Profile

Abstract

Coupling Large Eddy Simulations (LES) with Numerical Weather Prediction (NWP) models has promising applications for wind energy. Regional climatology, optimal siting of wind turbines as well as short term wind energy forecasts can be improved by considering all the energetic scales of atmospheric motions. However, the complexity of NWP-LES coupled simulations introduces challenges and uncertainties that need to be addressed. This study focuses on understanding the relative importance of different factors and assumptions in NWP-LES calculations for wind energy applications. Using a recent large ensemble of LES simulations driven by NWP boundary conditions over the complex terrain of the Perdigão area, our analysis reveals significant discrepancies in wind power estimates across ensemble members, particularly over hilltops. Depending on the model configuration and the coupling technique, instantaneous predictions can be as sensitive as 800kW for a 2MW wind turbine, in terms of ensemble standard deviation. On multi-day time averages, the model sensitivity is in the order of 150kW. We further analyze the main factors that lead to the observed model sensitivity. Results from a four-way ANOVA analysis identify topography and land use datasets as the primary drivers of variability, for time averaged estimates. Temporal analysis shows strong inter-daily variability and the importance of turbulence modeling and the coupling techniques for instantaneous predictions. Overall, most of the sensitivity is observed during day-to-night and synoptic transitions. By understanding the relative importance of different factors, future model development and applications can be guided to enhance the accuracy and reliability of wind energy assessments.