Stochastic simulation of turbulent inflow fields commonly used in wind turbine load computations is unable to account for contrasting states of atmospheric stability. Flow fields in the stable boundary layer, for instance, have characteristics such as enhanced wind speed and directional shear; these effects can influence loads on utility-scale wind turbines. To investigate these influences, we use large-eddy simulation (LES) to generate an extensive database of high-resolution ( ∼ 10 m), four-dimensional turbulent flow fields. Key atmospheric conditions (e.g., geostrophic wind) and surface conditions (e.g., aerodynamic roughness length) are systematically varied to generate a diverse range of physically realizable atmospheric stabilities. We show that turbine-scale variables (e.g., hub height wind speed, standard deviation of the longitudinal wind speed, wind speed shear, wind directional shear and Richardson number) are strongly interrelated. Thus, we strongly advocate that these variables should not be prescribed as independent degrees of freedom in any synthetic turbulent inflow generator but rather that any turbulence generation procedure should be able to bring about realistic sets of such physically realizable sets of turbine-scale flow variables. We demonstrate the utility of our LES-generated database in estimation of loads on a 5-MW wind turbine model. More importantly, we identify specific turbine-scale flow variables that are responsible for large turbine loads—e.g., wind speed shear is found to have a greater influence on out-of-plane blade bending moments for the turbine studied compared with its influence on other loads such as the tower-top yaw moment and the fore-aft tower base moment.
Overall, our study suggests that LES may be effectively used to model inflow fields, to study characteristics of flow fields under various atmospheric stability conditions and to assess turbine loads for conditions that are not typically examined in design standards. Copyright © 2013 John Wiley & Sons, Ltd.