This thesis contributes to the research on the challenging topic of wind turbine wake development in the atmospheric boundary layer. Through a combined theoretical and numerical approach, the focus is on developing new frameworks to predict the wind turbine wake flow under different inflows and operating conditions and to gain insight into the various aspects of the wake recovery process. In this thesis, four studies are presented to explore this topic from different perspectives.
In the first study, a new physics-based model is proposed and validated to predict the wind turbine wake expansion based on the incoming ambient turbulence and turbine operating conditions. The proposed model ensures mass and momentum conservation in the far wake and considers the effects of the ambient and turbine-induced turbulence on wake expansion. It is found that the predictions of the proposed model are in good agreement with the numerical data for a wide range of incoming turbulence levels. In particular, the proposed model performs well for low ambient turbulence conditions, for which the existing models are unable to account for the contribution of turbine-induced turbulence on the wake expansion.
In the second chapter, a new streamwise scaling is introduced by normalizing the streamwise distance with the near-wake length. With this new scaling, a collapse of the normalized wake velocity deficit profiles for different turbulence levels is obtained under different ambient turbulence levels. Then, the possibility of using the relationship obtained for the normalized maximum wake velocity deficit as a function of the normalized streamwise distance in the context of analytical wake modeling is explored, with a focus on computational efficiency. Overall, a good agreement between the simulation data and the model predictions is observed, along with significant savings in the computational cost of the models.
In the third study, we use large-eddy simulations to study the influence of the thrust coefficient variations on the wind turbine wake. For this purpose, the wake of an actuator disk is simulated using a thrust coefficient varying from 0.4 to 0.9. The simulation results show that the thrust coefficient variation leads to considerable differences in the distributions of the wake velocity deficit and added-turbulence intensity in the downwind distances close to the actuator disk. In addition, the dataset is used to evaluate the performance of several analytical and empirical wake models for different thrust coefficients. Overall, the models with physically-based approaches incorporated within their framework are robust to the thrust coefficient variations and provide more accurate predictions.
Finally, the last study focuses on the effect of inflow turbulent scales on wind turbine wake characteristics. Through a systematic approach and by varying the boundary layer height, a series of large-eddy simulations is generated to have a given rotor-averaged turbulence intensity and different inflow turbulent scales. In the first step, several metrics are used to evaluate the range of inflow turbulent scales in the simulations without the turbine. Next, several wake features are studied, and their sensitivity to variations in the inflow turbulent scales is discussed. Furthermore, a detailed analysis of the wake recovery mechanism under different inflow turbulence scales is presented.