Wind energy, together with other renewable energy sources, are expected to grow substantially in the coming decades and play a key role in mitigating climate change and achieving energy sustainability. One of the main challenges in optimizing the design, operation, control, and grid integration of wind farms is the prediction of their performance, owing to the complex multiscale two-way interactions between wind farms and the turbulent atmospheric boundary layer (ABL). From a fluid mechanical perspective, these interactions are complicated by the high Reynolds number of the ABL flow, its inherent unsteadiness due to the diurnal cycle and synoptic-forcing variability, the ubiquitous nature of thermal effects, and the heterogeneity of the terrain. Particularly important is the effect of ABL turbulence on wind-turbine wake flows and their superposition, as they are responsible for considerable turbine power losses and fatigue loads in wind farms. These flow interactions affect, in turn, the structure of the ABL and the turbulent fluxes of momentum and scalars. This review summarizes recent experimental, computational, and theoretical research efforts that have contributed to improving our understanding and ability to predict the interactions of ABL flow with wind turbines and wind farms.

This work is dedicated to systematically studying and predicting the wake characteristics of a yawed wind turbine immersed in a turbulent boundary layer. To achieve this goal, wind tunnel experiments were performed to characterize the wake of a horizontal-axis wind turbine model. A high-resolution stereoscopic particle image velocimetry system was used to measure the three velocity components in the turbine wake under different yaw angles and tip-speed ratios. Moreover, power and thrust measurements were carried out to analyse the performance of the wind turbine. These detailed wind tunnel measurements were then used to perform a budget study of the continuity and Reynolds-averaged Navier–Stokes equations for the wake of a yawed turbine. This theoretical analysis revealed some notable features of the wakes of yawed turbines, such as the asymmetric distribution of the wake skew angle with respect to the wake centre. Under highly yawed conditions, the formation of a counter-rotating vortex pair in the wake cross-section as well as the vertical displacement of the wake centre were shown and analysed. Finally, this study enabled us to develop general governing equations upon which a simple and computationally inexpensive analytical model was built. The proposed model aims at predicting the wake deflection and the far-wake velocity distribution for yawed turbines. Comparisons of model predictions with the wind tunnel measurements show that this simple model can acceptably predict the velocity distribution in the far wake of a yawed turbine. Apart from the ability of the model to predict wake flows in yawed conditions, it can provide valuable physical insight on the behaviour of turbine wakes in this complex situation.

Yaw angle control is known nowadays as a promising and effective technique to mitigate wake effects in wind farms. In this paper, we perform wind tunnel experiments to study the performance of a model wind farm with five turbine rows under a wide variety of yaw angle distributions. Electrical servo controllers are used to monitor and control the operating conditions of each model wind turbine, which consists of a recently developed, highly efficient rotor with a diameter of 15 cm. Each turbine is used as a sensor to detect its own inflow conditions. Using this method ensures us that all the turbines within the wind farm always operate with an optimal rotational velocity, regardless of their yaw angles or inflow conditions. Wind farm power measurements are carried out for more than 200 cases with different yaw angle distributions. Our results show that yaw angle control can increase the overall wind farm efficiency as much as 17% with respect to fully non-yawed conditions. Special emphasis is placed on studying yaw angle distributions with different levels of simplicity and power improvement. Among different yaw angle distributions, the most successful ones are those with a relatively large yaw angle value for the first turbine row, and then, the yaw angle decreases progressively for downwind rows until it eventually becomes zero for the last one. In addition, power measurements show that yaw angle control can improve the wind farm efficiency more noticeably for a larger number of turbine rows although this improvement is expected to reach a plateau after several rows.

Comprehensive wind tunnel experiments were carried out to study the interaction of a turbulent boundary layer with a wind turbine operating under different tip-speed ratios and yaw angles. Force and power measurements were performed to characterize the variation of thrust force (both magnitude and direction) and generated power of the wind turbine under different operating conditions. Moreover, flow measurements, collected using high-resolution particle-image velocimetry as well as hot-wire anemometry, were employed to systematically study the flow in the upwind, near-wake, and far-wake regions. These measurements provide new insights into the effect of turbine operating conditions on flow characteristics in these regions. For the upwind region, the results show a strong lateral asymmetry under yawed conditions. For the near-wake region, the evolution of tip and root vortices was studied with the use of both instantaneous and phase-averaged vorticity fields. The results suggest that the vortex breakdown position cannot be determined based on phase-averaged statistics, particularly for tip vortices under turbulent inflow conditions. Moreover, the measurements in the near-wake region indicate a complex velocity distribution with a speed-up region in the wake center, especially for higher tip-speed ratios. In order to elucidate the meandering tendency of far wakes, particular focus was placed on studying the characteristics of large turbulent structures in the boundary layer and their interaction with wind turbines. Although these structures are elongated in the streamwise direction, their cross sections are found to have a size comparable to the rotor area, so that they can be affected by the presence of the turbine. In addition, the study of spatial coherence in turbine wakes reveals that any statistics based on streamwise velocity fluctuations cannot provide reliable information about the size of large turbulent structures in turbine wakes due to the effect of wake meandering. The results also suggest that the magnitude of wake meandering does not depend on turbine-operating conditions. Finally, the suitability of the proper orthogonal decomposition for studying wake meandering is examined.

This paper solves an approximate form of conservation of mass and momentum for a turbine in a wind farm array. The solution is a fairly simple explicit relationship that predicts the streamwise velocity distribution within a wind farm with an arbitrary layout. As this model is obtained by solving flow-governing equations directly for a turbine that is subject to upwind turbine wakes, no ad hoc superposition technique is needed to predict wind farm flows. A suite of large-eddy simulations (LES) of wind farm arrays is used to examine self-similarity as well as validity of the so-called conservation of momentum deficit for turbine wakes in wind farms. The simulations are performed with and without the presence of some specific turbines in the wind farm. This allows us to systematically study some of the assumptions made to develop the analytical model. A modified version of the conservation of momentum deficit is also proposed to provide slightly better results at short downwind distances, as well as in the far wake of turbines deep inside a wind farm. Model predictions are validated against the LES data for turbines in both full-wake and partial-wake conditions. While our results highlight the limitation in capturing the flow speed-up between adjacent turbine columns, the model is overall able to acceptably predict flow distributions for a moderately sized wind farm. Finally, the paper employs the new model to provide insights on the accuracy of common wake superposition methods.