Session L30: Boundary Layers: Wind Turbine Interactions

Extending a row-averaged model to a turbine-specific model for wind farm control

This study builds upon a recently proposed model-based receding horizon control approach that enables wind farms to follow a reference power signal. In particular, we extend the underlying wake model to enable control of individual turbines in order to both generalize the approach to arbitrary wind farm configurations and account for spatially heterogeneous inflow conditions. We also develop the associated estimation techniques to account for the spatially varying inflow conditions within the control loop. The additional control authority introduced through the individual turbine control has the potential to improve the performance of the control over a wide range of wind conditions. Results demonstrate that accounting for the local wind farm conditions leads to improved estimates of the flow field and the ability to reproduce the transient behavior of the flow field in irregularly arranged wind turbine arrays. Finally, the trade-offs between the improvements in power tracking and estimation accuracy and the increased computational complexity of using the individual turbine versus the aggregate row approach are evaluated.   

Layout dependence on spatio-temporal correlations between turbine outputs in large wind plants

Turbulent boundary layers in which wind turbines operate are characterized by a strong correlation of the velocity field in the streamwise direction. Practically, a significant correlation allows a wind farm controller to use information from upstream turbines to estimate incoming flow conditions for downstream ones. Moreover, the spatio-temporal correlation of the flow is directly linked to the resulting fluctuations of the aggregate wind plant power.  It is evident that the interaction of a turbulent boundary layer with a cluster of large-scale turbines, and their low momentum wakes, is expected to influence the turbulent structure in the wind plant. Here, an existing wind tunnel data-set of a scaled wind farm with sixty instrumented porous disks, and one-hundred models in total (Bossuyt, J. et al. 2017, Exp in Fluids, 58:1), is used to study the effect of layout and wind plant length, on time-lag and correlation of power outputs. The data-set consists of fifty-six different layouts, covering both classical uniform and unconventional non-homogeneous turbine spacings.
  

A data-driven flow model for wind-farm control based on Koopman mode decomposition of large-eddy simulations

The promise of increasing farm performance through turbine control has caused significant interest in wind farm control research in recent years. However, an incompatibility between control model adequacy and computational cost persists up to date: standard engineering models fail to account for relevant nonlinear flow physics on the one hand, yet accurate nonlinear models such as large-eddy simulations (LES) are still too costly for real-time control purposes on the other hand. In the current work, we present a data-driven linear flow model for wind-farm control based on Koopman theory, in which observables of the underlying nonlinear flow dynamics are lifted into a space on which they evolve linearly through the Koopman operator. By approximating this operator using extended dynamic mode decomposition of LES data, we can incorporate nonlinear flow dynamics into a computationally efficient linear model. Performance of the Koopman model is compared to a nonlinear LES for an array of aligned wind turbines. Finally, opportunities and challenges for application in model-predictive control are discussed.    

Turbulence characteristics of wind in complex terrain at the CCSM site

Despite high wind potential sites often lying in complex terrain, few campaigns in complex terrain have both long enough duration and a sufficient number of sites to truly characterize the wind patterns.  The Chokecherry/Sierra Madre (CCSM) development site in south central Wyoming is under development with 1000 turbines to be spread over complex terrain with large variations in elevation (100s of meters) that often occur over small distances.  The site has been equipped with more than 30 meteorological towers of which several have run continuously for more than five years.  The long duration measurements allow for statistical analysis at a level not possible with shorter campaigns, and the number of towers allows for an understanding of the spatial effects of the terrain on the wind.  Using these data, it is possible to identify the unique behavior of the winds at different times of day in different parts of the year providing a better understanding of the effect of the complex terrain on the observed wind patterns.   

Effect of a synthetic low-level jet on the mean power and momentum transport of a model wind-turbine arra

Nocturnal low-level jet (LLJ) is a distinctive phenomenon at the top of stable boundary layers. Distinctive low-level velocity peak results in attractive power resource for wind turbines. However, a maximum in the mean wind speed profile implies the co-existence of positive and negative mean shear in the vicinity of the peak. To gain better understanding of the impacts of LLJ on wind turbines, well-controlled wind tunnel experiments were performed using particle imaging velocimetry (PIV) on model wind turbine arrays including power measurements.
The results from single turbine and 3×2 turbines array in the positive and negative shear regions of the synthetic LLJ were compared to canonic neutrally stratified boundary layer.
We will discuss on the role of energy entrainment induced by the positive and negative shear of the LLJ and the contribution to the vertical transport of momentum and kinetic energy across the  turbine array.   

Humidity variation in the shadow of a large wind farm: an LES investigation

Numerous studies have shown that wind turbine wakes within a large wind farm bring about changes to both the dynamic and thermal properties of the atmospheric boundary layer (ABL). Previously, we compared humidity field measurements in the near-wake region of a large wind turbine with LES results of a single turbine in a stable ABL and found good agreement. The effect of the compounding wakes within a large wind farm on the relative humidity was also investigated using LES. The relative humidity was found to decrease below and increase above the turbine hub height, with the downstream turbines magnifying the effect within the wind farm. This study will investigate how the areas of relative humidity variation, that were observed in the near-wake, develop downstream of a large wind farm. To this end, an LES study of a 7x4 turbine array operating in a stable ABL is carried out. Two wind farm layouts are considered: aligned and staggered. Vertical, streamwise and lateral profiles of the change in relative humidity between upstream and downstream of the turbine array will be presented and discussed.