Xiaolei Yang

Computational study and modeling of turbine spacing effects in infinite aligned wind farms

We study the turbine spacing effects in infinite, aligned wind-turbine arrays using large-eddy simulation(LES) with the wind turbine rotors parameterized as actuator disks. A series of simulations is carried out to systematically investigate the different effects of streamwise and spanwise turbine spacings on the array power output and turbulence intensities. We show that for the same turbine density, increasing the streamwise spacing is more beneficial than increasing the spanwise spacing. Larger streamwise turbine spacing increases the power extraction and lowers the turbulence intensity at each turbine more efficiently than when the spanwise turbine spacing is increased. The reason for the different effects of streamwise and spanwise turbine spacings on wind farm performance is that the wake recovery of wind turbines in infinite arrays depends on the area influenced by the wind-turbine wakes, rather than the land area occupied by each turbine. Based on this idea, an improved effective roughness height model is proposed, which can account for the different effects of streamwise and spanwise turbine spacings in infinite aligned wind farms. The predictive capabilities of the new model are demonstrated via extensive comparisons with results obtained from the LES and previously proposed roughness height models.

On the predictive capabilities of LES-actuator disk model in simulating turbulence past wind turbines and farms

In this paper, we validate an LES (large-eddy simulation)-actuator disk model by comparing the computed results with wind tunnel measurements for both stand-alone wind turbine case and wind farm case. In the actuator disk model, the forces are uniformly distributed. The effects of rotation and the effects of nacelle and tower are not taken into account. Dynamic subgrid scale model is employed in the LES. For the stand-alone wind turbine case, grid refinement studies with three different resolutions are performed. Discrepancies with the wind tunnel measurements are observed in the near wake locations (2D and 3D downstream). However, very good agreements with the wind tunnel measurements are obtained for further downstream locations. For the wind farm case, overall good agreements with the wind tunnel measurements are obtained for the streamwise variations of the streamwise mean velocity and the streamwise turbulence intensity at the bottom and top heights of wind turbines. Some discrepancies with the wind tunnel measurements are observed at the hub height of wind turbines.

Effects of a three-dimensional hill on the wake characteristics of a model wind turbine

The spatial evolution of a turbine wake downwind of a three-dimensional sinusoidal hill is studied using large-eddy simulations and wind tunnel measurements. The computed flow fields behind the hill show good agreement with wind tunnel measurements. Three different heights of the hill, i.e., hhill = zh − 0.5D, ≈ zh and =zh + 0.5D (where zh is the turbine hub height and D is the diameter of the turbine rotor), were considered. The effect of the hill turbine spacing was investigated through a comparative analysis with the turbine wake results in the undisturbed turbulent boundary layer. It is observed that the turbine wakes downwind of the hill with hhill ≈ zh and hhill = zh + 0.5D recover faster because of the increased entrainment of ambient flow into the turbine wake, which is due to the enhanced turbulent transport in both spanwise and vertical directions. In comparison with the turbine only case, significant increases in the turbulence kinetic energy (TKE) in the turbine wake are observed for the hill-...

Coherent dynamics in the rotor tip shear layer of utility-scale wind turbines

Recent field experiments conducted in the near-wake (up to 0.5 rotor diameters downwind of the rotor) of a 2.5 MW wind turbine using snow-based super-large-scale particle image velocimetery (SLPIV) (Hong et al., Nature Comm., vol. 5, 2014, no. 4216) were successful in visualizing tip vortex cores as areas devoid of snowflakes. The so-visualized snow voids, however, suggested tip vortex cores of complex shape consisting of circular cores with distinct elongated comet-like tails. We employ large-eddy simulation (LES) to elucidate the structure and dynamics of the complex tip vortices identified experimentally. The LES is shown to reproduce vortex cores in good qualitative agreement with the SLPIV results, essentially capturing all vortex core patterns observed in the field in the tip shear layer. We show that the visualized vortex patterns are the result of energetic coherent dynamics in the rotor tip shear layer driven by interactions between the tip vortices and a second set of counter-rotating spiral vortices intertwined with the tip vortices. We further show that the mean flow within the region where such rich coherent dynamics occur satisfies the instability criterion proposed by Leibovich and Stewartson (J. Fluid Mech., vol. 126, 1983, pp. 335--356), indicating that the instability uncovered by the SLPIV and the LES is of centrifugal type. This study highlights the feasibility of employing snow voids to visualize tip vortices and demonstrates the enormous potential of integrating SLPIV with LES as a powerful tool for gaining novel insights into the wakes of utility scale wind turbines.

On the use of spires for generating inflow conditions with energetic coherent structures in large eddy simulation

ABSTRACT While it has long been a practice to place spires near the inlet of a wind tunnel to quickly develop a turbulent boundary layer with similarities to an atmospheric boundary layer, this has not been the case for creating turbulent boundary layer inflow in large eddy simulations (LESs) of turbulent flows. We carry out LES with the curvilinear immersed boundary method to simulate the flow in a wind tunnel with a series of spires in order to investigate the feasibility of numerically developing inflow conditions from a precursory spire LES and assessing the similarities of the turbulence statistics to those of an atmospheric boundary layer. The simulated mean velocity field demonstrates that a turbulent boundary layer with height equal to the spire height develops very quickly, within five spire heights downstream. The major attribute of using spires for precursory simulations is the spatially evolving coherent structures that form downstream of the spires offering a range of length scales at both the vertical and streamwise directions allowing multiple turbulent inflow conditions to be extracted from a single simulation. While the distribution of length scales far from the spires resembles an atmospheric boundary layer, some turbulence statistics have some significant differences.

Effects of spanwise blade load distribution on wind turbine wake evolution.

In this paper, the effect of two different turbine blade designs on the wake characteristics was investigated using large-eddy simulation with an actuator line model. For the two different designs, the total axial load is nearly the same but the spanwise (radial) distributions are different. The one with higher load near the blade tip is denoted as Design A; the other is Design B. From the computed results, we observed that the velocity deficit from Design B is higher than that from Design A. The intensity of turbulence kinetic energy in the far wake is also higher for Design B. The effect of blade load distribution on the wind turbine axial and tangential induction factors was also investigated.

Wind turbine wake interactions at field scale: An LES study of the SWiFT facility

The University of Minnesota Virtual Wind Simulator (VWiS) code is employed to simulate turbine/atmosphere interactions in the Scaled Wind Farm Technology (SWiFT) facility developed by Sandia National Laboratories in Lubbock, TX, USA. The facility presently consists of three turbines and the simulations consider the case of wind blowing from South such that two turbines are in the free stream and the third turbine in the direct wake of one upstream turbine with separation of 5 rotor diameters. Large-eddy simulation (LES) on two successively finer grids is carried out to examine the sensitivity of the computed solutions to grid refinement. It is found that the details of the break-up of the tip vortices into small-scale turbulence structures can only be resolved on the finer grid. It is also shown that the power coefficient CP of the downwind turbine predicted on the coarse grid is somewhat higher than that obtained on the fine mesh. On the other hand, the rms (root-mean-square) of the CP fluctuations are nearly the same on both grids, although more small-scale turbulence structures are resolved upwind of the downwind turbine on the finer grid.

Wake meandering of a model wind turbine operating in two different regimes

The flows behind a model wind turbine under two turbine operating regimes are investigated using wind tunnel experiments and large-eddy simulations. Measurements from the model wind turbine experiment reveal that the power coefficient and turbine wake are affected by the operating regime. Simulations employing a new class of actuator surface methods which parameterize both the turbine blades and nacelle with and without a nacelle model are carried out for each operating condition to study the influence of the operating regime and nacelle. For simulations with a nacelle model, the mean flow field is composed of an outer wake, caused by energy extraction from the incoming wind by the turbine blades, and an inner wake directly behind the nacelle, while for the simulations without a nacelle model, the central region of the wake is occupied by a jet. The simulations with the nacelle model reveal an unstable helical hub vortex expanding outwards towards the outer wake; while the simulations without a nacelle model show a stable and columnar hub vortex. The hub vortex for the turbine operating in Region 3 remains in a tight helical spiral and intercepts the outer wake a few diameters further downstream than for the turbine operating in Region 2. Wake meandering commences in a region of high turbulence intensity for all simulations indicating that neither a nacelle model nor an unstable hub vortex is a necessary requirement for wake meandering. However, further analysis of the wake meandering using a filtering technique and dynamic mode decomposition show that the unstable hub vortex energizes the wake meandering. The turbine operating regime affects the shape and expansion of the hub vortex altering the location of the onset of the wake meandering and its oscillating intensity. The unstable hub vortex promotes a energetic meandering which cannot be predicted without a nacelle model.

Modeling the effect of control on the wake of a utility-scale turbine via large-eddy simulation

A model of the University of Minnesota EOLOS research turbine (Clipper Liberty C96) is developed, integrating the C96 torque control law with a high fidelity actuator line large- eddy simulation (LES) model. Good agreement with the blade element momentum theory is obtained for the power coefficient curve under uniform inflow. Three different cases, fixed rotor rotational speed ω, fixed tip-speed ratio (TSR) and generator torque control, have been simulated for turbulent inflow. With approximately the same time-averaged ω, the time- averaged power is in good agreement with measurements for all three cases. Although the time-averaged aerodynamic torque is nearly the same for the three cases, the root-mean-square (rms) of the aerodynamic torque fluctuations is significantly larger for the case with fixed ω. No significant differences have been observed for the time-averaged flow fields behind the turbine for these three cases.

LES investigation of infinite staggered wind-turbine arrays

The layouts of turbines affect the turbine wake interactions and thus the wind farm performance. The wake interactions in infinite staggered wind-turbine arrays are investigated and compared with infinite aligned turbine arrays in this paper. From the numerical results we identify three types of wake behaviours, which are significantly different from wakes in aligned wind-turbine arrays. For the first type, each turbine wake interferes with the pair of staggered downstream turbine wakes and the aligned downstream turbine. For the second type, each turbine wake interacts with the first two downstream turbine wakes but does not show significant interference with the second aligned downstream turbine. For the third type, each turbine wake recovers immediately after passing through the gap of the first two downstream turbines and has little interaction with the second downstream turbine wakes The extracted power density and power efficiency are also studied and compared with aligned wind-turbine arrays.