Bulletin of the American Physical Society
70th Annual Meeting of the APS Division of Fluid Dynamics
Volume 62, Number 14
Sunday–Tuesday, November 19–21, 2017; Denver, Colorado
Session M26: Wind Turbine InteractionEnergy
|
Hide Abstracts |
Chair: Ali Hamed, Union College Room: 707 |
Tuesday, November 21, 2017 8:00AM - 8:13AM |
M26.00001: The Restricted Non Linear - LES approach to simulations of large wind farms Joel Bretheim, Charles Meneveau, Dennice Gayme The restricted nonlinear (RNL) model has recently proven useful in numerical investigations of wall-bounded turbulent flows due to the much smaller number of streamwise Fourier modes required to describe the RNL dynamics. In this work, we extend a recently developed large eddy simulation framework for the RNL system (RNL-LES) to the application of wind farms. This new simulation framework exploits the simplified dynamics and computational tractability of the RNL-LES system to provide a computationally inexpensive tool for conducting studies of wind farms over a wide range of parameter variations; such studies may be cost-prohibitive for traditional wind farm simulation tools (e.g. LES). We report simulation results for different configurations of arrays of vertically staggered wind turbines (modeled as actuator disks). We also consider the effects on total farm power production and quantify various terms in the budget of mean kinetic energy, with special emphasis on vertical flux of mean kinetic energy. [Preview Abstract] |
Tuesday, November 21, 2017 8:13AM - 8:26AM |
M26.00002: Numerical investigation of interactions between marine atmospheric boundary layer and offshore wind farm Pin Lyu, Wenli Chen, Hui Li, Lian Shen In recent studies, Yang, Meneveau \& Shen (Physics of Fluids, 2014; Renewable Energy, 2014) developed a hybrid numerical framework for simulation of offshore wind farm. The framework consists of simulation of nonlinear surface waves using a high-order spectral method, large-eddy simulation of wind turbulence on a wave-surface-fitted curvilinear grid, and an actuator disk model for wind turbines. In the present study, several more precise wind turbine models, including the actuator line model, actuator disk model with rotation, and nacelle model, are introduced into the computation. Besides offshore wind turbines on fixed piles, the new computational framework has the capability to investigate the interaction among wind, waves, and floating wind turbines. In this study, onshore, offshore fixed pile, and offshore floating wind farms are compared in terms of flow field statistics and wind turbine power extraction rate. [Preview Abstract] |
Tuesday, November 21, 2017 8:26AM - 8:39AM |
M26.00003: Classification of Rotor Induced Shearing Events in the Near Wake of a Wind Turbine Array Boundary Layer Sarah Smith, Bianca Viggiano, Naseem Ali, Raul Bayoan Cal Flow perturbation induced by a turbine rotor imposes considerable turbulence and shearing effects in the near wake of a turbine, altering the efficiency of subsequent units within a wind farm array. Previous methods have characterized near wake vorticity of a turbine and recovery distance of various turbine array configurations. This study aims to build on previous analysis with respect to a turbine rotor within an array and develop a model to examine stress events and energy contribution in the near wake due to rotational effects. Hot wire anemometry was employed downstream of a turbine centrally located in the third row of a 3x3 array. Data considered points planar to the rotor and included simultaneous streamwise and wall-normal velocities as well as concurrent streamwise and transverse velocities. Conditional analysis of Reynolds stresses induced by the rotor agree with former near wake research, and examination of stresses in terms of streamwise and transverse velocity components depicts areas of significant rotational effects. Continued analysis includes spectral decomposition and conditional statistics to further characterize shearing events at various points considering the swept area of the rotor. [Preview Abstract] |
Tuesday, November 21, 2017 8:39AM - 8:52AM |
M26.00004: Producing Turbulent Wind Tunnel Inflows Relevant to Wind Turbines using an Active Grid Christopher Rumple, Matthew Welch, Jonathan Naughton The rise of industries like wind energy have provided motivation for generating realistic turbulent inflows in wind tunnels. Facilities with the ability to produce such inflows can study the interaction between the inflow turbulence and the flow of interest such as a wind turbine wake. An active grid – a system of actively driven elements - has gained increasing acceptance in turbulence research over the last 20 years. The ability to tailor the inflow turbulence quantities (e.g. turbulence intensities, integral length scale, and turbulence spectrum) is a driving reason for the growing use of active grids. An active grid with 40 independent axes located within the forward contraction of a low speed wind tunnel is used to explore the range of turbulent inflows possible using hot-wire anemometry to characterize the turbulence. Motor control algorithms (i.e. user waveform inputs) used to produce various turbulent inflows will be presented. Wind data available from meteorological towers are used to develop relevant inflows for wind turbines to demonstrate the usefulness of the active grid. [Preview Abstract] |
Tuesday, November 21, 2017 8:52AM - 9:05AM |
M26.00005: Modeling distortion of HIT by an Actuator Disk in a periodic domain Aditya Ghate, Niranjan Ghaisas, Sanjiva Lele We study the distortion of incompressible, homogeneous isotropic turbulence (HIT) by a dragging actuator disk with a fixed thrust coefficient (under the large Reynolds number limit), using Large Eddy Simulation (LES). The HIT inflow is tailored to ensure that the largest length scales in the flow are smaller than the actuator disk diameter in order to minimize the meandering of the turbulent wake and isolate the length scales that undergo distortion. The numerical scheme (Fourier collocation with dealiasing) and the SGS closure (anisotropic minimum dissipation model) are carefully selected to minimize numerical artifacts expected due to the inviscid assumption. The LES is used to characterize the following 3 properties of the flow a) distortion of HIT due to the expanding streamtube resulting in strong anisotropy, b) turbulent pressure modulation across the actuator disk, and the c) turbulent wake state. Finally, we attempt to model the initial distortion and the pressure modulation using a WKB variant of RDT solved numerically using a set of discrete Gabor modes. [Preview Abstract] |
Tuesday, November 21, 2017 9:05AM - 9:18AM |
M26.00006: Mesh Dependence on Shear Driven Boundary Layers in Stable Stratification Generated by Large Eddy-Simulation Jacob Berg, Edward G. Patton, Peter S. Sullivan The effect of mesh resolution and size on shear driven atmospheric boundary layers in a stable stratified environment is investigated with the NCAR pseudo-spectral LES model (J. Atmos. Sci. v68, p2395, 2011 and J. Atmos. Sci. v73, p1815, 2016). The model applies FFT in the two horizontal directions and finite differencing in the vertical direction. With vanishing heat flux at the surface and a capping inversion entraining potential temperature into the boundary layer the situation is often called the conditional neutral atmospheric boundary layer (ABL). Due to its relevance in high wind applications such as wind power meteorology, we emphasize on second order statistics important for wind turbines including spectral information. The simulations range from mesh sizes of $64^3$ to $1024^3$ grid points. Due to the non-stationarity of the problem, different simulations are compared at equal eddy-turnover times. Whereas grid convergence is mostly achieved in the middle portion of the ABL, statistics close to the surface of the ABL, where the presence of the ground limits the growth of the energy containing eddies, second order statistics are not converged on the studies meshes. Higher order structure functions also reveal non-Gaussian statistics highly dependent on the resolution. [Preview Abstract] |
Tuesday, November 21, 2017 9:18AM - 9:31AM |
M26.00007: A fast wind-farm boundary-layer model to investigate gravity wave effects and upstream flow deceleration Dries Allaerts, Johan Meyers Wind farm design and control often relies on fast analytical wake models to predict turbine wake interactions and associated power losses. Essential input to these models are the inflow velocity and turbulent intensity at hub height, which come from prior measurement campaigns or wind-atlas data. Recent LES studies [1,2] showed that in some situations large wind farms excite atmospheric gravity waves, which in turn affect the upstream wind conditions. In the current study, we develop a fast boundary-layer model that computes the excitation of gravity waves and the perturbation of the boundary-layer flow in response to an applied force. The core of the model is constituted by height-averaged, linearised Navier--Stokes equations for the inner and outer layer, and the effect of atmospheric gravity waves (excited by the boundary-layer displacement) is included via the pressure gradient. Coupling with analytical wake models allows us to study wind-farm wakes and upstream flow deceleration in various atmospheric conditions. Comparison with wind-farm LES results shows excellent agreement in terms of pressure and boundary-layer displacement levels. [1] Allaerts D. and Meyers J., J. Fluid Mech. 814, 95-130 (2017) [2] Allaerts D. and Meyers J., Boundary-Layer Meteorol. (Revision submitted) [Preview Abstract] |
Tuesday, November 21, 2017 9:31AM - 9:44AM |
M26.00008: On very-large-scale motions (VLSMs) and long-wavelength patterns in turbine wakes Asim Onder, Johan Meyers It is now widely accepted that very-large-scale motions (VLSMs) are a prominent feature of thermally-neutral atmospheric boundary layers (ABL). Up to date, the influence of these very long active motions on wind-energy harvesting is not sufficiently explored. This work is an effort in this direction. We perform large-eddy simulation of a turbine row operating under neutral conditions. The ABL data is produced separately in a very long domain of $240\delta$. VLSMs are isolated from smaller-scale ABL and wake motions using a spectral cutoff at streamwise wavelength $\lambda_x=3.125\delta$. Reynolds-averaging of low-pass filtered fields shows that the interaction of VLSMs and turbines produce very-long-wavelength motions in the wake region, which contain about $20\%$ of the Reynolds-shear stress, and $30\%$ of the streamwise kinetic energy. A conditional analysis of filtered fields further reveals that these long-wavelength wakes are produced by modification of very long velocity streaks in ABL. In particular, the turbine row acts as a sharp boundary between low and high velocity streaks, and accompanying roller structures remain relatively unaffected. This reorganization creates a two-way flux towards the wake region, which elucidates the side-way domination in turbulent transport. [Preview Abstract] |
Tuesday, November 21, 2017 9:44AM - 9:57AM |
M26.00009: Large eddy simulation study on the effect of vertical staggering in large-scale wind farms Richard Stevens, Mark Arendshorst, Mengqi Zhang We present results from large eddy simulations (LES) of extended aligned windfarms. We vary the hub height of consecutive downstream turbine rows in order to create vertically staggered windfarms and we study the effect of streamwise turbine spacing, turbine rotor diameter, and the hub height difference between consecutive rows on the average turbine power output. The results show that the production of the second turbine row increases when the first turbine row is lower than the second row. Consequently, we find that the average turbine power output increases significantly in the entrance region of the windfarm. However, we find that the relative benefit of vertical staggering, compared to the non-staggered case, decreases further inside the windfarm. The reason is that the vertical kinetic energy transfer, which brings high velocity fluid from above the windfarm towards the hub-height plane, does not significantly increase by vertically staggering wind turbines. This limits the potential benefit of vertical staggering in extended windfarms. We find, for a fixed hub height difference between consecutive rows, that vertical staggering is more beneficial for the average windfarm power output when the streamwise turbine spacing and turbine diameter are smaller. [Preview Abstract] |
Tuesday, November 21, 2017 9:57AM - 10:10AM |
M26.00010: Characterising the coherent structures in large eddy simulations of extended windfarms Mengqi Zhang, Richard Stevens The present work studies the large coherent structures in large eddy simulations of extended windfarms using proper orthogonal decomposition (POD) method. In order to evaluate the effect of wind turbines on the coherent structures, we simulate, in addition to a reference case of a neutral atmospheric boundary layer, periodic and developing windfarms with aligned and staggered wind turbines. The simulations are run for a long time, so as to generate a sufficient database for POD analysis. Both the coherent structures of the velocity field and the kinetic energy flux are investigated. In all the cases, elongated streamwise counter rotating roll structures, covering 1 or 2 turbines in spanwise direction, are identified to be the dominant mode in the POD results. Another pattern, varying in streamwise direction, also appears in all the cases. Besides, by applying a Fourier Transform to the physical database, we also studied the coherent structures at a certain frequency. Preliminary results show that for the most energetic frequency in the windfarms, the real part and imaginary part of the coherent structures are streamwise-varying and the amplitude manifests consistently a streamwise streak structure. [Preview Abstract] |
Follow Us |
Engage
Become an APS Member |
My APS
Renew Membership |
Information for |
About APSThe American Physical Society (APS) is a non-profit membership organization working to advance the knowledge of physics. |
© 2024 American Physical Society
| All rights reserved | Terms of Use
| Contact Us
Headquarters
1 Physics Ellipse, College Park, MD 20740-3844
(301) 209-3200
Editorial Office
100 Motor Pkwy, Suite 110, Hauppauge, NY 11788
(631) 591-4000
Office of Public Affairs
529 14th St NW, Suite 1050, Washington, D.C. 20045-2001
(202) 662-8700