Wind Energy

Flow Analysis, Inc., through its innovative Vorticity Confinement (VC) model investigates the downstream vortex wake flow, including effects from complex terrain, wake interference and wake losses. Aerodynamic noise predictions are also made for any site including reflections, refraction and diffraction. This can be applied for a large number of wind applications as follows:

1. Wind resourcing:

Estimated micro-scale wind maps are provided for any area of interest. Meteorological data/wind statistics are used at reference points for further calculations at smaller scales. An atmospheric boundary layer model is included in the computations and complex configuration for terrain and local obstacles are easily accommodated in the uniform Cartesian grid. With a conventional CFD method, even for attached flow, the boundary layer (BL) vorticity would quickly diffuse away unless the grid is fitted to the boundary and special numerical schemes employed. These schemes are far more complex than those employed with VC, where the grid is uniform Cartesian. However, the use of VC confines vorticity to 2-3 grid cells along the surface, when it is attached, even when the grid is not aligned with the surface. VC has proven to be a very robust and efficient way of modeling boundary layers.

VC is straightforward to implement. It effortlessly accommodates any geometrical configuration as it uses an uniform Cartesian grid, which greatly reduces the computational time. This method employs a “level set” representation unlike complicated adaptive grid methods and allows the surface to be simply ‘immersed’ in a uniform Cartesian grid. This results in thin vortical regions along the surface, which is smooth in the tangential direction, giving no “staircase” effects.

A Velocity contour map which includes regions of highest wind speed is provided. Such “hot spots” are easily predicted with VC and with no requirement to use a very fine grid. Additionally, “wind roses” are also provided for any turbine site location. The simulation model, which is fairly simple to implement requires no computational effort for grid generation. Therefore, it is possible to execute wind resource simulations using a laptop PC, needing only a reasonably small computational time.

2. Turbine Design

Wake development plays a dominant role in wind turbine aerodynamics. The flow behind the turbine may result in a skewed wake, which causes fluctuating loads on downwind turbines. Also, aerodynamic loads due to high winds and wakes from upstream turbines and obstacles must be computed. The phenomenon known as dynamic stall, which has a significant effect on rotor speed and power production, as well is also included in the modeling. Further, results from VC can be input to aero-elastic codes to study the structural dynamics at different flow conditions. Due to the use of uniform grids any blade/tower configuration is easily accommodated.

VC is also helpful to perform aerodynamic studies and improvise innovative models of windmills with minimal computational effort, as compared to cost and time incurred in setting up a wind tunnel test or using a highly complicated fine grid/adaptive grid/higher order schemes.

3. Micrositing

To maximize power production in a wind farm it is necessary to determine the exact position of each wind turbine. Not only does wind play an important role but the complex topographical conditions need to be taken into account as well. Simulations for numerous locations are done to determine the local wind field for an optimized farm. Wake losses from the upwind turbine energy extraction are included in the model to compute downwind and crosswind spacing of the wind turbine. Array losses depend on location and configuration of the turbines, wind regime and turbulent wakes. All of these are easily computed for a best or “near best” micro-siting decision.

The wind farm array depends on wind speed and direction. Computations for various flow conditions are performed with minimal computational effort/time. Fast, flexible, easy to use methods such as VC are necessary for computations involving wind energy, since the highly nonlinear dependence of power output precludes the attempt to use average wind models. Thus, many direct numerical simulations must be done for many wind conditions to effectively determine windmill siting.

4. Noise

As the demands of locating wind farms near the energy grid and resulting infringement on population increase, so too does rectifying the intrusive nature of noise that is generated by the wind farms. The major source of noise originates from flow interaction with blades. This is referred as aerodynamic noise and generally increases with tip speed/tip speed ratio. Each wind turbine produces noise to its own level. The levels of “unwanted” noise are computed at varying wind speeds. When large wind turbines are located close to highly populated areas, developing quiet wind turbines becomes very important. The unwanted noise level depends on background noise, distance between source and observer, reflections from multiple surfaces and refraction.

The classical theory of using spherical waves is not accurate for long distance propagation, due to the influence of wind, topography and temperature gradients. VC incorporates an Eulerian technique called, Wave Confinement (WC). Atmospheric absorption is also included in the model to compute the noise decay. Noise contours for any site of interest are generated. The temperature gradients near the surface can form evaporated ducts, which can propagate the noise for longer distances than expected. WC has the capability to capture these realistic effects.