Authored by
Stephen Ferguson
Application Engineer
CD-adapco
Melville, N. Y.

Edited by Leslie Gordon
leslie.gordon@penton.com

Resources:
CD-adapco, www.cd-adapco.com
University di Udine, www.dimi.uniud.it
Wind Giant, www.windgiant.com

For years, CFD software has helped engineers optimize designs in traditional industries such as oil and gas. So it’s probably no surprise the software has also taken a role in renewable energy. For example, the software can predict the efficiency of a wind turbine while simultaneously calculating its structural loading.

Recall that a wind turbine is essentially a large, inverse fan. Instead of using electricity to produce a breeze, a wind turbine relies on wind pressure acting on its blades to turn a rotor which, via a gearbox, drives a generator. In turn, populations of these turbines on wind farms feed larger electrical grids.

The design problem at hand is to extract the maximum amount of energy in the widest range of conditions at the least cost. This applies to individual units as well as entire farms. For example, turbines must be located to maximize their total energy output while minimizing the likelihood of damage. (Wind-turbine gearboxes are notorious for being vulnerable to wind gusts.)

CFD software such STAR-CCM+ comes into play by letting users simulate the airflow over a potential site with different turbine configurations. This helps wind-farm developers see complex wind patterns, identify areas with high wind speeds and turbulence, and optimize turbine positioning.

In the software, turbines come from a library of virtual models. They represent the influence of the turbine (for example, how its wake affects the flow downstream) without having to physically represent the complex geometry of the turbine. Users also import specific terrain geometry. The software then predicts what is called the wind-power density across the terrain. It also pinpoints local 3D flow features such as wind shear, wind veer, and gustiness that could potentially harm the gearbox. The software evaluates terrains with little or no manual input from the operator.

Previously, numerical simulation for wind-park prospecting oversimplified the underlying physics. The tools ignored the influence of the turbines themselves by neglecting wake interactions and they used overly simplistic terrains. The software was also too complex and time consuming to deploy realistically because users were forced to figure out how to make many different modules work together. Additionally, simulation runs took a long time — often more than two weeks. In contrast, STAR-CCM+ lets users analyze large terrains for 12 different wind directions in less than 3 hr.

Aerodynamics and structural loading
Most current wind-turbine designs use data from experimental testing on small physical models to maximize efficiency. This approach provides considerable insight into a particular design’s performance. The downside: Prototypes are costly and their construction is time consuming.

In contrast, CFD simulations are typically carried out at full scale. The computer model, thus, has the same dimensions as the actual production wind turbine. This lets users interpret results directly without undergoing scaling, which can introduce significant uncertainty, especially for transient phenomena such as the impact of extreme gusts.

In a recent project, our engineering services team used STAR-CCM+ to help Wind Giant, Schiedlberg, Austria, demonstrate the feasibility of its new wind-turbine concept. The turbine uses innovative technology in which wind flow accelerates through a shrouded multi-bladed fan. Simulations gave detailed insight into the turbine’s aerodynamics and force loadings, letting engineers quantify the effects of multiple design changes.

Compared to traditional turbines, the Wind Giant design is more compact while delivering higher energy per unit surface area at wind speeds as small as 1.5 m/sec. Additionally, CFD helped optimize the units for noise (they emit less than 40 dB), so they can reside in urban residential as well as industrial settings.

In another example, the University di Udine, Italy, used STAR-CD, another CFD tool, to study blade shrouding. “Shrouds act as static wings or sails,” says Hans Grassman of the University. “The idea is to exploit this effect to increase aerodynamic efficiency. Developers over the last decade have attempted to do this many times, but usually succeeded only in wind-tunnel studies, not in ambient air.”

Before the University entered the picture, a private company had built an expensive prototype that failed to meet expected design goals. “Our CFD studies of shrouded and unshrouded units immediately revealed complex relationships between flow forces on the turbine and the transfer of energy and linear momentum,” says Grassman. “Turns out, given a certain force, energy transfer is not dependant on flow velocity. However, momentum transfer is dependant. Therefore, it’s not possible to boost the power of a typical shrouded wind turbine beyond the theoretical limit for the same turbine without a shroud — the so called Betz limit. Knowing this in advance could potentially have cut the company’s project costs by millions of dollars.”

In the past, the shrouds were fitted closely to the propeller to minimize tip-vortex drag, says Grassman. “STAR-CD showed that increasing this area actually causes air to accelerate as it approaches the turbine,” he says. “Analysis showed the resulting pressure drop forces air through the propeller at a mean axial velocity of
7.2 m/sec, while ambient wind velocity is only 5 m/sec. The design ups the turbine’s peak power by a factor of four. These so-called ‘partially static turbines’ will allow the construction of hydro power plants in places where large dams are not feasible.”

Extreme loading
Other benefits of CFD software include an ability to exactly pinpoint extreme loading. This helps designers understand conditions that will ensure safe operation. In addition, a rapid turnaround time helps break dependence on preexisting design codes, letting designers more easily explore multiple “what-if” scenarios. Once a user has set up a CFD model, it is relatively simple to repeat calculations for multiple loading schemes.

Additionally, CFD provides much more data than that garnered from a few experimental monitoring probes. In fact, it provides data at every point on the turbine, at every discrete time interval. Users can see results from any angle and calculate instantaneous forces acting on any part of the structure.

Data from CFD calculations can also feed other types of analysis. For example, users can export the forces acting on a turbine to FEA software to pinpoint the stresses at which a blade might break. In extreme cases, such as where fluid forces cause large blade deflections, users can couple CFD simulations directly with FEA and perform stress and fluid simulations simultaneously, each simulation feeding updated boundary conditions to the other.

Offshore wind farms
When it comes to ocean wind farms, heavy gusts, saltwater corrosion, and high waves all make designing a durable, cost-effective solution difficult at best. Here, 3D flow simulation plays a part by making it simpler to model wave loading. Users can analyze different wave heights, sea states, and wind speeds, at full scale. Routinely used in the marine industry for a long time for boat design, CFD software now reduces the impact of harsh offshore settings on wind turbines.

What’s the Betz Limit?
Also called the Betz Law, the Betz Limit describes the theoretical limit of the percentage of power that can be extracted from fluid moving through a turbine. Betz proves that the average wind speed through a rotor (as shown to the left) is the average of the undisturbed wind speed before the turbine, V1, and the wind speed after passing through the rotor plane, V2, or (V1 + V2)/2.

The mass of the air moving through the rotor per second is:

m= PF(V1 +V2) / 2

where m = mass per second, Ρ = density of air, F = swept rotor area, and [(V1 + V2)/2] is the average wind speed through the rotor area.

According to Newton’s second law, the power extracted from the wind by the rotor is equal to the mass times the drop in the wind speed squared:

P = (1/ 2)m(V1 2 - V2 2 )

Substituting m into this expression from the first equation gives this expression for the power extracted from the wind:

P = (P / 4)(V1 2-V2 2)(V1+V2) F

Next, compare these results with the total power of the undisturbed wind traveling through the exact same area, F, without a rotor to block it. Calling this power P0:

P0 = (P / 2)V1 3 F

The ratio between the power extracted from the wind and the power in the undisturbed wind is:

(P/P0) = (1/2) (1 – (V2/V1)2) (1+ (V2/V1))

Plotting P/P0 as a function of V2/V1 shows the function reaches its maximum at V2/V1 = 1/3 and that the maximum value for the power extracted from the wind is 0.59 or 16/27 of the total power in the wind. A design’s success can be measured by how close it gets to this 0.59 efficiency figure.

© 2010 Penton Media, Inc.