Topography effects on the wake and performance of wind turbines: a wind tunnel study

The past two decades have seen a growing interest of the research community in the interaction between the atmospheric boundary layer (ABL) and wind turbines. Wind turbine performance is known to be affected by the characteristics of the incoming flow. In particular, the mean wind speed and the turbulence intensity levels are used in engineering applications to estimate turbine power production as well as the fatigue loads on the turbines. Furthermore, in wind farms, these flow statistics are strongly affected by wind-turbine wakes. In order to better understand and predict these effects, a number of issues have yet to be addressed, such as the influence of the stability of the atmosphere on the wakes, the interactions of the wakes either between themselves or with downstream wind turbines in a wind farm, and the effect of topography on turbine performance as well as turbine-wake structure.

From the ABL point of view, the surface of the earth is in most parts rough, thus slowing down the flow close to the surface and enhancing the turbulence intensity to the detriment of wind turbines. However, locally the wind velocity can also be increased by effect of topography such as the acceleration of the flow on the windward face of a hill, on an elevated plateau, or by the Venturi effect in valleys.

To investigate these effects on wind turbines, experiments are being performed in the wind tunnel of the Wind Engineering and Renewable Energy (WIRE) laboratory at the Ecole Polytechnique Fédérale de Lausanne (EPFL), Switzerland. It is a suck-down wind tunnel with a test section 28m long, 2.5m wide and 2m high and running with a free stream velocity of up to 8m/s. Angular velocity measurements were done followed by instantaneous velocity measurements with multi-hole pressure probes. Each probe contains 4 pressure taps closely packed but in different planes, and the three components of the instantaneous velocity can be calculated from the corresponding pressure measurements as long as the angle between the velocity direction and the direction the probe is pointed towards is less than 45°.

The first case tested is that of a stand-alone three-blade turbine with a rotor diameter D=127mm on a 107mm mast mounted on a cliff-like forward facing step 1D-high. The flow over a forward facing step has been extensively investigated and is highly inhomogeneous. Three main regions can be distinguished: a recirculation area in front of the step, a separation bubble starting at the edge of the step, and a re-attachment of the flow further downstream. Above the front edge of the step, the acceleration of the mean flow due to the constriction of the step could potentially be exploited by a wind turbine to increase its power output. However the strong shear associated with the recirculation area enhances the turbulence intensity which can also be detrimental to wind energy production.

Angular velocity (RPM) measurements were done for three different reference velocities with the turbine situated at different locations between 0.5D behind the front edge of the step to 0.5D from the back edge of the (16D-long) step. The results show a peak in RPM at 1D behind the front edge of the step, followed by a sharp decline as the turbine is moved downstream. A plateau is reached at 6D downstream followed by a slight increase up to end of the step. The peak in RPM represents an increase of about 50% compared to the minimum found at x/D=6. The wake measurements were subsequently done with the wind turbine positioned at x/D=1.

Preliminary measurements of the flow over the step (without turbine) were carried out with a free stream velocity of 8m/s and a turbulence intensity of about 6.5%. They confirmed the presence of an accelerated flow at hub height at a distance of x/D=1 downstream the edge of the step. The mean velocity is here 16% larger than it would be without step. The results also show a peak in turbulent intensity with a 5-fold increase compared to background turbulence. The peak is located 50mm above ground (the bottom tip of the turbine being at 43mm above ground) and sharply decreases with height. At hub height the turbulence intensity is 50% greater than that of the background flow.

Vertical and spanwise profiles of mean velocity, turbulence intensity and Reynolds stresses were obtained at multiple locations downstream the wind turbine. Despite the different conditions in mean flow and turbulent intensity between the top- and bottom halves of the wind turbine, the wake remains fairly symmetrical (once the incoming flow is subtracted) up to 2D downstream and does not appear to be displaced vertically as might have been expected with the presence of the recirculation area. Further downstream the recovery of the wake is faster in the upper part and the maximum velocity deficit is shifted toward the height of the bottom tip of the turbine blades. Considering the velocity at hub height and in comparison to a turbine placed in a boundary layer without topography the results also show a tendency to a faster recovery of the mean flow. With regards to the turbulence intensity, we can here distinguish two regions where the turbine has distinct effects. In the near wake (x/D<2) the turbulence is enhanced above hub height and dampened below, but further downstream the turbulence intensity is clearly dampened throughout the wake whereas wind turbines are typically considered to have an enhancing effect on turbulence levels.

These preliminary results show the complexity of the interaction of topography and wind turbines in the case of a forward-facing step. Different shapes of ramps bringing the flow to the top of the step are currently being tested and the most recent results will be presented in due course during the conference.