Turbine-to-turbine wake interaction has been widely accepted as a significant contributor to the underperformance and reduced reliability of modern wind plants. Wakes generated by an upstream turbine have been shown to propagate large distances, impacting the inflow conditions and loading of downstream turbines. This negative interaction is compounded by the fact that individual turbines within a wind plant operate autonomously. The control schemes currently implemented on each turbine have no inherent knowledge of the surrounding wind field and, therefore, cannot recognize and proactively respond to important incoming wind features, such as thunderstorm outflows, transient gusts and lulls, or wakes from an upstream turbine.

The use of scanning remote sensing technologies to measure wind plant complex flows provides the opportunity to construct real-time, comprehensive wind maps that could serve as an input into the next generation of individual turbine and wind plant control schemes.

One such technology is specialized Doppler radar, which uses the Doppler effect to obtain information about the wind field. The radar system emits a microwave pulse several thousand times each second, and these pulses are scattered by particles in the atmosphere (e.g., dust, pollen, insects and raindrops). Between each emitted pulse, the radar “listens” for scattered radiation to be returned.

By combining information from successive pulses, the radar can determine particle motion toward or away from the radar, yielding a line-of-site wind vector. Furthermore, the appropriate deployment of two radars providing independent looks to a point in space allows for the construction of dual-Doppler synthesis, yielding complete horizontal wind speed and direction wind fields.

Texas Tech University developed two specialized mobile Ka-band Doppler radar systems called TTUKa radars, which are used to study a wide variety of small-scale atmospheric phenomena, including the structure of thunderstorm outflows, tornadic circulations, hurricane winds at landfall and wind plant complex flows.

The radars provide excellent spatial resolution with a half-power beam width of 0.33°. They also take advantage of specialized pulse-compression frequency modulation techniques, yielding excellent along-beam range resolution of 9 to 15 meters. The TTUKa radars are also capable of scanning speeds of 30°/sec in the horizontal plane and 6°/sec in the vertical plane, providing rapid wind field revisits over a scanned sector every few seconds.

With cooperation from industry partners, the TTUKa radars have been used to investigate the impact of adjusting the controls of an individual turbine on the resulting wake structure of that turbine. Wind field visualization can be provided by a radar system coordinated during periods when operational turbine controls are changing. The radar wind fields can then be related back to the output data provided by the turbine to better understand how subtle changes in turbine operation can yield large impacts to the wakes they produce.


Changes in turbine pitch

Small changes in turbine pitch have been found to yield significant changes in wake structure on short time scales. As the turbine blade pitch changes, the amount of wind capture is altered, varying the turbine thrust and, therefore, modifying the resulting wake generated by the turbine. To highlight this effect, TTUKa measurements were collected every four seconds for a one-hour period (see Figure 1).


The radar is ground based, located 2,700 meters upstream of the turbine and is collecting a plan position indicator sector at a 1.8° elevation angle such that the radar beam intersects the turbine at hub height (80 meters). Throughout the one-hour period, environmental wind conditions were changing, and the turbine pitch control was operating to maximize the turbine’s performance. It can be seen that in the lower wind speed regimes (Panel A), the blade pitch angle is relatively small such that a large turbine thrust is occurring, yielding a well-defined wake that propagates beyond 1.5 km downstream of the turbine. As wind speeds increase and the turbine approaches rated power output, the blade pitch angle increases slightly, turbine thrust decreases, and the resulting turbine wake is hardly identifiable (Panel B).

While the turbine is operating near rated power and is reacting to gusts and lulls in the wind as they pass through the turbine, the blade pitch angle adjusts accordingly, yielding rapidly changing variability in the resulting wake structure (Panels C and D). Within minutes, a cohesive wake structure can completely disappear.


Changes in turbine yaw

Misalignment of turbine yaw position with the prevailing wind direction is known to cause a deflection in the downstream wake propagation. This deflection is a result of a variability in the induced velocity across the rotor sweep, where immediately behind the rotor, one side of the wake is propagating downstream at a different speed than the opposite side. The magnitude of the deflection is believed to be associated with the magnitude of the yaw position offset, where a larger offset will yield a larger deflection.


A TTUKa radar was deployed at the Scaled Wind Farm Technology facility to assess the impact of manual yaw offset on wake deflection. The focus was placed on a turbine pair, where the wake of an upstream turbine was directly impacting the inflow of a nearby downstream turbine during normal operation (Figure 2). The upstream turbine was then manually yawed 20° in the clockwise direction to create an intentional misalignment with the true wind direction. The result was a deflection of the wake immediately behind the turbine, mitigating the previously seen wake impact on the downstream turbine, which in turn, was experiencing inflow conditions characterized by higher wind speed (shown by radial velocity) and reduced turbulence (shown by spectrum width). The wake deflection can be visualized to occur within the distance of only a few rotor diameters downstream of the yawed turbine. The surrounding wind regime then acts to try to realign the wake with the true wind direction.

Remotely sensed wind maps offer situational awareness not provided by current turbine controls and offer the potential to improve future wind turbine and plant controls schemes. Real-time wind maps allow for anticipatory reaction to incoming wind features, such as relevant gusts and lulls, but also offer baseline intelligence to mitigate wake impacts. TTUKa wind field visualizations show that wake structure and propagation – and, thus, wake impacts on downstream turbines – can be modified by changes in turbine pitch and yaw. The trade-offs for implementing these control concepts need to be further investigated, as sacrifices in power output and/or increases in loading may occur to the turbine executing the control changes.

However, the positive impact for downstream turbines may yield a cost-benefit scenario that leads to increased production for the wind plant as a whole. These control improvements offer great potential to increase wind plant reliability and energy capture, ultimately contributing to the reduction of energy costs. w


Brian Hirth is a research professor at Texas Tech University’s National Wind Institute. John Schroeder is a professor at Texas Tech University’s Atmospheric Science Group. They can be reached at brian.hirth@ttu.edu and john.shroeder@ttu.edu, respectively.

Industry At Large: Wind Farm Reliability

How Turbine Controls Impact Wake Behavior

By Brian Hirth & John Schroeder

Using Doppler radar wind maps, researchers better understand how subtle changes in turbine operation have a big impact on turbine wakes.





NAW_body hyperlink NAW_body_i NAW_body_bi NAW_body_b_i NAW_body_b








NAW_SH norule