Session H2: Wind Turbines: Experiments

Fractional Flow Speedup from Porous Windbreaks for Enhanced Wind Turbine Power.

A wind tunnel experiment was performed to investigate the potential of porous windbreaks to increase the momentum into the swept area of a wind turbine, and thus power output. Planar particle-image velocimetry (PIV) along with linear perturbation theory is used quantify the effect of windbreak height in the changes in power output. Results show that far above the windbreak, perturbations reduce to potential flow, with a near-ground boundary condition defined by the recirculation zone behind the windbreak. Similarity in the windbreak flow is investigated and used to predict an increase in power which depends roughly linearly with windbreak height, which is corroborated by direct measurements of power from a model wind turbine. The flow field predicted by the linear theory is in broad agreement with the PIV measurements. By incorporating this result with a top-down wind turbine boundary layer approach which treats the windbreaks as additional roughness, it is found that there exists an inter-turbine spacing, on the order of 10 rotor diameters, for which windbreaks induce a net positive effect. This break-even spacing is dependent on surface roughness and the spanwise width of the windbreaks.   

Wake flow control using a dynamically controlled wind turbine

A wind tunnel based ``Hyper Accelerated Wind Farm Kinematic-Control Simulator'' (HAWKS) is being built at Texas Tech University to emulate controlled wind turbine flow physics. The HAWKS model turbine has pitch, yaw and speed control which is operated in real model time, similar to that of an equivalent full scale turbine. Also, similar to that of a full scale wind turbine, the controls are developed in a Matlab Simulink environment. The current diagnostic system consists of power, rotor position, rotor speed measurements and PIV wake characterization with four cameras. The setup allows up to 7D downstream of the rotor to be mapped. The purpose of HAWKS is to simulate control strategies at turnaround times much faster than CFD and full scale testing. The fundamental building blocks of the simulator have been tested, and demonstrate wake steering for both static and dynamic turbine actuation. Parameters which have been studied are yaw, rotor speed and combinations hereof. The measured wake deflections for static yaw cases are in agreement with previously reported research implying general applicability of the HAWKS platform for the purpose of manipulating the wake. In this presentation the general results will be introduced followed by an analysis of the wake turbulence and coherent structures when comparing static and dynamic flow cases. The outcome of such studies could ultimately support effective wind farm wake flow control strategies.   

Effective solidity in vertical axis wind turbines

The flow surrounding vertical axis wind turbines (VAWTs) is investigated using particle imaging velocimetry (PIV). This is done in a low-speed wind tunnel with a scale model that closely matches geometric and dynamic properties—tip-speed ratio and Reynolds number—of a full size turbine. Previous results have shown a strong dependance on the tip-speed ratio on the wake structure of the spinning turbine. However, it is not clear whether this is a speed or solidity effect. To determine this, we have measured the wakes of three turbines with different chord-to-diameter ratios, and a solid cylinder. The flow is visualized at the horizontal mid-plane as well as the vertical mid-plane behind the turbine. The results are both ensemble averaged and phase averaged by syncing the PIV system with the rotation of the turbine. By keeping the Reynolds number constant with both chord and diameter, we can determine how each effects the wake structure. As these parameters are varied there are distinct changes in the mean flow of the wake. Additionally, by looking at the vorticity in the phase averaged profiles we can see structural changes to the overall wake pattern.   

Low order physical models of vertical axis wind turbines

In order to examine the ability of low-order physical models of vertical axis wind turbines to accurately reproduce key flow characteristics, experiments were conducted on rotating turbine models, rotating solid cylinders, and stationary porous flat plates (of both uniform and non-uniform porosities). From examination of the patterns of mean flow, the wake turbulence spectra, and several quantitative metrics, it was concluded that the rotating cylinders represent a reasonably accurate analog for the rotating turbines. In contrast, from examination of the patterns of mean flow, it was found that the porous flat plates represent only a limited analog for rotating turbines (for the parameters examined). These findings have implications for both laboratory experiments and numerical simulations, which have previously used analogous low order models in order to reduce experimental/computational costs.   

Velocity Data in a Fully Developed Wind Turbine Array Boundary Layer

Results are reported from an experimental study of an array of porous disks simulating offshore wind turbines. The disks mimic power extraction of similarly scaled wind turbines via drag matching, and the array consists of 19x5 disks of 0.25 m diameter. The study was conducted in the UNH Flow Physics Facility (FPF), which has test section dimensions of 6.0 m wide, 2.7 m high and 72.0 m long. The FPF can achieve a boundary layer height on the order of 1 m at the entrance of the wind turbine array which puts the model turbines in the bottom third of the boundary layer, which is typical of field application. Careful consideration was given to an expanded uncertainty analysis, to determine possible measurements in this type of flow. For a given configuration (spacing, initial conditions, etc.), the velocity levels out and the wind farm approaches fully developed behavior, even within the maintained growth of the simulated atmospheric boundary layer. Benchmark pitot tube data was acquired in vertical profiles progressing streamwise behind the centered column at every row in the array.   

Wind tunnel measurements of wake structure and wind farm power for actuator disk model wind turbines in yaw

Reducing wake losses in wind farms by deflecting the wakes through turbine yawing has been shown to be a feasible wind farm control approach. In this work, the deflection and morphology of wakes behind a wind turbine operating in yawed conditions are studied using wind tunnel experiments of a wind turbine modeled as a porous disk in a uniform inflow. First, by measuring velocity distributions at various downstream positions and comparing with prior studies, we confirm that the non­-rotating wind turbine model in yaw generates realistic wake deflections. Second, we characterize the wake shape and make observations of what is termed a ``curled wake," displaying significant spanwise asymmetry. Through the use of a 100 porous disk micro-­wind farm, total wind farm power output is studied for a variety of yaw configurations. Strain gages on the tower of the porous disk models are used to measure the thrust force as a substitute for turbine power. The frequency response of these measurements goes up to the natural frequency of the model and allows studying the spatio­temporal characteristics of the power output under the effects of yawing.   

Wind tunnel measurements of a large wind farm model approaching the infinite wind farm regime

A scaled wind farm, with 100 porous disk models of wind turbines, is used to study the effect of wind farm layout on the wind farm power output and its variability, in a wind tunnel study. The wind farm consists of 20 rows and 5 columns. The porous disk models have a diameter of $0.03m$ and are instrumented with strain gages to measure the thrust force, as a surrogate for wind turbine power output. The frequency response of the measurements goes up to the natural frequency of the models and allows studying the spatio-temporal characteristics of the power output for different layouts. A variety of layouts are considered by shifting the individual rows in the spanwise direction. The reference layout has a regular streamwise spacing of $S_x/D=7$ and a spanwise spacing of $S_y/D=5$. The parameter space is further expanded by considering layouts with an uneven streamwise spacing: $S_x/D = 3.5 \& 10.5$ and $S_x/D = 1.5 \& 12.5$. We study how the mean row power changes as a function of wind farm layout and investigate the appearance of an asymptotic limiting behavior as previously described in the literature by application of the top-down model for the spatially averaged wind farm - boundary layer interaction.   

Achieving Full Dynamic Similarity with Small-Scale Wind Turbine Models

Power and thrust data as a function of Reynolds number and Tip Speed Ratio are presented at conditions matching those of a full scale turbine. Such data has traditionally been very difficult to acquire due to the large length-scales of wind turbines, and the limited size of conventional wind tunnels. Ongoing work at Princeton University employs a novel, high-pressure wind tunnel (up to 220 atmospheres of static pressure) which uses air as the working fluid. This facility allows adjustment of the Reynolds number (via the fluid density) independent of the Tip Speed Ratio, up to a Reynolds number (based on chord and velocity at the tip) of over 3 million. Achieving dynamic similarity using this approach implies very high power and thrust loading, which results in mechanical loads greater than 200 times those experienced by a similarly sized model in a conventional wind tunnel. In order to accurately report the power coefficients, a series of tests were carried out on a specially designed model turbine drive-train using an external testing bench to replicate tunnel loading. An accurate map of the drive-train performance at various operating conditions was determined. Finally, subsequent corrections to the power coefficient are discussed in detail.