Experimental aerodynamics (funded by DOE)

Aerodynamic analysis of wind turbines using special small scale turbine models that share important wake characteristics with utility scale wind turbines. Understanding the wind turbine wakes is important to improve the siting of individual turbines in wind farms for optimal output as shown in figure 1. A blade for a 2m rotor was manufactured by a PhD student and later used in the WINDEE facility (London, Ontario) to measure the turbine wake. 

Microscale modeling of wind resources in complex terrain (funded by DOE)

Most wind farms located in Wyoming are sited in complex terrain including hills, ridges and valleys. A detailed prediction of the influence of this topographic features on the local wind resource is important for optimal siting of the wind turbines. Lower fidelity (Reynolds Averaged Naiver Stokes – RANS) and higher fidelity approaches (Large Eddy Simulation) have been developed to predict the local wind resource using Computational Fluid Dynamics (CFD). For example, the Sierra Madre site (see figure 3 near Rawlins, WY, has been studied and the predicted wind distribution compared to available date form Met towers. Highly local flow features such as boundary layer separation and recirculation regions could be predicted as shown in figure 4.  

High fidelity CFD modeling of Wind Farms (funded by DOE):

Such predictions are possible using a multiscale modeling approach using different simulation tools. At the largest scale, we use the Weather Research and forecasting (WRF) model to predict winds on a continental scale with resolution of about 3km. The meso-scale WRF results then drive the inflow for an intermediate scale LES  (such as shown above) which generates the inflow for the finest level Wind Plant CFd (see figure 5 top right). The Wind Plant CFD is based on fully resolved blades using the overset mesh technology as shown see figure 5 top-right for s single 5MW NREL turbine. Single turbines are then combined to large wind farms to study the wake interaction.  Simulations for a many as 48 wind turbines (figure 5 bottom) have been conducted in the past requiring 1.55 billion degrees of freedom and using 22,464 computer cores in parallel on the NCAR-Wyoming Supercomputer (NWSC). 

 

Wind Resource Modeling (funded by NREL):

The primary goal of the Tools Assessing Performance (TAP) project is to enable widespread adoption of distributed wind (DW) technology by improving resource characterization capabilities, thereby reducing project performance risk and uncertainty. The team will develop a computational framework which will allow the DW community to access the wind resource data and modeling capabilities to perform timely and accurate performance assessments for DW projects at locations across the U.S. This will ultimately lead to a reduction in the DW levelized cost of energy (LCOE) and a significant increase in the installed capacity of DW across the U.S. 

A main objective of the TAP project is the development of a freely available national wind resource dataset for the Contiguous United States (CONUS) for 20 years using regional scale weather models. This new dataset will replace the existing wind resource dataset, the WIND Toolkit, that is widely used by the industry for making siting decisions for large scale wind energy generation plants, and for grid integration.  The new dataset will update the WIND Toolkit to include characterization of wind resources from surface to 80 m above ground for supporting distributed wind deployment, and to include uncertainty information. The uncertainty of the 20 year dataset will be estimated from shorter 1-3year ensemble runs.