AWS Truepower Improves Wind Project Modeling Techniques

“Our goal is to advance the industry’s ability to accurately model wind turbine wakes, especially in large wind farms,” stated Philippe Beaucage,
Senior Research Scientist at AWS Truepower. “AWS Truepower has developed two wake models over the past two years: 1) the Deep-Array Wake Model (DAWM), which is currently implemented in our openWind Enterprise software for wind park design, and 2) a research model based on Large-Eddy Simulations (LES). WakeBench gives us the opportunity to validate and
improve these models.”

A wake model attempts to simulate the wakes generated by wind turbines and their
impacts on other turbines.  As Beaucage
explains, “DAWM is adapted from a
conventional wake model, and represents an intermediate stage in model
advancement. Our more advanced LES model – which resolves most of the
energy-containing scales of atmospheric turbulence – simulates both the wind flow and
turbine wakes in a single, integrated solution.”

Each WakeBench participant will run their wake model using the same initial conditions in order to do a fair comparison. The models will be validated at several different sites. This project started in October 2011, will last for three years, and is being co-managed by the National
Renewable Energy Laboratory (NREL) and the National Renewable Energy Centre of Spain (CENER).  CENER, in Spain, is putting together all the
test cases and will gather the results from each model and perform the
comparison.

“We believe our participation in WakeBench is critical to advancing the industry’s capability in
this area,” stated Michael Brower, Chief Technical Officer at AWS Truepower.
“Our ultimate goal is to reduce the uncertainty in energy production estimates
by applying the best available modeling technology.”

In previous research, AWS Truepower confirmed that large wind turbine arrays
can experience greater wake losses than predicted by standard wake models. This
led to the development of the DAWM. The WakeBench
project is significant because it will expand the number of test cases
available for validating and fine-tuning models under a range of conditions.

For more information on wake effects,
download our technical paper at  http://www.awstruepower.com/2011/06/the-openwind-deep-array-wake-model-development-and-validation/