• Subscribe

Greener wind blowing in the Columbia River Gorge

Speed read
  • Columbia River Gorge offers an opportunity to harvest renewable wind energy
  • Complex coastal terrain is challenging for existing forecast models
  • Mira supercomputer enables higher-resolution models and more accurate forecasts

Winds in the Pacific Northwest’s Columbia River Gorge commonly reach speeds of 10-40 mph. Such strong winds are a nuisance for sightseers, but they’re a boon to wind surfers, paragliders, and sustainable energy enthusiasts.

Wind turbines already in place in the Gorge can generate up to 4500 megawatts of power, or the equivalent of more than five 800-megawatt nuclear power plants. But engineers are prevented from fully harnessing that power due to the wind’s unpredictable nature.<strong>Blowin' in the wind.</strong> A beautiful and diverse ecosystem, the Columbia River Gorge sports extreme wind conditions, making it an ideal resource for wind energy. Courtesy Cacophony.

Utility operators require reliable predictions of wind power because conventional coal and nuclear power plants take a long time to start up and cool down. If wind power floods the grid without warning, the energy is wasted.

However, the Gorge’s dramatic terrain includes mountains, canyons, and coastal areas which produce a variety of complex effects that influence wind conditions and hamper forecasts.

More accurate forecasts could help integrate more wind energy into the grid and reduce the cost of electricity.

Enter the Wind Forecast Improvement Project (WFIP2), a four-year joint project between the US Department of Energy (DOE), the National Oceanic and Atmospheric Administration (NOAA) and other organizations to improve wind forecasting models.

“If grid operators confidently know where and when the wind is blowing, they have the ability to turn off 'spinning reserves' such as coal plants that are kept online in case the wind forecasts are wrong and back-up energy is needed,” says Joe Olson of NOAA. “These back-up reserves are ultimately an extra cost that is paid for by citizens.”

Mira, mira on the wall…

NOAA’s National Center for Environmental Prediction currently runs two hourly-updating atmospheric forecast models at grid spacing of 13 km and 3 km, respectively. But these models are primarily built to simulate weather on flat terrain, like that found in the American Midwest. In order to resolve wind features in more complex topography like the Gorge, models with much finer 750 m grid spacing are needed, a 16-fold increase.

The first step in boosting the accuracy of models to run at grid spaces of 750 meters (about 1/6 the size of Central Park, or the area of an average wind farm) is to collect extensive real-world data for model testing and validation. To achieve this realistic modeling, the WFIP2 team placed more than 20 environmental sensors in the Gorge to record wind conditions every 10 minutes over an 18-month period.

The sensors will send the data to NOAA, where they are assimilated into a 3D representation of the atmosphere. Those initial conditions are then forwarded to Mira, the 10-petaFLOPS IBM Blue Gene/Q supercomputer at the Argonne Leadership Computing Facility (ALCF) to run experimental models.

<strong>What color is the wind?</strong> 750 meter High-Resolution Rapid-Refresh model from the WIF2 project region. Courtesy NOAA.The team chose Mira because its highly parallel architecture can efficiently simulate a large quantity of weather events at very high model resolutions.

“We compare forecasts against the observations collected in the Columbia River Basin to single out certain weather events that are particularly difficult to forecast well. We then re-run the model for those select case studies many times as we develop aspects of the model until we find improvements,” says Olson.

Modeling synergy

At the end of the field data collection period in March 2017, scientists will simulate an entire year of weather conditions with an emphasis on wind conditions in the Gorge, comparing the existing 3 km control model with a new model with improved physical parameters.

However, the team has encountered complications in predicting wind conditions for terrain with such a high degree of complexity.

“The biggest challenge is determining which part of the model forecast system is the cause of the differences found between the forecasts and the observations,” says Olson. “Our focus is on developing the model physics, but we’ve learned that errors also arise from inadequacies in the initial state of the model atmosphere or the model dynamics.”

“Each component of the numerical weather prediction model needs to work well with the other components in a synergistic fashion, in order to improve the skill of weather forecasts, especially for wind speeds in the turbine rotor layer (50-150 m above ground level),” says Olson.<strong>Fickle wind. </strong> The unreliablilty of wind energy limits its acceptance into the power grid. If utility operators know when to expect an influx of renewables, they will know when to take nuclear and coal plants offline. Courtesy Sam Beebe. <a href='https://creativecommons.org/licenses/by/2.0/legalcode'>(CC BY 2.0) </a>

Greener energy

At the end of the study, the team hopes to show solid improvements in low-level wind forecasts. These improved components will be candidates for implementation into the next upgrade of the operation version of NOAA’s existing hourly-updated forecast models, which support the National Weather Service, Federal Aviation Administration, the Forest Service, and many other agencies.

“Improved wind forecasts have the potential to limit the use of expensive energy reserves such as coal,” says Olson, “and reduce the output of greenhouse gases, which can help mitigate climate change.”

The winds in the Columbia River Gorge may blow off your hat, but soon they may also save you money and improve the quality of the air you breathe.

Join the conversation

Do you have story ideas or something to contribute?
Let us know!

Copyright © 2017 Science Node ™  |  Privacy Notice  |  Sitemap

Disclaimer: While Science Node ™ does its best to provide complete and up-to-date information, it does not warrant that the information is error-free and disclaims all liability with respect to results from the use of the information.


We encourage you to republish this article online and in print, it’s free under our creative commons attribution license, but please follow some simple guidelines:
  1. You have to credit our authors.
  2. You have to credit ScienceNode.org — where possible include our logo with a link back to the original article.
  3. You can simply run the first few lines of the article and then add: “Read the full article on ScienceNode.org” containing a link back to the original article.
  4. The easiest way to get the article on your site is to embed the code below.