High-Res Forecasts Could Help U.S. Expand Offshore Wind Power
A new NOAA dataset of wind forecasts could help the energy industry identify which offshore areas in the United States have the best potential for wind resource development. The data from this study, just published in the journal Wind Energy, could pinpoint areas of reduced risk and ultimately help minimize the cost of offshore wind power.
“This dataset provides information on turbine hub-height wind speeds in areas off the coasts where measurements are very sparse or nonexistent,” said study lead author Eric James, a scientist at the Cooperative Institute for Research in Environmental Sciences at the University of Colorado Boulder working in NOAA’s Global Systems Laboratory.
“Reducing uncertainty could improve companies’ bottom line,” added Melinda Marquis, Program Manager for Atmospheric Science for Renewable Energy at NOAA.
Offshore wind energy currently provides about 10 GW (1 GW= 1 billion watts) of electricity in Europe. In the United States, only a single 30‐MW (1 MW= 1 million watts) wind farm near Block Island, Rhode Island is currently operating. To accelerate the development of technologies used for offshore wind, the U.S. Department of Energy (DOE) has laid out a strategy for offshore wind development. Part of that strategy is identifying areas with a large wind resource—a prospecting process that requires high-quality wind measurements at the height of the wind turbines.
Block Island Wind Farm. Image: Ionna22/ Wikimedia Commons
It's expensive and difficult to measure wind properties over the ocean at the height of wind turbine rotors. So wind energy scientists and operators use a workaround: They get wind estimates from the output of computer weather models, called numerical weather prediction models. If that output were known to be accurate, scientists and operators could rely on those wind estimates, instead of deploying costly instruments.
NOAA and CIRES scientists have been developing a powerful tool that they believe could provide accurate wind estimates: NOAA’s High-Resolution Rapid Refresh (HRRR) weather model. HRRR uses observations from aircraft, satellite, and radar in its hourly weather predictions over the the entire lower 48 United States. The model produces high resolution forecasts— grid boxes are spaced just three kilometers (km), or less than two miles, apart. That means the model can depict small weather systems impacted by geography, including the complex topography along U.S. coastlines. Because the HRRR is more advanced than other NWP models used in wind resource studies, researchers wanted to see if it could produce better estimates of coastal winds.
On this map of 1- hour HRRR 80 m wind speed wind forecasts in meters per second (m/s), white stars roughly identify buoy locations, and outlined areas correspond to the offshore regions in the study. For reference, 4 m/s = 9 mph, 12 m/s = 27 mph, and 25 m/s = 56 mph.
Using the experimental version of the HRRR weather model, the researchers created a dataset of three years of hourly wind forecasts at 80 meters, the height of a typical turbine rotor. They focused on five regions: the northeast U.S. Atlantic coast, the Great Lakes, the coastal Carolinas, the south Texas Gulf coast, and the Pacific coast of Oregon and northern California. From the dataset, they also characterized the frequency of winds exceeding three critical speeds: 4, 12, and 25 m/s—the speeds at which a typical turbine starts producing power, is most efficient, and needs to shut down to avoid damage to the motor, respectively.
To validate that those forecast winds were close to reality, the researchers checked against data collected by existing offshore buoys, which measure winds close to the surface. The team used those data to estimate winds at 80 m high, and they concluded the model worked well: 80-m wind estimates were good enough to be used by the industry to accurately assess areas for offshore wind development.
Average 80m wind speed from 1-hour HRRR forecasts over the study period showing the northeast coast of the United States.
The research team is now looking at how offshore winds at the same five locations vary. “Our dataset, which starts in January 2012, continues to grow in length, and that will allow us to look for patterns in seasonal and day-versus-night winds,” said James.
Those differences in wind speed could be key to smart wind power generation. “In the summertime, off the Northeast and Mid-Atlantic coasts, daytime wind speeds tended to be higher closest to the coast,” James said. “So the timing of the diurnal wind cycle could help the region meet high electricity demands during the summer months.” Researchers are also interested in using HRRR wind data at sites across the continental United States to identify promising locations for wind development.
This study was available online in December, 2017 ahead of publication in the April, 2018 issue of Wind Energy.