Boulder scientists say a wind energy forecasting system they helped develop is saving Xcel Energy and its customers millions of dollars each year.
The system was developed by the Boulder-based National Center for Atmospheric Research – NCAR – with Xcel. It enables the utility to capture energy from turbines far more effectively and at a lower cost.
The system was formally taken over by Xcel last month.
By issuing forecasts that are 35 percent more accurate than previous forecasting methods, the system enables utility operators to constantly anticipate the amount of energy produced by wind farms across Xcel Energy’s service area, says NCAR.
As a result, the utility can make critical decision about saving money by powering down traditional coal and natural gas plants when possible while still meeting the needs of customers.
“The goal of this project is to make it more affordable for Xcel Energy to bring on more wind energy,” said William Mahoney, an NCAR program director overseeing the project. “Xcel Energy has been very proactive in adding wind energy to its system. But one of the major obstacles is the difficulty in predicting when and how strongly winds will blow at the location of the turbines.”
“Every fraction we can improve the forecasts results in real savings,” Mahoney said.
The system, which has become increasingly accurate since NCAR entered into a contract with Xcel Energy to begin developing it in 2009, saved the utility $6 million in 2010.
Future savings will vary from year to year, depending on such factors as prices of other energy sources and the amount of wind in a given year, NCAR said.
The system relies on highly detailed observations of atmospheric conditions and an ensemble of cutting-edge computer models. It issues frequent high-resolution wind energy forecasts, updated with new information every 15 minutes, for wind farm sites.
It is used for wind farms in states served by Xcel, including Colorado, Minnesota, New Mexico, Texas and Wisconsin.
Monte Whaley: 720-929-0907 or mwhaley@denverpost.com
