DENVER (AP) -- Xcel Energy Inc. has a new wind energy forecasting system that it says is saving the company money by helping it better predict when it can power down its natural gas-fired power plants and rely on wind turbines instead.
The system developed by the National Center for Atmospheric Research in Boulder helped Xcel save nearly $6 million last year, or enough to pay for itself, said Bill Mahoney, an NCAR program director who helped oversee the project.
"We're excited with how the system has helped us," Xcel spokesman Gabriel Romero said Thursday.
The system is billed as being 35 percent more accurate than previous forecasting methods. Xcel uses it at windfarms with a combined 3,000 turbines in Colorado, Minnesota, New Mexico, Texas and Wisconsin, Mahoney said.
The challenge for any utility adding wind power is predicting when and how winds will shift. That means knowing when to fire up natural gas-fired plants if there's barely a breeze and when to turn them off when strong winds threaten to overload the system. Starting up or shutting down a gas-fired plant can take hours.
"Xcel came to NCAR in late 2008 and was struggling with the fact that they really wanted to increase their windenergy portfolio," Mahoney said. "But errors from their wind energy prediction capabilities were costing them a lot of money. It was making it ever more costly to take on more wind."
The NCAR forecasting system improves on Xcel's old methods by measuring wind at the height that it hits a turbine's blades rather than at ground level, and gathering that data in real time. It also uses high-tech formulas to calculate how much energy can be generated when winds blow at different speeds and directions.
Romero said the better forecasts allow the company to be more aggressive in relying on wind power.
"We're getting to the point where a huge portion of our load can be covered specifically by wind, which is a whole different era for us," he said. "By doing that, you're talking about big-time savings for our customers."
Mahoney said scientists hope to keep fine-tuning their forecasting to improve predictions for what will happen when ice builds up on turbine blades, which can make them less efficient.
Other utilities and forecasters can use what NCAR has developed to improve their own predictions.
"Xcel had a goal to be a leader in renewable energy around the country. We've always had our mission to be a leader in atmospheric research," Mahoney said. "We're happy to be able to contribute to the country moving forward and adopting more wind energy on a national scale."