The National Center for Atmospheric Research is entering into a new phase of its partnership with Xcel Energy, aimed at providing the utility more sophisticated forecasting, which should enable Xcel to improve its renewable energy efficiency.

Under the agreement, NCAR scientists expect to develop custom forecasting programs that can predict abrupt changes in the wind, enable utilities to shut down turbines in advance of damaging icing events, and also predict the amount of energy being generated by solar panels.

The work is being funded through XCel energy.

The forecasting systems developed through NCAR will be used by Xcel Energy control centers in Denver, Minneapolis and Amarillo, Texas.

The initiative builds on a prior pioneering wind energy forecasting system developed at NCAR and credited with saving millions of dollars for Xcel customers in eight states around the country.

"We can improve the nowcasting capability," said NCAR program director Sue Ellen Haupt. "We previously optimized for the day-ahead forecasts, and that is especially good for the (utility's) trading, where the day before, they can trade power on the market.

"But now they are more interested in the ramp forecasting. Whenever (the wind) is going to increase or decrease over a short period of time, they need to know that, so they can bring down other units, such as the coal units, and allow the wind to be used."


Xcel Energy spokesman Gabriel Romero termed the utility's partnership with NCAR "extraordinary" and emphasized its importance in enabling the utility to boost its commitment to wind energy.

With a current wind-generated capacity within Colorado of 2,170 megawatts, that dwarfs the utility's capacity for energy drawn from other renewables. Its in-state capacity on solar energy is just 250 to 300 megawatts.

"What this allows us to do is to be more aggressive when we look at other sources," Romero said. "We are able to power the system using wind more often, and aggressively, than we have in the past.

"The system that we put together with NCAR really allows us to do this. ... Now we have new upgrades that allow us to push the envelope, so to speak, where it comes to renewable energies -- especially wind, of course."

NCAR started designing a wind energy prediction system for Xcel in 2009 that is credited with saving the utility's customers $6 million in 2010 alone. The system relies on a suite of tools including highly detailed observations of atmospheric conditions, a series of powerful computer models and artificial intelligence techniques to issue high-resolution forecasts for wind farm sites.

The next phase of NCAR's work will include focusing on so-called "ramp" events, major changes in wind energy due to passing fronts or other atmospheric events, through the use of the Variational Doppler Radar Analysis System.

Also, there will be greater focus on predicting ice and extreme temperatures within a 48-hour window that can affect icing of turbine blades, through a partnership with Pennsylvania State University.

Haupt believes it should all translate to more prompt and efficient recognition of weather "regimes," weather patterns which are accompanied by fluctuations in wind and solar energy that can be exploited by -- and prepared for by -- energy users.

Through use of the VDRAS system, she said, along with "some artificial intelligence techniques that we think can really help in the short term in particular, using new metrics, as well as expert systems that identify the regime that is occurring now, I believe we can do a better job of forecasting for it.

"This is pushing the state-of-the-art still further, using the latest science to enable Xcel Energy to generate energy from the atmosphere more effectively," Haupt said.

Contact Camera Staff Writer Charlie Brennan at 303-473-1327 or