Wind Energy Forecasts Go Skyward for More Accuracy

Real-time wind speed data measured up high could help independent system operators manage the U.S. power grid more efficiently.

by / November 16, 2010
Montana wind turbines stock photo

In a strange melding of technologies, cell phone towers are helping to make wind energy forecasts in Texas more accurate and predictable.

Niwot, Colo.-based wind data provider Onsemble recently announced the completion of a wind data network that tracks real-time wind speeds, direction and temperature from sensors placed on cell phone towers that are approximately 260 to 320 feet above ground — the same height as many wind turbines.

The sensors will track data throughout the Electric Reliability Council of Texas (ERCOT) region, which manages most of the state’s electricity grid and covers 95 percent of wind farms in Texas. The Public Utility Commission of Texas oversees ERCOT.

The project is part of the nationwide effort to better predict wind turbines’ electricity output, which typically is more unreliable and intermittent than other energy sources.

The sensors collect one-minute averages of wind data and send the information to a central hub every 10 minutes in order to predict a wind energy “ramp event” — when a large influx of wind energy will be introduced to the electricity grid. The system can predict such an event 12 to 24 hours before it happens.

This improved forecasting could help Texas — which according to private-sector data produces one-third of the country’s 45,000 megawatts of wind power — run the wind farms more efficiently and make better financial decisions, partly because prices in the Texas electrical grid update every 15 minutes based on the supply and demand of the prior 15 minutes. Consequently more accurate and frequent readings should equal more accurate prices. And if operators know when a boost of wind-powered energy is about to happen, they can reduce their reliance on reserve power sources, such as fossil fuels.

Most nationwide wind data used by the power industry is still collected from about 10 meters above ground. The readings are typically farther away from the source, so it’s not as accurate.

Government Involvement

One of the obstacles to implementing the more accurate systems was an assumption that governments would have to pay for the new sensor networks because there wasn’t a business model to back private funding. But according to experts, that’s no longer the case.

“With the recent development of wind energy and solar energy, and the need to more accurately predict the intermittency of those renewable energy resources, there is now a business case to be made for obtaining this data at those levels,” said Anish Parikh, co-founder and vice president of Onsemble.

In addition to ERCOT, Onsemble is currently operating a real-time data network in the Bonneville Power Administration, a federal agency in the Northwest; the Public Service Company of Colorado; and the Southwest Power Pool utility markets. A nationwide build-out of the network is planned through 2012.

Once the data is collected, it’s sold to agencies that makethe forecasts. Generally packages are sold as yearly subscriptions to either real-time or archived data sets. For government customers, prices are driven by the data package’s size and how frequently it’s updated.

The improved network is just one of many efforts to build out the country’s wind power data. Last month, IBM announced a partnership with two private wind-data collection companies to use the company’s existing software to develop an automated wind control development platform. In 2009, a private company funded Boston-based WindPole Ventures to install wind-measuring sensors on 1,150 communication towers over the next 22 years, although the company confirmed that only 600 were suitable for use.

Lauren Katims Nadeau

Lauren Katims previously served as a staff writer and contributing writer for Government Technology magazine.