On Aug. 21, when afternoon temperatures in Washington state soared, Avista Corp.’s (AVA) utility division asked customers to cut their electricity use while it scoured the region for power supplies. Utility operators were frustrated, in part, because the company’s supply of wind power was producing nothing, thanks to a lack of wind.
Avista wasn’t alone. Throughout the West during that August heat wave, a growing fleet of windmills met triple-digit temperatures with impotence. California’s grid operator was serving up a record amount of power that afternoon, too, while its 2,850 MW of wind turbines were churning out just 112 MW.
“Typically in the Northwest, very hot weather means high pressure, and high pressure means no wind,” said Steve Silkworth, who is in charge of buying renewable energy for Avista Utilities.
As investment in wind power grows and the nation increases its dependence on wind as a power resource, utilities and grid operators are trying to figure out how to manage its volatility.
Windmills deliver pollution-free electricity and reduce dependence on fossil fuels. They can be built quickly, and their costs have plummeted, while the technology has improved. But wind generally blows less in the afternoon and more at night, when much less power is needed and prices are lower. At utilities’ critical moment – a heat wave – winds tend to die.
These general trends aside, output from windmills is wildly unpredictable. Averaged over the course of a year, windmills produce about a third of their capacity, but actual output jumps from 0% to 100%, sometimes in a matter of hours. So they require backup generators, which usually burn natural gas and must be paid for – even to sit idle.
Wind output has always been intermittent, but historically it made up only about 1% of U.S. power supply so fluctuations were easily tolerated. But many utilities now plan to use wind for 10%-20% of their generating capacity, so the volatility is becoming critical. A sharp increase in wind energy has at times destabilized the power systems of Germany, Denmark, Spain and Portugal, according to a recent study by ABS Energy Research in London.
“What level is manageable, and what are the challenges in getting to that level? That’s the top assignment right now,” said Bill Bojorquez, director of system planning for Ercot, the grid operator for Texas.
The shift to alternative sources of power has been spurred by federal tax breaks, renewable energy quotas in 22 states, and a growing expectation that the country will move to curb emissions of greenhouse gases. Because it’s relatively inexpensive, wind power has become by far the dominant choice among the alternatives, which include solar panels, biomass generators and geothermal geysers.
In the territory controlled by PJM, the grid operator for mid-Atlantic and some Midwestern states, four new wind turbines are under construction for every windmill currently in operation. For every one under construction, permit applications have been filed to build 9 additional windmills.
What can all these windmills do to help prevent a blackout in a heat wave? Utilities’ estimates of that range widely. When figuring out how to keep the lights on during the coming summer’s hottest day, PJM pencils in 20% of wind capacity for serving peak load. The California Independent System Operator figures 5% will be there. For Texas, which has more windmills than any other state, Ercot counts on just 2.6% of capacity. Avista, like many utilities operating their own grid, doesn’t count on any wind power during the summer peak. As more windmills come on line, overestimating could mean a blackout, while underestimating could mean paying a lot of money for unneeded standby generators.
“There’s no question we can handle it,” said Joe Karp, attorney for the California Wind Energy Association. “Denmark gets 50% of its electricity from wind. There is a cost associated with it. Maybe at some point we don’t want any more wind than some particular amount.”
PJM, for example, doesn’t see any near-term problems, because it operates such a big system, said Joseph Kerecman, PJM’s manager of alternative generation. If a fourth of the windmills now in the permitting stage get built, PJM will see wind power go from 353 MW to 5,000 MW, but that’s still only 3% of PJM’s system.
More Wind, New Challenges
But wind’s impact on electric reliability in Texas and the western U.S. is more immediate. Most studies suggest an electric system can have about 10%-15% of its capacity in wind with moderate operating issues, according to Ercot’s Bojorquez. Texas could get to the 10% level as early as 2008 or 2009, he said.
“We know at 5,000 MW we’re going to be okay. We don’t know at 10,000 MW what challenges we’ll have, but the legislature says that’s the goal,” Bojorquez said in an interview. “I think we’ll go further.”
The Utility Wind Integration Group, formed by the three main U.S. utility associations, found that power systems can handle up to 20% wind capacity. Even that level isn’t too far in the future for California. The ISO expects its wind capacity to grow to 12,500 MW by 2012, which would be about 20% of capacity.
California has an advantage in that its wind power comes from five different areas, with different wind patterns, that tend to even out each other’s fluctuations, said David Hawkins, the ISO’s manager for integration of renewable resources. That’s not exactly how it worked out most of the afternoons during the July heat wave, though.
“Some of the hottest days, it wasn’t there for us,” said Hawkins. But he added that power systems don’t rely on wind for peak electricity demand. “Wind is an energy resource, not a peaking resource.”
In the Northwest, the Bonneville Power Administration, which operates the federal government’s huge system of hydroelectric dams in the Northwest, recently ceased offering a service that supplied extra power to utilities when their windmills slowed. Too many windmills are being built for Bonneville to do that anymore.
By Mark Golden
A Dow Jones Newswires Column
(Mark Golden has reported on electricity markets and policy for eight years.)
|Wind Watch relies entirely
on User Funding