There will still be times during the year when poor weather forecasts and high demand lead to drops in power that are too big to compensate with supplemental power plants. But according to NREL’s models, this will happen rarely–just 89 hours out of the year, or 1 percent of the time. For such a small amount of time, it’s not economical to build backup generators to make up the difference. But NREL found that a strategy called demand response could keep the grid from collapsing under the pressure. It’s already been used in Texas, for example, to make up for sudden drops in wind power.
In demand response, utilities send out signals to customers–typically businesses–asking them to cut their power use, in return for favorable electricity rates. That cuts demand enough to balance the grid. “It’s effective and a lot less expensive” than building backup power plants, Lew says.
Distributing renewable energy among several states would not require extensive new interstate transmission lines at first, Lew says, because in the West there is already enough transmission available–if operators are given ready access to spare capacity. No new interstate transmission will be needed to accommodate wind farms that supply up to 20 percent of the electricity in a year and solar power that provides up to 3 percent, although there will need to be additional power lines to convey power locally from new renewable power plants, which might be in a remote area.
This is in contrast to the eastern United States, where long-distance transmission would likely need to be built to carry power from the Midwest, where it is windy but electricity demand is low, to cities on the coast, where demand is high, she says. In the West, reaching 30 percent wind and 5 percent solar will require more transmission, but this could easily be built in the time it would take to reach such high levels of renewable energy.
In addition to sharing power over large areas, for example, utilities could significantly reduce variability by changing their scheduling practices. Utilities now use algorithms to predict demand a day ahead and then schedule hourly changes to power generation. These changes on the hour, when a utility might cut down the amount of power a coal plant generates in anticipation of lower demand, actually introduce some instability into the grid. Scheduling changes more frequently could smooth this out, and allow more accurate coordination with renewable energy resources. Scheduling more often isn’t a technical challenge, although it could require software upgrades–it’s mostly a matter of changing practice, Brooks says.
But utilities, notoriously slow to change, are showing signs of responsiveness in the face of renewable energy mandates, Brooks says. The need is urgent because some areas of the country already rely on relatively large amounts of wind power, at least part of the time. In some places in Texas, wind power, at times, can account for a quarter of the power on the grid, although it still only accounts for just over 6 percent of the total electricity in a year. In the Bonneville Power Administration in the Northwest, during some short periods of time, half of the power being generated comes from wind. “Utilities see more and more solar and wind coming on line and they’re scared that they can’t manage this on their own,” Lew says.