GE recently sold the first of a new line of “hybrid” wind turbines that comes with a battery attached. The turbine’s battery can store the equivalent of less than one minute of the turbine operating at full power. But, by pairing the battery with advanced wind-forecasting algorithms, wind farm operators could guarantee a certain amount of power output for up to an hour.

This could make integrating intermittent renewable energy far easier, and lower the cost of wind power. Indeed, even relatively small batteries could double the amount of renewable energy the power grid can handle.

The true cost of renewable energy will, to some extent, depend on how much energy storage is needed (see “Better Computer Models Needed for Mega Wind Farms”). Much attention has been on how to develop extremely cheap batteries to make it possible to, say, store wind power at night for use when it’s needed during the day. But depending on how fast the renewables are deployed and the nature of the wind and solar resources in different areas, such large batteries often won’t be needed for decades. Small and more affordable batteries can make it possible to use renewable energy for a large fraction of the power on the grid.

GE isn’t alone in recognizing that even small amounts of energy storage could have a big impact. One of the biggest problems with renewable energy is that the speed with which it can increase and decrease—as clouds obscure the sun or the wind drops—can be far faster than conventional power plants can change their output to compensate. But even small batteries can respond to these changes almost instantly.

When relatively little renewable energy is on an electricity grid, fluctuations in supply can easily be absorbed without any battery storage. As the fraction of renewable energy increases, however, existing methods for coping with variability will have an increasingly difficult time keeping up, which could lead to voltage problems on the grid or even blackouts. The exact point when renewables become a problem depends on a slew of factors. But in many cases, renewables can account for about 20 percent of the grid capacity without causing trouble.

In certain cases, increasing that limit to 30 to 40 percent could be as simple as adding 15 minutes of battery storage to wind farms, says Richard Fioravanti, vice president of storage applications at DNV KEMA, an energy consultancy. If batteries can compensate for changing power levels from a wind farm for 15 minutes, they can give grid operators time to ramp up (or lower) power from conventional power plants. “It’s a great way to catalyze larger amounts of renewables,” he says.

According to GE, you don’t need to store 15 minutes of power to guarantee power output for 15 minutes. For a 2.5-megawatt turbine, that would require a 625 kilowatt-hour battery. If you have good wind forecasting algorithms, just 25 kilowatt hours of storage—comparable to a battery in an electric vehicle—is enough to guarantee power output for 15 to 60 minutes, says Keith Longtin, general manager for GE’s wind product line.

GE uses proprietary algorithms to predict the power output of its wind turbines to a high degree of accuracy. Then it uses the battery to make up for the differences between its prediction and actual power output. If the wind turbine generates too little power, a quick 50-kilowatt boost could be enough to match the forecast.

GE hasn’t given a precise figure, but says it’s a small fraction of the millions of dollars each turbine costs. Similarly sized batteries for electric vehicles cost well under $20,000.

The battery could pay for itself in a number of ways, such as reducing fees that some utilities charge wind farm owners and generating revenue even when the wind isn’t blowing. It could also provide revenue-generating services to the grid, like delivering bursts of power when needed.

The true cost of renewable energy should also factor in the cost of having conventional power plants standing by to ramp up. That can be about one cent per kilowatt-hour, according to one estimate. (That’s a substantial increase—for reference, new natural gas plants in the United States are expected to generate electricity at 6.5 cents per kilowatt, on average, over their lifetime.)

Haresh Kamath, program manager for energy storage at the Electric Power Research Institute, says that in some areas, 15 minutes of energy storage won’t be enough—for example, where the mismatch between demand and renewable energy output or the amount of variation in the renewable energy is too great.

Achieving 100 percent renewable energy would require hours of battery storage and far cheaper and better batteries than we can build now. What’s more, there can be huge seasonal variations in renewable energy power output. Dark winters and summer doldrums would require not hours, but days, weeks, or even months of storage. (Other forms of energy storage and long-distance transmission lines to connect widely dispersed wind farms will also help—see “A Solution to Solar Power Intermittency” and “Supergrids.”)

But it will be possible to reach relatively high levels of renewable energy with a small amount of battery backup, providing time for new battery technologies to be developed. However, even at rapid paces of development, it will be decades before renewable energy exceeds 40 percent of generating capacity in most places.