To see the big obstacle confronting renewable energy, look at Denmark. The small nation has some of the world’s largest wind farms. Yet because consumer demand for electricity is often lowest when the winds blow hardest, Denmark has to sell its overflow of electrons to neighboring countries for pennies—only to buy energy back when demand rises, at much higher prices. As a result, Danish consumers pay some of the highest electricity rates on the planet.

Utilities in Texas and California face a similar mismatch between supply and demand; they sometimes have to pay customers to take energy from their windmills and solar farms. On paper, wind and sun could supply the U.S. and some other countries with all the electricity they require. In practice, however, both sources are too erratic to supply more than about 20 percent of a region’s total energy capacity, according to the U.S. Department of Energy. Beyond that point, balancing supply and demand becomes too difficult. What are needed are cheap and efficient ways of storing power, to be tapped later, that is generated when winds are howling and the sun is beating down.

Certain technologies such as superconducting magnets, supercapacitors and advanced flywheels are too expensive for that purpose or cannot efficiently hold power for extended periods. But Scientific American has examined five technologies that might do the trick. Each of them could possibly store, for days, the amounts of energy needed to keep an entire metropolis humming. We asked a panel of experts to rate each one based on three criteria: How well can the technology scale up? Is it cost-effective to build? Is it efficient to operate? No storage method can return the same amount of energy put into it, yet some systems do better than others.

The first two solutions—pumped hydro and compressed air—are somewhat mature and economically feasible. Each of the other contenders will require some kind of breakthrough, but the payoff could be huge. “Ten years from now I expect that we will see a lot of energy storage on the grid,” says Imre Gyuk, a physicist who manages the DOE’s storage program.


Pros: Efficient, cost-effective, highly reliable
Con: lack of suitable sites
*Average ratings, out of 5, by our panel of experts

Several countries already store considerable power—about 20 gigawatts in the U.S.—using pumped hydro. This century-old technique is essentially a hydroelectric dam that can operate in reverse. Excess electricity is used to pump water from a low reservoir to one higher uphill. When the water falls back down to the lower reservoir, it passes through turbine blades that turn a generator to create electricity. Round-trip efficiency—the energy that can be recovered, minus losses—can be as high as 80 percent.

In the U.S., 38 pumped-hydro facilities can store the equivalent of just over 2 percent of the country’s electrical generating capacity. That share is small compared with Europe’s (nearly 5 percent) and Japan’s (about 10 percent). But the industry has plans to build reservoirs close to existing power plants. “All you need is an elevation difference and some water,” says Rick Miller, a senior vice president at HDR in Omaha. Enough projects are being considered to double existing capacity, he says.

Among the most ambitious plans is the Eagle Mountain Pumped Storage Project in southern California. It would carve two reservoirs out of an abandoned iron surface mine to store energy from regional wind and solar farms, and it could return 1.3 gigawatts of power—as much as a large nuclear power station. In Montana, Grasslands Renewable Energy’s pro­posed hydro storage project would hold wind energy from the Great Plains in an artificial lake that would be built on top of a butte, with a drop of 400 meters.

Pumped hydro’s growth is limited primarily by topography. Large, elevated basins must be flooded, which can damage the ecosystem. Some places, such as Denmark and the Netherlands, are just too flat. For those regions Dutch energy consulting company Kema has come up with a radical “energy island” alternative: an artificial lagoon—in a shallow sea—with a circular wall that would be built from landfill. Excess electricity would pump seawater out of the lagoon and into the surrounding ocean. When energy is needed, water from the sea would flow back inside, through tunnels in the wall, passing through turbines. The ocean acts as the “upper” reservoir.

Gravity Power in Santa Barbara, Calif., has an option that could be deployed almost anywhere: a deep vertical shaft would be dug into the ground, and a heavy cylinder would rest at the bottom. Water would be pumped underneath the cylinder, lifting it. To recover energy, tunnels at the base would open, and the water would rush into them through turbines.


Pros: Cost-effective, tested
Con: May need to burn some natural gas

Deep under the ground in rural Alabama, a cavern half as large as the Empire State Building holds what could be the quickest fix for the world’s energy storage needs: air. Up on the surface, powerful electric pumps inject air at high pressure into the cavern when electricity supply exceeds demand. When the grid is running short, some of that compressed air is let out, blasting through turbines and spinning them. The facility, in McIntosh, Ala., run by the PowerSouth Energy Cooperative, can provide a respectable 110 megawatts for up to 26 hours. It is the only compressed-air operation in the U.S., but it has operated successfully for 20 years. German company E.ON Kraftwerke, based in Hannover, operates a similar plant in Huntorf in the state of Lower Saxony.

PowerSouth created the cavern by slowly dissolving a salt deposit with water, the same process that formed the U.S. Strategic Petroleum Reserve caverns. Salt deposits are plentiful throughout the southern U.S., and most states have geologic formations of one kind or another, including natural caverns and depleted gas fields, that could hold compressed air.

Proposals for compressed-air projects have popped up in several states, including New York and California. Yet recently a proposed $400-million Iowa Stored Energy Park near Des Moines was scrapped because detailed study showed that the permeability of the sandstone that would hold the air was unacceptable.

One practical hurdle is that air heats up when it is compressed and gets cold when it is allowed to expand. That means some of the energy that goes into compression is lost as waste heat. And if the air is simply let out, it can get so cold that it freezes everything it touches—including industrial-strength turbines. PowerSouth and E.ON therefore burn natural gas to create a hot gas stream that warms the cold air as it expands into the turbines, reducing overall energy efficiency and releasing carbon dioxide, which undermines some of the benefits of wind and solar power.

Because these complications limit the efficiency of compressed-air storage, engineers are devising countermeasures. One option is to insulate the cavern so that the air stays warm. The heat could also be transferred to a solid or liquid reservoir that could later reheat the expanding air. SustainX, a start-up based in Seabrook, N.H., sprays water droplets into the air during compression, which heat up and collect in a pool. The water is later sprayed back into expanding air, warming it. SustainX has demonstrated its process in above-ground tanks. General Compression in Newton, Mass., is developing a similar approach for underground storage and is planning a large demonstration plant in Texas. “We don’t need to burn gas, ever,” says president David Marcus.


Pros: Energy-efficient, reliable
Con: Expensive

Batteries may be the ideal storage medium for intermittent power sources, some experts say. They charge readily, turn on and off instantly, and can be scaled up easily. For decades utilities have provided backup power to remote recesses of the grid by stacking up racks of off-the-shelf batteries, including the lead-acid type found in cars. Some companies have experimented with molten sodium-sulfur batteries. Power company AES has installed more than 30 megawatts of lithium-ion batteries in Elkins, W.Va., to back up its 98 megawatts of wind turbines. Yet if batteries are to compete for large-scale storage, their cost must drop considerably.

A battery’s expense is driven by materials—the positive and negative electrodes and the electrolyte that separates them—as well as the process of manufacturing them into a compact package. Radical redesigns may have a better shot at sharply cutting costs than incremental improvements to common battery types.

Donald R. Sadoway, a chemist at the Massachusetts Institute of Technology, is developing one unusual design that he calls a liquid-metal battery. Its promise lies in its simplicity: a cylindrical vat kept at high temperature is filled with two molten metals, separated by a molten salt between them. The liquid metals are immiscible with the salt—“like oil and vinegar,” Sadoway says—and have different densities, so they naturally stack on top of each other. When the two metals are connected via an external circuit, an electric current flows. Ions of each metal dissolve into the molten salt, thickening that layer. To recharge the battery, excess current from the grid runs the process in reverse, forcing the dissolved ions back into their respective layers.

Sadoway has so far made “pizza box–size” batteries in the lab, but he thinks that the design could scale up economically, perhaps even becoming cheaper than the $100 per kilowatt-hour of pumped hydro. Sadoway will not know for sure what issues may arise with scaling until he tries it, but he is enthusiastic because, unlike the painstaking, costly manufacture of traditional batteries, his can be built in bulk simply by pouring the materials into a tank.

A more tried-and-true design is the flow battery. A solid-state membrane inside a container separates two liquid electrodes, which can store a lot of energy. Flow batteries are similar in spirit to a more recent technology nicknamed “Cambridge crude,” which uses nanoparticles as electrodes that are suspended in a fluid [see “Liquid Fuel for Electric Cars,” by Christopher Mims; World Changing Ideas, Scientific American, December 2011].

The flow battery has several advantages. It operates at room temperature, unlike the liquid-metal battery, which must be heated. To scale up, just make larger electrodes or add more containers. A defunct start-up company, VRB Power Systems, installed two flow batteries with solutions based on the metal vanadium—one in Moab, Utah, and one on a small Australian island—before selling its technology to Prudent Energy in Bethesda, Md. Other companies are trying to improve on the idea by making the ion flow across the membrane more efficient. Mike Perry, a chemical engineer at United Technologies Corporation (UTC) in Hartford, Conn., says his company is investing millions of dollars and betting that within five years or so flow batteries can become competitive with gas-fired plants used to satisfy peak utility demand. UTC has focused on vanadium, too, because it is a plentiful and inexpensive by-product of petroleum extraction. Energizer Resources in Toronto is also developing a large vanadium mine in Madagascar, which would ensure supply.


Pro: Can be sited anywhere
Cons: Expensive, Hard to hold energy for long periods

In regions that have steady sunshine, concentrated solar power stations can be an economical way to generate power as well as to store the sun’s energy. Rows of parabolic mirrors focus sunlight on long pipes that run parallel to the rows, heating a fluid such as mineral oil inside the pipe. The oil travels to a building where its heat converts water into steam, which turns a turbine to generate electricity. When the sun goes down, the fluid can be stored in tanks to produce more steam for at least several hours, until it slowly cools.

A number of concentrated solar power stations operate in the U.S. and Europe. To retain heat energy longer, however, Archimede Solar Energy in Italy has built a demonstration plant near the town of Syracuse in Sicily that uses molten salt instead of oils. Molten salt can be heated to nearly 550 degrees Celsius, compared with 400 degrees C for oil, so it can create more steam for more hours after sundown, says Paolo Martini, Archimede’s director of business development and sales. Five cubic meters of molten salt can store one megawatt-hour of energy, compared with 12 cubic meters of oil, Martini says. Solar Millennium in Germany has been operating the sizable Andasol 1 molten salt system in Andalusia, Spain, since 2008. And in June 2011 it achieved the milestone of 24-hour uninterrupted solar-electric generation.

Power from today’s concentrated solar power plants is about twice as expensive as that from a natural gas plant. Yet an industry road map predicts that by tweaking plant designs—including the chemistry of the fluids—and introducing economies of scale, concentrated solar energy could become competitive with natural gas within 10 years. Success might be most likely for plants built in places that rarely see clouds, such as the Sahara Desert.

Of course, excess energy generated from wind farms or other sources can also heat fluids that generate power later on. Thermal storage can involve cold instead of hot, too. Ice Energy, a start-up based in Windsor, Colo., sells systems that produce ice during the night when power is plentiful. During the day the ice melts to feed cooling fluid to HVAC systems for air-conditioning. Some commercial utility customers such as big-box stores are beginning to install the units, thereby lessening demand on the grid for air-conditioning power during the hottest hours of the day.


Pros: Efficient, lightweight
Con: Basic materials breakthroughs still needed

One long-shot method of storing energy would rely on homeowners instead of utility installations. For more than two centuries scientists have split water into hydrogen and oxygen by running an electric current through it. The hydrogen can later be consumed in a fuel cell to generate electricity. The challenge is to both split water and “burn” hydrogen efficiently, without producing too much waste heat.

The efficiency of splitting hydrogen could be much higher if sunlight were used directly, instead of power from the grid, the way plants harness the sun for hydrolysis during photosynthesis. Man-made hydrolytic cells that can do the same have existed for years, but they are inefficient and expensive. Chemists such as Daniel Nocera of M.I.T. and Nathan S. Lewis of the California Institute of Technology have been developing novel materials that could perform better—cobalt-based catalysts in Nocera’s case and nanorods in Lewis’s—but costs remain very high.

Whether one is using electricity or the sun directly, hurdles on the reconversion side are enormous as well. Fuel cells burn hydrogen efficiently, but they rely on expensive catalytic materials such as platinum. A unit that can power a car or light a building can cost tens of thousands of dollars. Thus, scientists are seeking alternative materials. Storing hydrogen adds another difficulty because the gas is explosive and must be liquefied or compressed.

If all these challenges could be overcome, homeowners could have their own small hydrogen power stations on their premises. When the local utility has excess wind or solar energy, homeowners would use it to split hydrogen, which would later power the home when the sun or winds fade. And because hydrogen’s energy density is even greater than that of gasoline, it could one day propel cars and trucks as well, leading to the long-envisioned hydrogen economy.

This article was published in print as "Gather the Wind."