The morning clouds had just burned off at 11 a.m. last Friday when Los Angeles Mayor Antonio Villaraigosa and former Vice President Al Gore strode to a lectern outside the city Department of Water and Power’s headquarters. The air was clear and cool, so the utility’s 1.3 million customers weren’t cranking their air conditioners and the utility didn’t need to generate anything near its total electrical capacity of 7,200 megawatts, which is almost as much as Ireland can produce. Villaraigosa announced that by 2025, the city would no longer get any of its electricity from coal—a dramatic change from just a few years ago, when half of L.A.’s power was coal-fueled. The city has already quadrupled its renewable energy use. California enacted a renewable portfolio standard in 2002 that ordered all utilities in the state to get a third of their electricity from wind, solar, or hydropower by 2020. Twenty-eight other states and the District of Columbia have similar requirements. But as states ramp up their use of renewables, they’ll run into a problem nobody at the lectern talked about: energy storage. Wind turbines and solar panels produce energy intermittently, often when the grid doesn’t need it. Much of the energy they make has to be put somewhere until demand rises. The energy storage solution now most commonly used is pumped-storage hydropower: Facilities send water up a hill when the grid is producing excess power, store the water behind a dam, then release it through a turbine when demand rises. But the system requires a lot of water, and water tends to be scarce where sun and wind are abundant. What’s more, all the good spots with the right topography in the United States are already taken. As of now, no proven and available energy storage technology can affordably meet all the demands of the electricity grid of tomorrow. “As solar expands in cities, we’re going to be in a heap of trouble if we don’t figure out energy storage,” LADWP Commissioner Jonathan Parfrey told me. Panama Batholomy, an adviser to the California Energy Commission, agreed: “It’s going to take a rotary phone-to-iPhone type of leap.” In February, the California Public Utilities Commission ordered Southern California Edison to contract for 50 megawatts of energy storage. It will be the first time a utility puts out a procurement request for energy storage—and it’s a request for a technology that doesn’t yet exist in a form that meets users’ needs. But a few companies are working on it, and some of their ideas are a mix of brilliant and bizarre. The ideal energy storage solution would have five qualities: It would put a lot of energy in a small space; it would be inexpensive; it would lose in transfer less than a fifth of the energy put into storage and taken back out; it would last decades; and it would release the energy quickly. The optimal energy storage technology would also be safe to transport and non-toxic to dispose of, as well as made of raw materials that can be obtained without causing major environmental damage. The solutions currently considered most viable to expand energy storage beyond pumped-storage hydropower (which accounts for 93 percent of energy storage worldwide) are batteries, flywheels, compressed air, and one brilliantly simple technology involving boxcars full of gravel. Compressed air works like this: Electricity drives a pump to pack air into a tank. As the molecules become more densely packed, they heat up. The heat is later converted back into electricity. The problem is that transfer is inefficient. Danielle Fong, co-founder of a Berkeley company called LightSail Energy, told Wired.com last year she’s invented a system that can get the efficiency up to 70 percent. (Randy Howard, LADWP director of power system planning and development, would like to see efficiency of between 85 and 90 percent.) Her prototype has impressed even investors skeptical of clean tech and attracted a fresh $37 million round of financing in November. Flywheels convert electricity to kinetic energy and back. Certain kinds can be up to 85 percent efficient, and they can run for decades with very little maintenance. But flywheel systems are expensive, becoming cost-effective only over a 20-to-30-year time horizon. Temporal Power, of Ontario, Canada, claims it has a technology that reduces energy losses; its first megawatt-size project is just getting off the ground. Batteries’ main problem is that they haven’t conquered energy density. In Lancaster, Calif., a desert city with far more watts of solar per capita than anywhere else in the state, I visited a new house built by KB Homes (the company that virtually invented urban sprawl) with solar panels on the roof and a refrigerator-size lithium-iron phosphate battery from Chinese manufacturer BYD in the garage. When the solar panels are producing more electricity than the house needs, the power charges the battery. When the grid needs electricity, it can pull power out of the battery, and the homeowner’s meter runs backward. Lithium-iron phosphate batteries have a lower energy density and slower discharge duration, but BYD America’s vice president, Micheal Austin, said they were the clear choice. “There’s no battery as fire-safe as iron phosphate.” Other types would effectively become a bomb in the event of a house fire. Nickel-metal hydride batteries (the type of rechargeable battery sold for small consumer electronics) and lead acid batteries (the ones in gasoline-powered cars) become toxic waste when they run out of charge cycles. The garage deployment could become part of a complex system of distributed energy storage that would include electric-vehicle charging stations, the vehicles themselves, and other grid-connected batteries. Such a network could meet much of utilities’ needs without a huge technological leap or major new transmission lines—which, as I wrote in Slate earlier this month, can be a nightmare to build in the United States (and which have their own efficiency issues). Jim Kelly thinks he has the energy storage solution. In his 38 years in various R&D and engineering executive positions at Southern California Edison, Kelly built several pumped-storage hydropower facilities. Next month, on a ranch in the Tehachapi Mountains owned by one of the founders of the wind energy industry, Kelly’s company, Advanced Rail Energy Storage, will begin testing a variation on pumped hydro. Except instead of dams, channels, and water, Kelly’s new system has rail yards, train tracks, and electric locomotives hauling boxcars full of gravel. These heavy-haul trains, borrowed from mining applications, use the same software as computerized trains at many airports. A motor hooked up to an electric third rail draws electricity from the grid to push the trains up a 7 to 8 percent slope; at the top, the energy is stored as potential energy. When the grid needs the watts back, the software allows the trains to run downhill at about 35 miles per hour, “releasing energy all the way,” Kelly explains. The locomotive’s motor becomes an electric generator, pushing the electricity back into the electrified rail and from there, to the grid. A large-scale storage facility that could handle 500 megawatts or more would take about 8 miles of track. The heavy boxcars are connected and disconnected according to how much power is being stored or sent back. The trains can store the power for an hour, a week, or a month with no loss over time—gravity doesn’t decay. And Kelly says they can achieve up to 90 percent efficiency. DWP’s Howard said that Kelly’s idea sounds “intriguing” and thinks it could work. It would still need transmission, though, which means that even if everything goes just right next month at the test site, utilities will likely continue their quest for a solution with high efficiency and energy density, low discharge time and cost, that’s safe when you’re using it and nontoxic when you’re done with it. If you have any ideas, let them know.