Solar farms are proliferating on undeveloped land, often harming ecosystems. But placing solar canopies on large parking lots offers a host of advantages — making use of land that is already cleared, producing electricity close to those who need it, and even shading cars.

Despite the green image, putting solar facilities on undeveloped land is often not much better than putting subdivisions there. Developers tend to bulldoze sites, “removing all of the above-ground vegetation,” says Rebecca Hernandez, an ecologist at the University of California at Davis.

The appeal of parking lots and rooftops, by contrast, is that they are abundant, close to customers, largely untapped for solar power generation, and on land that’s already been stripped of much of its biological value.

A typical Walmart supercenter, for instance, has a five-acre parking lot, and it’s a wasteland, especially if you have to sweat your way across it under an asphalt-bubbling sun. Put a canopy over it, though, and it could support a three-megawatt solar array, according to a recent study co-authored by Joshua Pearce of Western University in Ontario. In addition to providing power to the store, the neighboring community, or the cars sheltered underneath, says Pearce, the canopy would shade customers — and keep them shopping longer, as their car batteries top up. If Walmart did that at all 3,571 of its U.S. super centers, the total capacity would be 11.1 gigawatts of solar power — roughly equivalent to a dozen large coal-fired power plants. Taking account of the part-time nature of solar power, Pearce figures that would be enough to permanently shut down four of those power plants.

One reason such facilities are still scarce is that building solar on developed land can cost anywhere from two to five times as much as on open space. For a parking lot canopy, says Pearce, “you’re looking at more substantial structural steel with a fairly substantial concrete base.” It’s like putting up a building minus the walls. For a public company fixated on quarterly results, the payback time of 10 or 12 years can also seem discouragingly long. But that’s the wrong way to look at it, says Pearce. “If I can give you a greater-than-four-percent return on a guaranteed infrastructure investment that will last for 25 years minimum,” that’s a smart investment.