How to expand solar power without using precious land

Land is a finite resource, facing huge demands from a growing population that clamors for living space and food. Farming, meanwhile, is beset by soil degradation, water shortages, plummeting biodiversity, and climate change.

Open in viewer

So it’s unfortunate that solar power, an essential solution to climate change, should also be hungry for land. To generate as much energy as a conventional 1-gigawatt power station, an array of solar photovoltaic (PV) panels needs to cover about 80 square kilometers of land. Unsurprisingly, solar development faces increasingly organized resistance from many rural communities and activist groups, who see it as an enemy of farming.

Sign up for PNAS alerts.

Get alerts for new articles, or get an alert when an article is cited.

Yet there's no need for this confrontation. Properly designed solar installations can increase food harvests, reduce the need for irrigation, revive dying lakes, rescue pollinators, restore soils, and cool overheated humans—all while producing more power than conventional solar arrays.

That’s the promise of a wave of projects that aims to expand solar power without taking useful land out of commission. Symbiotic solar installations on farmland, lakes, and parking lots could enable solar to supply a large fraction of the world’s energy needs sooner than would otherwise be possible. “This can grease the skids for solar, by reducing conflict between food and energy,” says Greg Barron-Gafford, a plant ecologist at the University of Arizona.

These approaches still face a range of obstacles, including cost, convenience, and the need for collaboration between farmers and developers. But the signs are promising—researchers are developing symbiotic solar systems that are cheaper and more efficient, while governments are beginning to plow serious cash into the field. These efforts raise the possibility that symbiotic approaches might become commonplace.

Solar Farming

Solar power and farming often compete for the same precious land. It costs about $1 million to install a mile of electricity transmission lines, so most new solar power arrays are close to cities, where residents and industries need the power. But that puts solar installations in prime agricultural territory.

In 1982, researchers at the Fraunhofer Institute for Solar Energy (ISE) in Germany proposed a stunningly simple solution: set solar panels a few meters above the ground, and grow food underneath. Their original sketch shows angled panels with fairly large gaps in between, so the crops still get plenty of sunlight (1). This concept began to bear fruit in the early 2010s, with field trials in Japan and Europe. Japan now has about 3,000 farms with small solar installations set up on stilts, which are financially supported by government funding and known as solar sharing. In the United States and Europe, this idea is usually called agrivoltaics (AV), and it comes in a mouth-watering array of varieties.

The simplest approach is to plant grass under the panels and unleash some sheep. The United States already has more than 15,000 acres of solar grazing, including a huge 4,700-acre site at Topaz Solar Farm in California. The sheep gain shelter from the panels, and it saves on the cost of cutting the grass. With an eye on improving biodiversity, other projects plant native vegetation beneath their panels to support pollinating insects. This can also restore soils that have been depleted and compacted by decades of intensive farming, locking up carbon from the atmosphere. Both of these are low-maintenance options, and they work with panels set less than 1 meter above the ground, which keeps installation costs down.

Greater benefits can come from combining solar with food crops. The solar panels must be mounted higher up to allow workers and machinery to access the crops, making the setup more expensive. But this approach can help to offset those costs by boosting harvests.

In 2016, for example, Barron-Gafford’s team started an AV project growing cherry tomatoes, chiltepin peppers, and jalapeños—“things to make salsa, because if all else fails, you can still eat the science,” he says. The researchers found that the panels kept plants cooler during the day and warmer at night, and they held more moisture in the air. These less-stressed plants produced just as many jalapeños, twice the crop of tomatoes, and three times the amount of chiltepin peppers as those on a control plot (2). They also needed substantially less watering, a key benefit in a time of worsening water shortages around the world. Water evaporation from the plants even helped to cool the panels and increase electricity output.

Making Light Work

Some of the best crops for AV systems include root vegetables and leafy greens, which grow larger leaves in shady conditions. A US-wide study called InSPIRE (Innovative Solar Practices Integrated with Rural Economies and Ecosystems) has spent the past 7 years studying dozens of AV installations to provide a robust evidence base to guide crop selection, including what grows well under various climates and designs. “We had bits of AV starting to blossom in different parts of the country, but we don’t have time for academics to putz around in their own worlds; we need to work together,” says environmental scientist Jordan Macknick at the National Renewable Energy Laboratory (NREL) in Golden, Colorado, who leads the InSPIRE team.

The project’s first report, which came out in August, shows that details matter—some varieties of a given species do much better than others in given climate and soil conditions, for example (3). But the report also emphasizes that AV is most successful when local communities and solar developers collaborate, in order to ease legal difficulties and create installations that work for farmers, rather than getting in their way.

AV must also be tailored to local climate, landscape, and soil—for example, by spacing panels more widely in darker latitudes to let in more light. To fine-tune these designs, Alexis Pascaris at the Michigan company AgriSolar Consulting is working with Colorado developers Sandbox Solar to create a software tool called SPADE, which will calculate light conditions for a given location and panel configuration and then identify crops that could thrive there.

Although AV generally offers higher crop yields in sunny, water-stressed areas, it can reduce yields in damper, darker climes. Even so, AV can put less pressure on land than separate farm and solar installations. For example, if a hectare of AV in Germany produces as much food as 0.8 hectares of conventional farming, and as much power as 0.7 hectares of conventional solar, it still saves half a hectare of land.

The biggest AV rollout is happening in China, which hosts 30 gigawatts of capacity across 80,000 sites. But many of these power stations were not well designed for agriculture, says optoelectronic engineer Wen Liu at the University of Science and Technology Beijing, and some people have stopped farming at the sites. While the government is now pursuing stricter standards for AV projects, Liu is developing new designs to boost plant growth. In his Even-lighting Agrivoltaic System (EAS), grooved glass plates are set in between the solar cells (4). They refract sunlight onto the plants, providing uniform illumination that ensures crop yields are at least as high as in open fields.

Open in viewer

AV’s food- and water-security benefits are likely to be crucial in the developing world, where rising temperatures and increasingly erratic rainfall are expected to hit crop yields over the coming decades. “These challenges face rural communities that are also seeking electrification,” says environmental scientist Richard Randle-Boggis at the University of Sheffield in the United Kingdom. “About half of Africa’s population have no access to electricity, and if they do have power, it is diesel-generated, vulnerable to blackouts and price changes.”

Randle-Boggis is part of a team that has set up AV test sites in Kenya and Tanzania over the past 2 years. Both sites have shown increased yields of cabbages, helping to pay back investment costs faster than conventional PV installations. The panels provide welcome shade for farm workers, and with added guttering pipes and a tank, they also collect rainwater for irrigation (5). The team is now trying out other crops, including butternut squash, sweet potato, and spinach.

Floating a Solution

Solar has a second escape route from the clash with farming: It can take to the water. Pioneered in Japan, this approach is known as floating PV, or floatovoltaics. Placing panels on water cools them during the daytime, making them more efficient than land-based PV. When set on hydroelectric reservoirs, floatovoltaics can plug into existing transmission lines and reduce evaporation, so that there’s more water for the turbines. Or it can benefit open-water aquaculture by providing power for fish and seafood farms, while sheltering the residents.

Floatovoltaics has started to catch on over the past decade, especially in Southeast Asia. The largest installation, on a reservoir in Shandong province, China, has a capacity of 320 megawatts. Take-up has been slower in the United States, although the US Army launched a 1.1-megawatt installation at Fort Bragg in North Carolina in 2022, to help improve energy resilience and security.

Some floatovoltaic plants look like ordinary ground-based PV, with panels propped up on steel supports that are fixed to pontoons. Other systems mount panels on rigid polymer floats or thin flexible membranes, which can rotate to track the sun. Engineers say that the systems should be robust enough to withstand stormy weather. Overall costs tend to be around 15% higher than conventional ground-mounted PV, says environmental scientist Konstantin Ilgen at the Fraunhofer ISE, where researchers are testing various designs for cost, durability, and impact on aquatic ecosystems.

Using lightweight floats might help to lower costs. Materials engineer Joshua Pearce, at Western University in London, Ontario, Canada, is sticking foam onto the back of thin, flexible solar panels to make them buoyant. As well as being potentially cheaper, these lightweight structures could have an exceptionally low carbon footprint [11 kilograms of CO₂ per megawatt-hour generated, according to Pearce’s calculations (6)] and can be self-cleaning. “We set them just above the water, and in field testing, small waves clean them off,” Pearce says.

Floatovoltaics may be most appealing in hot, dry places because of its ability to reduce evaporation from lakes. Walker Lake in Nevada, for example, has lost so much water and become so salty that fish can no longer survive there. Pearce and his Western University colleague Koami Soulemane Hayibo have calculated that covering half of the lake with foam floatovoltaics would prevent more than 50 million cubic meters of water from evaporating each year—about a quarter of the total annual water loss—which would help restore the lake to health (7).

Design Dilemmas

After a slow start, AV and floatovoltaics are now growing fast. The total installed capacity of AV reached 14 gigawatts in 2021, while global floatovoltaic capacity is an estimated 5.2 gigawatts. Many researchers now think that symbiotic solar could eventually supply a large fraction of our energy needs. “I think any PV project built on land that supports vegetation could incorporate agrivoltaics in some way,” Macknick says.

A study led by environmental engineer Chad Higgins at Oregon State University suggests that putting panels on around 1% of all agricultural land would generate as much energy as the world uses today (8). Even in relatively cloudy countries, it could be a big part of the energy mix. “I think AV could reach 0.5–1% of agricultural land in Germany by 2040 to 2050,” says agricultural economist Max Trommsdorff at the Fraunhofer ISE in Freiburg. That could produce roughly one-third of the country’s total energy requirements today.

Meanwhile, NREL researchers have calculated that freshwater floatovoltaics covering just 4% of the area of manmade water bodies could replace about 10% of existing US power generation (9). The study ruled out all natural lakes and any reservoirs associated with recreation, navigation, or fishing and assumed that only 27% of each eligible body would be covered. “That is conservative to the point of absurdity,” Pearce says. “If we needed to, we could provide all of our power needs with floating PV.”

For now, though, symbiotic solar capacity is dwarfed by the 168 gigawatts of conventional PV installed in 2021 alone. In the United States, crop-based AV remains a niche market: The largest project covers a 10-acre blueberry field in Maine. To fulfill its potential, AV needs to become commercially viable on large mechanized farms—and it’s not quite there yet.

“Typical agrivoltaic design makes harvesting difficult,” Macknick explains. Often, the solar hardware gets in the way of agricultural machinery, with panels set too low or too close together. Solar developers need to collaborate closely with farmers to avoid this, he says. At Chatfield Farms on the outskirts of Denver, Colorado, InSPIRE is building a site designed to be farmer-friendly. Panels are 3 meters up, allowing access for most tractors, and as much as 7 meters apart.

Unfortunately, this arrangement increases costs. Even at Jack’s Solar Garden in Colorado, where the panels are only 2 meters above ground, the extra steel added about 10% to the installation cost compared with a conventional PV array, Barron-Gafford says. To be tractor-friendly, a Fraunhofer research project near Lake Constance in Germany sets the panels 5 meters up, with 19 meters between supports. This setup is intended to study the trade-offs between energy production, crop performance, and farmers’ convenience and would be much too expensive for commercial AV, Trommsdorff admits. “But there are a lot of potential solutions,” he adds.

Vertical panels could work. These “solar fences” don’t get in the way of machinery, and they don’t need much cleaning, but they tend to produce less power per hectare than overhead AV. Alternatively, the Agrivoltaico system, developed by Italian company REM TEC energy, saves on steel by using tensioned cables instead of bulky girders to provide structural strength, a little like the guy ropes that anchor a tent. The company has set up trial Agrivoltaico plants in the Po valley and says that the capital cost is about €1.5 million per megawatt. Although that’s much more expensive than ground-mounted PV, the system produces more power because its elevated design lets it track the sun more effectively. According to a modeling study by environmental scientist Alessandro Agostini, based at ENEA in Rome, over a 25-year lifetime, the Agrivoltaico system could generate electricity at a cost only a few percent higher than conventional ground-mounted PV (10).

Cabbages and Cash

In principle, the high costs of installing AV could be recouped by revenue from crops. But solar developers tend not to factor this in—their businesses are focused on making profits from power. “They’re not going to start selling cabbages,” Pascaris says. The diverse nature of agriculture means that there is no standardized business model to divide profits between developers and farmers. Nor are there cash rewards for enhancing ecosystem services, such as preserving groundwater and encouraging pollinators.

Government subsidies can help AV overcome the capital-cost barrier: Germany guarantees above-market rates for AV electricity, for example. And there are promising signs in the United States, where the US Inflation Reduction Act, passed in August 2022, includes $161 billion in clean energy tax credits. This has energized funds for both clean energy and agricultural resilience, Pascaris says, “so it could be a catalyst for agrivoltaics.” On a smaller scale, the US Department of Energy announced funding of $8 million for AV in May 2022, and Colorado has its own AV stimulus fund. Europe is stepping up too, with the Italian government investing more than €1 billion in AV as part of its COVID-19 recovery plan. Funding aside, speedier planning approval would also help, says Barron-Gafford, pointing to the many years it took for Jack’s Solar Garden to get the go-ahead.

AV also suffers an awareness problem, Barron-Gafford adds: “Most people don’t even know this is an option.” When the message finally gets out there, perhaps symbiotic solar will cool the confrontation between farming and clean energy—and make a significant contribution to our zero-carbon aspirations.