How Much Land Would It Take to Go Solar?
October 7, 2013
A report from Bloomberg New Energy Finance finds that, in 2013, solar photovoltaic (PV) power will add more capacity globally than wind for the first time. The group predicts that the world will build 36.7 gigawatts (GW) of solar PV in 2013, compared to 35.5 GW of onshore and offshore wind combined.
What does the expansion of solar power generation mean for land use? Is PV efficient enough to generate the electricity the world needs, without taking over all of the agricultural land and crowding us all out?
A highly-publicized New York Times opinion piece a few years ago by Robert Bryce of the Manhattan Institute for Policy Research popularized the idea that a transition to renewable energy for electrical generation will consume “vast amounts of natural resources -- most notably, land.” In his op-ed, Bryce cited a study by The Nature Conservancy (TNC) highlighting the problem of “energy sprawl” and its potential effects on the environment. In a Wall Street Journal op-ed, Senator Lamar Alexander invoked the TNC study in defense of nuclear as a land-efficient energy source.
The term “energy sprawl” has been taken up by critics of renewable energy in subsequent years, even though TNC’s Robert I. McDonald, the principal author of the study, wrote that “it’s unsettling sometimes to see the rhetorical uses others have found for this research, often far from its original context in a scientific journal.” McDonald stressed that “increased renewable energy production will have to be one of the ways America begins to reduce its greenhouse gas emissions” and that his report “simply shows that renewable energy production has the potential to take a significant amount of space, particularly biofuel production.” McDonald said that the “energy sprawl” challenge could be overcome through energy efficiency and “proper management.”
In the case of solar energy, a report from the U.S. Energy Department’s National Renewable Energy Laboratory (NREL) has firmed up the understanding of the land-use requirements of solar power plants. A research group led by NREL energy analyst Sean Ong examined data from 72 percent of the solar power plants in use and under construction in the U.S. and found that solar power for 1,000 homes would require on average 32 acres of land. The group found that direct land-use requirements for small and large PV installations averaged 6.9 acres per MW, ranging from 2.2 to 12.2 acres/MW.
Quoted in an agency announcement, researcher Paul Denholm emphasized that the purpose of the study wasn’t to compare land-use requirements for the various competing sources of energy. “The numbers aren’t good news or bad news,” he said. “It’s just that there was not an understanding of actual land-use requirements before this work.” The numbers, in fact, were very much in line with previous estimates for solar land use. These more definite figures can now be used by planners, modelers and analysts in evaluating the impacts of solar energy.Land-Use Comparisons Among Energy Sources
While the NREL researchers emphasized that they were not setting out to compare land-use requirements among energy sources, such research naturally invites comparisons.
In fact, a 2009 study by Vasilis Fthenakis and Hyung Chul Kim of Columbia University already compares lifecycle land requirements of conventional and renewable energy sources, including both direct and indirect impacts. The researchers found that land-use impacts varied, depending on geography and technology considerations. However, their overall conclusions were that
[T]he photovoltaic (PV) cycle requires the least amount of land among renewable-energy options, while the biomass cycle requires the largest amount. Moreover, we determined that, in most cases, ground-mount PV systems in areas of high insolation transform less land than the coal-fuel cycle coupled with surface mining. In terms of land occupation, the biomass-fuel cycle requires the greatest amount, followed by the nuclear-fuel cycle.
Solar PV compared well in the Columbia study, the researchers wrote, because, “Unlike conventional non-renewable technologies, the solar electric-fuel cycle generates electricity without fuel extraction.” Therefore, “the electricity generated from a given area of land cumulates proportionally to the lifetime of the solar power plant.”
Renewable energy in general enjoys some advantages over conventional sources when it comes to land use, Fthenakis and Kim found:
The land use of renewable-energy sources, like PV, wind and biomass, pose distinct features from conventional fuel cycles in that they use land statically. Once the infrastructure of renewable-energy technologies is constructed, there is no need for further extraction of resources. Moreover, PV and wind-power plants can be located on low quality lands (e.g., brownfields), and often be used for multiple purposes (e.g., grazing, shading).
Coal mining, on the other hand, “transforms the existing landscape, destroys the soil, and removes ground vegetation, all direct effects of land use. Furthermore, the usage of materials and the energy for operating coal mines and building infrastructures requires additional land during the upstream processes -- indirect land use.” Similar land-use effects arise with natural gas and nuclear power, both of which require extraction operations: “fossil- or nuclear-fuel cycles continuously must transform some land in search of fuels.” In addition, a nuclear plant requires a fairly large physical footprint.
An earlier study by the researchers at NREL considered land use for wind power, evaluating 80 percent of existing and planned utility-scale wind projects in the U.S. Land use impacts from wind power can be difficult to evaluate, the researchers found, as some impacts are permanent and others temporary; some are direct and others indirect. Permanent and direct impacts result from infrastructure like turbine pads, access roads and substations. Such impacts are more obvious and easier to quantify, whereas the total wind plant area has effects that are harder to define, such as aesthetics and habitat and avian disruptions.
The study found that permanent direct impact varied from 0.06 hectares/MW to 2.4 hectares/MW (a hectare is about 2.47 acres). Temporary impact during construction was higher, as was the harder-to-define total area impact. Total area requirements, the researchers found, could vary widely from 9 hectares/MW to 100 hectares/MW. The group found that “direct impact is mostly caused by road development, as opposed to the turbine pads and electrical support equipment.”Solar’s Manageable Footprint
Denholm and colleague NREL Robert Margolis estimate that if solar PV were used to meet 100 percent of U.S. electricity demand, it would take up about 0.6 percent of the total area of the country, or less than 2 percent of the land dedicated to cropland and grazing. That study finds the average footprint for solar energy to be about 181 square meters per person in a base-case scenario, although the actual figures could vary from 50 to 450 square meters depending on geography and other factors.Denholm and Margolis commented, “One of the strengths of PV is that it can be deployed in a wide range of applications and locations -- from central to distributed applications, and from rooftops to parking lots to field-mounted systems.” PV is suitable, they pointed out, for poorer-quality lands such as brownfields and is compatible with multiple uses such as grazing and shade-tolerant crops. Thus it can be incorporated in land-use plans where minimal environmental impact is desired.