(283d) On the Likely Competition for Land between Food and Energy in a Solar Economy

Miskin, C., Purdue University
Li, Y., Purdue University
Agrawal, R., Purdue University
The main purpose of this presentation is to demonstrate that contrary to the popular belief that ample land area is available to install photovoltaic modules, in an economy mainly supported by solar energy, there will be severe land availability constraints in many regions of the world—including many states in the USA. This will inevitably lead to competition for land between food and energy [1]. Our analysis will highlight the urgency with which solutions must be sought to avoid competition between these basic needs of humanity to enable an earth with more than 10 billion people in a few decades.

The vast potential for solar energy cannot be disputed. However, the land area requirements often stated to meet our energy needs from solar energy are often much lower than will actually be required in practice. A recent study by MacKay showed that the energy consumption per unit area in densely populated countries such as the UK, Germany, Republic of Korea, and Japan is approaching the energy generation per unit area of photovoltaic farms [2]. This suggests that a transition to a solar powered economy in these nations would require deployment of solar on a large fraction of the nation’s total land area, which will directly compete with land requirements for other needs (e.g. agriculture). Many other countries, including the United States are also trending toward this scenario [2]. A recent study by National Renewable Energy Laboratories confirms the large land area requirements of solar farms [3].

In this study, we estimate the United States’ land area requirements for an economy fully powered by solar energy by using actual production data of PV farms, which have an average production throughout the year of 4-11 W/m2. Using available data [4] we have calculated the energy use per square meter for the U.S. (excluding Alaska and Hawaii) to be 0.43 W/m2. We can thus estimate that the actual land requirement will likely be 1.9-5.2% of the U.S. land area, when reasonable allowances are made for nighttime storage losses. (Note that in performing these calculations we have accounted for the fact that a kWh of coal does not have the same intrinsic value as a kWh of electricity and have appropriately converted between various energy sources and end use sectors based on average conversion efficiencies of each energy transformation.) Unfortunately, much of the available land needed for deployment of PV is often located great distances from the population centers where the energy is needed. When the same analysis is performed on a state by state basis we find that the land area needed to meet individual states’ needs ranges from 0.2-0.5% (Montana) to 16.4-45.0% (New Jersey). While a solar economy is feasible for many of the large, sparsely populated states of the Plains States and Mountain West, a transition to a solar economy will be challenging for many other states.

A remaining important question is which of these states have sufficient available land not dedicated to other uses to meet these energy needs. We answer this using data from the United States Department of Agriculture on current land usage and estimate that as many as 31 states would be unable to internally meet their energy needs given the available land in those states not already dedicated to other uses. This calculation is based on what is categorized as “miscellaneous land” in each state—that land not currently devoted to cropland, pasture/range land, forest use, urban areas, and special uses (parks and wildlife areas). Note that miscellaneous land includes deserts, swamps, bare rock areas, rural residential, etc.—some of which will be unsuitable for PV, so land constraints may be more severe than estimated. The reason for the lack of available space is that much of the open land in the U.S. is devoted to cropland (21.5%), pasture land (32.3%), and forest land (30.4%). Only 3.6% of the land falls in the category of miscellaneous/other use land [5]. While some land currently devoted to other purposes (e.g. rooftops in urban areas) could be used for PV as well, many states would have to rely on long distance transmission to meet their energy needs under current strategies. Rooftop PV also has inherently higher cost compared with larger scale PV installations.

Assuming it will remain a priority to preserve the wildlife habitat embodied by forests and other special uses and in the absence of radical improvements in agricultural output that allow us to greatly reduce the percentage of land devoted to agriculture, innovation will be required to find ways of sharing PV with current land uses as the available space is filled up. One way this could be accomplished is by incorporating solar energy harvesting with agricultural activities, as has been done successfully with wind farms in many areas. While this seems counterintuitive, PV systems can be designed that allow large fractions of the visible portion of the solar spectrum needed for photosynthesis to pass through, while capturing other photons for electricity or heat. We will briefly describe some of possible configurations and material research needs to accomplish these objectives.


[1] Solar Projects Sow Tension, Wall Street Journal, (March 9, 2017) A3.

[2] D.J.C. MacKay, Solar energy in the context of energy use, energy transportation and energy storage., Philos. Trans. A. Math. Phys. Eng. Sci. 371 (2013) 20110431. doi:10.1098/rsta.2011.0431.

[3] S. Ong, C. Campbell, P. Denholm, R. Margolis, G. Heath, Land-Use Requirements for Solar Power Plants in the United States - Technical Report NREL/TP-6A20-56290, 2013.

[4] State Profiles and Energy Estimates, (n.d.). http://www.eia.gov/state/.

[5] Major Land Uses, (n.d.). https://www.ers.usda.gov/data-products/major-land-uses/.