What would the United States look like 69% of today’s electricity were generated from solar technologies, photovoltaic and concentrated solar power? A recent paper in Energy Policy (see 1 below: Fthenakis, Mason, and Zweibel., 2009) proposes this type of vision of generating 69% of US electricity from solar by 2050 (using also large quantities of energy storage technologies such as molten salts and compressed air energy storage). The authors note that 640,000 km2 of available land area exists in the Southwest US that can be used to for solar power stations (see link on National Renewable Energy Laboratory website for resource maps, but without land restrictions – NREL Solar Maps). This area is 48% of the total area of the included states of California, Nevada, Arizona, and New Mexico – which in total cover 1,320,000 km2.
In 2008 approximately 4,100 TWh of electricity were generated in the US. Thus, 69% is approximately 2,840 TWh. If we assume that installed solar systems convert sunlight (at an average of 6.4 kWh/m2/day for the Southwestern US) at 8% efficiency from sunlight to delivered electricity to the consumer, then approximately 2.4% of the available sunlight on the 640,000 km2 would be needed for solar systems. Note that current photovoltaic (PV) and concentrated solar power (CSP) systems can have solar-to-electric conversion efficiencies from 7%-40%, where the high end of the efficiency range is by laboratory multi-junction PV cells under concentrating lenses.
As stated by Fthenakis at al. (2009) their projected 2050 solar electricity generation (which is more than 2,870 TWh due to assumed increases in generation each year) would occur from an installed capacity of approximately 5.5 TW of solar generation – 1.5 TW from solar CSP and 4.0 TW from PV. For a 500 MW solar plant the authors estimate 10.6 km2 of land is needed (at 14% efficiency). Using this ratio of land for all solar installations projects the 5.5 TW of capacity would need approximately 110,000 km2 of land, or 17% of the available 640,000 km2.
Only 2.4% of the sunlight hitting the land needs to be converted to electricity, but 17% of the 640,000 km2 of land would be required for power plants. The reason is that the conversion efficiencies are calculated assuming that the PV panels are tilted toward the sun at an angle equal to the latitude of installation. This causes a PV panel to shade the next PV panel just to the North if it is placed too closely. Thus, there needs to be some spacing between panels and mirrors for CSP systems. If the solar power plants were equally distributed spatially throughout the Southwestern US (which is a practical impossibility and strategically undesirable) one could likely not walk on any path from southern California and Nevada throughout Arizona and New Mexico without seeing a solar power plant. Solar power plants would essentially be pervasive throughout the Southwestern United States.
This large scale of solar power pervasiveness brings to mind the question of whether there are enough materials to manufacture all of the solar panels and mirrors. A paper from this year in Environmental Science and Technology by Wadia, Alivisatos, and Kammen takes a first-order look at this question for the various semiconductor materials that can be used to create the functional component of the PV panels. They point out that the known reserves of the inorganic photovoltaics materials in ore deposits are sufficient for creating a broad variety of PV panels to generate the 17,000 TWh of electricity in the world today. The other necessary materials associated with circuitry (silver, copper, etc.) and mounting (aluminum) would also need to be available. Because there are many competing products for these resources (e.g. indium for flat panel displays) and some elements are only secondarily mined substances – meaning they are essentially byproducts associated with mining for other primary materials – we will only find out the future to what uses these elements are applied. It might also be interesting to think about how much energy (and power) would be required to mine all of necessary materials …
Perhaps we should use our fossil fuels to mine the fossil minerals and elements that are needed to stop using fossil fuels. Sounds like a chicken or the egg problem. We may not know which came first, but eventually, we’ll learn which one is last.
1. Fthenakis, V.; Mason, J.E.; and Zweibel, K. The technical, geographical, and economic feasibility for solar energy to supply the energy needs of the US. Energy Policy. 2009, 37, 387-399, doi:10.1016/j.enpol.2008.08.011.