04 December 2009

Water and Land Use Intensity of Solar Power vs. Biofuels

In continuing to grasp some basic land use and (ground)water use issues associated with energy production, it occurred to me during some recent discussions with my wondeful friends and colleagues Stephen Kaffka, Dan Putnam (both UC Davis), Margot Gerritsen (Stanford), and Wayne Spencer (Conservation Biology, San Diego) to compare the water intensity of solar power (see blog below) with that of another renewable energy source: biofuels. In the Southwest, all three (food&fiber production, biofuel production, and solar power) compete for land and water. Comparing water use and land use of a solar farm and an ethanol-corn farm in the Southwest on a per MWh-basis is not going to be favorable for biofuels, I figure - given all that irrigation water. But how bad is it?

In my recent blog, these are the numbers that I cited for solar power:

Water intensity: 35 gal/MWh - 175 gal/MWh (dry-cooled)

Land use intensity: 4,000 acres for 500 MW. Assuming an annual production of about 800,000 MWh from a 500 MW installation (I am going by the efficiency of my own new solar panels at home), this equals a land use intensity of 5 acres/1000 MWh (annual production).
To estimate water and land use intensity for ethanol produced from irrigated corn (perhaps the worst case of biofuel water intensity), I am using numbers from a report to Congress (p.61-62):

Irrigated corn needs an average (all of the U.S.) of 1.2 acre-feet irrigation water per acre and has an average yield of 178 bushels per acre. The average consumptive water use on irrigated corn in the U.S. is 2,200 gallons of water per bushel, ranging from 500 gal/bushel for Pennsylvania to 6,000 gal/bushel for Arizona (also see, e.g., this recent journal article by Stanley Mubako and Chris Lant at Southern Illinois U., and another by Andy Aden). Given typical ethanol production rates from corn and ethanol energy content, that is 2,500 to 29,000 gal per MMBtu (million Btu) - see the Figure below. 1 MWh equals 3.414 MMBtu. Hence, at 178 bushels/acre, we have

Water intensity: 730 gal/MWh - 8,500 gal/MWh
Land use intensity:  8 acres/1000 MWh (back calculating, e.g., reciprocal of 6,000 gal/bushel * 178 bushel/acre / 8,500 gal/MWh)

If the ethanol was used in a thermo-electric power plant, and assuming a typical efficiency of thermo-electric plants of about 30% (a number typically found for coal power plants), we have the following water and land use intensities:

Water intensity: 2,200 gal/MWh - 25,000 gal/MWh [+300 gal/MWh in consumptive use at the plant]
Land use intensity: 25 acres/1,000 MWh (annual production).

Michael Campana, aka Aquadoc, just a few weeks ago, listed a number of reports on the topic of water use in biofuels. The bottomline is that biofuels grown from irrigated crops in the Southwest or California have an enormous consumptive water use intensity. And their land use intensity is significantly higher than that of solar power farms.

This is not to argue against the use of biofuels (see my next blog). The use of both renewables, solar energy and biofuels is a matter of much more than water use or land use. For example, at low irrigation rates, biofuels could be generated in water positive fashion even by desalination of salt water (which takes a lot of energy). The issue of water and land use in the biofuel production is certainly not new and much has been written about it. There are alternative biofuel sources that are much less water intensive. Dana Larson and her colleagues at the Bren School, UC Santa Barbara, recently published a wonderful summary on the water usage in energy production, related policy issues, and current policy requirements. Quoting from the above-mentioned Report to Congress (p.44):

"[...] biofuel feedstock produced from crop residues in excess of those needed to maintain a healthy ecosystem, from feedstocks grown without irrigation, or from feedstocks grown with nontraditional water, will have minimal freshwater use intensity associated with production. This could provide significant volumes of bioenergy and biofuels in the future with low water use intensity (Perlak et al., 2005). In all cases, some water use is associated with processing, as shown in Figure V-4, but further technology development is likely to lower these values." [Figure below from the same page]


  1. good article on water footprints ..


  2. Thank you, Harris. I am still cross-checking the numbers on the two biofuel blogs and adding references. I will update as needed...