Is Biofuel Water-Positive?
An arcane but instructive way to evaluate corn ethanol, along with all biofuels, may not just be to audit their “net energy balance,” but also their “net water balance.” Evaluating whether or not a biofuel crop could be “water positive” is even more subjective to calculate than whether or not that crop is energy positive, but here goes:
Corn is one of the better temperate crops to use as a primary biofuel feedstock, since cellulosic extraction isn’t here yet and sugar cane doesn’t grow in Iowa. Using corn as an example, a good ethanol yield is about 480 gallons per acre per year, which is based on 160 bushels per acre, and 3.0 gallons of ethanol per bushel. How much water corn needs varies greatly, and the range we’ve arrived at for this analysis is between 300 and 900 cubic meters per ton. Our source for 900 m3/ton is from a reference to UNESCO’s “The Water Footprint of Nations,” and our source for 300 m3/ton is from Colorado State University.
Since a bushel of corn weighs about 70 pounds, based on a yield of 160 bushels per acre, expressed in tons the per acre yield of corn is about 5.6 tons. This means, at the lower figure of 300 cubic meters of water per ton of corn, the average acre of corn requires 1,680 tons of water per harvest cycle, which equates to 444,000 gallons of water for every 480 gallon yield of ethanol. Clearly, from this perspective, the 3-6 additional gallons of water required after harvest to refine each gallon of corn ethanol is not the critical factor - particularly when petroleum fuels also require water during their refining process.
If it takes 925 gallons of irrigation water to grow corn for every gallon of ethanol that can be distilled from corn, how much energy would it take to desalinate seawater to irrigate that corn? Would there be energy left over after the ethanol had been used to power the desalination plant that provided the fresh water for irrigating the corn? The answer is yes, but only when we use the lower figure - 300 cubic meters of water per ton of corn harvested.
Since 2.0 kilowatt-hours is necessary to desalinate a cubic meter of seawater, then at 300 cubic meters of water per ton, and 5.6 tons per acre, it takes 3,360 kilowatt-hours of electric power to desalinate enough water to irrigate an acre of corn for a year. Since ethanol has about 80,000 BTUs of energy per gallon, at a yield of 480 gallons per acre you will extract 38 million BTUs. Theoretically, 3,400 BTUs equals one kilowatt-hour, but even the best electric generating plants only succeed in capturing about 60% of those BTU’s. This means that in terms of electric power, corn ethanol is good for about 23 million BTUs, equating to 6,776 kilowatt-hours.
So is corn ethanol water positive? At 300 cubic meters of water per acre, you would require 50% of your corn ethanol yield per acre to power the desalination plant to irrigate the corn. At 900 cubic meters of water per acre, your corn crop would not yield enough ethanol to desalinate the water required to irrigate the corn. Under these assumptions, growing and refining corn ethanol is certainly not a decisively water-positive enterprise.
This reality points to another trade-off when considering whether or not to develop biofuel crops to scale. If the land status is changing - either from forest or from farmland - in order to produce biofuel crops, then it is valid to assume the opportunity cost of the biofuel includes reallocating the water - even if it was rainfall. The measurements explored here could be as good as any when determining the net water balance of biofuel crops.
June 6th, 2007 at 1:02 am
This is off-topic, but have you seen this?
United Nation’s Intergovernmental Panel on Climate Change scientists noted that rice production is perhaps a bigger source of global warming than carbon dioxide emissions of all industrial and power plants.
Story at Agnet and written at ABS-CBN.
So many factors!
July 19th, 2007 at 8:43 am
I was alerted to this posting yesterday (7/18/2007) by an individual at the USDA National Agricultural Library who was making an inquiry about corn water use. I am the senior author on the newsletter article from CSU from which the 300 m3/T figure was calculated. There were, unfortunately, a misunderstanding of the information in the article and a miscalculation that resulted in the erroneous 300 m3/T figure.
1. The first mistake was using a value of 70 lb/bu of corn. Corn weighs 56 lb/bu.
2. The calculation used the smallest value reported for corn water use (17.2 inches to produce 168 bu/a corn) to get the value of 300 m3/T. These data are from only one year of data, and did not include soil water extraction as part of the water use (only rainfall and irrigation were used to get the 17.2 inches). So it wasn’t really total water use. What should have done is use the first equation given in the article:
yield in bu/acre = 10.4 x (water use in inches - 9.1)
Had that been done, a value of 24.5 inches of water needed to produce 160 bu/a corn would have been obtained, which equates to 563 m3/T.
Using the proper values in the energy analysis shows that growing corn for ethanol production and using some of the ethanol to power the desalinization process would still leave an excess of ethanol produced (74% of the ethanol produced would be used for desalinization, 26% would be left to put into the market).
In any case, 300 m3 of water/Ton of corn produced is probably too low, and a more realistic number is about 560 m3/T for 160 bu/a corn. The value decreases to about 520 m3/T for 200 bu/a corn. The value continues to go down as yields go up because the amount of water needed to grow the plant to the point where it can start producing grain doesn’t change (about 9 inches), and becomes a smaller fraction of the total water used to produce the final grain yield.
The linear relationship shown above and used to generate these values changes from location to location, with slight changes in the slope of the relationship, and some more significant changes in the offset. For example, using a water/yield relationship generated in Nebraska, we would need about 25.8 inches of water to grow 200 bu/a corn, which would lead to a value of 473 m3 water/Ton of corn (vs 520 m3/Ton using the Colorado relationship shown above). Water use efficiency of corn grain production generally improves as we move east and north of Colorado.
But as you can see, the previously posted analysis using 300 m3/Ton was incorrect and too optimistic.
David Nielsen
Research Agronomist
USDA-ARS
Akron, CO
July 29th, 2007 at 6:56 pm
How much potable water do we really have left on this earth? In Wyoming they are pumping natural gas from deep wells in coal fields and bringing dirty water to the surface and pumping it into the aquifer used by the people for drinking water. One of many missuses of our water supply. How long can we keep up this kind of contamination? I am from St. Louis, MO. here we are lucky four rivers feeding a plentiful supply to us. A lot of our watershed has been diverted up stream, especially the Missouri river, to the point where barge traffic is stopped during the fall season. Well wait a moment, come to think about it we will not need the barge traffic at harvest time if we are not shipping the corn to export points. I guess that solves that problem doesn’t it. But what a solution!
Sorry for the rambling. I am just an old man thinking what the future will bring for our posterity. We can replace corn, and many other things, but what do we do for water? I guess President Bush has an answer up his sleeve, like how to get out of this mess in Iraq.
God Bless you for helping bring this to the attention of the public. where can we find another million like you to get the picture out? Harry Donnegan
January 10th, 2008 at 2:43 am
Where I live there is very little to no irrigation done. We get 40+ inches of rain a year; tilling to drain away excess is normal.
Only water that falls naturally, requires no energy or ’subtracting’ from rivers or aquafers.
There are many ethanol plants around me. Practically no acres are irrigated. That reduces the water requirement back to the 3-6 gallons the plant uses.
Even in irrigated areas of the West, 1/2 to 3/4 of the water ‘used’ to grow corn falls naturally from the sky.
I believe in worst case, this water ‘use’ is off by 1/2, and for much of the corn growing area of the USA, it is off by 95%. Natural rain should not be counted as a ‘use’ as it is not pumped but natural, and mostly continues on it’s natural course as it passes through the corn field - used to grow weeds, grass, or corn it’s all the same as it goes on to aquifers or rivers.
National corn yield is 153 bu/a, in the parts of Minnesota with many ethanol plants corn yield approches 200 bu/a. Corn weights 56 lbs/bu, and ethanol plants are getting about 2.9 gallons/bu.
No need to desalinate or pump any water other than that used in processing for most corn growing areas?
Paul
January 10th, 2008 at 8:55 am
Paul: Corn ethanol can make sense in areas where there is adequate summer rain. But where water requirements already approach or exceed water supply, such as California, these calculations are more relevant. Our position has always been that there is a strong argument in favor of biofuel crops, if they are grown on farmland in a region of water and food abundance. But subsidizing irrigated corn ethanol crops in a place like California’s central valley is insane, as is destroying the last tropical rainforests to make way for biofuel crops such as sugar cane and oil palms.