February 2008, Vol. 20, No.2

Biofuels Balancing Act


How much water does it take to grow a liter of ethanol? According to a report released by the Water Science and Technology Board of the U.S. National Research Council “there are likely to be significant regional and local impacts where water resources are already stressed.” While they require more research, some emerging biofuels are expected to be less water- and nutrient-intensive than those currently being grown

With ethanol and biodiesel use and demand growing and, therefore, more corn and soybeans being farmed, the Water Science and Technology Board of the U.S. National Research Council (NRC) held a colloquium last summer to examine the water resources implications of growing these biofuel crops.

The resulting report, Water Implications of Biofuels Production in the United States, concludes that during the next 5 to 10 years, biofuel production probably will not affect water resources nationwide. But “there are likely to be significant regional and local impacts where water resources are already stressed,” the report states. Furthermore, the report pinpoints some emerging biofuels, which — while they require more research — are expected to be less water- and nutrient-intensive than those currently being grown, painting a more sustainable long-term picture for renewable fuels.

Assessing the Need
The dramatic increase in biofuel production, specifically corn ethanol, has been driven by recent increases in oil prices in conjunction with subsidy policies that reflect a strong U.S. national interest in greater energy independence. In the past 6 years, the amount of ethanol produced in the United States has tripled from 6.2 million m³/yr (1630 million gal/yr) in 2000 to 18.4 million m³/yr (4855 million gal/yr) in 2006, according to the Renewable Fuels Association (Washington, D.C.).

In 2007, President George W. Bush called for U.S. production of ethanol to reach 132.5 million m³/yr (35 billion gal/yr) by 2017, which would displace 15% of the nation’s projected annual gasoline use. By 2030, the administration aims to increase that production to 227.1 million m³/yr (60 billion gal/yr).

Managing Supplies
When biofuel crop production will begin affecting water supplies is a very local question, according to Mark Alley, professor of agriculture at Virginia Polytechnic Institute and State University (Blacksburg, Va.) and a speaker at the colloquium. He said that water use associated with biofuel falls into two categories: irrigation water for crops and process water for biorefineries.

For crop production, if the number of irrigated acres is expanded for biofuel crops, then “you could conceivably say that the fuel production has required that extra amount of water,” Alley said. However, he added that it’s difficult to determine if that has happened yet, because the average number of irrigated acres has remained relatively steady.

The report specifically mentions that portions of the Ogallala (or High Plains) aquifer, which extends from western Texas north into South Dakota and Wyoming, shows water table declines of more than 31 m (100 ft).

Another colloquium speaker, Steve Kaffka, director of the Long-Term Research on Agricultural Systems Project, part of the Agricultural Sustainability Institute at the University of California–Davis, said that in this case, “the question becomes, What is the best use of water from that aquifer, both currently and over time?”

“Farmers are depleting [aquifers] even in the absence of a biofuel demand,” Kaffka said. “They might end up finding it extremely useful in the short term to pump more water to grow more corn because the prices are so high.” He speculated that increased demand for corn biofuel could have an adverse effect on aquifers.

“I actually think that biofuel development should be very region- and site-specific to allow for the greatest creativity and flexibility in responding to those needs,” Kaffka said. “Agriculture is a highly site-specific activity, and farms in different regions and farmers in different regions will be able to create quite different solutions to the same kind of problem.”

In fact, the report states that whether or not biofuel crops will require more or less water will depend on what crop is being substituted and where it is being grown. Corn generally uses less water than soybeans in the Pacific and Mountain regions, but the reverse is true in the Northern and Southern plains. Therefore, farmers switching from soybeans to corn will need more water in some regions and less water in others.

As an example, Alley cited Texas A&M University (College Station, Texas), where researchers are converting sugar cane into ethanol. That crop is very productive in the Texas climate but would fail completely in Minnesota or even Virginia. Alley said the key is “the right plant in the right place.”
With biorefineries, on the other hand, it’s easy to measure the additional water needed.

“When you site an ethanol plant, it has a large water requirement,” Alley said. “The question becomes at that local site, Is it a significant amount of the available water?” Siting the plants in areas with sufficient water supplies is the major factor to consider, he said.

In southwestern Minnesota, for example, a local water system turned down a proposed refinery that would have generated 378,500 m³/yr (100 million gal/yr) of ethanol, the report states. The water system could not supply the 3785 m³/d (1 mgd) of water needed by the ethanol plant. To put that amount in perspective, the report states, 3785 m³/d (1 mgd) is equivalent to the daily water supply for a town of 5000 people.

Totaling the Amounts
So what’s the overall water usage to grow and refine a liter of ethanol from corn? The answer is about 784 L (207 gal), with almost all of it needed to grow the corn.

The report cites data that in Nebraska it took about 7950 L (2100 gal) of irrigation water to grow a bushel of corn in 2003. Assuming the common figure of producing about 10.2 L (2.7 gal) of ethanol from one bushel of corn, it takes about 780 L (206 gal) of water to grow the corn for 1 L of ethanol, the report states.

In addition, it takes 3.8 L (1 gal) of water to produce 1 L of ethanol at the refinery, according to the report. In comparison, it takes only 1.5 L (0.4 gal) of water to produce 1 L of gasoline at a conventional petroleum refinery.

Could reclaimed wastewater be used for biofuel crop irrigation?

“Even some unreclaimed wastewater could be used, in my view,” Kaffka said. He said if the biomass is to be used to make fuel, “that doesn’t seem to require quite the level of concern of microorganisms or even trace elements that would be required if wastewater [were] to be used for vegetable production — to use the two extremes.”

Controlling Side Effects
In addition to extra water usage, additional agriculture also could affect nutrient and sedimentation loads entering streams and rivers. Moreover, “of the potential biofuel crops, the greatest application rates of both fertilizer and pesticides per hectare are for corn,” the report states.

At the moment, biofuel agriculture heavily favors corn. Ethanol derived from sorghum (<0.38 million m3 [<100 million gal] in 2006) and biodiesel derived from soybeans (0.34 million m3 [90 million gal] in 2006) each currently make up a very small fraction of U.S. biofuels.

Converting other crops or noncrop plants to corn likely will lead to much higher application rates of nitrogen.

The potential for additional corn-based ethanol production to increase the extent of hypoxic regions in the Gulf of Mexico and other waters is considerable, the report states.

However, the report concluded that water quality effects of biofuel crop production can be mitigated by the same farming best management practices and precision agriculture techniques used for food and other crops, said William Logan, senior program officer at NRC and director of the report.

Kaffka said that it’s difficult to establish agricultural policy for a continental-scale nation where the interests of the farmers vary with the region: A policy that is ideal in New England might be quite poor in California. For example, he said, California probably never will be a corn ethanol-based biofuel area, but it might produce sugar cane, sorghum, or perennial grasses to some degree.

Weighing the Options
According to the report, one of the ways to assess the environmental impact of biofuel production and use is to look at its net energy balance (NEB) — the energy content of the biofuel divided by the total fossil energy used throughout the full life cycle of the production of the feedstock, its conversion to biofuel, and transport.

The report lists the NEB values for several biofuels produced in the United States:

  • Corn ethanol has an NEB of 1.25 to 1.3; i.e., it returns about 25% to 30% more energy, as ethanol, than the total fossil energy used throughout its production life cycle.
  • The NEB for U.S. soybean biodiesel is about 1.8 to 2.0, or about a 100% net energy gain.
  • Switch-grass ethanol via fermentation is projected to have a much higher NEB (between 4 and 15).
  • Similarly high are the estimates for cellulosic ethanol (NEB = 5.5) and synthetic gasoline and diesel from certain mixtures of perennial prairie grasses, forbs, and legumes (NEB = 8.1).

If growing corn requires the most nutrients to be applied to the fields, uses more water (depending on the region), and has the lowest energy benefit of the biofuel crops produced, then why is corn ethanol so heavily favored?

“We grow corn efficiently, and we have a lot of it,” Kaffka said. “We’re fortunate to have large areas of the country where corn can be produced at high levels with high levels of efficiency under rainfall. We’ve developed an infrastructure for both economic and physical transport and management of corn. We can move millions and millions of tons of it. It also stores easily.”

However, “it’s not necessarily, in the long run, what will be best.” Kaffka said.

Extensive research is under way to make cellulosic ethanol viable on an industrial scale, Alley said. He explained that making cellulosic ethanol would be much more energy-efficient. The process turns the starches in plant walls into ethanol. The feedstock sources can be as varied as switch grass or wood chips.

The other advantage of cellulosic ethanol is that the feedstocks, such as perennial grasses, typically have nutrient and pesticide application requirements that are an order of magnitude lower than those of corn or soybeans, the report states. However, the report cautions that since such crops have very little history of large-scale cultivation, “even basic information such as water or nitrogen inputs needed, herbicide use, impact on soil erosion, and even overall yields is preliminary.”

“We had best be diversifying our research efforts and our analysis, because if you pick the wrong one, you’ve got a major problem,” Alley said. “The other part of that is what’s right for one area of the country is probably not going to be right for other areas of the country.”

Overseeing Progress
Even though the relationships among biofuel crop production, water quality, water quantity, and preserving other natural resource are vast and complex, Alley and Kaffka said that the system will balance itself.

“While we could subsidize something in every region to start, over the long term, what’s going to be sustainable is going to have to be economically sustainable in the marketplace,” Alley said. “Now, just like every other sustainable industry, we need to be concerned about environmental costs and regulate the environmental part of it.

“We’re talking about major areas of land to be devoted to this,” Alley continued. “We’re talking about major transportation issues. We’re talking about major demand issues that vary all over the country. We cannot be polluting water or air in the name of energy production.”

“Both entrepreneurial and scientific activity, and scientific invention and the activities of the marketplace ought to be allowed to develop,” Kaffka said. We might find “quite novel or so far unforeseen solutions to the needs of producing biomass for energy.”

The report was sponsored by the Energy Foundation (San Francisco), McKnight Foundation (Minneapolis), National Science Foundation, and U.S. Environmental Protection Agency. NRC also provided some funding, Logan said.

The report can be downloaded from the U.S. National Academies Web site at www.nap.edu/catalog.php?record_id=12039.

— Steve Spicer, WE&T