Water Intensity of Corn-Based Ethanol Needs to Include DDGS Credit

Study recalculates water consumption with expanded system boundaries and coproduct credit
By Gouri Shankar Mishra and Sonia Yeh | August 15, 2011

The growth in biofuels production, specifically corn ethanol, has raised the discussion about the sustainability of biofuels and the impact on natural resources. Recent studies have estimated the “consumptive water use,” “water embodied,” and “water footprint” of corn-based ethanol in the U.S, with numbers differing by orders of magnitude. Estimates for liters of water used in corn ethanol per vehicle kilometer traveled (VKT) vary from 1.1 to 335 L/VKT for Iowa and from 59 to 214 L/VKT for Nebraska. The major difference among these studies stems from the debate existing in the water life-cycle analysis (LCA) literature on how to handle the crop evapotranspiration needs met through precipitation, also referred to as green water.

While the use of blue water—water from rivers, canals and aquifers—is consistently included by all studies, we believe that estimation and reporting of green water use is very informative. The explicit reporting of green water requirements acknowledges competing demands for limited freshwater, and also the overall hydrological impacts of growing bioenergy crops in a given area instead of alternative uses of land such as pastures or other crops. Further, estimates that include total water use are more robust than those focusing on blue water use alone. Drought in any single year will necessitate application of more irrigation water to compensate for lower precipitation, although total evapotranspiration remains fairly constant.

One important variable is overlooked in most of these studies, however. When the proper credits are incorporated for the coproducts of corn ethanol production, the resulting water requirements in the LCA are significantly lower than previous estimates. DDGS displaces other animal feeds such as soybean meal and feed corn, and thus precludes the use of water to grow these feed sources.

Our study sought to analyze corn ethanol’s water intensity while explicitly stating the water source—green water versus blue water—and broadening the system boundaries to account for the types of uses, water losses and water inherent in energy use, as well as coproduct credits. We also attempted to estimate the water requirement in the corn cob, as a source of cellulosic ethanol feedstock, separately from the corn grain.

Figure 1 summarizes the various water requirements considered in our study. In addition to evapotranspiration by crops, and process and cooling water requirements at the ethanol conversion facility, there are water losses through evaporation from open irrigation canals or drift losses in sprinkler systems and water embodied in the various energy inputs across the life cycle. There are also non-consumptive water uses, such as the deep percolation of crop water below the root zone, runoff and seepage losses in irrigation systems. Such nonconsumptive use of water may not be considered as losses if the spatial boundary used for analysis is extended from the field or farm level to the level of a hydrological basin. Runoff and seepage from an irrigation system can be a loss at a local level but actually benefit downstream users. As an example, individual irrigation systems in the Nile basin in Eqypt are around 30 percent, while the overall efficiency of the entire Nile system is estimated at 80 percent.

Our analysis focused on five states—Illinois, Indiana, Iowa, Kansas and Nebraska—which accounted for more than 50 percent of the U.S. corn crop in 2009. In the heart of the Corn Belt, Indiana, Illinois and Iowa, more than 97 percent of the corn was rain-fed. Nebraska and Kansas have both rain-fed and irrigated corn crops, with irrigation relying upon center pivot sprinklers. We also considered California where the entire corn crop is irrigated. In our study, we modeled dry mill conversion, which accounts for 88 percent of ethanol operating capacity in the U.S. In calculating the water credit from the production of distillers grains, we used displacement ratios based on the widely used GREET model, which accounts for the relative nutrient content and market share of distillers grains and displaced product. We assumed that displaced corn and soybean are grown in the same region, except for California where distillers grains displaces soybeans shipped in from the Midwest.

Figure 2 summarizes the statewide average consumptive water requirements of ethanol from corn grain, and from both corn grain and cob. Most of the water requirements of ethanol from rainfed crops are met by precipitation; with a small amount of blue water used during ethanol conversion. The green water intensity of such ethanol varies from around 70 L/VKT in Iowa to 135 L/VKT in Kansas. The blue water intensity of ethanol from irrigated corn ranges from 50 L/VKT in Nebraska to around 150 L/VKT in California. Irrigated corn crops in Nebraska and Kansas are largely grown with groundwater. The difference in total water requirements reflects both differences in climatic conditions as well as corn yields.

Although we extended the system boundary to include additional water requirements such as conveyance losses of irrigation water, our estimates of corn grain ethanol’s blue and green water consumption are lower than those of previous studies This is due to the accounting of coproduct credits for water use, which we estimated to be 5 percent and 45 percent of the total blue water used to produce ethanol from rain-fed and irrigated corn, respectively, and around 50 percent of green water in both cases (shown in Figure 2). It is interesting to observe that these numbers are higher when compared to estimates in the literature of a 20 percent greenhouse gas emissions credit for coproducts. This can be explained by the fact that soybean requires significantly less fertilizer and pesticides (which emit greenhouse gases) but similar amounts of water compared to corn on a per bushel basis. As a result, dried distillers grains with solubles generates relatively fewer greenhouse gas emission credits but more water credits.

Our methodology to determine the water intensity of ethanol is the first step towards accurately assessing the water use impacts of biofuels. The blue and green water requirements of ethanol from corn grown in different regions provide useful information for local water resource management. For example, water use by ethanol can be compared with a region’s total water budget to identify potential water availability constraints and risks.

Our water intensity estimates, however, may not be comparable across states. Regions differ in terms of water availability, scarcity and potential impact on ecosystems as a result of cultivation of bioenergy crops and biofuels production. Our volumetric estimates do not capture such regional differences. In the LCA literature, “characterization factors” are being developed to convert volumetric water intensities to stress-weighted water intensities that will allow regional comparisons and help reach important policy-related conclusions.

Further, our water intensity estimates may not be aggregated to assess the water use impacts of policies encouraging use of biofuels. Our estimates are based on “average” water requirements of corn. Policymakers should consider “marginal” water requirements, which are likely to be higher than average. As suggested in literature, higher corn prices, as a result of ambitious production mandates in the renewable fuel standards, could lead to expansion in corn production to marginal lands with lower yield potentials. They could also result in intensification of corn cultivation in existing lands, which increases yield in the short run but could lower future yields. Since water intensity is negatively correlated with yield, such expansion and intensification will increase the water intensity of ethanol. Further, corn expansion is occurring disproportionately on land that requires irrigation, which according to our results has higher average total water (green plus blue water) and blue water consumptive intensities.

Future study could combine the methodology developed in this study with economic models to analyze the marginal impacts of ethanol production on water use. Work is ongoing to use the water intensity estimates of our model to undertake impact analysis and accurately assess the effects of biofuel production on water resources.

Authors: Gouri Shankar Mishra
Research Associate
Institute of Transportation Studies
University of California-Davis
(510) 378-3187

Sonia Yeh
Associate Researcher
Institute of Transportation Studies
University of California-Davis

Note: This article is a summation of a paper, “Life Cycle Water Consumption and Withdrawal Requirements of Ethanol from Corn Grain and Residues,” which appeared in the journal, Environmental Science & Technology (Volume 45, Issue 10), American Chemical Society. Figures 1 and 2 are reprinted from the article with permission.