Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Water Issues of Biofuel Production Plants I n addition to the water required to grow crops, biofuel facilities require significant process water. As noted earlier, existing U.S. biofuel facilities consist primarily of ethanol production from corn kernels and minor biodiesel from soybeans, and at the pilot or demonstration-scale, additional ethanol is planned from cellulosic crops such as switchgrass. HOW MUCH WATER DO BIOREFINERIES USE? A useful measure of performance from a water-efficiency standpoint is the net energy yield per unit of water withdrawn or consumed. Consump- tive use of water is largely due to evaporation losses from cooling towers and evaporators during the distillation of ethanol following fermentation. Consumptive use of water is difficult to directly measure because it depends on relative humidity, wind speed, and temperature in addition to the process configuration. However, water permits are generally required from state authorities to withdraw well water or surface water for industrial use, and this water is more or less continually metered. For that reason, this report considers water withdrawals as the measure of water use. This includes both consumptive and non-consumptive use, but as biorefineries increasingly incorporate water recycling, the difference between consumptive and total water use is decreasing. The water needs of each type of production system are discussed in the text below. Corn Ethanol Ethanol produced from corn kernels totaled 4.5 billion gallons in 2006. Production is growing rapidly in the United States and is expected to reach 6 billion gallons this year, but it still provides only a small fraction of total U.S. liquid transportation fuels. A typical process schematic and unit opera- 45
46 Water Implications of Biofuels Production in the United States tions of an ethanol plant are shown in Figure 5-1. Pure water is required for the slurry operation with whole corn, followed by liquefaction to liber- ate sugars from starch via hydrolysis. This is followed by fermentation and distillation operations. Current estimates of the consumptive water use from these facilities are in the range of 4 gallons of water per gallon of ethanol produced (gal/gal) (Pate et al., 2007). For perspective, consumptive water use in petroleum refining is about 1.5 gal/gal (Pate et al., 2007). Overall water use in biore- fineries may be as high as 7 gal/gal, but this number has been consistently decreasing over time and as of 2005 was only slightly over 4 gal/gal in 2005 (Phillips et al., 2007). Thus for a 100 million gallon per year plant, a little over 400 million gallons of water per year would be withdrawn from aquifers or surface water sources (1.1 million gallons per day). The overall water balance for a typical bioethanol plant using corn is shown in Figure FIGURE 5-1â Process schematic and unit operations of ethanol production facility from whole corn kernels. DDGS is âdry distillers grains with solubles.â SOURCE: Parkin et al. (2007). 5-1
Water Issues of Biofuel Production Plants 47 Delta-T Overall Water Balance Evaporation & Drift = 9,420 14,700 Cooling Deaerator & Leaks = 140 Tower 300 Blowdown = 5,280 Boiler Blowdown = 120 13,860 13,860 Water Vapor = 8,398 Fresh M ak eup 6,631 Ethanol Plant W ater Dryer 2,248 System DDGS = 502 Ethanol =19 H2O Corn W ater I n = 23,860 W ater Out (including consum ptive use) = 23,86 0 Ga l Wa t e r /Ga l Et h a n o l = 3 .6 3 FIGURE 5-2â The overall water balance of a typical 50 million gallon per year corn-based Dry Mill ethanol production facility. All figures are in gallons per hour. SOURCE: Reprinted, with permission, from Courtesy of Delta-T Corp. Type has been enlarged for readability 5-2. Ethanol could also be produced from crops other than corn. Potatoes, 5-2 sugar cane, sugar beets, or sweet sorghum could be used as a source of starch or sugar for fermentation, and these would alter the water require- ments somewhat. Sugar Fermentation of Cellulosic Ethanol Producing ethanol from cellulosic materials such as grasses, crop resi- dues, and wood requires a different process than for corn because they are not rich in starch or sugar. Rather, they are primarily made up of larger or more complex molecules such as cellulose, hemicellulose, and lignin, which must be converted to starch prior to processing. Additional enzymes are re-
48 Water Implications of Biofuels Production in the United States quired to break down these substances in cellulosic-ethanol production, and the biochemical pathways used by microbes in the guts of ruminants such as cattle, and in wood-boring insects like termites, are also being studied. Conventional wisdom is that a technology breakthrough is required for this process to become commercialized, and it may be five or more years in the future. Only demonstration- and pilot-scale plants are currently operating for cellulosic-ethanol production. The total water requirements for ethanol from cellulose are thought to be largeâabout 9.5 gal/gal (M. Holtzapple, Texas A&M, personal commun., July 12, 2007), but this likely will decline as efficiency increases with experi- ence at cellulosic-ethanol plants. Consumptive use is projected to be about 2 to 6 gal/gal (Pate et al., 2007). Thermochemical Conversion Thermochemical conversion of cellulosic materials could be the next generation of biofuel plants. The process begins with gasification of biomass. Various catalysts are used to obtain a wide variety of potential products including synthesis gas, hydrogen, methane, or mixed alcohols (including ethanol) for fuel. DuPont Chemical has invested heavily in the alcohol, biobutanol, as a potentially important transportation fuel. Biofuels are normally produced from homogeneous feedstocks, i.e., single-food crops like corn kernels, sugar beets, sugar cane, potatoes, cano- la, sunflower, and soybeans. But thermochemical conversion would allow the use of mixtures of feedstocks. In this technology, polycultures such as mixtures of native prairie plants could be used as a feedstock for transporta- tion fuels (Tilman et al., 2006). This is attractive, because the use of prairie polycultures may have a distinct advantage in terms of lower soil loss, less nutrient applications and runoff, and especially improvement in wildlife habitat (Chapter 3). The thermochemical conversion process holds the promise of much better energy yields and possibly lower water use. However, such technol- ogy is available today only at a demonstration scale; the infrastructure of automobile manufacturing and fuel delivery might need to be revamped to enable the use of biofuels from thermochemical conversion. Phillips et al. (2007) developed a design that would require about 2 gal/gal; this would be about half that required for corn ethanol plants (see above). Pate et al. (2007) estimate of 2 to 6 gal/gal consumptive use is lumped for several pro- cessing methods. Some of the water savings in Phillips et al.âs (2007) design
Water Issues of Biofuel Production Plants 49 is through improvements in cooling tower and boiler feed operations. Some of these efficiencies may be applicable to corn ethanol plants as well. Biodiesel Biodiesel, which in the United States is produced primarily from soy- beans, comprises several percent of the nationâs total biofuel production. Methanol and caustic (sodium hydroxide) are used in the production of biodiesel. Glycerin is a major co-product that has a low market value currently, in large part due to biodiesel production. Because of this, it is sometimes viewed as a major waste product, but greater commercial uses for glycerol could make biodiesel production more profitable. Biodiesel itself burns much cleaner than petrochemical diesel and enjoys considerable advantage in terms of lower air pollution. Biodiesel refining requires much less water per unit of energy produced than bioethanol. Overall, consumptive use is about 1 gallon of fresh water per gallon of biodiesel and overall water use may be up to 3 gal/gal (Pate et al., 2007). Still lower usage may be possible in the future with new tech- nologies, which include the possibility of using recycled waste water with various degrees of treatment. HOW DOES BIOREFINERY WATER USE COMPARE TO THE AMOUNT NEEDED TO GROW ITS FEEDSTOCK? Water withdrawals by biofuel production plants are similar to those of many other industries. They should be considered in the context of the total water cycle for the watershed or aquifer unit that is being utilized. Thus, bio- fuel plants can present local (or regional) problems depending upon where they are located. Even within the same state, the conditions can vary greatly; for example, aquifers in the northeastern part of Iowa tend to be quite pro- ductive, whereas those in the south have a much more limited yield. Siting of some ethanol plants is occurring where the water resource is already under duress. Figure 5-3 shows, for example, that many bioethanol plants that each require 0.1-1.0 million gallons per day are located on the High Plains aquifer. This aquifer is currently being pumped at a rate of more than 1.5 billion gallons per day for agriculture, municipalities, industry, and private citizens. Thus, 15 million gallons per day for bioethanol would repre- sent only 1 percent of total withdrawals. But it is an incremental withdrawal from an already unsustainable resource. Current water withdrawals are much greater than the aquiferâs recharge rate (about 0.02 to 0.05 foot per year in
50 Water Implications of Biofuels Production in the United States FIGURE 5-3â Existing and planned ethanol facilities (2007) and their estimated total water use mapped with the principal bedrock aquifers of the United States and total water use in year 2000. SOURCE: Janice Ward, U.S. Geological Survey, personal commun., July 12, 2007. 5-3 south-central Nebraska; McMahon et al., 2007), resulting in up to a 190-foot decline in the water table over the past 50 years. It is equivalent to âminingâ the water resource, and the loss of the resource is essentially irreversible. The situation is also of concern in some locations in the Midwest, which draw water from confined units like the Silurian-Devonian and the Cambrian-Ordovician aquifers. Counties west of Chicago, for example, have drawn down the Cambrian-Ordovician aquifer by more than 800 feet of water head since 1850. In southwestern Minnesota, a proposed 100 million gallon per year ethanol plant was turned down by the local water system, which could not supply the 350 million gallons of water per year (~1 mil-
Water Issues of Biofuel Production Plants 51 lion gallons per day) that would be needed by the plant. By comparison, per capita water use from public water supply nationally is about 180 gallons per person per day (Table 5, Hutson et al., 2004), so this is the equivalent to the water supply for a town of about 5,000 people. Compared to the water incorporated in the feedstock, water use for the biorefineries is quite small. For example, in neighboring Nebraska about 2,100 gallons of irrigation water were applied per bushel of corn in 2003 (Noel Gollehon, U.S. Department of Agriculture Economic Research Service, personal commun., July 12, 2007). Assuming the common figure of about 2.7 gallons of ethanol from one bushel of corn, 2,100 gallons of water/bushel Ã 1 bushel/2.7 gallons of ethanol = about 780 gallons of water per gallon of ethanol. This is about 200 times larger than the approximately 4 gal/gal given above for a corn ethanol biorefinery. This indicates that biorefineries themselves generate local, but often intense, water supply challenges, while irrigated agriculture can generate regional-scale problems. If, however, the agriculture is rainfed, water for the biorefinery may be the primary source of groundwater or surface water extraction in the area. WHAT WATER QUALITY ISSUES ARE ASSOCIATED WITH BIOREFINERIES? Ethanol plants have various waste streams. First, salts build up in cooling towers and boilers due to evaporation and scaling, and must be periodically discharged (âblowdownâ). Second, the technologies used to make the pure water needed for various parts of the process (e.g., reverse osmosis [RO], ion exchange, iron removal; not shown in Figure 5-1) result in a brine effluent. Under the National Pollutant Discharge Elimination System (NPDES) permits are required from the states to discharge this effluent. These permits often cover total dissolved solids (TDS), acidity, iron, residual chlorine, and total suspended solids. Table 5-1 gives chemical characteristics of waste water from the RO operation and from the cooling tower blowdown for two plants in Iowa. Some violations of NPDES permits have been reported in Iowa and Minnesota from ethanol facilities, primarily for TDS. Wastewater, potentially high in biochemical oxygen demand (BOD, the oxygen used when organic matter is decomposed by microbes), emanates from the processing of by-products such as thin stillage, wet distillersâ grains, and dry distillersâ grains with solubles (DDGS). Discharge of high-BOD wa- ter to rivers and lakes is problematical because decomposition can consume all of the dissolved oxygen, suffocating aquatic animals. DDGS is a valuable by-product that is rich in protein and especially good feed for animals such
52 Water Implications of Biofuels Production in the United States TABLE 5-1â Water Quality of Waste Streams from Two Existing Ethanol Facilities in Iowa Siouxland Ethanol Facility Little Sioux Ethanol Facility (Sioux Center, Iowa) Simulated Blowdown Big Sioux Constituenta Raw GW RO Reject Water Surface Water Tower Eff. TDS 2,113 7,288 703 3,240b Ca2+ 305 1,033 129 638 Mg2+ 138 458 58 185 K + 0 0 2 33 Na+ 148 485 20 297 Cl- 23 131 35 27 SO42- 1,420 4,716 107 2,265 a Concentrations in milligrams per liter. b Concentration in milligrams per liter as CaCO3. SOURCE: Parkin et al. (2007). as dairy cattle, steers, and sheep. Co-location of animal feeding operations with bioethanol production facilities could capture better efficiency in the overall operation compared to transporting the DDGS long distances to animals as is sometimes done. Cellulosic-ethanol plants would have similar water requirements and brine discharges as the current operating corn ethanol plants. There are two additional steps required in converting lignin and cellulose into starch, and these operations could produce wastewater streams that are high BOD and would require on-site treatment or treatment at publicly-owned treatment works (POTWs). Biodiesel has the potential to produce waste water discharges of high BOD, grease, and oils. Wastewater is normally transported to the local POTWs or treated on-site. If treated on-site, it is regulated by the U.S. Envi- ronmental Protection Agency as a bulk organic chemical production facility. Like ethanol plants, biodiesel plants also have waste streams from cooling tower blowdowns and water treatment reject streams. One final potential water quality impact of biofuels would occur well âdownstreamâ in a commercial sense. The increasing production of new mixtures of alcohol and gasoline, such as the 85:15 ratio known as E85, may create new challenges for groundwater in association with fuel spills. These spills might occur around gas stations, or from tanker truck or railcar accidents. While there is an extensive body of knowledge concerning the behavior of contaminants such as benzene in common gasoline spills, a mixture of 85 percent ethanol could alter this behavior considerably. While
Water Issues of Biofuel Production Plants 53 ethanol is completely soluble in water and rapidly biodegraded under most conditions, the presence of high ethanol concentrations enhances dissolu- tion of more toxic gasoline compounds. In addition, rapid biodegradation of ethanol may inhibit the biodegradation of these compounds, which might then migrate farther off-site (Rice and Depue, 2001). REFERENCES Hutson, S. S., N. L. Barber, J. F. Kenny, K. S. Linsey, D. S. Lumia, and M. A. Maupin. 2004. Estimated Use of Water in the United States in 2000. U.S. Geological Survey Circular 1268. Reston, VA: U.S. Geological Survey. McMahon, P. B., J. K. BÃ¶hlke, and C. P. Carney. 2007. Vertical Gradients in Water Chemistry and Age in the Northern High Plains Aquifer, Nebraska, 2003. U.S. Geological Survey Scientific Investigations Report 2006â5294. Reston, VA: U.S. Geological Survey. Parkin, G., P. Weyer, and C. L. Just. 2007. Riding the Bioeconomy Wave: Smooth Sailing or Rough Water for the Environment and Public Health? Proceedings of the 2007 Iowa Water ConferenceâWater and Bioenergy, March 6, 2007, Iowa State Center, Ames, Iowa. Available online at http://www.aep.iastate.edu/water/2007/parkin.html. Accessed on November 12, 2007. Pate, R., M. Hightower, C. Cameron, and W. Einfeld. 2007. Overview of Energy-Water Interdepen- dencies and the Emerging Energy Demands on Water Resources. ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ï¿½ Report SAND 2007-1349C. Los Alamos, NM: Sandia National Laboratories. Phillips, S., A. Aden, J. Jechura, D. Dayton, and T. Eggeman. 2007. Thermochemical Ethanol via Indirect Gasification and Mixed Alcohol Synthesis of Lignocellulosic Biomass. Technical Re- port NREL/TP-510-41168. Golden, CO: National Renewable Energy Laboratory. Rice, D. W., and R. T. Depue. 2001. Environmental Assessment of the Use of Ethanol as a Fuel Oxygenate: Subsurface Fate and Transport of Gasoline Containing Ethanol. Report UCRL- AR-145380 for the California State Water Resources Control Board. Livermore, CA: Lawrence Livermore National Laboratory. Available online at http://www-erd.llnl.gov/ethanol/etohdocII/. Accessed on July 13, 2007. Tilman, D., J. Hill, and C. Lehman. 2006. Carbon-negative Biofuels from Low-input High-diversity Grassland Biomass. Science 314: 1598-1600.
Photo by Brett Hampton, Agricultural Research Service, U.S. Department of Agriculture