Reduction of odors and water pollution as well as potential energy generation (CHP) on large livestock operations have stimulated increased interest in anaerobic digestion systems. Since the mid-1970s, interest in methane generation technology in the United States has varied. Early anaerobic digestion in the United States had been farm-based with primary efforts to develop appropriate technology, which would require low initial capital cost, low operating

costs, low maintenance, little operator time, and minimum management skills (Jewell et al. 1979). Based on a simplified reactor design and reported advantages of lower capital and operational costs, the plug-flow digester has been the most readily utilized design compared with a completely mixed design for dairy manure. A variation of the plug-flow digester (especially when food wastes are added) is a horizontal mixed digester used to maintain solids in suspension. A new development is the design of a vertically mixed digester tank of either concrete or steel. The mixing is achieved by several methods, such as slow rotational paddles at several depths, submersible pumps or propellers, and side-entry circulating systems with the motive force outside the tank. Multiple tanks may be needed for larger farms and/or when food wastes and other organic materials are introduced as a co-digestion feedstock.

The U. S. EPA (2008) through its AgSTAR program estimated in December 2008 that there are 121 farm-scale digesters operating at commercial livestock farms in the United States (Table 4.2). Six of these installations are systems that provide manure treatment for multiple farms. In 108 of the 121 operational systems, the captured biogas is used to generate electrical power, with many of the farms recovering waste heat for the electricity generating equipment for on — farm use. The U. S. EPA estimates that these systems generate ~218,000MWh of electricity per year. The remaining 13 systems burn the biogas in boilers, upgrade the gas for injection into the natural gas pipeline, or simply flare the captured gas.

Although a number of high-value products have been produced using glycerol, production of ethanol from glycerol provides a unique opportunity for the biodiesel industry to generate another biofuel from biodiesel waste. E. aerogenes HU101 isolated from methanogenic sludge produced ethanol and H2 when glycerol was used as a substrate (Ito et al. 2005). Low amounts of lactate, acetate, formate, and 1,3-PDO were also detected as coproducts. A maximum ethanol yield of 0.85 mol/mol-glycerol and concentration of 90mM was reported following anaerobic fermentation of glycerol (Ito et al. 2005). Although promising, E. aero- genes needs yeast extract and tryptone for enhanced glycerol utilization. The highest yield and productivity were achieved in a packed-bed reactor that used porous carrier matrix as support material to immobilize microbial cells (Ito et al. 2005).

Escherichia coli has been reported to metabolize glycerol fermentatively under favorable culture conditions (Dharmadi et al. 2006). Cells growing fermentatively on glycerol exhibited exponential growth at a maximum specific growth rate of 0.040 per hour (Murarka et al. 2008). Cell growth was not affected despite blocking of several respiratory processes, demonstrating the fermentation of glycerol by E. coli.

Pathways responsible for fermentative metabolism of glycerol in E. coli have been revealed (Gonzalez et al. 2008). E. coli mutants with disrupted respiratory genes glpD and glpA have been shown to ferment glycerol, indicating the existence of alternative pathways for the metabolism of glycerol in the absence of electron acceptors such as oxygen and fumarate. E. coli encodes a type II glycerol dehydrogenase enzyme, GldA, which was earlier thought to be cryptic in nature without any significant physiological role in wild-type strains (Jin et al. 1983). Nonetheless, this enzyme has the potential to oxidize glycerol into dihydroxyacetone (DHA) that can further be converted to DHAP by DHA kinase. When gldA (encoding GldA) and dhaKLM (encoding DHA kinase) mutants were evaluated for their ability to ferment glycerol, they failed to do so. This observation underscores the active role of GldA and DHA kinase in the glycerol fermentation pathway. In addition to this oxidative pathway for metabo­lizing glycerol, a parallel reductive pathway acting as an electron sink was also discovered in the form of 1,2-PDO (Figure 6.6). Disruption of the 1,2-PDO synthesis pathway decreased cell growth on glycerol, while overexpression of this pathway led to cell growth without any supplementation of rich media (Gonzalez et al. 2008; Murarka et al. 2008).

The grain harvest unit is assumed to be harvesting a field located 20 km from a grain storage. Reasonable combine efficiency will require two operators on infield hauling units and two truck drivers hauling with tractor-trailer trucks to the grain storage. Total crew required is then 1 (combine) + 2 (grain wagons) + 2 (truck drivers) = 5.

To achieve reasonable cotton harvester efficiency, the unit will require the following crew: 1 (harvester) + 2 (side-dump wagons) + 1 (operator for module builder) = 4. As compared with the grain operation, the crew size is reduced from five to four by disconnecting the field operations from the highway hauling. Remember, the cotton module is covered and left in roadside storage to be hauled later. Infield hauling is not delayed when a truck is not present to receive the seed cotton. Another key difference between the grain and cotton systems is that the highway hauling is done by the gin, not the farmer/producer.

A sugarcane harvesting unit is defined as 5 harvesters, 8 tractors pulling 3 wagons each for a total of 24 wagons, and 1 operator at the ramp for a total crew of 14. If the field is only 3.5 km from the sugar mill, the over-the-road hauling will require eight tractor-trailer trucks each hauling two bins. If haul distance is 15 km, then 15-18 trucks are required. At 40 km, it requires 25-28 trucks to keep the harvest unit moving.

Capacity for a harvest unit with good harvest management is given in Table 7.8. There are many production situations where the average capacity for an entire day’s operation will be only half the capacity shown in Table 7.8. As a reference point, the cost is also shown. These cost numbers are constantly changing, thus they should not be used out of context.

Hauling

Truck tractors used for over-the-road hauling are a mature technology. (There are over

250,0 trucking companies in the United States, and truck manufacturers work continuously

to improve the performance of their trucks and capture a greater share of this market.) Costs to operate the truck tractor are well defined.

It is interesting to compare the cost per dry t for the three systems. For this analysis, the cost to operate the truck is assumed to be $450/day. Labor cost for the driver is included; however, fuel cost is not included.

If a truck averages three loads of 20% moisture content grain per day, and the load size is 22.5 t, the hauling cost is

3 loads/day x 22.5 t/load x (1 — 0.2) = 54 dry t/day

If a truck (module hauler) averages four loads of cotton per day and the module is 7.2 t at 20% moisture content, the hauling cost is

4 loads/day x 7.2 t/load x (1 — 0.2) = 23 dry t/day

The truck on which the module hauler is mounted is basically the same machine (same engine, drive train, running gear) as the truck tractor used for grain hauling; thus, it is appro­priate to compare the hauling cost using the same per day truck cost.

If a truck averages 10 loads of sugarcane per day and the total load (two bins) is 24.5 t at 80% moisture content, the hauling cost is

10 loads/day x 24.5 t/load x (1 — 0.8) = 49 dry t/day

$450 day
49 dry mton. day

These results are summarized in Table 7.9 . Using the grain hauling as a reference, cotton is 2.3 times more expensive, and the sugarcane is 1.1 times more expensive. Unlike the harvest costs shown in Table 7.8. the hauling costs can be directly compared. Truck owner­ship and operating cost ($/day) for short-haul operations is approximately the same no matter what the truck is hauling.

Alkali Treatment

Alkali treatments, particularly the treatment with calcium hydroxide (overliming), have been widely used for improving the fermentability of ligocellulose hydrolysates. In an overliming process, lime is added to the hydrolysate, resulting in the formation of insoluble salts. Larsson et al. (1999) compared the effects of 12 detoxification methods, including alkali treatments with sodium hydroxide (NaOH) and calcium hydroxide (Ca(OH)2), on the chemical composi­tion and fermentability of a hydrolysate from spruce pretreated with dilute acid. Overliming was more efficient than NaOH in terms of removal of inhibitors under similar conditions (Larsson et al. 1999) . One drawback with overliming is the formation of calcium sulfate precipitate (gypsum). In addition, if the treatment is done at high pH and temperatures, a considerable degradation of fermentable sugars occurs (Cheng et al. 2008). The treatment has to be optimized to maximize the fermentability and lower sugar degradation.

Martinez et al. )2001) employed a titration method to predict the optimal amount of Ca(OH)2 for overliming at 60°C using 15 different batches of bagasse hemicellulose hydro­lysate. All 15 overlimed hydrolysates exhibited the same trend, despite differences in the amount of added Ca(OH)) . Total furans were reduced by 51% and soluble phenolic com­pounds were reduced by 41%. Presumably, these furans and phenolic compounds were converted to less toxic products by overliming. Total sugars were reduced by 8.7%. Although common and effective, overliming does not remove acetic acids, which are known to inhibit ethanol production at concentrations greater than 2g/L (Berson et al. 2006).

Ammonium ions do not form poorly soluble salts, and treatment with ammonium hydrox­ide compared favorably with overliming (Horvath et al. 2005) . NaOH would be another option, but under similar conditions NaOH treatment has so far been less efficient than over­liming. Optimal conditions were found to be in a range around pH 9.0/60°C for NH. OH treatments and in a narrow area stretching from pH 9.0/80°C to pH 12.0/30°C for NaOH treatments (Horvath et al. 2005).

Corn fiber is a coproduct of the corn wet milling industry. It is a mixture of corn kernel hulls and residual starch not extracted during the wet milling process. Corn fiber is composed of approximately 40% hemicellulose, 12% cellulose, 25% starch, 10% protein, 3% oil, and 10% other substances such as ash and lignin (Singh et al. 2003). Approximately 6.3 x 106 dry tons of corn fiber is produced annually in the United States. Typically 4.51b of corn fiber is obtained from a bushel (56 lb) of corn, which can be converted to about 3.0 lb of fermentable sugars (Ezeji et al. 2006). The major fermentable sugars from hydrolysis of corn fiber are glucose, xylose, and arabinose (Ezeji and Blaschek 2008a; Noureddini and Byun 2010). Small amounts of mannose (Ezeji and Blaschek 2008a) and galactose (Noureddini and Byun 2010) have also been measured in corn fiber hydrolysates.

Economically, it is important that these mixed sugars present in corn fiber hydrolysates be fermented to butanol for this renewable biomass to be used as feedstock for butanol produc­tion. Solventogenic Clostridium species have an added advantage over many other cultures as they can utilize both hexose and pentose sugars (Ezeji et al. 2007a, b; Ezeji and Blaschek 2008a) released from lignocellulosic biomass upon hydrolysis to produce butanol. C. beijer — inckii BA101 effectively ferments detoxified dilute acid pretreated corn hydrolysates to produce butanol (Table 3.2; Qureshi et al. 2008a). Similarly, Parekh et al. (1988) produced butanol from hydrolysates of corn stover using C. acetobutylicum P262. Marchal et al. (1986), in addition, fermented acid pretreated corn stover hydrolysates to produce 12.8 g/L acetone — butanol using C. acetobutylicum NCIB644 (Table 3.2). In another development, Zhu et al. (2002) produced butyric acid, an intermediate ABE fermentation product, from acid hydro­lysates of corn fiber by Clostridium tyrobutyricum in a fibrous — bed bioreactor. This fermenta­tion process can improve butanol yield, reduce butanol production costs, and improve the economics of butanol fermentation. Importantly, butyric acid can easily be incorporated into fermentation medium and converted to butanol during ABE fermentation (Tashiro et al. 2004 ).

The free enzymes comprise a catalytic module alone with no accessory modules or with a CBM. The simplest enzymes often specialize on degrading soluble oligosaccharide break­down products or bind to the polysaccharide substrate such as cellulose or xylan via the intrinsic affinity of its active site. In contrast, the polypeptide chain of many free enzymes includes both a catalytic module together with a CBM. This basic bi-modular arrangement can be further extended by the inclusion of additional types of modules or repeating units of the same module, all of which serve to modify the activity of the catalytic module on the polysaccharide substrate. The free enzyme, however, remains unattached to other enzymes and can work in an independent manner on a given substrate. The enzyme systems of aerobic fungi and bacteria usually contain numerous enzymes that are basically in the free state (Knowles et al. 1987; Wilson 1992; Warren 1993, 1996; Teeri 1997; Teeri et al. 1998; Wilson 2004; Viikari et al. 2007; Kumar et al. 2008).

Logistics of biomass supply involves an orderly flow of biomass from farm to factory. Figure 7.6 shows at least five options for the supply chain configurations to transfer baled biomass to biorefinery. In options 1 and 2, the baled biomass is transported directly from farm or from stacks next to the farm to the biorefinery. Biomass may be minimally processed (i. e., ground) before being shipped to the plant. In this case the biomass is generally supplied from the stacks where the biomass will be minimally processed. The biomass is trucked directly from farm to biorefinery if no processing is involved.

The supply options 3 and 4 transfer the biomass to a central location where the material is cumulated and dispatched to biorefinery later on. While in the depot, the biomass could be preprocessed minimally (i. e., ground) or extensively (pelletized). The depot also provides an opportunity to interface with rail or barge transport if that is an available option. The

choice of any of the options 1 to 5 depends on the economics and cultural practices. For example, in irrigated areas, there is always space on the farm (in unirrigated corners of the land in a center pivot irrigation scheme) where quantities of biomass can be stacked. In northern dry land farming, the farmer may allow storage of biomass on the field over winter until April but needs the land to seed the new crop.

Agricultural and forestry residues and organic portions of municipal solid waste can have a negative impact on the environment as they decay. On the other hand, converting these residues to ethanol can offer immediate and sustained GHG advantages and simultaneously enhance domestic fuel production (Paustian et al. 1998; Tilman et al. 2001; DiPardo 2002; Wyman 2003b; Zhang et al. 2007; BR&Di 2008; Fargione et al. 2008; Smith et al. 2008). In general, cellulosic ethanol can be a low carbon fuel and provide a valuable replacement for gasoline from petroleum. Although production and combustion of ethanol adds CO2 to the atmosphere, an equivalent amount of CO2 can be taken up as the next rotation of agriculture feedstocks is grown to replace that used to produce ethanol (Wyman 2003a2 . Thus, cellulosic ethanol provides an opportunity to recycle carbon instead of continually building up carbon in the atmosphere as fossil fuels do, and it has been estimated that ethanol from corn stover could reduce GHG emissions by over 80% compared to petroleum — derived fuels (Wang et al. 1999). Exporting the excess power produced by burning lignin and other portions of cellulosic biomass not utilized for making ethanol can reduce the amount of coal used to produce electricity in the grid, potentially resulting in negative emissions of CO2 compared to the status quo (Wyman 1994a). Moreover, because the large amount of virtually pure CO2 (around 300 kg CO2 per dry ton of corn stover) pro­duced during fermentation could be sequestered more easily than being considered for capturing CO2 from burning coal, ethanol could actually become a negative GHG emission fuel. Because these agricultural residues are associated with food production, they will be grown whether we use them for making fuel or not, largely avoiding many of the concerns about indirect land use now being hotly debated (EPA 2005; BR&Di 2008; Fargione et al. 20082 .

Even if GHG benefits are demonstrable, other sustainability and environmental consider­ations must be addressed for use of agricultural residues. In particular, removing agricultural residues from the land can impact soil cultivation and the needs for fertilizer, pesticides, and other chemicals, all of which can impact soil, water quality, air quality, site productivity, and GHG emissions (Kim and Dale 2004; EPA 2005; Smith et al. 2008). Some residues such as sugarcane bagasse, rice hulls, rice straw, and corn fiber that are typically removed from the field anyway can be employed without additional negative consequences when properly managed, and their use results in little, if any, net additional demands for cropland, fertilizer, pesticides, or water. On the other hand, leaving corn stover, wheat straw, and other plant matter remaining after food is harvested on the field helps maintain soil organic and inorganic matter and protect against erosion (Lal 2006). The amount of sustainably harvestable residues varies with location and depends upon climate, soil texture, rain fall, and the production practice used (Tilman et al. 2002; BR&Di 2008). For example, conventional till production of corn leaves more residues in the field than no-till systems (Kim and Dale 2005; BR&Di

2008) . However, estimates of the amounts of residues that can be removed sustainably are still being refined.

Hans P. Blaschek is a Professor of Microbiology and Director of the Center for Advanced BioEnergy Research (CABER) at the University of Illinois. He also serves as Assistant Dean of Biobased Research Initiatives in the Office of Research in the College of Agricultural, Consumer and Environmental Sciences, and is the theme leader of the Molecular Bio­engineering of Biomass Conversion Research Theme of the Institute for Genomic Biology. His research is focused on the acetone-butanol-ethanol fermentation and he is cofounder of a company called TetraVitae Biosciences that is currently commercializing the production of bio-butanol.

Thaddeus C. Ezeji received his PhD in Microbiology in 2001 from the University of Rostock Germany under the supervision of Prof. Dr. Hubert Bahl. He joined Dr. Hans Blaschek ’s laboratory at the University of Illinois Urbana-Champaign in 2001 as a postdoctoral research associate. Dr. Ezeji has been a faculty member of The Ohio State University since 2007, and his research has focused on fermentation, microbial strain development, metabolomics, and processes regulating the conversion of agricultural byproducts, coproducts, or wastes into biofuel and value-added products.

Jurgen Scheffran is a professor in climate change and security at KlimaCampus and the Institute for Geography of Hamburg University in Germany. Until summer 2009, he held positions at the University of Illinois at Urbana-Champaign (UIUC), in the Program in Arms Control, Disarmament and International Security, the Departments of Political Science and Atmospheric Sciences, and the Center for Advanced BioEnergy Research. After his physics PhD at Marburg University, he worked at Technical University of Darmstadt, the Potsdam Institute for Climate Impact Research. Recent activities include the Renewable Energy Initiative at UIUC and related projects funded by the Environmental Council, the Department of Energy, and the Energy Biosciences Institute.

Conflicting Reports on Mixing in Digesters

The literature contains conflicting information regarding the mixing intensity and duration that is necessary in digesters to treat a waste with a high concentration of VS. In general, workers agree that mixing is necessary to enhance substrate contact with the microbial com­munity, improve pH and temperature uniformity, prevent stratification and scum accumula­tion, facilitate the removal of biogas, and aid in particle size reduction (Stafford 1981) . However, mixing can be performed intermittently or continuously at different intensities (Hansen et al. 1999; Karim et al. 2005). On the one hand, Dague et al. (1970) showed that intermittent mixing resulted in an increased biogas production and increased COD and VS reduction efficiencies due to enhanced bioflocculation (settling) compared with continuous mixing. In addition, Stroot et al. (2001) and McMahon et al. (2001) reported that minimally mixed digesters demonstrated a much more stable operation than digesters that were continu­ously and vigorously mixed. They postulated that vigorous, continuous mixing inhibited relationships between syntrophic bacteria and their methanogenic partners, possibly by dis­rupting the spatial juxtaposition between these organisms. On the other hand, Lanting (2003) and Muller et al. (2007) showed that high mixing intensities results in particle size reduction and diffusion limitation reduction, which increased processing capacity for a digester treating waste-activated sludge. The choice in mixing intensity and duration is important, because continuous and vigorous mixing may significantly increase the necessary energy (i. e., elec­tricity) input for the digestion system, and thus reduce the net energy balance ratio. To shed light on the advantages and disadvantages of mixing intensity and duration for animal waste digestion, we conducted two different studies: one with swine waste in high))ate ASBRs,

which is described in the Mixing in a High-Rate System Treating Swine Waste section (Angenent et al. 2001- , and the other with dairy waste in the low-rate completely stirred anaerobic digester, which is described in the Mixing in a Low-Rate System Treating Dairy Waste section (Hoffmann et al. 2008- . We found that low-rate and high-rate reactors are fundamentally different in regard to mixing requirements, and that the use of different reactor types and configurations may explain the discrepancies in the digester literature in regard to mixing characteristics (Hoffmann et al. 2008).

Mixing in a High-Rate System Treating Swine Waste

We operated two lab-scale ASBR systems in parallel and the mixing duration and intensity were changed in one bioreactor from gentle, intermittent mixing to gentle, continuous mixing to vigorous, continuous mixing, while the other bioreactor (control) was operated throughout the experiment with gentle, intermittent mixing (Angenent et al. 2001). The substrate for this study was diluted swine waste (20 gVS/L) at a volumetric loading rate of ~4gVS/L/day. The inoculum of the ASBRs was from previously operated ASBRs that contained well-acclimated biomass with excellent settling characteristics (Angenent et al. 2002a) , and therefore the start-up period was less than 10 days (Figure 4.5A).