HTL of Biowaste Streams

In this section, two topics are summarized: one is on HTL of different kinds of waste streams, including manure, sewage sludge, urban waste, and agricultural wastes. The other is on the mechanisms study of HTL.

Appell et al. (1970) are among those who first started the study on the HTL process of waste streams. Urban refuse, cellulosic wastes, and sewage sludge were tried as feedstocks in a 500-mL autoclave. Effects of two kinds of initial gases, CO and H2, were compared. In the research, using CO as initial gas led to higher oil yield in the process. When using munici­pal refuse as feedstock, at 20 minutes, 380oC, and 1500 psi, the oil yield is 41% versus 18% for CO versus H2. At 500 psi CO, 1 hour, 250°C, 20% NaHCO3, oil yields of 37% and 35.5% were obtained from newsprint, pine needles, and twigs, respectively. For sewage sludge, without a catalyst, oil yield was 24.5%. Infrared and mass spectrometric analysis indicated the oils to be paraffinic and cycloparaffinic with the presence of carbonyl and carboxyl groups. High temperature product has a very small amount of aromatic material, probably phenolic, but the oil product at low temperature does not appear to have any aromatic compounds.

With cellulose as feedstock, at 2 hours, 350°C, 1500psi initial pressure, a much higher benzene-soluble oil yield was obtained when using CO rather than using H2 as initial process gas, 40% versus 5% solubility, respectively. CO as a process gas also resulted in a lower oxygen content in the oil (8.9% vs. 15.5%). Residue and oil have similar elemental composi­tion but different structures, making them appear differently. Water-soluble fraction resulting from hydrolysis of the cellulose is believed to be a precursor of the oil because it can be converted to oil by recycling the aqueous solution through the process with subsequent charges of refuse.

Fu et al. (1974) conducted the HTL process on bovine manure. Bovine manure was hydro­genated and liquefied with the existence of hydrogen or synthesis gas (H2 :CO = 1:1) at temperatures of 330-4252C and operating pressures of 1500-3000psi, in the presence of a recycled manure oil (or an alkylnaphthalene oil) and a Co-Mo catalyst. The Co-Mo catalyst could be a good choice for increasing oil yield. At 3802C/425oC, 3000psi, better oil yield was observed when the Co-Mo catalyst was present. Of all the variables investigated, tem­perature has the most dramatic effect on properties of the oil product. Although there were no significant improvements for conversion or oil yield, significant improvements in oil product quality were observed, with increased carbon content, decreased oxygen content, and reduced viscosity. High-resolution mass spectrometry analysis of oil produced at 380oC indicates that the main components are alicyclic hydrocarbons, N — containing heterocyclic compounds, and alkyl phenols with carbon numbers ranging from C8-C18. All carbon dioxide was produced before the reaction mixture reached 380°C (before zero time).

Minowa et al. (1995a) applied the HTL process to artificial garbage prepared by mixing cabbage, boiled rice, boiled and dried sardines, butter, and the shell of short-necked clams. Three temperatures (250, 300, and 340oC) and three retention times (0.1, 0.5, and 2 hours) were tested without a catalyst. Oil yield and its properties strongly depended on the catalyst addition and reaction temperature, while retention time showed no significant effect. Highest oil yield of 27.6% on an organic basis was obtained at 3400C, 18 MPa pressure and 0.5 h retention time with a catalyst. The oil had a heating value of 36MJ/kg and a viscosity of 53,000mPas at 50oC.

Suzuki et al. (1988) investigated optimum starting materials and catalyst loading for con­version of sewage sludge to heavy oil. Various kinds of sewage sludge were liquefied at 300oC, 12MPa, and a catalyst loading of 0-20 wt %. Of digested sludge, raw waste activated sludge, raw primary sludge, and raw mixed sludge, have higher oil yields averaging 43%. The nature of the sludge had no significant influence on the elemental composition and heating value of oils obtained. Catalyst loading had no significant influence on the properties of oil products. Results showed a nearly linear relationship between crude fat content and the amount of oil fraction in the starting materials. Calcium salts could possibly have some catalytic effect on the liquefaction reactions.

Murakami et al. (1990) converted activated sludge from a cornstarch processing plant into oil with the HTL process. A 100 cm3 autoclave and N2 initial gas were used, and the working pressure was maintained at the saturated vapor pressure of water at the required temperature plus 3.0 MPa. Effects of temperature, Na2CO3 catalyst loading, and holding time were studied. Results showed that maximum oil yield was 30% at 300oC, 60-minute retention time without a catalyst. Oil yield was not significantly affected by catalyst loading. Properties of the oil product were not influenced to any great extent by temperatures between 225-300oC, while the aqueous phase product yield and solid residue decreased as temperature rose. Heavy oil production at temperature as low as 250o C is possible, provided the reaction is carried out with sufficient retention time. According to their energy balance, the liquefaction could be a self-sufficient process under certain conditions.

A demonstration plant with a capacity for processing up to 5 t/d as dewatered sludge was operated at 300oC, 10 MPa, (feedstock, moisture content ~80%, VS ~80%; Itoh et al. 1994). The sludge had been dewatered by belt press dehydrator after adding a polyelectrolyte coagulant. As a result, 48% mass of the organic materials in the sludge were converted into heavy oil, and a quarter of the oil was separated from the reaction mixture by high pressure distillation with a distillate ratio of 0.33. Heating value of the heavy oils distilled were 37- 39MJ/kg, while that left in the bottom was 31-35MJ/kg. Energy balance was calculated based on the pilot plant data collected. For a plant of 60t/d dewatered sludge, the sludge is treated without any auxiliary fuel, and 1.5 tons of heavy oil is produced as surplus energy. In conclusion, the treatment of sewage sludge by this method could be sufficiently profitable.

He et al. (2000a, b; 2001a, b,c) studied HTL for swine manure. The process was evaluated by oil production efficiency and waste reduction efficiency. The oil product was analyzed for its benzene solubility, elemental composition, and heating values. Thermogrametric proper­ties and viscosity were also measured on selected oil samples. The difference of chemical oxygen demands before and after the HTL process was used as the waste reduction efficiency. The key factors of the HTL process were the operating temperature, the retention time, and the addition of a process gas. The operating temperature effect was studied in the range of 275~350°C. The suggested operating temperatures are between 295°C and 305°C. A process gas addition was necessary to achieve an oil product. Without the process gas, no oil products formed. The process gases investigated include carbon monoxide, hydrogen, nitrogen, carbon dioxide, and compressed air. Carbon monoxide was the most effective gas for the process. Retention time is another important factor. The necessary retention time to achieve an oil product was largely related to the operating temperatures. When the operating temperatures were 295°C-305°C, the retention time was 15—30 minutes. Based on an average of 135 dif­ferent oil samples, 62 wt % of the volatile solids were converted to oil. The waste strength was reduced by 60% to 70%. The highest oil production efficiency was 80 wt %. The average carbon and hydrogen contents were as high as 72 wt % and 9 wt %, respectively. The heating values for 80% of the oil products ranged from 32,000 to 36,700kJ/kg. The results showed that HTL of swine manure to produce oil is technically feasible, and could be a promising technology for waste reduction and renewable energy production.

Since a continuous system is more applicable for scale-up operations, a small-scale continuous HTL (CHTL) reactor system was developed (Ocfemia et al., 2006a) to aid in the evaluation of the technical feasibility and economical viability of a pilot plant that is capable of producing oil from swine manure. The effects of operating conditions, including temperature, pressure, hydraulic residence time, and use of process gas, were evaluated in order to determine the optimal process condition. The composition of the different product streams (i. e., oil, aqueous, and gas) was determined to better understand the mechanics of the reaction process and to provide information for further develop­ments. The CHTL reactor system was composed of a high-pressure slurry feeder, a process gas feeder, a continuous-stirred tank reactor, a products separation vessel, and process controllers. It had a capacity to process up to 48 kg of manure slurry per day. The operating parameters—temperature, pressure, residence time, and the use of CO—were all found to affect oil yield (Ocfemia et al., 2006b). The interaction between operating temperature and pressure was evident. The highest yield of 70% (based on volatile solids content of the manure feedstock) was found to be in the region where temperature was about 300°C and pressure was 10 MPa. Yield was found to increase with hydraulic resi­dence time, but there was a diminishing benefit after 60 minutes. The addition of CO in the process did not improve the oil yield, but produced a more fluid oil product. The heating value of the oil product ranged from 25,176kJ/kg to 33,065 kJ/kg with the highest value at the operating condition of 305°C, 10.3 MPa, and 80 minutes hydraulic residence time. The energy balance based on oil heating value and energy used for heating the feedstock material to the operating temperature showed that the process was a net energy producer.

Elemental analysis of the crude oil showed that the average (from all tests) carbon and hydrogen content of the oil was 62.7% ± 6.4% and 9.6% ± 0.4%, respectively. The nitrogen content was high with a value of 3.9% ± 0.3%. The sulfur content of the oil was low with a value of 0.3% ± 0.1%. The composition of the oil based on SARA analysis showed that the oil was primarily resins (—45%) and asphaltenes (—44%) with small amounts of saturates (—3%) and aromatics (—2%). The boiling point distribution of the oil showed that the majority of the compounds were in the boiling point range of 316-482°C.

In Ocfemia et al.’s study (2006b) , the aqueous product was found to contain volatile organic compounds, primarily ketones and benzenyl compounds. Of dominance was acetone, which accounted for 0.7 mg/L of the aqueous product. The total N content was 6.1 ± 1.9 g/L. Phosphorus as phosphate was 1.0 ± 0.3 g/L. The total aqueous K was 1.5 ± 0.8 g/L. The main gas product was CO, accounting for —98% of the total. Carbon monoxide accounted for about —2%. Hydrocarbons, including methane and ethane, were found to have a combined concentration of 299 ppmv. Very small amounts of aromatic compounds, including benzene, ethylbenzene, toluene, and styrene, were also detected.

Dote et al. (1992) studied the oil samples separated by steam distillation after liquefaction of sewage sludge. Oil was fractionated to strongly acidic, weakly acidic, basic, and neutral fractions by acid-base extraction. The total recovery of fractions was 77%. Each fraction was analyzed by gas chromatography and mass-spectrometry (GC — MS), and 71 types of com­pounds were identified with reasonable certainty. No strongly acidic fraction was obtained. The weak acidic fraction, comprising 2% of the oil, was exclusively composed of phenolic compounds. The basic fraction, comprising 20% of the oil, was exclusively composed of pyridines, pyrazines, quinolines, amines, and methylphenylacetamide. The neutral fraction, comprising 56% of the oil, was exclusively composed of aliphatic compounds, alicyclic compounds, alcohols, ketones, aromatic compounds, sulfur-containing compounds, nitrogen — containing compounds, and oxygen-containing heterocyclic compounds.

Dote et al. (1996) studied the distribution of nitrogen in the products for direct liquefac­tion of protein-contained biomass. Albumin from an egg was used as feedstock, and tests were run at 150-340°C, 0.5 hours, and 2 hours, w/o Na2CO3 as a catalyst. The maximum oil yield was 10%, much less than that for practical feedstocks (all above 30%). Nitrogen dis­tributed to oil was 5% at most, much less than that for practical feedstocks (30%-45%). Other practical feedstocks contain other elements such as cellulose and lipid, which may increase the amount of oil converted from protein or react with nitrogen-containing com­pounds produced from protein during the conversion. No distribution of nitrogen to oil occurred below 150-C, and the distribution was completed by 250-C. The majority of the nitrogen in albumin (80%) was distributed to the aqueous phase, and albumin was decom­posed to ammonia, not to amino acids. Sodium carbonate seemed to prevent the distribution of nitrogen to oil.

Minowa et al. (2004) used glucose and glycine as model compounds of carbohydrates and proteins, to study the mechanisms of hydrothermal reactions. There were 1.8 g glucose, 0.75 g glycine, and 30 mL distilled water charged into 100 mL autoclave. N2 was used as initial gas and was added to 3 MPa pressure. Temperatures of 150-350oC were studied. It was concluded that at 150oC, the main reaction was Maillard reaction and melanoidin was formed. At 200oC, produced melanoidin was decomposed to form char, gas, and aqueous-soluble materials. Oil production started at 250oC, and oil yield increased with reaction temperature. It appears that oil is formed through the secondary decomposition of the decomposed products in the aqueous phase and not directly formed from melanoidin. Char was formed in the low reaction temperature range from 150oC to 200oC, and char yield was almost the same over 200oC. No pathway from oil to char is significant. Produced ammonia could inhibit the char formation from oil.

It appears that fatty acid and lipid are the main reactants of HTL. The predominant HTL reaction below 300oC is considered to be distillation of aliphatic compounds. The existence of considerable straight chain compounds (C13-C22) suggests that the aliphatic compound is the main resource of derived oil. A large quantity of nitrogenous compounds (mainly composed of amide and cyanide) in the oil suggests that protein is widely involved in the HTL reaction, possibly by peptide bond splitting and amino acid conversion dehydration. Within 300-450oC, the protein conversion reaction intensifies, and its principal structural bond (peptide bond) begins to react. Saccharide reaction mainly belongs to the splitting of branched chain and group transfer while considerable dehydration and cyclization of the main chain still appears not to be dominant. A simplified reaction model of HTL for sewage sludge consists of two serial competitive reactions—producing volatile matter and char, respectively. Estimated Arrhenius kinetic parameters of the reaction model based on Thermogrametric testing results were introduced.

It is worth mentioning that Huber et al. (2006) conducted a comprehensive review on biofuel producing methods, including gasification, liquefaction, and pyrolysis. Some chemi­cal reaction mechanisms, oil synthesis, and upgrading methods were also included in that review.

Steam pretreatment is a process in which lignocellulosic biomass is treated with high-pressure (0.69-4.83 MPa) saturated steam (160-260°C) for several seconds to a few minutes (Sun and Cheng 2002). The pressure may be swiftly reduced so the material undergoes an explosive decompression (Sun and Cheng 2002). The only difference between steam pretreatment and explosion is the quick depressurization and cooling of the biomass after steaming at a pre­determined length of time. Steam pretreatment or explosion helps to disrupt the integrity of lignin sheath (Figure 3.1) and solubilize the hemicellulose component to enhance accessibility of cellulose to enzymatic hydrolysis. Grous et al. (1986) pretreated chopped poplar with steam explosion, and subsequent enzymatic hydrolysis resulted in greater than 90% efficiency in 24 hours, compared with 15% hydrolysis of untreated poplar. Residence time, biomass par­ticle size, temperature, and moisture content affect the efficiency of steam pretreatment and explosion (Duff and Murray 1996).

The most abundant source of carbohydrates in the biosphere is plant biomass, which harbors the lignocellulosic materials comprising the cell walls of all higher plants. Plant cell walls are composed of polymeric networks. The predominant polymer and most abundant form of structural polysaccharide is cellulose. The second most abundant form of plant cell wall polysaccharide comprises the hemicelluloses and the third includes the pectins. The most abundant cell wall proteins are known as extensins and are classified as strongly basic gly­coproteins. The non-polysaccharide aromatic polymer, lignin, is deposited after the cell has completed growing and is covalently cross — l inked to hemicelluloses (Preston 1974) . Plant cell walls are divided roughly into two types, the primary and the secondary cell walls.

The chemical composition and structure of primary and secondary cell walls are strikingly different. For example, cells that have only primary cell walls are often not lignified, such as in parenchyma tissue, whereas cells that have secondary cell walls may or may not be ligni — fied, depending on their function in the plant tissue. Secondary cell walls are always more abundant than the primary cell walls in most plants and therefore represent the major reservoir of fermentable sugars in the plant.

Cellulose is a linear P-(1~4)-D-glucan and comprises 30%-50% of the plant cell wall. Pectins are “chelating — agent — extractable polysaccharides plus chemically similar inextract­able polysaccharides” (Timell 1964, 1967). Pectins are rich in D-galacturonic acid and contain, in decreasing order, arabinose, galactose, and rhamnose. Hemicellulose is a general term used to refer to cell wall polysaccharides that are not celluloses or pectins. Hemicelluloses include a variety of compounds, including xylans, xyloglucans, arabinoxylans, and mannans. Hemicelluloses almost always are branched with a wide spectrum of substituents along the backbone polysaccharide. Hemicelluloses in grasses and hard woods are primarily arabinox­ylans, whereas those in so lit woods are primarily galactoglucomannans (Aspinall 1959). Hemicelluloses are thought to undergo hydrogen bonding interactions with cellulose, as well as to other hemicelluloses, which form the basis for the microfibrillar structure of modern plant cell walls (Figure 5.2). The cell wall matrix is also established as a consequence of the esterification of hemicelluloses to lignins via p-coumaroyl and feruloyl groups (Mueller — Harvey et al. 1986). Hemicelluloses can, however, be extracted with alkali or dimethyl

sulfoxide, which disrupts the hydrogen bonds and, for alkali extraction, saponifies any ester cross-linkages to lignin (Bacon et al. 1975).

Introduction

Typically, the feedstock cost constitutes about 35%-50% of the total production cost of ethanol or power. The actual percentage depends upon biomass species, yield, location, climate, local economy, and the type of systems used for harvesting, gathering and packaging, processing, storing, and transporting of biomass as a feedstock. The following is a list of feedstock requirements to ensure the success of biorefineries.

• Identify quantities, quality of biomass, delivery costs for the year-round supply of biomass.

• Conduct resource assessment considering mix of available biomass species, annual yield variations, environmental factors, seasonality, and competitive demands for biomass.

• Optimize for the least cost equipment and infrastructure for timely harvest, densifying, storing, and transporting of the biomass.

• Develop regional and national strategies for locating biorefineries and organizing supply chains with respect to biomass cost and availability.

Figure 7.1 shows a diagram of biomass-to-product thread from production to biorefinery. The type of biorefinery may range from biomass to heat and power or to production of chemicals and liquid fuels. Biomass production can be from agricultural and forestry activi­ties, and municipal and industrial wastes. The activities within the large oval identify the current and probable future biomass supply enterprises. Biomass is collected in a distributed system at the farm or at the forest level. The collected biomass is transported either a short distance (0-100 km) or a long distance, hundreds of kilometers for storage and/or preprocess­ing. Preprocessing may include one or a combination of several of size reduction, fraction­ation, sorting, and densification. The storage of wet biomass may also impart biochemical and physical modifications to the biomass. We call this as in-store preprocessing. The pre­processed biomass is transported to a biorefinery where it is fed directly into the conversion reactor.

The arrows on the diagram show the flow of material and information. The information flow (lines with arrows between unit operations) from the biorefinery to a biomass supply enterprise includes quality specifications for biomass, that is, moisture content, particle size,

composition (cellulose, hemicellulose, lignin, ash content, chlorines, silica, etc). Important information for logistics includes quantities and delivery schedules and price. In response to demands, the production side provides biomass to the supply system (biomass types, quanti­ties, format, and cost). The supply system uses energy and power to collect, preprocess, and transport biomass. The system will give off emissions that need to be minimized.

The objective of this section is to present an analysis of the cost of collecting and handling of biomass throughout the entire supply chain using several collection, storage, and transport options. The Integrated Biomass Analysis and Logistics (IBSAL) model is used for this analysis. The section also discusses energy input to the model and carbon emissions from powered equipment used in the supply chain.

Two options have been proposed for the collection of herbaceous biomass as a feedstock for bioenergy, chopping (with a forage harvester), and baling (with a round baler or square baler). A logistic system for the forage chopping option has basically the same challenges as the sugarcane system. As previously mentioned, this system can work well for plantation agriculture but cannot be widely adapted in the United States. Several hundred farmers

chopping biomass and delivering on their own schedule to the bioenergy plant is not a practical option. If some on-farm storage option is used for the chopped material, perhaps a bunker silo, then the delivery can be better managed. Moisture content is an issue. The moisture must be high enough to achieve ensiling for acceptable storage, but hauling of silage means the hauling of large amounts of water, the same problem encountered with the sugarcane system.

Baling provides a disconnect between the harvest and infield operations, which is a significant advantage. One operator can bale an entire field with no requirement to coor­dinate with any other operation. If the big square baler is used, there is a requirement to haul the bales before they get rained on. (This requirement is not a major factor in the West, but it is a significant factor in the Southeast.) Once the big square bales get rained on and the water penetrates, it is unsafe to store the bales in covered storage. The big round bale can be stored in a single-layer ambient storage, because the rounded top thatches and sheds water. A hay shed with big square bales stacked about four high gives a storage cost of about $8/t and a single-layer ambient storage of round bales on a gravel surface costs about $2/t.

The round bale option is chosen for more detailed analysis. The key disadvantage of this option is that the round bale was developed to be used on the farm where it is produced. Systems for efficient over-the road hauling have not been developed.

The following constraints are listed for consideration by someone designing a round-bale logistic system.

1. The equipment should provide for multi-bale handling. The labor productivity of an operator on a machine loading individual bales, or unloading individual bales, is simply too low.

2. The equipment for over-the-ioad hauling must provide some increase in load bulk density over the bulk density of an individual bale. This increase may be modest, perhaps as low as 10%, depending on the cost to achieve it.

3. The goal for truck load time should be 10 minutes, meaning that a tractor-trailer truck is loaded with round bales in 10 minutes. If the cost ($/t loading cost) to achieve this goal is too high in comparison with the truck cost reduction ($/t) achieved with a 10 minute load time, then the optimum compromise between load cost and truck cost must be determined.

4. The goal for truck unload time should be 10 minutes, thus the interaction with the receiving facility at the bioenergy plant is critical. This goal equals the unload time for other materials (grain, fuel chips, sugarcane, and cotton).

5. The system must provide for easy flow of material into, and out of, at-plant storage. A system that gives the lowest delivered cost for feedstock through the plant gate may not give the lowest cost for a continuous stream of biomass into the plant 24/7.

6. Ideally, the equipment system must provide a means to establish a single-file stream of round bales into the plant. This requirement facilitates the introduction of the bales to the size reduction unit. Size reduction is best done at the plant.

Mutant of Saccharomyces cerevisiae

Schneider (1996) used a mutant of S. cerevisiae YGSCD 308.3 to selectively remove acetic acid that inhibited D-xylose to ethanol conversion. The S. cerevisiae mutant grew on acetic acid but not on xylose, glucose, mannose, and fructose. The process reduced acetic acid to very low levels and caused only small changes in sugar concentration. The inability to phosphorylate glucose, xylose, mannose, fructose is the direct result of the presence of mutations in three genes hxkl, (hexokinase I), hxk2 (hexokinase II), and glc (glucokinase

I). In the presence of a normally functional hexokinase II gene, for example, enzymes for the metabolism of carbon sources other than D-glucose (acetic acid and D-galactose), are repressed to low levels when D-glucose is present in the medium and vice versa when hexokinase is dysfunctional (Schneider 1996- . The hydrolysate became fermentable after the treatment, and hexoses and D-xylose were subsequently converted to ethanol by the S. cerevisiae mutant, which might be applicable to obviation of acetic acid inhibition effects in ethanol production from hemicellulose hydrolysates.

Methanogenic Food Web

Anaerobic digestion of readily degradable organic feedstock (biomass) is a very efficient energy recovery system because the final products—methane and carbon dioxide (CO2)—are

Biofuels from Agricultural Wastes and Byproducts Edited by Hans P. Blaschek, Thaddeus C. Ezeji and Ju rgen Scheffran 39 © 2010 Blackwell Publishing. ISBN: 978-0-813-80252-7

automatically and constantly removed from the process by degassing (bubble formation), while cell yields for anaerobic microbial growth are low. This results in a relatively high level of energy recovery by the production of a reduced carbonaceous energy carrier— methane. No artificial fermentation product removal system is necessary, and under healthy digestion conditions all intermediate fermentation products are maintained at very low con­centrations (i. e., no other product that lowers efficiency is formed; De Schamphelaire and Verstraete 2009). Under these conditions, the anaerobic food web, which consists of hydro­lytic and fermentative bacteria, obligatory H2 — producing acetogenic bacteria, H2- utilizing acetogenic or homoacetogenic bacteria, hydrogenotrophic methanogenic archaea, and aceto- clastic methanogenic archaea, is balanced to maintain extremely low product concentrations at each trophic level, and in other words pulls all reactions to completion. This results in an optimum system to convert a complex and often nonsoluble substrate into methane via the intermediate fermentation products acetate and hydrogen. Because acetate, butyrate, and propionate are all intermediate products of this food web and are collectively referred to as carboxylates, anaerobic digestion is part of the carboxylate platform as an alternative to the better known sugar and thermochemical platforms of the biorefinery concept.

Cellulosomes are multi-enzyme complexes that catalyze the efficient hydrolysis of cellulosic substrates and constitute a major paradigm of prokaryotic degradation of cellulose and related plant cell wall polysaccharides. The first cellulosome was discovered in the anaerobic ther­mophilic bacterium, C. thermocellum (Bayer et al. 1983; Lamed et al. 1983a, b). During the past 25 years or so, the cellulosome concept has been subject to numerous reviews that have chronicled its discovery, development, and potential (Lamed et al. 1983b; Lamed and Bayer 1988a, b, 1991, 19931 Felix and Ljungdahl 19931 Doi et al. 1994i Bayer et al. 1996, 1998, 2004, 2006, 2008b; Begum and Lemaire 1996; Belaich et al. 1997; Karita et al. 1997; Shoham et al. 1999; Doi and Tamura 2001; Schwarz 2001; Doi and Kosugi 2004; Demain et al. 2005; Bayer and Lamed 2006).

In C. thermocellum, cellulosomes appear in both the cell-free and cell-bound forms, the latter being associated with polycellulosomal protuberance-like organelles on the cell surface (Bayer et al. 1985; Bayer and Lamed 1986). Later, cellulosomes were detected in other cel­lulolytic organisms (Lamed et al. 1987; Mayer et al. 1987), including Acetivibrio cellulolyti — cus, Bacteroides cellulosolvens, C. acetobutylicum, C. cellulolyticum, C. cellulovorans, C. papyrosolvens, and the two rumen bacteria: Ruminococcus albus and R. flavefaciens (Doi et al. 1994; Pohlschroder et al. 1994; Belaich et al. 1997; Ding et al. 1999, 2000, 2001; Morrison and Miron 2000; Ohara et al. 2000; Sabathe et al. 2002).

Cellulosome systems contain numerous structural and enzymatic components. A simplified schematic view of the C. thermocellum cellulosome and its interaction with cellulose is shown in Figure 5.3. The cellulosomal enzyme subunits are incorporated into the complex by means of a unique class of nonenzymatic, multi-modular polypeptide subunit, termed scaffoldin. The “primary” scaffoldins usually contain a familyGa CBM that provides the cellulosei binding function and multiple copies of a definitive type of cohesin module. On the other hand, the cellulosomal enzyme subunits carry a complementary type of module, called dock — erin. The cohesin-dockerin interaction provides the definitive molecular mechanism that integrates the enzyme subunits into the cellulosome complex (Tokatlidis et al. 1991, 19931 Salamitou et al. 1994b). A second, divergent type of cohesin-dockerin interaction serves to attach the enzyme-laden scaffoldin to the cell surface by virtue of a second type of “anchor — mg” scaffoldin that contains one or more cohesins and an SLH module (Salamitou et al. 1994a; Lemaire et al. 1995; Leibovitz and Beguin 1996; Leibovitz et al. 1997).

Until fairly recently, cohesins and dockerins were considered to be exclusive cellulosome “signature sequences”—that is, their presence is a good indication of a cellulosome in a given bacterium (Bayer et al. 1998). This perception has changed with the discovery of cohesins and dockerins, first in a non-cellulosomal archaeon (Bayer et al. 1999), and later in a broad number of non-cellulosomal archaea, bacteria, and primitive eukarya (Peer et al. 2009). The cohesins and dockerins are thus represented in all three domains of life, to the extent that their presence in the cellulosome might be the exception rather than the rule. Be that as it may, in the few known cellulolytic bacteria that produce cellulosomes, the cohesins and dockerins play a definitive role in cellulosome architecture and function.

The major difference between free and cellulosomal enzymes is that the free enzymes usually contain a CBM as an integral part of the polypeptide chain for guiding the catalytic module to the cellulosic substrate, whereas the cellulosomal enzymes bear a dockerin for their integration into the complex. Otherwise, both the free and cellulosomal enzymes can contain very similar types of catalytic modules. The cellulosomal enzymes collectively rely on the single scaffoldin-borne CBM3a for binding to the crystalline cellulose substrate.

In different cellulosomal bacteria, the modular architectures of the different scaffoldins, the location, and specificities of their various modular types—the cohesins, CBM, and/or dockerin—determine the overall composition and status of the cellulosome as a whole. Although the similarities among the different cellulosome species abound, their diversity and divergence from the C. thermocellum paradigm are striking. Each new cellulosome-producing bacterium provides new and valuable information regarding the diversity of cellulosomes in nature.

The cellulosome was originally suggested to provide the anaerobic cellulolytic bacterium with distinct advantages in their efficient degradation of cellulosic substrates (Lamed et al. 1983a, b). The presence of the CBM on the common scaffoldin serves to deliver the enzymes en block to the substrate. Moreover, their close proximity on the scaffoldin subunit ensures their enhanced synergistic action. In addition to these advantages, the fact that the cellulosome is essentially a cell-surface organelle and the bacterial cell itself is attached to the substrate by virtue of the scaffoldin-borne CBM ensures that the cellulose-degradation products are preferentially available to the parent bacterium.

Transport costs in IBSAL are calculated based upon specifying either a fixed distance or a variable distance. Fixed distance is the transport cost from particular satellite storage (or stacks) to the biorefinery. For example, 50001 of biomass is transported from the satellite storage (or depot) A to the biorefinery. The variable distance scenario is when we specify a total quantity of biomass to be collected from locations within a specified radius (or a minimum to maximum distance). For example, 5000 t (or any quantity) of biomass has to be supplied to a biorefinery from within the circle (supply radius). The biomass is supplied from locations A (maximum distance) or any location B or C closer to the biorefinery.

The cost of transporting baled biomass for a variable distance of 20-100 km was calculated by Sokhansanj et al. (2006) . The cost of transporting biomass for a fixed distance was also calculated. For transport analysis, the large square bales were loaded on a flat bed (36 bales). The bales are transported to the biorefinery where they are stacked. The bales are ground for entering the process line at the biorefinery. The cost of transporting a maximum fixed distance is higher than the cost of transporting a variable distance between the biorefinery P and storage A.

Transport cost is a strong function of bulk density. Table 7.5 is a list of transport costs for biomass in the form of grind and pellets. Experiments (Mani et al. 2004) with the bulk density of grind size of 2.5 mm shows that a bulk density of 120-180 kg/m3 can be achieved depend­ing on the method of fill and the vibration of the container. We assumed a bulk density of 140kg/m3 for dry grinds. Pellets can have a density as high as 650kg/m3. We assumed a bulk density of 580kg/m3 for pellets. Table 7.5 shows that the grind transport cost is also very dependent on the method of loading. In this analysis we use a front-end loader to load the 100m3 capacity truck. It is costly at $9.03/t. Pellets are loaded using the same method but cost only $2.71/t because of high bulk density. The total cost of biomass transport or pellets for a 20-100-km distance is roughly $6/t.

As one approach to harvesting, collection, and transport of corn stover, a simple two-step operation of baling and bale collection and delivery to the processor was identified to replace the former five-step procedure of raking, baling, field loading, hauling, and unload­ing (www. ceassist. com 2000). By turning off the spreader on the corn combine, a windrow was left behind that could be more easily baled, resulting in a collection of about 1.5-2.0 dry tons per acre of stover that could be increased to 2.5-3.5 dry tons per acre by adding a rake in front of the baler. Both round and square bales were collected with a target of 1200 dry lb per round bale and large rectangular (4′ x4’x 8′) bales and 650 dry tons

aCellulase loading unit FPU/g dry weight substrate.

bTotal xylan, glucan, and sugar yield based on original component content, including oligomeric and monomeric sugars.

for intermediate-sized square bales. Round bales were rapped with three layers of plastic net to ensure they did not break apart when collected and handled later in processing. The bales were picked up at speeds of 5-7 mi/h in the field by a tractor-loading arm operated by one person capable of picking up bales in any orientation, rotating them into the proper position, and loading them onto a “load and go” trailer that holds at least 17 round bales and folds to legal width when empty. Some haulers completed a loading cycle of about 21,000 dry lb (17 round bales) in less than 20 minutes. The unit could safely traverse corn fields, including crossing ditches, and still travel at highway speeds of up to 60mph for transport to the collection center where the load could be weighed, sampled for moisture, and unloaded in less than 10 minutes. The trailer could then return to the field for another load. Overall, this approach was said to be able to cut corn stover costs by up to nearly half. However, a potential concern is the likelihood of picking up large amounts of dirt and rocks when the material is taken from the field piles.

Various companies have collected corn stover, and particularly corncobs, and it is impor­tant to determine if these collection methods would work for making biofuels. For example, QO Chemical in conjunction with the Anderson Brothers of Columbus, Ohio, ran a “Cob Saver Program” in the late 1980s to collect corncobs in addition to the kernels from the field. This entailed substituting a sieve plate with larger diameter holes into combines so that the cobs would pass through the holes during harvesting rather than just the kernels; the stalks and husk passed over the plate and were returned to the field. The cobs were easily separated from the kernels when the truck was unloaded and stored in piles for future conversion to furfural and other products. This system was extensively tested in the field, and excellent cob recovery was demonstrated with little damage to the kernels. It also avoided gathering feed­stock off the ground after harvesting and the inevitable contamination with dirt and rocks that increase transportation costs for nonusable materials and can damage processing equipment.

Although the focus above was on corn stover and cobs because of the size of the resource and relevant experience, other feedstocks could be used for the commercial process. Because most Ag residues, herbaceous materials, and hardwoods behave similarly in pretreatment and fermentation systems with small adjustments in chemical (e. g., sulfuric acid) additions and residence times, it should not be a major issue to employ wheat straw, switchgrass, or other biomass sources abundant in the local processing area (Torget et al. 1992). Important con­siderations impacting feedstock flexibility can be such seemingly simple aspects as biomass conveying because the details of its design vary with bulk packing density and other feedstock properties, and failure to recognize these important distinctions can cripple a project and result in financial collapse (Wiltsee and Bain 2000).