Understanding the characteristics of LB is necessary for the effective application of a conversion technology or a pretreatment method. LB is the most abundant organic and renewable resource on the planet (Klass,

1998) . Man has been producing chemicals, materials, and energy from LB since his origin. These activities continue to be the most promising activities to pursue in order to address our contemporary challenges. For
example, we currently look to use LB to reduce depen­dence on fossil fuels.

There are three families of LBs: grassy plants, shrubs, and trees, each possess four primary components: cellu­lose, hemicellulose, lignin and extractives. Each species possesses these components in different proportions. Hardwood trees (angiosperms), shrubs and grassy plants (graminoids) usually possess less lignin than soft­wood trees (gymnosperms) (Liu, 2012).

Figure 27.1 illustrates the variation in cellular struc­ture between hardwood and softwood. Notice that in both images the structure is porous, where the pores are empty spaces and run longitudinally. Figure 27.2 is a compilation of sketches of wood cells. The cellular structure of each plant species is of different sizes and shapes and varies in the size and number of pores. From Figure 27.1, one can note that there is no space between the cells. Regardless of size and shape, each cell is glued tightly to its neighbors. The intercellular spaces are called middle lamellae. The great majority of the middle lamellae, over 80%, contain lignin (Liu, 2012). Lignin is the glue that binds the cells together and pro­vides the rigid structure of wood. The remaining volume of the middle lamellae consists of hemicellulose and extractives. Conversely, the cell wall is mostly made of cellulose and to a much lesser degree contains lignin and hemicellulose. The great majority of biomass dry weight is derived from the cell wall. A much smaller portion of the biomass dry weight comes from the middle lamellae. Most of the total lignin content of LB comes from the cell wall. Despite the high concentration of lignin in the middle lamellae, over 60% of the lignin from LB comes from the cell wall portion of the material (Liu, 2012).

Table 27.1 shows that cellulose is the largest portion of LB. Cellulose, which represents between 40% and 50% of the dry weight of wood, is a homopolymer of

b-D-glucopyranose where dehydration of the b-D — glucose units forms a linear chain with a degree of poly­merization (DP) between a few hundred and several thousand b-D-glycosidic bonds. The dehydration occurs between the one and four carbons of b-D-glucopyranose units and leaves an oxygen atom to join the two units, which is written as, b-1-O-4 glycosidic bonds. The for­mula for cellulose is H—(СбН^05)п—OH, where "n" rep­resents the DP. This highly ordered, tightly bound pattern is made of bonds that are quite strong and are difficult to break.

Cellulose grows into microfibrils with crystalline and amorphous regions. The crystalline portions of the molecule line up side by side. Hydrogen bonds, between the hydroxyl groups, provide strong, sturdy and stable links between and within these crystalline units. When these microfibrils form macrofibrils and interact with noncellulosic material in the cell walls of plants, the result is strength and rigidity.

While the crystalline regions are stable and strong, the amorphous regions provide an opportunity to break down the large structure into smaller saccharides. Sol­vents, reagents and enzymes may be used to penetrate and hydrolyze the structure. Hydration requires the addition of energy or a strong acid. Alternatively, en­zymes, such as cellulase, may facilitate the conversion. Enzymatic hydrolysis tends to be much slower than acid hydrolysis. Reducing the chip size or increasing
the exposed surface area of LB increases the effective­ness of these solvents, reagents and enzymes.

Hemicelluloses compose another large portion of LB, between 20% and 30% of the dry weight of wood, see Table 27.1. These are heteropolymers, or heterosacchar­ides of five — and six-carbon sugars. They are found mostly in the cell walls of LB. Common hemicellulose sugars are D-glucose, D-mannose, D-galactose, D-xylose, L-arabinose, and to a lesser degree, L-rhamnose. Hemi — cellulose has a low DP, around 100—200, and thus is more easily hydrolyzed into their monomeric sugar components (Glaudemans and Timell, 1958; Goring and Timell, 1960; Koshijima et al., 1965; Timell, 1960).

The structure of a hemicellulose tends to possess a primary backbone, off of which might hang a variety of residual units. These residual units are nonpolymeric acids and sugars. The degree of branching or number of residual units depends on the origin or species of the biomass. For hardwood, the backbone is xylan, contain­ing b-linked bonds at carbons one and four, like cellu­lose. Unlike cellulose, residual units can hang off from the other carbon positions. These residues may include those of acetic acid, glucuronic acid, mannose, arabinose and galactose. Softwood is even more vari­able in that the backbone may be made of more diverse materials. The backbone is typically made of galacto — glucomannan units or arabinoglucuronoxylan units. Galactoglucomannan is a polymer that is a primarily

TABLE 27.1 Major Components of Wood

Lignin

Phenolic —OH

Aliphatic —OH

Methoxyl —OCH3

Carbonyl >C=O

Hemicellulose

Galactoglucomannan

(1:1:3)

(Galacto)glucomannan

(0.1:1:4)

The OH groups in the xylose units were partially substituted by OAc on the C-2 or C-3 positions, i. e. R=CH3CO (Ac) or H

OH in the xylo-units were partially substituted by OAc on the C-2 or C-3 positions (about 7 in 10 xylo-units), i. e. R=CH3CO (Ac) or

5—8 2—4

Terpenes, terpenoids, esters, fatty acids, alcohols, etc.

Phenols: p-cresol, p-ethylphenol, guaiacol, salicyl alcohol, eugenol, vanillin, coniferyl aldehyde, acetovanillone, propioguaiacone, salicylic acid, ferulic acid, syringaldehyde, sinapaldehyde, and syringic acid; stilbenes: pinosylvin, pinosylvin monomethyl and dimethyl ethers, 4-hydroxystilbene, 4-hydroxystilbene monomethyl ether; lignans; hydrolyzable and condensed tannins; flavonoids; isoflavones or isoflavonoids

Arabinose, galactose, glucose, xylose, raffinose, starch, pectic material

Ca, K, Mg, Na, Fe, SO4~, CL, etc.

Cyclitols; tropolones; amino acids, protein, alkaloids, etc.

0.2—0.5 0.2—0.8

Source: Fengel, 1989.

linear and perhaps mildly branched chain. In hemicel — lulose, the residual units take the place of the strong hydrogen bonding that occurred with cellulose components.

Recall that cellulose is highly ordered and tightly bound and thus resistant to hydrolysis. Hemicellulose is not. Hemicellulose tends to be more randomly orga­nized with a more variable and loosely bound structure (amorphous). Therefore, it can be hydrolyzed by weaker or more dilute acids and bases, or at milder conditions.

Lignin is the third largest component of LB at 25—35% of the LB dry weight (Boerjan et al., 2003). Lignin is a het­eropolymer with methoxylated phenylpropylene alcohol units. Its structure tends to be amorphous and variable. These units are interconnected by stable ether and ester linkages. It is hydrophobic and aromatic. It covalently links to hemicelluloses and cross-links different plant polysaccharides giving mechanical strength to the cell wall (Mielenz, 2001). Additionally, lignin is highly resis­tant to biological degradation and thus it protects cellulose and hemicellulose from decay.

Lignin from different plant families vary in their alcohol content and composition. These lignins are thus defined by these components into different types. The lignin precursor in gymnosperms is coniferyl alcohol. The precursor in angiosperms is p-coumaryl alcohol and sinapyl alcohol. The corresponding lignins are guaiacyl (G), p-hydroxyphenyl (H) and syringal

(S), respectively. Grasses tend to contain G while palm trees contain mostly S (Sjostrom, 1993).

The next largest component of LB is the extractives. These make up between 2% and 8% of the total dry weight (Table 27.1). Extractives are compounds found in LB that are soluble in neutral organic solvents or water at standard temperatures and atmospheric condi­tions. Extractives vary in solubility. Some are lipophilic and others are hydrophilic. Lipophilic extractives that are soluble in nonpolar organic solvents are called resins. There is a large diversity in the number of extrac­tives. Additionally, the concentrations of extractives are highly variable throughout the plant depending on the tissue type, i. e. root, stem, bark, branch, needle or leaf. It is important to note that over 70 metal, earth elements, and inorganic compounds may be found in LB. The extractives are the first components that can be extracted from wood. This is advantageous for using LB as a biore­mediation for toxic soil and wastewater in addition to being a source for biofuel and other products.

LB is bulky and much of the volume is "empty" or saturated by air. Air in LB can be replaced by liquid water at high temperature and pressure. When water-saturated LB is suddenly exposed to low pres­sures, liquid water suddenly expands when vaporized forcing LB to disintegrate into fine particles. The Masonite process was invented in 1926 (Mason, 1926) employing this water to steam explosion process. Since then, the steam explosion pretreatment (SEP) has been a common technique (Mason, 1928). SEP is used to break the crystalline and lignocellulosic structure of biomass into its three major components, cellulose, hemicellulose and lignin. SEP enhances the resultant cellulose’s sus­ceptibility to enzymatic hydrolysis. High-pressure, satu­rated steam is applied to biomass for a brief period and then allowed to rapidly decompress to atmospheric pressure, hence the term explosion.

The explosion breaks up solid particles and is used as a standard practice in chemical pulping operations. The steam is vented and the biomass is discharged to a larger vessel for rapid flash cooling (Mosier et al., 2005). SEP is as much a mechanical process as it is a thermal process (Holtzapple et al., 1989). Regardless, the explosion per se, whether it causes particle disintegration or not, does not play a significant role in producing a product that is easily digested by enzymes (Brownell et al., 1986). A more likely mechanism at play is the treat­ment’s effect in removing hemicellulose (Mosier et al.,

2005) . Applying acid catalysts, usually SO2 (or sulfite), enhances this effect by reacting, in conjunction with wa­ter, within the interstitial spaces to form sulfuric acid and thus catalyze hemicellulose degradation (Gregg and Saddler, 1996b).

SEP effectiveness and the chemical changes that take place depend on residence time, temperature, chip size, and moisture content. Effectiveness is determined by the amount of hemicellulose solubilized and the rate of sub­sequent enzymatic hydrolysis. Optimal outcomes are obtained when pretreatment occurs at either high tem­perature or short residence time, such as 270 °C for 1 min, or at lower temperature and longer residence time, such as 190 °C for 10 min. Generally, initial treat­ment pressures range from 0.69 to 4.83 MPa and treat­ment temperature ranges from 160 °C to 260 °C.

At high temperatures water acts as an acid. Thus, dur­ing the treatment time, the hemicellulose hydrolyzes into soluble sugars. The hemicellulose is considered to auto — hydrolyze as a result of exposure to the acetyl groups in the organic acids formed at these high temperatures. Acetic acid is formed from the acetylated hemicelluloses. The pH during SEP is kept quite low, near pH 3—4. SEP degrades a significant portion of the hemicellulose (Sun et al., 2005). However, degradation of hemicellulose may not stop at this point. If the treatment conditions are severe, the solubilized hemicellulose may undergo a series of secondary reactions that yield furfural and HMF. These severe conditions may be high temperature or a long incubation time. Furfural and HMF are

Water with dissolved extracts

FIGURE 27.6 Schematic diagram illustrating three types of liquid hot water reactors: (a) cocurrent, (b) countercurrent, (c) flow through. (For color version of this figure, the reader is referred to the online version of this book.)

undesirable products that inhibit enzymatic hydrolysis and limit the effectiveness of fermentation.

Meanwhile, lignin is partially depolymerized, some lignin is redistributed within the material and some may be removed completely from the fibers, each of which contribute to an improved exposure of the cellu­lose domains (Chen and Qiu, 2010). The reduction in hemicellulose and partial removal of lignin exposes the cellulose surface and thus improves the ability of the enzyme to attack the cellulose microfibrils (Alvira et al., 2010). If the treatment conditions are severe, some degradation of cellulose to glucose can occur. One study reported an enzymatic hydrolysis efficiency of 90% over 24 h using poplar chips using SEP. This was significantly better than the control where the enzy­matic hydrolysis efficiency was 15% from untreated
poplar chips (Grous et al., 1986). That said, SEP is more effective on agricultural residues than in wood as a result of the lower acetic acid content in the hemicellu — lose portion of the biomass.

Adding a supplemental acid to the SEP reduces both residence time and temperature. Adding an acid such as H2SO4 (or SO2) or CO2, typically 0.3—3% (w/w), im­proves hydrolysis, decreases the production of inhibi­tory compounds and leads to a more complete removal of hemicellulose (Kumar et al., 2009a). For the effective treatment of softwoods, adding an acid catalyst is essential to make the substrate susceptible to enzy­matic hydrolysis. Adding a supplemental acid also im­proves the enzymatic hydrolysis of the residual solids and decreases the production of inhibitory compounds (Morjanoff and Gray, 1987).

These three parameters, the level of H2SO4 (or SO2) or CO2, the residence time and the temperature, are the most influential parameters on total sugar yield. For SEP treatment of sugarcane bagasse, the optimal condi­tions are 1% H2SO4, 220 °C; 30 s residence time, and a water-to-solids ratio of 2:1 (Holtzapple et al., 1989). After SEP treatment under these conditions sugar production was determined to be 65.1 g sugar/100 g.

A two-step SEP is a good pretreatment for softwood (Soderstram et al., 2003). In this case, the first step is to optimize the amount of hydrolyzed hemicellulose by employing low severity conditions where the biomass is treated at 180 °C for 10 min with 0.5% H2SO4. In the second step, the solid material from the first step is washed and impregnated again with H2SO4. SEP is applied again using more severe conditions. This time the biomass is treated at 180 °C—220 °C for a longer time, between 2 and 10 min, and with a higher concen­tration of acid catalyst, 1—2% H2SO4. These treatments appear to hydrolyze a portion of the cellulose and make it more accessible to enzymatic attack (Sassner et al., 2008). The most favorable conditions for Salix wood is to impregnate it with 0.5% H2SO4 at 200 °C for between 4 and 8 min. The yield is thus

55.6 g glucose and xylose per 100 g dry biomass (Sassner et al., 2008).

If one uses SO2 as the impregnating agent in spruce chips, the sugar yield is almost independent of impreg­nation time and slightly increases with decreasing chip size (Monavari et al., 2009). Shorter impregnation times result in slightly lower mannose yields in larger chips. The optimum pretreatment conditions when using SO2-catalyzed SEP for lodgepole and Douglas fir pine is 200 °C for 5 min with 4% SO2 (w/w) (Ewanick et al., 2007; Kumar et al., 2010).

Another option for an impregnating agent is to use a weak organic acid, in particular, lactic acid (Monavari et al., 2011). It was observed that it was not efficient and resulted in lower sugar yields in spruce, with or without the addition of SO2. However, using a weak organic acid is more environmentally friendly than us­ing an inorganic acid as it would biodegrade in a waste stream or be used for production of a biogas such as methane.

Particle size, by itself, is not a significant contributor to SEP effectiveness. Some studies report that larger par­ticle size may improve the outcomes from SEP (Cullis et al., 2004). In these studies, pretreated Douglas fir, a softwood, was milled to three particle sizes: <0.422 mm screenings, 1.5 x 1.5 cm and 5 x 5 cm. They were then steam exploded using SO2. It was observed that the largest particle size suffered less from pretreat­ment severity and had the highest cellulose recovery. It had larger quantities of solubilized carbohydrate and contained fewer furan degradation products. The smaller particle sizes produced outcomes containing more solubilized hemicellulose and lignin. If the resul­tant biomass is further refined to particles of a finer size by plate milling, with a 0.178 mm gap, the initially larger particle size showed a higher lignin removal with peroxide washing and a greater rate of enzymatic hydrolysis. This is likely due to the reduction in lignin redeposition as a result of treatment severity.

These findings were substantiated in studies of steam-exploded pine, another softwood (Ballesteros et al., 2000). The largest of the sizes exhibited a higher cellulose recovery and also a higher content of solubi­lized hemicellulose. Conversion of cellulose to glucose was only slightly higher from the larger particle sizes. However, the total recovered glucose, including the sol­ubilized glucose from the steam explosion, was much higher when starting with the largest particle size (Ballesteros et al., 2000).

Starting with a herbaceous feedstock, such as Brassica carinata residues, produced different results. Although the cellulose recovery was still higher for the 8 and 12 mm fractions, the smaller particle sizes performed better during enzyme hydrolysis. The 5—8 mm and 2—5 mm fractions yielded 100% while the 8—12 mm fraction produced 85% (Ballesteros et al., 2002). This suggests that lignin condensation is not as influential in herbaceous feedstock. It is a critical factor when pretreating softwood.

In softwood, the larger particle size produces a higher maximum glucose yield, over 80%, compared to smaller particle sizes with yields under 70% (Ballesteros et al.,

2002) . SEP-treated hardwood exhibited no difference in either enzyme digestibility or ethanol yield between disparate particle sizes, in particular between 2—5 mm and 12—15 mm (Negro et al., 2003). This indicates that the severity of the treatment plays a larger role than par­ticle size when using softwood. In these cases, smaller particles increase the lignin condensation and recalci­trance to enzyme hydrolysis.

Overall, SEP is an attractive pretreatment method considering the low amount of energy required to reduce the biomass size compared to mechanical comminution. Conventional mechanical size reduction methods require 70% more energy than SEP to achieve the same particle size reduction (Mittal et al., 2009). Additionally, SEP is attractive because there are no recy­cling or environmental concerns. This too lowers cost. SEP is thus recognized as one of the most cost-effective pretreatment methods for hardwoods and agricultural wastes. It has been extensively tested on a wide array of lignocellulosic feedstock. It has been observed that SEP is less effective for the pretreatment of softwoods.

The most significant limitations include the partial destruction of xylan, incomplete disruption of the lignin-carbohydrate matrix, low lignin removal, and lignin redistributes over the surface of cellulose (Chen et al., 2010). Additionally, there is a risk of producing un­desirable compounds such as furfural, HMF and other soluble phenolic compounds. These undesirables inhibit microbial growth and enzymatic hydrolysis. Thus, prior to fermentation, SEP-treated LB must be washed with water to remove these undesirable materials along with water-soluble hemicellulose. Unfortunately, this wash lowers the overall effectiveness as it washes away around 20—25% of the initial dry matter and a portion of the soluble sugars (Sun and Cheng, 2002).

Mark P. McHenry

School of Engineering and Information Technology, Murdoch University, Perth, Western Australia, Australia

email: mpmchenry@gmail. com

OUTLINE

Introduction 447

Theoretical Income Streams 448

Renewable Energy and Fuel Generation 448

Carbon Sequestration of Biochars and Carbon Markets 449

Agricultural Benefits 450

Economic Analysis 451

Can Biochar Be a Cost-effective Fertilizer Substitute? 451 Can Biochar Be a Cost-Effective Approach to Increase Grain Crop Primary Productivity? 452

Can Biochars Increase Livestock Growth Rates, or Provide a New Market for Semiarid Forestry? 453

A Comparison of Biochar Carbon Value for Different Potential Income Streams 454

Conclusion 454

Disclaimers 455

References 455

INTRODUCTION

Rural biomass energy and carbon options seem to offer increased financial resilience to agricultural enter­prises relative to fluctuating seasonal growing condi­tions and uncertain market prices of inputs, products, and exchange rates. The projected increases in farming costs from any future inclusion of the agricultural sector from carbon pricing may be offset by additional net in­come from such rural biomass-based sequestration and renewable energy activities. Cellulose, hemicelluloses, and lignin are the main components of wood and crop residues of known potential for bioenergy and stable carbon forms, and the management of which requires detailed agronomic, technical, and market information. Thus, there is a synergistic match between growing
food and growing biomass for energy and carbon in the same rural enterprise.

Modern concepts of biochar-agricultural systems and their respective projected financial viabilities have been outlined in the existing literature (Lehmann and Joseph,

2009) . These systems commonly incorporate complex semi-industrial operations with rural and forestry biomass as well as small-scale low-technology concepts with farm waste and domestic heating. To narrow research specificity, this work focused on the West Midlands of the Northern Agricultural Region of Western Australia (WA), and uses Australian dollars. (At the time of writing the Australian and US currencies were roughly parity.) To date, this region is one of the few regions of Australia that has exhibited economically encouraging agricultural responses from biochar addition, and has an established

Bioenergy Research: Advances and Applications

http://dx. doi. org/10.1016/B978-0-444-59561-4.00026-7

448

practice of profitable grazing leguminous fodder shrubs, which is a potentially large and sustainable biomass supply.

Conversion refers to the collection of processes employed to modify a feedstock into desired product(s).

Given that LB is composed of a number of distinct components, there are a variety of treatment options available that one can use to change these components into fuel, chemicals and other products.

With a harsh condition (high temperature, strong acid/base, strong solvents, or a combination of these agents), LB can be turned into small molecular units (such as C, CO2, CO, H2 and H2O) and then further con­verted to a desired product. Thermochemical conver­sion technologies usually employ this strategy to break down LB unselectively to accommodate further conver­sions either catalytically or biologically. Therefore, ther­mochemical conversion technologies can be versatile.

The structure of wood (or LB in general) is sufficiently strong and complex that it is not feasible to attack the whole complex at mild conditions in a single step, nor is it feasible to isolate the components and attack them individually. When mild conditions are desired, one must attack at least one portion of the whole structure and weaken it. Follow up with another treatment to break down the first component or attack a second component. Continue to treat the biomass until the desired composition is obtained. These treatment op­tions are classified into mechanical, thermal, chemical, or biological processes. These are not discrete classifica­tions. In other words, a process can be considered to belong to more than one category. For example, if one were to saturate LB with water, then heat it under high pressure and rapidly release the pressure, the hot water could vaporize into steam and thus explode apart the woody cells it had penetrated. This is called steam explosion and it uses a thermal process to accomplish a mechanical breakdown of the woody material. Steam explosion will be discussed later in this chapter.

While there are several LB conversion technologies available, this chapter will focus on biochemical conver­sion technologies with some discussions in thermo­chemical conversions.

These treatments may be applied at various points across the process. The typical process to acquire fuel products using a bioconversion methodology is gener­ally described in four parts: pretreatment, hydrolysis, fermentation and distillation, separation and filtration.

As discussed previously, cellulose, hemicellulose and lignin are strong, stable structures. These structures are challenging for one to convert into fermentable compo­nents (Mielenz, 2001). Of these three components, hemi — cellulose is the most vulnerable and easiest to degrade. Recall that compared to cellulose, hemicellulose is a lower molecular weight and is less uniform as it is composed of a variety of sugar polymers and residual units.

In bioconversions, the objective of pretreatment is to, as efficiently as possible, prepare LB for fermentation into products. The amount of energy required to break

Thermal-

mechanical

pretreatment

Gasification

Syngas

>“Coke

FIGURE 27.3 Schemes of biochemical conversion to materials, chemicals and fuels: (1) sequential incremental deconstruction; (2) two-step saccharification and fermentation; (3) simultaneous saccharification and fermentation; (4) gasification and fermentation. Source: Wang et al., 2012. (For color version of this figure, the reader is referred to the online version of this book.)

down LB is fixed. No matter what suite of treatment options used to convert LB to product, the thermody­namic barrier is the same. It requires the same amount of energy to completely biologically degrade wood as it does to chemically treat it or gasify it. One trades off time for allowing organisms to invest energy on one’s behalf versus applying heat or concentrated chemicals to accomplish the same task more quickly. Additionally, biological methods allow for greater selection of the portion of biomass to convert it into product, and usu­ally by selecting a descending route of molecular chem­ical energy or intermediates that are not down to simplistic building blocks if possible (green chemistry). By selecting only a portion of the LB to convert, one lowers the amount of investment energy required. Biochemical processes operate at moderate or low tem­peratures. These milder conditions may be slightly more efficient than their thermochemical counterparts. However, a burden of biological or biochemical pro­cesses could arrive for the need of detoxification. One must often remove toxic components resulting from the pretreatment methods employed.

Figure 27.3 illustrates a set of four treatment path­ways to convert LB into various products. These are not the only methods available but merely an example of commonly used methods. This pathway represents one of the most popular biorefinery designs used to bio­chemically convert LB into biofuels and bioproducts.

Pathway 4 shows a gasification process to produce syn­gas. This is a thermochemical process. The sugars in syn­gas are subsequently fermented into liquid fuels similar to those produced by the more biochemical methods. The four pathways shown vary in the number of steps, or time, required to acquire the product.

Figure 27.4 provides a slightly more detailed look at pathway 1 from Figure 27.3. Different pretreatment methods have different desired characteristics (Limayem and Ricke, 2012). A summary of pretreatment methods to be discussed in this chapter and their characteristics is shown in Table 27.2.

Another explosion pretreatment is the ammonia fiber explosion (AFEX) process. Instead of using liquid water under high pressure, liquid ammonia is used. AFEX is an effective and somewhat economically attractive method to increase the yields of fermentable sugars from LB (Holtzapple et al., 1991; Holtzapple et al., 1992). In this method LB is exposed to liquid ammonia, not ammonium hydroxide (i. e. no water/moisture), at moderate temperatures and elevated pressures for a longer period of time. After the appropriate residence time, the system is rapidly vented allowing the liquid to vaporize and literally explode the fibrous material. Typically, 1—2 kg of liquid ammonia is used for each kg of dry biomass. The system operates at temperatures below 100 °C, pressures above 3 MPa, and is quite tolerant of pH. Any pH under 12 appears suitable. The residence time is between 10 and 60 min. Under these conditions, the system forms few degraded sugar prod­ucts yet gives a high yield of desirable sugar products (Mosier et al., 2005).

AFEX is an attractive treatment method for a variety of herbaceous crops and grasses as it significantly im­proves the saccharification rates. It has been tested on a variety of LB including aspen chips, softwood and kenaf newspaper, alfalfa, wheat chaff, wheat straw, barley straw, rice straw, bagasse, coastal Bermuda grass, switchgrass, corn stover, and municipal solid waste. One of the benefits is that AFEX only solubilizes a trivial amount of solid material. Also, compared to acid pre­treatment and acid-catalyzed steam explosion, very little hemicellulose or lignin is removed. Lastly, the structure of the material changes such that the result is an increase in water-holding capacity and improved digestibility. Although physically modified, the chemical composi­tion of the material following AFEX pretreatment is essentially unchanged from its original condition. The benefit is illustrated as follows: over 90% hydrolysis of the cellulose and hemicellulose may be obtained after AFEX pretreatment of Bermuda grass where 5% of that is lignin. The result is similar for bagasse except 15% of the hydrolysate is lignin (Holtzapple et al., 1992). These low-lignin containing biomasses readily hydro­lyze at near theoretical yields of sugars. The resulting sugars ferment rapidly with a high yield into a variety of desired products. Since the AFEX treatment produces very few inhibitors to the downstream biological pro­cesses, a water wash is not necessary (Dale et al., 1984; Mes-Hartree et al., 1988).

Materials with a high lignin content, around 25%, have proved to be recalcitrant to AFEX. Therefore, AFEX is a less effective pretreatment method for hard­wood chips, some newspaper material, and nut shells (Teymouri et al., 2005). AFEX does not require a small particle size for it to be an effective treatment option (Larson and King, 1986) like steam explosion and hot water treatments.

The most significant cost is that associated with recy­cling the ammonia following pretreatment (Kumar et al., 2009a). Since pure ammonia is used in the process, more stringent environmental and recovery procedures are required. Thus, recycling is necessary to reduce the envi­ronmental impact and the cost of the procedure. To recover the ammonia, a superheated ammonia vapor, at temperatures upward of 200 °C, is used to vaporize and strip the residual ammonia from the pretreated biomass. The evaporated ammonia is then drawn off the system by a pressure controller for final recovery (Holtzapple et al., 1990). Using this recovery method has demonstrated that over 99% of the ammonia can be recycled successfully. Even so, the overall capital and operating costs are higher than other comparable methods.

The potential income streams in the West Midlands from above ground rural biomass include renewable energy and fuel generation, carbon sequestration of bio­chars, and agricultural benefits from the use of biochar and ash from energy and fuel generation or charring alone. It is likely that tradable sequestered carbon will be reliant on the supplies from bioenergy generation plants that are able to comply with both emission and biochar quality standards. However, a price on carbon may help offset the additional costs of the coproduction of electricity and biochar from biomass. (Table 26.1 outlines key benefits, costs and barriers to biochar compliance to carbon markets.) The Australian Farm Institute (2011) estimates an income reduction of be­tween 1.4% and 1.6% from a carbon price based on elec­tricity consumption for a WA mixed farming enterprise of 4900 ha (2400 ha cropped and about 2000 head of live­stock, mainly sheep), assuming agriculture and trans­port fuels are excluded from any carbon liability (Australian Farm Institute, 2011). In contrast to concerns of a carbon price reducing agricultural profitability, this work presents the case that integrating new sequestra­tion options into conventional production systems from low-cost biochars produced from agricultural wastes (with sufficient operational safety consider­ations) may offset costs in the West Midlands. The prof­itability of income streams (presented in Table 26.1) are highly sensitive to (and often dependent upon) govern­ment subsidies for renewable energy, a carbon price, and also the location-specific demand for biochar and energy. Similarly, agricultural effects of biochar addition will vary more with soil type, seasonal conditions, and animal nutrition characteristics. In complex and uncer­tain circumstances, predictive modeling can become particularly challenging. However, the agricultural effects (where they occur) will likely provide a more solid basis for emerging industry development than the highly sensitive and evolving carbon and electricity markets. If agricultural benefits, initially at least, exhibit less risky investments to individual farms than bio­energy cooperatives or carbon sequestration pooling activities, then agricultural benefits may be a more suit­able foundation for the establishment of biomass-based industrial developments than energy or carbon seques­tration policies in the initial phases.

The first and most important step in any conversion process is to reduce the physical size of LB. In order to obtain the high yields required for commercial success in bioconversion operations, it is vital to pretreat and reduce the biomass into an effective size (Mosier et al.,

2005) . Reducing the LB size from a log to wood chips to even fine powders improves mass and heat transfer as well as increases the surface area of the particle. Increasing the surface area exposes a higher percentage of the glycosidic or ester bonds to the agents in solution

FIGURE 27.4 Schematic flow sheet for biomass conversion to bioproducts. Source: US DOE, 2006. (For color version of this figure, the reader is referred to the online version of this book.)

(Mosier et al., 2005). Catalysts, such as a proton or an enzyme, can only access active chemical bonds when exposed at the solid—liquid interface (Liu, 2003; Yang and Liu, 2005). Smaller particles translate into faster, more uniform reactions and a more complete conversion.

The energy required to reduce the biomass into a treatable size depends on the density of the biomass source. Herbaceous materials do not require as much processing to achieve the needed particle size as it does to reduce wood (Cadoche and Lopez, 1989). Since LB reduction is much more energy intensive, it is imper­ative to adequately define the reduction process. This re­quires an understanding of the quality and condition of the source materials. Qualities such as moisture content, soil particles, foreign matter, and initial cut length will impact efficiency, energy requirements and downstream treatment conditions and requirements. Pretreatment is costly and greatly influences the cost and effectiveness of downstream operations. It affects fermentation toxicity, the rate of enzymatic hydrolysis, enzyme load, powder mix, product concentration, product purity, waste treatment requirements, energy requirements and a host of other process variables (Zhu and Pan, 2010). Thus, it is important to begin the design process with the end in mind. Much effort should be invested to design the whole process up front with specific source materials and conditions defined.

To address the expense of the AFEX method, the su­percritical carbon dioxide explosion method was devel­oped. Compared to steam explosion, the supercritical CO2 explosion method produces fewer inhibitory compounds. Additionally, CO2 is much more environ­mental friendly than organic solvents used in the orga — nosolv method and the ammonia used in the AFEX method. Because carbon dioxide is nontoxic, physiolog­ically safe and inexpensive it is used in a variety of in­dustries, for example, in food and pharmaceutical production. The critical temperature of CO2 is 31.1 °C and its critical pressure is 73 atm. The term supercritical refers to a fluid that at standard temperature and pres­sure would exist in its gaseous state. However, when compressed using high pressures and at temperatures above the critical point, the gas condenses into a liquid­like density. In this state it retains the characteristics of mass transfer that are "gaslike" but with the solvating power that is "liquidlike" (Kim and Hong, 2001). Carbon dioxide molecules are small, like water and ammonia, and thus it penetrates the small pores of LB. It is believed that CO2 forms carbonic acid and thus it should increase the hydrolysis rate. Furthermore, at low temperatures it is thought to prevent significant decomposition of the monosaccharides by the weak acid. However, the pri­mary effect of supercritical carbon dioxide explosion is from the explosion whereby it disrupts the biomass structure and increases the surface area and improves its vulnerability to enzymatic attack (Conner and Lorenz, 1986; Zheng et al., 1998).

Despite these advantages, the operating and capital costs of the supercritical carbon dioxide explosion pre­treatment option remain prohibitive.

In simple terms, wood-fired stoves, barbeques, or wa­ter heaters are a biomass-based renewable energy system. Yet, the growing range of new medium- to large-scale bioenergy technologies include gasifier and pyrolysis power stations coproducing electricity, heat, and a range of biofuels. Nonetheless, all traditional and new technol­ogies convert the complex hydrocarbon molecules in biomass to hydrogen, methane, carbon monoxide, carbon dioxide, and numerous other gasses, including polyaro­matic hydrocarbons and dioxins. Some technologies also produce liquid and soil fuels (such as biochar) from the same biomass. In general, while small-scale and simple technology designs have less control and efficiency, they exhibit lower capital and operating costs, although they are usually more labor intensive per unit production of output (McHenry, 2012b). At the regional scale biomass power plant technology choices often

include gasifiers (which optimize gas production), and slow pyrolysers (which optimize biochar production). A general outline of the variations in biomass renewable energy technologies are shown in Table 26.2. In terms of developing a regional energy/biochar industry, medium-sized biochar production units may address concerns of soil nutrient loss from harvested biomass. Despite the generally high costs of transporting timber trees, transporting returned biochar is relatively efficient on a weight basis, as the biochar mass is 70—80% less than the original dry biomass (Lehmann, 2007). Nonetheless, industrial biochar production and use will require a number of safeguards. Handling risks include flamma­bility concerns, and the dusts can spontaneously combust in enclosed spaces and is comparable to the risk of handling some metals, foods (flour, etc.), coal, plastics, and woods (Joseph, 2007).

At the time of harvest, an operation is performed in the field to presize the LB. Herbaceous biomass is pre­pared by shredding or forage cutting. Chipping is the
preferred method for reducing the size of wood. Chip­ping reduces wood to 10—50 mm in two dimensions and 5—15 mm in the third (Zhu and Pan, 2010). This is the minimum treatment necessary to begin conversion. However, additional reduction is often performed. For example, wood chips may subsequently be refined to fi­bers such as that in fiber production, pulverized into wood fibers or wood flour (Zhu and Pan, 2010). Pulver­ization requires much more energy than chipping (Zhu and Pan, 2010).

In addition to chipping and shredding, hammer mill­ing, knife milling, disk or attrition milling, and ball mill­ing are viable alternatives to reduce biomass sizes. Large-scale reduction operations have favored hammer and disk milling (Tienvieri et al., 1999). Chip refining is also an alternative as it can have a large throughput.

Hammer milling is primarily used for making wood flours for composites and pellets. Disk milling is used for wood fiber production at a commercial scale, around 1000 tons per day. Disk milling operations are depen­dent on environmental conditions and the quality of source materials. The energy requirement and the wood particle size and shape depend on these opera­tional parameters (Tienvieri et al., 1999).

Milling operations have a significant impact on downstream energy requirements and the efficiency of enzymatic cellulose saccharification. Since the goal of a biorefinery is to optimize the conversion process, to reduce energy requirement and maximize the enzymatic cellulose saccharification, it is important to attend to the biomass size reduction portion of the process. Failure at this stage amplifies the cost of energy requirements and reduces the effectiveness of subsequent treatments.

Since these mechanical processes can produce a range of particle sizes it is often necessary to control the

particle size used in the biorefinery. Size characterization is accomplished using sieves, screens and imaging analysis. The particle surface area is the most relevant determination of effectiveness, and thus, it is the quality to be controlled. Specific surface area correlates to energy consumption and the efficiencies of a variety of size reduction processes have been compared (Holtzapple et al., 1989).

There is a limit to the effectiveness of size reduction. At this point, additional surface area increases, or parti­cle size reductions will not improve substrate enzymatic digestibility. This critical size is proportional to the pore size in and along the wood cells. Refer to Figures 27.1 and 27.2. A common target size is one that maintains the cell structure while allowing for lignin removal from the middle lamellae.

Size reduction below the cell size will provide a more efficient conversion. To reduce particle size to this smaller level is done by comminution (Vidal et al.,

2011) . Comminution of biomass, especially at the final sizing stage, is energy intensive and the product is of low value. Thus, there is much interest in finding the most efficient milling processes.

To that end, ball milling has been extensively studied. It has been shown to deliver excellent results in terms of the hydrolysis rate and sugar yield. Additionally, this pretreatment method is clean and easy to do. Vibratory ball milling has been shown to be more effective at breaking down the crystallinity of cellulose and improving the digestibility of the biomass over ball mill­ing alone (Millet et al., 1976). Mechanical milling requires long operation times and a large amount of energy (Lynd et al., 1996). The smaller the desired particle size the greater the comminution requirements will be in terms of time and energy (Cadoche and Lopez, 1989).