The first detailed economic costing of a lignocellulosic ethanol process envisaged a process similar in outline to that eventually adopted by Iogen in Canada, where wheat straw was first pretreated with acid before fungal cellulase was used to digest the cellulose for a Saccharomyces fermentation of the liberated glucose (no pentose sugars were included as substrates at that date). Results from laboratory studies were extrapolated to a 25-million-gallon/year facility that was designed to be stand­alone and capable of generating the required fungal cellulase on site (table 5.2).

The calculated factory gate plant was much higher than that estimated for corn — derived ethanol, that is, $3.34/gallon — this cost included raw materials, utilities,

maintenance materials and labor, operating labor and supplies, the facility labora­tory, plant overhead, taxes, insurance, and depreciation but made no estimates for general and administrative, sales or research costs, profit, or any by-product credits or disposal charges. Of this total, 61% was attributed to materials, of which the wheat straw feedstock accounted for 11-12%. The other components of the price estimate and various options for reducing the total were assessed:

• Utilities and capital costs were each 14% of the total.

• Labor costs were 11% of the total.

• A high-cost peptone (proteased protein) nitrogen source was included in the fermentation media; adopting a lower-price product could reduce the cost of the product by 9%.

• The conversion of cellulose to soluble sugars was 45% (w/w basis, calcu­lated on the straw weight); increasing this to 60% could effect a 6% reduc­tion in product cost.

• Reducing the enzyme loading (ratio of enzyme to cellulose) by fourfold could reduce the product price to approximately $2/gallon.

No allowance is made in table 5.2 for capital costs or profit; on the basis of 15%/year of fixed capital for general and administrative, sales, and research costs, a 15-year life expectancy of the plant, and a 48% tax rate, the selling price for the 95% ethanol product would have to have been $4.90 or $4.50/gallon for a discounted cash flow return on investment (after taxes) or 15% or 10%, respectively.12

An independent estimate for ethanol production from corn stover, funded by the DOE for work carried out at the University of Berkeley, California, was first announced at a symposium in 1978.13 Scaling up an advanced process option with cell recycling (chapter 4, section 4.4.1) and converting the postdistillation stillage into methane by anaerobic fermentation to generate a combustible source of steam generation, a total production cost of $3.38/gallon was estimated, again with no allowance for plant profitability at a 14-million-gallon/year scale of production.

The main conclusion was, therefore, that the costs of generating fermentable sugars by enzymic hydrolysis dominated production (i. e., not-for-profit) economics; consequently, any major reduction in the cost of generating the cellulase or increase in its specific activity (or stability or ease of recovery for recycling) would be highly effective in lowering the production costs. Although federal and state initiatives in operation by 1979 offered some mitigation of the high costs of biomass-derived ethanol, the sums were small: for example, under the National Energy Act of 1978, alcohol fuels were eligible for DOE entitlements worth 50/gallon, and 16 states had reduced or eliminated entirely state gasoline taxes on gasohol mixtures, the largest amount being that in Arkansas, worth 9.50/gallon.

The fixation of gaseous molecular nitrogen, N2, to biologically utilizable nitrogen is performed as an essential part of the global nitrogen cycle by bacteria that may be either free living or exist in symbiotic associations with plants, bivalves, or marine diatoms.21 The central enzymic reduction of nitrogen to ammonia is that catalyzed by nitrogenase:

N2 + 8H+ + 8e — ^ 2NH3 + H2

Molecular hydrogen is an obligatory product of the overall reaction — which, how­ever, is so energy-expensive that most nitrogen fixers rarely evolve H2, employing another hydrogenase (“uptake hydrogenase”) to recycle the H2 via oxidation by O2 and conservation of part of the potentially released energy to support nitrogenase
action. For this reason, nitrogenase-containing microbes are not viewed as likely sources of H2 for biofuels.

5.5.1 Definitions and Semantics

Much of the public and scientific debate about biofuels has assumed that the pro­duction of bioethanol as the lead biofuel is inherently “sustainable,” mostly on the grounds that any agricultural activity is renewable, whereas the extraction of crude oil is necessarily a once-only activity, given the extremely long geological time scale of oil’s generation. As a term, “sustainability” suffers from inexactness.7879 As Patzek and Pimentel79 point out, an exact definition can be deduced from thermodynamics and is consequently defined in terms of mathematical and physical properties, that is, a cyclic process is sustainable if (and only if):

1. It is capable of being maintained indefinitely without interruption, weaken­ing, or loss of quality

2. The environment on which this process depends and into which the process expels any “waste” material is itself equally renewable and maintainable

These are very strict criteria and are not exemplified by, for example, an annual replanting of a crop plant such as maize, which depends on outlays of fossil fuel energy (for fertilizers, etc.) and which may seriously deplete the soil or minerals and contribute to soil erosion — although such a system appears to be renewed every year, that is “only” within living memory or the history of agricultural production on Earth, on a geological time scale a minuscule length of time. It is a sad fact of human agricultural activity that periodic crises have accompanied large-scale, organized farming for millennia in (among many other examples) the erosion and salt accumulation that caused the downfall of the Mesopotamian civilization, the overgrazing and poor cultivation practices that have caused the expansion of the Sahara, and the overintensive cultivation of fragile tropical rainforest soils that con­tributed to the collapse of the Mayan economy and society.80 Industrial agriculture and intensive farming are relatively recent arrivals, within the last century, and they have brought accelerated land degradation and soil erosion; agrochemical residues in water courses are a global problem, leading to eutrophication of freshwater fishing grounds and threatening the collapse of fragile ecosystems. To environmental lobby­ists, therefore, the prospect of energy crop plantations is highly unwelcome if such agronomy requires and is dependent on the heavy use of fertilizers, insecticides, and pesticides, the depletion of soil organic carbon, and (worst of all) the sequestration of limited farmland.

A more analytically useful definition of the renewability of “renewable” energy could focus on a more analyzable time scale, perhaps 160 years (approximately, the life span of the industrial activity that underpins modern society).79 Perhaps, even more appropriate would be a century, that is, most of the time in which automobiles have been fuelled by gasoline and diesel. For either option, generating a quanti­tative framework is crucial to any assessment of the practicalities of a biofuels program.

Glycerol is an obligatory coproduct of biodiesel production, glycerol forming 10% by weight of the triglycerides that act as substrates for chemical or enzymic transesteri­fication (figure 6.1). The waste stream is, however, highly impure, with the glycerol mixed with alkali, unreacted methanol, free fatty acids, and chemical degradation products. The glycerol can, at the cost of considerable expenditure of energy in gas sparging and flash distillation, be recovered at more than 99.5% purity.74 Much sim­pler is to mix the crude glycerol with methanol and substitute waste petroleum oil and heavy fuel oil as a direct fuel.75

Glycerol represents a valuable chemical resource as a potential feedstock.15 Research into value-added utilization options for biodiesel-derived glycerol has, therefore, ranged widely in the search for commercial applications using both chem­ical and biotechnological methods (table 8.3).76 For the production of 1,3-PD for polymer manufacture, glycerol as source via microbial fermentation was (5-6 years ago) more expensive than either glucose or chemical routes from ethylene oxide or acolein.3 Those economics have significantly altered now, and glycerol represents the shortest, most direct route for bioproduction of 1,3-PD, a two-reaction sequence comprising an enzyme-catalyzed dehydration followed by a reduction:

glycerol ^ 3-hydroxypropionaldehyde ^ 1,3-PD

Microbial studies have focused on the fermentative production of 1,3-propandiol by clostridial species, but this is greatly complicated by the multiplicity of other products, including и-butanol, ethanol, and acids.77 Bioproduction of a nutraceutical fatty acid derivative (marketed as a dietary supplement) by microalgae is a route to a higher-value product than bulk chemicals.78 As examples of chemical engineer­ing, recent industrial patent documents have disclosed methods for converting the biodiesel waste glycerol into dichloropropanol and an alcohol-permeable membrane to generate a purified mixture of esters, glycerol, and unreacted alcohol.7980

For many years, glycerol was not considered fermentable by E. coli but only by a limited number of related bacterial species; a landmark publication in 2006, how­ever, reported that E. coli could efficiently ferment glycerol to ethanol (and a small amount of succinic acid) provided high pH in the culture was avoided — this is a cru­cial point because growth from glycerol requires an anaplerotic step (section 8.3.1), the CO2 being generated by the pyruvate formate lyase reaction whose activity may be much reduced by high pH in the growth medium (figure 8.6).81 The application of genetically engineered E. coli with superior ethanologenic potential (chapter 3,

Chemical and Biotechnological Transformations of Biodiesel-Derived Glycerol

Fermentation route

Clostridium butyricum, Klebsiella pneumoniae None

Gluconobacter oxydans Anaerobiospirillum succiniciproducens Enterobacter aerogenes None None

Various osmophilic microbial species

section 3.3.2.1), combined with further manipulations to eliminate competing path­ways of acid accumulation, could result in ethanol productivity approaching the theoretical molar production (mole per mole) from glycerol.

After herbicide resistance in major crop species, the first target area for GM crop development was that of protection against plant pests.283 Although tolerance of modern herbicides is usually located in the amino acid sequences of a handful of target genes in biosynthetic pathways, plants have multiple inducible mechanisms to fight back against microbial pathogens:284 [44]

Wheat is prone to attack by the rust pathogens, Puccinia graminis and P. tritici; stem and leaf rusts are considered to be major constraints to wheat production world­wide.285 More immediately alarming from the perspective of Iogen’s dependence on wheat biomass is that new and highly infectious variants of the pathogen have been noted in Africa for some years, with newspaper reports in the first quarter of 2007 describing its spread into Asia. A single gene (Sr2) has been identified as a broad — spectrum resistance locus for more than 80 years; recently, this gene (or two tightly linked genes) confer resistance and the associated dark pigmentation traits, pseudo­black chaff.286 Stem rust-susceptible barley has been transformed into a resistant form by an Agrobacterium plasmid containing the barley resistance gene, RpgP; a single copy of the gene is sufficient to confer resistance against stem rust.287

Of the abiotic stresses that plants experience, drought is a serious limiting factor on growth and productivity even in the Northern Hemisphere, and cereal crops are highly prone to fluctuating yield depending on seasonal rainfall and average tempera­tures in the growing season; to be dependent on a monoculture crop such as wheat (for starch or straw) runs the risk of uncertain prices as well as variable feedstock availability. With the evolution of domesticated cereal species over millennia, genetic diversity has been lost; using the natural genetic diversity of wild species is an invalu­able resource because wild types harbor very broad ranges of tolerance characteris­tics; other exploitable traits include those for salt tolerance (allowing saline water to be the source of irrigation) and, especially if global temperatures increase because of global warming, flowering times and other growth parameters more typical of Medi­terranean regions for transfer to cultivars grown further north.288 Drought appears to exert physiological effects via oxidative stress signaling pathways, a property shared with freezing, heat, and salinity stresses; protein kinases are often associated with signaling pathways and expression of a protein kinase gene (NPK1) from tobacco in maize protects kernel weights when the water supply is reduced.289

6.3.1 Methanol: Thermochemical and Biological Routes

The versatility of the FT process is that almost any hydrocarbon produced that can be derived from petroleum can be made from syngas, not only alkanes and alkenes but oxygenated compounds; the exact mixture of products obtained can be varied by choices of catalyst, pressure, and temperature, and straight-chain alcohols are pro­duced in the “synthol” reaction at 400-450°C and 14-MPa pressure in the presence of an iron catalyst.94 Industrial production from natural gas has, however, been dom­inated since the 1960s by a lower temperature and pressure process invented by Imperial Chemical Industries in which CO, CO2, and H2, derived by steam reform­ing, are reacted over a mixed Cu/ZnO/Al2O3 catalyst at 250°C and 50-10 MPa when two reactions occur:95

CO + 2H2 ^ CH3OH
CO2 + 3H2 ^ CH3OH + H2O

A recent development has been to combine syngas production from methane with the reduction of ZnO to metallic zinc in a metallurgical plant; the syngas has a H2:CO ratio of approximately 2:1, highly suitable for methanol production.96 A renewable-resource route for methanol (one of the largest bulk chemicals in the contemporary world) is, however, entirely feasible via biomass gasification as an intermediate step, and this would be entirely appropriate given methanol’s older name of “wood alcohol,” indicative of its historical provenance by incomplete combustion.

As an “energy carrier,” methanol is inferior to ethanol, with an energy content of only 75% (on either a weight or a volume basis) compared with ethanol and approximately 50% compared with conventional gasoline.97 Blends of methanol with conventional gasoline up to 20% can be tolerated without the need for engine modifications, that is, as a fuel extender; the corrosive effect of methanol on some engine materials limits the extent of this substitution.98 Methanol would have been an excellent replacement for MTBE as a gasoline oxygenate additive (chapter 1, section 1.4), but its acute neurotoxicity is well known and a barrier to several poten­tial uses. One notable exception, however, is as an at-site (or on-board) source of hydrogen for fuel cells (chapter 7, section 7.1): between 1983 and 2000, nearly 50 patents were granted for catalytic methanol “reforming” systems to automobile producers (such as General Motors, Daimler-Benz, DaimlerChrysler, and Honda), chemical multinationals (DuPont, BASF, etc.), and major oil companies (CON­OCO, Standard Oil Company, etc.).95 Combined reforming with liquid water and gaseous oxygen has been intensively investigated for use in mobile applications for transportation:

(s+p)CH3OH (l) + sH2O (l) + 0.5pO2 ^ (s+p)CO2 + (3s+2p)H2

since the composition of the reactant feed can be varied and the process carried out under a wide range of operating conditions.[61]

The first pilot plant for testing and evaluating the production process for “bio­methanol” was established in a program that commenced in 2000 between the Minis­try of Agriculture, Forestry, and Fisheries of Japan and Mitsubishi Heavy Industries at Nagasaki (Japan); various feedstocks have been investigated, including wood, rice husks, rice bran, and rice straw.99 The test plant consisted of

• A drier and grinder for the biomass input (crushed waste wood)

• A syngas generator

• A gas purifier

• A methanol synthesis vessel (with an unspecified catalyst)

The pilot plant was designed for a capacity of 240 kg/day, with a methanol yield (weight of methanol produced per unit dry weight of material) of 9-13%. A larger plant (100-tonne daily capacity) is predicted to have a much higher methanol yield (38-50%).

No economic analysis of the Japanese pilot facility has been published, but a the­oretical study of methanol production via the syngas route suggested that methanol from biomass (by 2002) had production costs approximately twice those of conven­tional gasoline on an equal energy basis.100 The surge in crude oil and gasoline refin­ery gate costs subsequently (figure 5.1) implies that methanol production costs from biomass sources would now (mid-2008) be competitive. This encouraging result is very timely because direct methanol fuel cell (DMFC) technology has reached the stage where Toshiba in Japan has announced the development of a micro-DMFC suitable for powering MP3 players.101 A U. S. patent covering aspects of DMFC con­struction was also issued in March 2007 to Creare (Hanover, New Hampshire).102 The Japanese invention utilizes a polymer electrolyte membrane device with the electrochemical reactions:

CH3OH + H2O ^ CO2 + 6H+ + 6e — (anode)

1.5O2 + 6H+ + 6e — ^ 3H2O (cathode)

The inputs are concentrated methanol and air (O2); the only outputs are water and CO2 and electricity (100 mW) sufficient to power a portable device for 20 hours on a 2-cm3 charge of solvent. The prospects for large DMFCs for heavier duty use are presently unclear.

Another thermochemical route has been explored to convert methanol to another biofuel, dimethylether (DME), (CH3)2O, a highly volatile liquid that is a suitable fuel for diesel engines because of its low self-ignition temperature and high cetane num — ber.103 Although bio-DME has only half the energy content of conventional diesel, diesel engines can easily be retrofitted for bio-DME use. Well-to-wheel analyses showed that bio-DME was a little inferior to FT diesel for total fossil fuel substitu­tion and pollutant emissions.9192

A 14-million-gallon/year facility for 95% ethanol production from sugarcane molas­ses was technically the easiest production process to design and cost in the 1970s (table 5.3). To simplify even further, although coproducts (CO2 and fusel oils) were considered, no economic calculations were made for these; sale of the yeasts grown in

the fermentations were, however, included to effect a cost reduction of approximately 6.5%. The final calculated manufacturing price of 95% ethanol was $0.995/gallon, “a cost which is today very comparable to producing alcohol from ethylene.”12

Of the production costs, the raw material molasses was the dominating factor, accounting for 63% of the total — no exact geographical location for the hypotheti­cal facility was given but the quoted molasses price ($50/ton) was the 1978 summer average of molasses delivered to the East Coast and Midwest of the United States.

The major challenge in developing commercial H2 generation by “dark” biotechnology, that is, by fermentation processes, has been the low productivity of natural H2- evolving microbes, and this can be resolved into two distinct limitations: the molar yield from a fermentable substrate and the expected low growth rates and cell densi­ties of microbial producers under energy-poor environmental conditions.

On the basis of data presented in table 7.2, a maximum of 4 mol of H2 per mole of fermentable glucose substrate equates (on a mass balance) to only 8 g/180 g of sugar consumed. In cell-free biotransformations, using mixtures of enzymes, a nearly threefold higher productivity has been demonstrated.41 If the pentose phosphate pathway (figure 3.2) can be run in a cyclic manner to completely oxidize glucose 6-phosphate to CO2 and H2O, each mole of glucose consumed (in six “turns” of the cycle) can generate 12 mol of reduced cofactor:

6C6H12O6 + 12 NADP+ ^ 5C6H12O6 + 6CO2 + 12NADPH + 12H+

Coupling the reoxidation of NADPH to the hydrogenase from Pyrococcus furiosus (table 7.2), one of only a few hydrogenases known to accept NADPH as a reducing agent, generated 116 mol H2/mol of phosphorylated glucose oxidized.41

Japanese researchers genetically modified a strain of E. coli to overexpress the gene for formate hydrogen lyase, an enzyme catalyzing the reaction:

HCOOH ^ CO2 + H2

By growing the cells to high cell densities under glucose-supported aerobic conditions before transfer to an anaerobic fermentor, a high catalytic potential for formate transformation to H2 was established, reaching 300 L H2/hr/l culture.42 This rate of H2 production could support a 1-kW fuel cell operating at 50% efficiency using only 2 l of culture medium maintained under continuous conditions by a feed of formic acid. Further strain construction (deleting lactate dehydrogenase and fumarate reductase genes) has improved the induction of the formate hydrogen lyase activity.43 The same genetic manipulations have eliminated side reactions of (phosphoenol)pyruvic acid in glucose-grown E. coli, maximizing the transformation of pyruvate to formate via the pyruvate formate lyase-catalyzed step:

CH3COCOOH + CoASH ^ CH3CO-SCoA + HCOOH

Such a genetic background (with formate hydrogen lyase as the next step) increases the production of H2 from glucose as the fermentation substrate, although only rates of approximately 20 l/hr/l culture have been achieved.44

Thermophilic and hyperthermophilic microbes are obvious choices for production strains to be cultured at high temperatures to accelerate H2 formation.

A strain of the bacterium Klebsiella oxytoca isolated from a hot spring in China could produce H2 even in the presence of 10% O2 in the gas phase but had a low molar yield (1 mol/mol glucose consumed).45 The extreme thermophile Caldicel — lulosiruptor saccharolyticus shows up to 92% of the theoretical H2 yield from glu­cose (4 mol/mol) at low growth rates at 72-73°C, indicating possible applications in long-term free or immobilized cultures.46 This organism also can produce H2 from hydrolyzed paper sludge industrial waste as the sole carbon source and is unusual in that it can utilize xylose faster than glucose.4748

An advanced bioprocess option for H2 production utilizes a membrane bioreactor to maintain the bacteria inside the reactor while allowing fluids to exit.49 This design could be the optimal methodology to restrict the growth of methanogenic bacteria that consume H2 and generate CH4, a gas with only 42 percent of the energy content of H2.50 Restricting the residence time of materials in a continuous flow reactor system allows H2 producers to outcompete the slower-growing methanogens. With a 12-hr residence time, glucose could be utilized as a substrate for H2 production with an over­all consumption of 98% and an efficiency (assuming 4 mol H2/mol glucose) of 25%, accumulating H2 at a concentration of 57-60% (volume basis) in the headspace.49

The production of H2 need not be based on pure bacterial cultures, mixed cul­tures, even ones with only indirect evidence of the microbial flora present, being suitable for wastewater treatment or for local production sites in isolated rural locali­ties. A clostridial population (on the basis of the spectrum of metabolites produced in parallel to H2) provided a system capable of stable and prolonged production, with H2 reaching 51 percent in the gas phase and with no methanogenesis observed.51 Some process control is, however, unavoidable to maintain H2-evolving capacity, particu­larly pH: maintaining a pH of 6.0 may inhibit the growth of lactobacilli in a mixed culture of Clostridium and Coprothermobacter species that could utilize untreated sludge and lake sediment material as substrates.52 The choice of pH regulant may, however, be crucial, and the accumulation of sodium to toxic levels was a severe limitation in a continuous biohydrogen system from sucrose-supplemented anaerobic sewage sludge.53 At a constant pH, the combination of substrate-material retention time and temperature (in the 30-37°C range) can have a marked effect on the bal­ance between different clostridial species, the appearance of nonclostridial bacterial species and the overall molar yield of H2 from carbohydrates.54 In a molasses waste­water treatment plant in China, the H2 production rate was highest in an ethanol­forming stage of the process, and at least six types of H2-producing microbe were present, predominantly a novel species, Ethanoligenes harbinese.55 Such a complex microbial ecology may be highly adaptable to differing types and compositions of carbon sources during production cycles or when seasonally available.

The use of advanced reactor types has been explored; for example, the fluidized bed reactor design has been explored with a mixed community that rapidly estab­lished H2 production from C. butyricum; instability developed during the course of time as propionate producers gradually took over, and biofilm-type reactors may not be the optimal design because of the efficient adhesion of H2-consuming microbial species to the carrier.56 A trickle-bed reactor packed with glass beads inoculated with a pure culture of C. acetobutylicum certainly gave high H2 gas concentrations but soon (60-72 hours) clogged because of bacterial growth.57

Irrespective of the long-term prospects for the industrial production of H2, dark fermentations are very likely to be permanent features of wastewater treatment technologies and as an alternative to methane for local “biogas” production — indeed an obvious application of the trickle-bed reactor may be for the treatment of high-carbohydrate wastewaters, requiring no energy input for stirring a conventional mixed tank and producing H2 as a recoverable fuel gas as well as a more dilute stream for conventional biogas production.57 The use of water streams with lower organic loading may, however, be advantageous for H2 production because supersaturation of the gas space inside bioreactors may feed back to inhibit H2 synthesis.58 Removing CO2 (e. g., by the use of KOH to absorb the gas) is also beneficial to H2 production, probably by minimizing the flow of utilizable substrates to acetogenic bacteria capable of synthesizing acetic acid from H2 and CO2.59 Animal waste — contaminated water can also be made acceptable to biological H2 producers if the ammonia concentration can be reduced and maintained below toxic concentrations in continuous flow systems, especially if the microbial community can be gradually adapted to increased ammonia levels.60 Food processing aqueous streams with high chemical oxygen demands can support biohydrogen production at 100 times the rates possible with domestic wastewaters — often reaching commercially viable amounts of H2 if used on-site as a heating fuel.61

As an excellent example of the opportunistic use of H2-producing microbes in biofuels production, a strain of Enterobacter aero genes was shown to be highly adept at producing both H2 and ethanol from glycerol-containing wastewaters from biodiesel production; continuous production in a packed bed reactor using porous ceramic support material maximized the H2 production rate.62 But, as a final twist, H2 producers may have an unexpected role in assisting a microbial community of methanogens to achieve full productivity, that is, a syntropic relationship may be established to provide the methanogens with a readily utilizable substrate; adding mesophilic or thermophilic H2-producing cultures increases biogas production from animal manure slurry, and the added species persist for several months of semicontinuous operation.63

In 2000 (but using data from 1996), a review, two of whose authors had affiliations to ExxonMobil, concluded that sugarcane grown under Brazilian conditions could generate above-ground and harvestable biomass of 932 GJ/hectare/year; a total of 9.3 x 108 hectares of land could substitute the full global primary energy of fossil fuels (3.2 x 1011 GJ/year), even if only sugarcane stems were used as a fuel with a quan­titative extraction of the energy inside the plant material, that is, 343 GJ/hectare.81 Based on a total area of land used to grow crops worldwide of 1.4 x 109 hectares, the fossil fuel demand could, in principle, be met with 67% of cultivatable land dedi­cated to sugarcane as an energy crop, or only 24% if all of the harvestable material were used. If, however, the sugarcane stems were to be used for ethanol production, only a third of the useful energy would be converted to the biofuel, and more than twice the global area of land used to grow crops would be needed, that is, ethanol production could not substitute for the world’s appetite for fossil fuels.

This is an interesting but disingenuous set of calculations because ethanol had never been advocated as anything other than a convenient energy carrier as a replace­ment (partial or otherwise) for gasoline in automobiles. From the most recent data published by the International Energy Agency, oil used for transport is expected to continue being approximately 20% of the yearly fossil fuel demand or to slowly increase to more than 50% of the crude oil extraction rate (figure 5.8). To substitute the global demand for oil as a transport fuel, therefore, sugarcane grown for ethanol would need to occupy 40% of the world’s arable land (stems only as a feedstock) or 20-25% if the entire harvestable biomass were to be used for ethanol production. The long-perceived potential conflict between land use for food crop production and for bioenergy is highly likely to be a reality, and genetic engineering or other innovative

technologies cannot avoid this without massive increases over current bioprocess — limited abilities to transform biomass into ethanol (or any other liquid biofuel).82 Any conventional or molecular “improvement” of a major crop species must avoid greater dependence on fertilizer and other agrochemical application to achieve greater yields. A particular drawback is the reliance on energy crops with large inputs of nonrenew­able energy for the production of biofuels: corn-derived ethanol and, for biodiesel, rape (canola) and soybean seeds (see chapter 6, section 6.1).83

Equally, the use of agricultural residues for biofuel production has severe agro­nomic limitations. Although a worldwide potential residue harvest of 3.8 x 109 tonnes/year has been estimated, equivalent to 7.5 billion barrels of diesel, removing more than 30% of this could greatly increase soil erosion as well as depleting the soil organic carbon content. One solution to this problem is to establish new bioenergy plantations on uncultivated land, perhaps 250 million hectares worldwide.84 This could be supplemented by the exploitation of weed infestations, for example, water hyacinth in lakes and waterways, naturally occurring sugar-rich forest flowers, or other adventitious resources.85 86 Of the major candidate energy crops, switchgrass and other prairie grasses appear to offer the best balance among high net energy yields, low nutrient demands, and high soil and water conservation; with its high total root mass, switchgrass can replace soil carbon lost during decades of previous tilling within (perhaps) 20 years.87 Under U. S. farm conditions, there is mounting evidence that dedicated energy crops would significantly reduce erosion and chemi­cal runoff in comparison with conventional monoculture crops.88-90 Replacing con­tinuous corn cropping with a corn-wheat rotation and no-till field operations might maximize agricultural residue availability.91

Doubts persist, however, that any of these options represent true sustainability, even when consideration is extended to prolific plantations of acacia and eucalyptus
trees, both of which rapidly exhaust tropical soils of nutrients and fall far short of even 100 years of high productivity.79 The nearest approach to true sustainability could be the immediate-locale use of sun-dried wood from well-managed energy planta­tions, using highly efficient wood-burning stoves; the necessary high rates of biomass “mining” (probably requiring some method of energy-dependent biomass drying) to support predicted rates of growth of gasoline and biofuel consumption appears to be mathematically unsustainable.79 In any case, intensive energy crop plantations would certainly require a high degree of environmental monitoring and management to avoid biological collapse unless lavish amounts of fertilizers are applied and maxi­mum yields are ensured by equally large supplies of insecticides and pesticides.

A final twist is the impact of global climate change: could global warming (at least for a limited time, perhaps as long as 50-100 years) increase plant growth to such an extent that biomass extraction targets would be more closely met? Even accurately predicting biomass yields from monocultures of fast-growing tree species such as willow in the northeastern United States suffers from methodological weaknesses.92 The effects of changes in daily minimum and maximum temperatures are complex because they differentially influence crop yield parameters; this is a major area of uncertainty for projecting yield responses to climate change.93 For three major cereal species (wheat, corn, and barley), and an increase in annual global temperatures since 1980 of approximately 0.4°C, there is evidence for decrease in yield; the magnitude of the effect is small in comparison with the technological yield gains during the same period but suggests that rising temperature might cancel out the expected increase in yield because of increased CO2 concentrations.9495 Using the global environment as an uncontrolled experimental system has its methodological drawbacks, but the results of the one long-term field trial to attempt to isolate the effects of temperature on rice growth indicate that a 15% reduction in yield could result from each 1°C rise in temperature, a much greater effect than predicted by simulation models.96 The International Panel on Climate Change has, in stark contrast, predicted that CO2 benefits will exceed temperature-induced yield reductions with a modest rise in tem­perature. Because this conclusion has received wide media coverage, it is important to examine the detailed conclusions for food, fiber, and forest products: [53]

• Globally, commercial timber productivity rises modestly with climate change in the short to medium term, with large regional variability around the global trend.

Taken together, these International Panel on Climate Change prognoses suggest some short-term improvement in the productivity of a range of plant species, including timber grown as feedstocks for lignocellulosic ethanol, but the predictions become more unreliable as the geographical area narrows and related effects of climate are considered. This highlights the obvious conclusion that geostatistical surveys and controlled experiments should both be pursued vigorously as priority issues for agronomy and plant physiology in the next ten years.

Producing low-volume, high-cost products or, at least, middle-volume and middle — price coproducts from a biorefinery will be essential in establishing crop — and bio­mass-dependent biorefineries as components of the industrial landscape. The presence of up to 500 biorefineries in North America and 500-1000 across Europe brings the option of large numbers of small — to medium-capacity sugar streams in widely scat­tered locations, often in agricultural or afforested regions. Each site could house a fermentation facility for fine chemicals as well as being a site for biofuels — the costs of transporting biomass and of high-volume substrate solutions long distances by road or rail are unlikely to appeal either to industry or to environmentalists. The implication is that large numbers of production sites will arise where many commer­cial products for the chemical, food, and other industries will be synthesized. In turn, this suggests a reversal of the process by which large-scale fermentation production for antibiotics, enzymes, vitamins, food flavors, and acids has been exported from the United States and Europe to areas with lower labor and construction costs (India, China, etc.). This has been described as “restructuring the traditional fermentation industry into viable biorefineries.”82 Such a vision will almost certainly challenge the imaginations of industrial fermentation companies and demand that much closer attention is paid to the practical economics of the biorefinery concept(s).

Nevertheless, long-term strategic research around the world has begun to explore what might be achieved with such an abundant supply of pentose sugars from lig- nocellulosic biomass and of the metabolic and biosynthetic uses that might result.

Corynebacterium glutamicum has, as its name suggests, been much used for the industrial production of amino acids such as glutamic acid and lysine for the food and agricultural feed sectors. To broaden its substrate utilization range to include xylose, a two-gene xylose catabolic pathway was constructed using the E. coli xylA gene (encoding xylose isomerase) with either the E. coli xylB gene (for xylulokinase) or a corynebacterial gene for this enzyme; recombinants could grow in minimal media with xylose as the sole carbon source under aerobic conditions or, when O2- limited, utilize xylose alone or in combination with glucose.83

An unusual Lactobacillus sp. strain MONT4, isolated from a high-temperature fermenting grape must, is uniquely capable of fermenting L-arabinose to a mixture of d — and l-lactic acids; the organism contains two separate genes encoding lactate dehydrogenase with differing stereochemistries.84 D-lactic acid, along with “unnatu­ral” D-acids and amino acids, is contained in a series of bioactives developed as antiworming agents.85 Chemical transformations of lactic acids can yield a variety of chemical intermediates and feedstocks, notably acrylic acid and propylene glycol, compounds with major existing petrochemical-derived markets.86

Few (if any) major industrial-scale processes could fail to utilize individual sugars or mixtures of the carbohydrates emanating from lignocellulosic biomass processing or biodiesel production — even as atypical a microbe as Streptomyces clavuligerus (unable to metabolize glucose) can use glycerol for fermentations for the medically important в-lactam inhibitor clavulanic acid.87 The extensive experience of adapt­ing microbes to growing and functioning in lignocellulosic hydrolysates (chapters 3 and 4) should be readily transferable to industrial strains that already are expected to produce large amounts of high-value products in extremely concentrated media with sugars, oligosaccharides, or plant oils as substrates and with vegetable pro­teins or high concentrations of ammonium salts as nitrogen sources. Conversely, the genomes of already adapted ethanologens could (with removal of genes for ethanol formation) provide platform hosts for the expression of other biosynthetic pathways.

Only biopharmaceuticals, that is, recombinant proteins expressed in and produced by animal cell cultures, yeast, or E. coli, are impossible to merge with biorefineries because of regulatory restrictions and the absolute requirement for sterile cleanliness at the production facility; with the rise of biogeneric products, the production of biopharmaceuticals, already becoming global, will have completely left its early exclusive bases in southern California and western Europe.88 Less highly regulated bioprocesses, including all fermentation-derived chemicals not intended for biomed­ical use, are likely to migrate to sites of cheap and abundant carbon, nitrogen, and mineral nutrients, probably assisted by grants and incentives to provide employment in rural or isolated areas.

As an example of a tightly closed circular biorefinery process, consider biodiesel production from a vegetable oil — or, as work in Brazil indicates, intact oil-bearing seeds89 — using enzyme-catalyzed transesterification rather than alkaline hydroly- sis/methanolysis (chapter 6, section 6.1.2). The glycerol-containing effluent could be a carbon source for the methylotrophic yeast Pichia pastoris (able also to catabo — lize any contaminating methanol) that is widely used for expressing heterologous proteins, including enzymes; if the esterase used in the biodiesel process were to be produced on-site by a Pichia fermentation, the facility could be self-sufficient biotechnologically (figure 8.7). More ambitiously, any solid materials from the fer­mentation together with any “waste” plant material could provide the substrates for solid-state fermentations producing bioinsecticides, biopesticides, and biofertilizers for application to the farmland used to grow the oil crop itself.90

Microbial enzymes and single-cell proteins (as well as xylitol, lactic acid, and other fermentation products) have become targets for future innovations in utiliz­ing sugarcane bagasse in Brazil.91 The bioproduction of bacterial polyhydroxyal — kanoate polymers to replace petrochemical-based plastics is another goal of projects aiming to define mass-market uses of agricultural biomass resources in Brazil and elsewhere.92 Polyhydroxybutyrate and related polymers have had a long but unsuc­cessful record of searching for commercialization on a large scale on account of their uncompetitive economics and not always entirely industry-friendly chemical and physical properties, but high oil prices will increasingly sideline the former, whereas continued research into the vast number of possible biosynthetic structures may alight on unsuspected properties and uses — this also illustrates an impor­tant distinction made by the founders of “biocommodity engineering,” that is, that replacement of a fossil resource-derived product by a biomass-derived compound of identical composition is a conservative strategy, whereas a more radical one is that of substituting the existing chemical product with a biochemical with equivalent functionalities but with a distinct composition, which involves a more protracted transition but could be more promising in the long term.93

Using blue-green algae, not only to generate photobiohydrogen, but as microor­ganisms capable of much wider (and mostly unexplored) metabolic and biosynthetic capabilities, also offers innovative polymeric products, that is, exopolysaccharides whose massive cellular production has elicited interest in exploring their properties

and uses as industrial gums, bioflocculents, soil conditioners, emulsifiers and stabi­lizers, and vehicles for the removal and recovery of dissolved heavy metals.94

The biggest and most challenging leap for feedstocks for biorefineries, however, is that of using known microbial biochemistry to adsorb excess CO2 in the atmo­sphere. This is one of the two means — the other being the genetic manipulation of higher plants to increase their photosynthetic efficiencies (chapter 4, section 4.7) — that biotechnology could make decisive contributions to the global campaign to reduce atmospheric CO2 levels. Although plants use the Calvin-Benson cycle to fix CO2 into organic carbon compounds (initially sugar phosphates), CO2 fixation by “dark” metabolic processes are considerably less well publicized. The most recently discovered pathway was defined as recently as 1989 in an archeon, a type of single — celled microbe that is bacteria-like but evolved as an ancient line quite separately from the eubacteria and blue-green algae and whose members usually inhabit extreme environments (figure 8.8).95 The extensive cultivation of photobioreactors anywhere but in climates and locations with long guaranteed daily hours of intense sunlight is inefficient except in thin films, energy-requiring (e. g., to maintain an optimum temperature of 24°C or more[67]), and always limited by variations in the light/dark cycle.96 Light-independent bacterial bioprocesses avoid such limitations; if combined with biosynthetic pathways for high-value chemicals, bioreactors supplied with CO2 pumped into underground storage or with chemically adsorbed and released CO2 could be a primary technology for the later twenty-first century, whereas known and planned biofuels (other than H2) may prove disappointing in the extent to which their production and use over their full life cycles actually reduces transportation green­house gas emissions (chapter 1, section 1.6.2, and chapter 6, section 6.1.4), applying biotechnology to use accumulated CO2 as a process input would be a more widely applauded achievement.

Chemical routes to carbon capture and sequestration have already begun to establish themselves, especially for the most pertinent application of removing (as far as possible) CO2 emissions from power stations; trapping CO2 with amine-based systems has, however, been questioned as to its economics and its use of energy.97,98 Pumping trapped CO2 into sandstone deposits underneath the North Sea is practiced by the Norwegian oil industry but is of doubtful legality as well as requiring constant monitoring — it is, in many ways, a Faustian bargain with uncertain implications for the future despite its technical feasibility.99