Technical advances will define to what extent agricultural and wastewater resources can be transformed into quantities of electricity and liquid and gaseous fuels as energy carriers and to what extent these will contribute to local and national power demands, but certain industries could benefit greatly from considering their wastes as sources of bioenergy, either for immediate use or as a saleable commodity. To illustrate this, a bilateral study between Japan and Malaysia of the waste streams from the Malaysian palm oil industry (42.5 million tons in 2001) demonstrated that, with Enterobacter and Clostridium strains evolving H2 or producing acetone, buta­nol, and ethanol (chapter 6, section 6.3.3), more than 62,000 tons of oil equivalent in total energy could be generated.113

Biohydrogen (however it is produced by living cells) has the potential to mar­ginalize all other biofuels, from ethanol and biodiesel to all presently contemplated “bio” options for mass transportation — if, that is, onboard fuel cells achieve their cost and vehicle range targets. In one possible future, mariculture units growing cya­nobacteria in coastal waters will be the sources of renewable energy from water and sunlight, providing the H2 as the most environmentally friendly energy carrier.114 In another (Chinese) model, cultures of Propionibacterium, Clostridium, and Bacteri — odes species relentlessly ferment whatever substrates can be made available to them to form H2 and ethanol and/or mixtures of acids, quite independently of sunlight, and in processes that can be managed by simply adjusting the pH to determine the product stream.115

Even the hydrogen economy has its critics, however, and in another possible future, hybrid gasoline-electric vehicles will dominate the highways by as early as 2020, reducing gasoline and greenhouse gas emissions by 30-50%, with no major investments in fuel infrastructure; they may even be dually hybrid, being able to run on gasoline-biofuel blends, traveling 500 miles on a gallon of gasoline mixed with five gallons of cellulosic ethanol.116

When, in 1975, Ballard occupied a derelict motel in southern Arizona at the beginning of the quest to develop a viable technology to power an electric vehicle and so reduce dependency on fossil fuels, optimism may have been tempered with the realization that a long and uncertain journey had just commenced, but optimism was certainly rampant by the late 1990s.117 As an “entirely unauthorized” biography of Ballard (the early pacesetters in fuel cell technology for transportation) noted, the starting pistol for the race to develop a marketable fuel cell-powered automobile was fired in April 1997 when Daimler-Benz paid nearly $200 million for a 25% share in Ballard and committed itself to invest a further $300 million.118 A decade later, prog­ress continues worldwide, but projections about who crosses the winning line first and the eventual date of mass use of such ecofriendly vehicles tends to relentlessly slip back into “decades away” — inventors have, necessarily, to be ever hopeful about the futures of their brain children but underestimate production costs and the full sequence of engineering and other events that lead to commercialization.

Even if developments in the next two to three decades render both fuel ethanol and biodiesel obsolete, however, the many advances made in the biotechnology of the bioproduction of biofuels will not prove to be wasted. To return once more to the preface, biomass as the main supply of chemical feedstocks may be unavoidable in the twenty-first century as increased demand for gasoline in the rapidly developing economies of Asia and South America applies a price tourniquet to petrochemicals, particularly rapidly if oil reserves prove smaller than estimated or if an accelerat­ing CO2-dominated climate change forces political action to restrain CO2-producing industries.118 Can agricultural sources ever be justified as substrates for the produc­tion of transportation fuels? Or solely for automobile fuels?

Or can biofuels gain global approval as part of the mix of products emanating from biorefineries, in a flexible output that could replace petrochemicals, provide biofuels for blends according to market demands, and provide fuels for multiple types of fuel cells? The next and final chapter explores how the biorefinery concept emerged in the 1990s to be the beacon of a radically different vision of how biotech­nology and commodity chemical production can merge in another blueprint for a sustainable mobile and industrial society.

Contamination was a consideration in the poor take-up of continuous fermentation technologies by potable alcohol producers, especially because holding prepared wort for sometimes lengthy periods without yeast inoculation provided an excellent growth medium for adventitious microbes in the brewery.177 With the accumulation of operating experience in fuel alcohol facilities, bacterial populations have been identified that not only reduce yield but also can prove difficult to eradicate; some bacteria (including lactic acid producers) form biofilms under laboratory conditions and can colonize many (perhaps every) available surface in complex sequences of linked fermentors and the associated pipe work.190

TABLE 4.2

Immobilized and Free Cell Systems for Fuel Ethanol Production: Critical Parameters for Process Efficiency

When bacterial contaminants reach 106-107 cells/ml, the economic losses for ethanol production can reach 3% of volumetric capacity; if profitability is marginal, this will have a serious impact, and antibiotic regimes have been devised to pulse controlling agents through continuous processes.191 This prophylactic approach has been applied to continuous ethanol facilities where the total losses will be greater because continuous operations have begun to dominate the larger (>40 million gal — lons/year) production plants — an antibiotic such as penicillin G is not metabolized and degraded by S. cerevisiae, and its addition rate can be poised against its expected chemical degradation at the low pH of the fermentation broth.[40] Outside the spectrum of known antibiotics, a useful alternative is the curious (and little known) chemical adjunct between urea and hydrogen peroxide; this bacteriocidal agent can effectively control lactobacilli in wheat mash and provides useful levels of readily assimilable nitrogen and O2 (by enzyme-catalyzed decomposition of the peroxide) to enhance yeast growth and fermentative capacity.192

Biodiesel is unique among biofuels in not being a single, defined chemical compound but a variable mixture, even from a monoculture crop source. The triglycerides in any plant oil are a mixture of unsaturated and saturated fatty acids esterified to glycerol; fatty materials from land animals have much higher contents of saturated fatty acids (table 6.2).6 This variability has one far reaching implication: reducing the content of saturated fatty acid methyl esters in biodiesel reduces the cloud point, the tempera­ture below which crystallization becomes sufficiently advanced to plug fuel lines; a diesel suitable for winter use may have a cloud point below -11°C, and “winteriza­tion” (treatment at low temperature and removal of solidified material) of biodiesel generates a product with similar improved operability and startup characteristics.7,8

The idiosyncratic fatty acid content of canola seed oil, with its preponderance of the very long chain erucic acid (table 6.2), has a quite different significance. Erucic acid has been known since the 1950s to stimulate cholesterol synthesis by animals.9 The potential adverse health effects (increased risk of circulatory disease) led to leg­islation on the erucic acid content of edible oils and the development of low-erucic acid cultivars, whereas, by contrast, high-erucic acid oils have a market (estimated to be more than $120 million in 2004) because erucic acid and its derivatives are

Saturated Unsaturated

a Erucic (canola), C14 monoethenoic (soybean)

feedstocks for the manufacture of slip-promoting agents, surfactants, and other spe­cialized chemicals.10 High-erucic acid oils would be either desirable or neutral for biodiesel production, but low-erucic cultivars are higher yielding — and, in any case, legal requirements were in place in the European Union by 1992 to geographically separate the two types of “oilseed rape” cultivation to minimize cross-pollination and contamination of agricultural products intended for human consumption.4

The majority of the biodiesel producers continue to employ a base-catalyzed reaction with sodium or potassium hydroxide (figure 6.3).11 This has the economic attractions of low temperatures and pressures in the reaction, high conversion efficiencies in a single step, and no requirement for exotic materials in the construc­tion of the chemical reactor. The liberation of glycerol (sometimes referred to as “glycerine” or “glycerin”) in the transesterification reaction generates a potentially saleable coproduct (see section 6.3). The generation of fatty acid methyl esters is the same reaction as that to form volatile derivatives of fatty acids before their analysis by gas liquid chromatographic methods, and the key parameters for optimization are reaction time, temperature, and the molar ratio of oil to alcohol, but choices of

the type of catalyst used and the short-chain alcohol coreactant can also be made.12 Different oil types of plant origin have been the subject of process optimization studies; five recent examples, exemplifying the global nature of R&D activities with biodiesels, are summarized in table 6.3.13-17 More subtle factors include differential effects on product yield and purity; for example, temperature has a significant posi­tive effect on biodiesel purity but a negative influence on biodiesel yield, and the alcohol:oil molar ratio is only significant for biodiesel purity (with a positive influ­ence).17 Although the biodiesel yield increased at decreasing catalyst concentration and temperature, the methanol:oil ratio did not affect the material balance.18

A variety of novel catalysts have been explored, partly to avoid the use of caustic materials but also to facilitate catalyst recovery and reuse:

• Sulfonated amorphous carbon19

• Ion-exchange resins20

• Sodium ethoxide21

• Solid acid catalysts (e. g., ZnO)14

Indeed, the requirement for a catalyst can be eliminated if high temperatures and pressures are used to generate “supercritical” fluid conditions, under which alcohols can either react directly with triglycerides or (in two-stage procedures) with fatty acids liberated from triglycerides.22-24

Far greater attention has, however, been paid to developing a biotechnological approach to biodiesel production, employing enzyme catalysts, usually lipases, and employing their catalytic abilities to carry out transesterification (or alcohololysis) rather than straightforward hydrolyses of triglycerides to liberate free fatty acids and glycerol.25 The principal process advantage of the enzyme-based approach is the ability to use low to moderate temperatures and atmospheric pressure in the reaction vessel while ensuring little or no chemical decomposition (i. e., a high product purity); the main drawback is the much longer incubation times to achieve more than 90% conversion of the triglycerides, that is, up to 120 hours.26 The barrier to full commer­cialization is maintaining the (relatively expensive) enzyme active during repeated

batch use. Rival enzyme products show differing stability, and methanol appears to induce a faster loss of activity than does ethanol.27 Examples of enzyme-catalyzed processes using oils of plant origin and with prolonged survival of the lipases are summarized in table 6.4.26-29 Lipase from Pseudomonas cepacia was used in an immobilized form within a chemically inert, hydrophobic sol-gel support; under optimal conditions with soybean oil, high methyl and ethyl ester formations were achieved within a 1-hr reaction, and the immobilized lipase was consistently more active than the free enzyme, losing little activity when subjected to repeated uses.30 Stepwise addition of methanol to delay inactivation of the enzyme is another pos­sible strategy.31 Reversal of methanol-mediated inactivation of immobilized lipase has been demonstrated with higher alcohols (secondary and tertiary butanols).32 In the long term, molecular evolution technologies will develop lipases with reduced sensitivity to methanol and increased specific activities; in the short term, whole-cell biocatalysis has obvious potential for industrial application, offering on-site genera­tion of lipase activity in cell lines that could be selected to be robust for oil trans — esterification.33-35 As with cellulases (chapter 2, section 2.4.3), investigation of newly discovered microbes or extremophiles may reveal enzymes with properties particu­larly well suited for industrial use. A lipase-producing bacterium strain screened from soil samples of China, identified as Pseudomonas fluorescens, contains a novel psychrophilic lipase (with a temperature optimum of only 20°C); this may represent a highly competitive energy-saving biocatalyst because lipase-mediated biodiesel production is normally carried out at 35-50°C.36

Commercially available lipases and lipases identified in a wider spectrum of microbial enzyme producers can efficiently use different low-molecular-weight alcohols as substrates for transesterification. Substituting higher alcohols for metha­nol can maintain active lipase for much longer periods of continuous batch opera — tion.37 A research group in Italy has also exploited this lax substrate specificity to produce fatty acid esters from a mixture of linear and branched short-chain alcohols

that mimics the residual fusel oil left after ethanol; not only can this utilize a waste product from bioethanol production but the fatty acid esters are potentially impor­tant for biodiesels because they improve low-temperature properties.38

Crude oils can give poor transesterification rates because of their contents of free fatty acids and other components lost during refining; the free fatty acids (up to 3% of the oils) react with the alkaline catalysts and form saponified products during the transesterification. Crude soybean oil could be converted into methyl esters as well as refined oil if a lipase process was used, although a lengthy incubation period was again required.39 With an Indonesian seed oil (from Jatropha curcas) exhibiting very high free fatty acids (15%), a two-step pretreatment process was devised: the first step was carried out with sulfuric acid as catalyst in a reaction at 50°C and removing the methanol-water layer; the second step was a conventional transesterification using an alkaline catalyst to produce biodiesel at 65°C.40 Rice bran stored at room tempera­ture can show extensive (>75%) hydrolysis of triglycerides to free fatty acids; the suc­cessful processing of the oil fraction also required a two-step methanolysis process (but both steps being acid-catalyzed), resulting in a 98% methyl ester formation in less than 8 hr, and the coproduction of residue with high contents of nutraceuticals such as y-oryzanol and phytosterols.41 Supercritical methanol treatment (without any catalyst) at 350°C can generate esters from both triglycerides and free fatty acids, thus giving a simpler process with a higher total yield of biodeisel.42

Other innovations in biodiesel production have included the following:

• A six-stage continuous reactor for transesterification of palm oil in Thai­land, claimed to produce saleable biodiesel within residence time of six minutes in a laboratory prototype with a production capacity of 17.3 l/hr43

• A Romanian bench-scale continuous process for the manufacture of biodiesel from crude vegetable oils under high-power, low-frequency ultrasonic irradiation44

• A two-phase membrane reactor developed to produce biodiesel from canola oil and methanol (this combination is immiscible, providing a mass-trans­fer challenge in the early stages of the transesterification); this Canadian design of reactor is particularly useful in removing unreacted oil from the product, yielding high-purity biodiesel and shifting the reaction equilib­rium to the product side45

• A novel enzyme-catalyzed biodiesel process was developed to avoid the liberation of glycerol from triglycerides, maximizing the carbon recovery in the product; methyl acetate replaced methanol, and the resulting triacetylglycerol had no negative effect on the fuel properties of the biodiesel46

After biodiesel production, the fuel’s thermal properties have been improved — in this case, to reduce the onset of volatilization (table 6.1) of soybean-derived biodiesel to below that of conventional diesel — by ozonolysis; the onset freezing temperature of ozonated methyl soyate was reduced from -63°C to -86°C.47

The most radical development in biodiesel production has, however, been in Brazil where PETROBRAS has combined mineral and biological oils in the H-BIO
process; few details have been made public, but the essential step is to add a vegetable oil to the straight-run diesel, gasoil, and coker gasoil fractions from the refining pro­cess, the total streams then being catalytically hydrogenated.48 The triglycerides are transformed into linear hydrocarbon chains, similar to the hydrocarbons in the petro­leum oil streams; the conversion of triglycerides is high (at least 95%), with a small propane coproduct. Because the process takes advantage of the existing infrastruc­ture of an oil refinery, the potential exists for an orderly transition from conventional diesel to biodiesel blends, with a gradual increase in the “bio” input if (as widely predicted) oil reserves dwindle (chapter 5, section 5.6).

the Petrochemical Industry?

There is little doubt that biotechnology has — or will — produce viable solutions to all the manifold problems inherent in processing lignocellulosic biomass with maximum efficiency into ethanol and other advanced biofuels. But are these tech­niques and methodologies simply too elaborate and overtechnologized for the task in hand? Although academic groups insert increasingly complex and inclusive arrays of genes into microorganisms to metabolize the whole gamut of sugars pres­ent in woody feedstocks, is this the realistic technology toolbox for the massive amounts of primary chemical building blocks and biofuels that will (eventually) be required?

Of the three established economically significant biofuels presently produced, two (starch — and sugar-derived ethanol) have required only modest advances in carbohy­drate-processing enzymes as essential inputs from industrial biotechnology, whereas the third (biodiesel) has evolved as an entirely chemical operation. Recent applica­tions of chemical catalysis to the production of fuels and other value-added chemicals from biomass-derived oxygenated feedstocks have now opened up routes involving well-defined reactions of organic chemistry (dehydration/dehydrogenation, aldol con­densation, etc.).119 To the process chemist’s eyes, biorefineries are entirely analogous to petrochemical refineries in that a limited range of major feedstocks — biomass hydrolysates, vegetable oils, pyrolytic bio-oils, lignin, and others (in biorefineries) and oils and diesel fuel (in a petrochemical refinery) — are convertible by cata­lytic cracking, hydrotreating, and hydrocracking to a range of fuels (gasoline, diesel, aviation fuel, and liquid petroleum gas) and chemicals (olefins, etc.).120 The logical extrapolation is that biomass-derived and fossil fuel inputs to refineries could be mixed — and trials of that concept have indeed already begun, with oil compa­nies exploring cofeeding of biomass and petroleum feedstocks, the production of biofuels in petroleum refineries, and the direct production of diesel fuel from veg­etable oils by a hydrotreatment process. Eventually (and perhaps within the next 20-60 years), the massive infrastructure of oil refineries may find their conventional feedstocks dwindling; the massive accumulated experience of chemical engineer­ing could easily accommodate new feedstock types and modified chemical catalytic processes. Although biotechnologists have focused for the past decade on refining ethanologens and other biofuels producers to utilize more of the available carbon in highly processed lignocellulosic substrates, the chemists can offer pyrolytic and thermochemical technologies that can transform all the available carbon in crude biomass to forms that can either be catalytically modified to value-added chemicals or act as inputs to Fischer-Tropsch fuel production streams. At high temperatures and pressures, given well-designed catalysts and predictable downstream operations, no material is “refractory.” If the investment in the chemical hardware has already been made, why not ease into a new feedstock regime in a form of “reeducation,” gradu­ally restructuring the industry to a sustainable future?

Between 20 and 60 years: either limit is allowed by present versions of the “Hubbert curve” before the question of how to replace dwindling oil reserves in the production of the myriad chemicals on which Western industrialized societies rely becomes acute.121 What biotechnology now makes very clear is that understanding the metabolism of even as ancient a “domesticated” microbe as the common etha — nologenic yeast cannot be achieved purely in terms of genes and gene function but must include fermentation media and operating engineering in any attempt to map the biochemistry and molecular biology accompanying ethanol formation and accu­mulation in real-world and real-time fermentors.122,123

That the two final scientific references in this volume explore the computational re-creation of metabolism inside the fungus whose secretion of cotton-destroying enzymes initiated modern scientific exploration of cellulases and in ethanol-forming yeast cells is entirely appropriate. Nine millennia have passed since ethanol-forming microbes were first consciously — and (possibly) wisely — used by Homo sapiens; in 2007, we learned how to construct computer simulations of yeast metabolic networks and perform dynamic analyses and were able to conclude that genetic information alone is unable to optimize any biochemical pathway or maximize the yield of etha­nol or any other fermentation product. Mathematical modeling, genetic manipulations, and the application of advanced chemical engineering can combine to achieve optima but only on the clear understanding that any conclusion may not be portable to simple change in circumstances (with glucose and xylose as cosubstrates rather than when individually consumed) or to each and every running of a complex series of events, that is, the biologically variable growth and metabolism of a microorganism under multiple physiological stresses in concentrated media and driven to maximal productivity.

Twenty, 40, 50, or 100 years: yeast cells (or at least, those in laboratory strains) have had little more than 20 years of deliberate, goal-driven molecular evolution, and this despite an incomplete understanding of gene expression and its endogenous regulation. “Scientific” studies of beer and wine yeasts, their inheritance patterns, stability, and metabolic capabilities may extend to a century. Pasteur was defining modern concepts of fermentation little more 150 years ago. The present generation of research scientists may be unique: it is the first to work with the sense of a severe time limit, not racing to launch a product or file a patent, but with a deadline, the realization that — however imprecisely computed is that future date — novel tech­nologies must be in place before petrochemicals become increasingly scarce and all ready-at-hand energy sources become valuable commodities to be bartered, in other words, an ideal time in which to put to good use all the knowledge gained over nine millennia of evolving biotechnological craft and science.

[1] Chinese texts from ca. 1000 BC warn against overindulgence in distilled spirits.

• Whisky (or whiskey) was widely known in Ireland by the time of the Nor­man invasion of 1170-1172.

• Arnold de Villeneuve, a French chemist, wrote the first treatise on distilla­tion, ca. 1310.

• A comprehensive text on distilling was published in Frankfurt-am-Main (Germany) in 1556.

• The production of brandies by the distillation of grape wines became widespread in France in the seventeenth century.

• The first recorded production of grain spirits in North America was that by the director general of the colony of New Netherland in 1640 (on Staten Island).

• In 1779, 1,152 stills had been registered in Ireland — this number had fallen drastically to 246 by 1790 as illicit “moonshine” pot stills flourished.

• In 1826, a continuously operating still was patented by Robert Stein of Clackmannanshire, Scotland.

• The twin-column distillation apparatus devised by the Irishman Aeneas Coffey was accepted by the Bureau of Excise of the United Kingdom in 1830; this apparatus, with many variations and improvements to the basic design, continues to yield high-proof ethanol (94-96% by volume).

[2] The Automobile Club of America sponsored a competition for alcohol-powered vehicles in 1906.

[3] High demand for oil exceeded predictions in 1970.

• The continued closure of the Suez Canal after the 1967 war between Israel and Egypt was confounded by a shortage of tanker tonnage for the much longer voyage around South Africa.

[4] In 1990, the Instituto do Agucar e do Alcool, the body through which govern­mental policy for ethanol production had been exercised, was abolished.

• In 1993, a law was passed that all gasoline sold in Brazil would have a minimum of 20% ethanol by volume.

[5] In 2004-2005, Brazil was the world’s largest producer of ethanol, with 37% of the total, that is, 4.5 billion gallons.

• Brazil exported 15% of its total ethanol production in 2005.

[6] The 1970 Clean Air Act (amended in 1977 and 1990) began the requirement for cleaner burning gasoline and (eventually) the mandatory inclusion of “oxygenates,” that is, oxygen-rich additives.

• The 1988 Alternative Motor Fuels Act promoted the development of ethanol and other alternative fuels and alternative-fuel vehicles (AFVs).

• The 1992 Energy Policy Act defined a broad range of alternative fuels but, more urgently, required that the federal vehicle fleet include an increasing number of AFVs and that they be powered by domestically produced alternative fuels.

[7] Six Canadian companies, starting operations between 1981 and 2006, with total capacity of 418 million liters of ethanol per year, mostly from wheat starch, with a further 390 million liters under construction (Canadian Renewable Fuels Strategy, Canadian Renewable Fuels Association, Toronto, Ontario, Canada, April 2006).

[8] Climatic factors including seasonal hours of sunlight, precipitation, average temperature, and others cer­tainly affect crop yield and will, therefore, impact on the economics of ethanol production as the cost of, for example, corn grain as an essential input into the ethanol production process varies from year to year.

0

oxygenated hydrocarbons. It is only at temperatures above 1100°C that the thermal conversion steps are efficient and only CO is a significant waste product of incomplete combustion (although even this can be minimized above a threshold residence time in the zone of highest temperatures); below 700°C, the use of catalytic conversion is essential to avoid serious air pollution with gases and volatilized material including tars.2 With catalytic combustors in residential wood burners designed to high engineering standards, overall efficiencies can be 80% of the theoretical maximum and particulate emissions as low as 1 g/kg.

Direct combustion is, however, only one of three thermochemical options for biomass utilization. Gasification (incomplete combustion) yields different mixes of products depending on the conditions used:2,3

• With pure oxygen as the combustant, a producer (or synthesis) gas with a high CO content is produced.

• The use of air rather than oxygen reduces the heating value because nitrogen dilutes the mixture of gases.

[10] If water is present and high temperatures are reached, hydrogen may also be formed, but excess water tends to result in high CO2 concentrations and greatly reduces the heating value of the gaseous product.

Producer/synthesis gas resulting from gasification technologies generally has a low heating value (4-10 MJ/L) and is best suited to in situ power heat and/ or generation. The third thermochemical method, pyrolysis (i. e., heating in the absence of air or oxygen) can be an efficient means of generating a gas high in hydrogen and CO but can also yield charcoal, a material with many (sometimes ancient) industrial uses.

[11] Ethanol is formed if a culture growing below the crucial rate of glucose entry is transferred to anaero­bic conditions; the faster the subcritical rate of feeding, the faster ethanol is formed after transfer to a “fermentation” environment.12

[12] Steam explosion was originally patented for fiberboard production in 1926. In the 1970s, batch and continuous processes were developed by the Iotech Corporation and the Stake Corporation, respec­tively, in Canada.

TABLE 2.2

Chemical Pretreatment Methods for Lignocellulosic Biomass

Method Principle Pilot plant use? a

Acids Hemicellulose solubilization

dilute sulfuric —

dilute hydrochloric —

dilute nitric —

dilute phosphoric —

steaming with sulfuric acid +

impregnation

steam explosion/sulfuric acid —

impregnation

steam explosion/sulfur dioxide —

steam explosion/carbon dioxide —

Alkalis Delignifi cation + hemicellulose

removal

sodium hydroxide —

sodium hydroxide + peroxide —

steam explosion/sodium —

hydroxide

aqueous ammonia —

calcium hydroxide —

Solvents Delignifi cation

methanol —

ethanol —

acetone —

a Data from Hsu.19

[13] The conversion of renewable substrates into single-cell protein was the topic in the 1960s and 1970s that most clearly resembles bioethanol production 30-40 years later (see, for example, reference 23). Historically it was the first major failure of industrial biotechnology.

[14] Historically, this organism (designated until recently as Trichoderma reesei) represents the beginning of biotechnological interest in cellulase, as it caused the U. S. Army major equipment and supply prob­lems in the Pacific during World War II by digesting military cotton garments.

[15] The susceptibility of sophorose to P-glucoside-catalyzed degradation precludes the possibility of the disaccharide being used as a recoverable “catalyst” for cellulase expression, although a chemical analog lacking the glycosidic linkage may be more stable and recyclable.

[16] Terrestrial and marine bacteria

• Yeasts and fungi

• Rumen bacteria and protozoa

• Marine algae

• Wood-digesting insects

• Molluscs and crustaceans

• Higher plants (in particular, in germinating seeds)

[17] Much of the report was focused on projections for meeting a much larger biomass supply by combination of altered land use and increased productivity. These topics are covered later, in chapter 5, section 5.4.

[18] cerevisiae is, from the standpoint of classical microbial physiology, best

described as “facultatively fermentative.” That is, it can metabolize sugars such as

glucose either entirely to CO2 and water given an adequate O2 supply or (under micro­

aerobic conditions) generate large amounts of ethanol; this ability for dual metabolism

is exhibited by a large number of yeast species.10 In complete anaerobiosis, however,

[23] There is nothing archaic about this area of microbial physiology, one of the effects being emphasized in the title of a 2002 patent (Production of Lactate Using Crabtree Negative Organisms in Varying Culture Conditions, U. S. Patent 6,485,947, November 26, 2002) awarded to Cargill Dow Polymers, LLC, Minnetonka, MN.

[24] So similar are P. stipitis and C. shehatae that they were at one time described as anamorphs of each other but the application of modern nucleic acid analytical techniques established that they are distinct biological species; from their ribosomal RNA sequences, C. shehatae has only recently (in biological time) diverged from the Pichia group, but both are well seperated from S. cerevisiae.57

[25] This fermentative route is not available to a strictly aerobic yeast such as Schizosaccharomyces pombe, which contains a mitochondrial DHOdehase and requires a fully functional mitochondrial electron transport chain for DHOdehase activity.137

[26] Redox balance is achieved because ethanol (oxidation value -2) and CO2 (oxidation value +2) effec­tively cancel one another out, lactate and acetate having zero values, thus leaving a small inaccuracy for the trace of succinate (oxidation value +1) formed in the slightly “leaky” mutant.

[27] Tropical counterparts of yogurts and fermented milk drinks prepared using Z. mobilis were known to the Aztecs for their therapeutic properties.194

[28] Symport is the simultaneous transport of a substrate and a cation (Na+, H+, etc.) in the same direction, while antiport is the exchange of two like-charged compounds (for example, Na+ and H+) via a common carrier; 40% of substrate uptake into bacterial cells requires one or the other of these two types of ion-driven transport.154

[29] High conversion yield

• High ethanol tolerance

• Tolerance to hydrolysates

• No O2 requirement (i. e., a facultative anaerobe)

[30] The assignee of an early U. S. patent for a microbe (a Kluyveromyces yeast) capable of producing alcohol from xylose and cellobiose (table 3.6).

[31] Acid prehydrolysis times had been described as “less than 1 minute” in 1999 but a more flexible regime had been instituted after the facility became operational (figure 4.1).

[32] Including Bacillus species then being developed by Agrol Ltd. (later, Agrol Biotechnologies Ltd.), a spinout company from the University of London, England.

[33] Denmark is entirely above latitude 54°N; the southernmost regions of Canada reach below 50°N; both are too far north to grow biomass crops other than some cereal species and forest tress (and Denmark has a limited land availability for large-scale tree plantations).

• Liquid flow-through enhances hemicellulose sugar yields, increases cel­lulose digestibility to enzyme treatment, and reduces unwanted chemi­cal reactions but with the associated penalties of high water and energy use; some of the benefits of flow-through can be achieved by limited fluid movement and exchange early in the acid digestion process.42-48’56-57

• Pretreated corn stovers appear to be much less toxic to ethanologens than other agricultural substrates.44- 4647

• Moreover — removal of acetic acid (a degradation product acetylated hemi­cellulose sugars) has been demonstrated at 25-35°C using activated carbon powder — and a natural fungus has been identified to metabolize furans and actively grow in dilute acid hydrolysates from corn stover.52-53

[35] This Chinese work is particularly relevant for improving the economics of bioethanol production because sugarcane bagasse represents the most abundant lignocellulosic agricultural material in southern China.97

[36] It is sometimes asserted that brewers disliked the increased emphasis on hygiene and “scientific” precision but offering such precise, reproducible technology was a positive and populat selling point in the 1970s, explicitly stated in advertisements for major brands of nationally sold beers and lagers in the United Kingdom and elsewhere.

[37] Utilizing the cell recycling principle developed in New Zealand but adding a vacuum-enhanced continuous recovery of ethanol from the fermentation

[38] Avoiding high temperature “excursions” in the stage of ethanol produc­tion — glycerol formation becomes uncoupled to growth and glycerol

[39] A continuously productive immobilized system with cane sugar molasses as the substrate proved to be stable for up to 60 days at a wide range of dilution rates, 0.05-3.00/hr.

• The sugar utilization efficiency was 75% with an ethanol yield of 86% of the theoretical maximum from the supplied mix of carbohydrates.

• An ethanol outflow concentration of more than 45 g/l was measured at a dilution rate of 0.06/hr, decreasing to 43 g/l at 0.11/hr.

[40] It remains to be seen if bacterial populations secreting в-lactamase evolve in bioethanol plants, although a wide range of antibiotics could (in principle) to be used to mitigate this resistance.

[41] Fusarium wilt is a disease that affects more than a hundred species of plants; the fungus colonizes the xylem vessels, blocking water transport to the leaves.

[42] Often (but confusingly for discussions of biofuels) abbreviated to SSF, although arguably with a claim to prior use in this area of biotechnology.

[43] Bagasse combustion in steam turbines generates electricity at 1 MWh/m3 of produced alcohol.

• Anaerobic digestion of vinasse can produce enough biogas for 0.5 MWh/m3 of produced alcohol, and both processes have been applied at full scale at distilleries.

• Laboratory studies show that anaerobic digestion would also be beneficial for bagasse, increasing the power output to 2.25 MWh/m3 of produced alcohol if the nonbiodegradable residue is burned.

• The total potential power generation from biogas and combustion routes would be equivalent of 4% of national power demand.

[44] Permanent (“constitutive”) expression of genes encoding chitinases or a ribosomal-inactivating protein confers partial protection against fungal attack.

• Enhancing lignin deposition in response to fungal or bacterial invasion is a possible multigene defense.

• Overexpressing genes encoding biosynthesis of phytoalexin antibiotics have been explored together with the introduction of novel phytoalexin pathways by interspecies gene transfer.

• Specific natural plant disease resistance genes are beginning to be identi­fied and cloned for transfer into susceptible plants.

[45] This thornless tree, known by many different names but including false koa, horse tamarind, or jumbie bean, has been included in the list of the 100 worst invaders, forming dense thickets and difficult to eradi­cate once established. Being “corralled” in energy plantations may be an appropriate use for it — but adjacent farmers and horticulturists may vehemently disagree.

[46] Corn-based manufacturing facilities at scales of 10 million up to 100 mil­lion gallons/year

[47] “It is conceivable that as the price of gasoline increases to a point greater than the price of ethanol, producers could raise the price of ethanol to equalize the prices of the two liquid fuels.”12

The operating profit was, however, entirely represented by sales of the fermenta­tion stillage delivered (by truck) to neighbors within 5 miles of the farm. The selling price of such stillage would be much depressed if a large brewery or distiller were located nearby; if no net income could be generated by these means, then the facility would run at a loss. The capture of CO2 from the fermentation was not considered because the capital cost of the equipment was too high to give a good return on the investment. A localized and small-scale production of fuel ethanol could, therefore, provide all the fuel requirements for running a farm’s gasoline-consuming opera­tions and provide a reasonable financial return as a commercial venture — but only if the “agrobusiness” was run as an early example of a biorefinery (chapter 8), pro­ducing not only ethanol but a saleable fermentation — and corn-derived coproduct.

Published in 1982, a second survey of technology and economics for farm-scale ethanol production at more than 100 gallons/hr (or, up to 1 million gallons/year) estimated a total annual cost of $1.97/gallon as a breakeven figure.14 The technical aspects of the process had been investigated in a facility at the South Dakota State University with fermentation vessels of up to 5750 l capacity. The projected price included a $0.41/gallon sales income from the wet grain coproduct, annual amortized capital cost, operating costs, and fixed costs (including insurance, maintenance, and

advances, including the introduction of corn hybrids with properties tailored for wet milling (e. g., accelerated steeping) and improved fermentor designs allowing cell entrapment were estimated to offer cost reductions of 4-70/gallon.

The “missing” analysis of lignocellulosic ethanol in the mid-1990s was (in all probability) supplied by two reviews also published in 1996.2223 Two scenarios were considered in the cost modeling: a base case and an advanced technology option (although without a precise date for implementation); the data are summarized in table 5.10. Using plausible technology for the mid-1990s, a production cost of $1.18/ gallon was computed; a fourfold increase in the capacity of the facility, together with innovative bioprocess technologies, was predicted to reduce the production costs to approximately 500/gallon.

This organism has been known first as Termobacterium mobile and subsequently as Pseudomonas linderi since 1912; although it was first known in Europe as a spoiling agent in cider, its function in the making of potable beverages such as palm wines is well established in Africa, Central[27] and South America, the Middle East, South Asia, and the Pacific islands and can ferment the sugar sap of the Agave cactus to yield pulque.194 The species was described as “undoubtedly one of the most unique bacterium [sic] within the microbial world.”195 Its unusual biochemistry has already

U. S. Patents Awarded for Yeast and Bacterial Ethanol Producer Strains Capable of Utilizing Lignocellulosic Substrates

Date Title Assignee Patent Number January 11, 1983 Direct fermentation of D-xylose to ethanol by a xylose-fermenting yeast mutant Purdue Research Foundation, West Lafayette, IN 4368268 September 18, 1984 Process for manufacturing alcohol by fermentation Kyowa Hakko Kogyo Co. Ltd. 4472501 December 25, 1984 Production of ethanol by yeast using xylulose Purdue Research Foundation, West Lafayette, IN 4490468 April 16, 1985 Direct fermentation of D-xylose to ethanol by a xylose-fermenting yeast mutant Purdue Research Foundation, West Lafayette, IN 4511656 May 5, 1987 Process for enhanced fermentation of xylose to ethanol United States 4663284 June 20, 1989 Process for producing ethanol from plant biomass using the fungus Paecilomyces sp. United States 4840903 March 19, 1991 Ethanol production by Escherichia coli strains co-expressing Zymomonas PDC and ADH genes University of Florida, Gainesville, FL 5000000 December 13, 1994 Combined enzyme mediated fermentation of cellulous [sic] and xylose to ethanol… United States 5372939 May 2, 1995 Xylose isomerase gene of Thermus aquaticus Nihon Shokuhin Kako Co. Ltd., Tokyo, Japan 5411886 June 13, 1995 Ethanol production by recombinant hosts University of Florida, Gainesville, FL 5424202 October 3, 1995 Process for producing alcohol CSIR, New Delhi, India 5455163 January 9, 1996 Ethanol production in Gram-positive microbes University of Florida, Gainesville, FL 5482846 May 7, 1996 Recombinant Zymomonas for pentose fermentation Midwest Research Institute, Kansas City, MO 5514583 January 27, 1998 Pentose fermentation by recombinant Zymomonas Midwest Research Institute, Kansas City, MO 5712133 March 10, 1998 Recombinant Zymomonas for pentose fermentation Midwest Research Institute, Kansas City, MO 5726053 August 4, 1998 Recombinant yeasts for effective fermentation of glucose and xylose Purdue Research Foundation, West Lafayette, IN 5789210 December 1, 1998 Single Zymomonas mobilis strain for xylose and arabinose fermentation Midwest Research Institute, Kansas City, MO 5843760

Xylose utilization by recombinant yeasts

Ethanol production in Gram-positive microbes

Recombinant cells that highly express chromosomally integrated heterologous genes

Recombinant organisms capable of fermenting cellobiose

Stabilization of PET operon plasmids and ethanol production in bacterial strains…

Genetically modified cyanobacteria for the production of ethanol…

SHAM-insensitive terminal-oxidase gene from xylose-fermenting yeast

Pentose fermentation of normally toxic lignocellulose prehydrolysate with strain of Pichia stipitis…

Recombinant Zymomonas mobilis with improved xylose ultilization

Production of ethanol from xylose

Ethanol production in recombinant hosts

Genetically modified cyanobacteria for the production of ethanol…

High-speed, consecutive batch or continuous, low-effluent process…

Transformed microorganisms with improved properties

Xyrofin Oy, Helsinki, Finland University of Florida, Gainesville, FL University of Florida, Gainesville, FL

University of Florida Research Foundation, Inc., Gainesville, FL United States

Enol Energy, Inc., Toronto, Canada

Wisconsin Alumni Research Foundation, Madison, WI

Midwest Research Institute, Kansas City, MO

Midwest Research Institute, Kansas City, MO Xyrofin Oy, Helsinki, Finland University of Florida Research Foundation, Inc., Gainesville, FL

Enol Energy, Inc., Toronto, Canada Bio-Process Innovation, Inc., West Lafayette, IN Valtion Teknillin Tutkimuskeskus, Espoo, Finland

been described (section 3.3.1), but multiple curious features of its metabolism made it a promising target for industrial process development:

• Lacking an oxidative electron transport chain, the species is energetically grossly incompetent, that is, it can capture very little of the potential bioenergy in glucose — in other words, it is nearly ideal from the ethanol fermentation standpoint.

• What little energy production is achieved can be uncoupled from growth by an intracellular wastage (ATPase).

• It shows no Pasteur effect, seemingly oblivious to O2 levels regarding glu­cose metabolism — but acetate, acetaldehyde, and acetoin are accumulated with increasing oxygenation.

During the 1970s, biotechnological interest in Z. mobilis became intense.* A patent for its use in ethanol production from sucrose and fructose was granted in mid-1989 (table 3.6). In the same year, researchers at the University of Queensland, Brisbane, Australia, demonstrated the ability of Z. mobilis to ferment industrial substrates such as potato mash and wheat starch to ethanol, with 95-98% conversion efficiencies at ethanol concentrations up to 13.5% (v/v).196 The Australian process for producing ethanol from starch was scaled up to more than 13,000 l.39

The capability to utilize pentose sugars for ethanol production — with lignocel — lulosic substrates as the goal — was engineered into a strain recognized in 1981 as a superior ethanologen; strain CP4 (originally isolated from fermenting sugarcane juice) exhibited the most rapid rate of ethanol formation from glucose, achieved the highest concentration (>80 g/l from 200 g/l glucose), could ferment both glucose and sucrose at temperatures up to 42°C, and formed less polymeric fructose (levan) from sucrose than the other good ethanol producers. On transfer to high-glucose medium, CP4 had the shortest lag time before growth commenced and one of the shortest doubling times of the strains tested.197 Researchers at the University of Sydney then undertook a series of studies of the microbial physiology and biochemistry of the organism and upscaling fermentations from the laboratory:

• Pilot-scale (500 l)-evaluated mutant strains selected for increased ethanol tolerance while improving ethanol production from sucrose and molasses became targets for strain development.198

• To reduce malodorous H2S evolution by candidate strains, cysteine auxo — trophs were isolated from studies of sulfur-containing amino acids.198

• Technical and engineering developments greatly increased the productiv­ity of selected Z. mobilis strains (as discussed in the context of bioprocess technologies in chapter 4).198

• Direct genetic manipulation was explored to broaden the substrate range.199

• High-resolution 31P nuclear magnetic resonance (NMR) of intracellular phos­phate esters in cells fermenting glucose to ethanol showed that kinetic limita­tions could be deduced early in the ED pathway (figure 3.4), in the conversion of glucose 6-phosphate to 6-phosphonogluconate, and in the glycolytic pathway (phosphoglyceromutase), defining targets for rational genetic intervention.200

By 1993, a review on Z. mobilis could already reference 362 publications.

Strains of Z. mobilis were first engineered to catabolize xylose at the National Renewable Resources Laboratory, Golden, Colorado. Four genes for xylose utilization by E. coli were introduced into Z. mobilis strain CP4 and expressed: xylose isom- erase (xylA), xylulokinase (xylB), transketolase (tktA), and transaldolase (talB) on a plasmid under the control of strong constitutive promoters from Z. mobilis.201202 The transformant CP4 (pZB5) could grow on xylose as the carbon source with an ethanol yield of 86% of the theoretical maximum; crucially, xylose and glucose could be taken up by the cells simultaneously using a permease because no active (energy — expending), selective transport system for glucose exists in Z. mobilis; the transport “facilitator” for glucose is highly specific, and only mannose and (weakly) galactose, xylose, sucrose, and fructose appear to be taken up by this mechanism.203 Using a plasmid containing five genes from E. coli, araA (encoding L-arabinose isomerase), araB (L-ribulose kinase), araD (L-ribulose 5-phosphate-4-epimerase), plus tkta and talB, a strain (ATCC39767[pZB206]) was engineered to ferment l-arabinose and produce ethanol with a very high yield (96%) but at a slow rate, ascribed to the low affinity of the permease uptake mechanism for l-arabinose.204 A third NERL strain was ATCC39767 (identified as a good candidate for lignocellulose conversion based on the evidence of its growth in yellow poplar wood acid hydrolysates) trans­formed with a plasmid introducing genes for xylose metabolism and subsequently adapted for improved growth in the presence of hydrolysate inhibitors by serial sub­culture in progressively higher concentrations of the wood hydrolyate.202205

A strain cofermenting glucose, xylose, and arabinose was constructed by chro­mosomal integration of the genes; this strain (AX101, derived from ATCC39576) was genetically stable, fermented glucose and xylose much more rapidly than it did arabinose, but produced ethanol at a high efficiency (0.46 g/g sugar consumed) and with only minor accumulations of xylitol, lactic acid, and acetic acid.206-208 The major practical drawback for the AX101 strain is its sensitivity to acetic acid (formed in lignocellulosic hydrolysates by the breakdown of acetylated sugars); this sensitivity was demonstrated in trials of the strain with an agricultural waste (oat hulls) sub­strate pretreated by the two-stage acid process developed by the Iogen Corporation in Canada, although the bacterial ethanologen outperformed a yeast in both volumet­ric productivity and glucose to ethanol conversion.209

The University of Sydney researchers have also transformed their best candidate ethanologen with the NERL pZB5 plasmid to introduce xylose utilization; strain ZM4(pZB5) produced 62 g/l of ethanol from a medium of 65 g/l of both glucose and xylose, but its ethanol tolerance was lower than that of the Z. mobilis wild type.210 The recombinant Z. mobilis shares the energy limitation on xylose observed with E. coli.180 NMR examination of strain ZM4(pZB5) growing on glucose-xylose mixtures demonstrated low levels of nucleotide phosphate sugars inside the cells when xylose was mainly supporting metabolism; because these intracellular components are bio­synthetic precursors for cell replication, the energy limitation has a clear biochemical mechanism for growth restriction.211 In addition to the metabolic burden imposed by the plasmids, the production of unwanted by-products (xylitol, acetate, lactate, acetoin, and dihydroxyacetone) and the formation of xylitol phosphate as a possible inhibitor of enzyme-catalyzed processes may all contribute to the poorer fermenta­tion performance on xylose as a carbon source. Further NMR investigations showed that acetic acid at growth-inhibitory concentrations decreased nucleotide phosphate sugars inside the cells and caused acidification of the cytoplasm, both complex bio­chemical factors difficult to remedy by genetic manipulation.212 Taking one step back, a mutant of the ZM4 strain with greater tolerance to acetate was isolated by classical selection procedures; electroporating the pZB5 plasmid into this AcR strain resulted in a xylose-fermenting strain with demonstrably improved resistance to sodium acetate at a concentration of 12 g/l.213 Overexpressing a heterologous xylulokinase gene under the control of a native Z. mobilis promoter did not, however, increase growth or xylose metabolism on a xylose-containing medium, indicating that constraints on xylose uti­lization reside elsewhere in the catabolic pathway or in xylose uptake.214 The xylulo — kinase-catalyzed step was more convincingly rate-limiting for xylose utilization with a Z. mobilis strain constructed at the Forschungszentrum Julich (Germany) with K. pneumoniae XI and XK, as well as E. coli transketolase and transaldolase genes; overexpression of XK was deduced to be necessary and sufficient to generate strains capable of fermenting xylose to ethanol at up to 93% of the theoretical yield.215

The potential impact of the acetate inhibition of Z. mobilis is so severe with com­mercial process that investigations of the effect have continued to explore new molecu­lar targets for its abatement. With starting acetate concentrations in the range 0-8 g/l in fermentations of glucose and xylose mixtures, high acetate slowed the increase in intracellular ATP (as a measure of bioenergetic “health”).216 Expressing a gene from E. coli encoding a 24-amino acid proton-buffering peptide protects Z. mobilis strain CP4 from both low pH (<3.0) and acetic acid; optimization of this strategy may be success­ful with high-productivity strains for lignocellulose hydrolysate fermentation.217

5.1 BIOETHANOL MARKET FORCES IN 2007

5.1.1 The Impact of Oil Prices on the "Future" of Biofuels after 1980

“Economists attract ridicule and resentment in equal measures.”1

The most telling aspect of the above quote is not that it derives from a recent col­lection of essays originally published in The Economist, one of the leading opinion formers in Western liberal economic thought, but that it is the first sentence in the Introduction to that volume. Graphical representations from many economic sources share one common factor, a short time axis. In the world of practical economics, hours, days, weeks, and months dominate the art of telling the near future — for price movements in stock markets, in the profitability of major corporations and their mergers, the collapse of currencies, or surges in commodity prices. Extrapo­lations are usually linear extrapolations from a small historical database of recent trends. Pundits predict much, but whatever the outcome, their predictions may very soon be forgotten. Economic models may, with hindsight, be wildly optimistic or inaccurate, but by the time “hindsight” is raised as a debating issue, the original set of parameters may have become completely irrelevant.

The history of biofuels since the early 1970s exhibits such cycles of optimism and pessimism, of exaggerated claims or dire prognostications; a series of funding programs have blossomed but — sometimes equally rapidly — faded.2 The prime mover in that sequence of boom and bust, evangelism and hostility, newspaper head­lines and indifference, has invariably been the market price of oil. However undesir­able a driver this is in the ongoing discussions on the development of biofuels, from the viewpoint of the biotechnologist in the scientific research community, it can never be ignored.3 Even with supporting arguments based on energy security and ameliora­tion of greenhouse gas emissions, a high cost of any biofuel relative to that of gaso­line, diesel fuel, and heating oil is the main plank in the logic used by skeptics: that however worthy are the goal and vision of biofuels for the future, they simply can­not be afforded and — in a sophisticated twist of the argument — may themselves contribute to the continuing deprivation of energy-poor nations and societies while the energy-rich developed economies impose rationing of fossil fuel use and access to maintain their privileged position. Although lobbyists for the global oil industry clearly have a vested interest in continuously challenging the economic costs of bio­fuel production, an underlying fear that, whether significant climate change could be lessened by the adoption of biofuels for private transportation (a debatable but quantifiable scenario — see chapter 1, section 1.6.2) and whether “energy security” is simply a novel means of subsidizing inefficient farmers to grow larger and larger harvests of monoculture crops to maintain agricultural incomes and/or employment for a few decades more, only a clear understanding of the financial implications of biofuels can help fix the agenda for rational choices to be made about investing in new technologies across the wide spectrum of rival possible biofuel options in the twenty-first century.

It is, however, undeniable that oil price volatility can shake the confidence of any investor in bioenergy. In the two decades after 1983, the average retail price for gasoline (averaged over all available grades) was 830/gallon, but transient peaks and troughs reached 1290 and 550, respectively (figure 5.1). Slumping oil prices almost wiped out the young sugar ethanol-fueled car fleet in Brazil in the 1980s (chapter 1, section 1.2). Since 2003, the conclusion that the era of cheap oil is irreversibly over has been increasingly voiced as the demands of the burgeoning economies of India and China place unavoidable stresses on oil availability and market price.4 If accu­rate, that prediction would be the single most important aid and rationale for biofuels as a commercial reality.

The cultivation of photosynthetic microalgae under dark conditions, supplied with organic carbon, closely resembles typical microbial fermentations. Because several bac­terial species are well known as accumulators of triglycerides (oils) and esters of fatty acids with long-chain alcohols (waxes), the logical conclusion was to combine these biosynthetic abilities with that of ethanol formation to generate the precursors of tri­glycerides in microbial production systems, that is, “microdiesel” produced without any need for a chemically or enzymatically catalyzed transesterification.9 The simple bacte­rium Escherichia coli was used as host for the Zymomonas mobilis pyruvate decarbox­ylase and alcohol dehydrogenase genes for ethanol production (chapter 3, section 3.3.2) together with the gene encoding an unspecific wax ester synthase/acyl-CoA: diacylg — lycerol transferase from a bacterial strain (Acinetobacter baylyi) known to accumulate lipid as an internal cell storage reserve. The resulting recombinant could accumulate ethyl esters of fatty acids at up to 26% of the cellular dry mass in fermentations fed with glucose. Insomuch as glucose is a fully renewable carbohydrate supply (via, e. g., cel­lulose or starch), microdiesel is a genuinely sustainable source of preformed transporta­tion fuel — although the chemical engineering aspects of its extraction from bacterial cells and the economics of its production systems require further definition.

A refinement preliminary to industrial feasibility studies would be to transfer to a host capable of higher endogenous accumulation of lipids; many of these are Gram-negative species (like E. coli), and ethanol production in such species is a well-understood area of biochemistry.1011

Corn stover is the above-ground plant from which the corn grain has been removed, and the constituent parts are leaves, stalk, tassel, corn cob, and shuck (the husk around the grains when in the intact cob); up to 30% by dry weight of the harvested plant is repre­sented by the collected grain. In one of the earliest technological and economic reviews of corn-based fuel alcohol production, corn stover was included for consideration—but solely as an alternative to coal as a boiler fuel for distillation. In late 1978, the report for the U. S. DOE estimated that corn stover would increase the final cost of fuel ethanol by 40/gallon as the use of corn stover as a fuel entailed costs roughly double those of local Illinois coal.36 The use of corn stover was, therefore, considered to be “justified only if the plant is located in an area where transportation cost would cause a doubling of the coal cost, or environmental considerations would rule against the use of coal; neither of which is very likely.” Such arguments left corn stover in the field as an aid against soil erosion for over a decade until the option of lignocellulosic ethanol began to be seriously considered. By 2003, the National Renewable Energy Laboratory, Golden, Colorado, estimated the annual and sustainable production of corn stover as 80-100 million dry tonnes/year, of which 20% might be utilized in the manufacture of “fiber” products and fine chemicals (e. g., furfural); 60-80 million dry tonnes would remain as a substrate for bioethanol production.37 Five years earlier, an estimate of total corn stover availability had been more than 250 million tonnes, with 30 million being left on the fields for erosion control, leaving 100 million available for biofuels production.38 With expand­ing corn acreage and a definite future for corn-based ethanol, a supply of corn stover is ensured — and commercial drivers may direct that starch ethanol and “bioethanol” facilities might be best sited adjacent to one another (see chapter 2, section 2.6).

With corn stover rising up the rankings of biomass substrates for ethanol produc­tion in the United States and elsewhere, experimental investigations of pretreatment technologies has proliferated since 2002.39-50 From this impressive corpus of practi­cal knowledge, some reinforced conclusions are apparent: [34]

temperatures (25-55°C), the enzymic digestibility of the resulting cellulose was highly influenced by both the removal of acetylated hemicellulose residues and delignification, but deacetylation was not seriously influenced by the levels of O2 or the temperature.58 Adding a water washing to ammonia-pretreated material removed lignophenolic extractives and enhanced cellulose digestibility.59 Grinding into smaller particles increased the cellulose digestibility after ammonia fiber explosion, but the chemical compositions of the different particle size classes showed major changes in the contents of xylans and low-molecular-weight compounds (figure 4.4). This could be explained by the various fractions of corn stover being differentially degraded in smaller or larger particles on grinding; for example, the cobs are rela­tively refractive to size reduction; the smaller particle sizes after AFEX treatment were more cellulase-degradable than were larger particles. Electron microscopic chemical analysis of the surface of the pretreated material provided evidence that lignin-carbohydrate complexes (chapter 2, section 2.3.2) had been disrupted.59 The high hemicellulose content of corn cobs has been exploited in a development where aqueous ethanol-pretreated material is washed and then hydrolyzed with an endox — ylanase; food-grade xylooligosaccharides can easily be purified, and the cellulosic material is readily digestible with cellulase.60 An additional advantage of corn cobs is that they can be packed at high density, thus reducing the required water inputs and giving a high concentration in the xylan product stream.

The dominance of inorganic acids for acid pretreatment of biomass substrates has only recently been challenged by the use of maleic acid, one of the strongest organic dicarboxylic acids and a potential mimic of the active sites of hydrolase enzymes with two adjacent carboxylic acid residues at their active sites.61 In com­parison with dilute sulfuric acid, maleic acid use resulted in a greatly reduced loss of xylose at high solids loadings (150-200 g dry stover/l), resulting in 95% xylose

yields, only traces of furfural, and unconditioned hydrolysates that could be used by recombinant yeast for ethanol production; 90% of the maximum glucose release could be achieved by cellulase digestion of the pretreated stover within 160 hours.

Examination of (and experiments with) the cellulase digestion of pretreated corn stover have led to other conclusions for industrial applications:

• Studies of the binding of cellobiohydrolase to pretreated corn stover identi­fied access to the cellulose in cell wall fragments and the crystallinity of the cellulose microfibrils after pretreatment to be crucial.62

• Adding small amounts of surfactant-emulsifiers during cellulase digestion of pretreated corn stover also increased the conversion of cellulose, xylan, and total polysaccharide to sugars, by acting to disrupt lignocellulose, sta­bilize the enzyme, and improve the absorption of the enzyme to the mac­roscopic substrates.63

• With steam-pretreated corn stover, near-theoretical glucose yields could be achieved by combining xylanases with cellulase to degrade residual hemi — cellulose bound to lignocellulosic components.64

• The initial rate of cellulase catalyzed hydrolysis is influenced strongly by the cellulose crystallinity whereas the extent of cellulose digestion is most influ­enced by the residual lignin.65 Modern methods of polymer analysis (e. g., diffusive reflectance infrared and fluorescence techniques) used in this work may be adaptable to on-site monitoring of pretreated biomass substrates.

• The formation of glucose from pretreated corn stover catalyzed by cellulase is subject to product inhibition, and the effects of substrate concentration and the amount (“loading”) of the enzyme are important in determining kinetic parameters.66

• Cellulase and cellobiohydrolase can both be effectively recovered from pre­treated and digested corn stover and recycled with consequent cost savings of approximately 15% (50% if a 90% enzyme recovery could be achieved).67

• The solid material used for cellulase-catalyzed hydrolysis itself is a source of potential toxic compounds produced during pretreatment but trapped in the bulk solids; activated carbon is (as discussed above) effective in removing acidic inhibitors from the liquid phase resulting from digestion of the reintroduced substrate.68

A comparative study of several methods for corn stover pretreatment concluded that alkaline methodologies had the potential to reduce the quantities of cellulase necessary in cellulose digestion but that hemicellulase activities may require supplementation.69

Outside primary scientific journals, data from a range of sources (including reports prepared for governments and conference proceedings) were compiled on the basis of 2003 costings as a baseline for future cost modeling.32 Ethanol produced from sugar, starch (grain), and lignocellulosic sources covered production cost estimates from less than $1/gallon to more than $4/gallon (table 5.15). Even with the lower production costs for lignocellulosic ethanol in the United States, taking into account financial outlays and risks ($260 million for a 50-million-gallon annual production plant), an ethanol price of $2.75/gallon would be more realistic.33

The International Energy Agency’s most recent assessment of sugar — and starch- derived ethanol (2005 reference basis) is that Brazil enjoys the lowest unit costs ($0.20/l, or $0.76/gallon), starch-based ethanol in the United States costs (after pro­duction subsidies) an average of around $0.30/l (or $1.14/gallon), and a European cost (including all subsidies) is $0.55/l (or $2.08/gallon).34 Brazilian production costs for fuel alcohol, close to $100/barrel in 1980, decreased rapidly in the 1980s, and then more slowly, but only a severe shortage of sugarcane or a marked rise in sugar prices would interrupt the downward trend in production costs.35,36

With due allowance of the lower fuel value of ethanol, therefore, the historical trend of fuel ethanol production costs versus refinery gate price[50] of gasoline is show­ing some degree of convergence (figure 5.5). In particular, the real production costs of both sugar — and corn-derived ethanol have fallen so that the production costs (with all tax incentives in place, where appropriate) now is probably competitive with the production cost of gasoline, as predicted for biomass ethanol in 1999.37 Critics of the corn ethanol program have, however, argued that the price of fuel ethanol is arti­ficially low because total subsidies amount to $0.79/gallon for production costs of $1.21/gallon, that is, some $3 billion are expended in subsidizing the substitution of only 1% of the total oil use in the United States.38 Although incentives for domestic
„12-500-1

о ‘I и Л c

.2 400

a 300

0

ethanol use in Brazil were discontinued by 1999 (except as part of development poli­cies in the northeast region), a cross-subsidy was created during the 1990s to subsi­dize ethanol production through taxation on gasoline and diesel; this was operated via a tight government control of the sales prices of gasoline, diesel, and ethanol, and the monopoly represented in the country by PETROBRAS.36

Brazilian sugarcane ethanol has reached the stage of being an importable com­modity to the United States, promoting the development of shipping, port handling, and distribution network infrastructures. In Europe, grain alcohol might be cost — competitive — with the much higher tax rates prevailing in Europe, the scope for regulating the end user price is much higher. In contrast, the economics of lignocel — lulosic ethanol remain problematic, although it is possible that in the United States, at least, production costs may become competitive with gasoline within the next five to ten years unless, that is, crude oil prices decrease significantly again.

8.1 THE BIOREFINERY CONCEPT

As a neologism, “biorefinery” was probably coined in the early 1990s by Charles A. Abbas of the Archers Daniel Midland Company, Decatur, Illinois, extrapolating the practices implicit in the fractionation of corn and soybean — the wet milling process was an excellent example of a protobiorefinery (figure 1.20). Certainly, by the late 1990s, the word (or, in an occasional variant usage “biomass refinery”) was becoming increasingly popular.1 The concept has carried different meanings according to the user, but the central proposition has been that of a comparison with the petrochemical refin­ery that produces not only gasoline and other conventional fuels but also petrochemical feedstock compounds for the chemical industry: from a biorefinery, on this formal anal­ogy, the fuels would include ethanol, biodiesel, biohydrogen, and/or syngas products, whereas the range of fine chemicals is potentially enormous, reflecting the spectrum of materials that bacterial metabolism can fashion from carbohydrates and other mono­mers present in plant polysaccharides, proteins, and other macromolecules (figure 8.1).

The capacity to process biomass material through to a mixture of products (includ­ing biofuels) for resale distinguishes a biorefinery from, for example, a “traditional” fermentation facility manufacturing acids, amino acids, enzymes, or antibiotics, indus­trial sites that may use plant-derived inputs (corn steep liquor, soybean oil, soy protein, etc.) or from either of the two modern polymer processes producing any one output from biomass resources that are often discussed in the context of biorefineries:

• Cargill Dow’s patented process for polylactic acid (“Natureworks PLA”), pioneered at a site in Blair, Nebraska; this was the first commodity plastic to incorporate the principles of reduced energy consumption, waste genera­tion, and emission of greenhouse gases and was awarded the 2002 Presi­dential Green Chemistry award.2

• 1,3-Propanediol (1,3-PD) produced from glucose by highly genetically engineered Escherichia coli carrying genes from baker’s yeast and Kleb­siella pneumoniae in a process developed by a DuPont/Tate & Lye joint venture; 1,3-PD is a building block for the polymethylene terphthalate poly­mers used in textile manufacture.3

However good are these example of the use of modern biotechnology to support the bulk chemistry industry, they center on single-product fermentations (for lactic acid and 1,3-PD, respectively) that are not significantly different from many earlier bacte­rial bioprocesses — in particular, lactic acid has a very long history as a microbial ingredient of yogurts and is used in the food industry to control pH, add flavor, and

control microbial growth in products as diverse as alcoholic beverages, frozen desserts, and processed meat; the lactic acid production sector has major manufacturers in China, the United States, and Europe that utilize lactobaccili, bacilli, or Rhizopus molds in large-scale fermentations.

The following three definitions for biorefineries focus on the multiproduct (usu­ally) biofuel-associated nature of the envisaged successors to fossil-based units:

1. The U. S. Department of Energy: “A biorefinery is an overall concept of a processing plant where biomass feedstocks are converted and extracted into a spectrum of valuable products.”4

2. The National Renewable Energy Laboratory: “A biorefinery is a facility that integrates biomass conversion processes and equipment to produce fuels, power, and chemicals from biomass. The biomass concept is analogous to today’s petroleum refineries, which produce multiple fuels and products from petroleum.”5

3. “Third generation (generation-III) and more advanced biorefineries … will use agricultural or forest biomass to produce multiple product streams, for example ethanol for fuels, chemicals, and plastics.”1

In early 2008, no such biorefineries exist, but the concept provides a fascinating insight into how biofuel-production facilities could develop as stepping stones toward the global production of chemical intermediates from biomass resources if lignocel — lulosic ethanol fails to meet commercial targets or if other developments (e. g., the successful emergence of a global hydrogen economy) render liquid biofuels such as bioethanol and biodiesel short-lived experiments in industrial innovation.[64]

The sheer scale of chemical endeavor possible from biomass resources is moreover extremely persuasive:6

• By 2040, a world population of 10 billion could be supported by 2 billion hectares of land for food production leaving 800 million hectares for nonfoods.

• With a “modest” increase in agricultural productivity to 40 tonnes/ hectare/year, this land surplus to food production could yield 32 billion tonnes/year.

• Adding in 12 billion tonnes annually from forests and other agricultural waste streams yields 50 billion tonnes.

Of this total, only 1 billion tonnes would be required to generate all the organics[65] required as chemical feedstocks — leaving the rest for biofuels, including traditional biomass as a direct source of power and heat.

Calculations prepared from and for German industry show that agricultural waste only, that is, cereal straw, could match the total demand for E10:gasoline blends as well as all the ethylene manufactured for the national chemical and plas­tics industries plus surplus ethanol for use in E85 blends and other chemical uses (figure 8.2).7

Included in the range of roles proposed for biorefineries, as codified by biorefin — ery. nl, the umbrella organization in the Netherlands tasked with developing strategic aspects of biorefineries (www. biorefinery. nl), are the following:

1. Primary processing units for waste streams from existing agricultural endeavors

2. Essential technologies for ensuring that biomass-derived ethanol and other biofuels can be produced at costs competitive with conventional fuels

3. New additions to be integrated with the infrastructure of agricultural processing — these might include (in Europe) beet sugar refineries

As corollaries and (probably) axiomatic truths, biorefineries will only become “interest­ing” (as players in the industrial economy) when they reach large scales of operation and contribute significant amounts of materials to widely used and/or specialist chemistry platforms while being driven not essentially or solely as means to reduce greenhouse gas emissions[66] but by considerations of the future depletion of fossil fuel reserves and the desire to broaden the substrate base, with governments being instrumental in cata­lyzing these developments by favorable taxation regimes and economic subsidies.