How to define yield with energy crops? This is both more flexible and less precise than with, for example, crop yield as measured by grain size or weight per plant or per unit area. For a feedstock such as wheat straw, increasing plant biomass accumulation per plant may suffice, and this implies a change in nutrient utilization or absorption from the soil, but this may paradoxically reverse the trend toward “dwarf” crops (i. e., more grain, less stem/stalk) that has been emphasized in the green revolution type of agronomy.290 In principle, simply achieving higher leaf, stalk, and stem mass per plant is a straightforward target that is not limited by considerations of morphology for crops dedicated to energy supply and/or biofu­els. Mass clonal propagation of commercial trees is now well advanced, and gene transfer technologies have been devised for conifers and hardwoods, that is, forest biotechnology has emerged (probably irreversibly) out of the laboratory and into the global ecosystem.291

Traditional breeding and marker-assisted selection can identify genes involved in nutrient use efficiency that can then be used in gene transfer programs to improve features of plant nutrition — for crop plants in intensive agriculture, nitrogen assimilation and recycling within the plant over the stages of plant development are crucial.292 Although plant biochemistry is increasingly well understood at the molecular level, what is much less clear is how to accurately modulate gene expres­sion (single genes or whole pathways) to achieve harvestable yield increases.290 As with other “higher” organisms, a greater understanding of how regulatory circuits and networks control metabolism at organ and whole-plant levels as an exercise in systems biology will be necessary before metabolic engineering for yield in crop plants becomes routine.294

Nevertheless, successes are now being reported for “gene therapy” with the goal of improving the assimilation of CO2 into biomass, however this is defined:

• Transgenic rice plants with genes for phosphoeno/pyruvate carboxylase and pyruvate, orthophosphate dikinase from maize, where the two enzymes are key to high photosynthetic carbon fixation under tropical conditions (“C4 metabolism”), increases photosynthetic efficiency and grain yield by up to 35% and has the potential to enhance stress tolerance.295

• Once CO2 is “fixed” by green plants, some of the organic carbon is lost by respiratory pathways shared with microorganisms, and there are several reports that partially disabling the oxidative pathways of glucose metabo­lism enhances photosynthetic performance and overall growth: for exam­ple, in transgenic tomato plants with targeted decreases in the activity of mitochondrial malate dehydrogenase, CO2 assimilation rates increased by up to 11% and total plant dry matter by 19%.296

• Starch synthesis in developing seeds requires ADP-glucose phosphorylase; expressing a mutant maize gene for this enzyme in wheat increases both seed number and total plant biomass, these effects being dependent on increased photosynthetic rates early in seed development.297

Much attention has been given to improving the catalytic properties of the primary enzyme of CO2 fixation, ribulose 1,5-bisphosphate carboxylase (Rubisco), the most abundant single protein on Earth and one with a chronically poor kinetic efficiency for catalysis; although much knowledge has been garnered pertaining to the natural variation in rubisco’s catalytic properties from different plant species and in developing the molecular genetics for gene transfer among plants, positive effects on carbon metabolism as a direct result of varying the amounts of the enzyme in leaves have proved very slow to materialize.298 A more radical approach offers far greater benefits: accepting the inevitable side reaction catalyzed by Rubisco, that is, the formation of phosphoglyceric acid (figure 4.15), transgenic plants were constructed to contain a bacterial pathway to recycle the “lost” carbon entirely inside the chloro — plast rather than the route present in plant biochemistry that involved the concerted actions of enzyme in three plant cell organelles (chloroplast, mitochondrion, and peroxisome); transgenic plants grew faster, produced more biomass (in shoot and roots), and had elevated sugar contents.299

Is there an upper limit to plant productivity? A temperate zone crop such as wheat is physiologically and genetically capable of much higher productivity and efficiency of converting light and CO2 into biomass than can be achieved in a “real”

environment, that is, in hydroponics and with optimal mineral nutrition; again, most studies focus on yield parameters such as grain yield (in mass per unit area) but total leaf mass (a component of straw or stover) will also increase under such ideal conditions.300 The higher the light intensity, the greater the plant response, but natural environments have measurable total hours of sunshine per year, and the climate imposes average, minimum, and maximum rainfall and temperatures; supplementary lighting is expensive — but, given unlimited renewable energy resources, substantial increases in plant productivity are theoretically possible. From a shorter-term perspective, however, the choice of biomass refuses to be erased from the agenda, and agricultural wastes simply cannot compete with the “best” energy crops. For example, using biofuel yield as the metric, the following ranking can be computed from relative annual yields (liters per hectare):301

Corn stover (1.0) < poplar (2.9) < switchgrass (3.2) < elephant grass (Miscanthus, 4.4)

Rather than mobilizing molecular resources against the vagaries of climate, concen­trating effort on maximizing biomass supply from a portfolio of crops other than those most presently abundant would pay dividends. Overreliance on a few species (vulnerable to pests and climatic variability) would be minimized by expanding the range of such crops.

Glycerol represents 10% by weight of typical triglycerides, and biodiesel production generates large amounts of this coproduct (figure 6.3). This obviously marginalizes any chemical routes to glycerol synthesis that are either cost-inefficient or otherwise suboptimal in a competitive manufacturing environment.82 By any manufacturing route, glycerol production provides a good input to FT conversion to liquid alkanes because glycerol can be converted over platinum-based catalysts into syngas at rela­tively low temperatures, 225-350°C.104 The gas mixture from glycerol conversion at 300-450°C has a molar excess of H2 over CO (up to 1.83:1), a high ratio (up to 90:1) between CO and CO2, and only traces of methane. With a subsequent FT step, the overall conversion of glycerol to hydrocarbons can be written as

25C3O3H8 ^ 7C8H8 + 19CO2 + 37H2O

and is a mildly exothermic process (enthalpy change -63 kJ/mol glycerol). High rates of conversion of glycerol into syngas were observed using glycerol concentra­tions of 20-30% (w/w).

Unlike ethanol, a major fermentation product of only a selected few microbial species, glycerol is ubiquitous because of its incorporation into the triglycerides that are essential components of cellular membranes, as well as being accumulated in vegetable oils (figure 6.1). Unlike methanol, which is toxic to microbial species as well as higher animals and has no biosynthetic pathway, glycerol is benign and has a well-characterized route from glucose and other sugars (figure 2.3). In fermentations for potable ethanol, the priority is to regulate glycerol accumulation as its formation is a waste of metabolic potential in fuel alcohol production — and much research effort has, therefore, been devoted to minimizing glycerol formation by yeasts by, for example, regulating the glucose feeding rate to maintain an optimal balance of CO2 production and O2 consumption.105

Conversely, maximizing glycerol production by yeasts is also straightforward, successful strategies including:106

• Adding bisulfite to trap acetaldehyde (an intermediate in the formation of ethanol), thus inhibiting ethanol production and forcing glycerol accumula­tion to restore the balance of intracellular redox cofactors

• Growing yeast cultures at much higher pH values (7 or above) than tradi­tionally used for ethanol fermentations

• Using osmotolerant yeasts — glycerol is often accumulated inside yeast cells to counteract the adverse effects of high osmotic pressures

Suitable osmotolerant strains can accumulate 13% (w/v) glycerol within four to five days. Even more productive is the osmophilic yeast Pichia farinosa that was reported to produce glycerol at up to 30% (w/v) within 192 hours in a fed-batch fermentation with glucose as the carbon source as a molar yield of glycerol from glucose of 0.90.107

Fixing a maximum theoretical yield of glycerol from glucose is difficult because it is highly dependent on the totality of biochemical routes available to the producer cells. With Saccharomyces cerevisiae genetically manipulated to overproduce glycerol, the glycerol yield was 0.50 g/g of glucose, that is, a molar yield of 0.98.108 With this genetic background, a yield of 1 mol of glycerol/mol of glucose consumed may be the maximum obtainable (figure 6.10). Although an osmotolerant Saccharomyces strain isolated from sugarcane molasses could accumulate higher levels of glycerol (up to 260 g/l), the molar yield was still 0.92.109 An osmotolerant Candida glycerinogenes was, however, reported to produce glycerol with a molar yield of 1.25.110 Breaking through the limitation imposed by redox cofactors can, therefore, be accomplished by natural biochemistry. Some yeasts harbor a well-characterized “short circuit” for NADH oxidation, bypassing an energy conservation mechanism via an alternative oxidase, a mitochondrial enzyme constitutively expressed in industrial strains of Aspergillus niger used in citric acid prorduction.111113 Similar enzymes are known in plants, fungi, and many types of yeast; overexpressing the alternative oxidase gene in another yeast of industrial relevance (Pichia pastoris) resulted in a small increase in growth rate.114 Because the functioning of the alternative oxidase system is thought to circumvent any restrictions imposed on productivity by redox cofac­tor regeneration in citric acid producers, a bioconversion of each mole of glucose to produce nearly 2 mol of glycerol might be achievable.

Even with presently known levels of glycerol production, it is technically feasible to process a harvested filtered culture broth directly for syngas formation from glyc­erol.104 Because 50% of the annual costs of FT fuels from biomass were considered to be capital-related, combined syngas/FT conversion of glycerol would be expected to entail markedly lower costs, eliminating the need for a biomass gasifier and gas­cleaning steps.89,104

How small can a viable rural ethanol production site be? The final section of the 1978-1980 collection of case studies projected fuel ethanol production from corn on a family-run farm in the Midwest. The cost calculations were very different, however, because there was already a working market for fuel alcohol — and a known price ($1.74/gallon in Iowa, November 1979). Using this figure,[47] a net operat­ing profit of total revenues could be projected for the first year of operation, ignoring factors such as finished goods and work-in-process inventories that would reduce the actual production cost of ethanol (table 5.4).

property taxes). Also designed for farm-scale use in South Dakota, a solid-state fer­mentation process (chapter 4, section 4.6.2) using sweet sorghum as the feedstock and upscaled to produce in principle 83 l of 95% ethanol/hr was predicted to have produc­tion costs of $1.80/gallon; the single largest contributor (61%) to production costs was that of the sorghum feedstock.15

In comparison with fermentor-based H2 production, the use of photosynthetic organisms has received wider publicity because envisaged bioprocesses have convincing environmental credentials, that is, the ability to produce a carbonless fuel using only water, light, and air (CO2) as inputs. Hydrogen photobiology is, how­ever, highly problematic because of the incompatibility of the two essential steps:

• In the first stage, water is split to produce O2.

• In the second stage, the photoproduced electrons are combined with proteins to form H2 by either a hydrogenase or a nitrogenase — and O2 is a potent inhibitor of such an “anaerobic” system.

Nature provides two related solutions to this dilemma.64 First, filamentous cyanobacteria (e. g., Anabaena cylindrica) that compartmentalize the two reactions into different types of cell: vegetative cells for generating O2 from water and using the reducing power to fix CO2 into organic carbon compounds that then pass to specialized nitrogenase-containing heterocyst cells that evolve H2 when N2 reduction is blocked by low ambient concentrations of N2. The second scenario is that of non­heterocystous cyanobacteria that separate O2 and H2 evolution temporally (in day and night cycles), although the same overall effect could be achieved using sepa­rate light and dark reactors. With either type of nitrogen-fixing organism, however, the high energy requirement of nitrogenase would lower solar energy conversion efficiencies to unacceptably low levels.

Hydrogenase is the logical choice of biocatalyst for H2 production, and nearly 35 years have now passed since the remarkable experimental demonstration that simply mixing chloroplasts isolated from spinach leaves with hydrogenase and ferredoxin isolated by cells of Clostridium kluyveri generated a laboratory system capable of direct photolysis of water and H2 production.65 The overall reaction sequence in that “hybrid” biochemical arrangement was

H2O + light ^ /02 + 2e — + 2H+ ^ ferredoxin ^ hydrogenase ^ H2

Light-induced photolysis of water produced electrons that traveled via the photo­systems of the chloroplast preparations to reduce ferredoxin before hydrogenase catalyzing the reunion of electrons and protons to form molecular hydrogen. For over a quarter of a century, therefore, the nagging knowledge that direct photolytic H2 production is technically feasible has both tantalized and spurred on research into solar energy conversion.

Thirty years ago, the abilities of some unicellular green (chlorophyll-containing) algae, that is, microalgae, to generate H2 under unusual (02-free) conditions where hydrogenase was synthesized had already been defined.66 Such microalgae can, when illuminated at low light intensities in thin films (5-20 cellular monolayers), show conversion efficiencies of up to 24% of the photosynthetically active radiation.67

What is the biological function of hydrogenase in such highly aerobic organisms? An induction period with darkness and anaerobiosis appears to be essential.68 Photosynthetic H2 production is also enhanced if the concentration of C02 is low, suggesting that the hydrogenase pathway is competitive with the normal CO2- fixing activity of chloroplasts.69 Because the electron transport via the hydrogenase pathway is still coupled to bioenergy conservation (photosynthetic phosphorylation), hydrogenase may represent an “emergency” strategy in response to adverse environmental conditions, for example, in normally above-ground plant parts subject to water logging and anaerobiosis where essential maintenance and cellular repair reactions can still operate with a continuing source of energy. It naturally follows that reintroduction of O2 and CO2 would render such a function of hydrogenase superfluous — and explains the inhibitory effect of 02 on hydrogenase and the ability of even background levels of O2 to act as an electron acceptor in direct competi­tion with hydrogenase-mediated H2 production.70 Genetic manipulation and directed evolution of algal hydrogenases with reduced or (in the extreme case) no sensitivity to O2 is, therefore, unavoidable if maximal and sustained rates of photohydrogen production can be achieved in microalgal systems.

Progress has begun to be made on the molecular biology of microalgal hydrog — enases, including the isolation and cloning of the two genes for the homologous iron hydrogenases in the green alga Chlamydomonas reinhardtiV1 Random mutagenesis of hydrogenase genes could rapidly isolate novel forms retaining activity in the pres­ence of O2 and/or improved hydrogenase kinetics. Screening mutants of Chlamydo — monas reinhardtii has, however, revealed unexpected biochemical complexities, in particular the requirement for functional starch metabolism in H2 photoproduction.72 Several changes were indeed identified during the successful improvement of H2 photoproduction by this alga:73

• There was rational selection of mutants with altered electron transport activities with maximized electron flow to hydrogenase.

• Isolates were then screened for increased H2 production rates, leading to a mutant with reduced cellular O2 concentrations, thus having less inhibition of hydrogenase activity.

• The most productive mutant also had large starch reserves.

Using the conventional representation of electron transport inside chloroplast membrane systems, the possible interactions of photohydrogen production and other photosynthetic activities can be visualized (figure 7.7).74 Active endogenous metabolism could remove photoproduced O2 by using O2 as the terminal electron

acceptor in mitochondria; the problem of different spatial sites for O2 production and O2 utilization still, however, requires a reduced sensitivity of hydrogenase to O2 as the gas cannot be removed instantaneously — only in “test-tube” systems can O2-removing chemicals be supplied, for example, as glucose plus glucose oxidase to form gluconic acid by reaction between glucose and O2.65

The obvious implication of the redox chemistry of figure 7.7 is that the normal pro­cesses of photosynthesis, involving reduction of NADP for the subsequent reduction of CO2 to sugars, can be separated in time, with light-dependent O2 evolution and dark H2 production or if H2 production can proceed with inhibited O2 evolution, that is, “indirect biophotolysis.”64 The particular advantage of this arrangement is that the light-dependent stage can be operated in open pools to maximize productivity at mini­mal cost. Sustained H2 production could be achieved over approximately 100 hours after transfer of light-grown C. reinhardtii cells to a medium deficient in sulfur; these conditions reversibly inactivated Photosystem II and O2 evolution, whereas oxidative respiration in the continued light depleted O2, thus inducing hydrogenase.75 The subse­quent H2 production only occurred in the light and was probably a means of generating energy by Photosystem I activity (figure 7.7). Starch and protein were consumed while a small amount of acetic acid was accumulated.

This was the first reported account of a single-organism, two-stage photobiological production process for H2, although a prototype light/dark device using three stages (one light and two dark) with a marine microalga and a marine photosynthetic bacterium was tested in Japan in the 1990s.76 How much H2 could a microalgae-based approach produce? With C. reinhardtii cells given an average irradiance of 50 mol photons/m2/ day (a possible value in temperate latitudes, although highly variable on a day-to-day and seasonal basis), the maximum H2 production would be 20 g/m2/day, equivalent to 80 kg/acre/day (or 200 kg/hectare/day) — but the likely value, allowing for low yields of H2 production measured under laboratory conditions, the far from complete absorp­tion of incident light, and other factors, is only 10% of this.74 In a further refinement of this approach, the sulfate-limited microalgae were shown to form a stable process for 4000 hours: two automated photobioreactors were coupled to first grow the cells aero­bically before being continuously delivered to the second, anaerobic stage.77 Until all the biological and physical limitations can be overcome, however, large infrastructural investments in high and predictable sunlight regions would be required, and the capital costs for such solar power stations would be high, but the technical complexity may only approximate that of installing extensive photovoltaic cell banks for the direct produc­tion of electricity, an option vigorously advocated by critics of biofuels programs.78,79

Cyanobacteria (“blue-green algae”) are prokaryotes but share with higher pho­tosynthetic organisms the basic electron transport chains of Photosystems I and II (figure 7.7). The molecular biology and biochemistry of hydrogenases in cyanobac­teria is well understood, the complete genomes of several such organisms have been sequenced, and interspecies gene transfer is established.80 Much of the research has unfortunately concentrated on nitrogenase as a source of H2, but many cyanobacteria contain hydrogenases catalyzing the reversible formation of H2, a route with far more biotechnological potential for commercial H2 generation, and protein engineering has begun to reduce the O2 sensitivities of cyanobacterial hydrogenases.81 The physi­ological role of hydrogenase in cyanobacteria has been debated for decades; recent results suggest some kind of safety valve function under low O2 condition when a light-to-dark transition occurs, and inactivating quinol oxidase (an enzyme with a similar hypothetical function) and nitrate reductase (a third electron “sink”) increase photohydrogen evolution rates.82

Thermophilic cyanobacteria are known to be capable of H2 photoproduction at up to 50°C in open-air cultures maintained for more than 3 weeks.8384 If a fermentable carbon source is supplied, a sustained photoevolution of H2 can be achieved, with photolysis of water (a Photosystem II activity — see figure 7.7), whereas carbohy­drate-mediated reduction of the plastoquinone pool continues independently.85 This H2 production system has been termed “photofermentation”; in principle, relatively little light energy is required to drive the reaction because of the energy input from the fermentable substrate.64 The green alga C. reinhardtii shares this pattern of metabo­lism with cyanobacteria, behaving under photofermentative conditions much like an enteric bacterium such as E. coli, exhibiting pyruvate formate lyase activity and accu­mulating formate, ethanol, acetate, CO2, and H2 as well as glycerol and lactate.86

The overlapping molecular structures of cyanobacteria and nonphotosynthetic bacteria were exemplified by the coupling (both in vivo and in vitro) between cyano — bacterial photosynthetic electron transport components with clostridial hydrogenase; even more remarkable was the expression in a Synechococcus strain of the hydrogenase gene from C. pasteurianum, the enzyme being active in the cyanobacterial host.87 As a possible pointer to the future of designing an improved photosynthetic organism for H2 production, the “hard wiring” of a bacterial hydrogenase with a peripheral subunit of a Photosystem I subunit of the cyanobacterium Thermosynechococcus elongatus resulted in a fusion protein that could associate functionally with the rest of the Pho­tosystem I complex in the cyanobacterium and display light-driven H2 evolution.88

Photosynthetic bacteria differ from other photosynthetic organisms in using bacteriochlorophyll rather than chlorophyll as the central pigment for light-induced electron transport; they also lack Photosystem II (figure 7.7) and perform anoxy — genic photosynthesis and require electron donors more reduced than water, includ­ing reduced sulfur and organic compounds.89 Being able to fix gaseous nitrogen, the photosynthetic bacteria contain nitrogenase in addition to hydrogenase and occur globally in widely different habitats, including fresh, brackish, and sea waters, hot sulfur springs, paddy fields, wastewaters, and even in Antarctica. Hydrogen can be photoproduced in the presence of an organic substrate, sometimes with high effi­ciencies deduced from the maximum theoretical H2 production on a molar basis (table 7.3). Both free and immobilized cells have been used to produce H2 during extended periods (table 7.4). All photosynthetic bacteria can use H2 as a reductant for the fixation of CO2 into organic carbon, and considerable reengineering of the molecular biochemistry is unavoidable if the cells are to be evolved into biological H2 producers.27 Photofermentations are also known, and Rhodobacter capsulatus has been used as a test organism to evaluate photobioreactor designs potentially reaching 3.7% conversion efficiency of absorbed light energy into H2 fuel energy.90

Photosynthetic bacteria may have the additional capability of catalyzing the “water shift” reaction (chapter 6, section 6.2.1):

CO + H2O ^ CO2 + H2

Photofermentative Hydrogen Production by Photosynthetic Bacteria

TABLE 7.4

Photofermentative Hydrogen Production by Immobilized Cells

H2 Evolution rate

Species Electron donor Immobilization method (ml/hr/g dry weight)

but, unlike the thermochemical process, at moderate temperatures and without multiple passages of gases through the reaction vessel.64 A continuous process was devised for Rhodospirillum rubrum with illumination supplied by a tungsten light.91 With biomass as the substrate for gasification, a substantially (if not entirely) biological process for H2 production can be envisaged. A National Renewable Energy Laboratory report concluded that a biological reactor would be larger and slower but could achieve comparable efficiencies of heat recovery in integrated systems; the most likely niche market use would occur in facilities where the water gas shift was an option occasionally (but gainfully) employed but where the start-up time for a thermal catalytic step would be undesirable.92

Patents describing processes for H2 production using photosynthetic microbes cover at least 23 years, and include topics as diverse as their basic biology, molecular and enzymic components, and analytical methodologies. In the last 5 years, however, several patents have also appeared focusing on biohydrogen production by fermentative organisms (table 7.5).

The case of the mostly widely applauded biofuel scheme to date, that of sugar­cane ethanol in Brazil, also has major doubts from the environmental perspective, although the first decade (1976-1985) of the program probably achieved a reason­able soil balance by recycling fermentor stillage as fertilizer, a valuable source of minerals, particularly potassium.20 36 With the great expansion of the industry sub­sequently, however, significant pollution problems have emerged. The volume of stillage that can be applied varies from location to location, and in regions with near-surface groundwater, much less stillage can be applied without contaminat­ing the water supply.36 In the case of the Ipojuca river in northeast Brazil, sugar cultivation and adjacent ethanol production plants use stillage extensively for both fertilization and irrigation, and this has led to water heating, acidification, increased turbidity, O2 imbalance, and increased coliform bacteria levels.98 The authors of this joint German-Brazilian study urged that a critical evaluation be made of the pres­ent environmental status of the sugar alcohol industry, focusing on developing more environmentally friendly cultivation methods, waste-reducing technologies, and water recycling to protect the region’s water resources.

The preservation of surface and groundwater in Brazil in general as a con­sequence of the sugar alcohol industry’s activities and development was ranked “uncertain, but probably possible” (table 5.17).99 Sugarcane plantations have been found to rank well for soil erosion and runoff criteria in some locations in Sao Paolo state, although the experimental results date from the 1950s (table 5.18). A much more recent study included in the second, Dutch-Brazilian report showed much poorer results for sugarcane in comparison with other monoculture crops (fig­ure 5.9). Nevertheless, although Brazilian sugarcane alcohol (viewed as an indus­trial process) makes massive demands on the water supply (21 m3/tonne of cane input), much of this water can (in principle) be recycled; in addition, Brazil enjoys such a large natural supply of freshwater from its eight major water basins (covering an area of 8.5 million km2) that the ratio of water extracted to supply is, on a global basis, exceedingly small: approximately 1%/annum, equivalent to 30-fold less than comparable data for Europe. Local seasonal shortages may, however, occur, and two of the four main sugar production regions have relatively low rainfalls (figure 5.10). Although sugar cultivation has mainly been rain-fed, irrigation is becoming more common.

Annual Soil Losses by Erosion and Runoff in Experimental Stations in Brazil

North and
northeast (sugar)

1.0 1.5

Rainfall (mm/km2/year)

FIGURE 5.10 Annual rainfall in main sugar-producing and other regions of Brazil. (Data from Smeets et al.99)

the enormous consumption levels of the Global North will not lead the Brazilian countryside out of poverty or help attain food sovereignty for its citizens.”100 On the other hand, to achieve poverty alleviation and the eradication of social exclusion and with support from environmentalists, Brazil proposed the Brazilian Energy Initiative at the 2002 World Summit on Sustainable Development (Johannesburg, South Africa) aiming at the establishment of global targets and timeframes of min­imum shares of energy from renewable sources.101 Headline figures for the global numbers of malnourished people known to international agencies are another datum point with a large uncertainty: from below 1 billion to 3.7 billion.3895 As an economist from the Earth Policy Institute was quoted as saying: “The competition for grain between the world’s 800 million motorists to maintain their mobility and its two billion poorest people who are simply trying to stay alive is emerging as an epic issue.”102

Brazil became the global leader in ethanol exports in 2006, exporting 19% (3 bil­lion liters) of its production — 1.7 billion liters of which were imported by the United States — and plans to export 200 billion liters annually by 2025, increasing sugar­cane planting to cover 30 million hectares.100 Sugar for ethanol will increasingly be viewed by nations without a strong industrial base but with suitable climatic condi­tions for sugarcane growth as a cash crop, in exactly the manner that Brazil regards coffee or soybeans; the example provided by Brazil in creating rural employment at low cost, reducing the economic burden of oil imports, and developing national industrial infrastructure will be one difficult to resist, especially if major sugar pro­ducers, including Brazil, India, Cuba, Thailand, South Africa, and Australia, unite to create an expanding alternative fuel market with sugar-derived ethanol.36 South Africa, for example, has a great and acknowledged need to improve its sugarcane

economy, where 97% of its sugarcane growers are small scale, achieving only a quarter of the productivity realized by commercial operators; sugar is produced in a surplus, most of which is exported, but a national plan to encourage biofuels usage is in place, and a first ethanol plant is planned for construction by a South African sugar producer[54] in neighboring Mozambique.103

Academic economists and agronomists are calling (and will continue to call) for an informed debate about land use in the context of increasingly large areas of highly fertile or marginal land being reallocated for energy crops.104 Although there is good evidence that sugarcane-derived ethanol in Brazil shows the highest agri­cultural land efficiency in both replacing fossil energy for transportation and avoid­ing greenhouse gas emissions (figure 5.11), impacts on acidification and human and ecological toxicity and deleterious environmental effects occurring mostly during the growing and processing of biomass are more often ranked as unfavorable than favorable in surveys.105

The principal economic drivers toward greater biofuel production in develop­ing economies are, however (and paradoxically), those widely accepted programs to reduce greenhouse gas emissions, increase energy security, and move to a scientifi­cally biobased economy by promoting the use of biofuels (table 5.19). If a new orga­nization of ethanol exporting countries, mostly in the Southern Hemisphere, arises to make up any shortfall in the production of endogenous biofuels in major OECD economies, only a sustained effort to require and enforce agronomically sound and environmentally safe practices on the part of those net importers will provide

200

5.12 The impact of fuel economy on projected demand for ethanol in various gasoline blends. (Data from Morrow et al.76)

Selected Policies on Light-Duty Vehicle Fuel Economy

• Shifting our reliance on petroleum products to biobased products that gen­erally have fewer harmful environmental effects

When another principle is added — strengthening rural economies and increasing demand for agricultural commodities — the main issues of the political agenda that has emerged post-2000 in both the United States and OECD economies in general are clear. There is one final argument, however, and one that commenced in the 1950s, that, instead of rendering the question of economic price of biofuels irrel­evant, reformulates the question to ask: how will biofuels affect the cost of living and personal disposable income in the twenty-first century?

Vast caverns of CO2-absorbing bacterial fermentations producing high carbon-con­tent products with immediate human use — including bacterial cellulose as fiber, single-cell protein, or bioplastics — may be industrial realities for the later years of the present century but perhaps a more compressed timescale should occupy a high priority on the biofuels and climate change agenda. Over the coming 25 years, “hard

truths” about the global energy future will be (or are) unavoidable, and the role of biomass and other renewables in the emerging technological mix is a key issue.100

Any individual’s ranking of the immediate challenges that could be met by biotechnology is biased and partial but a useful departure point may be the prior­ity list of discussion items in a major international conference on biofuels held in 2007.

Intended for Biofuel Production

With a lignocellulosic platform for bioethanol production, one obvious target is (as just discussed) the management of energy crop productivity to maximize the capture of solar energy and atmospheric CO2; the chemical composition of the bio­mass is, however, of great practical significance for the industrial bioprocessing of

feedstocks:302,303

1. Developing crop varieties with reduced lignin contents (especially with softwoods)

2. Crops with increased cellulose and, arguably, hemicellulose contents

3. Plants with the increased capability to degrade cellulose, hemicellulose, and lignin — after harvest (i. e., in a controlled manner capable of minimiz­ing biomass pretreatment)

Of these, modifying lignin content has been the most successful — classical genetics suggests that defining quantitative traits and their genetic loci is relatively easy, and (even better) some of these loci are those for increased cellulose biosynthesis.304 As collateral, there is the confidence-building conclusion that lignin contents of commer­cial forest trees have been reduced to improve pulping for the paper industry; the genetic fine-tuning of lignin content, composition, or both is now technically feasible.305

Reductions in plant lignin content have been claimed using both single — and multiple-gene modifications (figure 4.16):

• Down-regulating either of the initial two enzymes of lignin biosynthesis, phenylalanine ammonia lyase and cinnamate 4-hydroxylase (C4H), reduces lignin content and impairs vascular integrity in the structural tissues of

plants.305

• Deletion of the second activity of the bifunctional C4H enzyme, coumarate 3-hydroxylase, results in reduced lignin deposition.306

• Later enzymes in the lignin pathway were considered to be less amena­ble for inhibiting lignification but multiple-gene down-regulation could

be effective.307,308

• Inactivating O-methyltransferase activity with an aspen gene incorporated into a transmissible plasmid in the antisense orientation reduced lignin for­mation in Leucaena leucocephata[45] by 28%, increased monomeric phenolic levels, and increased the cellulose content by 9% but did not visibly affect the plant phenotype.309

Are “lignin-light” plants biologically viable for commercial cultivation? Altered stem lignin biosynthesis in aspen has a large effect on plant growth, reducing total leaf area and resulting in 30% less total carbon per plant; root growth was also

4-OH’Coumaric ||
acid

СОСоА

I

СН

4’OH’COumarylCoA II

4’OH’Coumarylaldehyde 4-OHcoumaryl alcohol

FIGURE 4.16 Outline of biosynthesis of lignin precursors: PAL, phenylalanine ammonia lyase; C4H, cinnamate 4-hydroxylase; C3H, 4-coumarate 3-hydroxylase; COMT, caffeate O-methyltransferase; CCoAOMT, caffeoyl-CoA O-methyltransferase; CCR, cinnamoyl-CoA reductase; CAD, cinnamyl alcohol dehydrogenase; F5H, ferulate 5-hydroxylase. (After Hertzberg et al.312)

compromised.310 Vascular impairment can lead to stunted growth.307 On the other hand, aspen wood in reduced-lignin transgenics was mechanically strong because less lignin was compensated for by increased xylem vessel cellulose.308 Smaller plants may be grown, as energy crops, in denser plantations; alternatively, plants with reduced stature may be easier to harvest, and various practical compromises between morphology and use can be imagined — this can be seen as analogous to the introduction of dwarfing rootstocks for fruit trees that greatly reduced plant height and canopy spread and facilitated manual and mechanical harvesting.

Also without obvious effects on plant growth and development was the introduc­tion and heterologous expression in rice of the gene from Acidothermus cellulolyticus encoding a thermostable endo-1,4-P-glucanase; this protein constituted approximately 5% of the total soluble protein in the plant and was used to hydrolyze cellulose in ammonia fiber explosion-pretreated rice and maize.311 More ambitiously, enzymes of polysaccharide depolymerization are being actively targeted by plant biotech compa­nies for new generations of crops intended for biofuels. A large number of genes are under strict developmental stage-specific transcriptional regulation for wood forma­tion in species such as hybrid aspen; at least 200 genes are of unknown function, pos­sibly undefined enzymes and transcription factors, but this implies that heterologous glucanases and other enzymes could be produced during plant senescence to provide lignocellulose processing in plants either before the preparation of substrates for con­ventional ethanol fermentation or in solid-phase bioprocesses (section 4.6.2).312

Glycerol was produced on an industrial scale by fermentation in the first quarter of the twentieth century (especially during World War I) but then declined, unable to compete with chemical synthesis from petrochemical feedstocks.106 A similar historical fate occurred with the ABE fermentation-producing “solvents,” that is, acetone, butanol, and ethanol in various proportions. Beginning (as with fuel etha­nol) with the oil crises of the 1970s, renewed interest was evinced in the technology, aided greatly by the accelerating advance of microbial physiology and genetics at that time.115116 The microbial species capable of this multiproduct biosynthesis are clostridia, which also have remarkable appetites for cellulosic and hemicellulosic polymers, able to metabolize hexose sugars and pentoses (usually, both xylose and arabinose).117,118 This again parallels the drive to produce ethanol from lignocellu — losic biomass substrates (chapter 3, section 3.3.2.5). It came as no surprise, there­fore, when the neologism “biobutanol” (for и-butanol, C4H9OH) appeared. DuPont, Wilmington, Delaware, and British Petroleum are the companies most associated with the development of butanol as an advanced biofuel and which aim to market biobutanol by the end of 2007; according to the DuPont publicity material (www2. dupont. com), biobutanol’s advantages are persuasive:

• Butanol has a higher energy content than ethanol and can be blended with gasoline at higher concentrations for use in standard vehicle engines (11.5% in the United States, with the potential to increase to 16%).

• Suitable for transport in pipelines, butanol has the potential to be intro­duced into gasoline easily and without additional supply infrastructure.

• Butanol/gasoline mixtures are less susceptible to separate in the presence of water than ethanol/gasoline blends, demanding no essential modifications to blending facilities, storage tanks, or retail station pumps.

• Butanol’s low vapor pressure (lower than gasoline) means that vapor pres­sure specifications do not need to be compromised.

• Production routes from conventional agricultural feedstocks (corn, wheat, sugarcane, beet sugar, cassava, and sorghum) are all possible, supporting global implementation.

• Lignocellulosics from fast-growing energy crops (e. g., grasses) or agricul­tural “wastes” (e. g., corn stover) are also feasible feedstocks.

The principal hurdles to process optimization were in manipulating cultures and strains to improve product specificity (figure 6.11) and yield and in reduc­ing the toxicity of butanol and O2 (the fermentation must be strictly anaerobic)

Carbon Source

FIGURE 6.11 Variation in butanol production with two strains of Clostridium acetobutylicum grown on six different carbon sources. (Data from Singh and Mishra.118)

to producing cells.118119 Notable among advances made in the last decade are the following:

• Isolation of hyperproducing strains — Clostridium beijerinckii BA101 expresses high activities of amylase when grown in starch-containing media, accumulating solvents up to 29 g/l and as high as 165 g/l when adapted to a fed-batch fermentation with product recovery by pervaporation using a silicone membrane.120122

• Gas stripping has also been developed as a cost-effective means to remove butanol and reduce any product inhibition.123

• At the molecular level, the high product yields with hyperproducing strains can be ascribed to a defective glucose transport system exhibiting poor regu­lation and a more efficient use of glucose during the solventogenic stage.124

• The demonstration that the ABE fermentation can utilize corn fiber sugars (glucose, xylose, arabinose, and galactose) and is not inhibited by major sugar degradation products of pretreated lignocellulosic substrates.125126

• Overexpression of a single clostridial gene to increase both solvent produc­tion and producer cell tolerance of product accumulation.127

• Improved understanding of the molecular events causing loss of productiv­ity in solventogenic strains spontaneously or during repeated subculturing or continuous fermentation.128129

A technoeconomic evaluation of a production facility with an annual capacity of 153,000 tonnes published in 2001 estimated production costs for butanol of $0.29/kg ($0.24/l, assuming a density of 0.8098 kg/l, or $0.89/gallon), assuming a conversion
efficiency of 0.50 g products per gram of glucose and corn as the feedstock.130 The calculations were noted to be very sensitive to the price paid for the corn, the worst — case scenario costs reaching $1.07/kg; with the best-case scenario, the production costs were probably competitive with conventional gasoline (at that time showing a high degree of price instability), allowing for a lower energy content (figure 5.1).

The downstream processing operations for the ABE fermentation are necessar­ily more complex than for fermentations with single product, for example, ethanol. Not only can the insoluble materials from the harvested fermentation be used as a source for animal feed production, but the fermentation broth must be efficiently fractionated to maximize the economic returns possible from three saleable solvent products. Detailed analysis of a conventional downstream process modeled solvent extraction (by 2-ethyl-1-hexanol), solvent stripping, and two distillation steps to recover 96% of the butanol from a butanol-dominated mix of products.131 An optimal arrangement of these downstream steps could reduce the operating costs by 22%.

Advanced bioprocess options have included the following:

• A continuous two-stage fermentation design to maintain the producing cells in the solventogenic stage132

• Packed bed biofilm reactors with C. acetobutylicum and C. beijerinckii133

• A continuous production system with a high cell density obtained by cell recycling and capable of operation for more than 200 hours without strain degeneration or loss of productivity134

• Simultaneous saccharification and fermentation processes have been inves­tigated by adding exogenous cellulase to poorly cellulolytic strains135

A novel feedstock for biobutanol production is sludge, that is, the waste product in activated sludge processes for wastewater treatments; this material is generated at 4 x 107 m3/year in Japan and most is discharged by dumping.136 Adding glucose to the sludge supported growth and butanol production and a marked reduction in the content of suspended solids within 24 hours. In the Netherlands, domestic organic waste, that is, food residues, have been tested as substrates for the clostridial ABE fermentation, using chemical and enzymic pretreatments; growth and ABE forma­tion were supported mainly by soluble sugars, and steam pretreatment produced inhibitors of either growth or solvent formation.137138

Echoing the theme of recycling is the MixAlco process, developed at Texas A&M University, College Station, Texas; this can accept sewage and industrial sludges, manure, agricultural residues, or sorted municipal waste) as a feedstock, treated with lime and mixed with acid-forming organism from a saline environment to produce a mixture of alcohols that are subsequently thermally converted to ketones and hydro­genated to alcohols, predominately propanol but including higher alcohols.139 This is another fermentation technology awaiting testing at a practical commercial scale.

A key change in the pricing structure of industrial alcohol in the United States occurred in the decade after 1975: the price of petrochemical ethylene showed an increase of nearly tenfold, and this steep rise in feedstock costs pushed the price of synthetic industrial alcohol from 150/l (570/gallon) to 530/l ($2.01/gallon).16 Corn prices fell significantly (from $129/ton to $87/ton) between 1984 and early 1988; because the coproduct costs were increasing as a percentage of the corn feedstock cost at that time, the “net corn cost” for ethanol production (the net cost as delivered to the ethanol production plant minus the revenue obtained by selling the coprod­ucts) and the net corn cost per unit volume of ethanol were both halved (figure 5.3).

By 1988, the costs involved in corn-derived ethanol production were entirely competitive with those of synthetic industrial alcohol (table 5.5). The major concern was that unexpectedly high investment costs could place a great strain on the eco­nomics of the process if the selling price for ethanol dipped: in general, such costs could be minimized by adding on anhydrous ethanol capacity to an existing bever­age alcohol plant or adding an ethanol production process to a starch or corn syrup plant, but expensive grassroots projects could face financial problems. Across the whole range of production facilities (small and large, new or with added capacity,

and with varying investment burdens), a manufacturing price for ethanol could be as low as 180/l (680/gallon) or as high as 420/l ($1.59/gallon).16 By 1988, the average fuel ethanol selling price had fallen below 3O0/l ($1.14/gallon), an economic move­ment that would have placed severe pressures on farm-scale production business plans (see section 5.2.1.4). As an incentive to fuel ethanol production and continuing the developments noted above (section 5.2.1.2), federal excise tax concession of 160/l (150/gallon), discounting by individual states by as much as 210/l (610/gallon), and direct payments by states to producers amounting to as much as 110/l (420/gallon), in conjunction with loan guarantees and urban development grants, encouraged the development of production by grassroots initiatives.16 Industrial-size facilities, built without special incentives, were already reaching capacities higher than 1 billion gallons/year as large corporations began to realize the earning potential of fuel etha­nol in what might become a consumer-led and consumer-oriented market.

It is highly doubtful that industrial biohydrogen processes will be the entry points for the widespread use of H2 as a fuel. Despite a number of major national and inter­national initiatives and research programs, fossil fuel-based and alternative energy processes are widely considered to be essential before 2030, or even as late as 2050. Of these nonbiological technologies, H2 production by coal gasification is clearly the worst alternative in terms of fossil energy use and greenhouse gas emissions (figure 7.8).93 Nevertheless, gasification and electricity-powered electrolytic routes to H2 offer the promise of production costs rivaling or even less than those of conventional gasoline for use in fuel cell-powered vehicles with an anticipated fuel economy approximately twice that of conventional internal combustion engines (figure 7.9). As a carbonless production route, the internationally accepted “route map” is the sulfur-iodine cycle based on the three reactions:

H2SO4 ^ SO2 + H2O + /O2 [850°C]

I2 + SO2 + 2H2O ^ 2HI + H2SO4 [120°C]

2HI ^ H2 + I2 [220-330°C]

The high temperatures required for the first reaction have prompted research pro­grams investigating solar-furnace splitting of sulfuric acid, for example, in the five — nation project HYTHEC (HYdrogen THErmochemical Cycles), involving research teams from France, Germany, Spain, Italy, and the United Kingdom in the “search for a long-term massive hydrogen production route” that would be sustainable and independent of fossil fuel reserves (www. hythec. org).

Electrolysis (wind)

The enormous added bonus of biohydrogen would be the use of other highly renewable resources as well as avoiding undue reliance on nuclear technology (an alternative means of providing the power for very-high-temperature reactors), highly persuasive rationales for the continuing interest in biohydrogen energy in the twenty-first century as exemplified by the International Energy Agency’s Hydrogen Implementing Agreement whose Task 15 involves Canada, Japan, Norway, Sweden, the Netherlands, the United Kingdom, and the United States in four R&D areas:94

• Light-driven H2 production by microalgae

• Maximizing photosynthetic efficiencies

• H2 fermentations

• Improving photobioreactors for H2 production

In Japan, all the major automobile manufacturers are active in the development of fuel cell-powered vehicles: Toyota, Honda, Nissan, Mazda, Daihatsu, Mitsubishi, and Suzuki.94

In Europe, HYVOLUTION is a program with partners from 11 European Union countries, Russia, and Turkey, funded by approximately $9.5 million, and aiming to establish decentralized H2 production from biomass, maximize the number and diversity of H2 production routes, and increase energy security of supply at both local and regional levels (www. biohydrogen. nl/hyvolution). The approach is based on combined bioprocesses with thermophilic and phototrophic bacteria to provide H2
production with high efficiencies in small-scale, cost-effective industries to reduce H2 production costs to $10/GJ by 2020 — with production costs in the $5-7/GJ range, biomass-derived H2 would be highly competitive with conventional fuels or biofuels.95 Principal subobjectives for HYVOLUTION include the following:

• Pretreatment technologies to optimize biodegradation of energy crops

• Maximized conversion of biomass to H2

• Assessment of installations for optimal gas cleaning

• Minimum energy demand and maximal product output

• Identification of market opportunities for a broad feedstock range

Based in Sweden, the SOLAR H program links molecular genetics and biomimetic chemistry to explore radically innovative approaches to renewable H2 production, including artificial photosynthesis in manmade systems (www. fotmol. uu. se). Japanese research has already explored aspects of this interface between industrial chemistry and photobiology, for example, incorporating an artificial chlorophyll (with a zinc ion replacing the green plant choice of magnesium) in a laboratory system with sucrose, the enzymes invertase and glucose oxidase, together with a platinum colloid to photoevolve H2.96

The size of the investment required to bring the hydrogen economy to fruition remains, however, daunting: from several billion to a few trillion dollars for several decades.97 The International Energy Agency also estimates that H2 production costs must be reduced by three — to tenfold and fuel cell costs by ten — to fiftyfold. Stationary fuel cells could represent 2-3% of global generating capacity by 2050, and total H2 use could reach 15.7 EJ by then. There are some appreciated risks in these prognos­tications, with governments holding back from imposing fuel taxes on H2 but impos­ing high CO2 penalties being strongly positive for increasing the possible use of H2, whereas high fuel cell prices for automobiles will be equally negative (figure 7.10).