I. INTRODUCTION

As late as the mid 1800s, biomass supplied the vast majority of the world’s energy and fuel needs and only started to be phased out in industrialized countries as the fossil fuel era began, slowly at first and then at a rapid rate. But with the onset of the First Oil Shock in the mid 1970s, biomass was again realized by many governments and policy makers to be a viable, domestic, energy resource that has the potential of reducing oil consumption and imports and improving the balance of payments and deficit problems caused by depen­dency on imported oil. For example, the contribution of biomass energy to U. S. energy consumption in the late 1970s was over 850,000 BOE/day, or more than 2% of total energy consumption at that time. By 1990, it had increased to about 1.4 million BOE/day, or 3.3% of energy consumption, and conservative projections indicate that by the year 2000, biomass energy consumption is expected to increase to 2.0 million BOE/day (Klass, 1994). Other industrialized countries have also increased biomass energy consump­tion. Canada, for example, consumed about 134,000 BOE/day of biomass energy, or 3% of its total energy demand in the late 1970s, and by 1992 had

increased consumption to 250,000 BOE/day, or 4.4% of total energy demand. Although biomass energy has continued to be utilized in Third World countries as a source of fuels and energy for many years, it has become a renewable carbon resource for energy and fuels once again for industrialized countries and is expected to exhibit substantial growth in the twenty-first century. In this chapter, the concept of virgin and waste biomass as an alternative source of supply for energy and fuels is examined and the potential of biomass energy and its market penetration are evaluated.

Grasses are very abundant forms of biomass (U. S. Dept, of Agriculture, 1948). About 400 genera and 6000 species are distributed all over the world and grow in all land habitats capable of supporting higher forms of plant life. Grasslands cover over one-half the continental United States, and about two — thirds of this land is privately owned. Grass, as a family (Gramineae), includes the great fruit crops, wheat, rice, corn, sugarcane, sorghum, millet, barley, and oats. Grass also includes the many species of sod crops that provide forage or pasturage for all types of farm animals. In the concept of grassland agriculture, grass also includes grass-related species such as the legumes family—the clo­vers, alfalfas, and many others. Grasses are grown as farm crops, for decorative purposes, for preserving the balance of productive capacity of lands by crop rotation, for controlling erosion on sloping lands, for the protection of water sheds, and for the stabilization of arid areas. Many advances in grassland agriculture have been made since the 1940s through breeding and the use of improved species of grass, alone or in seeding mixtures; cultural practices, including amending the soil to promote herbage growth best suited for specific purposes; and the adoption of better harvesting and storage techniques. Until the mid-1980s, very little of this effort had been directed to energy applications. A few examples of energy applications of grasses can be found such as the combustion of bagasse for steam and electric power, but many other opportuni­ties exist that have not been developed.

Perennial grasses have been suggested as candidate feedstocks for conversion to synfuels. Most perennial grasses can be grown vegetatively, and they reestab-

lish themselves rapidly after harvesting. Also, more than one harvest can usually be obtained per year. The warm-season grasses are preferred over the cool-season grasses because their growth increases rather than declines as the temperature rises to its maximum in the summer months. In certain areas, rainfall is adequate to permit harvesting every 3 to 4 weeks from late February into November, and yields between about 18 and 24 t/ha-year of dry grasses may be obtainable in managed grasslands. Some tropical and semitropical grasses are very productive and can yield as high as 50 to 60 t/ha-year on good sites (Westlake, 1963). The tropical fodder grass Digitaria decumbens has been grown at yields of organic matter as high as 85 t/ha-year (Westlake, 1963). Table 4.11 lists some promising grasses that have been proposed as energy resources in the United States (Cushman and Turhollow, 1991).

An example of a tropical grass that has been grown commercially as a combination foodstuff and fuel crop for many years is sugarcane (Saccharum

“Cushman and Turhollow (1991). bAveraged by site; data are for range of sites.

Thick-stem grass.

Thin-stem grass.

“Productivity rates after 1- to 2-year establishment period. Productivity rates after 1- to 2-year extablishment period.

spp.), but rising production costs, alternative sweeteners, and the nebulous mixture of changing social, political, and agricultural policy issues have not been kind to insular sugar planters (Alexander, 1993). A great deal of informa­tion has been accumulated about sugarcane, and it might well be used as a model tropical herbaceous crop for other biomass energy systems. It grows rapidly and produces high yields; the fibrous bagasse is used as boiler fuel for the generation of electric power; and sugar-derived ethanol is used as a motor fuel in gasoline blends (gasohol). Sugarcane plantations and the associated sugar processing and ethanol plants are in reality biomass fuel farms. About one-half of the organic material in sugarcane is sugar and the other half is fiber. Total cane biomass yields have been reported to range as high as 80 to 85 dry t/ha-year. Normal cultivation provides yields of about 50 to 59 dry t/ha-yr (Westlake, 1963). Studies on sugarcane managed specifically as an energy crop have been underway in Puerto Rico for several years. “First- generation energy cane” consisting of conventional varieties managed for opti­mal growth with irrigation averaged 186 green t/ha-year of whole cane includ­ing detached trash, whereas second generation yields exceeded 269 green t/ha-year (Alexander, 1983). At an average of 40 wt % dry matter, these yields range from 74 to 108 dry t/ha-year. Presuming the energy content of energy cane is about 18.5 GJ/dry t, the energy yields correspond to 1369 to 1998 GJ/ ha-year, or the equivalent of 232 to 338 BOE/ha-year, a very high yield.

In moderate climates, switchgrass (Panicum virgatum) has been recom­mended as one of the model biomass energy crops for North America because of its high yield potential, adaptation to marginal sites, and tolerance to water and nutrient limitations (Sanderson et ah, 1995). It is a warm-season grass native to much of North America and is a major species in tall prairie grasses. Average yields are reported to range from 5.5 to 11.3 dry t/ha-year in the midwestern and eastern United States (Wright, 1994). In the southwestern United States, evaluation of eight switchgrass cultivars showed that for six locations in Texas, single harvests of fertilized plots of the Alamo cultivar afforded the highest aver­age yields, 10.7 to 15.7 dry t/ha-year (Sanderson et ah, 1995).

Other productive grasses that have been given serious consideration as raw materials for the production of energy and synfuels include the perennials Reed canary grass, tall fescue, crested wheatgrass, weeping lovegrass, and Bermuda grass, the annual sorghum and its hybrids, and others. It is apparent that there are many grasses and related biomass species that can be considered for energy applications. They have many of the desirable characteristics needed for terrestrial biomass energy systems.

A. Fundamentals

Baling has long been used to densify hays, straws, and other agricultural crops such as cotton to simplify removal from the field and to reduce storage space and transportation costs. Baled straw has a density of 70 to 90 kg/m3 at 10 to 15 wt % moisture content, whereas the bulk density of piled straw is about 5 to 15% of this density range. When straws are compressed to form pellets, briquettes, or cubes in specially designed dies and presses, the density can be increased to 350 to 1200 kg/m3. In contrast, dried wood has a density of 600-700 kg/m3 and a bulk density of about 350 to 450 kg/m3, whereas the bulk densities and densities of wood briquettes are 700 to 800 kg/m3 and up to 1400 kg/m3, respectively.

One of the original uses of biomass pellets in the United States was as fodder. Alfalfa, other grasses, and some straws were pelletized and sold as livestock feed. Biomass densification appears to have the greatest use for up­grading agricultural and forestry residues that might otherwise be lost or that require disposal at additional cost. The potential advantages for energy and feedstock applications of densified waste biomass are evident. High-density, fabricated biomass shapes simplify the logistics of handling and storage, im­prove biomass stability, facilitate the feeding of solid biomass fuels to furnaces and feedstocks to reactors, and offer higher energy density, cleaner burning solid fuels that in some cases can approach the heating value of coals. However, the basic problem often encountered in the use of densified biomass fuels and feedstocks is production cost. Some of the economic factors are discussed in Part D.

The heating value depends on the moisture and ash contents of the densified material and is usually in the range of 15 to 17 MJ/kg. The use of asphaltic binders or pelletizing conditions that result in some carbonization can yield densified products that have higher heating values. Pellets, briquettes, and logs have been manufactured by densification methods from biomass for many years. Prestologs” made from waste wood and sawdust were marketed before 1940 in North America, and the market for pellet fuels made from wood sawdust, shavings, and chips for residential pellet-burning stoves has grown significantly since the 1980s (Pickering, 1995; Folk and Govett, 1992). Numer­ous commercial processes for production of densified fuels in the form of logs, briquettes, and pellets from a wide range of biomass provide domestic fuels for space heating; industry uses the pellets and briquettes as boiler fuels (Edwards, 1991). In Europe, briquettes made from waste biomass are commer­cially available and are used for both residential and industrial applications. In Spain, households consume 80% of the total production of briquettes for use in furnaces, fireplaces, and barbecues. Bakeries use them in furnaces, and small industries such as ceramic plants use them as boiler fuel (Ortiz, Miguez, and Granada, 1996).

In contrast to the pyrolysis conditions needed to increase charcoal yields, the conditions for increasing the yields of organic liquid products would be ex­pected to involve short heat-up and reaction times and rapid removal and quenching of the organic volatiles before they are carbonized. Further, if the pyrolysis temperature at which maximum devolatilization occurs is chosen, but which is insufficient to gasify the volatiles (i. e., convert them to light fuel gases), more liquid products should be produced. When the reaction temperature is too low, less devolatilization and more char and tar formation occur, and when the temperature is too high, gasification reactions are domi­nant. Intuitively, pyrolysis at short reaction times and intermediate tempera­tures would be expected to promote higher organic liquid yields at the expense of other products. Char and gas yields should be low under these conditions. Numerous studies have indeed demonstrated that short-residence-time pyroly­sis, or flash pyrolysis, can be performed with biomass feedstocks to maximize liquid yields (с/. Bridgwater and Bridge, 1991; Bridgwater and Peacocke, 1995).

In some of the early work on the continuous flash pyrolysis of biomass at atmospheric pressure, it was shown that at optimum temperatures, liquid yields are maximized (Scott and Piskorz, 1983). With entrained flow injection of biomass feedstock of —250 jam to +105 ju, m particle size into a mini — fluidized-bed reactor with sand heat carrier and vapor residence times of 0.44 s, it was found that the maximum yield of liquid products occurs at the optimum temperature, and that yield drops off sharply on both sides of this maximum. Pure cellulose was found to have an optimum temperature for production of liquids at 500°C, whereas the wheat straw and wood species tested had optimum pyrolysis temperatures for maximum liquids at 600°C and 500°C, respectively. The yields of organic liquids were of the order of 55 to 65% of the dry weight of the biomass fed. The liquids contained relatively large quantities of organic acids. Pilot plant tests verified these observations. As research progressed on the conversion of biomass by flash pyrolysis, the optimum conditions for maximum liquid yields were found to be temperatures within the range 400 to 600°C, vapor residence times within the range 0.1 to 2.0 s, particle sizes less than 2 mm and a maximum of 5 mm for wood feeds, and an oxygen-free gaseous atmosphere such as recycled flue gas in the pyrolysis zone (Scott, Piskorz, and Radlein, 1993). Any reactor that can be operated under these conditions and that provides for biomass heating so that the particle temperatures can exceed about 450°C before 10% weight loss occurs can be used as a flash pyrolysis reactor. Such designs include fluidized — bed reactors, circulating fluid beds, transport or entrained flow reactors with or without a solid heat carrier, ablative reactors, and reduced-pressure reactors. Maximum organic liquid yields were projected to be 55 to 65 wt % of woody feedstocks and 40 to 65 wt % of grass feedstocks. The product yields from the flash pyrolysis of softwood (white spruce) and hardwood (poplar) are shown in Table 8.10. The total pyrolytic liquid yields are about the same for each feed, 66.5 and 65.7% of the feed dry weight, but they include several sugars and polysaccharide derivatives that are normally solids at ambient conditions. About 50% of the pyrolytic liquid product is water soluble. The largest fraction is the pyrolytic lignin, the insoluble fraction that remains after water extraction of the pyrolytic liquid. In other research, it has been shown that because of differences in oxidation rates between the anhydrosugar and aromatic lignin-derived pyrolysis products, which are formed in the fast pyroly­sis of deionized or prehydrolyzed wood, it is possible to carry out flash pyrolysis with controlled levels of oxygen to selectively oxidize lignins with relatively little effect on the anhydrosugar yields (Piskorz et ah, 1995). The recovery of the anydrosugars for use as fermentable sugars or as chemicals is expected to be simplified because their concentrations in the pyrolysis liquid are then significantly increased. The complex nature of the products and the poor selectivity of pyrolysis are evident, but with suitable refining, a wide range of chemicals could be manufactured from the products. Indeed, similar technol­ogy has already been commercialized for the production of fuels and chemicals.

A commercial, flash pyrolysis plant (RTP™) built in the United States in 1989 is believed to be the first successful plant in the world based on fast pyrolysis (Graham, Freel, and Bergougnou, 1991). This plant had a capacity of 100 kg/h of particulate hardwood feedstocks and is an upflow design that incorporates complete recirculation of the solid heat carrier in a reactor system capable of operating between 450 and 600°C at a residence time in the range of 0.6 to 1.1 s. The plant is used for the production of boiler fuel and specialty chemicals such as flavorings and natural colorings. The liquid is pourable and pumpable at room temperature and has an HHV ranging from 15 to 19 MJ/ kg, including the contained moisture, or approximately the same heating value as the feedstock entering the conversion unit. Typical liquid yields from representative hardwoods at 10-15% moisture content are about 73 wt % of the feedstock, including contained moisture (Graham and Huffman, 1995). In general, the yield increases slightly with an increase in the cellulose content of the feedstock and decreases slightly with an increase in feedstock lignin. However, the energy yield is almost constant since lignin-derived liquids have a higher energy content than cellulose-derived liquids. The liquids are pro­duced as a single phase unlike the heavy tars produced by conventional biomass pyrolysis. Experimentation has shown that by manipulating the vapor con­denser operating conditions, the product liquid can be tailored for chemicals and boiler, diesel, or turbine fuel applications without altering the pyrolysis

process itself. Since 1989, larger commercial plants of capacity up to 70 green t/day of feedstock have been built in North America, demonstration plants have been completed in Europe, and others that range in feedstock capacity from 100 to 350 t/day are in the advanced planning stages (Graham, Freel, and Kravetz, 1996).

In the mid-1990s, the economics of the RTP process for a 100-t/day plant were viable at a product fuel oil price of $5.27/GJ if the biomass feedstocks were available at zero cost. This can occur in several situations, especially with captive sources of waste biomass. The fuel oil price covers all fixed and variable operating costs, the annual capital costs to finance debt (75% of total capital), and an acceptable rate of return on the equity investment (25% of total capital). It is assumed that wood is available in the form of wet chips, and that the drying and grinding equipment and pyrolysis unit are included in the capital cost estimates. Expressed in other terms, the cost of converting 1 dry tonne of feedstock to fuel oil by this process is about $60, which means that the cost of the fuel oil is then about $10.54/GJ if the cost of the biomass feedstock is $60/dry t. For zero-cost fuel oil, a tipping fee of $60/dry t of waste feedstock is required. All of this means that the economics can be favorable in locations where there is an abundance of feedstock at zero or negative cost. In other words, the technology is very site specific, or the cost of competitive products, petroleum fuel oils in this case, must be high enough to justify commercial plants.

An example of one of the first flash pyrolysis processes developed for waste biomass is shown in Fig. 8.4 (U. S. Environmental Protection Agency, 1975; Preston, 1976). In this process, MSW is separated by a sequence of steps to obtain refuse-derived fuel (RDF) and recyclables. The sequence consists of shredding of MSW and air classification to obtain the RDF, magnetic separation of the ferrous metals, screening and froth flotation to recover a glass cullet, and aluminum separation by an aluminum magnet. The RDF is dried in a rotary kiln to about 4 wt % moisture content, and finely divided to a particle size of which 80% is smaller than 14 mesh (1200 /u. m). The feed, about 0.23 kg of recycled char preheated to 760°C per kilogram of this finely divided material, is rapidly passed through the pyrolysis reactor at atmospheric pres­sure. The raw product mixture, which consists of product gas and liquid, the char fed to the reactor, and new char formed on pyrolysis, leaves the reactor at about 510°C. Separation of the gas and liquid from the char and rapid quenching to about 80°C yields the liquid fuel. The remaining gas is passed through a series of cleanup steps for in-plant use. Part of the gas is used as an oxygen-free solids transport medium for pyrolysis and part of it as fuel. The raw product yields are about 10 wt % water, 20 wt % char, 30 wt % gas, and 40 wt % liquid fuel. The product char has a heating value of about 20.9 MJ/kg, contains about 30 wt % ash, and is produced at an overall yield

of about 7.5 wt % of the dry feed. The corresponding values for the liquid fuel are about 24.4 MJ/kg, 0.2 to 0.4% ash, and 22.5 wt % of dry feed as received (approximately 1 bbl/ton of raw refuse). This product was proposed for use as a heating oil; its properties are compared with those of a typical No. 6 fuel oil in Table 8.11. It is apparent that some major differences exist, but successful combustion trials in a utility boiler with the liquid fuel were performed. A plant designed to process 181 t/day of MSW was built in the United States in 1977, operated for one year, and then after producing several thousand liters of oil, was mothballed because of operating problems in the MSW separation unit and flash pyrolysis reactor.

Another flash pyrolysis process (GTEFP process) operating at atmospheric pressure was also developed that affords liquid yields in excess of 60% on a dry basis from hardwoods (O’Neil, Kovac, and Gorton, 1990). Yields as high as 70% were projected for commercial plants. The GTEFP process was devel­oped in bench-scale studies and a large-scale PDU in which the particulate feed is entrained in hot combustion gases and pyrolyzed at millisecond resi­dence times and temperatures in the range 500°C. Typical higher heating values of the product oils were 22 MJ/kg.

Short-residence-time pyrolysis of biomass at reduced pressure has been found to improve the yields of liquid products (Roy et al, 1985, 1990). In this research, a large-scale, electrically heated, multiple-hearth PDU afforded pyrolytic oil yields in the range of about 50 wt % of the feedstock with a wide range of wood species. Pyrolysis temperatures in the last hearth were about 450°C. The optimum temperature range was found to be between 350 and 400°C at residence times of about 2 to 30 s (с/. Bridgwater and Bridge, 1991). A yield of 60 wt % on a dry, ash-free basis of pyrolytic oil was obtained at an average heating rate of 10°C/min at a total system pressure of 5 to 41 kPa. The process was operated at feed rates of 30 kg/h in a multiple-hearth pilot plant, which was shown to offer advantages in product separation and fraction­ation because of the primary condensing units attached to each hearth. How­ever, the char and gas yields still comprised about one-third of the products, so it is probable that the liquid yields could be improved further at shorter residence times. A multiple-hearth reactor may not be suitable for biomass pyrolysis at millisecond reaction times.

Still another fast biomass pyrolysis configuration for increasing liquid yields involves what is termed ablative pyrolysis (Diebold et al, 1987, 1990). In this system, biomass particles are entrained tangentially into a vortex tube with a jet of carrier gas at velocities over 100 m/s. This causes the solid particles to be centrifuged to the hot wall of the vortex reactor, where very rapid heat transfer occurs to the surface of the particles. Ablative or surface pyrolysis takes place at high rates essentially independent of the feedstock particle size. This type of conversion process favors chain-cleavage reactions to form oxygenated, organic vapors rather than chars and gases, and is expected to make it possible to design small reactors having high throughput rates. With temperatures of 625°C on the reactor walls, 60 to 70 wt % of the primary vapor products are composed of oxygenated organic compounds and polymer fragments. They condense to form acidic, water-soluble liquids that have nearly the same elemental composition as the feedstock. For softwood feeds, the vortex reactor produces about 58 to 67 wt % of the dry feedstock as primary pyrolysis oils, 10 to 12 wt % as char, 10 to 14 wt % as gases, and 13 to 16 wt % as water. A plant for conversion of 33 t/day of wood wastes has been designed for construction in the Midwest to evaluate this process (Johnson, Tomberlin, and Ayres, 1991; Johnson et al, 1993). Dry wood feedstock (13 mm) is discharged from a hopper and pneumatically transported by recy­cled pyrolysis gases to a vortex reactor enclosed in a furnace designed to burn pyrolysis gas and natural gas. Natural gas is used for startup. The vapor-rich gases leave the vortex reactor along with the char particles entrained in the gas stream and pass through a cyclone, which is designed to remove 99.5% of the solids. The vapors are condensed at the rate of 11 L/min (dry). This plant is projected to cost $1.5 million, to produce 870 L/h (wet) of oil and 0.21 t/h of char, and to have a net annual revenue before taxes of $194,000, presuming the tipping fee for accepting the waste wood is $ll/t, and the oil and char can be sold for $0.11/L and $88/t on the open market. Larger plants having capacities of 90 and 227 t/day are being designed to gain economies of scale.

Other fast biomass pyrolysis techniques for liquid products have been exam­ined wherein a reactive atmosphere is present during conversion to attempt to affect yields and product compositions. Examples are flash pyrolysis in an atmosphere of hydrogen (hydropyrolysis) (Bodle and Wright, 1982; Sundaram, Steinberg, and Fallon, 1984) and methane (methanolysis) (Steinberg, Fallon, and Sundaram, 1983). In hydrogen atmospheres, hydrogenation might be expected to occur, at least to some extent, to yield liquids with lower oxygen and higher hydrogen contents. The experimental data indicated only moderate improvements in product yields and compositions under the test conditions used in an entrained flow pyrolysis reactor. However, substantial changes were observed in methane atmospheres. Flash methanolysis of dried wood particles at residence times of 1 to 2.8 s in an entrained flow pyrolysis reactor at pressures of 138 to 1379 kPa and temperatures between 800 and 1050°C

afforded benzene, toluene, and the xylenes (BTX), a heavy oil, ethylene, and carbon monoxide. As much as 12% of the available carbon in the wood feed­stock was converted to BTX, 21% to ethylene, and 48% to carbon monoxide at 345 kPa and 1000°C. The maximum heavy oil yield was observed at 345 kPa and 800°C. Data obtained with a methane blanket alone under the same pyrolysis conditions showed that no products were formed. It was concluded from this research that the optimum conditions for maximum production of both BTX and ethylene are a reactor pressure of 345 kPa, a temperature of 1000°C, a residence time of less than Is, and a methane-to-wood feed ratio of about 5.

Several small-scale, fixed-bed, updraft gasifiers are operated commercially in Sweden and Finland for the gasification of a wide range of biomass feedstocks, including wood chips, saw mill residues, straw, and RDF (Patel, 1996). The gasifiers are marketed by Carbona Corporation and are refractory-lined shaft furnaces that are fed from the top by a hydraulically operated feeder. The units are equipped with hydraulically rotated mechanical grates at the bottom. Ash sintering is prevented by water vapor contained in the gasification air, and the ash is removed through an ash discharge system installed at the bottom. The moisture content of the feeds can range from 0 to 45%, and the corresponding heating values of the product gas are about 5.5 to 3.8 MJ/m3 (n). The gasifiers can be connected to a hot water or steam boiler depending on whether heat or electric power is desired, or alternatively, the product gas can be used for hot gas generation for kilns and dryers.

The discussion of photosynthesis to this point has concentrated more on the light reactions that occur in photosynthesis. The organic components in bio­mass are formed during the dark reactions. Some discussion of the biochemical pathways and organic intermediates involved in the reduction of C02 to sugars is beneficial because they play a significant role in our understanding of the molecular events of biomass growth, and in differentiating between the various kinds of biomass.

Before discussion of these pathways, it is important to note that the photo­synthetic pathways also involve several dark reactions that occur in the glycoly­sis of glucose. The metabolic pathways provide energy for cellular maintenance and growth, form 3-phosphoglyceric acid and 3-phosphoglyceraldehyde, from which almost all other cellular organic components are synthesized, and are virtually identical in a large variety of living organisms, for example, in corn, wheat, oats, legumes, algae, many bacteria, the muscle, brain, liver, and other organs of humans and animals, and many birds, insects, and reptiles. Several reactions in the dark reactions on C02 uptake are the same as those that occur in the metabolism of foodstuffs by these organisms. Seven reactions of photosynthesis are common to the Embden-Meyerhof metabolic pathway, and three reactions are common to the pentose phosphate metabolic pathway. Each pathway converts glucose to pyruvic acid.

In photosynthesis, C02 generally enters the leaves or stems of biomass through the stoma, the small intercellular openings in the epidermis. These openings provide the main route for both photosynthetic gas exchange and for water vapor loss in transpiration. At least three different biochemical path­ways can occur during C02 reduction to sugars (Rabinovitch, 1956; Loomis et ah, 1971; Osmond, 1978).

One pathway is called the Calvin or Calvin-Benson cycle and involves the three-carbon intermediate 3-phosphoglyceric acid. This cycle, which is sometimes referred to as the reductive pentose phosphate cycle, is used by autotrophic photochemolithotrophic bacteria, algae, and green plants. As shown in Fig. 3.2, ribulose-1,5-diphosphate (1) and C02 react to form 3- phosphoglyceric acid (11), which in turn is converted via 1,3-diphosphoglyceric acid (111) and 3-phosphoglyceraldehyde (IV) to glucose (V) and ribulose-5- phosphate (VI), from which I is regenerated. For every 6 molecules of C02 converted to 1 molecule of glucose in a dark reaction, 18 ATP, 12 NADPH, and 24 Fd molecules are required. Twelve molecules of II are formed in the chloroplasts from 6 molecules each of C02 and I. After these carboxylation reactions, a reductive phase occurs in which 12 molecules of II are successively transformed into 12 molecules of III and 12 molecules of IV, a triose phosphate. Ten molecules of the triose phosphate are then used to regenerate 6 molecules of I, which initiates the cycle again. The other 2 triose phosphate molecules are used to generate glucose.

Plant biomass species that use the Calvin-Benson cycle are called C3 plants. The cycle is common in many fruits, legumes, grains, and vegetables. C3 plants usually exhibit low rates of photosynthesis at light saturation, low light

CHO

6 co2

FIGURE 3.2 Biochemical pathway from carbon dioxide to glucose for Cj biomass. (Net process: 3 ATP, 2 NADPH2, 4 Fd+2/C02 assimilated.) saturation points, sensitivity to oxygen concentration, rapid photorespiration, and high C02 compensation points (about 50 ppm). The C02 compensation point is the C02 concentration in the surrounding environment below which more C02 is respired by the plant than is photosynthetically fixed. Typical C3

biomass species are alfalfa, barley, Chlorella, cotton, Eucalyptus, Euphorbia lathyris, oats, peas, potato, rice, soybean, spinach, sugar beet, sunflower, tall fescue, tobacco, and wheat.

The second pathway is called the C4 cycle because C02 is initially converted to the four-carbon dicarboxylic acids, malic or aspartic acids (Fig. 3.3). Phos — phoenolpyruvic acid (I) reacts with one molecule of C02 to form oxaloacetic acid (II) in the mesophyll of the biomass, and then malic or aspartic acid (III) is formed. The C4 acid is transported to the bundle sheath cells, where decarboxylation occurs to regenerate pyruvic acid (IV), which is returned to

C02H

C=0

I

CH3 IV. ,

NADP

co2

FIGURE 3.3 Biochemical pathway from carbon dioxide to glucose for C4 biomass. (Net process: 5 ATP, 2 NADPH2, 4 Fd+2/C02 assimilated.)

the mesophyll cells to initiate another cycle. The C02 liberated in the bundle sheath cells enters the C3 cycle in the usual manner. Thus, no net C02 is fixed in the portion of the C4 cycle shown in Fig. 3.3, and it is the combination with the C3 cycle which ultimately results in C02 fixation.

The subtle differences between the C4 and C3 cycles are believed responsible for the wide variations in biomass properties. In contrast to C3 biomass, C4 biomass is usually produced at higher yields and has higher rates of photosyn­thesis, high light saturation points, insensitivity to atmospheric oxygen concen­trations below 21 mol %, low levels of respiration, low C02 compensation points, and greater efficiency of water usage. C4 biomass often occurs in areas of high insolation, hot daytime temperatures, and seasonal dry periods. Typical C4 biomass includes important crops such as com, sugarcane, and sorghum, and forage species and tropical grasses such as Bermuda grass. Even crabgrass is a C4 biomass. At least 100 genera in 10 plant families are known to exhibit the C4 cycle.

The third pathway is called crassulacean acid metabolism, or CAM. CAM refers to the capacity of chloroplast-containing biomass tissues to fix C02 via phosphoenolpyruvate carboxylase in dark reactions leading to the synthesis of free malic acid. The mechanism involves the /З-carboxylation of phosphoe — nolpyruvic acid by this enzyme and the subsequent reduction of oxaloacetic acid by malate dehydrogenase. CAM has been documented in at least 18 families, including the family Crassulaceae, and at least 109 genera of the Angeospermae. Biomass species in the CAM category are typically adapted to arid environments, have low photosynthesis rates, and have high water usage efficiencies. Examples are cactus plants and the succulents, such as pineapple. The information developed to date on CAM biomass indicates that CAM has evolved so that the initial C02 fixation can take place in the dark with much less water loss than the C3 and C4 pathways. CAM biomass also conserves carbon by recycling endogenously formed C02. Several CAM species show temperature optima in the range 12 to 17°C for C02 fixation in the dark. The stomates in CAM plants open at night to allow entry of C02 and then close by day to minimize water loss. The carboxylic acids formed in the dark are converted to sugars when the radiant energy is available during the day. Relatively few CAM plants have been exploited commercially.

I. INTRODUCTION

Up to the mid-1990s, only a few commercial virgin biomass energy systems in which dedicated biomass is grown for use as an energy resource were in operation in industrialized countries. The technology is available or under development and is slowly being incorporated into regional, national, and world energy markets. More rapid deployment awaits the inevitable effects on renewable energy usage of fossil fuel depletion and environmental issues. Most of the contribution of biomass to primary energy demand in the 1990s comes from waste biomass. Waste biomass is energy-containing materials that are discarded or disposed of and that are mainly derived from or have their origin in virgin biomass. They are lower in cost than virgin biomass and often have negative costs. Some are quite abundant, and some can be disposed of in a manner that provides economic benefits to reduce disposal costs. Waste bio­mass is generated by anthropological activities and some natural events. It includes municipal solid waste (urban refuse); municipal biosolids (sewage); wood wastes and related residues produced in the forests and logging and forestry operations; agricultural wastes such as animal manures and crop resi­dues produced in farming, ranching, and related operations; and the wastes produced by certain industries such as the pulp and paper industry and those involved with processing foodstuffs. In this chapter, the production of these wastes and their energy potential and availabilities are addressed. The United States is used as the model country because U. S. data are available on most waste biomass to illustrate its role in the development of biomass energy on a national basis. But the conclusions regarding waste biomass as an energy resource in the United States are generally similar for other industrialized coun­tries.

A theoretical model of the combustion of biomass is illustrated by the complete oxidation of giant brown kelp. Note that kelp, for which complete analytical data were available, is used here simply to illustrate the utility of the model, which is applicable to all biomass species. Based on the empirical formula derived from the elemental analysis of dry kelp at an assumed molecular weight of 100, the combustion stoichiometry is

C2.6iH4.63N0.i0S0.01O2 23 ash26.7 + 2.762502 —» 2.61C02 + 0.10NO2 + 0.01SO2

+ 2.315H20 + 26.7ash.

The experimentally measured ash content is assumed to be present in the original biomass and to be carried through the process unchanged. This is not strictly true since oxygen is chemically taken up as metal oxides are formed during standard ash determinations. The ash content is calculated as the difference between the weight of the residue after ashing the sample and the original sample weight, so it does not correspond to the actual ash-forming, metallic elements in the original, dry sample. But for purposes of illustrating the stoichiometry of complete combustion, this equation is adequate. The heat evolved by combustion of this particular sample of kelp is 12.39 MJ/kg (296.1 kcal/g-mol) with product water in the liquid state (Chapter 3). Since on the average, air is 20.95 mol % oxygen, the stoichiometric air requirement for complete combustion is 13.19 mol of air per mole of kelp, or an air-to — kelp mass ratio of 3.805. The ultimate concentration of C02 in the dry flue gas is 19.85 mol %.

Except for submerged combustion processes that are used for treatment of aqueous dissolved and suspended biosolids and a few other special combustion processes, the combustion of virgin and waste biomass involves solid fuels. Stoichiometric combustion data for four types of biomass, two coals, and one coke are compared in Table 7.2. Each of the biomass fuels is assumed to contain 15.0 wt % moisture. The stoichiometric air requirements are considerably less for biomass than for coals and cokes. The reason for this is that the C-to-H mass ratios of biomass are much less than those of fossil fuels. Also, most of the carbon in biomass is, effectively, already partially oxidized. Less oxygen is needed for complete oxidation. For the data in Table 7.2, it is assumed that organic nitrogen and sulfur in each solid fuel are oxidized to N02 and S02 and that nitrogen in air is inert. The calculated amounts of N02 and S02 formed on complete combustion are more than might be expected for a biomass fuel. The relatively high concentrations of organic nitrogen and sulfur in each biomass sample, except the pine wood sample, could potentially cause air pollution problems that require NOx and SOx removal from the combustion products before the flue gases are exhausted to the atmosphere. This will be discussed later. It is sufficient to state here that agricultural and forestry residues, wood chips, bagasse generated in sugarcane plantations, MSW, and RDF have been used as fuels for combustion systems for many years.

A comparison of the heating values and compositions of the raw product gases from selected coal gasification processes is shown in Table 9.4. Some of these processes, a few of which are used for synthesis gas production, have been commercialized. Some have been developed to the point where they might be termed near-commercial, and a few are under development. It is evident that a wide range of gas compositions can be produced by coal gasification. It is also evident that several of the gas compositions and operating conditions can be correlated with thermodynamic principles and the thermodynamics of the carbon-oxygen-steam system. Methane and carbon dioxide yields are generally higher at lower temperatures and higher pressures, as illustrated by the raw gas compositions for the Synthane process, whereas higher temperatures and lower pressures favor carbon monoxide and hydrogen, as illustrated by the raw gas compositions reported for the Koppers-Totzek process. Interestingly, the heating values of the product gases for processes supplied with steam — oxygen coreactants are generally in the same range despite the wide range of operating conditions. The heating values of the product gases from processes supplied with steam-air coreactants are also in the same range, although they are lower than those of the product gases produced by coal-steam-oxygen processes. Arithmetic adjustment of the heating values by deducting nitrogen from the product gases shows that all of them are in the same range.

Biodiesel Fuels

Natural glycerides have been investigated as alternative fuels for the compression-ignition engine since the late 1800s. Rudolph Diesel, the inventor of the engine that bears his name, demonstrated a diesel-cycle engine fueled with peanut oil in 1900 at the Paris Exposition. In 1912, Diesel wrote: “The use of vegetable oils may become in the course of time as important as petro­leum and the coal tar products of the present time.” This obviously has not happened, since the majority of heavy-duty farm and construction vehicles and large trucks are powered by diesel engines that operate on petroleum diesel fuels. Of the two basic types of diesel engines, direct injection, in which the diesel fuel is injected directly into the combustion chamber, and indirect injection, in which the fuel is injected into a precombustion chamber, the direct injection design is more commonly used for the larger vehicles. The direct-injected engines are easier to start and less expensive than the more elaborate but quiet indirect-injected engine (Quick, 1989).

Many natural glycerides can be used with little or no difficulty as diesel fuels for vehicles equipped with indirect-injection engines. Several problems arise when they are used to fuel direct-injection engines. One of the problems is caused by the higher viscosity and lower volatility of natural glycerides as compared with the corresponding properties of petroleum diesel fuels (cf. Krawczyk, 1996). A comparison of selected properties of No. 2 diesel fuel, soybean and rapeseed oils, their corresponding transesterification products with methanol and ethanol, and the fatty acid makeup of the glyceride oils and esters is shown in Table 10.1.

Performance of the glycerides as a diesel fuel is improved by conversion of the fatty acid moieties of the glycerides to the corresponding methyl or ethyl esters. Fouling problems are significantly reduced, and the viscosities, pour points, and combustion characteristics of the esters in blends with diesel fuel or as neat liquids are superior to those of the natural glycerides. The cetane numbers of the methyl and ethyl esters are about 50 to 65, and they have lower ash, sulfur, and volatilities and higher flash points than conventional diesel fuels. The cetane numbers of the methyl and ethyl esters of the pure fatty acids in the oils have been correlated with the chain length and degree of saturation (Freedman et ah, 1990; Clements, 1996; Knothe, Bagby, and Ryan, 1996). For the ethyl esters of the C18 fatty acids, stearic, oleic, linoleic, and linolenic acids, the reported cetane numbers are 77, 54, 37, and 27, respectively. The corresponding cetane numbers determined for the free acids

TABLE 10.1 Comparison of Some Typical Properties of Diesel, Soybean Oil, Rapeseed Oil, and Ester Fuels11 Property No. 2 diesel Soybean oil Rapeseed oil Oil Methyl ester Ethyl ester Oil Methyl ester Ethyl ester Specific gravity 0.8495 0.92 0.886 0.881 0.91 0.880 0.876 Viscosity at 40°C, mm2/s 2.98 33 3.891 4.493 51 5.65 6.17 Cloud point, °С -12 -4 3 0 0 -2 Pour point, °С -23 -12 -3 -3 -21 -15 -10 Flash point, °С 74 188 171 179 124 Boiling point, °С 191 339 357 347 273 Water & sediment, vol % <0.005 <0.005 <0.005 <0.005 <0.005 Carbon residue, wt % 0.16 0.068 0.071 0.08 0.06 Ash, wt % 0.002 0 0 0.002 0.002 Sulfur, wt % 0.036 0.01 0.012 0.008 0.01 0.012 0.014 Cetane number 49 38 55 53 32 62 65 Copper corrosion 1A 1A 1A 1A 1A Higher heating value, MJ/kg 45.42 39.3 39.77 39.96 40.17 40.54 40.51 MJ/L 38.58 36.2 35.24 35.20 36.60 35.68 38.00 Fatty acid composition, wt % Palmitic (16:0) 9.8 9.9 10.0 1 2.2 2.6 Stearic (18:0) 2.4 3.8 3.8 0.9 0.9 Oleic (18:1) 28.9 19.1 18.9 32 12.6 12.8 Linoleic (18:2) 50.7 55.6 55.7 12.1 11.9 Linolenic (18:3) 6.5 10.2 10.2 15 8 7.7 Eicosenoic (20:1) 0.2 0.2 7.4 7.3 Behenic (22:0) 0.3 0.3 0.7 0.7 Erucic (22:1) 0.0 0.0 50 49.8 49.5 Others 1.7 0.8 0.9 2 6.3 6.6 “Adapted from Cruz, Stanfill, and Powaukee (1996); Peterson et al. (1995); Reed (1993); Shay (1993); Stumborg et al. (1993); Auld, Peterson, and Korns (1989). The figures in parentheses after each fatty acid denote the number of carbon atoms and double bonds in the fatty acid. The figures for “Oil” are

are 62,46,31, and 20. These trends would be expected because of the increasing degree of unsaturation from stearic to linolenic acid and the fact that more paraffinic hydrocarbons usually have higher cetane numbers. Similar correla­tions were found for other esters.

Since the methanol and ethanol transesterification products of a variety of commercial biomass-derived oils, such as soybean, rapeseed, peanut, palm, and sunflower oils, and even waste animal fats have generally been found to be suitable as additives or neat fuels for diesel-powered engines without engine modifications, programs were started, first in several European countries and then the United States, to develop and market what has been termed “biodiesel.” Biodiesel is defined as monoalkyl esters of long-chain fatty acids derived from renewable feedstocks, such as vegetable oils and animal fats, for use in compres­sion ignition engines. The emphasis in European biodiesel programs has been on the methyl ester of rapeseed oil, and in the United States, the methyl esters of soybean and rapeseed oils have received the most attention (с/. Krawczyk, 1996). In Europe, the most advanced program is in Austria. Pilot plants, small — scale farm cooperatives, and industrial-scale plants (10,000 to 30,000 t/year capacities) have been built, and fuel standards for rapeseed oil methyl ester have been adopted. In Malaysia, the methyl ester of palm oil is being developed as biodiesel, and in Nicaragua, biodiesel applications of the oil from Jatropha curcas, a large shrub or tree native to the American tropics, are under investiga­tion (с/. Jones and Miller, 1991).

Many projects are underway in the United States to complete the U. S. database on the properties and performance of biodiesel. Included in this work are studies on the emissions and performance of engines and vehicles fueled with biodiesel as additives, neat fuels, and fuel blends; the effects of oxygenated additives in biodiesel on cetane number; the effects of interesterification of different vegetable oils followed by transesterification with methanol on bio­diesel properties; and the low-temperature flow properties, long-term storage effects, and toxicides of a variety of methyl and ethyl esters. The information being developed in this work continues to support the marketing of biodiesel. Some engine test results and fields trials with diesel vehicles indicate that a few issues must be addressed before biodiesel can be successfully commercial­ized. In order of importance and frequency of occurrence, these issues concern fuel quality, fuel filter plugging, injector failure, material compatibility, and fuel economy (Schumacher and Van Gerpen, 1996; Schumacher, Howell, and Weber, 1996). It appears that the fledgling biodiesel industry is approaching a stage similar to that of the petroleum industry many years ago, when detailed motor fuel specifications and test procedures were needed to facilitate fuel compatibility and interchangability from different suppliers across the country. The tentative specifications being developed for biodiesel by the American

“Schumacher, Howell, and Weber (1996); Weber and Johannes (1996). These are tenative specifi­cations for 100% biodiesel and are being evaluated by the American Society for Testing and Ma­terials as of March 4, 1996. A considerable amount of experience exists in the United States with blends of 20% biodiesel and 80% diesel fuels. Although biodiesel can be used in this form, the use of blends over 20% biodiesel should be evaluated on a case-by-case basis until further expe­rience is available. The GC method for glycerol is the Austrian update (Christiana Plane) of the U. S. Department of Agriculture method.

Society for Testing and Materials (ASTM) are shown in Table 10.2. After fine — tuning and approval, the ASTM standards are expected to eliminate many of the potential problems that can result from the large number of vegetable oils and animal fats that can be used to produce biodiesel. In other words, when a particular manufacturing source supplies biodiesel that meets the ASTM standards, it is assumed it will be fully compatible with all other sources of biodiesel and exhibit the same end-use characteristics even though the feed­stock may not be the same at each manufacturing site.

Market projection studies indicate that biodiesel which meets the ASTM standards will be manufactured and distributed initially for niche market applications such as government bus and truck fleets where the mandates of the Clean Air Act of 1990 require the use of cleaner burning alternative fuels. A major difficulty regarding large-scale marketing of biodiesel is consumer cost. This subject will be examined later in Section IV, but is not expected to be an insurmountable barrier in the long term.

Potential Sources of Biodiesel

Various oilseed crops afford natural triglyceride esters several of which are under development as diesel fuels as already mentioned. Some biomass species produce natural esters of fatty acids and fatty alcohols, such as the desert shrub jojoba (Simmondsia chinensis [Link] Schneider) which yields esterification products of C20-C22 straight-chain, monounsaturated fatty alcohols and acids (National Academy of Sciences, 1977). These esters have not been considered as a source of biodiesel, but could be readily converted to normal paraffinic hydrocarbons. The largest source of feedstock for conversion to biodiesel is the seed oils. Seed oils are largely triglycerides and typically contain three long-chain fatty acids, each bound to one of the carbon atoms of glycerol via an ester linkage. The fatty acid distributions of soybean and rapeseed triglycer­ides in Table 10.1 are typical of most seed oils. An analysis of the average and potential annual yields of oilseeds and seed oils is shown in Table 10.3. With the exception of the Chinese tallow tree (Sapium sebiferum), the range in potential oil yields is about 400 to 1800 L/ha-year (2.5 to 11.3 bbl/ha-year), but average commercial yields are about 30 to 60% of these values. The estimates of potential yields are based on discussions with experts in breeding and agronomy of the individual oilseeds, analyses of potential yields under experimental conditions, and knowledge of crop improvement considerations (Lipinsky et ah, 1984). In addition, the confidence limits associated with potential yields of commercial oilseeds such as corn and soybeans are much narrower than are the estimates for the newer oilseed crops, such as the Chinese tallow tree.

Soybeans are the most widely planted oilseed in the United States and account for more than 50% of the world’s oilseed output. Annual soybean oil production in the United States and worldwide in the mid-1990s was about 7 million and 12 million tonnes, which in petroleum industry terms is about 48 million and 82 million bbl/year, a substantial amount of potential liquid fuel. Worldwide, annual commercial production in the mid-1990s of biomass — based glycerides, which includes nonseed oils such as palm, olive, and coconut oil production of about 14 million tonnes, was about 62 million tonnes. Essentially all production is used for foodstuffs and specialized industrial applications. So although the amounts of biomass glycerides that could poten­tially be converted to biodiesel are significant, it is obvious that large-scale biodiesel manufacture from existing commercial production under ordinary market conditions is problematic. Development of a large-scale biodiesel indus­try in almost any industrialized country would in all likelihood require dedi­cated, incremental production of biomass glycerides over that which is required for foodstuffs and industrial use.

“Adapted from Lipinsky et al. (1984). Growth is under dry-land conditions except for cotton, which is irrigated. The yield for the Chinese tallow tree is one reported yield equivalent to 6270 L/ha of oil plus tallow and is not an average yield from several sources. It is believed that the yield would be substantially less than this in managed dense stands, but still higher than that of conventional oilseed crops.

The Chinese tallow tree, which has been cultivated and naturalized in the South Atlantic and the Gulf States, deserves special consideration because the yields of potential biodiesel feedstocks are much higher than those of conventional oilseed crops, and because the seeds are not currently harvested for commercial processing or foodstuff usage in the United States. The tree is a member of the Euphorbiaceae family and is native to subtropical China, where it has been cultivated for 14 centuries as a specialty oilseed crop, medicinal plant, and source of vegetable dye, and for uses similar to those of linseed oils (Morgan and Schultz, 1981; Scheld, 1986). Plantations can be established from cuttings, seedlings, or seed. The most convenient and econom­ical method of stand establishment is direct planting of seeds by standard, mechanical, row-crop planters. The tree grows well in saline lands which are marginal for conventional agriculture. Dense plantations of 0.60 to 0.75 m spacing are practical, and coppicing is a feasible system of management. The seed pods yield both a hard vegetable tallow on the outside and a liquid oil on the inside, which together comprise 45 to 50 wt % of the seed. Expressed as weight percentages of the seed, the tallow is typically 25 to 30% of the seed, the oil is 15 to 20%, the hull is about 40%, and the remainder is seed protein and fiber. In some strains, the tallow mainly consists of a single triglyceride, the palmitic-oleic-palmitic triglyceride. The oil is composed largely of glycerides of oleic, linoleic, and linolenic acids and is chemically similar to linseed oil and dehydrated castor oil. Both the tallow and oil have good potential as a source of biodiesel. The annual yields of tallow and oil are also remarkably high, each probably near 14.8 bbl/ha-year from a planting with seeds of about 11,200 kg/ ha (10,000 lb/ac).

Certain microalgae represent another source of natural triglycerides. Much research has been done on the culture and compositional characteristics of certain microalgae as a source of “algal oils” (cf. Klass, 1983, 1984, 1985; Brown, 1993). This work has been focused on the growth of the organisms under conditions that can promote glyceride formation. The oils are high in triglycerides and can be transesterified to form biodiesel in the same manner as other natural triglycerides (Hill and Feinberg, 1984). As mentioned in Chapter 4, the production of natural triglycerides from microalgae can some­times eliminate the high cost of cell harvest and extraction because it may be possible to separate the lipids by simple flotation or extraction from the culture media if they are leaked as extracellular products. As already mentioned, stressed growth conditions can often be used to increase the formation of natural lipids as shown for several microalgae in Table 10.4. The stress was caused by either the use of nutrient-deficient media or the addition of excess salt to nutrient-enriched media. The combination of both nutrient deficiency and salt enrichment appears to enhance lipid formation with Isochrysis sp., but to reduce it with Dunaliella salina. Interestingly, the free glycerol content can apparently be quite high for Dunaliella sp. This suggests either that all of the intracellular glycerol was not esterified, or that the esters were hydrolyzed after formation. Botryococcus braunii exhibited relatively high lipid contents under each set of growth conditions, but were the highest, 54.2 dry wt %, under nutrient-deficient growth conditions. Other microalgae have also been found to exhibit similar lipid contents. For example, the lipid content increased from 28 to over 50% of the cell dry weight with increasing nitrogen deficiency for Nannochloropsis sp. when grown in media under nitrogen-limited condi­tions (Tillett and Benemann, 1988). Lipid productivity was demonstrated to have a maximum of 150 mg/L-day at 5 to 6% cell nitrogen and was apparently independent of the light supply and cell density when the initial nitrogen concentration exceeded 25 mg/L. The key to the practical use of triglycerides from microalgae as feedstock for biodiesel production is the rate of triglyceride formation during the growth of microalgae. The yield on conversion of carbon

“Adapted from Tables 1-9 of National Renewable Energy Laboratory (1983). The culture media are denoted by FW, freshwater; NE, nutrient enriched; ND, nutrient deficient. The analyses were conducted on five different cultures of each species; statistical error analysis showed no standard error of more than 10%.

dioxide to microalgae is important, but high yields are not essential to obtain high triglyceride productivities. They are also a function of the growth rate of the particular microalgae and its triglyceride content. Commercial net produc­tivities of triglycerides from microalgae per unit growth area were projected to be 23 g/m2-day or 66 t/ha-year several years ago, and improvements through continued research were estimated to raise these figures to 43 g/m2-day or 124 t/ha-year (Hill and Feinberg, 1984). Although many species of microalgae have since been isolated and characterized, and small outdoor test facilities have been operated to collect data on optimal growth conditions, the sustained production of microalgae and the harvesting of the triglycerides in integrated, large-scale systems have not yet been demonstrated (Brown, 1993). Very high cell densities of the order of 46 million/ml have been achieved in a shallow, outdoor, 50-m2 raceway in Hawaii for the marine diatom Phaeodactylum tricor — nutum, the lipid content of which was found to be as high as 80% of its dry weight (Glenn, 1982; National Renewable Energy Laboratory, 1982; Shupe, 1982). This and similar projects on the growth of microalgae in fresh and saline water in California and the Southwest have since been terminated. Methods of maintaining optimal levels of nutrients, carbon dioxide, salinity, and temperature for continuous, outdoor microalgal growth, as well as eco­nomic methods of recovering triglycerides, must be developed and tested to allow design of large-scale systems.

Research on aquatic biomass by French researchers has resulted in several interesting results with the microalga B. braunii. In laboratory studies, this microalga, which under the growth conditions used is reported to have a hydrocarbon content as high as 75% of its dry weight, was reported to have been cultured at hydrocarbon productivities per unit growth area up to 15 g/ m2-day (Casadevall and Largeau, 1982). The dominant hydrocarbons produced are branched-chain olefins. In subsequent work, the French confirmed the high hydrocarbon content of B. braunii (Casadevall, 1984; Brenckmann, 1985; Brenckmann et al., 1985; Metzger et ah, 1985). Nitrogen limitation was not found to be necessary for high hydrocarbon production, the highest productivi­ties of which were observed during exponential growth. Light intensity did not affect the structure of the hydrocarbons, mainly C27, C29, and C3i alkadienes, which were produced in continuously illuminated batch cultures. But adjust­ment of the light intensity to the proper level gave maximum biomass and hydrocarbon yields. One strain was reported to yield straight-chain alkadienes and trienes as odd-numbered chains from C23 to C31, and another strain pro­duces triterpenes of the generic formula CnH2n_10 where n varies from 30 to 37.

The French results with B. braunii contrast sharply with the U. S. results and are unexpected because most of the literature on this organism describes lipids and not hydrocarbons as metabolites. The formation of lipids and hydro­carbons together for certain biomass species is not unique, as indicated in this chapter. It is surprising, however, that an organism is reported to produce higher lipid content under stressed growth conditions than that produced under nonstressed growth conditions with no mention of hydrocarbons, and high hydrocarbon production under nonstressed growth conditions with no mention of lipids. It may be that the unknown organic components in B, braunii (Table 10.4) are hydrocarbons, but it is unlikely. It is more probable that a major shift in biochemical metabolite formation can occur from one strain to another. In any case, it appears that certain microalgae are capable of high productivities of lipids or hydrocarbons. Such organisms could provide significant benefits in processes designed for the manufacture of liquid fuels from biomass.