The chemistry of coal gasification is usually depicted to involve the following reactions of carbon, oxygen, and steam (cf. Bodle and Schora, 1979). The standard enthalpy change (gram molecules) at 298 К is shown for each reaction.

(9) C02 + 4H2 -» CH4 + H20 — 165.0 kj

(10) C + 2H2 -* CH4 — 74.8 kj.

In theory, gasification processes can be designed so that the exothermic and endothermic reactions are thermally balanced. For example, consider reactions 2 and 5. The feed rates could be controlled so that the heat released balances the heat requirement. In this hypothetical case, the amount of oxygen required is 0.27 mol/mol of carbon, the amount of steam required is 0.45 mol/ mol of carbon, and the oxygen-to-steam molar ratio is 0.6:

C 4- H20 —* CO + H2 4- 131.3 kj 1.2C 4- 0.6O2 -* 1.2CO — 131.3 kj Net: 2.2C + H20 4- 0.6O2 -» 2.2CO 4- H2.

Many reactions occur simultaneously in coal gasification systems and it is not possible to control the process precisely as indicated here. But by careful selection of temperature, pressure, reactant and recycle product feed rates, reaction times, and oxygen-steam ratios, it is often possible to maximize certain desired products. When high-energy fuel gas is the desired product, selective utilization of high pressure, low temperature, and recycled hydrogen can result in practically all of the net fuel gas production in the form of methane.

The oxygen-steam ratios required to maintain zero net enthalpy change are given in Table 9.3 for several temperatures and pressures (Parent and Katz, 1948). With increased pressure, the ratio necessary to preserve a zero net enthalpy change diminishes. The decrease is most pronounced at low pressures. The effect of temperature change at constant pressure is also shown in Table 9.3. At lower temperatures, the oxygen-steam ratio doubles for each temperature

interval of 100 K. At higher temperatures, the increase diminishes and finally becomes very small.

The thermodynamic equilibrium compositions and enthalpy changes for the carbon-oxygen-steam system are graphically illustrated at several represen­tative temperatures and pressures in Figs. 9.1 to 9.4 (Parent and Katz, 1948). Increasing pressures tend to lower the equilibrium concentrations of hydrogen and carbon monoxide and increase the methane and carbon dioxide concentra­tions (Fig. 9.1). Methane and carbon dioxide formation are favored at lower temperatures, and at higher temperatures, carbon monoxide and hydrogen are the dominant equilibrium products (Figs. 9.2 and 9.3). At high temperatures, the reactions occurring in the system are thermodynamically equivalent to reactions 2 and 5. It is also apparent that hydrogen-to-carbon monoxide molar ratios of 1.0 or more are thermodynamically feasible at lower feed ratios of oxygen to steam and low pressure (Fig. 9.4).

Although the utility of thermodynamic data to optimize the operating condi­tions of a gasification process is of considerable importance, thermodynamics ignore kinetic and catalytic effects and the mechanisms by which processes occur. The data presented here, however, provide valuable guidelines for the design of gasification processes. For coal gasification, the type of coal and reactant contact conditions in the gasifier produce large differences in the raw product gas compositions. In general, the same principles and conclusions apply to biomass gasification. Where experimental conditions are favorable, equilibrium may be approached by prolonged contact of the reactants or by use of catalysts. Where neither of these conditions offers a convenient solution, a compromise between idealized equilibrium and kinetics is necessary.

Modern, combined-cycle electric power generation systems using gas turbines as the primary generators offer higher thermal efficiencies than conventional steam-turbine systems. Many of the commercial plants in operation today use natural gas-fired, combined-cycle systems in which the hot exhaust from the gas turbines is processed in heat recovery steam generators to afford steam for injection into steam turbines for additional power generation and improved efficiency. Steam injection into the gas turbines along with combustion gases adds further efficiency improvements. Overall thermal efficiencies to electric power are up to twice those of conventional fuel-fired steam turbine systems. Availabilities can be high, the environmental characteristics are excellent, and capital costs are considerably less per unit of electric power capacity compared to the costs of conventional coal-fired plants. One of the largest combined cycle, natural gas-fired plants in the world—a 2000-MW central station plant in Japan—operates at 95% availabilities (adjusted for mandated inspections).

Integration of coal gasification processes with combined-cycle technologies has opened the way for coal to fuel similar power generation plants at high efficiencies. Integrated gasification-combined cycle (IGCC) systems are in oper­ation throughout the world and have made it possible to resurrect the use of low-cost, high-sulfur fossil fuels for power generation because the gasification process is, in effect, a desulfurization process. Oxygen-blown gasification plants have dominated both commercial and demonstration coal gasification units since as far back as the 1920s. Seventeen commercial plants, having a total of 153 coal gasifiers, are reported to be in commercial operation worldwide (с/. Simbeck and Karp in Swanekamp, 1996). Oxygen is used rather than air in these plants because they produce synthesis gas-based chemicals and premium fuels. In the United States, modern air — and oxygen-blown, fluid-bed gasifica­tion processes equipped with hot-gas cleanup systems are being perfected for use with coal feedstocks in IGCC plants. These plants are expected to have good emissions characteristics with one exception—carbon dioxide emissions per unit of fuel will be about the same as those of conventional fossil-fueled power plants. Biomass fuels, because of their relatively short recycling time, would avoid this problem.

It is apparent from the discussion of biomass gasification in this chapter that innovative processes for producing low — and medium-energy fuel gases have been developed for virgin and waste biomass feedstocks and are either about to be or have already been commercialized. These technologies are much improved over conventional, air-blown gasification processes. The availability of suitable fuel gases from modem biomass gasification processes facilitates their coupling with combined cycle power plants in the same manner as fossil — fueled IGCC plants. Biomass-fueled IGCC plants (BIGCC), particularly those having smaller capacities and those used for combined waste disposal and energy recovery, are expected to contribute to the expected 600 GW of new electric generating capacity needed worldwide over the next several years. IGCC plants fueled with both coal and biomass as sequential or combined feedstocks would appear to be a viable alternative because, as already pointed out, some gasification processes are capable of converting both feedstocks. The heat load for conventional Rankine steam-cycle power production using boilers and steam turbines is about 14.8 to 16.9 MJ/kWh; BIGCC technology should have about 25% less heat load and therefore considerably improved economics. The economics of BIGCC systems even as small as 1 to 10 MW in capacity can be very site-specific, but appear to be capable of reasonable rates of return (Craig and Purvis, 1995). Larger biomass integrated-gasification/ steam-injected gas-turbine (BIG/STIG) cogeneration plants are projected to be

attractive investments for sugar producers, for example, who can use sugarcane bagasse as fuel (Larson et al, 1991).

A 30-MW power plant fueled with eucalyptus wood from short-rotation energy plantations is planned in Brazil to demonstrate BIGCC technology (Carpentieri, 1993; de Queiroz and do Nascimento, 1993). This plant is pro­jected to operate at an availability of 80% and an overall thermal efficiency of 43% to produce 210 GWh/year from 205,835 m3/year of wood chips contain­ing 35 wt % moisture. The energy cost is estimated to be $0,045 to $0.065/kWh. The first plant is estimated to have a capital cost of $60 million to $75 million (U. S.); subsequent plants are estimated to cost $39 million to $45 million (U. S.). For sugarcane bagasse, which will be tested as a potential feedstock, the heat rate is estimated to be 8.368 GJ/MWh with fuel consumption at 50 wt % moisture content of 1.021 kg/kWh. It is estimated that the cost of electric power production in 53-MW BIG/STIG plants in Brazil using briquetted sugarcane bagasse is $0,032 to $0.058/kWh (Larson et al., 1991).

Other advanced technologies that are receiving considerable attention in­clude improved designs for combining small biomass gasifiers with motor — generator sets or gas turbines in the multiple-kilowatt range and in the 1 to 5 MW range. Numerous configurations are being developed, although some assessments have ruled out conventional steam turbines because of their rela­tively low efficiency and high cost at small sizes. Examples of small systems under development include a 1-MW system consisting of a fixed-bed, down­draft gasifier, a gas cleaning system, and a spark-ignited gas engine-generator set; and a 1-MW system consisting of a pressurized fluid-bed gasifier, a hot- gas cleanup system, and a gas turbine (Purvis et al, 1996).

Another advanced technology that can use biomass gasification for power generation employs fuel cells. Fuel cells are devices that electrochemically convert the chemical energy contained in the fuel into direct current electricity and the oxidation products of the fuel. The fuels can be natural gas and the product gases from the gasification of solid fuels, including biomass and de­rived fuels such as hydrogen, and intermediate liquid fuels such as hydrocarbons and ethanol. In one sense, fuel cells are similar to electric batter­ies, but the tuei and oxidant are continuously supplied Irom external sources. So, unlike batteries, fuel cells are not consumed or depleted in the process. Also, because fuel cells are not heat engines, they are not Carnot limited and can achieve high fuel-energy-to-electric power conversion efficiencies that can be above 60% based on the energy content of the fuel supplied to the fuel cell. Among the fuel cell configurations, three different types are being developed for power generation by units 100 kW to 25 MW in capacity. They are differenti­ated by the electrolytes used within the cell—phosphoric acid, molten carbon­ate, and solid oxide. Some designs such as those that use molten carbonate and solid oxide electrolytes are operated at sufficiently elevated temperatures to be suitable for use in cogeneration applications. A few of these designs are believed to be operable at overall efficiencies as high as 85% based on the energy content of the fuel supplied to the fuel cell. A few small-scale power units using biomass fuels for specialty applications may become available in the next few years, but large-scale fuel-cell power plants are not expected to be available for generating central station power until well into the twenty- first century.

Typical organic components in representative, mature biomass species are shown in Table 3.9 along with the corresponding ash contents. With few exceptions, the order of abundance of the major organic components in whole — plant samples of terrestrial biomass is celluloses, hemicelluloses, lignins, and proteins. Aquatic biomass does not appear to follow this trend. The cellulosic components are often much lower in concentration than the hemicelluloses as illustrated by the data for water hyacinth. Other carbohydrates and deriva­tives are dominant in species such as giant brown kelp to almost complete exclusion of the celluloses. The hemicelluloses and lignins have not been found in this species.

Biomass often undergoes compositional changes, some of which can be subtle or pronounced, during growth and sometimes after harvesting depend­ing on age of the biomass and environmental factors. An example of this phenomenon is the gradual decrease in sugar content and the gradual increase in hydrocarbon content during the maturation of E. lathyris (Ayerbe et ai, 1984). Another phenomenon that is quite common during biomass growth is the nonuniform distribution of organic components in various plant parts. For example, the hydrocarbon content in the leaves of E. lathyris is more than twice the amount in the stems (Sachs et al, 1981). All of these factors must be considered in some detail when biomass is utilized for production of cer­tain organic compounds or as a feedstock for conversion to fuels and energy products.

Alpha cellulose or cellulose as it is more generally known, is the chief structural element and a major constituent of many biomass species. In trees, cellulose is generally about 40 to 50% of the dry weight. As a general rule, the major organic components on a moisture and ash-free basis in woody biomass are about 50 wt % cellulosics, 25 wt % hemicelluloses, and 25 wt % lignins. However, cellulose is not always the dominant component in the carbohydrate fraction of biomass. As just mentioned and as shown in Table 3.9, it is one of the minor components in giant brown kelp. Mannitol, a hexahydric alcohol that can be formed by reduction of the aldehyde group of D-glucose to a methylol group, and alginic acid, a polymer of mannuronic and glucuronic acids, are the major carbohydrates.

The lipid and protein fractions of plant biomass are normally much less on a percentage basis than the carbohydrate components. The lipids are usually present at the lowest concentration, while the protein fraction is somewhat

higher, but still lower than the carbohydrate fraction. Crude protein values can be approximated by multiplying the organic nitrogen analyses by 6.25. This factor is used because the average weight percentage of nitrogen in pure dry protein is about 16%, although the protein content of each biomass species can best be determined by amino acid assay. The calculated crude protein values of the dry biomass species in Table 3.9 range from a low of about 0 wt % for pine wood to a high of about 30 wt % for Kentucky bluegrass. For grasses, the protein content is strongly dependent on the growing procedures used before harvest, particularly the fertilization methods. Some biomass spe­cies such as the legumes, however, fix nitrogen from the ambient atmosphere and often contain high protein concentrations.

It is apparent from the data in Table 3.5 that the sulfur content of virgin and waste biomass ranges from very low to about 1 wt % for primary biosolids. This sulfur level is similar to the sulfur content of high-sulfur Illinois bitumi­nous coal. Woody biomass generally contains very little sulfur.

The national inventory of forestry residues generated in the United States in the mid-1970s is one of the first comprehensive assessments to be done on a county-by-county basis for a large country (Stanford Research Institute, 1976). Data were compiled on residues from logging operations and on wood and bark residues produced at primary wood processing milk. The regional offices of the U. S. Forest Service supplied most of the data on the residues generated by the mills. Most of the data on logging residues were obtained by applying residue factors to industrial roundwood production figures for each country. The factors used for different regions of the country were obtained from published reports of actual logging residue studies conducted in the field. For cases in which specific county data were not provided, it was necessary to apportion multicounty regional data or total state data among the counties having primary wood processing mills. Detailed data on bark residues were sometimes limited. The results of this survey indicated that a total of 105 million dry t/year of forestry residues are generated in the contiguous United States. The percentage distribution of forestry residues by region was 33.2% in the Pacific area, 6.4% in the Mountain area, 13.0% in the West South Central area, 14.2% in the East South Central area, 19.8% in the South Atlantic area, 2.2% in the West North Central area, 5.2% in the East North Central area, and 6.1% in the New England and Mid-Atlantic area. The Southeast and the Northwest are the areas that produce most of the U. S. timber and hence most of the forestry residues. Of the total forestry residues, 35.4 million dry t/year or 33.7% consisted of logging slash, and 69.6 million dry t/year or 66.3% consisted of bark and wood mill residues.

In another inventory of various sources of wood wastes in the United States (McKeever, 1995), it was found that for 1991, 26.0 million dry tonnes of bark and 74.5 million dry tonnes of wood residues were generated at primary lumber processing mills. Only 5% of the bark and 6% of the wood residue were wasted and not used. The projection for 1993 based on these findings was that 5.7 million dry tonnes of bark and wood residues were available for recov­ery and use as an energy resource. The total of 100.5 million dry t/year gener­ated at the mills is about 44% higher than the mill residues found in the Stanford assessment.

Energy Potential

Assuming the results of the Stanford assessment of logging slash residues in the continental United States would provide approximately the same results today, the energy potential of 35.4 million dry t/year of slash is about 0.66 EJ/year unadjusted for availability. Similarly, 100.5 million dry t/year of mill residue is about 1.86 EJ/year.

Availability

Of the forestry residues collected during normal operations in the United States in the mid-1970s according to the Stanford assessment, most of it was wood and bark residues that accumulated at primary lumber mills. It was reported that of the total forestry residues generated, about 33% was sold, 16% was used as fuel, and 51% was wasted. Most of the wasted residue was generated by logging operations. The fuel usage was represented to be an energy contribu­tion of 0.32 to 0.42 EJ/year to primary energy demand. The lumber and wood products industry has consistently used about 0.35 to 0.50 EJ/year of forestry wastes as fuel for the past few decades (Klass, 1990) and is expected to continue to use them at about the same rate up to at least 2040 (Skog, 1993). If adjustments are made to the total energy potential of U. S. forest residues for the amounts not used but available as a waste biomass resource, most of the logging slash is available (0.66 EJ/year) and 5.7 million dry t/year of available, unused mill residues correspond to about 0.11 EJ/year, or a total from available forestry residues of 0.77 EJ/year. Since the mass of logging residue varies from about 25 to 45% of the wood cut (Howlett and Gamache, 1977), it is highly probable that an energy availability estimate of only 0.66 EJ/year from logging residues is too low. A detailed assessment of the recoverable energy potential of forestry residues in only one state, Georgia, is estimated to be 0.20 EJ/year for logging slash and 0.08 EJ/year for wood manufacturing residues, or a total of 0.28 EJ/year (Riall and Bouffier, 1990). These data suggest that a more realistic estimate of the available energy value from logging slash should be at least 1.0 EJ/year.

Perfect combustion of biomass and most other fuels would help minimize the emissions of particulate matter, and gaseous and toxic compounds. In practice, this can be difficult to attain, even with natural gas, the dominant component of which is methane, the simplest of all organic fuels (Chisholm and Klass, 1966). The carbon monoxide content of the flue gas is a good indicator of the completeness of combustion. The less carbon monoxide, the more complete the combustion process and the lower the emissions of organic compounds. As already indicated, a wide range of variables and independent parameters affect how closely perfect combustion can be approached for solid biomass fuels: particle size range and distribution, moisture content, composition, and heating value; air-fuel ratios, mixing, and reaction time; temperature and sometimes pressure range in the combustion chamber; and furnace design and heat-transfer methods. A compromise is often made between the operating parameters that promote complete combustion and those that minimize the emissions of inorganic derivatives. Under the best of conditions, a solid biomass fuel should consist of small, uniform particles, be low in moisture and ash contents, and have zero to very low chlorine, nitrogen, and sulfur contents. With the proper processing, virgin biomass such as wood can approximate these characteristics, whereas municipal solid waste is representative of biomass fuels that might be the furthest removed from them.

Emissions from biomass-fueled boilers can be controlled by a variety of methods. The control systems needed depend mainly on the composition of the feedstock. First, good combustion control is essential to maximize combustion and to minimize emissions of unburned hydrocarbons and carbon monoxide. Efficient removal of particulate matter in the flue gases can be achieved by various combinations of cyclonic separation, electrostatic precipi­tation, agglomeration, and filtration. Removal of acid gas emissions can be achieved by flue gas scrubbing and treatment with lime. There are several approaches to the control of NOx emissions (Clearwater and Hill, 1991). Combustion control techniques include use of staged combustion, low excess air, and flue gas recirculation. Staged combustion involves reduction of the maximum attainable flame temperature and the control of residence time. In the primary stage of combustion, the maximum flame temperature and thermal NOx formation are reduced by the transfer of heat, which is not returned, or by combustion with substoichiometric amounts of air. The formation of NO from chemically bound nitrogen is also largely avoided in the primary stage under these conditions if the residence time is sufficient to permit nitrogen to form. Combustion is then completed in the second stage with excess second­ary air at short residence times to minimize NO formation. Potential add-on controls include selective noncatalytic and catalytic reduction and natural gas reburning techniques. Selective noncatalytic reduction, such as ammonia injection, is one of the preferred methods because it has been effectively employed in several MSW combustion plants and has been shown to afford NOx emissions in the range of 120 to 200 ppmv. As discussed at the end of this section, staged combustion can be carried out without actually using hardware designed for staged combustion. The technique is quite effective for minimizing NOx in the Burlington, Vermont power plant.

Toxic polychlorinated dibenzo-p-dioxins (“dioxins” or PCDDs) and poly­chlorinated dibenzofurans (“furans” or PCDFs) can form on combustion of chlorine-containing biomass and be emitted in the flue gases and possi­bly in adsorbed form on flyash or in particulate matter. The isomer 2,3,7,8- tetrachlorodibenzo-p-dioxin, a strong carcinogen and a contaminant found in the defoliant Agent Orange used in the Vietnam War, is claimed to be the most lethally toxic of man-made chemicals when administered to guinea pigs (Esposito, Tiernan, and Dryden, 1980). The compound is destroyed when the temperature is 1000°C for as little as a millisecond or at lower temperatures for longer periods (Barnes, 1983). At stack gas temperatures below 200°C., the PCDDs and PCDFs are predominantly found on the particulate matter. These data suggest that control of combustion and downstream temperatures coupled with particulate matter reduction measures can reduce or eliminate PCDD and PCDF emissions.

MSW-fueled boiler systems present greater air pollution problems than most other biomass-fueled plants. Some of the advanced emission control systems used with a 350-t/day plant that meets California’s stringent South Coast Air Quality Requirements are described here to illustrate how effective the controls are. This particular plant is believed to have the lowest emissions of any refuse — to-energy plant in the world (Moore and Cooper, 1990). It is designed to generate 52,200 kg/h of steam at 435°C and 4483 kPa to supply a turbine generator developing 11.4 MW of electrical power. Combustion of MSW results in the formation of acid gases derived from chemically bound chlorine, fluorine, and sulfur in the refuse. These gases and particulate carryover from the boiler

must be removed before the flue gases are exhausted to the atmosphere. The flue gases enter the bottom of a dry scrubber through a cyclonic section designed to remove flyash particles larger than 150 /cm. From the cyclone, the gases flow upward through a spray section where atomized lime slurry is introduced. The lime reacts with the acid gases to produce nonacidic salts. Water in the slurry is completely evaporated by the flue gas, which lowers the temperature of the gas leaving the unit. The resulting dry reaction product falls out into a scrubber hopper. The remaining particulate matter in the flue gas is conditioned by a material (Tesisorb) that promotes particle agglomeration for subsequent removal by glass fabric filters. The removal efficiencies of the acid gases and particulates are in the high 90s on a percentage basis. Nitrogen oxides are reduced by ammonia injection above the combustion zone, and careful control of combustion conditions minimizes carbon monoxide, dioxin, and furan emissions. The carbon monoxide and nitrogen oxide emissions ranged from 18 to 25 ppmv and 48 to 69 ppmv, respectively. The dioxin emissions meet California’s requirements, assuming the lower analytical detec­tion limit is the actual emission.

It cannot be emphasized enough that the combustion process in biomass — fueled power plants should always be controlled with the objective of maximiz­ing boiler efficiency and minimizing stack gas emissions. These goals might be considered to be contradictory, since high-efficiency combustion generally means higher flame temperatures, which can result in higher NOx emissions. However, in a power steam generator firing whole tree chips, it is quite possible to achieve rated boiler efficiency and low NOx formation at the same time. Operation of the 50-MW plant in Burlington, Vermont under the proper conditions with 100% green, whole tree chips containing 40 wt % moisture afforded NOx emissions as low as 0.062 kg/GJ while still achieving a boiler efficiency in the range 68 to 73%, which is in the high end of the design range, without the use of postcombustion treatment or flue gas recirculation (Tewksbury, 1991). At full load, this plant is designed to burn 90.7 t/h of green wood fuel; the nominal steam capacity is 217,687 kg/h at 8828 kPa and 510°C. This performance was achieved at full load by careful control of the fuel distribution on the grates and the air-fuel ratio, and by balancing the overfire air and underfire air. Substoichiometric firing of the wood on the grates kept flame temperature and NOx formation low, but generated a high level of CO. A second level of combustion higher in the fire box occurs when additional air is added to complete the combustion process at a temperature where little or no further NOx is created. This operating mode simulated a two-stage combustion system. When the plant must be operated at minimum load, about 33% of normal load, a similar operating mode provided the best results, although the NOx emissions were slightly higher than before.

In this process, gasification is carried out in the presence of hydrogen. Most of the research on hydrogasification has targeted methane as the final product. One approach involves the sequential production of synthesis gas and then methanation of the carbon monoxide with hydrogen to yield methane. Another route involves the direct reaction of the feed with hydrogen (Feldmann et ah,
1981). In this process (Fig. 9.7), shredded feed is converted with hydrogen — containing gas to a gas containing relatively high methane concentrations in the first-stage reactor. The product char from the first stage is used in a second — stage reactor to generate the hydrogen-rich synthesis gas for the first stage. From experimental results obtained with the first-stage hydrogasifier operated in the free-fall and moving-bed modes at 1.72 MPa and 870°C with pure hydrogen, calculations shown in Table 9.8 were made to estimate the composi­tion and yield of the high-methane gas produced when the first stage is inte­grated with an entrained-char gasifier as the second stage. Note that although the methane content of the raw product gas is projected to be higher in the moving-bed reactor than in the falling-bed reactor, the gas from the first stage must still be reacted in a shift converter to adjust the H2/CO ratio, scrubbed to remove C02, and methanated to obtain SNG.

Other research shows that internally generated hydrogen for hydroconver­sion can be obtained in a single-stage, noncatalytic, fluidized bed reactor (Babu, Tran, and Singh, 1980). In this work, hydroconversion was envisaged to occur in a series of steps: nearly instantaneous thermal decomposition of biomass followed by gas-phase hydrogenation of volatile products to yield hydrocarbon

gases, hydrogen, carbon oxides, water, and hydrocarbon liquids; rapid conver­sion of a part of the devolatilized biomass char to methane at appropriate gasification conditions; slow residual biomass char gasification with hydrogen and steam to yield methane, hydrogen, and the carbon oxides; and combustion of residual biomass char, which supplies the energy for the endothermic char gasification reactions. Examination of hydroconversion under a variety of pressure and temperature conditions with woody biomass and hydrogen, steam, and hydrogen-steam mixtures and study of the kinetics of the slower steam-char reactions led to a conceptual process called RENUGAS®, which will be described in more detail later. Biomass is converted in the reactor, which is operated at about 2.2 MPa, 800°C, and residence times of a few minutes with steam-oxygen injection. About 95% carbon conversion is anticipated to produce a medium-energy gas which can be subjected to the shift reaction, scrubbing, and methanation to form SNG. The cold gas thermal efficiencies are estimated to be about 60%. Since this initial work, RENUGAS has been tested at the pilot, PDU, and demonstration scales, and is being commercialized.

Comparative studies on the gasification of wood in the presence of steam and hydrogen have shown that steam gasification proceeds at a much higher rate than hydrogasification (Feldmann et al, 1981). Carbon conversions 30 to 40% higher than those achieved with hydrogen can be achieved with steam at comparable residence times. Steam/wood weight ratios up to 0.45 promoted increased carbon conversion, but had little effect on methane concentration. Other experiments show that potassium carbonate-catalyzed steam gasification of wood in combination with commercial methanation and cracking catalysts can yield gas mixtures containing essentially equal volumes of methane and

carbon dioxide at steam/wood weight ratios below 0.25 and atmospheric pres­sure and temperatures near 700°C (Mudge et al, 1979). Other catalyst combina­tions produced high yields of product gas containing about 2:1 hydrogen/ carbon monoxide and little methane at steam/wood weight ratios of about 0.75 and a temperature of 750°C. Typical results for both of these studies are shown in Table 9.9. The steam/wood ratios and the catalysts used can have major effects on the product gas compositions. The composition of the product gas can also be manipulated depending on whether a synthesis gas or a fuel gas is desired.

It is of interest to examine potential sources of atmospheric C02 by analysis of the global distribution of carbon in all its forms. The data presented in Table 1.8 show that atmospheric carbon, which can be assumed to be essentially all in the form of C02 (i. e., 700 Gt carbon equals 2570 Gt of COz) comprises

only about 1.6% of total global carbon, excluding lithospheric carbon. Obvious sources of direct or indirect additions of C02 to the atmosphere are therefore fossil fuel deposits, since portions of them are combusted each year as fuels, and terrestrial biomass. Biomass, the photosynthetic sink for removal of C02 from the atmosphere, is important because any changes that modify natural biomass growth can affect ambient C02 concentration. Reducing the size of the photosynthetic sink by such practices as slash-and-bum agriculture, large — scale wood burning, and rain-forest destruction causes an overall reduction in the amount of natural photosynthesis.

To develop more quantitative information regarding atmospheric C02, the emissions on combustion of coal, oil, and natural gas per energy input unit (Appendix C) were used to calculate the C02 generated from fossil fuel combus­tion for the world’s regions and each of the top 10 energy-consuming countries (Table 1.9). Oil is the largest C02 source, followed by coal and natural gas. It is obvious that the largest energy-consuming regions of the world generate relatively more fossil-based C02, and that the world’s 10 top energy-consuming

countries generate almost two-thirds of the world’s total C02 emitted on combustion of fossil fuels. This kind of information has led to several national plans and international agreements to attempt to lower or at least maintain atmospheric C02 by reducing fossil fuel consumption through such mecha­nisms as fossil carbon consumption taxes and higher-efficiency hardware. A

variety of technologies for removal of C02 from the environment have also been proposed.

Although the position has been supported with limited and sometimes questionable data, it has come to be accepted as fact by manv if not most climate change specialists that fossil fuel consumption is the major cause of atmospheric C02 buildup. The C02 in the atmosphere is estimated to have a mass of about 2640 Gt (Table 1.7). Uncertainty is a factor because it is only by inference that the mass is calculated. But many direct analyses of atmo­spheric C02 have been made at different locations throughout the world. Analysis of air trapped in ancient ice cores shows that about 160,000 years ago, atmospheric C02 concentration was about 200 ppm and then peaked at about 300 ppm 130,000 and 10,000 years ago. The concentration then began to increase from an apparent equilibrium value of about 280 ppm in the eighteenth century to its present level of about 360 ppm, the highest concentra­tion in the past 160,000 years. Atmospheric C02 concentration has increased at least 50 ppm since 1860 and is currently increasing at an annual rate of about 1.5 ppm according to analyses carried out continuously over the last several decades. Presuming the atmospheric mass of 2640 Gt is correct, this corresponds to an annual increase of about 11.3 Gt/year.

Compared to other carbon flows, C02 emissions from fossil fuel consump­tion by country are perhaps the most accurate, large-scale carbon flux calcu­lations that can be performed. The reason for this is that detailed data on fos­sil fuel production and consumption are compiled and reported worldwide. Since the mid-1800s, fossil fuel usage has increased significantly, notably since World War II as discussed earlier, to over 300 EJ/year (Fig. 1.6). Global C02 emissions from fossil fuel combustion have been calculated and reported to four significant figures for many years; the annual average from 1978 to 1987 was 18.91 Gt/year (Klass, 1993) and is in the 22-Gt/year range in the 1990s (Table 1.9). So fossil fuel emissions are about twice the annual atmospheric C02 buildup. This type of “factual data” comprises the essence of the argument that fossil fuel consumption is the primary cause of C02 buildup in the atmo­sphere, and sic climate change. Much of the additional evidence is qualitative and uncertain because the study of global C02 buildup is inextricably related to global carbon cycles and reservoirs and the myriad of processes that take place over time on a living planet. The problem from an investigative standpoint is extremely difficult to elaborate. Few direct measurements can be made with precision and then be reproduced. Broad use is made of modeling, and real- world confirmation of the conclusions is often anecdotal. As will be shown later (Chapter 2), biomass has a very important role in atmospheric C02 fluxes and may affect ambient concentrations much more than fossil fuel consumption alone. Because of the environmental trends today, it appears that international agreements to limit fossil fuel consumption will be implemented sometime in the twenty-first century. This will require much greater usage of alternative fuels, especially renewable biomass energy and biofuels manufactured from biomass.

CHAPTER

About one-third of the world’s land area is forestland. Broad-leaved evergreen trees are a dominant species in tropical rain forests near the equator (Spurr, 1979). In the northern hemisphere, stands of coniferous softwood trees such as spruce, fir, and larch dominate in the boreal forests at the higher latitudes, while both the broad-leaved deciduous hardwoods such as oak, beach, and maple and the conifers such as pine and fir are found in the middle latitudes. Silviculture, or the growth of trees, is practiced by five basic methods: exploi­tative, conventional extensive, conventional intensive, naturalistic, and short- rotation (Spurr, 1979). The exploitative method is simply the harvesting of trees without regard to regeneration. The conventional extensive method is the harvesting of mature trees so that natural regeneration is encouraged. Conventional intensive silviculture is the growing and harvesting of commer­cial tree species in essentially pure stands such as Douglas fir and pine on tree farms. The naturalistic method has been defined as the growth of selected mixed tree species, including hardwoods, in which the species are selected to match the ecology of the site. The last method, short-rotation silviculture (short-rotation woody crop or SRWC, short-rotation intensive culture or SRIC), has been suggested as the most suitable method for energy applications. In this technique, trees that grow quickly are harvested every few years, in contrast to once every 20 or more years. Fast-growing trees such as cottonwood, red alder, and aspen are intensively cultivated and mechanically harvested every 3 to 6 years when they are 3 to 6 m high and only a few centimeters in diameter. The young trees are converted into chips for further processing or direct fuel use and the small remaining stems or stumps form new sprouts by vegetative growth (coppicing) and are intensively cultivated again. SRWC production affords dry yields of several tons of biomass per hectare annually, often without large energy inputs for fertilization, irrigation, cultivation, and harvesting, so that the energy balance is positive.

It should be noted that although the prime purpose is to produce wood fiber for the manufacture of paper products, the pulp and paper companies have operated large tree plantations that yield energy as a by-product for decades. Heat, steam, and electricity are produced from wood wastes and also black liquor, which is generated in the paper manufacturing process (Chapter 5). Almost two-thirds of all renewable fuels consumed by the U. S. industrial sector is accounted for by the industry’s use of black liquor (U. S. Dept, of Energy, 1995). The pulp and paper industry produces well over half of its own energy needs and clearly has a great interest in sustained-yield forestry. In the United States, several pulp and paper companies are developing SRWC technology to provide improved methods for supplying fiber to pulp mills and by-product energy (Stokes and Hartsough, 1994). In 1994, approximately 20,000 ha of SRWC systems were operated in the United States by the pulp and paper industry; 40,000 to 80,000 ha were projected to be operated by the year 2000.

Historically, trees are important resources and still serve as major energy resources in many developing countries. No fewer than 1.5 billion people in developing countries derive at least 90% of their energy requirements from wood and charcoal, and at least another billion people meet at least 50% of their energy needs this way (National Academy of Sciences, 1980). Hundreds of species in the seven genera Acacia, Casuanna, Eucalyptus, Pinus, Prosopis, and Trema are used as fuelwood in developing countries (Little, 1980). Several studies of temperate forests indicate productivities from about 9 to 28 t/ha — year, while the corresponding yields of tropical forests are higher, ranging from about 20 to 50 t/ha-year (Nichiporovich, 1967). These yields are obtained using conventional forestry methods over long periods of time, 20 to 50 years or more. Productivity is initially low in a new forest, slowly increases for about the first 20 years, and then begins to decline. Coniferous forests will grow

even in the winter months if the temperatures are not too low; they do not exhibit the yield fluctuations characteristic of deciduous forests.

One of the tree species that has been studied in great detail as a renewable energy resource is the eucalyptus (Mariani, 1978), evergreen hardwood trees that belong to the myrtle family, Myrtaceae, and the genus Eucalyptus. There are approximately 450 to over 700 identifiable species in the genus. The eucalyptus is a rapidly growing tree native to Australia and New Guinea, and is widely grown in the United States, especially in Southern California and Hawaii for a variety of construction purposes. High-density plantings (17,790 trees/ha) in Southern California of E. grandis harvested twice annually have been reported to yield in excess of 22 dry t/ha-year (Sachs, Gilpin, and Mock, 1980). It appears to be a prime candidate for energy use because it reaches a size suitable for harvesting in about 7 years. Several species have the ability to coppice after harvesting, and as many as four harvests can be obtained from a single stump before replanting is necessary. In several South American countries, eucalyptus trees are converted to charcoal and used as fuel. Eucalyp­tus wood has also been used to power integrated sawmill, wood distillation, and charcoal-iron plants in western Australia. Several large areas of marginal land in the United States may be suitable for establishing eucalyptus energy farms. These areas are in the western and central regions of California and the southeastern United States.

Various species and hybrids of the genus Populus are some of the more promising candidates for SRWC growth and harvesting as an energy resource (Sajdak et al, 1981). The group has long been cultivated in Europe and more recently in the eastern United States and Canada. Populus hybrids are easily developed and the resulting progeny are propagated vegetatively using stem cuttings. Consequently, there are hundreds of numbered or named clones established throughout the eastern United States. Summaries of record SRWC small-plot yields for Populus hybrids have shown production levels of 15-20 dry t/ha-year (Hansen, 1988) and yields of 30-40 dry t/ha-year have been projected as attainable goals through genetic engineering (Ranney, Wright, and Layton, 1987). SRWC growth of hybrid poplar clone D-01 has been reported to afford yields of biomass that range as high as 112-202 green t/ha — year (56-101 dry t/ha-year at 50 wt % moisture) (Dula, 1984). These results were reported with very high-density plantings that have been termed wood — grass in which the crop is grown like grass and is harvested several times each growing season. However, there is some dispute regarding the benefits of woodgrass growth vs SRWC growth (Wright et al, 1989). Several investigators have not been able to reproduce these results (с/. DeBell and Clendenen, 1991), although the high-density planting technique seems to have some potential benefits.

It was concluded from early studies that deciduous trees are preferred over conifers for the production of woody biomass for conversion to biofuels (InterTechnology Corp., 1975). Conifers are used as fuel in many parts of the world, including the United States, but the long-term research effort to develop woody species as dedicated energy crops emphasizes mostly deciduous species (с/. Ferrell et al, 1993; Wright, 1994; Ferrell, Wright, and Tuskan, 1995). Several deciduous species can be started readily from clones, resprout copiously and vigorously from their stumps at least five or six times without loss of vigor, and exhibit rapid initial growth. They can also be grown on sites with slopes as steep as 25%, where precipitation is 50 cm or more per year. It has been estimated that yields between about 18 and 22 dry t/ha-year are possible on a sustained basis almost anywhere in the Eastern and Central time zones in the United States from deciduous trees grown in dense plantings. Table 4.10 lists deciduous trees that were judged in early work to have desirable growth characteristics for plantation culture and that have been shown to grow satisfactorily at high planting densities for short and repeated harvest cycles.

The treatment of wood chips with steam at elevated pressures and temperatures for short time periods followed by rapid decompression changes the physical state of the woody structure by defibration. Although some chemical changes occur with the hemicelluloses and lignins in this process, the particle sizes are reduced and surface areas and pore volumes are increased, so some discus­sion of the technique is warranted. The process was originally developed in 1925 and has been extensively used in the manufacture of hardboard (Spalt, 1977). The commercial process involves pressurization with saturated steam at pressures up to about 7 MPa. The process has also been proposed for the pretreatment of lignocellulosic feedstocks in the production of fermentation ethanol (Chapter 11) because of the large increase in accessibility of the cellulosic fraction to enzymatic hydrolysis (cf. Schultz, Biermann, and McGin­nis, 1983; Mes-Hartree, Hogan, and Saddler, 1987; Foody and Foody, 1991).

In a series of steam-explosion experiments with different wood chips, sugar­cane bagasse, ground corn stover, and ground rice hulls at selected tempera­tures between 190 and 250°C and treatment times of 1 or 2 min, the results were as follows (Schultz and McGinnis, 1984): All material was defibrated; the hemicelluloses were at least partially degraded, while the remaining hemi­celluloses were extractable with hot water; the lignins were depolymerized by cleavage of the ether linkages, while at the higher temperatures, they became moderately condensed; and the enzymatic hydrolysis rates of the steam — exploded material increased dramatically for all materials except corn stover. Corn stover hydrolyzed at high rates without treatment. This study and others on steam explosion suggest that the technique can be used for several different biomass applications ranging from modifying the fibrous structure and particle sizes alone at the lower temperatures to a combination of physical and chemical changes at the higher temperatures.

As already mentioned, the older production methods for conversion of biomass to charcoal are slow processes and the yields are low. Several days are required to complete the process in earthen pits with seasoned wood, and the yields are only about 10 to 13% of the dry wood weight because most of the volatile organics leave the pyrolysis zone before carbonization occurs. But note that the coalification of biomass in nature is truly a long process compared to human-controlled processes. Coal is a product of the gradual, natural decompo­sition of cellulosic biomass, without free access to air, under the influence of pressure and temperature; it is formed through the successive stages of peat, lignite or brown coal, bituminous or soft coal, and anthracite or hard coal, characterized by increasing carbon content (с/. Fieser and Fieser, 1950).

As shown in Tables 8.6 and 8.7, the volatile matter in biomass as measured by the ASTM and TG procedures is much higher than the fixed carbon content, so to significantly increase charcoal yields, the volatiles must be carbonized as well. Closed reactors can be designed to keep the volatiles in the pyrolysis zone for longer periods and increase carbonization. The use of beehive kilns, for example, affords charcoal yields up to 35%, but the process still requires several days for completion (с/. Antal et al., 1996). The Ford Motor Company process in Badger-Stafford retorts was performed over 24-h cycles and the charcoal yields were about 27% (Table 8.8).

From a theoretical perspective, pure cellulose contains 44.4 wt % carbon, so the maximum theoretical yield of charcoal, assuming all of the cellulosics can be carbonized, is 44.4 wt %. But with dry wood chars containing about 60 to 70 wt % fixed carbon (с/. Table 8.6), the theoretical maximum yields of charcoal including volatile matter and ash from wood feedstocks then corre­spond to about 65 to 75% by weight of the dry wood. It is evident that if charcoal is the desired product, considerable process improvements should be possible.

A batch process that affords higher charcoal yields with biomass feedstocks over a relatively short reaction time has been developed (Antal et al, 1996). The biomass feedstock, usually logs, wood chips, or nutshells, is maintained under pressure up to about 0.7 MPa at typical temperatures of 450°C for pyrolysis times of 15 min to 2 h. The yields of dry charcoal have ranged from 42 to 62% as shown in Table 8.9. The PDU (process development unit) designed to demonstrate this technology employs a cylindrical steel canister covered by a lid. The canister is charged with biomass that is not predried and is placed within the pressure-tight steel vessel of the PDU, which for demonstration purposes was electrically heated. Startup involves heating the PDU to pyrolysis temperatures and then maintaining the PDU at that temperature. The steam formed from the contained moisture is released as needed to maintain the pressure at 0.7 MPa. Commercial systems will probably use the fuel gas emitted to supply heat. However, the results obtained with the PDU illustrate how the charcoal yields approach the theoretical limits, so the yields of other pyrolysis products may be too low to supply any needed fuel. The results from the PDU suggest that both the moisture in the biomass and the pressure can be manipulated to maximize charcoal yields. The vapors emitted during this batch process are kept in contact with the solid biomass undergoing pyrolysis at the internal pressure of the PDU. These conditions result in increased char and low tar yields at short reaction times compared to those that have been employed in most other processes.