Many processes can be used to produce energy or gaseous, liquid, and solid fuels from virgin and waste biomass. In addition, chemicals can be produced from biomass by a wide range of processes. It is evident from the data and information presented in this chapter, however, that the characteristics of potential feedstocks, particularly their moisture and energy contents, can have profound effects on the utility of specific biomass species and waste biomass. Table 3.10 is a summary of the principal feedstock, process, and product types that are considered in developing a synfuel-from-biomass process. There are many interacting parameters and possible feedstock—process—product combi­nations, but not all are feasible from a practical standpoint. For example, the separation of small amounts of metals present in biomass and the direct

TABLE 3.10 Summary of Feedstock, Conversion Process, and Primary Energy Product Types

Feedstock Conversion process Primary energy product

combustion of high-moisture-content algae are technically possible, but ener­getically unfavorable.

Moisture content of the biomass chosen is especially important in the selection of suitable conversion processes. The giant brown kelp, Macrocystis pyrifera, contains as much as 97 wt % intracellular water, so thermal gasification techniques such as pyrolysis and hydrogasification cannot be used directly without first drying the algae. Anaerobic fermentation is preferred because the water does not need to be removed. Wood, on the other hand, can often be processed by several different thermal conversion techniques without drying. Figure 3.7 illustrates the effects of thermal drying on biomass used for produc­tion of synthetic natural gas (SNG). A large portion of the feed’s equivalent energy content can be expended for drying, so the properties of a potential feedstock must be considered carefully in relation to the conversion process.

Table 3.11 lists the important feedstock characteristics to be examined when developing a conversion process for a specific virgin or waste biomass feedstock. A particular process also may have specific requirements within a given process type. For example, biological gasification and alcoholic fermenta-

Feed moisture content, %

tion are both microbiological conversion processes, but animal manure, which has a relatively high biodegradability, is not equally applicable as a feedstock for both processes.

In summary, it is not a simple matter to select the proper conversion process for a given biomass feedstock. Both biomass properties and process

Process type

requirements must be examined together and in depth to develop a technically and economically feasible system for producing the desired synfuels and energy products. These subjects will be examined in some detail in subsequent chapters.

The U. S. pulp and paper industry consumes about 2.6 EJ (2.5 quad) of energy per year, or about 2.9% of total annual U. S. energy consumption (U. S. Dept, of Energy, 1995). Energy consumption per tonne of paper produced is about 36 GJ or 6 BOE, a disproportionately large amount, especially when consid­ered in terms of the energy content of conventional paper products, about 17.6 GJ/dry t. In the mid-1990s, U. S. paper production capacity was about 30% of the world’s total capacity and accounted for 40% of all electric power cogenerated by U. S. manufacturing. Yet the paper industry still spent $5.5 billion on energy in 1991, or about 4.3% of the value of its shipments (U. S. Dept, of Energy, 1995). It is not unexpected, therefore, that the paper industry has made a great effort to become as energy self-sufficient as possible.

Black liquor is a major waste biomass resource and a by-product of the paper industry. In the pulping step of the paper manufacturing process, cellu — Iosic fibers are separated from the debarked, chipped wood. The dominant process used for pulping in the United States is the kraft process, which involves cooking the chips at elevated temperature and pressure with a solution of sodium hydroxide and sodium sulfide. Most of the lignins are dissolved in this process, and the resulting pulp is washed to remove the chemicals before further processing of the pulp into paper. The mixture of dissolved wood components and used pulping chemicals in the extract is called black or spent liquor. It is currently burned in recovery boilers to recover the pulping chemicals and to generate steam. The wood residues serve as boiler fuel and the spent chemicals are the bottoms that are processed for reuse in the pulping step.

In 1993, the U. S. paper industry manufactured about 77.1 million tonnes of paper and paperboard products. The average usable energy yield of black liquor for the industry then corresponds to about 14 GJ/t of paper manufactured (2.6 EJ/[77.1 X 106 t]), or about 2.3 BOE/t. This clearly illustrates the need to recover energy from the black liquor, if only to minimize production costs. Black liquor can almost be considered to be a co-product rather than a waste by-product. If boiler efficiencies were included in this calculation, the need to recover and use the energy in the paper manufacturing process would be even more apparent. However, these figures also suggest that there is consider­

able room for improving the efficiency of the chemical pulping process. The fact remains that approximately one-half of the annual primary energy consumed as biomass in the United States in the 1990s is attributed to the paper industry. Black liquor is available in large quantities and is the mainstay of this consump­tion pattern.

Conventional pyrolysis (carbonization, destructive distillation, dry distillation, retorting) consists of the slow, irreversible, thermal degradation of the organic components in biomass, most of which are lignocellulosic polymers, in the absence of oxygen. Slow pyrolysis has traditionally been used for the production of charcoal. Detailed studies of biomass pyrolysis beginning in the 1970s have led to methods of controlling the selectivities and yields of the gaseous, liquid, and solid products by controlling the pyrolysis temperature and heating rate (cf. Stevens, 1994). Short-residence-time pyrolysis (flash, rapid, ultra pyrolysis) of biomass at moderate temperatures can afford up to 70 wt % yields of liquid products (cf. Bridgwater and Bridge, 1991). Pyrolysis conditions can be used that provide high yields of gas or liquid products and char yields of less than 5%. One configuration of an advanced biomass pyrolysis system, for example, involves an ablative, vortex reactor for pyrolysis at biomass residence times of fractions of a second coupled to a downstream vapor cracker (Diebold and

Scahill, 1988). The products can be varied to yield up to about 56% liquids (dry) or 90% gases; the char yields are about 10-15% in each case. Advanced pyrolysis processes are discussed in more detail in Section III.

A. Feedstock Composition Impacts

As alluded to in Chapter 8, the ideal biomass feedstock for thermal conversion, whether it be combustion, gasification, or a combination of both, is one that contains low or zero levels of elements such as nitrogen, sulfur, or chlorine, which can form undesirable pollutants and acids that cause corrosion, and no mineral elements that can form inorganic ash and particulates. Ash formation, especially from alkali metals such as potassium and sodium, can lead to fouling of heat exchange surfaces and erosion of turbine blades, in the case of power production systems that use gas turbines, and cause efficiency losses and plant upsets. In addition to undesirable emissions that form acids (SOx), sulfur can also form compounds that deactivate methanol synthesis catalysts, whereas chlorine can be transformed into toxic chlorinated organic derivatives as well as acids.

Biomass is similar to some coals with respect to total ash content as discussed in Chapter 3, but because of the diversity of biomass, several species and types have relatively low ash and also low sulfur contents. Woody biomass is one of the feedstocks of choice for thermal gasification processes. The ash contents are low compared to those of coal, and the sulfur contents are the lowest of almost all biomass species. Grasses and straws are relatively high in ash content compared to most other terrestrial biomass, and when used as feedstocks for thermal conversion systems, such biomass has been found to cause a few fouling problems.

The high moisture contents of aquatic and marine biomass species make it unlikely that they would be considered as feedstocks for thermal gasification processes. However, a few processes can be performed with aqueous slurries or do not require dry biomass feedstocks as described earlier. As harvested, aquatic and marine biomass species often have moisture contents greater than 90% of the total plant weight. In addition to the relatively high ash contents of herbaceous feedstocks, the nitrogen content is an important factor. Grasses are higher in protein nitrogen than woody feedstocks and can increase nitro­gen oxide (NO) emissions on gasification.

The compositions of wood compared to those of other potential biomass feedtocks make woody biomass a preferred feedstock for thermal gasification. Although not shown here, most woody biomass species, especially those indige­nous to the contiguous United States, are similar in composition. It is important to emphasize that quantitative ash analyses of biomass feedstocks sampled at the plant gate and from storage should be carried out periodically and some­times continually to provide real-time data needed for process control. There can be large differences in the amounts of specific mineral components in biomass.

A major mechanism of the fouling of heat exchanger surfaces with biomass feedstocks, particularly the straws and herbaceous residues, is the formation in the thermal conversion zone of low-fusion-point alkali metal salt eutectics such as the alkali metal silicates. The problems caused by these salts and the control methods for combustion and thermal gasification systems were discussed in Chapter 8. Several experienced designers of biomass gasifiers and the operators of commercial plants operated on biomass feedstocks have indicated that the problem is usually not severe with gasification systems, but can be with combustion systems. Temperature control to reduce slagging and the formation of molten agglomerates and equipment designs that avoid contact of the internals with hot gases that may contain low-fusion-point particulates are the preferred control methods for minimizing these problems. For biomass gasifiers that are used to supply fuel for gas turbines, the control methods are similar. Some biomass, although high in minerals, may be low in alkali metals. Fouling by sticky particulates is therefore much less with this type of feedstock.

Some gasification process designers claim to have developed proprietary gas processing systems that yield product gases from biomass gasifiers “cleaner than natural gas” using conventional desulfurization processes for sulfur re­moval and cyclones and proprietary filters to remove ash and char fines. Electrostatic precipitators are not used, and scrubbers are claimed to be optional for some of these systems. These statements are difficult to support without public dissemination of full-scale test results. They are probably true, however, because there are many emissions — and ash-removal systems that have been installed and effectively operated in large-scale commercial biomass combus­tion plants that meet all requirements. Some of these plants are designed to meet California’s stringent South Coast regulations. Much of this experience and technology can be drawn upon to design environmentally clean biomass gasification plants.

Many of the commercial or near-commercial biomass gasification facilities that have been built and operated use green or partially dried feedstocks in which the moisture content of the feedstock to the gasifiers is not specified. The steam-carbon reactions that occur are undoubtedly one of the main reasons for variation in product gas compositions from these systems. Since the carbon content of dry biomass is about 45 wt %, green wood contains about 2.2 kg moisture/kg of carbon. Table 9.10 shows the effects of the moisture content of poplar wood when gasified in an air-blown, downdraft gasifier. As the moisture content of the wood decreases from 34 to 13 wt %, thermal efficiency, product gas heating value, dry gas yield, and the proportion of the combustible components in the dry gas each increase. These data illustrate the importance of specifying feedstock moisture content. Feedstock dryers are essential for some biomass gasification plants depending on the feedstock’s moisture content and variation, as well as on the end uses of the product gases.

The terminology “renewable carbon resource” for virgin and waste biomass is actually a misnomer because the earth’s carbon is in a perpetual state of flux. Carbon is not consumed in the sense that it is no longer available in any form. Many reversible and irreversible chemical reactions occur in such a manner that the carbon cycle makes all forms of carbon, including fossil carbon resources, renewable. It is simply a matter of time that makes one form of carbon more renewable than another. If society could wait several million years so that natural processes could replenish depleted petroleum or natural gas deposits, presuming that replacement occurs, there would never be a shortage of organic fuels as they are distributed and accepted in the world’s energy markets. Unfortunately, this cannot be done, so fixed carbon-containing materials that renew themselves over a time span short enough to make them continuously available in large quantities are needed to maintain and supplement energy supplies. Biomass is a major source of carbon that meets these requirements.

The capture of solar energy as fixed carbon in biomass via photosynthesis, during which carbon dioxide (C02) is converted to organic compounds, is the key initial step in the growth of biomass and is depicted by the equation

C02 + H20 + light 4- chlorophyll —> (CH20) 4- 02.

Carbohydrate, represented by the building block (CH20), is the primary or­ganic product. For each gram mole of carbon fixed, about 470 kj (112 kcal) is absorbed. Oxygen liberated in the process comes exclusively from the water, according to radioactive tracer experiments. Although there are still many unanswered questions regarding the detailed molecular mechanisms of photo­synthesis, the prerequisites for virgin biomass growth are well established; C02, light in the visible region of the electromagnetic spectrum, the sensitizing catalyst chlorophyll, and a living plant are essential. The upper limit of the capture efficiency of the incident solar radiation in biomass has been variously estimated to range from about 8% to as high as 15%, but in most actual situations, it is generally in the 1% range or less (Klass, 1974).

Many other terrestrial biomass species have been proposed as renewable energy resources for their high-energy components that can be used as fuels, for the components capable of conversion to biofuels and chemicals, or for the con­tained energy (с/. Buchannan and Otey, 1978; Cherney et al, 1989; DeLong et al, 1995; Gavett, Van Dyne, and Blase, 1993; Klass, 1974; McLaughlin, Kingsolver, and Hoffmann, 1983; Nemethy, Otvos, and Calvin, 1981; O’Hair, 1982; Schneider, 1973; Shultz and Bragg, 1995; Stauffer, Chubey, and Dorell, 1981; Taylor, 1993). Among them are kenaf (Hibiscus cannabinus), an annual plant reproducing by seed only; sunflower (Helianthus annuus L.), an annual oil seed crop grown in several parts of North America; Eurphorbia lathyris, a sesquiterpene-containing plant species that grows in the semiarid climates of the Southwest and California; Buffalo gourd (Curcurbita foetidissima), a perennial root crop native to arid and semiarid regions of the southwest­ern United States; other root crops such as Jerusalem artichoke (Helian­thus tuberosus), fodder beet (Beta vulgaris), and cassava (Manihot esculenta); alfalfa (Medicago saliva), a perennial legume that grows well on good sites in many parts of North America; soybean (Glycine max) and rapeseed (Brassica campestris), oilseed crops that produce high-quality oil and protein; and many other biomass species that are potentially suitable as renewable energy resources or multipurpose crops including energy and biofuels. Kenaf, for example, is highly fibrous and exhibits rapid growth, high yields, and high cellulose content. It is a pulp crop and is several times more productive than pulpwood trees. Maximum economic growth usually occurs in less than 6 months, and consequently two croppings may be possible in certain regions of the United States. Without irrigation, heights of 4 to 5 meters are average in Florida and Louisiana, but 6-m plants have been observed under near-optimum growth conditions. Yields as high as 45 t/ha-year have been observed on experimental test plots in Florida, and it has been suggested that similar yields could be achieved in the Southwest with irrigation. Another example is the sunflower, which is a good candidate for biomass energy applica­tions too because of its rapid growth, wide adaptability, drought tolerance, short growing season, massive vegetative production, and adaptability to root harvesting. Dry yields have been projected to be as high as 34 t/ha per growing season. Rapeseed is another example; its seeds normally yield 38 to 44 wt % high quality protein and over 40 wt % oil, which affords high-quality biodiesel fuel at the rate of 750 to 900 L/ha-year on extraction and transesterification. Still another example is alfalfa, a well-known and widely-planted herbaceous crop that offers environmental and soil conservation advantages when grown as a 4-year segment in a 7-year rotation with corn and soybeans. With alfalfa yields of about 9 dry t/ha-year and the alfalfa leaf fraction sold as a high-value animal feed, the remaining alfalfa stem fraction can be used as feedstock for power production.

Numerous devices and methods of fabricating solid fuel pellets and briquettes from a variety of biomass, especially RDF, wood, and wood and agricultural residues, have been developed and patented. The pellets and briquettes are manufactured by extrusion and other techniques. A binding agent such as a thermoplastic resin may be incorporated during fabrication. A ring-die extru­sion or die and roller mill is the most widely used machine type in wood pelleting, although punch and die technology has been developed (Folk and Govett, 1992). Other types of pelleting machines include disk pelletizers, drum and rotary cylinder pelletizers, tablet presses, compacting and briquetting rolls, piston-type briquetters, cubers, and screw extruders.

An exemplary method for production of pellets was developed in 1977 (Gunnerman, 1977). A raw material of random particle size such as sawdust or other wood residue, from which rocks, tramp metal, and other foreign materials are removed, is conveyed to a hammermill where particle size is adjusted to a uniform maximum dimension that is about 85% or less of the minimum thickness of the pellets desired. The milled product is then dried in a rotary drum dryer to a moisture content of about 14 to 22 wt % and fed through a ring-shaped die capable of generating pressures between 55 and 275 MPa to afford the desired shape and diameter. The pellet mill die and roller assembly must be capable of producing sufficient compression within the die to raise the temperature of the material to about 160-177°C. The products from the mill have a low, uniform moisture content, a maximum cross-sectional dimension of 13 mm, a density of 400 kg/m3, and a heating value of 19.8 to 20.9 MJ/kg. It is not necessary to add a binder to the particles, providing the pressure during pelleting produces the necessary temperature increase. During extrusion, the lignins in the biomass migrate to the pellet surface and form a skin on cooling that protects the pellet from shattering and from any rapid change in moisture content before use. This same basic procedure has been used over the years in several different hardware designs.

Briquettes are formed by similar procedures except the products are usually larger in diameter and length than pellets. Briquetting is described to consist of subjecting wood residues containing 8 to 15 wt % moisture at a maximum particle size of 0.5 to 1.0 cm to a pressure of about 200 MPa, which increases the temperature about 100-150°C (Ortiz, Migues, and Granada, 1996). The major machine types used to manufacture briquettes are impact, extrusion, hydraulic, pneumatic, and double-roll presses, and die presses that can also be used for pellet production. Briquette production rates are 200 to 1500 kg/ h for impact presses, but some models can produce 2000 to 6000 kg/h; 500 to 2500 kg/h for extrusion presses; and up to 5000 kg/h for hydraulic and pneumatic presses. The pellet machines suitable for pellet or briquette pro­duction contain annular or flat dies. The production rates are as high as 25,000 kg/h.

A few examples of typical biomass densifiers, feedstocks, and densified products are shown in Table 6.4. The first six examples in this table are commercial or commercially available systems, the last of which, Biotruck 2000, is unique (Sutor, 1995). It is a moving vehicle of special design that

continuously performs all of the operations in the field from harvesting agricul­tural virgin biomass to pellet production. The operating sequence consists of the integration into one machine of continuous crop harvesting, size reduction to about 0.6-mm pieces, heating the pieces to temperatures between 80 and 120°C using the waste heat of the engine, and compressing the heated pieces in a toothed-wheel pelleting press. No binder is used. The production rate of pelletized cereal crops is about 8000 kg/h and the bulk density is 500 to 700 kg/m3. In addition to cereal crops, the agricultural biomass suitable for harvesting and conversion to pellets by this system include grasses such as Chinese silvergrass, switchgrass, and hays and straws. Pellets for both feed and fuel applications are produced with Biotruck 2000.

Another unique example of densification listed in Table 6.4 is the production of high-density, moisture-resistant briquettes from wet wood residues without predrying or the use of binders (Edwards, 1991). The briquettes do not disinte­grate when wet and retain a maximum of about 40 wt % moisture after immersion in water. They are made from wood and bark alone or from mixtures in a pilot extruder at operating ram pressures typically ranging from 30 to 50 MPa at a maximum surface temperature of about 210°C. Moisture-resistant briquettes were made in tests from Western hemlock sawdust, a 50:50 mixture of Western hemlock and red cedar sawdusts, and Western hemlock bark hog fuel. The feed contains up to about 65 wt % moisture and must be sized so that the maximum size is less than 80% of the barrel diameter. The key to using wet biomass appears to be the simultaneous removal of excess moisture in the initial portion of the extruder while the feedstock is heated under pressure as it moves through the barrel, and the reduction of the temperature to less than 100°C before the briquettes leave the barrel to avoid the risk of explosive flash evaporation. Briquettes made in this manner contained about 35 wt % moisture. Over a 24-h period of immersion in water, they exhibited 0% swelling and only small changes in density and moisture content. The upper limit of the moisture content after immersion was consistently near 40 wt %.

Although not listed in Table 6.4, the combustible fraction of municipal solid waste, RDF, is commercially available as pellets that are similar to those produced from agricultural and woody residues (cf. Davis and Koep, 1990). The pellets have heating values of about 16.3 to 18.6 MJ/kg, moisture contents of 8 to 10 wt %, less than 10 wt % ash, and densities of 600 to 700 kg/m3.

Development of other densification methods for certain agricultural residues is expected to lead to improvements in soil growth characteristics as well as advanced residue recovery systems for energy applications. For example, cotton is a major crop in the state of Arizona. State law requires that cotton plant residue must be buried to prevent it from serving as an overwintering site for insect pests such as the pink bollworm. Research is underway to develop two systems for collecting and densifying this residue to facilitate removal from the field (Coates, 1995). The stalks are first pulled with an implement developed for the purpose. They are then baled using equipment that produces large round bales, or chopped with a forage harvester and converted into modules. The bales are either 1.2 m in diameter x 1.2 m long, or 1.8 m in diameter x 1.5 m long, depending on the baler used. The modules measure 2.1 m x 2.2 m in cross section and are up to 9.6 m long. The densities of the round bales are 93 to 168 kg/m3. The modules have densities of 168 to 252 kg/m3. The energy required to harvest and densify the residues is 9.2 kWh/t for the bales, and 8.6 kWh/t for the modules, and the heating values of the densified residues are about the same as those of wood. The module system produced a denser package than the baling system, and also made loading easier using truck-mounted module movers.

Because of the broad scope of direct biomass pyrolysis, the basic technologies and principal products are tabulated in Table 8.12 to facilitate easy comparison. The conversion conditions and major products shown in this table are typical, but subject to considerable variation. There are several commonalities among the different pyrolysis methods. Pyrolysis time and temperature are clearly the key operating parameters that have the most influence on product yields and distributions. Moderate but optimized temperatures are needed at short residence times to maximize liquid yields, whereas long residence times and

low temperatures are needed to maximize char yields. Biomass gasification, which can involve pyrolysis at the higher temperatures, is treated in Chapter 9. With few exceptions, the selectivities of specific pyrolysis products are poor. Essentially all of the product liquids produced by direct pyrolysis are highly oxygenated, acidic, generally unstable over time, and contain many com­pounds, and the product gases are low-energy gaseous mixtures. Some specialty and commodity chemicals can be extracted from the product liquids for market. The oils can often be used directly as fuel for power and steam production. The charcoals can be readily separated from the product mix for sale or captive use, and the product gases can be used as fuel and in some cases as sources of chemicals. Overall, however, most biomass pyrolysis plants will have to deal with complex product separations, waste disposal, and further product refining if pure chemicals and other products are required. The costs of the additional operations will have to be justified based on the markets for the various products that can be manufactured by biomass pyrolysis. Multiproduct slates will dominate most of these plants. This can be quite beneficial from a revenue standpoint when markets fluctuate provided the plant operating conditions can be readily changed to take advantage of the markets for specific products, much like a petroleum refinery.

Up to the 1930s, biomass pyrolysis processes were utilized on a large scale to produce several commodity chemicals from pyrolytic oils. These processes have since been largely replaced by nonpyrolytic processes based on petroleum and natural gas feedstocks. Note that only a few building blocks from fossil feedstocks—ethylene, propylene, butadiene, benzene, toluene, xylene, and synthesis gas—are used to manufacture the vast majority of the organic chemi­cals sold today (Chapter 13). Since the First Oil Shock in 1973, hundreds and perhaps thousands of research projects have been carried out to perfect and develop biomass pyrolysis technologies for the production of petroleum substi­tutes. The history of commercializing modern biomass pyrolysis systems in industrialized countries, however, since the First Oil Shock is not very encoura­ging. In the United States, for example, a few advanced design, biomass pyroly­sis plants were built in the 1980s and then closed down, generally because of operating problems and poor economics.

The properties of pyrolytic liquids from biomass such as their high oxygen contents and acidity generally limit their fuel uses to heating oils. The emphasis over the past several years has therefore been to develop methods for upgrading them to more hydrocarbon-like liquids for use as motor fuels and motor fuel components. One approach has focused on catalytic hydrogenation. Hydrogen, which can either be generated from the biomass feed or the conversion prod­ucts, or be obtained from an independent source, is reacted directly with the pyrolytic liquids or intermediate process streams at elevated pressures and temperatures to yield substitute fuels with higher hydrogen-to-carbon ratios.

In theory, highly oxygenated feedstocks should be capable of reduction to liquid and gaseous fuels at any level between the initial oxidation state of the feed and methane:

R(OH)x + у H2 -» RH/OH)^ + у H20 R — R’ + H2 —> RH + R’H.

For a cellulosic material containing hydroxyl groups, the reactions might consist of dehydroxylation and depolymerization by hydrogenolysis, during which there is a transition from solid to liquid to gas. In early work to produce hydrocarbon fuels, hydroliquefaction of biomass or wastes was achieved by direct hydrogenation of wood chips on treatment at 10,132 kPa and 340 to 350°C with water and Raney nickel catalyst (Boocock and Mackay, 1980). The wood is completely converted to an oily liquid, methane, and other hydrocar­bon gases. Batch reaction times of 4 h give oil yields of about 35 wt % of the feedstock. The oil still contains substantial oxygen, about 12 wt %, but has a heating value of about 37.2 MJ/kg. Distillation yields a major fraction that boils in the same range as diesel fuel and is completely miscible with it.

Catalytic hydrogenation of the pyrolytic liquids at elevated pressures and temperatures and deoxygenation with molecular sieve catalysts that yield hy­drocarbon liquids with higher hydrogen-to-carbon ratios than the liquid feed­stocks have been studied (с/. Elliott and Baker, 1987; Rajai et ah, 1991; Bakhshi, Kaitikaneni, and Adjaye, 1995; Laurent, Maggi, and Delmon, 1995; Horne, Nugranad, and Williams, 1995). Several systems are quite effective for convert­ing pyrolytic liquids to hydrocarbons suitable for use as motor fuels or motor fuel additives. Catalytic hydrogenation is well-developed commercial technol­ogy in the petroleum industry and can be applied to pyrolytic oils to obtain partial or complete reduction to hydrocarbons. Deoxygenation with molecular sieve catalysts has the advantage that a source of hydrogen is not needed and BTX and light olefins can be produced by direct passage of the pyrolytic vapors over zeolite catalysts, although conversion conditions must be carefully controlled to avoid formation of undesirable by-products such as coke. High yields of BTX and olefins approaching the theoretical limits can be obtained at high weight hourly space velocities and low steam-to-biomass ratios. This route to olefins appears to offer an economic route to methyl t-butyl ether (MTBE) and ethyl t-butyl ether (ETBE) from the light olefins for use as oxygen­ates in gasolines (Bain et ah, 1993).

It would seem that a more practical approach to the upgrading of pyrolytic liquids from biomass is to utilize what is already on hand, namely, the oxygen­ated product liquids. Instead of conversion to hydrocarbons, which usually requires severe reaction conditions, why not convert the liquids by simple chemistry to other liquids that are suitable for use as motor fuels or additives? Although not directly related to pyrolysis, this approach has been pursued in the development of oxygenated diesel fuels (biodiesel) with natural biomass- derived oils, as will be discussed in Chapter 10. With the possible exception of the lignin components in biomass, the overall thermal efficiencies of convert­ing a highly oxygenated organic feedstock in which the majority of the carbon atoms are bonded to oxygen atoms to other oxygenates should be much more favorable, and lower in cost, than conversion to hydrocarbons. However, there may even be exceptions to this rationale, as will be shown in the following section. Also, there are several opportunities to produce chemical additives suitable for use in modern gasolines as oxygenates by thermochemical process­ing of biomass (Chapter 13).

Downdraft Gasifier

The National Renewable Energy Laboratory (Solar Energy Research Institute at that time) designed, built, and operated a 0.9-t/day prototype, downdraft biomass gasifier between 1980 and 1985 (Reed, Levie, and Markson, 1984; Schiefelbein, 1985; Babu and Bain, 1991). In 1985, Syn-Gas, Inc., scaled this process to a 22-t/day plant to develop the concept for the commercial produc­tion of methanol. Feedstocks included wood chips, urban wood waste, and densified RDF. Tests in the 22-t/day plant at 870 to 930°C with cedar wood feedstock and oxygen gave 87 to 91% carbon conversions and dry gas analyses of 39 to 45 mol % carbon monoxide, 24 to 30 mol % carbon dioxide, 5 to 6 mol % methane, and 21 to 22 mol % hydrogen; the remainder was C2-C3 hydrocarbons. The product gas had a lower heating value (wet) of 8.3 to 9.8 MJ/m3 (n).

Directly Heated, High-Temperature, Steam-Oxygen Fluid-Bed Gasification

The Rheinbraun High-Temperature Winkler process is an outgrowth of the successful operation of two atmospheric Winkler gasifiers operated on lig­nite feedstocks in Germany from 1956 to 1964 with a combined capacity of 34,000 m3/h of synthesis gas, and subsequent operation of a 1.3-t/h pilot plant beginning in 1978 (Schrader et ah, 1984). The process was developed by Rheinische Braunkohlenwerke AG and consists of gasification in a pressurized fluid-bed system supplied with oxygen and steam. Operating pressures and temperatures range up to 1013 kPa and 1100°C. The operating results with lignite at 1013 kPa and 1000°C, and oxygen and steam at 0.36 m3 (n)/kg and 0.41 kg/kg of dry lignite, gave 96% carbon conversion and a combined hydrogen-carbon monoxide yield of 1406 m3 (n)/t. At this steam-to-lignite ratio and an exit gas temperature of 900°C, the raw gas contained about 2 mol % methane. These tests provided the information and data needed to construct a demonstration plant to produce 300 million m3 (n)/year of synthesis gas for methanol synthesis at Rheinbraun’s facility. Feedstock tests were con­ducted for customers worldwide with wood, peat, lignite, and coal feedstocks. Rheinbraun reported that each of these feedstocks is suitable for gasification by their process. Wood, especially, can be converted at high reactor through­put rates.