I. INTRODUCTION

As discussed in previous chapters, there are numerous aquatic and terrestrial virgin biomass species and many types of waste biomass that are potential fuels or feedstocks. With the exception of microalgae and some high-moisture — content biomass, essentially all are solid materials. They contain organic com­pounds, minerals, and moisture. Some of the compositional differences are large. The aquatics, municipal biosolids, and animal manures are high in moisture content; the terrestrial species contain relatively small amounts of moisture. The ash contents of woody biomass species are small; some aquatics and agricultural crops contain large amounts of ash. On a moisture- and ash­free basis, the heating value of most biomass is in the same range, but on a dry basis, these materials can exhibit wide variations.

Because of these broad differences, many of the possible feedstock-process — energy product combinations are not feasible. For example, untreated mu­nicipal biosolids contain very large amounts of moisture and are normally unsuitable for thermochemical conversion. Such feedstocks do not support self-sustained combustion under conventional conditions unless the moisture is reduced by a considerable amount, a high-cost process in wastewater treat­ment plants. Biosolids are more suited for microbial conversion in aqueous systems where a liquid water medium is essential. In contrast, woody biomass is often suitable for direct use as a solid fuel or as a feedstock for thermochemical conversion. Predrying to remove some of the moisture, if needed, is readily accomplished at low cost.

Chapters 6 to 12 address specific groups of processes and methods employed for converting biomass to energy and fuels. In this chapter, the physical pro­cesses employed to prepare biomass for use as fuel or as a feedstock for a conversion process are discussed. The processes examined are dewatering and drying, size reduction, densification, and separation. The physical process, a few specific examples of the process, and its relationship to the thermochemical or microbial process that may be used for subsequent conversion are described.

Considerable experimental work has been done with cellulose to clarify the kinetics of biomass pyrolysis. Most kinetic studies on cellulose pyrolysis have

built on the multistep model proposed in the early work with cellulose and described the evolution of volatiles by a single, pseudo-first-order reaction of the type (с/. Broido, 1976; Bradbury, Sakai, and Shafizadeh, 1979; Zaror and Pyle, 1982; Lai, and Krieger-Brockett, 1993; Antal and Varhegyi, 1993)

dm/dt = A exp (— E/RT)(1 — m),

where A is the preexponential factor (time~ *), E is the apparent activation energy (J/mol), R is the ideal gas constant (J/mol-K), T is the absolute tempera­ture (K), and m is the fraction of volatiles produced at time, t. This expression should be quite useful for designing advanced pyrolysis systems, but unfortu­nately, the reported kinetic factors vary widely and several results are in conflict. However, it was found that experimental thermogravimetric data obtained on prolonged, low-temperature, isothermal treatment of pure cellu­lose fit a slightly modified “Broido-Shafizadeh” model (Antal and Varhegyi, 1995). Under conditions of commercial interest, small samples of pure cellulose are reported to undergo thermal decomposition by a single, first-order, high- activation-energy, rate-determining step. The activation energy is estimated to be 238 kj/g-mol. As our knowledge of biomass pyrolysis expands, it is expected that detailed kinetic parameters for more biomass components will be eluci­dated and applied to fine-tune practical biomass pyrolysis systems and designs.

Most of the experimental evidence accumulated over many years of study to understand biomass pyrolysis indicates that there are four basic kinds of processes. All of them involve the formation of large amounts of water vapor. One type is dominant at low heating rates and relatively low pyrolysis tempera­tures below about 250 to 300°C; chars, tars, and dehydration products are the primary products and some volatiles and gases are formed. One type is domi­nant at conventional pyrolysis conditions of intermediate heating rates and temperatures in the 300 to 600°C range; chars, tars, volatiles, and gases are formed in reasonable yields. The third type is dominant at fast heating rates and temperatures in the range 450 to 600°C; volatile liquids are the primary products, but some chars, tars, and gases are formed. The fourth type of pyrolysis occurs at temperatures above 600°C where gasification reactions begin to dominate. Devolatilization of biomass is important in each of these pyrolysis types.

Union Carbide Corporation developed a partial oxidation process called the PUROX System for converting MSW to fuel gas and an inert slag (Fisher, Kasbohm, and Rivero, 1976). The process was scaled up in the mid 1970s from a pilot plant to a commercial, 181-t/day plant at Union Carbide’s facility in West Virginia. A 68-t/day plant was also built in Japan. The plant in West Virginia was operated successfully on MSW. One tonne of refuse required about 0.18 t of oxygen and produced 0.6 t of medium-energy gas with a higher heating value of about 14.5 MJ/m3 (n), 2 t of sterile aggregate residue, and 0.25 t of wastewater. Within the process, 0.03 t of oil was separated in the gas-cleaning train and recycled to the furnace for cracking into additional gas. A typical gas analysis was 40 mol % carbon monoxide, 23 mol % carbon dioxide, 5 mol % methane, 5 mol % C2+, 26 mol % hydrogen, and 1 mol % nitrogen. The energy balance expressed in terms of percent of the energy in the feedstock was a net 68% recovered in the product gas, 21% lost on conver­sion, and 11% used for in-plant electric power generation. The 181-t/day plant included front-end shredding and separation equipment for ferrous metal recovery, liquid separation equipment for recycling the condensed oil to the reactor, provision for removal of the slag from the hearth and quenching in a water bath, and treatment of the product gas by water scrubbing and electro­static precipitation. The reactor was a three-zoned, vertical shaft furnace operat­ing at about 50 cm of water. The RDF was fed to the top of the furnace through a gas seal and oxygen was injected at the bottom. The furnace was maintained essentially full of RDF which continually descends through the reactor. The oxygen in the hearth reacts with char to generate slagging temperatures to melt the glass and metals. Projections indicated that with a 1360-t/day plant composed of 181 to 317-t/day modular units, about 114 million L/year of methanol could be manufactured.

Detailed estimation of the amounts of biomass carbon on the earth’s surface is the ultimate problem in global statistical analysis. Yet what appear to be reasonable projections have been made using available data, maps, and surveys. The validity of the conclusions in their entirety is difficult to support with hard data because of the nature of the problem. But such analyses must be performed to assess the practical feasibility of biomass energy systems and the gross types of biomass that might be available for energy applications.

The results of one such study are summarized in Table 2.2. Ignoring the changes in agricultural practice and the deforestation that have taken place over the last few decades, this is perhaps one of the better attempts to con­duct an analysis of the earth’s biomass carbon distribution (Whittaker and Likens, 1975). Each ecosystem on the earth is considered in terms of area, mean net carbon production per year, and standing biomass carbon. Standing biomass carbon is that contained in biomass on the earth’s surface and does not include the carbon stored in biomass underground. A condensation of this data (Table 2.3) facilitates interpretation. Of the total net carbon fixed on the earth each year, forest biomass, which is produced on only 9.5% of the earth’s surface, contributes more than any other source. Marine sources of net fixed carbon are also high, as might be expected because of the large area of the earth occupied by water. But the high turnover rates of carbon in a marine environ­ment result in relatively small steady-state quantities of standing carbon. In contrast, the low turnover rates of forest biomass make it the largest contributor

to standing carbon reserves. According to this assessment, the forests produce about 43% of the net carbon fixed each year and contain over 89% of the standing biomass carbon of the earth. Tropical forests are the largest sources of these carbon reserves. Temperate deciduous and evergreen forests are also major sources of biomass carbon. Next in order of biomass carbon supply would probably be the savanna and grasslands. Note that cultivated land is one of the smaller producers of fixed carbon and is only about 9% of the total terrestrial area of the earth.

It is necessary to emphasize that anthropological activities and the increasing population, particularly in developing and Third World countries, continue to make it more difficult to sustain the world’s biomass growth areas. It has been estimated that tropical forests are disappearing at a rate of tens of thousands of square miles per year. Satellite imaging and field surveys show that Brazil alone has a deforestation rate of about 8 X 106 ha/year (19.8 X 106 ac/year; 30,888 mi. Vyear) (Repetto, 1990). At mean net biomass carbon yields of 9.90 t/ha-year for tropical rain forests (Table 2.2), this rate of deforestation corresponds to a loss of 79.2 x 106 t/year of net biomass carbon productivity.

The remaining carbon transport mechanisms on earth are primarily physical mechanisms, such as the solution of carbonate sediments in the sea and the release of dissolved C02 to the atmosphere by the hydrosphere. Because of the relatively short lifetimes of live biomass (phytoplankton and zooplankton) in the oceans compared to those of land biomass, there is a much larger amount of carbon in viable land biomass at any given time. The great bulk of carbon, however, is contained in the lithosphere as carbonates in rock. The carbon deposits that contain little or no stored chemical energy, although some high — temperature deposits can provide considerable thermal energy, consist of lith­ospheric sediments and atmospheric and hydrospheric C02. Together, these carbon sources comprise 99.96% of the total carbon estimated to exist on the earth (Table 2.4). The carbon in fossil fuel deposits is only about 0.02% of the total, and live and dead biomass carbon makes up the remainder, about 0.02%. Biomass carbon is thus a very small fraction of the total carbon inventory of the earth, but it is an extremely important fraction. It helps to maintain the delicate balance among the atmosphere, hydrosphere, and biosphere necessary to support all life forms, and is essential to maintain the diversity of species that inhabit the earth and to sustain their gene pools. Any large-scale utilization of biomass carbon, especially virgin material, therefore requires that it be replaced, preferably as it is consumed so that the biomass reservoirs are not

reduced. Indeed, enlargement of these reservoirs may become necessary as the world’s population expands and climate changes occur.

Microalgae have long been under development as renewable energy resources and other useful products (Benemann and Weissman, 1993). Almost 20,000 species are known. Unicellular algae such as the species Chlorella and Scenedes — mus have been produced by continuous processes in outdoor light at high photosynthesis efficiencies. Chlorella has been reported to be produced at a rate as high as 1.0 dry t/ha-day. This corresponds to an annual rate of 401dry t/ha-year presuming growth can be sustained (Retovsky, 1966). These figures are probably in error, but there is no theoretical reason why yields cannot achieve very high values because the process of producing algae can be almost totally controlled. Also, production is not composed only of surface growth. Algae are produced as slurries in lakes, ponds, and custom-designed raceways so that the depth of the biomass-producing area as well as plant yield per unit volume of water are important parameters. The nutrients for algae production can be supplied by municipal biosolids and other wastewaters. It should be pointed out that most unicellular algae are grown in fresh water, which tends to limit their energy applications to small-scale algae farms. The high water content of unicellular algae also tends to limit the conversion processes to biological methods. But this can be an advantage in some cases where the particular microalgae exudes triglycerides without cell destruction so that the product oil is continuously formed and can be easily recovered from the water surface.

Macroscopic multicellular algae, or seaweeds, have also been considered as renewable energy resources for many years. Some of the candidates are the giant brown kelp Macrocystis pyrifera (Bryce, 1978; North, 1971; North, Gerard, and Kubawabara, 1981), the red benthic alga Gracilaria tikvahiae (LaPointe and Hanisak, 1985; Ryther and DeBusk, 1982), and the floating, brown pelagic algae Sargassum natans and S. ftuitans (LaPointe and Hanisak, 1985). Giant brown kelp has been studied in great detail and is harvested commercially off the California coast. Because of its high potassium content, giant brown kelp was used as a commercial source of potash during World War I and is used today as a commercial source of organic gums, thickening agents, and alginic acid derivatives. Off the East Coast, Laminaria seaweed is harvested for the manufacture of alginic acid derivatives. In tropical seas not cooled by upwelled water, species of the Sargassum variety of algae may be suitable as renewable energy sources. Several species of Sargassum grow naturally around reefs sur­

rounding the Hawaiian Islands. Unfortunately, only a small amount of research has been done on Sargassum and little detailed information is available about this alga. A considerable amount of data on yields and growth requirements are available, however, on the Macrocystis and Laminaria varieties. Again, the very high water content of macroscopic algae suggests that biological conver­sion processes rather than thermochemical conversion processes should be used for synfuel manufacture. The manufacture of co-products from macro­scopic algae, such as polysaccharide derivatives, along with biofuel might make it feasible to use thermochemical processing techniques on intermediate process streams.

A. Fundamentals

It is sometimes desirable to physically separate potential biomass feedstocks into two or more components for different applications. The subject is quite broad in scope because of the wide range of biomass types processed and the variety of separation methods that are used. Even the harvesting of virgin biomass involves physical separation technologies. Examples are the separation of agricultural biomass into foodstuffs and residues that may serve as fuel or as a raw material for synfuel manufacture, the separation of forest biomass into the darker bark-containing fraction and the pulpable components, the separation of marine biomass to isolate various chemicals, the separation of urban refuse into RDF and metals, glass, and plastics for recycling, and the separation of oils from oilseeds. Common operations such as screening, air classification, magnetic separation, extraction, mechanical expression under pressure, distillation, filtration, and crystallization are often used as well as industry-specific methods characteristic of farming, forest products, and spe­cialized industries. Since the biomass types are so numerous and the physical separation methods are usually customized, some details of a few specific examples are described here to illustrate the scope of the subject, and how separation is performed. A few potential applications of physical separation methods are also described.

One of the more innovative low-temperature, low-pressure, thermochem­ical techniques of directly liquefying biomass in water involves the use of 57 wt % aqueous hydriodic acid (HI), the azeotrope boiling at 127°C (Douglas and Sabade, 1985). When treated at 127°C with the azeotrope in a stoichiomet­ric excess of 1.6 to 3.8 of the amount required for complete reduction, cellulose is rapidly hydrolyzed and converted to hydrocarbon-like molecules. The yields reach 60 to 70% at reaction times as short as 0.5 min. The laboratory data are consistent with chemistry in which HI acts to form alkyl iodide intermediates that are then converted to hydrocarbons and molecular iodine by further reaction with HI. The stoichiometry developed from the experimental data with cellulose is

C6H10O5 + 8.28HI -> C6H912I018O0,40 + 4.52H20 + 4.08I2.

Products corresponding to 50% deoxygenation in 1 min, 75% in 30 min, and 92% in 24 h are obtained; charcoal is not formed. Up to 98% of the HI reacted appears as molecular iodine. Ether extraction yields a material that has H: C and I: C ratios of 1.52 and 0.03, and yet there is a 90% reduction in the О: C ratio.

Hydriodic acid is a powerful reducing agent that can even be employed for conversion of benzene to cyclohexane. It is also well known that alcohols can be converted to alkyl iodides on reaction with HI and that alkyl iodides react with HI to form hydrocarbons. Since the experimental data obtained with cellulose shows that most of the HI is ultimately converted to molecular iodine, a cyclic process can be conceptualized in which HI is regenerated. One scheme might use hydrogen sulfide to regenerate HI. Another might use hydrogen. Thus,

ROH 4- HI —» RI 4- H20 RI 4- HI —> RH 4- I2 (I2 4- H2S -> 2HI 4- S)

(I2 + H2 ^ 2HI).

It appears that further research on HI chemistry could lead to processes for direct conversion of biomass to hydrocarbons without the economic penalty associated with operation at high pressure and temperature. The key to the value of such developments resides in the ability to recycle HI. Note that loss of only a small amount of the HI reacted can make the process quite uneconomical, so if it is developed to the point of commercial use, iodine recoveries would have to be substantially improved.

This process was developed in the 1970s and early 1980s to the pilot plant stage (0.6 m diameter by 12.2 m, rotating, inclined kiln) by Wright-Malta Corporation (Hooverman and Coffman, 1976; Coffman and Speicher, 1993). Since then, the process has been improved by using a stationary kiln having an internal rotor with vanes. Much of the development work was performed with a stationary kiln that is 0.3 m inside diameter by 3.7 m long. The process is reported to convert as-harvested green wood or any other wet biomass into medium-energy gas of heating value 15.7 to 19.6 MJ/m3 (n) in the self — pressurized kiln at pressures of about 2027 kPa at 590°C and residence times of about 1 hour. Steam is generated from the moisture in the feedstock and is normally not supplied to the kiln. No air or oxygen is used, and recycled wood ash serves as catalyst. As the biomass moves through the kiln from the cool feed end, it is gradually heated and partially dried, yielding steam. It then undergoes pyrolysis, yielding gas, liquids, tars, and char, all of which move cocurrently down the kiln where they undergo steam gasification and reforming to yield more gas. The inorganic residue is discharged at the hot end, and the hot gas is removed at the cold end after passage through heat-transfer coils in the kiln. The wood decomposition exotherm is reported to be sufficient to sustain the process after initial heat-up by an auxiliary boiler. Work in a small kiln showed that at pressures of 1378 to 2736 kPa and temperatures of 590 to 620°C with sodium carbonate catalyst, any type of green biomass can be gasified to 95 to 98% completion as long as it contains sufficient moisture. Dry gas compositions were about 5 to 10 mol % carbon monoxide, 40 to 50 mol % carbon dioxide, 15 to 22 mol % methane, and 20 to 28 mol % hydrogen. It was estimated that 907 t/day of green biomass at 11.6 MJ/kg would provide an output of 329 t/day of methanol.

Lipids are esters of the triol, glycerol, and long chain fatty acids. The fatty acids are any of a variety of monobasic fatty acids such as palmitic and oleic acids. The esters are formed in a large variety of oilseed crops, green plants, and some microalgae. Examples are soybean, cottonseeed, and corn oils. One pathway to the lipids produces glycerol and the other produces fatty acids, which can then combine to afford the triglycerides. In the chloroplasts, the Calvin-Benson cycle produces phosphoglyceric acid, which undergoes successive noncyclic photophosphorylation and isomerization to yield 3- phosphoglyceraldehyde and dihydroxyacetone phosphate. Three molecules of the aldehyde and one molecule of the phosphate are translocated out of the chloroplasts and combine to form fructose-1,6-diphosphate, which then suc­cessively undergoes a series of hydrolysis, isomerization, and condensation reactions to yield the disaccharide sucrose from glucose and fructose intermedi­ates. This pathway to glycerol involves reduction of dihydroxyacetone phos­phate to glycerol-1-phosphate. The fatty acids are derived from pyruvic acid formed on glycolyis of glucose. The pathway involves decarboxylation of pyru­vic acid, the formation of acetyl coenzyme A, which is involved in the synthesis (and breakdown) of fatty acids, and the buildup of fatty acid chains by insertion of two-carbon units into the growing chain.

Abundance

Municipal wastewater treatment plants in industrialized countries receive wastewaters from residential sources, industry, groundwater infiltration, and stormwater runoff. The pollutants associated with these sources include a wide range of suspended and dissolved compounds and oxygen-demanding materials, many of which are toxic. Pathogenic components are present, includ­ing certain bacteria, viruses, organic compounds, inorganic nutrients, and heavy metals. The purpose of most wastewater treatment processes is to remove or reduce these components, other pollutants, and biological oxygen demand before discharge to receiving waters. In the 1970s and 1980s, about 70 to 75% of the U. S. population was served by wastewater treatment facilities (U. S. Environmental Protection Agency, 1985; 1990). In 1992, more than 20,000 treatment and collection facilities served 180.6 million people or 71% of the population (U. S. Environmental Protection Agency, 1993). Of the 20,000 facili­ties, 15,613 provided treatment; the design capacity was 149 billion L/day. The need for new wastewater treatment capacity is expected to increase with growth of the sewered population. The need to treat 172 billion IVday is projected for 2012.

Primary biosolids (settleable and suspended solids) are present at a level of a few percent in the influent wastewater and are produced at a rate of about 0.091 dry kg/person-day (0.20 dry lb/person-day). Per million population, this corresponds to the production of 33,200 t/year of primary biosolids. After conventional primary and secondary wastewater treatment, the digested, dewa­tered biosolids are reduced to the equivalent of about 0.063 dry kg/person — day (0.14 dry lb/person-day), or 23,000 dry t/year per million population. For the United States in 1995, primary and treated biosolids production were about 8.6 and 5.9 million dry tonnes.

About 40 to 45% of the treated biosolids are disposed of in municipal landfills, 30% is applied to land or distributed or marketed as fertilizer, 20% is incinerated, and the remainder is disposed of in dedicated landfills or by a few other methods.

Availability

Again, there is no question of the physical availability of biosolids. They are collected in municipal wastewater systems and are therefore available in centralized locations. But in this case, treatment is essential for health reasons and protection of the public. Unless processes exist that can be used to treat and stabilize the waste and at the same time recover energy, it does not make much sense to use untreated biosolids as a waste biomass feedstock. In fact, such processes exist and will be discussed in some detail in later chapters. The other option to consider is the utilization of treated biosolids as a waste biomass.