The largest commercial RDF gasification plant in Europe is believed to be the system built in 1992 in Italy that produces low-energy gas at a feed rate of 180 t/day of RDF, which is obtained from 600 t/day of MSW (Barducci et ah, 1995, 1996). The hot product gas from the fluid-bed gasifiers is burned in an on-site boiler or is used as industrial fuel. The flue gas from the boiler is cleaned in a three-stage dry scrubber system before being exhausted through the stack. The steam raised in the boiler drives a 6.7-MW condensing steam turbine. Alternatively, the product gas is supplied as fuel to a neighboring cement plant. When the plant is eventually completed by addition of other gasification units, the facility will have the capacity to process 1300 t/day of MSW. The gasification system for this plant was developed by TPS Termiska Processer AB (Morris, 1996). It consists of two circulating fluid-bed gasifiers, each of 54 GJ/h capacity. A downstream cleanup process for the hot product gas is expected to be installed in future plant additions. In this process, the tars in the product gas are catalytically cracked at about 900°C in a dolomite — containing vessel located immediately downstream of the gasifier, and the particulates and alkalis are removed. This is expected to eliminate equipment contamination and filter clogging by tar condensation as the product gas is cooled. The TPS technology is expected to be used for similar projects in The Netherlands and the United Kingdom, and also for the biomass-fueled, integrated gasification, combined-cycle (BIGCC) power plant in Brazil. The B1GCC plant is expected to demonstrate the commercial viability of producing electric power from eucalyptus in an advanced technology plant of 30 MW capacity (see Section V, E).

Another large source of renewable carbon supplies is waste biomass. It consists of a wide range of materials and includes municipal solid wastes (MSW), municipal biosolids (sewage), industrial wastes, animal manures, agricultural crop and forestry residues, landscaping and tree clippings and trash, and dead biomass that results from nature’s life cycles. Several of these wastes can cause serious health or environmental problems if they are not disposed of properly. Some wastes such as MSW can be considered to be a source of recyclables such as metals and glass in addition to energy. Thus, waste biomass is a potential energy resource in the same manner as virgin biomass.

To assess the potential impact of energy from waste biomass on supplying energy demand, it is necessary to consider the amounts of the different types of wastes generated, their energy contents, and their availabilities. Every person in the United States, for example, discards about 2.3 kg (5 lb) of MSW per day. From an energy standpoint, one short ton of MSW has an as-received energy content of about 9.5 GJ (9.0 X 106 million Btu), so about 2.2 EJ/year (2.1 quad/year) of energy potential resides in the MSW generated in the United States.

As for the amount of energy that can actually be recovered from a given waste and utilized, much depends on the waste type. The amount of available MSW, for example, is larger than the total amounts of available agricultural wastes even though much larger quantities of agricultural wastes are generated. This is caused by the fact that a larger fraction of MSW is collected for centralized disposal than the corresponding amounts of agricultural wastes, most of which are left in the fields where generated. The collection costs are prohibitive for most of these wastes. Note that municipal biosolids on a dry solids basis is generated in the smallest quantity of all wastes. Its disposal, however, is among the most costly and difficult of all waste treatment opera­tions.

Many studies have been carried out to estimate the potential of available virgin and waste biomass as energy resources. One is presented in Table 2.7 for the United States for the year 2000 (Klass, 1990). The estimated energy potential of the recoverable materials is about 25% of the theoretical maximum. Wood and wood wastes are about 70% of the total recoverable energy potential and 50% of the estimated maximum energy potential. These estimates of virgin and waste biomass energy potential are based on existing, sustainable biomass production and do not include new, dedicated biomass energy plantations that might be developed and placed in commercial operation.

An assessment of the energy potential of waste biomass that is more localized can often provide better leads for the development of biomass energy supplies. The results of one such preliminary study performed for the state of Indiana are summarized in Table 2.8 (Klass, 1981). Indiana is a farm state. More than 60% of the state area was devoted to cropland at the time of the study and about 52% of state area was under active cultivation. The major agricultural crop and farm animal wastes as well as forestry and municipal wastes were therefore selected for the assessment of waste biomass energy potential. The waste biomass generated in the state each year was first inventoried, and each waste was then converted to gross energy content using generic conversion factors as a first approximation of energy potential. Comparison of the results with annual commercial energy utilization in the form of petroleum motor

TABLE 2.7 Potential Biomass Energy Available in United States in 2000“

“Klass (1990). The energy values are the higher heating values of the indicated biomass or derived methane. The conversion of biomass or methane to another biofuel or to steam, heat, or electric power requires that the process efficiency be used to reduce the potential energy available. These figures do not include additional biomass that could be grown as a dedicated energy crop.

fuels indicated that grain crop residues, particularly corn and soybean residues, and cattle manures have the largest potential as feedstocks for conversion to substitute motor fuels. Most of the other wastes are generated in insufficient quantities to make a large contribution. This simple assessment provided direction to the initiation of programs to develop systems using waste biomass feedstocks generated in the state of Indiana.

An example of a different type of assessment of waste and virgin biomass energy potential is one performed for the state of Wisconsin, another farm state in the Corn Belt of the United States. This assessment evaluated the economic impacts of shifting a portion of Wisconsin’s future energy investment from imported fossil fuels toward renewable energy resources. It assumed a 75% increase in the state’s renewable energy use by 2010—775 MW of new electric generating capacity to supply electricity to 500,000 Wisconsin homes, and 100 million gallons per year of new ethanol production to supply gasohol (blends of 10 vol % ethanol and 90 vol % gasoline) to 45% of Wisconsin’s automobiles (Clemmer and Wichert, 1994). This scenario generated about three times more jobs, earnings, and output (sales) in Wisconsin than the same level of imported fossil fuel usage and investment, and was equivalent to 63,234 more job-years of net employment, $1.2 billion in higher wages, and $4.6 billion in additional output. Over the operating life of the technologies

analyzed, about $2 billion in avoided payments for imported fossil fuels would remain in Wisconsin to pay for the state-supplied renewable resources, labor, and technologies. Wood, corn, and waste biomass contributed 47% of the increase in net employment.

This review of the concept of utilizing biomass energy shows that when sufficient supplies of renewable carbon are available, virgin and waste biomass have the potential of becoming basic energy resources. Presuming that suitable conversion processes are available, and that the demand for energy and estab­

lished organic fuels and intermediates continues, an industry based on renew­able biomass fuels and feedstocks that can supply a significant portion of this demand is, at the very least, a technically feasible concept.

A. Effects of Fossil Fuel Prices

The practical value of biomass energy ultimately depends on the costs of salable energy and biofuels to the end users. Consequently, many economic analyses have been performed on biomass production, conversion, and integrated bio­fuels systems. Conflicts usually abound when attempts are made to compare the results developed by two or more groups for the same biomass feedstock or biofuel because the methodologies are not the same. The assumptions made by each group are sometimes so different that valid comparisons cannot be made even when the same economic ground rules are employed. Comparative analyses, especially for hypothetical processes conducted by an individual or group of individuals working together, should be more indicative of the eco­nomic performance and ranking of biomass energy systems. However, several generalizations can be made that are quite important. The first is that fossil fuel prices are well documented and can be considered to be the primary competition for biomass energy. Table 4.13 summarizes U. S. tabulations of average, consumption-weighted, delivered fossil fuel prices by end-use sector in the mid-1990s (U. S. Energy Information Administration, December 1995). It is evident that the delivered price of a given fossil fuel is not the same to each end-use sector. The residential sector normally pays more for fuels than the other sectors, and the large end users pay less.

In the context of virgin biomass energy costs, dry woody and fibrous biomass species have an energy content on a dry basis of approximately 18.5 MJ/kg (7959 Btu/lb) or 18.5 GJ/t (16 MBtu/ton). For comparison purposes, if such types of biomass were available at delivered costs of $1.00/GJ ($1.054/MBtu), or $18.50/dry t ($16.78/dry ton), biomass on a strict energy content basis without conversion would cost less than most of the delivered fossil fuels listed in Table 4.13. The U. S. Department of Energy has set cost goals of delivered virgin biomass energy crops at $1.90-2.13/GJ ($2.00-2.25/MBtu), which corresponds to $35.15 to 39.41/dry t of virgin biomass (Fraser, 1993).

In the mid-1990s, few virgin biomass species were grown and harvested in the United States specifically for energy or conversion to biofuels, with the possible exceptions of feedstocks for fuel ethanol and a few tree plantations. This is not difficult to understand from an economic standpoint, especially if conversion costs are included. The nominal price of natural gas in the United

States in 1994 at the wellhead (not end-use cost) was estimated to be $1.74/ GJ ($1.83/MBtu) (U. S. Energy Information Administration, July 1995). For virgin biomass to compete as a feedstock for methane production on an equiva­lent basis, it would have to be grown, harvested, and gasified to produce methane at the same or lower cost. Assuming a gasification cost of zero and biomass conversion to substitute natural gas at 100% thermal efficiencies, both assumptions of which are totally unrealistic but which will help illustrate the best-case economics, the maximum market price of the biomass feedstock cost at the conversion plant gate including profit is then $32.19/dry t (at 18.5 GJ/ dry t X $1.74/GJ). At an optimistic yield of 22.4 dry t/ha-year (10 dry ton/ac — year), the biomass producer who supplies the gasification plant with feedstock would then realize not more than $721/ha-year ($292/ac-year), a marginal amount to permit a net return on an energy crop without other incentives. Similar calculations for the production and conversion of virgin biomass to liquid petroleum substitutes at zero conversion cost and 100% thermal effi­ciencies at the average U. S. nominal wellhead price of crude oil of $13.19/bbl ($2.234/GJ) in 1994 (U. S. Energy Information Administration, July 1995) correspond to a maximum market price for virgin biomass feedstock of $41.33/ dry t ($37.49/dry ton). The average price of hay, for example, received by farmers across the United States in 1994 was $95.59/t ($86.70/ton) (U. S. Dept, of Commerce, 1996). This indicates that the production of hay, and probably most grasses, as energy crops for conversion to liquid biofuels in direct competi­tion with petroleum liquids was not economically feasible at that time. These simplistic calculations emphasize the effect of fossil fuel prices on dedicated biomass energy crops. Inclusion of gasification or liquefaction costs and conver­sion efficiency factors by the processor would result in still lower market prices that the processor would be willing to pay for biomass feedstocks. Negative feedstock costs (wastes), substantial by-product credits, captive uses, other markets and uses, environmental credits, and/or tax incentives would be needed to justify dedicated energy crop production on strict economic grounds.

The production of virgin biomass for food and feed has progressed from very labor-intensive, low-efficiency agricultural practices over the 1800s and 1900s to what some consider to be a modern miracle. The invention of numerous agricultural machines in the late 1700s and 1800s that can seed the earth and reap the harvests with minimal labor and energy inputs made it possible to continuously produce biomass in quantity to help meet the massive demand for foodstuffs and other farm products caused by the growing population. In the United States today, only a few percent of the population living on farms is sufficient to produce enough food to meet all the nation’s demands for foodstuffs as well as supply surplus amounts for export. Farm equipment is available so that almost all row and grain crops can be continuously planted and harvested and separated into foodstuffs, feed, and residual materials. Eli Whitney’s cotton gin and Cyrus McCormick’s reaper are just two of the devices that helped mechanize agriculture and change the course of history by provid­ing non-labor-intensive methods of physically separating the desired products, cotton and grain for these particular inventions, from biomass. As candidate energy crops evolve, such as several of the thick-stemmed grasses that are difficult to harvest at high growth densities, new agricultural equipment designs and adaptations of existing machinery are expected to solve these problems also.

Simultaneously with the advancement of agriculture, although not via the same pathway, new hardware and improved methodologies were developed

for the planting, managing, and harvesting of trees that made large-scale com­mercial forestry operations more economic and less dependent on labor. Better methods of land clearing, thinning, and growth management, and improved hardware for harvesting, such as feller-bunchers, which were first used in the early 1970s, resulted in a modern forest products industry that supplies commercial and industrial needs for wood and wood products. As the use of trees for energy and feedstocks expands, it is expected that much of the existing commercial hardware and improvements will be applied to meet these needs.

Equipment and methods for the harvesting of the smaller short-rotation woody crops at low cost are also expected to be developed. Much of the ongoing work to design improved equipment for SRWC is directed to feller — bunchers that perform severing, bunching, and off-loading functions. The results from systems analysis studies indicate that prototype feller-buncher harvesters can be balanced with two or three small grapple skidders to move bunched SRWC to a landing for chipping or just loading in the case of whole trees (Perlack et al, 1996). However, because of the high skidding costs, a whole-tree, direct-load system for use with a track-type feller-buncher is pre­ferred.

A few of the nonmanual separation methods used for woody biomass pro­cessing that have use in energy applications are briefly described here. Delimb — ing and debarking of trees is an old technology. For the smaller trees where fiber in the form of white wood chips is the desired product, the trees can be debarked and delimbed by the use of chain flails, which remove the outer bark layer, leaving the white wood behind. Hammermilling then yields a homogeneous product (Hudson and Mitchell, 1992). In most thermochemical energy applications, however, separation of the bark and wood is not necessary. But where it is necessary to remove the bark, some efforts have been made to recover the residues for fuel from flail machines by using them together with tub grinders (Stokes, 1992). A tub grinder operating simultaneously with a chain flail was successfully used to comminute the residues (Baughman, Stokes, and Watson, 1990). The green weight of the fuel residues was about one — fourth to one-third of the total clean chip-plus-fuel weight.

In a few installations that bum hogged wood, disc and shaker screens have been employed to separate preselected, oversize pieces for subsequent size reduction and return to the fuel stream. Finely divided wood fuels such as sawdust and sanderdust are also sometimes screened to remove the larger pieces.

By-product hulls from the production of rice, cotton, peanut, soybean, and similar crops that have outer shells covering small seed or fruit are sometimes used directly as fuels or feedstocks. After the shells are fractured, most of the hulls can be separated with vibrating screens or rotating trommels having appropriately sized openings. The by-product hulls that have high ash contents

and bulk densities present a few difficulties on direct combustion or gasifica­tion, but specially designed systems are available to eliminate these problems (с/. King and Chastain, 1985; Bailey, 1990; Bailey and Bailey 1996).

A detailed comparative economic analysis of the production costs of fuel oil and synthetic gasoline and diesel fuel has been performed for direct liquefaction of biomass (Elliott et ah, 1990). To bring this analysis to today’s dollars, adjustments should be made to account for inflation and other factors. But since the treatment is a comparative analysis, the differences should remain the same. The basic parameters used for this analysis are listed in Table 8.14.

Three processing steps were involved in the assessment: liquefaction of wood — chip feedstock to a primary crude oil, catalytic hydrotreatment of the crude oil to a deoxygenated product oil, and refining of the product oil to gasoline or diesel fuel. Atmospheric flash pyrolysis (cf. Scott and Piskorz, 1983; O’Neil, Kovac, and Gorton, 1990) and pressurized solvent liquefaction (с/. Appell et al, 1975; Thigpen and Berry, 1982) were analyzed. Two versions of each process were used—one based on the technology as developed (present tech­nology) and one based on anticipated future improvements (potential tech­nology).

The atmospheric flash pyrolysis process for the present technology case is illustrated in the accompanying flowsheet (Fig. 8.5). One-millimeter particle size wood fibers are rapidly pyrolyzed in a fluidized-bed reactor at 500°C to vapors and char. The condensed vapors form the primary oil product, which contains approximately 39% oxygen on a dry basis. The second and third steps are not shown in Fig. 8.5. In the second step, the primary oil is upgraded in a two-stage catalytic hydrotreatment process using a conventional sulfided cobalt/molybdenum-on-alumina petroleum hydrotreatment catalyst. In the third step, the upgraded product is subjected to distillation, hydrodeoxygen­ation of the light fraction, hydrocracking of the heavy fraction, catalytic reform­ing of the gasoline fraction, and steam reforming of the hydrocarbon gas product to produce hydrogen for the process. A gasoline and diesel product slate is produced. For the potential technology case, pyrolysis takes place in a circulating fluidized-bed reactor, which has the advantage of much greater throughput than the present technology case. An advanced three-stage catalytic hydrotreatment is used in the second step and is assumed to yield a high- octane gasoline requiring no further processing other than fractionation.

The present technology case for pressurized solvent liquefaction is illus­trated in Fig. 8.6. Wood chips are ground to less than 0.5 mm and mixed with recycled wood-derived oil. A sodium carbonate solution and synthesis gas are added to the slurry prior to preheating. Liquefaction takes place in a tubular, upflow reactor at 350°C, 20.5 MPa, and a 20-min residence time. Gas is flashed from the reactor effluent. A portion of the liquid is recycled. Water is separated from the primary oil and is treated before discharge. The synthesis gas is obtained from a portion of the feedstock, which is gasified in an oxygen-blown gasifier. The product oil is upgraded and refined in a manner similar to the flash pyrolysis oil. The potential case for this process uses an extruder feeder to feed high concentration wood slurries. The oil phase of the slurry consists of recycled vacuum distillate bottoms. Superheated steam is added to the reactor to provide the reactor heat requirement. Sodium carbonate is assumed to be recycled entirely in the distillate bottoms, and no reducing gas is added to the reactor. The liquid product stream is separated into a distillate product and recycled bottoms in a vacuum distillation tower. Catalytic hydrotreatment

Atmospheric flash
pyrolysis

“Adapted from Elliott et al. (1990). The product energy value is the average U. S. spot market price from 1977 to 1987 for comparable petroleum-based liquid fuel. The average price of U. S. crude oil at the wellhead in 1987 was S2.61/GJ ($15.40/bbl).

is used to upgrade the primary oil as in the present technology case. The upgraded product is assumed to be a high-octane gasoline requiring no further processing other than fractionation.

Detailed estimates were prepared of the capital and operating costs of each of the four processes, which were designed around the results obtained from laboratory, PDU, and pilot-scale tests. A summary of these costs is presented in Table 8.15. The atmospheric flash pyrolysis process is clearly more economi­cal than the pressurized solvent liquefaction process for production of similar products. Using the average U. S. spot market price for fuel oil in the period 1977 to 1987, the ratio of primary oil cost to energy value is about 1.4 at a green wood chip price of $30/t for the present technology case for atmospheric flash pyrolysis. This means that the product is 40% more costly than comparable petroleum fuel. The potential technology case for atmospheric flash pyrolysis is the only process design of the four analyzed that appears capable of producing primary oil product competitive with comparable petroleum fuel. On the basis of sensitivity studies, each process appears to be more sensitive to feedstock cost than to capital cost. Although thermochemical conversion processes are generally capital intensive, the range of capital costs examined had little effect on the final product cost. However, sensitivity analysis of finished product cost and feedstock price showed that when the green wood chip price is $10/t, the ratio is less than 0.9.

It was concluded from this assessment that the most promising process for gasoline production by direct liquefaction of biomass is atmospheric flash pyrolysis. The high-pressure process may have the same future potential, but the uncertainties are much greater.

A dual fluid-bed process for biomass was developed in the United States in the 1970s and early 1980s by the EEE Corporation (Bailie, 1980, 1981). It was commercialized in Japan by EBARA Corporation. This process, called the Bailie process after its inventor, consists of two circulating fluid-bed reactors that permit the use of air instead of oxygen for conversion of biomass to medium-energy gas. In one bed, feedstock combustion occurs with air to heat the sand bed. The hot sand is circulated to the other reactor where steam gasification of fresh feed and recycled char occurs. The cooled sand is recircu­lated to the combustion reactor for reheating. This configuration produces a product gas with a heating value of 11.8 MJ/m3 (n) or more. The composition of the gas from the gasifier operated at 650 to 750°C in one of the pilot plants fed with RDF in Japan was 34.7 mol % carbon monoxide, 11.2 mol % carbon dioxide, 12.7 mol % methane, 8.0 mol % other hydrocarbons, 30.0 mol % hydrogen, 2.5 mol % nitrogen, and 0.9 mol % oxygen. The heating value was 17.6 MJ/m3 (n). The pilot plant data indicated that 60% of the carbon resided in the medium-energy gas, 30% was converted to char, and the remaining 10% formed liquid and char. The energy yield as medium-energy gas was between 50 and 60%. The plants operated with RDF feedstocks in Japan were a 36-t/ day pilot plant, a 91-t/d demonstration plant, and a 408-t/day commercial plant. These plants have been shut down.

Indirectly Heated, Dual Fluid-Bed, Steam Gasification

This process was developed by Battelle in the 1980s in a dual-bed PDU having a capacity of 20 to 25 t/day (Paisley, Feldmann, and Appelbaum, 1984; Paisley, Litt, and Creamer, 1991). Heat is supplied by recirculating a stream of hot sand between the separate combustion vessel and the gasifier. The PDU used a conventional fluid-bed combustor. In a commercial plant, both the gasifier and the combustor would be operated in the entrained mode to achieve higher throughputs. Tests have been conducted with wood and RDF. The operating ranges of the gasifier in the PDU were 630 to 1015°C at near-atmospheric pressure. The largest gasifier used was 0.25 m inside diameter and had a maximum wood throughput of 1.7 t/h. The heating value of the product gas was 17.7 to 19.6 MJ/m3 (n) and was reported to be independent of the moisture level of the feed. A thermally balanced operation with wood feedstock was achieved at throughputs of 1.5 t/h. Combustor carbon utilization was complete at temperatures above 980°C, and gasifier carbon conversion to gas was 50 to 80% at temperatures above 705°C. Typical nitrogen-free gas compositions were 50.4 mol % carbon monoxide, 9.4 mol % carbon dioxide, 15.5 mol % methane, 7.2 mol % ethane and ethylene, and 17.5 mol % hydrogen. Carbon conversions with RDF were similar to those of wood over a temperature range of 650 to 870°C. The heating values of the product gases were about 21.6 to 23.6 MJ/ m3 (n). A commercial plant based on this process has been built to supply

fuel gas to a central station power plant in Vermont (Paisley and Farris, 1995; Farris and Weeks, 1996).

Major differences in net photosynthetic assimilation of C02 are apparent be­tween C3, C4, and CAM biomass species. Biomass species that fix C02 by the C4 pathway usually exhibit higher rates of photosynthesis at warm tempera­tures. At cool temperatures, C3 biomass species usually have higher rates of photosynthesis. Plants that grow well in the early spring such as forage grasses and wheat, are all C3 species, whereas many desert plants, tropical species, and species originating in the tropics such as sugarcane are C4 plants. One of the major reasons for the generally lower yields of C3 biomass is its higher rate of photorespiration. If the photorespiration rate could be reduced, the net yield of biomass would increase. Considerable research has been done to achieve this rate reduction by chemical and genetic methods, but only limited yield improvements have been made. Such an achievement if broadly applicable to C3 biomass would be expected to be very beneficial for both the production of foodstuffs and biomass energy. Another advancement that will probably evolve from research concerns increasing the yields of the secondary derivatives such as the liquid terpene hydrocarbons and triglycerides produced by certain biomass species to make direct fuel production from biomass practical and economically competitive. Detailed study and manipulation of the biochemical pathways involved will undoubtedly be neccessary to achieve some of these improvements, particularly since most of the advancements that have been made by controlling growth conditions and trying to select improved strains of biomass have not been very successful in increasing the natural production of liquid fuels.

The C02-fixing pathways used by a specific biomass species will affect the efficiency of photosynthesis, so from a biomass energy standpoint, it is desirable to choose species that exhibit high photosynthesis rates to maximize the yields of biomass in the shortest possible time. Obviously, however, there are numerous factors that affect the efficiency of photosynthesis other than the carbon dioxide-fixing pathway. Insolation; the amounts of available water and macronutrients and micronutrients; the C02 in the surrounding environment; the atmospheric concentration of C02, which is normally about 0.03 mol %; the temperature; and the transmission, reflection, and biochemical energy losses within or near the plant affect the efficiency of photosynthesis. For lower plants such as the green algae, many of these parameters can be controlled, but for conventional biomass growth that is subjected to the natural elements, it is not feasible to control all of them.

The maximum efficiency with which photosynthesis can occur has been estimated by several methods. The upper limit has been projected to range from about 8 to 15%, depending on the assumptions made, that is, the maxi­mum amount of solar energy trapped as chemical energy in the biomass is 8 to 15% of the energy content of the incident solar radiation. It is worthwhile to examine the rationale in support of this efficiency limitation because it will help to point out some aspects of biomass production as they relate to energy applications.

The relationship of the energy and frequency of a photon is given by

e = (he)/A,

where e = energy content of one photon, J; h = Planck’s constant, 6.626 X 10-34 J • s; c = velocity of light, 3.00 X 108 m/s; and A = wavelength of light, nm. Assume that the wavelength of the light absorbed is 575 nm and is equivalent to the light absorbed between the blue (400 nm) and red (700 nm) ends of the visible spectrum. This assumption has been made for green plants by several investigators to calculate the upper limit of photosynthesis efficiency. The energy absorbed in the fixation of 1 mol C02, which requires 8 photons per molecule, is then given by

Energy absorbed = (6.626 X 10_34)(3.00 X 108)(575 X 10~9)-1(8) (6.023 X 1023)

= 1.67 MJ (399 kcal).

Since 0.47 MJ of solar energy is trapped as chemical energy in this process, the maximum efficiency for total white light absorption is 28.1%. Further adjustments are usually made to account for the percentages of photosyntheti­cally active radiation in white light that can actually be absorbed, and respira­tion. The fraction of photosynthetically active radiation in solar radiation that reaches the earth is estimated to be about 43%. The fraction of the incident light absorbed is a function of many factors such as leaf size, canopy shape, and reflectance of the plant; it is estimated to have an upper limit of 80%. This effectively corresponds to the utilization of 8 photons out of every 10 in the active incident radiation. The third factor results from biomass respiration. A portion of the stored energy is used by the plant, the amount of which

depends on the properties of the particular biomass species and the environ­ment. For purposes of calculation, assume that about 25% of the solar energy trapped as chemical energy is used by the plant, thereby resulting in an upper limit for retention of the nonrespired energy of 75%. The upper limit for the efficiency of photosynthetic fixation of biomass can then be estimated to be 7.2% (0.281 X 0.43 X 0.80 X 0.75). For the case where little or no en­ergy is lost by respiration, the upper limit is estimated to be 9.7% (0.281 X 0.43 X 0.80). The low-efficiency limit might correspond to terrestrial biomass, while the higher efficiency limit might be closer to the efficiency of aquatic biomass such as unicellular algae. These figures can be converted to dry biomass yields by assuming that all of the C02 fixed is contained in the biomass as cellulose, -(C6H10O5)x -, from the equation

Y = (CIE)/F,

where Y = yield of dry biomass, t/ha-year; C = constant, 3.1536; I = average insolation, W/m2; E = solar energy capture efficiency, %; and F = energy content of dry biomass, MJ/kg.

Thus, for high-cellulose dry biomass, an average isolation of 184 W/m2 (1404 Btu/ft2-day), which is the average insolation for the continental United States, a solar energy capture efficiency of 7.2%, and a higher heat of combustion of 17.51 MJ/kg for cellulose, the yield of dry biomass is 239 t/ha-year (107 ton/ac-year). The corresponding value for an energy capture efficiency of 9.7% is 321 t/ha-year (143 ton/ac-year). These yields of organic matter can be viewed as an approximation of the theoretical upper limits for land — and water-based biomass. Some estimates of maximum yield reported by others are higher and some are lower than these figures, depending on the values used for I, E, and F, but they serve as a guideline to indicate the highest theoretical yields of a biomass production system. Unfortunately, real biomass yields rarely approach these limits. Sugarcane, for example, which is one of the high-yielding biomass species, typically produces total dry plant matter at yields of about 80 t/ha- year (36 ton/ac-year).

Yield is plotted against solar energy capture efficiency in Fig. 3.5 for insol­ation values of 150 and 250 W/m2 (1142 and 1904 Btu/ft2-day), which span the range commonly encountered in the United States, and for dry biomass energy values of 12 and 19 MJ/kg (5160 and 8170 Btu/lb). The higher the efficiency of photosynthesis, the higher the biomass yield. But it is interesting to note that for a given solar energy capture efficiency and incident solar radiation, the yield is projected to be lower at the higher biomass energy values (curves A and C, curves В and D). From a gross energy production standpoint, this simply means that a higher-energy-content biomass could be harvested at lower yield levels and still compete with higher-yielding but lower-energy — content biomass species. It is also apparent that for a given solar energy capture

efficiency, yields similar to those obtained with higher-energy-content species should be possible with a lower-energy-content species even when it is grown at lower insolation (curves В and C). Finally, at the solar energy capture efficiency usually encountered in the field, the spread in yields is much less than at the higher-energy-capture efficiencies. It is important to emphasize that this interpretation of biomass yield as functions of insolation, energy content, and energy capture efficiency, although based on sound principles, is still a theoretical analysis of living systems that can exhibit unexpected be­havior.

Because of the many uncontrollable factors, such as climatic changes and the fact that the atmosphere only contains 0.03 mol% C02, biomass production outdoors generally corresponds to photosynthesis efficiencies in the 0.1 to 1.0% range. Significant departures from the norm can be obtained, however, with certain plants such as sugarcane, napier grass, algae, maize, and water hyacinth (Tables 3.1 and 3.2). The average insolation values at the locations corresponding to the biomass growth areas listed in Table 3.1 were used to calculate solar energy capture efficiency at the reported annual dry yields. Other than insolation, all environmental factors and the nonfuel components in the biomass were ignored for the solar energy capture efficiency estimates listed in Table 3.1 and it was assumed that all dry matter is organic and has an energy content of 18.6 MJ/dry kg (16 million Btu/dry ton). There is still a reasonably good correlation between dry biomass yield and solar energy capture efficiency in Table 3.1. The estimates of the efficiencies are only approximations and most are probably higher than the actual values. They indicate, however, that C4 biomass species are usually better photosynthesizers than C3 biomass species and that high insolation alone does not necessarily correlate with high biomass yield and solar energy capture efficiency. The biomass production data shown in Table 3.2 are some of the high daily rates of biomass photosynthesis reported for the indicated species. It has been estimated that water hyacinth could be produced at rates up to about 150 t/ha-year (67 ton/ac-year) if the plant were grown in a good climate, the young plants always predomi­nated, and the water surface was always completely covered (Westlake, 1963). Some evidence has been obtained to support these estimates (McGarry, 1971; Yount and Grossman, 1970). Unicellular algae, such as the species Chlorella and Scenedesmus, have been produced by continuous processes in outdoor light at high photosynthesis efficiencies (Burlew, 1953; Enebo, 1969; Kok and Van Oorschot, 1954; Oswald, 1969). Growth rates as high as 1.10 dry t/ha-day have been reported for Chlorella (Retovsky, 1966). In tropical climates, this rate might be sustainable over most of the year, in which case the annual yield might be expected to approach 401 dry t/ha — year. This yield range is beyond the theoretical upper limit estimated here

“Dry matter yield data from Berguson et al. (1990), Bransby and Sladden (1991), Burlew (1953), Cooper (1970), Loomis et al. (1963,1971), Lipinsky (1978), Rodin and Brazilevich (1967), Sachs et al. (1981), Sanderson et al. (1995), Schneider (1973), Westlake (1963).

from the basic chemistry of photosynthesis. It will be shown in later chapters that there are many species of biomass that can be grown at sufficiently high yields in moderate climates to make them promising candidates as biomass energy crops.

As indicated in the discussion of the chemistry of photosynthesis, PS II alone instead of both PS II and I is sufficient for photosynthesis of a green algal mutant under certain conditions (Greenbaum et al, 1995). The investigators suggested that the maximum thermodynamic conversion efficiency of light energy into chemical energy can be potentially doubled because a single photon rather than two is required to span the potential difference between water oxidation/oxygen evolution and proton reduction/hydrogen evolution. Com­parison of the experimental results from the wild strain that contained both PS II and I and the mutant indicated this did not occur because the quantum efficiencies are similar. Also, the phenomenon was not observed under aerobic conditions. But this research still suggests that yields can be improved if the single-photon system can be incorporated into other biomass that grows under atmospheric conditions. For example, such biomass might exhibit higher pro­ductivity not only because of more efficient usage of the solar energy that is absorbed, but also because more C02 could be fixed at lower insolation values due to longer equivalent growth times during the growing season. However, it has been suggested that since both PS II and PS I are required for photosynthe­sis under normal aerobic conditions, the validity of the Z-scheme remains secure (Barber, 1995).

The U. S. Department of Agriculture’s 1938 Yearbook of Agriculture contains these statements: “One billion tons of manure, the annual product of livestock on American farms, is capable of producing $3,000,000,000 worth of increase in crops. The potential value of this agricultural resource is three times that of the nation’s wheat crop and equivalent to $440 for each of the country’s 6,800,000 farm operators.” Since then, animal wastes have been transformed from a definite asset to a liability. By 1965, the disposal of animal excreta had become a serious problem (American Chemical Society, 1969). At any given time, an estimated 11 million cattle were on feedlots, the capacities of which ranged from 1000 to 50,000 head.

The problem has become much more severe today. Application of animal wastes to land is one of the most economical choices for disposal as well as providing fertilizing benefits. However, the utilization of livestock and poultry manures as waste biomass resources for energy applications could help mitigate pollution and at the same time open new markets. This possibility is examined here. Agricultural crop residues are included in the assessment.

Before discussion of advanced, biomass combustion systems, it is in order to consider electric power generation with biomass fuels because several advanced technologies are being used or are planned for this application. A typical utility boiler consists of a furnace, where heat is transferred to enclosed water-cooled tubes, and a convection section, where more heat is transferred to the water tubes. Steam superheating can occur, and various economizers and recupera­tors may be installed. The steam is produced at rates of about 100 to 4500 t/ h and converted to electric power in high-speed steam turbine generators, which range in capacity from about 20 to 1300 MW. In the United States, a very high percentage of electric power generation by utilities is by turbine — generator systems in which steam is expanded in variations of the Rankine cycle (Miller and Allen, 1985). This cycle, originally developed with steam engines, closely approximates the Carnot cycle when used with low-pressure steam. As pressures increase to obtain higher saturated steam temperatures, the Rankine cycle does not improve as much as the Carnot cycle because the low-temperature heat being added to bring the condensate back to boiler saturation temperature becomes a major portion of the total heat content of the saturated steam. But with regenerative feedwater heating and reheating the steam after it has been partially expanded through the turbine, Rankine efficiencies can approach Carnot efficiencies. Overall thermal efficiencies for power production usually range from about 28 to 34%. Some plants have been reported to operate at up to 40% overall efficiencies. The thermal energy in terms of fuel consumption needed to generate 1000 kWh of electricity is assumed in most U. S. tabulations in non-SI units to be about 1.8 bbl of crude oil, 0.47 ton of coal, 0.6 ton of dry biomass, or 10,000 ft3 of natural gas. This is equivalent to thermal energy consumption of about 11 MJ/kWh (10,400 Btu/kWh). Fossil-fueled steam-electric plants typically use about 10.5 to 12.7 MJ (10,000 to 12,000 Btu) of fuel input per kilowatt-hour gen­erated.

At full load, one of the largest single-boiler, stoker grate, wood-fueled, electric utility plants—a central station power plant in Burlington, Vermont that generates 50 MW of net production—consumes dry equivalent wood at a rate of 925 t/day (Tewksbury, 1987). Net electrical production was re­ported to be 280,137,900 kWh for a total green wood fuel consumption of 394,612.9 tonne (435,060.7 ton) at 50 wt % moisture over a 1-year period. At an average energy density of 18.6 GJ/t (16.0 million Btu/ton) for dry wood, this corresponds to thermal energy consumption of 13.10 MJ/kWh (12,424 Btu/kWh) generated and a thermal efficiency of 27.5%. The opportuni­ties for fuel savings in conventional electric power generation facilities are obvious. Note that at 100% resistance heating efficiencies, 1.0 kWh is equiva­lent to 3.60 MJ (3412 Btu) of thermal energy independent of the generating process.

Modem fossil-fired plants typically have capacities from 300 to 900 MW; 600 MW is the approximate average for U. S. utilities (Miller and Allen, 1985). Some plants have been built with capacities of 1300 MW. Steam conditions have effectively been standardized at 16,500 kPa and initial temperatures of 538°C with reheat to 538°C. Some plants utilize supercritical pressures of approximately 24,100 kPa, mostly with steam temperatures at 538°C/538°C. Some plants also utilize double-reheat and steam temperatures up to 565°C. A few advanced plants were designed to operate with steam pressures up to 34,500 kPa and steam temperatures up to 650°C. The net heat rate and the labor cost and investment per kilowatt-hour decrease with increasing plant size, so larger plants are desirable.

Biomass-fired boilers are typically limited to steam production rates up to 227 to 273 t/h (250 to 300 ton/h) according to some analysts because of fuel availability, fuel cost considerations, and materials handling difficulties associated with low-density fuels (Tillman, 1985). This restriction in turn limits the maximum economical pressure to about 10,300 kPa compared to coal-fired units, which range from 16,500 to 24,100 kPa, increases the steam rate requirement, and limits the number of feedwater heaters to 1 to 4, com­pared to the 8 feedwater heaters commonly associated with fossil-fired units. The characteristics of biomass power plants shown in Table 7.4 illustrate how these limitations can affect the technology. Biomass-fired cogeneration power plants usually have capacities in the range 5 to 25 MW, whereas condensing power plants have capacities up to 60 MW. Cogeneration is the simultaneous conversion of thermal energy into electrical energy and some other form of

energy. For example, steam produced in a boiler drives a steam turbine to generate electric power, and the waste heat is recovered and used for heat or process steam production. The overall thermal efficiency is higher because of the recovery of additional, useful energy. From a practical standpoint, the availability of fuel at a sustainable, competitive price is probably the most important factor that determines plant size.

The 50-MW plant in Burlington, Vermont, was limited in capacity by the wood fuel available within the area circumscribed by a radius of 80 km (50 mi.) from the plant. This is considered by most energy specialists to be the maximum distance that wood fuel can be obtained and economically transported to the plant by truck or rail. For captive sources of biomass fuels, the capacity can be larger. One example is the 60-MW, wood waste-fueled power plant located in Williams Lake, British Columbia (Baker, 1995). This plant is located in the center of a major lumber industry region that has five large sawmills located within 5 km of each other. The mills produce more than 540,000 green tonnes of bark, sawdust, and other wood waste products per year.

Efficiency improvements in the conversion of thermal energy to electric power are a direct route to increasing power plant capacity. Several techniques have been developed that offer large improvements in efficiency. Among them is the combined cycle configuration. In one configuration of a combined cycle plant, a combustion turbine drives a generator and the hot exhaust is fed to a heat recovery steam generator. The steam from this unit drives a steam turbine generator and the exhaust is used to provide process steam or is condensed and returned to the heat recovery steam generator. There are many variations of this design. Integrated gasification-combined cycle (IGCC) con­figurations for coal-fired systems are an example. IGCC systems are also appli­cable to biomass feedstocks and will be discussed in Chapter 9. The systems can be designed to operate at an overall energy conversion efficiency consider­ably larger than the sum of the efficiencies of separate systems that convert the same total quantity of fuel to electric power. Some projections indicate that overall thermal efficiencies as high as 70% might be possible.

Another approach to increasing power plant efficiencies is to use a nonther­mal conversion method for power production, such as fuel cells. Fuel cells rely on electrochemical conversion of the chemical energy in the fuel to electric power. In the cogeneration mode, these systems have been reported to be operable at overall efficiencies as high as 85% (Schora, 1991). Large-scale power plants based on fuel cells have not been developed yet and are not expected to be available for generating central station power until well into the twenty-first century.

The U. S. Department of Energy has developed a strategic plan that delineates how electric power generation from biomass can be significantly increased in

the U. S.A. (U. S. Dept, of Energy, 1996). The U. S. biomass power industry in the mid-1990s represented an investment base of $15 billion and supported about 66,000 jobs. DOE’s projections indicated the potential for biomass power to grow to an industry of 30,000 MW employing 150,000 persons in mainly rural areas and producing 150 to 200 billion kWh by the year 2020. This would require 127 million tonnes of dry biomass fuel annually according to DOE’s estimates, which is equivalent on a gross energy content basis to about 2.36 EJ/year and the annual gross generation of about 223 billion kWh at 85% availability and 33% overall thermal efficiency. If the required fuels were all dedicated biomass energy crops, 80,940 km2 (8.1 million ha) of growth area would be required at a conservative yield of 15.7 dry t/ha-year. Various strate­gies have been proposed to achieve the 30,000-MW target, including the cofiring of biomass fuels and coals as a bridging strategy. For example, the cofiring of wood wastes in coal-fired utility boilers has the potential to reduce fuel costs, support local economic development, and address environmental concerns (с/. Tillman et ah, 1995).

The most realistic approach to attainment of 30,000 MW of on-line biomass power in the United States within the next few decades is to develop large — scale, integrated biomass production-conversion systems that operate at high overall thermal and net energy production efficiencies (Chapter 14). This is perhaps the only practical approach, although efficiency improvements in power generation via advanced combined-cycle schemes and high-efficiency, nonthermal generation with fuel cells will help reduce the amounts of dedicated energy crops needed. Energy crop yields and costs are most certainly primary factors in achieving the 30,000-MW target.

A contrasting viewpoint is that large-scale, biomass-fueled power generation systems are unlikely to be economically competitive with natural gas or coal — fired generation, but that they can fill important niche markets, especially via distributed generation (Whittier, Haase, and Badger, 1996). Distributed generation is defined as any modular technology that is sited throughout a utility’s service area—interconnected to the distribution or subtransmission system—to lower the cost of service. They typically have capacities less than 50 MW. Distributed generation is claimed to provide multiple benefits to utilities and end users, including lower capital costs and reduced financial risk compared to those of the larger generation systems; deferral of upgrades to substations; provision of power in increments that match projected demand patterns; and various forms of grid support. Other advantages are that the logistics of sustaining operations are simplified and most of the biomass conver­sion technologies qualify as distributed generation candidates.

Assessments of commercial biomass power technologies indicate that oppor­tunities exist, particularly for niche market applications, when the business conditions are right. Federal legislation can have a large impact on these opportunities. For example, the U. S. Public Utilities Regulatory Policies Act of 1978 (PURPA, PL 95-617) created a utility market for independent, nonutility power producers by requiring public utilities to purchase power from them at the so-called avoided cost, or the utility’s cost of purchasing or generating the power itself. Many small power producers and cogenerators took advantage of this arrangement by generating power for on-site use and selling the surplus to the local utility. One technology that fared quite well under PURPA, when Standard Offer contracts in California allowed independent power producers to lock in payments that started at $0.08 to $0.09/kWh, is bubbling, fluid-bed combustion. It was a key technology that allowed plants to achieve favorable economics in a changing regulatory and fuel price environment. The flexibility of 3 such plants (net capacities of 10,10, and 25 MW), for example, permitted them to accept a very wide range of biomass fuels and to meet California’s strict emissions requirements using ammonia injection and limestone additive for NOx and SOx control over a 7-year operating history (Ferris, 1996). When scheduled shutdowns and reduced loads were required by the utility that purchased the power, the advanced designs of these plants made it possible for them to be operated as peaking units after the utility offered payments up to $0.06/kWh under curtailment contracts.

A. Introduction

The effort to develop and commercialize advanced biomass gasification systems is not nearly as extensive as the effort to develop coal gasification. However, considerable research and pilot plant studies have been carried out since about 1970 on biomass gasification for the production of fuel gases and synthesis gases (с/. Stevens, 1994). Several processes have been commercialized. Basic studies on the effects of various operating conditions and reactor configurations have been performed in the laboratory and at the PDU (process development unit) and pilot scales on pyrolytic, air-blown, oxygen-blown, steam, steam — oxygen, and steam-air gasification, and on hydrogasification. The thermal gasification of biomass in liquid water slurries has also been studied.

The chemistry of biomass gasification is very similar to that of coal gasifica­tion in the sense that thermal decomposition of both solids occurs to yield a mixture of essentially the same gases. But as pointed out in the Introduction, biomass is much more reactive than most coals. Biomass contains more volatile matter than coal, gasification occurs under much less severe operating condi­tions for biomass than for coal feedstocks, and the pyrolytic chars from biomass are more reactive than pyrolytic coal chars. The thermodynamic equilibrium concentrations of specific gases in the mixture depend on the abundance of carbon, hydrogen, and oxygen, the temperature, and the pressure. As in the case of coal feedstocks, increasing pressures tend to lower the equilibrium concentrations of hydrogen and carbon monoxide, and increase the methane and carbon dioxide concentrations. Also, as in the case of coal feedstocks, methane formation is favored at lower temperatures, and carbon monoxide and hydrogen are dominant at high temperatures. Biomass is gasified at lower temperatures than coal because its main constituents, the high-oxygen cellulos — ics and hemicellulosics, have higher reactivities than the oxygen-deficient, carbonaceous materials in coal. The addition of coreactants to the biomass system, such as oxygen and steam, can result in large changes in reaction rates, product gas compositions and yields, and selectivities as in coal conver­sion.

Biomass feedstocks contain a high proportion of volatile material, 70 to 90% for wood compared to 30 to 45% for typical coals. A relatively large fraction of most biomass feedstocks can be devolatilized rapidly at low to moderate temperatures, and the organic volatiles can be rapidly converted to gaseous products. The chars formed on pyrolytic gasification of most biomass feedstocks have high reactivity and gasify rapidly. Heat for pyrolysis is usually generated by combusting fuel gas either in a firebox surrounding the reaction chamber or in fire tubes inserted into the reaction chamber. As discussed in Chapter 8, chars, tars and oily liquids, gases, and water vapor are formed in varying amounts, depending particularly on the feedstock composition, heating rate, pyrolysis temperature, and residence time in the reactor. For biomass and waste biomass, steam gasification generally starts at temperatures near 300 to 375°C.

Undesirable emissions and by-products from the thermal gasification of biomass can include particulates, alkali and heavy metals, oils, tars, and aque­ous condensates. One of the high-priority research efforts is aimed at the development of hot-gas-cleanup methods that will permit biomass gasification to supply suitable fuel gas for advanced power cycles that employ gas turbines without cooling the gas after it leaves the gasifiers (International Energy Agency, 1991, 1992). It is important to avoid gas turbine blade erosion and corrosion by removing undesirable particulates that may be present. The re­moval of tars and condensables may also be necessary. Furthermore, utilization of the sensible heat in the product gas improves the overall thermal operating efficiencies. Nonturbine applications of the gas may also be able to take advan­tage of processes that provide clean, pressurized hot gas, such as certain down­stream chemical syntheses and fuel uses. Special filtration and catalytic systems are being developed for hot-gas cleanup. Some of the other research needs that have been identified include versatile feed-handling systems for a wide range of biomass feedstocks; biomass feeding systems for high-pressure gasifi­ers; determination of the effects of additives, including catalysts for minimizing tar production and materials that capture the contaminants; and suitable ash disposal and wastewater treatment technologies. Research on thermal biomass gasification in North America has tended to concentrate on medium-energy gas production, scale-up of advanced process concepts that have been evaluated at the PDU scale, and the problems that need to be solved to permit large — scale thermal biomass gasifiers to be operated in a reliable fashion for power production, especially for advanced power cycles. Research to develop biomass gasification processes for chemical production via synthesis gas waned in the mid-1980s because of low petroleum and natural gas prices. More attention was given to the subject in the 1990s when the market prices for these fossil fuels began to increase.

Examples of the various types of biomass gasification processes are reviewed in the next few sections before commercial and near-commercial processes are described.