Energy from Biological Processes: Volume Iiñtechnical And

Energy from Biological Processes: Volume II—Technical and Environmental Analyses

September 1980

NTIS order #PB81-134769 Library of Congress Catalog Card Number 80-600118

For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 Stock No. 052-003-00782-7 Foreword

In this volume of Energy From Biological Processes, OTA presents the technical and environmental analyses on which the conclusions in volume I are based. The “Part 1: Biomass Resource Base” includes forestry, agriculture, processing wastes, and various unconventional sources including oil-bearing and aquatic plants. “Part 11: Conversion Technologies and End Uses” considers thermochemical conversions, fermentation for ethanol production, anaerobic digestion, use of alcohol fuels, select energy balances, and a brief description of chemicals from biomass. In each case, appropriate technical, economic, and environmental details are presented and analyzed.

JOHN H. GIBBONS Director Energy From Biological Processes Advisory Panel

Thomas Ratchford, Chairman Associate Executive Director, American Association for the Advancement of Science Henry Art Robert Hodam Michael Neushul Center for Environmental Studies California Energy Commission Marine Science Institute Williams College Kip Hewlett University of California, Santa Barbara Stanley Barber Georgia Pacific Corp. William Scheller Department of Chemical Engineering Department of Agronomy Ralph Kienker Purdue University Monsanto Co. University of Nebraska Kenneth Smith John Benemann Dean Kleckner Off ice of Appropriate Technology Sanitary Engineering Laboratory President State of California University of California, Richmond Iowa Farm Bureau Federation Wallace Tyner Paul F. Bente, Jr. Kevin Markey Department of Agricultural Economics Executive Director Friends of the Earth The Bio-Energy Council Purdue University Jacques Maroni Calvin Burwell Energy Planning Manager Oak Ridge National Laboratory Ford Motor Co. Robert Hirsch EXXON Research and Engineering Co.

- NOTE. The Advisory Panel provided advice and comment throughout the assessment, but the members do not necessarily approve, disapprove, or endorse the report for which OTA assumes full responsibility

Working Group on Photosynthetic Efficiency and Plant Growth

One Bjorkman Gary Heichel Richard Radmer Carnegie Institution North Central Region Martin-Marietta Laboratory Stanford University U.S. Department of Agriculture University of Minnesota Glenn Burton Southern Region Edgar Lemon U.S. Department of Agriculture Northeastern Region U S. Department of Agriculture Cornell University

Contractors and Consultants

The Baham Corp. Purdue University, Departments of University of California at Davis, College of Charles Berg Agricultural Economics, Agricultural Agricultural and Environmental Sciences California Energy Commission Engineering, and Agronomy University of California, Richmond Field Otto Doering Ill Santa Clara University, Department of Station Douglas Frederick Mechanical Engineering University of Pennsylvania, Department of Charles Hewett State of California, Office of Appropriate Chemical and Biochemical Engineering 1 E Associates Technology University of Texas at Austin, Center for Larry Jahn T B Taylor, Associates Energy Studies Robert Kellison Texas A&M University, Departments of University of Washington, College of Neushul Mariculture, Inc. Agricultural Economics and Agricultural Forest Resources Participation Publishers Engineering RonaId Zweig Princeton University, Department of Texas Tech University, Department of Aerospace and Mechanical Sciences Chemical Engineering

iv Energy From Biological Processes Project Staff

Lionel S. Johns, Assistant Director, OTA Energy, Materials, and International Security Division Richard E. Rowberg, Energy Program Manager Thomas E. Bull, Project Director A. Jenifer Robison, Assistant Project Director Audrey Buyrn* Steven Plotkin, Environmental Effects Richard Thoreson, Economics Franklin Tugwell, Policy Analysis Peter Johnson, Ocean Kelp Farms Mark Gibson, Federal Programs

Administrative Staff

Marian Growchowski Lisa Jacobson Lillian Quigg Yvonne White

Supplements to Staff David Sheridan, Editor Stanley Clark

OTA Publishing Staff

John C. Holmes, Publishing Officer Kathie S. Boss Debra M. Datcher Joanne Mattingly

*Project director from April 1978 to December 1978 Acknowledgments

OTA thanks the following people who took time to provide information or review part or all of the study Don Augenstein, Flow Laboratories Tim Clidden, Dartmouth College Edward Nolan, General Electric Co Edgar E. Bailey, Davy McKee Corp. Irving Goldstein, North Carolina State John Nystrom, Arthur D. Little, Inc. Richard Bailie, Environmental Energy University Ralph Overend, National Research Council Engineering, Inc. John Goss, University of California of Canada Weldon Barton, U.S. Department of Roy Gray, Soil Conservation Service Billy Page, U.S. Forest Service Agriculture Loren Habegger, Argonne National R. Max Peterson, U S. Forest Service Charles Bendersky, Pyres, Inc. Laboratory David Pimentel, Cornell University Edward Bentz, National Alcohol Fuels John Harkness, Argonne National L. H. Pincen, U.S. Department of Commission Laboratory Agriculture Beverly Berger, U.S. Department of Energy Sanford Harris, U.S. Department of Energy Harry Potter, Purdue University Brian Blythe, Davy McKee Corp. Marilyn Herman, National Alcohol Fuels T. B. Reed, Solar Energy Research Institute Hugh Bollinger, Plant Resources Institute Commission Mark Rey, National Forest Products Diane Bonnert, Soil Conservation Service Edward Hiler, Texas A&M University Association Carroll Bottum, Purdue University Dexter Hinckley, Flow Resources Corp. Robert San Martin, U.S. Department of Robert Buckman, U.S. Forest Service Wally Hopp, Pacific Northwest Laboratory Energy Fred Buttel, Cornell University John Hornick, U.S. Forest Service Kyosti Sarkanen, University of Washington James Childress, National Alcohol Fuels William Jewell, Cornell (University John Schaeffer, Schaeffer and Roland, Inc. Commission Fred Kant, EXXON Research and Engineer- Rolf Skrinde, Olympic Associates Raymond Costello, Mittlehauser Corp. ing Co, Thomas Sladek, Colorado School of Mines Gregory D’Allessio, U S Department of J L. Keller, Union Oil of California Research Institute Energy Don Klass, Institute of Gas Technology Frank Sprow, EXXON Research and Ray Dideriksen, Soil Conservation Service J. A. Klein, Oak Ridge National Laboratory Engineering Co. Richard Doctor, Argonne National Al Kozinski, Amoco Oil Co George Staebler, Weyerhauser Corp. Laboratory Suk Moon Ko, Mitre Corp. Terry Surles, Argonne National Laboratory James Dollard, U.S. Department of Energy Kit Krickensberger, Mitre Corp. Robert Tracy, U.S. Forest Service Warren Doolittle, U.S. Forest Service Barbara Levi, Georgia Institute of R Thomas Van Arsdall, U.S. Department Ernest Dunwoody, Mittlehauser Corp. Technology of Agriculture Ed Edelson, Pacific Northwest Laboratory Les Levine, U.S. Department of Energy R. I Van Hook, Oak Ridge National Eugene Eklund, U.S. Department of Energy Edward Lipinsky, Battelle Columbus Laboratory George Emert, University of Arkansas Laboratory Thomas Weil, Amoco Chemicals John Erickson, U.S. Forest Service William Lockeretz, Northeast Solar Energy Donald Wise, Dynatech R&D William Farrell, EXXON Research and Center Robert Wolf, Congressional Research Engineering Co Dwight Miller, U.S. Department of Service Winston C. Ferguson, Conservation Con- Agriculture Robert Yeck, U S Department of sultants of New England John Milliken, U.S. Environmental Protec- Agriculture Kenneth Foster, University of Arizona tion Agency John Zerbe, U S. Department of Douglas Frederick, North Carolina State Larry Newman, Mitre Corp. Agriculture University Ralph E C. Fredrickson, Raphael Katzen Associates

vi Contents

Chapter Page Part 1: Biomass Resource Base 1. Introduction and Summary ...... 5 2. Forestry ...... 9 3. Agriculture...... 51 4. Unconventional Biomass Production...... 91 5. Biomass Wastes ...... 111

Part II: Conversion Technologies and End Use 6. Introduction and Summary ...... 119 7. Thermochemical Conversion...... 123 8. Fermentation...... 159 9. Anaerobic Digester ...... 181 10. Use of Alcohol Fuels ...... 201 11. Energy BaIances for Alcohol Fuels...... 219 12. Chemicals From Biomass...... 227

vii Part I.

Biomass Resource Base Chapter 1 INTRODUCTION AND SUMMARY Chapter 1 INTRODUCTION AND SUMMARY

The biomass resource base potentially in- ward energy production is needed to compare cludes hundreds of thousands of different the options and to establish cultivation re- pIant species and various animal wastes. In quirements and yields. principle, plants can be cultivated anywhere In addition to energy crops, substantial there is a favorable climate with sufficient quantities of crop residues can be collected water, sunlight, and nutrients. In practice, and used for energy without exceeding crop- there are numerous limitations, and the most Iand erosion standards. important of these appears to be the soil type for land-based plants and cultivation and har- The third major land category is rangeland, vesting techniques for aquatic plants. which vary from highly productive wetlands to deserts. Cultivation and harvesting techniques The largest area of underutilized land that is and plant growth are uncertain, and in drier welI suited to plant growth is the Nation’s for- regions the lack of water will Iimit yields unless estland. Through more intensive forest man- the crops are irrigated. However, irrigation agement— particularly on privately owned greatly increases the energy needed for farm- lands–the supply of wood for energy, as well ing and it is uncertain whether it will be social- as for traditional wood products, could be sub- ly acceptable to use the available water for stantially increased. However, haphazard energy production. wood harvest could cause severe environmental damage and reduce the available supply of Aside from natural wetlands, there are other wood. areas where freshwater plants might be grown and there are vast areas of ocean in which The highest quality land suitable for inten- ocean farms might be built. Cultivation and sive cultivation of plants is the Nation’s crop- harvesting techniques and crop yields are Iand. The best cropland is dedicated to food highly uncertain. product ion, but there is some underutilized hayland and pastureland as well as land that In addition to crop cultivation and residues can be converted to cropland. To a certain ex- there is biomass potential from processing and tent, grains—especially corn — can be grown animal wastes. * Most processing wastes cur- for ethanol production and the distillery by- rently are used for energy, animal feed, or product used as an animal feed to displace chemical production, but much of the remain- soybean production. More grain can then be der could be used for energy. Moreover, most grown on the former soybean land. As the etha- of the manure from animals in confined live- nol production level grows, however, the ani- stock operations could be used for energy. mal feed market for the distillery byproduct These biomass sources and various other will become saturated and grass production aspects of the resource base are considered in quickly will become a more effective energy the following chapters. option for the cropland. To a certain extent, environmental damage appears to be practi- cally unavoidable with grain production, but grass cultivation is more environmentally benign. *Wastes are defined as byproducts of biomass processing that The candidates for bioenergy crops are nu- are not dispersed over a wide area and therefore need not be col- merous, but crop development directed to- lected Residues must be collected

5 Chapter 2 FORESTRY Chapter 2.- FORESTRY

Page Page Introduction ● 9 3. Fuelwood Harvests in 1976...... 13 Present Forestland...... 10 4. Current and Expected Annual Stand Present Cutting of Wood...... 10 Improvements ...... 14 Forest Products Industry Roundwood 5. Estimated Above ground Standing Biomass of Harvesting ...... 11 Timber in U.S. Commercial Forestland. . . . . 15 household Fuelwood...... 13 6. Assumptions for Whole-Tree Harvesting Stand improvements ...... 13 Equipment ...... 19 Clearing of Forestland ...... 14 7 Annual Whole-Tree Chipping System Costs. 19 Summary of Current Cutting of Wood ...... 14 8. Assumptions for Cable Yarding Equipment . 20 Present lnventory of Forest Biomass...... 14 9. Cable Logging System Costs...... 20 Noncommercial Forestland ...... 14 10. Factors Affecting Logging Productivity . . . . 21 Commercial Forestland...... 15 11 . Potential Environmental Effects of Logging Quantity Suitable for Stand Improvement. . . 15 and Forestry ...... 28 Present and Potential Growth of Biomass in 12. Environmental impacts of Harvesting U.S. Commercial Forestso ...... 16 Forest Residues ...... 34 Forest Biomass Harvesting...... 18 13, Control Methods ...... 43 Factors Affecting Wood Availability ...... 21 Landownership ...... 21 Alternate Uses for the Land ...... 22 Public Opinion ...... 22 Alternate Uses for Wood ...... 22 Other Factors and Constraints ...... 23 Net Resource Potential ...... 23 Enviromental Impacts...... 25 FIGURES Introduction...... 25 Environmental Effects of Conventional Page Silviculture ...... 27 l. Forestland as a Percentage of Total Land Potential Environmental Effects of Area ...... 11 Harvesting Wood for Energy ...... 32 2. Area of Commercial Timberland by Region Controlling Negative Impacts ...... 41 and Commercial Growth Capability as of Potential Environmental Effects–Summary. 45 January 1,1977...... 12 R&D Needs ● **,.**...... ***,., 46 3. Forest Biomass inventory, Growth, and Use. 17 4. Supply Curve for Forest Chip Residues for Northern Wisconsin and Upper Michigan . . 21 TABLES S. Materials Flow Diagram for Felled Timber Page During Late 1970’s ...... 24

l. Logging Residues Estimate—summary . . . . 12 6. Environmental Characteristics of Forest 2. Logging Residue Estimates...... 13 Regions ...... 33 Chapter 2 FORESTRY

Introduction

The use of wood for fuel is at least as old as Wars I and 11, when conventional fuels became civilization. Worldwide, wood is still a very im- scarce. After the crises abated fuelwood use portant source of energy. The U.N. Food and dwindled rapidly, even though the reemerg- Agricultural Organization estimates that the ence of the same conditions in the near future total annual world harvest of wood in 1975 was may have been expected. 90 billion ft3 (about 25 Quads) of which nearly During 1917-18, for example, the Eastern one-half was used directly for fuel. ’ Much of United States suffered a shortage of coal. FueI- the wood that is processed into other products wood was used whenever possible to replace is available for fuel when the product is dis- coal, as were sawdust briquettes and other carded from its original use, and indeed large combustible biomass. Individual towns in New but unknown quantities are used in this man- England organized “cutting bees” and “cut a ner. cord” clubs for gathering wood fuel to offset Wood has been a very important fuel in the the shortage of coal. Between 1916 and 1917 United States, having been used for home the price of fuelwood increased by about 20 to heating and cooking, locomotive fuel, the gen- 30 percent. eration of electricity for home, business, and The U.S. Forest Service prepared a publica- industrial use, and for the generation of steam tion explaining, among other things, how wood for industry. According to Reynolds and Pier- could be used as fuel to conserve coal.4 It was son, more than half of the wood harvested thought at the time that coal reserves in the from U.S. forests for the 300 years of American z United States were dangerously low and that history preceding 1940 was used as fuel. Con- the war-induced shortage of 1917-18 had mere- sumption of wood fuel reached its peak in the ly emphasized the inevitable need to conserve United States in 1880 when 146 million cords them. This publication advocated a broad (2.3 Quads) were used according to Panshin, et Government policy for development of a fuel- al.3 The same authors report that per capita wood industry. The role that cutting fuelwood consumption of wood fuel peaked in 1860 at could play in forest management was consid- 4.5 cords/yr. During the past 100 years, the ered, and an analysis of the economics of cut- direct use of wood for fuel declined in the ting and gathering, etc., was given. The report United States to about 30 million cords/yr (0.5 Quad/yr). It was used primarily as a fuel by the concluded that a fuelwood industry could be profitable and could benefit the forest in other forest products industries, which used manu- ways as well. The document was published facturing residues, and for home fireplaces March 10, 1919, by which time the war had and outdoor cooking, which created demand ended, and the Nation’s fuel situation was al- for charcoal and hardwood roundwood. ready beginning to return to prewar condi- There have, however, been periodic revivals tions. There is no evidence that any of the rec- of fuelwood use to replace conventional fuels ommendations were folIowed. in the United States. They have usually oc- Since World War 11, the major emphasis on curred during times of crises, such as World wood use has been for lumber and paper pulp. The annual harvest of commercial wood (wood 1 Yearbook of Forest Products 1%4-1975 (Rome: Food and Agri- appropriate for the forest products industry) culture Organization of the United Nations, 1977) ‘R V Reynolds and A. H. Pierson, “Fuelwood Used In the U.S. grew by 22 percent between 1952 and 1976. 16301930, ” USDA Clr. 641, 1942 During this same period, the net growth of 3A J Panshin, E. S H arrar, J S Bethel, and W. J. Baker, Forest Products (New York: McGraw-Hill, 1962) ‘USDA Bulletin 753, Forest Service, Mar 10,1919

9 10 ● VOI. II—Energy From Biological Processes commercial wood (total growth of commercial increased by 20 percent from 1952 to 1976. timber less mortality of commercial timber) in- Thus, increased harvests of wood do not neces- creased by 56 percent. In only one region in sarily imply that the forests are being depleted. the country, the Pacific Coast, did the inventory of live commercial wood on commercial forestlands decline. In the Pacific Coast re- The growth of wood depends not only on the gion, however, the growth, as a percentage of climate and soil type, but also on the type and the standing inventory, is the lowest in the age of trees and the way the forest is managed. country due to the old age of the timber. Na- In this chapter, the potential for fuelwood pro- tionwide the inventory of commercial timber duction from the Nation’s forests is examined.

Present Forestland

Forestland is defined as land that is at least in the West. Despite the low-productivity clas- 10-percent stocked with forest trees or has sification, however, timber is harvested from been in the recent past and is not permanently many areas of land in this category. converted to other uses. The forestlands are Most of the forestland in the East is privately divided into two categories: commercial and owned, while about 70 percent of the western noncommercial. Commercial forestland is de- forestland is publicly owned and managed by fined as forestland that is capable of produc- 3 the Federal Government or State and local ing at least 20 ft /acre-yr (0.3 dry ton/acre-yr) of authorities. commercial timber in naturally stocked stands and is not withheld from timber production The U.S. Department of Agriculture (USDA) (e.g., parks or wilderness areas). The rest is projects that the forestland area will decrease termed noncommercial. by 3 percent by the year 2030 (about 0.4 million acres/yr or a total of 20 million acres). s In The forest regions of the United States and the 1980’s, a significant portion of the decline the percentage of the total land area of each will result from conversion to cropland, par- State that is forestland are shown in figure 1. ticularly in the Southeast. USDA projects that Currently, there are 740 million acres of forest- in the 1990’s, most of the conversion will be to Iand in the United States, with about half in reservoirs, urban areas, highway and airport the East (i.e., North plus South) and half in the construction, and surface mining sites. West. About 490 million acres are classified as commercial forestland and nearly three-quar- However, about 32 million acres of potential ters of this are in the East. The productive cropland are now classified as forestland (see potential of commercial forestlands is shown ch. 3). Consequently, if a strong demand de- in figure 2. velops for cropland, then the decrease in forest area will be somewhat larger than USDA’s pro- In addition, there are 205 million acres of jection. noncommercial forestland, which are classified this way because of their low productive 3 potential (i. e., less than 20 ft /acre-yr). Prac- ‘An Assessment of the Forest and Range Land Situation in the United States, tically all of the noncommercial forestland is review .draft, USDA Forest Service, 1979

Present Cutting of Wood

Forest wood is currently being cut for four dustry roundwood, 2) production of household purposes: 1) production of forest products in- fuelwood, 3) timber stand improvements, and Ch. 2—Forestry ● 11

Figure 1.— Forestland as a Percentage of Total Land Area

SOURCE Forest Service u S Department of Agriculture

4) clearing of timberland for other uses. Each left at the logging site after the commercial of these produces wood that can be or is used roundwood is removed. These residues are for energy. branches, small trees, rough and rotten wood, tops of harvested trees, etc. Forest Products Industry The statistics on logging residues reported Roundwood Harvesting for 1970 and 1976 by the Forest Service underestimate the total quantity of residues gener- Currently, the forest products industry is har- ated by harvesting activities. The Forest Serv- vesting 200 million dry ton/yr (3.1 Quads/yr) for ice data only include wood logging residues lumber, plywood, pulp, round mine timber, from growing stock trees. * etc.). During the processing of this wood, 90 Not reported are: million ton/yr of primary and secondary manufacturing wastes are produced. These wastes 1. bark — most studies of logging residue pre- are discussed later under “Biomass Processing sent volumes without bark; Wastes” in chapter 5. 2. residues from: ● nongrowing stock trees on logged-over In addition, the process of harvesting the areas, wood generates considerable logging residue. ‘Commercial stock trees that are 1 ) at least 5-inch diameter at The logging residue consists of the material breast height (dbh) and 2) not classified as rough or rotten 12 ● Vol. II-Energy From Biological Processes

Figure 2.—Area of Commercial Timberland by From various sources’ 78 and OTA estimates, Region and Commercial Growth Capability the ratios of growing stock residues to total as of January 1, 1977 biomass residues were derived. ’ Using these ratios and the Forest Service data for growing No. of acres s (millions) stock residues, the quantity of logging residues 180 was estimated to be about 84 million dry tons 170 (1.3 Quads) in 1976. The regional breakdown is shown in table 1, and a more detailed break- 160 N = North S = South down is shown in table 2. 150 PC = Pacific Coast. Alaska, Hawaii 140 RM = Rocky Mountain and Great Plains Table 1.–Logging Residues Estimatea–Summary (in million dry tons) 130

b 120 Region Softwood Hardwood Total 110 North ...... 2.9 13.2 “ 16.0 South ...... 17.6 15.2 32.8 100

90 Total b ...... 54.5 29.5 84.1 80 aFrom a Ig76 harvest of 130 mllllon dry tons of softwood and 54 mllllon dry tom Of hardwood 70 bsurns may not agree due to round off error 60 SOURCE J S 8ethel, et al,, “Energy From Wood, ” College of Forest Resources, University of Washington, Seattle, contractor report to OTA, April 1979. 50

40 There is some uncertainty as to whether various logging residue studies are in agreement 30 as to what constitutes nongrowing stock log- 20 ging residue. Loggers may avoid cutting non- 10 growing stock trees that hold little or no economic value. This practice would be common 2050 50-85 85120 120 + Total in selective logging. In many logging residue Growth capability studies, it is unclear whether or not such uncut 3 (ft commercial timber/acre-yr) trees were considered residue. Some of the dif- SOURCE Data from Forest Statistics, 1977, Forest Service, U S Department of Agriculture, 1978 ferences observed in logging residue factors reported by various authors in the same region may be due largely to these methodological trees of growing stock species and qual- differences. There is a danger that if uncut ity, but less than 5-inch diameter at nongrowing stock is counted as a logging resi- breast height, due, it might again be counted as part of the trees that would be growing stock trees biomass that should be removed by various sil- except that they are classified as rough vicultural stand improvements. Every effort or rotten, and was made to avoid this type of double count- trees of noncommercial species; ing. 3. tops and branches; and 4. stumps. All of these logging residue components, as ‘J O Howard, “Forest Residues– Their Voiume, Vaiue and Use, ” Part 2: Volume of Residues From Logging Forest Industries, well as the residue presented in the aforemen- 98(1 2), 1971 tioned reports, are potentially usable as fuel. * ‘R L Weich, “Predicting Logging Residues for the Southeast, ” USDA Forest Service Research Note SE-263, 1978 “J. T Bones, “Residues for Energy in New Engiand,” Northern Logger and Timer Processor 25(1 2), 1977 ‘j S Bethei, et al , “Energy From Wood, ” Coiiege of Forest Re- sources, University of Washington, Seattle, contractor report to *In this report, the stumpwood component IS not considered OTA, Apni 1979 Ch. 2—Forestry . 13

Table 2.–Logging Residue Estimates (thousand dry tons)

From growing stock From nongrowing stock Tops and Harvest branches in 1976 Wood Bark Total Wood Bark Total incl. bark Total Softwoods North...... 7,448 823 908 597 64 661 1,323 2,892’ South...... 63031 3,756 393 4,149 2,697 314 2,993 10,415 17,557 W.Pine ...... 16,500 1,548 181 1,729 2,022 236 2,258 3,000 6,987 Coast ...... 43,190 7,496 876 8,372 8.117 949 9,066 9,668 27,106 Total...... 130,169 13.623 1,535 15,158 13,433 1,563 14,978 24,406 54,542 Hardwoods North...... 24,546 4,214 313 4,527 1,410 100 1,510 7,147 13,184 South...... 27,974 4,984 381 5,275 1,637 123 1,760 8,185 15,220 W.Pine ...... 34 3 — 3 2 — 2 11 16 Coast ...... 1,094 345 36 381 255 27 282 458 1,121 Total ...... 53,648 9,456 730 10,186 3,304 250 3,554 15,801 29,541

SOURCE J S Bethel, et al, ’’Energy From Wood, “College of Forest Resources, University of Washington, Seattle, contract or reporf to OTA, April 1979

Household Fuelwood space for the higher quality trees. These cutting activities are generally referred to as stand The harvest of roundwood for use as house- improvements, and include stand conversions* hold fuel was estimated in 1976 to be 657 mil- and thinning operations. Wood from these ac- lion ft3, or approximately 10 million dry tons tivities or sources is suitable for fuel. (0.16 Quad). These figures are similar to the results reported by EIlis, who found that 600 The data on the amount of current stand im- mill ion ft3 of roundwood, excluding bark, were provement activity are very limited and do not harvested for fuelwood. ’” Allowing a IO-per- allow a detailed analysis. During the 1968-71 cent increase for bark, this becomes 660 mil- period, various practices, such as precommer- lion ft3. The regional breakdown is shown in cial and commercial thinning, species conver- table 3. The quantity harvested in more recent sion, weed control, and other stand improvements were carried out on a total of 1.4 million years is considerably larger, however, acres. This represents only 0.3 percent of the commercial timberland. Generally these prac- Table 3.–Fuelwood Harvests in 1976 (in million dry tons) tices are carried out irregularly, or on a when- Region Softwood Hardwood Total a and where-needed basis. Undoubtedly most of North, ., ...... 0.05 3.7 3.8 the activity is carried out on industry lands South ...... 1.3 4.2 5.7 where intensive forest management is most ad- Rocky Mountains 0.43 0.01 0.44 Pacific Coast...... 0.33 0.11 0.39 vanced. A recent survey of forest industry Total ...... 2.3 8.2 10.2 firms that manage their own lands revealed the current level of these practices. ’ These are asums may not agree due to round off errors summarized in table 4. SOURCE J S Bethel et al Energy From Wood, ‘ College of Forest Resources, Umverslty of Washmgtorr Seattle contractor report toOTA Aprd f979 In addition, there are timber stand improvements (excluding thinnings), species conver- Stand Improvements sion, and weed control items, on about 1.7 million acres of low-quality stands per year. Yields In normal forestry operations, there may be would vary tremendously among these prac- several times during the growth of a stand of *Stand conversion is the practice of eliminating tree species trees that malformed, rough, or otherwise un- currently occupying a stand and replacing them with other species. desirable trees are cut to make more growing ‘ID S DeBell, A P Brunette, and D C Schweitzer, “Expecta- tions From Intensive Culture on Industrial Forest Lands,” ). For., ‘OT H EIIIs, “FuelwoOd,” unpublished manuscript, 1978 January 1977 14 ● Vol. II—Energy From Biological Processes

Table 4.–Current and Expected Annual Combining these two sources results in 47 Stand Improvements million ton/yr (0.7 Quad/yr) of residues from Percent of stand improvements. Percent of firms expecting Industry Estimated to maintain or lands acres increase level Clearing of Forestland Treatment treated treateda of treatment Precommercial thinning ., 0.2 135,000 53 Clearing of forest land for other uses can pro- Timber stand vide a temporary, but potentially significant, Improvement, . 1 8 1,212,000 69 Commercial throning ... 25 1,684,000 92 local supply of wood. The yield per acre har- Species conversion 04 269,000 65 vested varies widely with the locality. Assum- Weed control . . ., 0.3 202,000 50 ing 30 ton/acre cleared, then USDA projections aPercenl 01 lands treated hmes toial acreage owned by industry for forestland clearing would provide about SOURCE O S OeBell A P Brunel[e and O C Schweitzer Expectations From Intenswe 0.2 Quad/yr to 2030. If the forestland with a Culture on Induslnal Forest Lands J For January 1977 high and medium potential for conversion to cropland is all cleared over the next 15 years, then this would provide 1 Quad/yr of wood for tices, but assuming 17 dry ton/acre (as derived these 15 years. Most of this would occur in the by Bethel for rough, rotten, and salvageable Southeast (see ch. 3). trees in the South), this amounts to 29 million dry ton/yr. Thinnings were also carried out on 1.8 mil- Summary of Current Cutting of Wood lion acres, but there is little information re- The forest products industry currently har- garding the amounts of residue produced. vests about 200 million dry ton/yr (3.1 Quads/ Yields have been reported of 2.2 dry ton/acre yr) of wood for lumber, plywood, paper pulp, 1 in 4-year-old Ioblolly pine thinning, 2 and 17 to and other products. The process generates an 28 dry ton/acre in pole timber hardwoods in additional 84 million ton/yr (1.3 Quads/yr) of the North. ” If a national average of 10 ton/ logging residues. Another 10 million dry tons acre is assumed, thinning would provide 18 (0.2 Quad/yr) are harvested for fuelwood, and milIion dry ton/yr of residue. about 47 dry ton/yr (0.7 Quad/yr) are cut during stand improvements. This results in a total harvest of about 340 million dry ton/yr or the eqiv- ‘‘Si/vicu/ture Biomass Farms (McLean, Va The MITRE Corp , a lent of 5.3 Quads/yr. Another 0.2 Quad/yr is 1977) ‘‘F E Blltoner, W A Hlllstrom, H M Stelnhlll, and R M Gad- obtained from clearing and converting forest- mar, USDA Fore$t Service Research Paper NC-1 37, 1976 Iands to other uses.

Present Inventory of Forest Biomass

It is not a simple matter to derive the total base, the present inventory of forest biomass forest biomass inventory from the Forest Serv- can only be estimated. ice surveys. As noted earlier, this lack of an adequate” census base stems from the tradi- Noncommercial Forestland tional practice of evaluating the wood in a forest only in terms of what is assumed to be mer- As mentioned above, of the one-quarter bil- chantable, rather than on a whole-tree or lion acres of noncommercial land, 24 million whole-biomass basis. Furthermore, the Forest acres (about 10 percent) are so classified be- Service does not survey noncommercial forest- cause they are recreation or wilderness areas, lands (about one-third of the total forest area). or are being studied for these uses. These lands As a result of this inadequate information are not included in the inventory of standing Ch. 2—Forestry ● 15 timber. Approximately 205 million acres are For the purposes of this study, an estimate classified as noncommercial because they are of total whole-stem biomass for the United considered incapable of producing as much as States was developed, based on Forest Statis- 20 ft3 of commercial wood per acre-year. This tics for the United States, 1977.5 Table 5 shows criterion, is an arbitrary one, however, and the result of this analysis for commercial for- timber is, in fact, harvested from many areas estland. The details of these computations and of land in this category. For this reason, the lat- more extensive tables are given in OTA’s con- ter category of noncommercial forestlands is tractor report “Energy From Wood. ”16 included in the inventory of standing timber. Assuming that these 205 million acres pro- Table 5.-Estimated Above-ground Standing 3 Biomass of Timber in U.S. Commercial Forestland duce an average of 10 ft /acre-yr of commer- (excluding foliage and stumps, in billion dry tonsa) cial wood, that they are mature stands (80 years old or more), and that the aboveground Region Hardwood Softwood Total biomass is 1.5 times the amount of commercial North. 5.2 1,3 6.5 timber, the inventory of these noncommercial South . . . : : : : : ., 4.6 2,3 6.9 Rocky Mountains ., 0.2 24 2.6 lands is 3.7 billion dry tons (57 Quads). Pacific Coast, ., 06 42 48 Alaska, ., ., 0.08 1,3 14 In addition, 23 million acres, mostly in Total ., . . 10.6 11 5 221 Alaska, were classified in 1978 as noncommercial because they were considered inaccessi- asum~ may noI agree due to round off errors SOURCE J S Bethel, et al Energy From Wood College of Forest Resources Umverslty of ble. Assuming a production capability of 35 Washmgfon Seaftle contractor reporl to OTA April 1979 ft3/acre-yr and the same assumptions as above, the inventory from these lands is 1.4 billion dry Adding commercial and noncommercial tons (22 Quads). land inventories gives 27 billion tons (430 These two categories result in an inventory Quads), which is estimated to be the inventory on 1978 noncommercial lands of about 5 bil- of biomass in U.S. forests, excluding stumps, lion dry tons (80 Quads). foliage, and roots and the biomass in parks and wilderness areas, or areas being considered for these uses. * Commercial Forestland Approximately 488 million acres of forest- Quantity Suitable for land are classified by the Forest Service as Stand Improvement commercial forestland for purposes of report- ing a national forest survey. It is possible to Of the 27 billion tons of standing biomass, estimate a fuel inventory from commercial for- some of the wood is of the type that would be estland, using national forest survey data, with removed in stand improvements. This would much more precision than was the case for include brush, rough, rotten, salvageable dead noncommercial lands. wood, and Iow-quality hardwood stands occu- Two options were considered for developing pying former conifer sites. In Alaska, there are roughly 330 miIIion tons of this type of wood. 7 estimates of total biomass on commercial for- In the rest of the Pacific Coast region, there are estland based on the national forest survey. 565 million tons, and in the Rocky Mountain One procedure involved the assumption of region, 324 million tons. The North and South multipliers that would convert the basic prod- have 822 million and 978 million tons, respec- uct inventory data to whole-stem biomass estimates. A second method involved the use of 1‘Forest Statfstlcs of the U S . J 977, USDA Iorest Service, re- stand tables from the national forest survey view draft and allometric regression equations for esti- ‘013ethel, Of) clt 4 * For the purpose> of this report, ~turnps, root$, and follage are mating biomass for various tree components. exc I uded from who le-~tenl b Iom as~ “llethel, op clt ‘‘1 bld 16 ● VOI. Il—Energy From Biological processes tively.18 The total is 3.1 billion tons (49 Quads) or all of the cuttings that could be used to con- of wood that would be appropriate for remov- vert stands of one kind of trees to a more pro- al in stand improvements on commercial for- ductive type. Consequently, this is a conserva- estlands. This figure does not include foliage tive estimate of the biomass available from ‘“lbd stand improvements.

Present and Potential Growth of Biomass in U.S. Commercial Forests

Current gross annual biomass growth in that the actual current biomass growth on commercial U.S. forests has been estimated commercial forestland is one to two times the from Forest Service data to be 570 million dry values derived from normal yield tables, or 570 ton/yr, of which 120 million ton/yr are mortali- million to 1,140 million dry ton/yr (9 to 18 ty, and 450 million ton/yr net growth. * The Quads/y r). (See figure 3.) usual method of determining the productivity These estimates do not take into account of a particular stand occupying a site is by ref- the productive potential of the forestland. For- erence to normal yield tables. These tables are est site productivity is estimated on the basis models used to predict growth of active nat- of the vegetation currently occupying the area ural stands, and are based on stands of “full” at the time of the survey. But over 20 million or “normal “ stock. acres of commercial forestland are unstocked, Because of the utilization assumptions built and much more land is stocked with species into normal yield tables, however, productivity that are growing more slowly than could be may consistently be assigned a low, and mis- achieved with species better suited to the site. leading rating. For example, when the actual The forest survey indicates that, due to these growth in 131 Douglas-fir plots scattered factors, current growth is about half the throughout western Washington and Oregon growth that could be achieved with full stock- was compared with Forest Service BuIIetin nor- ing of highly productive tree types (i. e., current mal yield tables for Douglas fir, it was found growth is estimated by the Forest Service at 38 that the yield tables consistently underesti- ft3/acre-yr while the land capability is estimated the actual growth. Actual growth in mated by USDA at 74 ft3/acre-yr). OTA there- some age-site combinations was more than fore estimates the potential growth to be double the normal yield table value, and the about two to four times that derived from nor- overall average growth exceeded the yield mal yield tables, or 1.1 billion to 2.3 billion dry table by nearly 40 percent.20 Furthermore, in ton/yr (18 to 36 Quads/yr) with full stocking of parts of the Rocky Mountains where Forest productive tree species on commercial forest- Service and industry lands are co-mingled, in- Iand. This corresponds to slightly more than 2 dustry representatives report that measure- to 4 ton/acre-yr on the average. ments of actual growth are two to three times Beyond the potential growth with unfertil- the productivity assigned by normal yield ized timber, studies in the Southeast indicate tables” Because of the errors associated with that fertilizers and genetic hybrids could in- estimating tree types, their number, and their crease the ‘biomass growth by 30 percent.22 size from normal yield tables, OTA estimates However, not all of the potential growth is physically accessible or economically attrac- ‘“H Wahlgren and T EIIIs, “Potential Resource Avallablltty With Whole Tree Uttllzatlon, ” TAPPI, VOI 61, No 11, 1978 tive as discussed below. ● The 120 million tons of annual mortallty are from growing stock trees only Mortallty from nongrowtng stock trees IS not known Under Intensive management, much of the mortality loss cou Id potentla I Iy be captured for product Ive use “)Bethel , Or) Clt J ‘ I bld Ch. 2—Forestry ● 17

Figure 3.—Forest Biomass Inventory, Growth, and Use (billion dry tons with equivalent values in Quads)

I mortality billion dry tons \

0.45-0.90 billion dry tons 7.0-14.0 Quads land Commercial / trees forestland 4 billion dry tons / — 64 Quads /

/’

O 01 billion dry tons

Industrial wood cutting

SOURCE Office of Technology Assessment 18 ● VOI. II—Energy From Biological Processes

Forest Biomass Harvesting

Variations on the current harvesting tech- they represent a disposal problem. Also the niques (described below) are likely to be com- whole-tree and tree length methods tend to do mon with fuelwood harvests and stand im- more damage to the timber being skidded and provement activities that produce residues to the residual stand. If there is thick under- suitable for fuel. Nevertheless development of brush, the who/e-tree method may be difficult new techniques and equipment designed for or impossible. A weighing of the various fac- fuelwood harvests and stand improvements tors appropriate to the site being logged results could lower the cost. in the method used. If markets for the limbs develop, however, then more who/e-tree skid- Intensive forest management might typical- ding may be used than is now the case. ly consist of the following: The stand would be clearcut, and the slash (or logging residue) re- Once the wood is at the roadside, it can be moved. The stand would then be replanted cut and loaded or loaded directly into trucks with the desired trees. After 5 to 20 years the for transport to the mill or conversion site. stand would be thinned so as to provide more Alternatively, the wood can be chipped at the space for the remaining trees. The stand would roadside with the chips being blown into a van then continue to be thinned at about 10-year for transport. intervals, by removing diseased, rough, rotten, Two large-scale harvesting systems consid- and otherwise undesirable trees and brush. In ered here are whole-tree harvesting and cable very intensively managed stands, the trees logging. In the whole-tree chip system, the might also be pruned to avoid the formation of trees are felled by a vehicle called a feller- large knots in the stem of the tree (e.g., for buncher, which grabs the tree and uses a hy- veneer). These periodic thinnings and (possi- draulic shear to cut the tree at its base. The ble) prunings would continue until the stand is tree is then lowered to the ground for skidding. again clearcut and the entire cycle repeated. This method is most appropriate for relatively For each operation mentioned above (except flat land and smaller trees (i.e., less than 20- the replanting), some woodchips suitable for inch diameter). energy could be made available. The method In the cable logging method, cables are ex- chosen for harvesting the fuelwood would de- tended from a central tower and the felled pend on a number of site-specific factors. The trees are dragged to a central point, where they primary objective would be to fell and trans- are sorted and skidded to the roadside. This port the selected trees or to transport the slash method is used primarily on terrain with steep in the most cost-effective manner, while doing slopes and large trees. Estimates for the equip- a minimum of damage to the remaining stand. ment and annual operating costs of these two systems are shown in tables 6 to 9. There are Currently there are four basic methods of other logging systems, but these two methods logging, each of which is designed to accom- are fairly representative of the range of exist- modate a number of physical and economic ing systems. factors peculiar to the logging site. Once the tree is felled: 1 ) it can be skidded (dragged) to a The major difference between the harvest- roadside as a who/e tree, 2) it can be delimbed ing of various categories of wood (e. g., resi- and the top cut off, and the entire stem or tree dues from logging, stand improvements, or pri- length skidded to the roadside, 3) it can be de- mary logging) is the quantity of wood that can Iimbed, topped, and cut (bucked) into long be removed from a site per unit of time, i.e., logs which are skidded, or 4) it can be cut into the logging productivity. Several factors affect shorter logs or short wood which are skidded. the logging productivity, and the most impor- The whole-tree skidding brings out the most tant of these are shown in table 10. The pro- biomass. However, if the limbs cannot be used duction of the logging operations discussed Ch. 2—Forestry ● 19 above might range from 15,000 to 75,000 green Table 6.–Assumptions for Whole-Tree ton/yr, leading to harvesting costs from about Harvesting Equipment $5 to $30/green ton. In addition, transporta- Initial Salvage tion, possible roadbuilding, and stumpage fees cost value Life Labora (fees paid to the landowner for the right to har- Equipment (dollars) (percent) years $/hour vest the wood) must be included. Transporta- Whole-tree chipper 380 hp ...... $115,000 20 5 $4.62 tion ranges from $0.06 to $0.20/ton-mile, and 600 hp ...... 132,000 20 5 4.62 where road building is necessary, the costs will Feller-buncher ...... 100,000 20 5 4.62 be considerably higher. Stumpage fees for Skidder (each) ...... 55,000 20 4 4.20 Used skidder ...... 10,000 10 3 fuelwood have been estimated at $0.40 to Lowboy trailer...... 10,000 10 10 $1 .00/green ton in New England, *J but these Used crawler ...... 30,000 20 5 will change with the market. Equipment moving truck ...... 1,680/yr 3 /4-ton crew cab pickup . . 8,400/yr The costs of whole-tree chipping 33 stands in 1 / 2 .ton pickup . 7,862/yr northern Wisconsin and the Michigan penin- Chain saws (3) ...... 3,024/yr Other labor sula 24have been modeled by computer simula- tion. In each case, the center of the country deck hands (2) . . . . . 7.20 foreman ...... 8.40 was assumed to be the destination for the supervisor . . . . . 2.81 wood. The supply curve for these stands is aSouth, includes payroll benefits shown in figure 4, exclusive of stumpage fees. SOURCE J S Bethel et al , Energy From Wood College of Forest Resources Umverslty of The cost average varies from $6 to $1 5/green Washington, Seattle, contractor report to OTA April 1979 ton ($1 2 to $30/dry ton) in 1978 dollars. The Where logging, transportation, and other costs range of delivered costs included relogging of logging residues ($16.50 to $20.30/green ton), are low, stumpage fees will be high and vice versa. The market will determine these fees, as thinning ($10.00 to $1 3.80/green ton), and inte- well as the quantity and types of wood that grated logging for lumber and residue chipping can be economically harvested. ($9.75 to $12.30/green ton). An equalizing factor in the delivered cost is the stumpage fee. The 1979 delivered cost of fuel chips was about $12 to $18/green ton in New England. 25 A I K Hewett, school of Forestry, Ya Ie Unlversltv, Private com- detailed national cost curve, however, would munication require a survey of all potential logging sites, 14j A Mattson, D P Bradley, and E M Carpenter, “Harvesting Forest Residues for E r-serrgy,” Proceedings of the Sec- 1’Connectl~ ut Valley Chipping, Plymouth, N H , L W ond Annual Fue/s From B/ornas$ S yrnposiurn (Troy, N Y Rensse- Hawhensen, president, letter to Conservation Consultants of laer Polytechnic Institute, June 20-22, 1978) New Eng[dnci, Dec 20, 1979

Table 7.–Annual Whole-Tree Chipping System Costs Annual costs to pay all expenses and earn 15% aftertax ROI, shown in thousands of dollars (values in columns are shown only when a change occurs). Annual Fuel, Local taxes Miscellaneous Region capital cost Maintenance lube, etc. and insurance Labora equipment b Total System based on 380-hp chipper Initial investment: $375,000 North. ., . . . . $221 $35 $44 $7 $83 $21 $442 South 221 — — — 62 — 390 West ...... 221 — — — 104 — 432 System based on 600-hp chipper Initial investment $447,000 North...... 264 40 50 9 94 21 478 South ...... , . 264 — — — 70 — 454 West ., ...... 264 — — — 118 — 502 alncludes foreman and superwsor bplckup trucks chamsaws elc

SOURCE J S Bethel et al Energy From Wood College of Forest Resources Unwersl!y of Washington Seattle contractor report to OTA April 1979 20 . VoI. II-Energy From Biological Processes

Table 8.–Assumptions for Cable Yarding Equipment which is not available. Nevertheless, some fuelwood can be had for as little as $10/green Initial Salvage 26 cost value Life Labora ton ($20/dry ton) plus stumpage fees. In un- Equipment (dollars) (percent) years $/hour favorable circumstances, the wood could cost Yarder with 50-ft tower $180,000 20 8 $10.29 as much as $30/green ton ($60/dry ton for relog- Yarder with 90-ft tower 228,000 20 8 10.29 ging of logging- residues in the Northwest).27 Radio and accessories . . . 11,386 0 4 — Whole-tree chipper Thus, fuelwood chips may vary in price from 380 hp ...... 115,000 20 5 7.76 about $20 to $60/dry ton which is in substantial 600 hp . ., ., . . . 132,000 20 5 7.76 agreement with the cost estimates based on Skidder (each) . . . ., 55,000 20 4 7.06 Hydraulic loader . . . . 207,000 20 6 10,64 harvesting costs. Used skidder ...... 10,000 10 3 Lowboy trailer, ., ., 10,000 10 10 In each category of wood there will be small Used crawler ., ., ., 30,000 20 5 businesses or individuals who are willing to Equipment moving work at lower rates, who are figuring only truck ., ., ...... 1,680/yr 3/4-ton crew cab pickup 8,400/Yr marginal costs, and/or who own the land and 1/2-ton pickup 7,862/yr assign a zero stumpage fee. In other words, Chain saws (3) . . . . . 3,024/yr there will always be limited supplies of wood Other labor yarding crew. . . . 47,84 below the average market price. (5 men) The A val/abi/lty of LVood for a 50 MW Wood F/red foreman ...... 14, 11 ‘(’C Hewett, Power P/ant In Northern Vermont, report to Vermont State Fn- supervisor (1/3 time). 4.72 erg~ Off Ice under grant No 01-6-01659 awest, includes payroll benehts ‘i IK ,P Hewlett, ‘~eorgla Paclf IC Corp , At Ianta, (;a , Prlvat@ communlcat ion, 1979 SOURCE Off Ice of Technology Assessment

Table 9.-Cable Logging System Costs Annual requirement to pay all expenses and earn 15% aftertax ROI (thousands of dollars) Annual Local capital Fuel, taxes and Miscellaneous Equipment cost Maintenance lube, etc. insurance Labora equipment Total 50-ft tower/ 380-hp chipper investment: $466,000 ., . . . . $200 $43 $44 $9 $158 $21 $475 90-ft tower/600-hp chipper $531,000 ...... 228 61 47 11 158 21 526 alnlcudes foreman and Supervisor bplCkup trucks chamsaws e[c

SOURCE Olflce of Technology Assessment Ch. 2—Forestry ● 21

Table 10.–Factors Affecting Logging Productivity Figure 4.—Supply Curve for Forest Chip Residues for Northern Wisconsin and Upper Michigan Road space. Slope and slope changes-slope steepness and whether logging is uphill or downhill. Size and shape of landing. Skidding distances–both loaded and return if different, affected by the tract shape and its relation to the road system 20 Skid trail preparation. Timber character– 15 Species–Especially hardwood v. softwood. Volume and number of trees per acre to be removed–The size of trees 10 and logs IS a very Important consideration. Quality– More defective timber IS likely to result in more breakage, in- 5 creasing materials handling problems Residual stand, if any, m terms of number of trees and volume per 1 I I I I I acre This IS prescribed by the silvicultural method. 5 10 15 20 25 30 Cutting policy–appropriate for the stand Cumulative volume—millions of tons Felling and logging methods–whole tree, shortwood, tree length, etc. Brush height and density. SOURCE: J. A. Mattson, D. P Bradley, and E M. Carpenter, “Harvesting Forest Condition and number of windfalls, old stumps, and slash per acre. Restdues for Energy, ” Proceedings of the Second Annual 13!omass Drainage and stream crossings. Sympos/um (Troy, N. Y.: Rensselaer Polytechnic Institute), Season. June 20-22, 1978. Crew size and aggressiveness. Wage plan. Equipment types, functions, and balance–especially the number of places handled per cycle, Maintenance policy. Environmental regulations–may prescribe certain practices or preclude certain equipment from areas with sensitive mixes of soils, slopes, and/or drainage thereby reducing production or Increasing costs In the West these regulations have caused a shift in the mix of tractor v. cable Iogging as well as shifts within each of these general categories

SOURCE J S Bethel et al Energy From Wood College of Fores! Resources Un[verslty of Wasmngton Seatlle conlracfor report 10 OTA April 1979

Factors Affecting Wood Availability

The presence of nearby roads, the concen- only 6 percent. In the East as a whole, 14 per- tration of wood on the logging site, and the ter- cent of the commercial forestland is publicly rain (steepness of the slope) are the most im- owned, while 7 percent is federally owned. portant physical factors affecting the econom- Ownership patterns in the West are reversed, ics, and thus the availability, of harvested with 68 percent of the commercial forestland wood, Nevertheless, landownership, alternate being publicly owned (96 percent in Arizona) uses for the land, taxation, and some sub- and 58 percent in Federal ownership. sidiary benefits and constraints also play an important role in wood availability. These Although patterns in the West permit log- other factors are discussed below. ging firms to deal with a limited number of large landowners, other restrictions may be placed on the logging operations. One exam- Landownership ple is the Federal requirement that logging residues be removed from or otherwise dis- One of the more important features distin- posed of on national forests in the West, to guishing the various forest regions in the coun- minimize the risks of forest fires. try is landownership. In New England 2 percent of the commercial forestland is federally Logging firms in the East must deal with a owned, and public ownership accounts for larger number of landowners, and in the North- 22 . VOI. II-Energy From Biological Processes

east, forestlands are often owned for recrea- There will be fewer overmature trees, the trees tional or investment purposes. It may be dif- will be more uniform in appearance and spac- ficult to determine who owns the land, to con- ing, and the forest floor will have less debris tact the owner, or to interest the owner in using and “extraneous” vegetation. The managed the land for logging. In the South this is less of forest will not look like a natural forest, and a problem. Large areas of forestland owned by the difference in appearance can be quite relatively small landowners are managed by large. the forest products industry and are available for logging. This change in appearance, together with various environmental uncertainties (see “En- vironmental Impacts”), leads to widely varying Alternate Uses for the Land opinions about the benefits of forest management. If the citizenry affected by increased The fact that a tract of land is forested and management cannot effectively participate in designated commercial does not necessarily the process of deciding where and how inten- mean that it can be logged. The owner may sively the forests will be managed, and if busi- have esthetic objections to logging, may use ness and Government officials are not sensi- the land for recreational purposes, or, in the tive to the concerns of the citizenry, then the case of an investor, may feel that it would be political atmosphere surrounding forest man- more difficult to sell the land after logging. in agement for energy could become polarized. New Hampshire and Vermont, for example, a Public opposition could then seriously restrict recent study concluded that only 6 percent of the use of forests for energy. the owners of commercial forestland considered timber production as a reason for owning Forest management, however, is not an ab- forestland, and only 1.3 percent listed it as the solute. There are many ways to manage forest- most important reason. 28 (This 6 percent owns Iands, from wood plantations to the occasion- 21 percent of the commercial forestland in the al gathering of fallen trees and branches. The two States). Nevertheless, 10 percent of the ability of political leaders to convey this fact private owners (representing 53 percent of the to the public, and the ability of Government to forestland) intended to harvest their timber aid in striking an equitable balance between within 10 years and over one-third of the environmental and esthetic concerns and the owners (representing 87 percent of the land) in- economics of wood harvesting, may prove to tended to harvest “some day. ” About half of be one of the most significant factors affecting the landowners (owning 9 percent of the land) an increased availability of wood for energy indicated that they would not harvest the land outside of the forest products industry. because of its scenic value or because their tracts were too small. Alternate Uses for Wood Public Opinion Much of the wood that will be used for energy in the near future is less suitable for ma- While proper management of a forest can terials (e. g., particle board or paper) than the improve the health and vitality of the trees, im- wood currently used for these products. If proper management can have severe environ- there is a strong demand for wood products, mental consequences. (See “Environmental however, some of this lower quality wood will Impacts”. ) In any event, an intensively man- be drawn into the materials market. Similarly, aged forest will look like it is being managed. technical advances in wood chemistry may create an additional demand for wood to be processed into chemicals. IJN p Klng+y and T W Birch, “The Forest-Land Owners of New Hampshire and Vermont, ” USDA Forest Service Resource It must be remembered that a strong wood Bulletln NE-51, 1977 energy market would provide an incentive to Ch. 2—Forestry ● 23 increase the number of stand improvements. be disposed of to minimize the risks of forest This will result in an increased supply of what fires. Stumpage fees for logging national for- is considered commercial-grade timber. Fur- estlands are therefore lower than for compara- thermore, some stands that cannot now be har- ble private lands in the region, in order to vested economically for only lumber or pulp- cover the cost of disposing of the residues. In wood will become economically attractive for the early 1970’s, as a result of a strong demand a combined harvest of lumber, pulpwood, and for paper, some of these residues were col- wood chips for energy. lected and chipped for paper pulp. Currently, however, the residues (about 0.2 Quad/yr) are In the very long term, competition for wood disposed of onsite by burning and other tech- may develop between the energy and materi- niques. If a strong energy market existed, much als/chemicals markets. For the next 20 years, of this could be chipped and used for energy. however, a wood energy market–properly managed—will increase the supply of wood It has been common practice in site prepara- for other uses over what would occur in its tion to use herbicides to kill unwanted plants absence, and indeed this situation is likely to so that preferred trees could regenerate either prevail for at least 50 to 60 years. naturally or artificially. Increasingly, however, the use of herbicides for this purpose is being restricted and in some cases banned (e. g., 2, 4, Other Factors and Constraints 5-T). A strong energy market would provide an additional incentive to harvest the brush and As noted previously, the Forest Service re- other low-quality wood and thereby minimize quires that logging residues on national forests the use of these controversial chemicals.

Net Resource Potential

There is no simple way to assess accurately cut but not used (see figure 5), and there is at the impacts of the various and sometimes con- least 40 Quads (total) of unmerchantable tradictory factors affecting the availability of standing timber. wood for energy. Many of the important fac- Assuming that the demand for traditional tors, such as public opinion, the way the for- forest products doubles by 2000, then 3.4 ests are managed, and the presence of roads, Quads/yr will be needed for finished wood will depend on actions taken in the future. As- products, and 3.9 to 11.2 Quads/yr could be suming, however, that 40 percent of the growth used for energy, provided increased forest potential of the U.S. commercial forestland is management occurs. If, however, the forest eventually accessible, 450 million to 900 mil- products industry becomes energy self-suffi- lion dry ton/yr (7.3 to 14.6 Quads/yr) could be cient by 2000, it could require as much energy available for harvest. as the lower limit of available wood energy, In terms of energy, the forest products in- but three factors will probably alter this simple dustry currently cuts 5.1 Quads/yr of wood, in- projection. First, the increased demand for cluding logging residues (1.3 Quads/yr) and wood products is likely to increase the number stand improvement cutting (0.7 Quad/y r). Of of stand improvements. Second, the energy ef- this total, 1.7 Quads/yr are converted into ficiency of the forest products industry will products sold by primary or secondary manu- probably increase as a result of higher energy facturers, and 1.2 Quads/yr, supplied by wood prices and new processes (such as anthraqui- wastes, satisfies over 45 percent of the in- none catalyzed paper pulping). Third, if the dustries direct energy needs. This leaves about forest products industry requires most of the 2.2 Quads/yr of wood that are currently being available output of 40 percent of the commer- 24 . Vol. Il—Energy From Biological Processes cial forestlands to supply its needs, then addi- These estimates are admittedly approxi- tional roads would be built to access more tim- mate, but a more precise estimate would re- berland. Additional wood that is not of high quire a survey of potential logging sites, land enough quality for lumber, veneer, paper pulp, capability, road availability, and the costs of etc., would therefore become available. In harvesting. light of these factors, it is likely that significant quantities of wood will become available for The results of such a survey could change energy uses outside of the forest products in- these estimates, but 5 to 10 Quads/yr is OTA’s dustry, but this industry could be the major best estimate of the energy potential from ex- user isting commercial forestland

Figure 5.—Materials Flow Diagram for Felled Timber During Late 1970’s (Quads/yr)

Felled timber

Returned to soil Forest products by bacterial industry harvest decomposition or 3.1 burned in forest 2.0

Pulpwood harvests

Primary and secondary manufacturing 0.7 Paper and pulp

, Residues of primary and secondary manufacturing

Total left in forest 2.0 Quads/yr Total used as energy 1.5 Quads/yr Unused residues 0.14 Quad/yr Total products 1.7 Quads/yr

SOURCE Office of Technology Assessment Ch. 2—Forestry . 25

Environmental Impacts

Introduction but may be perceived as detrimental by those who cherish the same forest in its original A forest may be perceived as: state. Hence, it is likely that a wood-for-energy ● a natural ecosystem deserving protection; strategy that increases the areal extent or in- ● a source of materials — renewable or oth- tensity of forestry management will promote a erwise; wide range of reactions . . . even if the physical ● a physical buffer to protect adjacent impacts are fully predictable and if forecasts of areas from erosion, flood, pollution, etc.; these physical changes are believed by all par- ● a source of esthetic beauty; ties. ● a wiIdlife preserve; Environmental evaluation is further compli- ● a source of recreation — hiking, hunting, cated both by difficulties in predicting the etc.; physical impacts and by the strong possibility ● a temporary land use; that even those predictions that can be accu- ● a place to retreat from civiIization; or rately made will not be accepted as credible ● an obstacle to another desired land use by all major interest groups. such as mining or agriculture. The problem of credibility stems largely This range of perceptions is complicated by from the history of logging activities in the the fact that individuals do not perceive all United States and the negative impact it has forests to be alike, and few would attach the had on public perceptions of logging. The same perspective—or value— to all forests. adaptation of the steam engine to logging Thus, the keenest environmentalist may com- around 1870 began an era (lasting into the 20th fortably accept a managed, single-aged pine century) when America’s forest resource was forest in the same terms as he accepts a wheat- mined and devastated .29 The dependence of field, while a lumber company president may logging on the railroads and on cumbersome view a preserve of giant Sequoias with as much steam engines— capital-intensive equipment reverence as a Sierra Club conservationist. that could not easily be moved from site to These perceptual differences make an eval- site–led to the cutting of vast contiguous uation of the environmental effects of a wood- areas. There was virtualIy no attention to refor- for-energy strategy difficult, because many of estation. In fact, it was then thought that most the effects may be valued by some groups as of this land would be used for agriculture, and positive and by others as negative. In other that clearcutting enhanced the value of the words, although some potential effects of land. It also was thought that the timber re- growing and harvesting operations (e.g., ef- source was essentially unlimited and that it fects such as impaired future forest productivi- was unnecessary to worry about regeneration. ty or extensive soil erosion) are clearly nega- Massive cutting followed by repeated fires tive or (in the case of restoration of lands led to the destruction of tens of millions of damaged by mining) positive, other effects are acres of hardwood (in the South and East) and more ambiguous. Changes in such forest char- softwood (in the Lake States, Rockies, and part acteristics as wildlife mix, physical appear- of the Northwest) forest and their replacement ance, accessibility to hikers, and water storage by far less valuable tree types or by grassland. capabilities may be viewed as detrimental or This massive destruction led to a considerable beneficial depending on one’s objectives or public revulsion towards logging, much of esthetic sense. For instance, measures that in- which still survives. It also led to a revulsion crease forest productivity by substituting softwood for hardwood production would be con- 29M Smith, “Appendix L, Maintaining Timber Supply in a Sound Environment” In Report of the Presidents Advisory Pane/ sidered as strongly beneficial by those who orI Timber and the .Environment (Washington, D C Forest Serv- value the forest mainly for its product output, ice, U S Department of Agriculture, 1973)

67-968 0 - 80 - 3 26 . Vol. II-Energy From Biological Processes against clearcutting and even-aged manage- 1. The environmental effects that occur ment within the forestry profession which when different kinds of silvicultural oper- lasted for 20 years; 30 although clearcutting (at ations (including different kinds and in- least the very limited version used today, tensities of cuts, regeneration practices, which involves very much smaller areas than roadbuilding methods, basic management were routinely cut in the past) is now an ac- practices, etc.) are practiced on different cepted and even popular practice in the pro- forest types and land conditions. fession, the attitudes formed by attempts at 2 The kinds and amounts of Iand likely to public education about forest values in the be harvested and their physical-environ- 1930’s and 1940’s linger on. Furthermore, there mental condition. have been enough reports of unsound forest 3 The types of practices, controls, etc., like- management and widely publicized environ- ly to be adopted by those harvesting this mental fights over such management in the in- land. tervening decades to create a sizable constitu- There is an extensive literature describing ency that is generally very skeptical about log- factor #1. However, the range of forest ecosys- ging practices. As a result, assessments that tems and possible silvicultural practices is far focus on the potential positive effects of in- greater than the range of existing research, and creased forest management may be greeted there are as well substantial gaps in the knowl- with skepticism by large segments of the pub- edge of some important cause-effect relation- lic. ships such as the effect of whole-tree removal The prediction of environmental changes and short rotations on nutrient cycling, or, that might occur in American forest areas if de- more generally, the ecosystem response to mand for wood energy grows is extremely diffi- physical pollutants such as sediments and pes- cult. The potential for wood energy identified ticides. previously is based on a “scenario”-a vision Identification and characterization of the of a possible future—that assumes an in- land base most likely to be affected by increased collection of wood residues that are creased wood demand (#2) are complicated by now left to rot in the forest as well as an as- a lack of good land resource data, the lack of sumed intensification of silvicultural manage- information on the precise nature of the future ment on suitable land that would increase wood market, and the complexity of incentives growth rates and timber quality, increasing the that affect the decisionmaking of small wood- supply of nonenergy wood products while pro- land owners. viding a steady supply of wood fuel. This type of strategy could lessen harvesting pressures Predicting the types of practices and envi- on wilderness areas and other vulnerable for- ronmental controls likely to be adopted (#3) is estlands. It probably would be perceived by difficult because State and local regulatory many groups as environmentally beneficial, al- controls generally do not specify or effectively though it would lead to esthetic and ecosys- enforce “best management practices. ” Thus, tem changes on those lands where manage- existing regulations cannot be used as a guide ment was intensified. Given the present institu- to actual practices. Also, although knowledge tional arrangements, however, there is no guar- about the present environmental performance antee that this assumed “scenario” will unfold of the forest industry might provide a starting as outlined. Instead, a combination of Federal, point for gaining an understanding of what to State, business, and other private interests will expect in the future (because most wood-for- respond to a complex market amid a variety of energy operations are more intensive exten- institutional constraints. In order to predict the sions of conventional forestry), it is surprising- environmental outcome of such a response, ly difficult to produce a clear picture of how the following factors must be understood: well the forestry industry is performing. With the exception of a few isolated State surveys ‘“I bid and a detailed survey of erosion parameters Ch. 2—Forestry ● 27

(percentage of bare ground, compaction, etc.) ated with alternative wood-for-energy systems. in the Southeast,31 there appears to be a severe In each case, the discussion will attempt to lack of surveys or credible assessments of ac- draw a distinction between clearly positive or tual forestry operations and their environmen- negative pollution and land degradation and tal impacts. As a result, a critical part of the restoration impacts and the more ambiguous basis for an adequate environmental assess- ecosystem and esthetic impacts. Because the ment is unavailable. environmental effects of silviculture are ex- ceedingly varied and complex and because a Because of these limitations, this discussion number of good reviews are available, the dis- generally is limited to a description of poten- cussion highlights only the major and most tial impacts, although a few of the impacts widespread impacts. It is stressed that few if described are inevitable. The economic and any of the environmental relationships de- other incentives that influence the behavior of scribed in the discussion are applicable to all those engaging in forestry are examined to situations. determine how probable some of these impacts are. The types of controls and practices available to moderate or eliminate the nega- Environmental Effects of tive impacts also are described. Conventional Silviculture As discussed above, wood for energy may be The practice of silviculture can have both obtained from several sources. With the positive and negative effects on the soils, growth of a wood-for-fuel market, the residue wildlife, water quality, and other components of slash from logging may be removed and of both the forest ecosystem and adjacent chipped. Thinning operations may become lands. Table 11 provides a partial list of the more widespread because the wood obtained potential environmental effects of convention- will have considerable value as fuel. Stand al silviculture. The magnitude of these impacts conversions— clearing of low-quality trees fol- in any situation, however, depends almost en- lowed by controlled regeneration–as well as tirely on management practices and on the harvesting of low-quality wood on marginal physical characteristics of the site, i.e., type of lands may increase, also because of the in- trees and other vegetation, age of the forest, creased value of the fuelwood gained. New soil quality, rainfall, slope, etc. It is also impor- harvesting practices such as whole-tree remov- tant to remember that most of the negative im- al may become more common. Waste wood pacts generally are short term and last only a from milling and other wood-processing opera- few years (or less) over each rotational cycle. tions will certainly be more fully utilized. Finally, wood “crops” may be grown on large Erosion has always been a concern in silvi- energy farms. culture, especially in logging operations (and particularly in road construction). Undisturbed Many of these activities are similar to forests generally have extremely small erosion (though usually more intensive than) conven- rates — often less than 75 lb of soil per acre per tional logging. In addition, other activities as- 32 year — and in fact tree planting is often used sociated with using wood as a long-term ener- to protect erosion-prone land. * Increased ero- gy supply — including tree planting, pesticide sion caused by logging, however, varies from and fertilizer application, etc. — are similar or negligible (light thinning and favorable condi- even identical to “ordinary” silvicultural activities. This section, therefore, first discusses the ‘zEnvironmenta/ Implications of Trends in Agriculture and $Ilvi- general impacts of silviculture and then de- cu/ture, Vo/ume 1 Trend Identification and Evacuation (Washing- scribes any changes or added effects associ- ton, D C Environmental ProtectIon Agency, December 1978), E PA-600/3 -77-l 21 *However, from a historical perspective, all land forms go “C Dlssmeyer and K Stump, “Predicted Erosion Rates for through natural erosional cycles that produce much htgher rates Forest Management Actlvltles and Conditions Sampled In the of soil loss These rates are often drwen by natural catastrophic Southeast,” USDA Forest Service, April 1978 events Including wildfire and storms i--

28 . Vol. II—Energy From Biological Processes

Table 11.-Potential Environmental Effects of and rill erosion rate on intensively managed Logging and Forestry agricultural land averages 6.3 ton/acre-yr. Water Most forestland is harvested at most once ● increased flow of sediments into surface waters from logging erosion (especially from roads and skid trails) every several decades and the increased ero- ● clogging of sfreams from logging residue sion generally lasts only a year or two on the ● leaching of nutrients into surface and ground waters majority of the affected acreage. Increased ● potential improvement of water quality and more even flow from for- estation of depleted or mined lands erosion from poorly constructed roads, how- ● herbicide-pesticide pollution from runoff and aerial application (from ever, may last longer. a small percentage of forested acreage) ● warming of streams from loss of shading when vegetation adjacent The processes involved in erosion of forest- to streams is removed Iand are stream cutting, sheet and gully ero- Air sion, and mass movement of soil. Erosion dan- . fugitive dust, primarily from roads and skid trails ● emissions from harvesting and transport equipment ger increases sharply with the steepness of the ● effects on atmospheric CO2 concentrations, especially if forested landscape, and the most common form of this land is permanently converted to cropland or other (lower biomass) erosion is mass movement. Mass movement use or vice-versa ● airpollution from prescribed burning “includes abrupt or violent events such as Land landslides, slumps, flows and debris ava- ● compaction of soils from roads and heavy equipment (leading to fol- lanches, as well as continuous, almost imper- lowing two impacts) 35 ● surface erosion of forest soils from roads, skid trails, other disturb- ceptible creep phenomena." Occurrence of ances mass movements is most often associated with ● loss of some long-term water storage capacity of forest, increased steep slope conditions where the forest soil is flooding potential (or increased water availability downstream) until underlaid with impermeable rock.36 These revegetation occurs ● changes in fire hazard, especially from debris movements are natural processes associated ● possible loss of forest to alternative use or to regenerative failure with the downwearing of these steep slopes, ● possible reduction in soil quality/nutrient and organic level from but they can be triggered by man’s activities. short rotations and/or residue removal (inadequately understood) ● positive effects of reforestation–reduced erosion, increase in water In contrast, sheet and gully erosion are rare in retention, rehabilitation of strip-mined land, drastically improved undisturbed forests, but they can be triggered esthetic quality, etc. by soil disturbances caused by careless road ● slumps and landslides from loss of root support or improper road design construction or logging practices. ● temporary degrading of esthetic quality Ecological The major causes of erosion problems in ● changes in wildlife from transient effect of cutting and changes in forestry operations are the construction and forest type use of roads and other activities that may com- ● temporary degradation of aquatic ecosystems 37 ● change in forest type or improved forest from stand conversion pact or expose soil or concentrate water. The compaction caused by the operation of heavy SOURCE Office of Technology Assessment machinery can reduce the porosity and water- holding capacity of the soil, encouraging ero- tions) to hundreds of tons per acre per year sion and restricting vegetation that eventually (poorly managed clearcuts on steep slopes in would reduce erosion. Roads and skid trails high rainfall areas) .33 comprise up to 20 percent of the harvest A recent Environmental Protection Agency area, 38 and the total area that may be com- (EPA) report suggests the loss of 7 or 8 tons of pacted at a site may range up to 29 percent in sediment per acre per year as a mean value for recently harvested forests, although the varia- “Earl Stone, “The Impact of Timber Harvest on Soils and tion around this mean is very large. ” To place Water, ” Report of the President’s Advisory Panel on Timber and the Environment (app M, Washington, D C Forest Service, U S this rate in perspective, the continuous sheet Department of Agriculture, April 1973) “Environmental Implications of Trends in Agriculture and Si/vicu/ture, vo/. 1., op cit “Environmental Readiness Document, Wood Commercializa- “Stone, op cit tion, draft (Washington, D.C : Department of Energy, 1979). “Draft 208 Preliminary Non-Point Source Assessment Report J4Envlronmenta/ /mp/;ca(/on5 of Trends in Agriculture and SIIVI- (Augusta, Maine Land Use Regulatory Commission, State of cuhure, vol. 1, op cit Maine, 1978 Ch. 2—Forestry ● 29 some instances. 39 Although in most areas the ding” logs may expose the subsoil, or compact thawing and freezing cycle allows compacted the soil. Exposing the surface is a problem soil to recover in 3 to 10 years, recovery takes when the soil is highly erosive or when water far longer when, as in parts of the Southeast, concentrates, but is usually not a major ero- this cycle does not occur.40 Also, when com- sion problem. The deeper disturbances of compaction is very severe, recovery may take con- paction and of cutting into the soil create siderably longer than 10 years; old logging more significant erosion problems, especially roads are still visible in the Northeast, even when they occur parallel to the flow of water. with the frost cycle. Most surveys of logging have concluded that the hauling or skidding of logs “generally does The vulnerability of logging roads to erosion not lead to appreciable soil erosion or im- is related to topography and soil type as well paired stream quality;”42 however, the same as to road design. Roads developed on gentle surveys conclude that “exceptions are com- to moderate slopes in stable topography pose men, ” and logging in vulnerable areas, under few problems with the exception of careless wet weather conditions, or with inappropriate movements of soil during construction. Large equipment are thought to be important prob- areas of forestland served by such roads draw 41 lems in the industry. little attention or criticism. Erosion caused by the actual cutting of the The great majority of difficulties and haz- trees generally is considered to be relatively ards arise, however, when roads are con- unimportant. Vegetation usually regenerates structed on steep terrain, cut into erosive soils quickly and reestablishes a protective cover on or unstable slopes, or encroach on stream the land, preventing surface erosion except in channels. Steep land conditions present a areas where other components of the logging dilemma for road development, and criteria operation have damaged the soil. “Many ob- for location, design, and construction that are servations and several studies on experimental satisfactory on even moderate slopes may lead watersheds demonstrate that sheet and gully to intolerable levels of disturbance on steep erosion simply do not occur as a result of tree lands. Building a road on a slope involves cut- cutting alone, even on slopes as steep as 70 ting into the slope to provide a level surface. percent. ”43 The soil removed from the cut is used as fill or dumped. The steeper the slope, the more soil However, land that is vulnerable to mass that must be disposed of and the more difficult movements may be damaged by tree cutting. is the job of stabilizing this soil. In the absence The decay of the old root systems will remove of proper attention to soil and geology, road crucial support from a vulnerable slope faster design (especially alinement and drainage), than it can be replaced by the root systems of and other factors, surface erosion from road new growth; within 4 to 5 years after tree cut- and fill surfaces can continue for years. Road- ting (or fire), mass movement potential may in- building on steep slopes may also remove crease dramatically. Forests in the Northwest enough support from the higher elevations to United States and coastal Alaska are the main cause mass failures; problems created by the areas for this type of damage potential .44 road cut may be aggravated by inadequate The method of clearing for forest regenera- drainage allowing further cutting away of sup- tion may also affect erosion potential. Inten- porting soil. sive mechanical preparation of land before Aside from roads, the movement of logs tree planting (i.e., use of rakes, blades, and from the harvest site to loading points may other devices to reduce a forest to bare ground present considerable erosion potential. “Skid- to favor reproduction of pine) can cause very

“fnvlronmental Implications of Trends in Agriculture and Silvi- cu//ure, vo/ 1 , op clt “lbld ‘“lbd ‘Jlbld “Stone, OP clt “’lbld 30 . VOI. II—Energy From Biological Processes serious erosion problems. This practice is oc- ent loading is usually short-lived, because re- curring on hilly sites in the South that have vegetation of the site slows runoff and leach- been depleted by intensive cotton production ing, increases nutrient uptake, and, by shading in the past; it “may foster a dangerous cycle of and cooling the soil, slows the decomposition topsoil and nutrient loss and increased sedi- of organic material and consequent nutrient ment loading in streams. ”45 Poorly managed release. raking may have adverse effects on forest pro- The effects of nutrient enrichment are ag- ductility. ” Area burning can also badly dam- gravated by the decomposition of organic mat- age forest soils if managed improperly or if ter from slash that is swept into streams, and used on improper soils. Although suitable for by any water temperature increases caused by highly porous, moist soils (where much of the loss of streambank shading* (the temperature surface cover is not consumed), poorly man- increases speed up eutrophication and further aged burning may consume most of the cover reduce oxygen content of the water). Tempera- and leave the soil exposed to surface erosion. ture increases may also directly harm some (However, area burning is considered to have a freshwater ecosystems by affecting feeding lesser potential to degrade productivity than behavior and disease incidence of cold water raking.”) Burning may also represent a signifi- fish. cant local source of air pollution. On the other hand, “controlled” burning may reduce future Logging operations affect water supply and fire hazard by reducing slash buildup and may may decrease a watershed’s ability to absorb favor regeneration on the site of fire-resistant high-intensity storm waters without flooding trees. (although this problem may have been exaggerated somewhat in the past). The sediment resulting from the erosion described in this section is “the major cause of The possibility of increased flooding stems impaired water quality associated with log- from two causes. First, cutting the forest re- ging.” 48 These sediments are directly responsi- duces the very substantial removal by trans- ble for water turbidity, destruction of stream piration of water from underground storage. bottom organisms by scouring and suffoca- During the period before substantial revegetation, and the destruction of fish reproductive tion has taken place, the amount of this long- habitat. Sediments also carry nutrients from term “retention storage” capacity available to the soil. Nutrient pollution is further increased absorb floodwaters will be lessened and peak by increased leaching and runoff as increased stream flows may rise. For example, increases solar radiation reaches the forest floor and in peak flows of 9 to 21 percent in the East and warms it, microbial activity (which transforms 30 percent in Oregon following clearcutting nutrients to soluble forms) accelerates and nu- have been reported. These increases are usual- trient availability increases (this soil heating ef- ly observed only during or right after the grow- fect also has been known to retard regenera- ing season, where continual drawdown of stor- tion, especially on south-facing slopes, by kill- age would be occurring had the trees not been ing off seedlings). The increased nutrient load- cut (floods occurring during the winter, as in ing of streams may have a variety of effects, in- the Northwest, may be unaffected or less af- cluding accelerated eutrophication and oxy- fected because drawdown would not normally gen depletion. Fortunately, the increased nutri- be occurring). This decrease in storage capacity apparently is not significant unless at least 45 Environrrrenta/ Effects of Trends, vo/. 2 (Washington, D C En- 20 percent of the canopy is removed.49 Second, vironmental Protection Agency, December 1978), E PA-600/ 3-77-121 “Stone, op clt *The extent of any Increases depends on stream volume, “lbld degree of removal of understory vegetation, and several other 4“Silviculture Activities and Non-Point Pollution Abatement: A factors In many cases, no significant effects occur Cost-Effectiveness Ana/ysis Procedure (Washington, D C Forest “An Assessment of the Forest and Range Land Situation in the Service, US Department of Agriculture, November 1977), United States (Washington, D C Forest Service, U S Depart- E PA-600/8-77-Ol 8 ment of Agriculture, 1979), review draft Ch. 2—Forestry ● 31

damage to forest soil increases runoff and in- ging and agriculture as well as “unnatural” for- hibits the action of even the temporary “deten- est fire suppression that gradually replaced tion storage” potential wherein water is tem- conifer forests with mixed hardwoods. porarily stored in pores in the upper soil layers All types of replanting are accompanied by and can be delayed from reaching streams for major changes in habitats available for wild- anywhere from several minutes to several life. In the short term, any wood-harvesting days. Although treecutting, even clearcutting, operation, other than large area clearcutting, is not likely to affect this temporary storage usually increases wildlife populations because capacity, compaction of the soil by roadbuild- mature forests normally do not support as ing, log skidding, and operation of heavy ma- great a total population of wildlife as do young chinery may reduce the infiltration of water growing forests. Many species require both into the soil if the compaction occurs over a cleared and forested area to survive, and thus, wide area50 and thus drastically reduce stor- the “edges” created by logging operations are age. Area burning on coarse-textured soils can particularly attractive to deer and other spe- create a water-repellant layer that would also cies. Other species dependent on subclimax decrease this infiltration and thus reduce the habitats (such as eastern cottontails) will also soils’ capacity for detention storage. 51 increase following logging, while species de- The reduction of transpiration that is caused pendent on mature climax forests (e. g., wolver- by timber harvest may be beneficial by in- ine, pileated woodpecker) will decline.54 creasing stream flow and groundwater supplies Although the desirability of the ecosystem in water short areas. Also, carefully structured changes caused by logging may always be sub- cuts can be used to trap and maintain snow ject to one’s point of view, different forestry accumulation, greatly reducing evaporation practices tend to have varying effects that may losses, It is claimed that by using such tech- be judged unambiguously from the standpoint niques, water yield from commercial forest- of wildlife diversity and abundance: Iand in the West could be increased, supplying millions of additional acre feet at a cost of a Forest management practices that reduce few dollars per acre foot. 52 structural diversity of habitat, such as extensive old growth clearcutting, the removal of Large-scale forestry operations often dras- snags that provide wildlife food and nesting tically alter local ecosystems, even for the long sites, and conversion to plantation manage- term. Wetlands in the South are being drained ment will generalIy reduce wildlife abundance and pine forests are being created with the aid and diversity by reducing habitat essential to of substantial applications of phosphate fertil- many species. Conversely, animal diversity izers. In the process, aquatic ecosystems are and wildlife abundance generally will be in- being replaced by terrestrial ones and some creased by opening up dense stands, making critical wildlife habitats, especially for water- small patch cuts, or by conducting other tim- fowl, are being destroyed.53 In the Pacific ber management activities that increase structural diversity and provide a wide mix of hab- Northwest, old stands of Douglas fir are being itat types. 55 replaced by single-aged plantings of the same species. EIsewhere, mixed hardwood forests Current pesticide and fertilizer use in U.S. are being replaced by plantations of conifers. forests is low, In 1972, insecticides were used In many cases, however, the ecosystems being on only 0.002 percent of commercial forest- replaced are themselves the result of past log- Iands, and fertilizers were used on less than

500,000 acres. 5’ Because long-rotation logging “’Stone, op clt and removal of only boles generally do not “R M Rice, et al , “Erosional Consequences of Timber Har- vesting An Appraisal, ” Watershed jr-i Trarrs/t/on, (Urbana, Ill “An Assessment of the Forest and Range Land Sltuatlon In f~e American Water Res Assoc Proc Ser 14, 1972) United States, op clt “An As~essment of the Fore$t and Range Land S/ fuat/on In the “lblci Un/ted State\, op cit ‘6Vo/ 1 Errvironmenta/ lrnpllcatlon~ of Trends In Agrlcu/ture 5‘Vo/ //, fnv(ronmenta/ Effects of Trend\, op clt and S\lv/cu/ture, op c(t 32 . VOI. II—Energy From Biological Processes deplete nutrients from forest soils, the most ception of the environmental effects of clear- important use of fertilizers is on soils that are cutting in particular and logging in general. As naturally deficient in nutrients or that have noted earlier, this perception has been exag- been depleted by past farming practices. For gerated by a number of factors including the example, intensive cotton production in the grim history (1870-1930) of forest exploitation Southeast seriously depleted soils and much of in the United States, the former revulsion this land was abandoned long ago. Phosphate against clearcutting practices within the for- fertilization has allowed this land to become estry profession itself during the 1930’s and productive in the growth of softwood forests. 1940’s, and continued attacks against logging Pesticides generally are used in forest manage- by the environmental community. Although a ment to control weed vegetation during refor- logged-over area may be no uglier, objectively estation or to combat serious outbreaks of in- speaking, than a harvested field, the public sect pests. There is considerable controversy perception of the two vistas is vastly different. over aerial spraying of insecticides to control the gypsy moth and other damaging insects. All reviews of logging and general forestry Also, circumstantial evidence exists that cer- impacts stress the importance of regional dif- tain herbicides in recent use may have caused ferences –as well as extensive site-specific dif- outbreaks of birth defects and other damage ferences – in determining the existence and when inadvertently sprayed over populated magnitude of environmental effects. Figure 6 areas. Although the existence of these effects presents a summary of those characteristics of has been vigorously denied by the manufactur- U.S. forest regions that are most relevant to ers, and although pesticide use in forests is a potential silvicultural impacts. Because the tiny fraction of the use in food production and descriptions in figure 6 are, of necessity, much is likely to remain so,57 this use is likely to con- oversimplified, they are meant to give some tinue to be a source of disquiet accompanying perspective of the general range of environ- intensive management of forests. mental conditions and problems in American forestlands and should not be considered as Silvicultural activities, and especially inten- fully representing all of the major conditions sive harvesting operations, strongly affect the and problems in these lands. esthetic appeal of forests. The immediate af- termath of intensive logging is universally considered to be visually unattractive, especially Potential Environmental Effects of where large amounts of slash are left on the Harvesting Wood for Energy site. Therefore, wood harvesting has a strong potential to conflict with other forest uses This section discusses the activities— har- such as recreation or wiIderness. vesting logging residues, whole-tree removal, intensifying and expanding silvicultural man- The significance of any negative effects de- agement, and harvesting for the residential pends on the nearness of logging sites to activi- space-heating market — which are characteris- ty areas or to scenic vistas, the rapidity of re- tic of an expansion in the use of wood as an vegetation, and the extensiveness of the oper- energy source. ation. Therefore, the Forest Service seeks to route trails away from active harvesting sites, to avoid interrupting vistas, and to plan the ex- Harvesting Logging Residues tent and shape of the areas to minimize visual and Whole-Tree Removal impacts. The harvesting of logging residues for an en- The negative effect on the esthetic and rec- ergy feedstock has potential for both positive reational quality of forests caused by logging and negative environmental impacts depend- may be aggravated by a negative public per- ing on the nature of the forest ecosystem and the previous manner of handling these resi- “lbld dues. Ch.2-Forestry • 33

Figure 6.-Envlronmental Characteristics 01 Forest Regions

SOURCE: Silviculture Activities and Nonpoint Pollution Abatement: A Cost·Effectiveness Analysis Procedure (Washington, D.C.: Forest Service, U.S. Department of Agriculture, November 1977). 34 ● Vol. II—Energy From Biological Processes

In forests where wood residues–tops, limbs, Table 12.-Environmental Impacts of and possibly leaves and understory — are rou- Harvesting Forest Residues tinely gathered into piles for open burning (this Water is required in forest fire prone areas of the decrease in clogging of streams caused by entry of slash West), residue use for energy production is en- increased short-term flow of sediments into streams because of loss of erosion control provided by residues, soil damage caused by removal vironmentally beneficial. It eliminates the air operations; somewhat counteracted by decline in broadcast burning, pollution caused by this burning and has essen- which at times destroys surface cover and causes erosion potential to tially no additional adverse impacts except increase possible changes in long-term flow of sediments where residue those incurred in physically moving the residue removal affects revegetation; this effect is mixed out of the forest (and burning it, with controls, changes in herbicide usage–on the one hand, chemical destruction of in a boiler). In forests where residues would growing residues (valueless trees) will cease; on the other, broadcast burning no longer effective in retarding vegetative competition to new otherwise be broadcast burned, physical re- tree growth, herbicide use may increase moval prevents some of the potential adverse increased short-term nutrient leaching because of increased soil tem- effects of burning — especially destruction of a peratures, accelerated decomposition portion of the organic soil layer. The removal Air ● reduction in air pollution from forest fires does, however, subject the soil to compaction ● reduction in air pollution from open burning of residues (if the residues or scraping damage by the mechanical remov- normally are broadcast burned or burned after collection) al process that would otherwise be avoided. ● dust from decreased land cover, harvesting operations Also, broadcast burning is, at times, used to Land ● potential depletion of nutrients and organic matter from forest soils and control weed vegetation, and in some circum- possible long-term loss of productivity (inadequately understood) stances herbicide use may be substituted if ● short-term increase in erosion and loss of topsoil, possible long-term burning cannot be practiced. decrease or increase ● reduction in forest fire hazard Where logging residues are normally left in ● short-term decreased water retention, increased runoff (and flooding hazard) until revegetation takes place; aggravated by any soil compac- the forest, institution of a residue removal pro- tion caused by removal operation gram will have mixed environmental effects Other which are summarized in table 12. ● change in wildlife habitat—bad for small animals and birds, good for large animals unless serious erosion results A worrisome effect of residue removal is the ● changes in tree species that can regrow increased potential for long-term depletion of ● esthetic change, usually considered beneficial when slash is heavy ● reduction in bark beef/es and other pathogens that are harbored by nutrients from the forest soils and consequent residues declines in forest productivity. These effects SOURCE Office of Technology Assessment. are not well understood and although nutrient cycling in natural and managed forests has been extensively studied, few of these studies Further study and careful soil monitoring have included the effects of residue removal.58 would allow the use of fertilizers to compen- The existing studies indicate that short-rota- sate for nutrient depletion, but fertilizer ap- tion Southern forests may be more susceptible plication is energy intensive; it may increase to depletion than longer rotation Northern the flow of nutrients to neighboring streams, forests, and that marginal sites suffer far more and its correct use may be difficult to ad- heavily than forests with fertile soils. 59 60 61 62 minister for smaller stands. Also, successful application may be difficult unless the nutri- “C. J. High and S E. Knight, “Environmental Impact of Har- vesting Noncommercial Wood for Energy Research Problems, ” ent depletion is a simple one involving only

Thayer School of Engineering, Dartmouth College paper DSD one or a few nutrient types< No. 101, October 1977. “E H White, “Whole-Tree Harvesting Depletes SOII Nutri- Residues serve a number of ecological func- ents, ” Can 1. Forest. Res. 4.530-535,1974 tions in addition to nutrient replenishment, ‘“J R Jorgensen, et al., “The Nutrient Cycle Key to Con- tinuous Forest Production, ” /. Forestry 73.400-403, 1975. and their removal will eliminate or alter these “J R Boyle, et al , “Whole-Tree Harvesting. Nutrient Budget functions. They provide shelter and food to Evaluation, ” ). Forestry 71 760-762 small mammals and birds, provide a temporary “C F Weetman and B Webber, “The Influence of Wood Har- vesting on the Nutrient Status of Two Spruce Stands, ” Can. /. For- food supply for deer and other larger mam- est. Res. 2 351-69, 1972 mals, moderate soil temperature increases that Ch. 2—Forestry ● 35 normally occur after logging, provide some is generally considered to be negative when protection to the forest floor against erosion, they are left at the logging site; when densely and are a source of organic matter for forest forested areas are cut, residues will completely soils. Thus, removal of residues will reduce cer- cover the ground with several feet of unsightly tain wildlife habitats and may expose the for- slash. Therefore, removal of residue will, in a est floor to some additional erosion above and positive sense, reduce the number and severity beyond that caused by conventional logging. of forest fires and pest infestations, improve Higher soil temperatures resulting from loss of esthetics, and reduce the potential for stream the shade provided by residue cover will accel- clogging. erate organic decomposition activity and may “Whole-tree harvesting” is really a variation lead to a period of increased nutrient leaching of residue removal with the bole and “resi- before revegetation commences. Also, the in- due”- branches, leaves, twigs– removed in creased rate of organic decomposition cou- one integrated operation. It is most likely to pled with the removal of a primary source of occur when the entire tree is to be chipped for organic matter may lower the organic content fuel or some other use. of forest soils. Declines in soil organic matter are expected to be accompanied by declines in The problems of long-term nutrient and or- nitrogen-fixing capacity, soil microbial activity ganic matter depletion from whole-tree har- rates, and cation exchange capacity, all con- vesting are basically the same as those of sidered to be important determinants of long- residue removal, and whole-tree logging simi- term forest health. 63 64 The present scientific larly removes far greater nutrients and organic understanding of organic matter removal is, matter from forest soils than do other conven- however, insufficient to allow a determination tional methods. Whole-tree removal of Nor- of the significance of these possible effects. way spruce, for example, results in a loss of 2 to 4 times more nitrogen, 2 to 5 times more The extensive residue left on the forest floor phosphorus, 1.5 to 3.5 times more potassium, after cutting dense stands can inhibit revegeta- and 1.5 to 2.5 times more calcium than convention, especially in softwood forests. To the ex- tional Iogging. 65 In addition, ground disturb- tent that residue removal may promote new ance from the actual tree removal is likely to vegetation, this will counteract the removal’s be worse with whole-tree harvesting when the short-term negative erosion and nutrient-leach- fully branched trees are dragged off the log- ing effect (as long as removal is not so com- ging site, eradicating understory vegetation in plete as to eliminate the light mulch necessary the process. This disturbance, besides promot- to shade the surface and maintain soil mois- ing erosion, will accelerate organic matter de- ture). composition. As noted previously, however, Residues also provide a habitat for disease the effects of these organic matter and nutri- and pest organisms such as the bark beetle ent removals on long-term forest productivity and, when washed into neighboring streams, are poorly understood. may clog their channels and degrade water quality. They add considerably to the inci- Intensifying and Expanding dence and intensity of forest fires, especially in Silvicultural Activities the West. Also, the esthetic impact of residues The creation of new energy markets for *‘E L Stone, “Nutrient Removals by Intensive Harvest— Some wood will have a significant effect on the eco- Research Gaps and Opportunttles, ” Proceedings Impact of Har- nomics of managing forested land, including vesting on Forest Nutr\ent Cycle (Syracuse, N Y State Unlverslty of New York, College of Environmental Science and Forestry, land not currently considered to be high-grade 1 979) “E H Wh/te and A E Harvey, “Modlficatlon of Intensive Management Practices to Protect Forest Nutrient Cycles, ” Pro “E Malkonen, “The Effects of Fuller Biomass Harvesting on ceedlngs /rnpact of Har\,e$t/ng on Forest Nutr/ent Cyc/e [Syra- SOI1 Fertlllty, ” SyrnposIurn on the Harvest~ng of a Larger Parf of cuse, N Y State Untverslty of New York, College of Environ- the Forest Biomass (Hyvlnkaa, Finland E conomlc Commlsslon mental Science and Forestrv, 1979) for Europe, Food and Agriculture Organlzatlon, 1976) 36 ● Vol. II—Energy From Biological Processes

forest. New lands will be harvested and silvi- may provide important environmental benefits cultural practices will intensify. in the form of decreasing logging pressures on lands that combine high-quality timber with One effect will be the expansion of logging competing values that would be compromised onto lands that are not now in the wood mar- by logging. Unfortunately, a decrease in log- ketplace. The operational costs of logging ging pressures on one segment of America’s some of these lands cannot, at present, be re- forests may be coupled with an undesirable in- couped through increased property values, the creased pressure on another segment. sale of the harvested wood, or the value of future growth of a regenerated forest. Additional A particular fear associated with the rise in lands that currently are economically attrac- demand for “low quality” wood is that mar- tive targets for logging activities (stand conver- ginal, environmentally vulnerable lands with sion, clearing for nonforest use, etc.) are with- stands of such wood may become targets for held by their owners for a variety of reasons logging. Much of this land that may be vulner- (their higher valuation of the land’s recre- able to logging for energy, although “poor” ational potential, fear of environmental dam- from the standpoint of commercial productivi- age, etc.). As an energy market for wood devel- ty, is valuable for esthetic, recreational, water- ops, however, harvesting part or all of the shed protection, and other alternative forest wood resource on these lands will become in- uses. These forest values may be lost or com- creasingly attractive. promised by permanent clearing or by harvesting on sites where regeneration may be a prob- The logging of some forests that would lem. For example, forests in areas with margi- otherwise be untouched (or, perhaps more nal rainfall —e. g., in the Southwest— may be realistically, that would only be logged at particularly vulnerable to regeneration failures some later time) may be viewed as beneficial and thus may be endangered by a growth in by some groups. Most reviews depict American wood demand. On lands with poor soils and forests as being characterized by “overmature steep slopes, clearcutting and other intensive stands of old-growth timber, especially in the forms of harvesting create a high potential for West, and . . . many stands, mainly in the East nutrient depletion, mass movement, and other and South, that were repeatedly mined of good 66 problems as described earlier. Because, as dis- trees in earlier, more reckless times. ” Conver- sion of such stands is often characterized as a cussed later, the Federal Government main- step towards a healthier forest, because tree tains supervisory control over forest opera- growth generally is enhanced and more “desir- tions on federally owned lands, this potential problem is likely to be concentrated on private able” tree species are introduced. Where whole-tree harvesting or residue removal is lands. The overall danger is somewhat miti- practiced, the forest may become more acces- gated, therefore, by the Federal Government’s sible to hikers and may be more esthetically ownership of a significant percentage of the most vulnerable land. appealing. The extent to which all this is considered a benefit depends heavily on one’s per- It is difficult to predict whether wood-for- spective, however, and optimizing commercial energy operations will tend to gravitate to the value is not necessarily synonymous with opti- poorer quality and more vulnerable lands. The mizing other values such as ecosystem mainte- several factors that will determine the tenden- nance or wildlife diversity. cy of wood-for-energy harvesting to gravitate As discussed later, expansion of silvicultural to vulnerable lands include: management onto suitable lands, combined 1. The direct cost of wood harvesting, – De- with an increase in the intensity of manage- velopment of more versatile harvesting ment on existing commercially managed lands, equipment can lower the cost of operating on steep slopes and promote harvest- “Smith, op. cit ing on vulnerable lands. Ch. 2—Forestry ● 37

2. The stringency and enforcement of envi- fy energy demands could be expected to lead ronmental standards. –The stronger the to exploitation of lands particularly vulnerable controls, the more likely it is that loggers to environmental damage. will avoid the more vulnerable stands. These references may have overlooked sev- 3. The price of woodchips for energy. –At a eral factors, however: high enough price, the “value-added” to the land by clearing will become less im- 1. As noted previously, there is considerable portant, and poorer quality lands will be- forest acreage of high quality— low come more attractive targets for harvest- slopes, rich and nonerosive soils, ade- ing. quate rainfall —with low-quality timber 4. The price of agricultural land and “high growing on it. This is especially true in the value” forestland. –At high prices, wood East. harvesting for energy would tend to gravi- 2. The cost of harvesting timber on flatter— tate to higher quality, less erosion-prone/ and thus less erosive —slopes is consid- depletion-prone lands because clearing erably less than on steep-sloped lands. for agriculture or stand conversion will be These flatter lands presumably would be more profitable. the first choice for harvesting. 5. The distribution of different soil/slope/ 3. The higher quality, less vulnerable sites rainfall conditions in forestland potential- offer the landowner the economic incen- ly available for cutting. tive of an added return from regrowth of 6. The attitude of private landowners, who high-quality timber or else alternative currently own much of the land available land uses such as farming. for clearing but who often are reluctant to 4. Increases in land prices for rural acreage allow harvesting. with high recreational and esthetic value 7. The cost of transporting wood. – Because have increased the economic incentive to the higher this cost, the more likely it is guard against environmental damage that that local shortages could force harvest- would compromise these values. ing onto vulnerable lands. On balance, it would appear that market Except for (1) and (5), these factors may be ex- pressures would tend to favor the harvesting of tremely volatile and will themselves depend on the less environmentally vulnerable lands. the availability of alternate fuels, the state of However, variations of land availability from the economy, etc. Except for forestland in the region to region, landowner decisions based on Southeast, the data necessary to define (5) are other than land suitability grounds, and other not available. factors are likely to lead to some level of inappropriate harvesting — especially if the current The Department of Energy, in its draft state of regulatory “laissez-faire” continues “Wood Commercialization Environmental 67 (see discussion on “The Institutional Climate Readiness Document,’’ asserts that the sites for Environmental Control”). with “nutrient deficiencies and delicate nutrient balances, and subsequently low productiv- A second effect of new energy markets for ity . . . are the non-commercial forests that wood will be an intensification of forest man- often are considered available for whole-tree agement—especially of thinning— because harvest for energy.” And a recent EPA report part or all of its cost will be recouped through asserts that “areas previously left unlogged use or sale of the collected wood. Residue . . . are most often increasingly steep with dif- removal or whole-tree harvesting, discussed ficult terrain.’’” Both of these statements im- previously, are likely to be another facet of ply that an areal expansion of logging to satis- this management intensification. The process of removing trees that are dead ‘7 Environmental Read/ness Document, Wood Commercializa- tion, op clt or diseased, stunted, poorly shaped, or of “un- bOVo/ II, En vjronmenta/ Effects of Trends, OP c It desirable” species is considered by foresters to 38 . VOl. II-Energy From Biological Processes be beneficial to the forest. Thinning allows in- quality forest and the increased use of thinning creased growth in the remaining trees, esthetic- on commercial forestlands may have, as its ally and physically “opens up” the forest, and most important effect, a decrease in the pres- may allow some additional growth of under- sures to log forests that have both high-value story vegetation if the thinning is extensive timber and strong nontimber values —recrea- enough. If heavy machinery is used, however, tion, esthetic, watershed protection, etc. — and resulting soil compaction can cause adverse that may be quite vulnerable to environmental impacts, and care must be taken during the damage. Analysts such as Marion Clawson of thinning operation to avoid damaging the trees Resources for the Future have long argued that that remain. the management of American forestland is extremely inefficient, that by concentrating in- A critical argument in favor of thinning and tensive management practices on the most , other logging operations is that these activities productive lands we could increase harvest result in increased wildlife populations and di- yields while withdrawing from silviculture less versity. The definition of “diversity” is critical productive or more environmentally vulner- to this argument. There is a substantial differ- 69 70 able lands. An expansion of wood use for ence between maximizing diversity in a single energy and the consequent creation of a strong forest stand and maximizing it in the forest market for “low quality” wood may have this system composed of many forest stands in a beneficial effect. region. The first definition may be well served by more intensive management because such OTA estimates that placing 200 million management provides more “edges” and un- acres of commercial forestland into intensive derstory vegetation for browse. On the other management (full stocking, thinnings every 10 hand, many species will suffer from such man- years, 30- to 40-year rotations) could allow agement. A great many species depend for wood energy use to reach 10 Quads annually their food and shelter on “unhealthy” –dead, while the availability of wood for nonenergy dying, rotten —trees that would be removed in products might double its 1979 value. Alter- a managed forest, and other species cannot natively, the same result might be achieved by tolerate the level of disturbance that would be using less intensive management on a larger caused by thinning operations. Maintaining di- acreage. The nature of any actual benefits, versity in a forest system must include protect- however, are dependent on the following con- ing these species by deliberately leaving un- siderations: managed substantial portions of the forest or a ● Major effects on the availability of high- percentage of the individual stands within the quality timber probably would not occur entire system. in regions where officially desig- for a number of years. Some additional nated wilderness areas or other protective high-quality wood might be available im- measures are adequate, intensive management mediately from stand conversions and on the remaining stands may be considered harvest of noncommercial timber, and (even by environmental groups) as benign or some in about 20 years from timber beneficial if good management practices are growth in stands that required only thin- carefully followed. In other regions, especially ning for stand improvement. The quan- in the East, intensive management may con- tities would not peak, however, before ceivably work to the detriment of species di- about 30 to 40 years as stands that had versity although it may increase the total wild- been cleared and replanted began to life population. Even in these regions, how- reach harvesting age, By this time, most of ever, there is a possibility that large numbers the old-growth stands accessible to log- of property owners may choose to leave their ging already may have been harvested, al- lands unmanaged because of personal prefer- ences. This would serve to protect diversity. “M. Clawson, “The National Forests, ” Science, VOI 20, Febru- ary 1976 The potential for added growth of high- ‘“M Clawson, “Forests in the Long Sweep of American His- quality timber from stand conversions of low- tory, ” Science, VOI 204, June 15, 1979 Ch. 2—Forestry ● 39

though significant benefits from reducing curring in an unmanaged forest give that forest logging pressures on other valuable or substantial resilience, because the diversity of fragile lands would still be available. vegetation and wildlife of the more mature ● Although the increased availability of states of the forest cycle as well as the diversi- high-quality timber might negate argu- ty created by the heterogeneous mix of stages ments that these valuable or fragiIe stands tend “to buffer the system against drastic must be cut to provide sufficient wood to change as by diluting the effects of pests on meet demand, there is no guarantee that single species.’’” Ecologists often have argued the wood made available from intensified that man pays a significant price in moving too management will be less expensive than far from this natural state: that obtainable from these stands, and The whole history of agriculture, and later, economic pressure to harvest them might forestry, is basically a continuous effort to continue. create simplifed ecosystems in which special- Although the long-range economic goals of ized crops are kept free of other species which intensive management provide an incentive interfere with the harvest through competi- against poor environmental practices, careless tion . . . diversified systems have built-in insur- ances against major failures, while the simpli- logging and regeneration practices will still oc- fied systems need constant care. ’z cur on a portion of the managed sites. Poor management may be practiced on a smaller In relation to human needs, the human strat- proportion of sites than would have been the egy can be viewed as a reversal of the successional sequence, creating and maintaining case without an expansion of wood for energy, early successional types of ecosystem where but the effects of such management may be gross production exceeds community respira- aggravated with such an expansion because: tion. Such . . . ecosystems, despite their high yield to mankind, carry with them the disad- ● more acreage wilI be logged each year, vantages of all immature ecosystems, in par- ● most affected sites will have fewer years ticular they lack the ability to perform essen- to recover before they are logged again, tially protective functions in terms of nutrient and cycling, soil conservation and population reg- ● the removal of maximum biomass and ulation. The functioning of the system is thus subsequent soil depletion may reduce the dependent upon continued human interven- sites’ ability to recover. tion. 73 Thus, the impacts associated with conven- There are, of course, counterarguments to tional logging— including erosion and soil deg- the thesis that this simplification of ecosys- radation, damage to water quality, esthetic tems places these systems under significant damage, and other impacts–are likely to oc- risk. One argument is that much of silviculture cur with even greater severity on a portion of duplicates natural events, and purposely so; those lands devoted to wood production for for example, clearcutting, sometimes followed energy. Unfortunately, because of the lack of by broadcast burning, is said to duplicate the data on logging practices and the very mixed effects of severe storms or catastrophic fires. 74 nature of the incentives for good management, Another is that professional silviculturalists it is impossible to make a good quantitative can compensate for any tendency towards a prediction of the size of this portion. “Smith, op cit A basic— and difficult to resolve— issue “A H Hoffman, “Comprehensive Planning and Management concerning the wisdom of moving to a very of the CountrVslde A Step Towards Perpetuation of an Ecologi- cal Balance, ” C/oba/ Perspectives on Eco/ogy, Thomas C Em- high level of intensive management of U.S. for- mel, ed (May field Publlshlng Co , 1977) estland is the possibility that the long-term ‘‘R Manners, “The Environmental Impacts of Modern Agricul- viability of these forests may be harmed. The tural Technologies, ” Perspectives orI Environment, I R Manners and M W Mlkesell, eds (Association of American Geographers, possibility of soil depletion is only one aspect 1974), publication No 13 of this. The cycles of natural succession oc- 74 Smith, op clt 40 . VOI. II-Energy From Biological Processes decline in resiliency. In its extreme, this argu- the resources of this assessment. Also, the high ment is particularly unacceptable to those level of emotional commitment that is at- who are skeptical of placing too great a faith tached to the alternative views of how far in science: nature can be safely manipulated makes it unlikely that such a gathering of evidence will We ought to believe that we can excel over nature; and if we do, we should not be re- change many minds. However, it is at /east stricted to blind imitation of her methods clear that a substantial increase in intensive . . . we have the chance to sift nature’s truths, management must be accompanied by a thor- and recombine them into a new order in which ough research program stressing examination not only survival, but enhanced productivity of such critical factors as nutrient cycling, the are the ruling criteria . . . (we) must look to role of soil organic material vis-a-vis resistance near-domestication of our forests . . . we must to tree disease, and other factors affecting sys- move forestry close to agriculture. 75 tem resiliency. The possibility that forest via- The strongest argument that can be made, bility might be at excessive risk if hundreds of however, is that past forestry experience has millions of acres in the United States were demonstrated that temperate forests can ab- placed in intensive management should not be sorb an unusual amount of stress without suf- automatically rejected, even though some de- fering long-term damage. For example, large gree of success in such management appar- acreages in Europe as well as the United States ently has been achieved elsewhere. that today are densely forested were intensely exploited as agricultural land in the past. In Harvesting for the Residential Market many instances, foresters can point to inten- The rapidly expanding demand for wood sive management practices in European for- fuel for residential use currently is satisfied ests that have continued to provide high pro- largely by harvesting of wood by homeowners ductivity of lumber for a hundred or more and by local entrepreneurs. The high price of years. In counterpoint to these arguments, wood for residential use is an incentive for some environmentalists are worried about the larger scale loggers to enter the market, and a future of Europe’s forests and point to increas- trend in this direction probably should be ex- ingly high external costs in terms of polluted pected in the future. The identity of the sup- water and increasing incidence of disease epi- plier may be an important component in deter- 76 demics. Also, insufficient data is available to mining the environmental effects of satisfying indicate whether or not small but significant a high residential demand for wood fuel. drops in long-term productivity may have occurred because of such past practices. An expansion of the residential wood market represents an opportunity for improved forest A similar argument rages about high-yield management because of the value it places on agriculture: yield levels in the Western coun- lower quality wood, which in turn should stim- tries have climbed steadily over the past cen- ulate an increase in thinning activities. The tury, with temporary setbacks that have thus potential benefits are the same as those de- far been dealt with by further adjusting the sys- scribed for the increase in intensive manage- tem, but environmentalists as well as many ment: an increase in productivity and timber agronomists are worried about increasing num- value on the affected lands. This opportunity bers of pesticide-resistant insects and rising en- exists on woodlands ranging from smalI private vironmental costs. woodlots to federally- and State-managed for- Pursuit of the evidence on both sides of this ests. The latter could use homeowners as a argument may be worthwhile, but it is beyond “free” work force to harvest selected trees, a practice that is already in operation in many ‘% Staebler, “The Forest and the Railroad, ” brochure pub areas. Ilshed by Weyerhaeuser Co , December 1975 “Goldsmith, “The Future of Tree Diseases, ” The Eco/ogist, No Unfortunately, a rising demand for wood 4[5, July-August 1979 will bring with it a potential for significant Ch. 2—Forestry ● 41 negative effects on woodlands. High prices for Tree Plantations wood fuels are likely to stimulate an increased The concept of an energy farm or plantation incidence of illegal cutting of wood. “Timber where trees are grown and harvested on short rustling” apparently is frequently encountered rotations like agricultural crops is a logical ex- in stands of very high-quality timber such as tension of current intensive single-aged man- redwood and walnut. More substantial cutting agement of forests. In fact, the growing of involving multiple acres at a time must be ex- Christmas trees on plantations is a more inten- pected as wood demand grows and prices in- sively managed activity than an energy farm is crease; remote areas, or areas where property likely to be, because the level of “manage- boundaries are not well marked should be par- ment” — including pesticide and fertilizer use ticularly vulnerable. (Illegal mining of coal —will tend to increase with the unit value of may be an analogous and somewhat prophetic the crop. In addition, a Christmas tree farmer example. Although it takes considerable time cannot tolerate relatively minor levels of pest and effort to expose and mine a coal seam, or drought damage because his crop value is coal poaching is not at all unusual in Appa- strongly dependent on appearance, and thus lachia, and some examples involving millions he must apply pesticides or irrigation water of dollars worth of coal have been reported re- during episodes that the energy “farmer” may cently. Poaching timber is going to be a lot be able to ignore. easier than poaching coal. ) In areas where wood stoves are oversold or where forest prod- The land requirements, growing needs and ucts companies occasionally enter the (lower harvesting techniques associated with energy quality) wood market, temporary fuelwood farms appear to be very similar to those of a shortages or price escalation may further stim- large agricultural enterprise growing perennial ulate illegal cutting, especially among poorer food crops. Because of this resemblance, the homeowners or those who cannot shift to an environmental impacts are not treated in this alternative fuel for space heating. section. The chapter on agricultural biomass production should provide sufficient informa- The same forces that stimulate illegal cut- tion about these impacts. ting, especially where coupled with ignorance of forest management, are likely to result in a variety of poor practices: improper harvesting Controlling Negative Impacts techniques leading to damage to adjacent A common theme running through reviews trees or to forest soils, incorrect tree selection, of silvicultural practices by the forestry es- overcutting, etc. tablishment—the wood products industry, The balance between beneficial and adverse schools of forestry, and the Forest Service— is effects of a rising demand for wood as a resi- that these practices may have negative envi- dential fuel is uncertain. Positive measures ronmental consequences but that the conse- such as an increased availability of trained quences are readily controlled, that significant foresters to provide assistance to small wood- environmental damages today are the excep- Jot owners, better dissemination of information rather than the rule, and that in those tion on woodlot management, and the organi- cases where damages occur they are almost al- zation of efficent and competitive retail sup- ways short lived, i.e., the forest quickly recov- pliers would help to limit adverse impacts. On ers and normal forest dynamics are restored. the other hand, the combination of a sharply The President’s Advisory Panel on Timber increased demand for wood coupled with a re- and the Environment reported that: 77 source base that is accessible and vulnerable “Fred A Seaton, et al , Report of the Presjdenr’s Advisory to illegal or poorly managed cutting appears to Pane/ on Timber and the Environment (Washington, D C Presi- be virtually a guaranteed source of trouble. dent’s Advisory Panel, April 1973)

67-968 0 - 80 - 4 42 . VOl.II-Energy From Biological Processes

A careful review . . . revealed that most of control techniques such as using water bars . . . (the environmental) damage caused by log- and artificially seeding erodible areas “are ging can be avoided or minimized. Many of (done) so infrequently that the role of these the fears that have been expressed are un- convenient erosion control devices in prevent- founded, misleading, or exaggerated, often ing postlogging degradation of water quality is due to extrapolation from an isolated case to minimal at present. ” 80 forest lands in general. Properly executed timber harvesting and Given the lack of knowledge of current for- other silvicultural procedures need not result est practices and the hints of environmental in important long-term losses of soil nutrients, problems provided by the limited data, Con- deterioration of the soil, nor cause other phys- gress should consider both the availability of ical environmental damage. Damage that has control measures and the institutional climate occurred resulted primarily from erosion asso- for putting these measures into practice before ciated with logging road construction and use, attempting to stimulate the increased use of skidding of logs downhilI or across streams, or wood for energy. harvesting on steep slopes where removal of vegetative cover caused slides. With updated methods, such difficulties will become rare ex- Control Capability ceptions. Such damage as has occurred will be The technical capability exists to control or corrected through natural processes as the reduce the negative effects of logging and, forest grows back. (Emphasis added.) more generally, of all silvicultural activities. The problem with statements such as these Table 13 presents a partial list of the control is that they do not acknowledge the current methods available to the forester. Some of the paucity of information on actual logging prac- more critical are: tices and effects. As noted in the introduction ● Site selection/identification and possibly to this section, there are few credible as- avoidance of problem areas. — Because sessments of forestry operations on a state- many of the environmental problems of wide or regional basis, The few that have been logging are strongly site-dependent, iden- attempted are limited in scope; for instance, a tification of problem areas followed by survey of practices in Maine in support of the revision or abandonment of logging plans 208 program (sec. 208, Public Law 92-500/ is a critical environmental control strat- Federal Clean Water Act) is limited to record- egy. Avoidance of steeply sloped sites ing the occurrence of gullying and the use or with unstable soils is important for mini- nonuse of simple erosion controls. 78 mizing erosion. This often coincides with The limited information that is available seems economic incentives, because the more to indicate that the generally optimistic tone of efficient heavy equipment cannot operate most reviews of forestry impacts should be on steep slopes. Geologic surveying of the viewed with caution. An interesting conclusion site can often detect vulnerable soil/rock/ of the Maine study was that “the area wide slope formations, although this capability magnitude of the (erosion and sedimentation) is not fully developed. Temporary avoid- problem is somewhere between the positions ance of some areas, for example, during espoused by the industry representatives on rainy conditions, can avoid major prob- the one hand, and groups and agencies con- lems of soil compaction and destruction cerned with maintaining environmental quali- of soil structure. Other site conditions ty on the other hand. 79 The survey found that that must be treated with special care or simple — and supposedly standard — erosion avoided include nutrient-deficient and thin soils, and sites in immediate proximi- ““A Survey of Erosion and Sedlmentatlon Problems Associ- ty to lakes and streams. In the latter case, ated With Logging In Maine, ” Land Use Regulation Commission, a buffer strip of smaller trees and shrubs State of Maine, for the Maine Department of Environmental Pro- tection, May 1979 ‘q I bid ‘[)lbld Ch. 2—Forestry ● 43

Table 13.–Control Methods

Mitiaative a Preventive Surface protection: System design and maintenance: ● Access: seeding, mulching, riprap, or mat on cut-and-fill slopes ● Access: minimize cuts and fills, roadway widths and slopes; control ● Timber harvest: maintenance of vegetative cover; distribution of slash road density ● Cultural treatments: seeding; planting; fertilization ● Timber harvest: minimize soil compaction from equipment operation; Flow diversion and energy: use site-compatible log removal system; control harvested volume ● Access: berms above cut slopes; benches on cut slopes; checkdams within a watershed; limit harvest on unstable slopes; shape openings in ditches; drop structure at culvert ends; water bars on road surface; for minimum esthetic impact, avoid cutting next to recreational activity flow diversion from potential mass failures or at mid-slope areas ● Timber harvest: buffer strips; water bars on skid trails ● Cultural treatments: minimize reentry disturbances; fire control ● Cultural practices: plowing, furrowing, bedding Timing: Access design modification ● Access: closure of temporary roads; limited access; closure during adverse conditions ● Timber harvest: limit operation during adverse climatic conditions; site preparations during favorable conditions ● Cultural treatments: intensity and number of thinnings acontrol~ can be described as preventive’ or ‘‘mltlgal[ve” according to the mode of applications Preverrwe COfIfrO/S apply to the prelmplementatlon Phase Of an oPeratlon These controls involve stopping or changing the actwlty before the sod-dlsturbmg actlwty has a chance fo occur Mmgalwe corwok include vegetatwe or chemical measures or physmal structures which alter the response of the soIl dlsturbmg achwty after it has occurred SOURCE S/lvlcullureAcl/vll/es and NonpovMPo//ul/onAb aternent A Cost-E/(ecOvenes sAna/ys/s Procedure (Washington, D C Forest Serwce, USDA, November 1977)

along the shoreline may be sufficient to only harvesting allowed on soils subject to provide shading and some sediment pro- mass movement. tection to the body of water. ● Erosion/sediment control measures. — Al- ● Selection of harvesting system. — Control though a certain amount of erosion from of erosion, esthetic, and other impacts soil compaction and mineral soil exposure can be achieved by matching the harvest- is inevitable in logging operations, it can ing system to the site conditions. For ex- be reduced by using lighter equipment to ample, the type of forest regenerated at avoid compaction, by using overhead or the site can be controlled by the harvest even aerial (balloon or helicopter) log col- system, because different degrees of dis- lection methods (although these methods turbance favor different tree species. are economically feasible only for very Clearcutting and residue removal favor high-quality timber), by properly design- species that need maximum disturbance ing roads and minimizing their overall to grow (e. g., Douglas fir, jack and lodge- length, by mulching the site, and by a pole pine, paper birch, red alder, and cot- variety of other methods. Furthermore, tonwood 81) and shelterwood cutting the erosion that cannot be controlled can (which leaves residual trees in sufficient be prevented from damaging water quali- numbers to shade new seedlings) favors ty by using buffer strips, sediment traps, species (such as true firs, spruces, and and other means. maples) that require light shade to thrive. The harvesting system may also be used to The Institutional Climate for avoid some of the negative effects to Environmental Control which the site is particularly vulnerable. Despite the generally resilient nature of the Clearcutting, for example, would be in- forests and considerable scientific knowledge dicated for old, decrepit stands in which of forest ecology and regeneration, forest en- residual trees would be likely to blow vironments may be threatened in the future be- down in the first severe storm following cause certain market forces or institutional harvest. Shelterwood cutting would be ap- constraints discourage adequate environmen- propriate for stands important to scenic tal protection. These problems include a lack views. A light selection cut may be the of expertise in the logging community, a vola- “Smith, op clt tile market that hinders adequate planning in i—

44 . VOI. //—Energy From Biological Processes

certain segments of the industry, and a lack of 3. Lack of incentive. —These are four reasons sufficient incentives to practice environmen- why a logger would pay strict attention to tally sound management. minimizing environmental damages: ● personal environmental or esthetic ideal- 1. Lack of expertise. –Although the majority of ism, negative impacts may occur because of ● economic incentive, failure to follow well-recognized guidelines, ● regulatory controls, or others occur because of failures of judg- ● public relations ment; forest environments are extremely complex and often require expert judgment Idealism–and the role of education in fos- about site conditions to select correct tering it—should not be ignored in predicting harvesting strategies. Some important im- impacts and attempting to mitigate them. The pacts can be avoided only if the logger can strengthening of existing programs to educate recognize subtle clues to the existence of potential wood harvesters about the adverse vulnerable conditions. For example, many environmental effects of careless harvesting unstable soil conditions may be recogniz- may be useful in tapping the vein of environ- able only to a soils expert. This type of ex- mental idealism in the United States. Idealism pertise usually is not available to the small is clearly insufficient to assure environmental operator, except possibly where local and protection, however, and more selfish incen- State governments offer preoperation in- tives are needed. spections and guidance (e. g., in Oregon]. I This poses a special problem if the residen- The long time period needed to recoup the tial market for wood expands considerably, benefits of protective measures and the tend- because small operators may be expected to ency of many of the benefits to accrue to adja- satisfy much of this new demand. cent landowners or the general public reduce 2. Insufficient time for proper planning. – In the economic incentive of environmental pro- current mill operations in Maine, many “mill tection. The shorter rotation periods that may managers commonly call on short notice for be used for obtaining wood for energy may en- a certain volume of a given type of product hance the economic incentive, especially for from the firms’ logging division . . . A com- owners of large tracts of land (because they are mon result is that a considerable amount of the “adjacent landowners”). Also, some “best the haul road construction is done on short management” measures do yield immediate notice . . . (without) . . . proper planning and returns to loggers, for example, measures that correctly installed and maintained drainage minimize road length or that prevent roadbeds structures. ”82 It is not clear that problems of from washing away. this nature will be as severe for wood har- Finally, to the extent that poor management vesting for energy, because demand for the of logging does long-term physical and esthetic wood as a feedstock may be more uniform damage to the forest, the value of forested and predictable than the demand for tradi- land as a recreational and esthetic asset offers tional forest products (it also is not certain a strong incentive to the landowner to insist on that the Maine experience is widely ap- sound practices. This incentive will be partic- plicable). Nonetheless, most operations will ularly strong in areas that have seen recent in- combine lumber and energy feedstock oper- creases in market value because of their envi- ations— removing the high-quality wood, ronmental value. This incentive will be effec- and then clearing to harvest the remainder tive, however, only where the landowner main- of the biomass for energy users. To the ex- tains close supervisory control over the logger. tent that the timing of these operations depends on the demand for the (higher value) Regulatory control of wood harvesting oper- lumber, this problem may remain. ations in the United States is very uneven. Al- though the Forest Service can exert considera- ‘*”A Survey of Erosion and Sedimentation Problems Associ- ated With Logging in Maine, ” op cit ble control of logging operations on Federal Ch. 2—Forestry ● 45 lands, logging on private lands is largely un- problem is discussed with insight in Brown controlled or very loosely controlled. 1976:84 The 1976 National Forest Management Act The difficulty is that rules specific to the includes requirements that federally owned wide variety of situations encountered would timber “be harvested only where soil or . . . often be difficult to write and cumbersome to water conditions will not be irreversibly dam- enforce for a great many problem areas, par- aged, that harvests be on a sustained yield ba- ticularly within the context of our present sis, that silvicultural prescriptions be written to state of technology. Field personnel recognize the dilemma of rules so vaguely written that ensure that stands of trees will generally not be they provide no control versus rules so spe- harvested until they are mature (although thin- cific that they prohibit flexibility and prevent ning and other stand improvement work is per- forest practice officers and operators from ad- mitted), that clearcutting meet certain stand- justing methods to meet complex or highly ards, and that land management plans be writ- varying situations. Given the option, most field ten with public participation. ”83 The Multiple people prefer to have flexibility at the risk of Use Act of 1960, by defining environmentally losing some control. oriented uses (such as wildlife protection) as Finally, many State forestry agencies have legislated uses of the national forests, requires concentrated their attention on forest fire management practices in these forests to con- prevention and control and not on forest man- sider environmental protection as a direct re- agement. Hence, the experience, interest, and quirement. In response to these mandates, the expertise of present State forestry personnel Forest Service enforces strict standards for may not provide a good base on which to build harvesting lumber on Federal lands. a strong management-oriented program. The degree of control exerted on non-Fed- The public’s increasing awareness of envi- eral forests — especially privately owned for- ronmental problems and willingness to act ests — is noticeably weaker. Water quality im- may serve as a strong incentive for the larger pacts from wood harvesting theoretically forest products companies to consider the should be regulated through the development public relations implications of their decisions. of nonpoint source control plans under section Companies like Weyerhaeuser spend large 208 of the Federal Water Pollution Control sums of money explaining their activities in Act. As discussed in volume 1, however, im- sophisticated advertisements; presumably, this plementation of section 208 generally has awareness of the importance of public approv- been disappointingly slow, and the eventual al affects their decision making and operations. effectiveness of the 208 plans is highly uncertain. Also, few States have comprehensive forest practices legislation or the manpower to Potential Environmental Effects— enforce such legislation. A major problem fac- Summary ing States wishing to control forest practices is the complexity and site-specific nature of the The use of wood as an energy feedstock environmental impacts, forcing the difficult holds considerable potential for reducing the choice of using either a substantial force of adverse impacts associated with fossil fuel highly trained foresters enforcing loosely writ- use. It also offers the potential for some impor- ten performance guidelines or else a more tant environmental benefits to forests, includ- (economically) manageable agency enforcing ing: rigid — and perhaps impractical — rules. This ● decreased logging pressures on some envi- ronmentalIy valuable forests;

‘“Brown, et al , Meet/rig Water Qua//ty Objecr/~es Through the n IE ~ “lronmenta/ Readjne55 Document, wood CommerC/a /;~a- Oregon Forest Pract/ces Act, (Oregon State Department of tlon, op clt “ Forestry, 1976) 46 ● Vol. II—Energy From Biological Processes

● improved management of forests that tally sound harvesting plans, especially by have been mismanaged in the past, with smalI operators. consequent improvements in productivi- If careful environmental management is not ty, esthetics, and other values; and practiced, the result might be: ● reduced incidence of forest fires. ● increased erosion of forest soils and con- There is considerable uncertainty, however, sequent degradation of water quality, about the extent to which a significant in- ● significant losses in esthetic and recrea- crease in the use of wood for energy will ac- tional values in forested areas, tually result in these benefits and avoid the ● possible long-term drop in forest produc- negative impacts that could also accompany tivity, such an increase. There are important econom- ● decline in forested area, and ic incentives for good management, including ● reduction of forest ecosystem diversity increased production of high-value timber and and loss of valued ecosystems and their avoidance of losses in land values. There are a wiIdlife. number of factors, on the other hand, that must be interpreted as warning signals: Because the quality of forest management 1. Environmental regulation of forestry oper- and the capacity for environmental regulation ations, especially on private lands, gen- currently span the entire range from very low eralIy is weak or nonexistent. (or nonexistent) to high, the expected result of 2. Some of the existing economic incentives a “business as usual” approach to wood-for- may induce cutting of vulnerable lands or energy environmental management would un- neglect of best management practices. doubtedly be a complex mix of the above im- 3. Important gaps in the knowledge of the pacts and benefits —with the marketplace de- effects of intensive silvicultural activity— termining the balance between positive and for example, of the nutrient and organic negative effects. Government action — includ- matter changes in the soil caused by ing improved programs for local management whole-tree logging — may deter environ- assistance, increased research on the effects of mentally sound choices from being made. intensive management, and increased incen- 4. The complexity and site-specificity of the tives (economic or regulatory) for good man- harvesting choices that “must be made agement — may be capable of shifting this bal- may complicate adoption of environ men- ance more towards the positive.

R&D Needs

The primary R&D needs in the area of wood there is a need to develop low-cost techniques supplies from forests fall into the categories of and equipment for harvesting smaller pieces of harvesting technology, growth potential, en- wood and brush on more varied terrains and at vironmental impacts, and surveys. Traditional greater distances from roads. harvesting technologies are geared toward removing large pieces of wood in a way that is Most research into forest growth potential is appropriate for lumber or paper pulp produc- aimed at producing large straight trees suit- tion. The wood that can be harvested for fuel, able for the traditional forest products in- however, is considerably more varied, involv- dustry. Although some of this is research ap- ing brush, rough and rotten timber, and the plicable to the production of wood for energy, smaller pieces associated with logging resi- the conditions and techniques for enhancing dues. Although the whole-tree chip method commercial timber growth are not the same as seems to work well on relatively flat land, those for enhancing total biomass growth. As Ch. 2—Forestry . 47 an example, thinning of tree stands reduces the provide a repertoire of techniques that can be total leaf surface and, with it, the amount of used where soil erosion may be a problem, sunlight that is being captured by plants. This such as in steeper slope terrains. Harvesting reduces the total biomass growth on the stand, techniques that decrease the degree of soil although it tends to increase the growth of compaction should also be developed. Further- commercial timber. If a strong wood energy more, the entire forest ecosystem needs to be market develops, the ideal forest composition better understood if the environmental im- could involve a mix of tree types, sizes, and pacts of various types of forest activities are to qualities. Various strategies for achieving in- be appropriately managed. tegrated and economical energy-commercial timber operations need to be investigated. The national forest survey is primarily in- Tree hybrids, for example, should be devel- tended as a survey of commercial timber. The oped with both commercial timber production assumptions as to what is commercial should and biomass production as dual goals. be separated from the survey of the biomass inventory and growth potential, in order to There are a number of uncertainties regard- have an accurate assessment of the quantities ing the environmental impacts of increased available for all uses. The survey should in- logging for energy. The nutrient balance in corporate noncommercial forest lands which forests, as noted, needs to be better under- are classified that way because of low growth stood in order to better define the types and potential. quantities of wood that can be removed without depleting the soil’s nutrients. The effects A thorough assessment of the energy poten- of high biomass removal on soil carbon con- tial of the forests should also include a qualita- tent and any subsequent long-term impacts on tive assessment of the conditions of the stock- productivity or on forest viability require con- ing on forestlands and the silvicultural activ- siderable research. The relationship between ities (e. g., stand improvements) that could be the diversity of tree and understory species in a carried out to increase the yield. The survey forest and the forest’s resilience to environ- data should include environmental conditions mental stresses must be better understood be- such as soil types, rainfall, and other parame- fore highly intensive management is allowed ters. Finally, the size of tract is an important to expand to a majority of the commercial for- factor affecting the availability of the wood. est acreage. Alternative harvesting techniques Consequently, the farm and miscellaneous for- such as strip cutting (or the cutting of strips of est landowner classifications in forest surveys trees through the forest rather than clearcut- should be subdivided according to tract size ting a large area) should be pursued in order to and ownership. Chapter 3 AGRICULTURE Chapter 3.—AGRICULTURE

Page Introduction . . * . . * * . . . * * * ...... 51 23. Energy lnputs for Various Crops ...... 62 Plant Growth, Crop Yields, and Crop 24. Energy inputs and Outputs for Corn in U.S. Production ...... 51 Corn Belt ...... 62 Theoretical Maximum Yield...... 51 25. Estimated Costs of Producing Grass Historical Yield Trends. ., ., ...... 53 Herbage at Three Yield Levels ...... 63 Record Yields ...... 54 26. Total Crop Residues in the United States Future Yields ...... 54 for 10 Major Crops...... 65 land Availability ...... 54 27. Total Usable Crop Residue by Crop ...... 68 Types of Crops ...... 60 28. State Average Estimated Usable Group Current Agricultural Practices and Energy and Residue Quantities and Costs...... 69 Economic Costs...... 61 29. Environmental impacts of Agriculture, . . . . 70 Energy Potential From Conventional Crops ...... 64 30. 1977 Cropland and Potential Cropland Energy Potential From Crop Residues ...... 67 Erosion Potential ...... 77 Environmental impacts of Agricultural Biomass 31. Mean National Erosion Rates by Production .0.,...... ● ...... *** 70 Capability Class and Subclass ...... 77 Introduction ...... 70 32. Agricultural Production Practices That The Environmental lmpacts of American Reduce Environmental Impacts ...... 79 Agriculture. .,...... 70 33. Enviromental Pollution Effects of 1973-74: A Case Study in Increased Agricultural Conservation Practices ...... 81 Cropland Use ...... 75 34. Ecological Effects of Agricultural Potential impacts of Production of Biomass Conservation Practices...... 83 for Energy Feedstocks ...... 76 35. Environmental impacts of Plant Residue Environmental lmpacts of Harvesting Agricultural Removal ...... 86 Residues ...... 86 R&D Needs ...... 88

FIGURES TABLES Page Page 14. Photosynthetic Efficiency Summary ...... 52 7. U.S. Average Corn Yield ...... 53 15. Factors Limiting Plant Growth ...... 52 8. U.S. Average Crop Output ...... 54 16. Cropland Use in 1977 ...... 56 9. Farm Production Regions of the 17. Potential Cropland...... 58 United States ...... 55 18. Cropland Balance Sheet ...... 58 10. Cropland as a Percentage of Total Land 19. Cropland Available for Biomass Production 59 Area by Farm Production Region ..,...... 56 20. Annual Production and Disposition of 11. Present Use of Land With High and Corn for Grain in the United States, 1966-75. 60 Medium Potential for Conversion to 21. Annual Production and Disposition of Cropland by Farm Production Region . . . . . 57 Wheat in the United States, 1966-75...... 60 12, Net Displacement of Premium Fuel per 22. Estimated per Acre Production Costs in Acre of New Cropland Brought Into lndiana, 1979 ...... 62 Production ...... 66 Chapter 3 AGRICULTURE

Introduction

Agriculture was originally developed to pro- accounted for over 13 percent of total farm vide a reliable source of food. Later feed was production. included in farm production and animals provided large parts of the population’s energy needs. Although animals are rarely used for en- Many of the food and feed crops as well as ergy on U.S. farms today, agriculture has ex- farming byproducts can also be used to pro- panded to include the production of nonfood duce fuels or be combusted directly. In this commodities, including cotton, tobacco, paint chapter, the technical aspects of conventional solvents, specialty chemicals, and various in- agriculture are considered, leading to esti- dustrial oils. In 1977, these nonfood products mates of its potential for supplying energy.

Plant Growth, Crop Yields, and Crop Production

Harvested yields of many crops have in- Sixteen nutrients are essential for plant creased dramatically over the past 30 years as growth and two or three more may increase a result of the development of genetically im- yields but are not essential for the plant to proved crop strains, as well as increased use of complete its growth cycle. Of the 16, nitrogen, fertilizers and irrigation. Also, increased ap- phosphorus, and potassium are the 3 main nu- plication of chemicals for control of insects, trients needed in large quantities to supple- diseases, and weeds; further mechanization so ment the soil supply in order to obtain high that operations can be timely; improved tillage crop yields. Calcium and magnesium are ap- and harvesting operations; and other forms of plied where needed as finely ground lime- improved management have also helped to stone. Sulfur is added as elemental sulfur or as raise yields. sulfates when needed. Carbon, hydrogen, and oxygen come from the air and water and the re- Photosynthesis is the basic process provid- maining seven are used in extremely small ing energy for plant growth. Solar energy is ab- amounts and are absorbed from the soil. All of sorbed by the green chlorophyll in the leaf and these nutrients play essential roles in the used to combine carbon dioxide (CO ) from the 2 growth processes within the plant. air with water from the soil into stored chemical energy in the form of glucose. Glucose is used in the formation of compounds like aden- osine triphosphate which provides energy for Theoretical Maximum Yield the synthesis of the various materials needed The theoretical maximum photosynthetic ef- in the plant such as cellulose and Iignins for ficiency can be estimated as follows: cell walls and the structural parts of the plant and various amino acids (protein components). Ten percent of the light striking a leaf is Glucose is respired to provide energy for pro- reflected. Only 43 percent of the light that duction of other compounds, plant growth, penetrates the leaf is of a proper energy to and absorption of nutrients from the soil. As stimulate the chlorophyll. The basic chemical the plant matures, carbohydrates are stored in reactions (10 photon process) which use stimu- the seed to provide energy for the growth of lated (“excited”) chlorophyll to convert CO2 new plants. and water to glucose have an overall efficien-

51 52 ● Vol. I/—Energy From Biological Processes cy of 22.6 percent. The net result of these fac- Table 15.–Factors Limiting Plant Growth tors is, in theory, a maximum photosynthetic ● Water availability, efficiency of about 9 percent. A summary of ● Light saturation–a tendency for the photosynthetic efficiency to drop the various cases of photosynthetic efficiency as the incident light intensity increases above values as low as about is presented in table 14. 10% of peak solar radiation intensity. ● Ambient temperature, especially wide fluctuations from ideal. ● Mismatch between plant growth cycle and annual weather cycle. ● Length of photoperiod (hours of significant illumination per day). Table 14.–Photosynthotic Efficiency Summary ● Plant respiration, ● Leaf area index–completeness of coverage of illuminated area by Average PSEa during leaves or other photosensitive surfaces. growth cycle ● Availability of primary nutrients–especially nitrogen, phosphorus, and (percent) potassium, ● Maximum theoretical ...... 8.7 Availability of trace chemicals necessary for growth. ● Highest laboratory short-term PSEa ...... -9 Physical characteristics of growth medium. ● Laboratory single leaves, high CO or Acidity of growth medium. 2 ● low O , C-3 plants, c 7% full sunlight. . . . . 6.3 Aging of photosynthetically active parts of plants. 2 ● Same as above, for C-4 plants...... 4.4 Wind speed. ● Corn canopy, single day, no respiration . . . . . 5,0 Exposure to heavy rain or hail or icing conditions. ● Record U.S. corn (345 bu of grain/acre, Plant diseases and plant pests. ● 120-day crop) ...... 3.0 Changes in light, absorption by leaves due to accumulations of water Record sugar cane (Texas) ...... 3.0 film, dirt, or other absorbers or reflectors on surfaces of leaves or any Record Napier grass (El Salvador) ...... 2.5 glazing cover. ● Record U.S. State average for corn (128 Nonuniformity of maturity of plants in crop. ● bu/acre, Illinois, 1979)...... 1.1 Toxic chemicals in growth medium, air, or water, such as pollutants Record U.S. average for corn (108 released by human activity. ● Availability of CO . bu/acre, 1979 ...... 0.9 2 ● Adjustment to rapid fluctuations in insolation or other environmental apSE.pholo5ynthetlc efficiency variables —i. e., ‘‘inertia’ of plant response to changing conditions. SOURCE OftIce of Technology Assessment SOURCE. Off Ice of Technology Assessment

An efficiency approaching the theoretical In an untended system, the environmental maximum appears to have been achieved for a factors are left to chance. Consequently, in short time under laboratory conditions using any given year, some areas of the country will an alga or single-celled water plant. These experience a favorable combination of factors, results are controversial, however, and in prac- resulting in more plant growth, while in other tice there are always several other factors that areas environmental factors will be unfavor- limit the efficiency of photosynthesis, and the able, resulting in less plant growth. The exact transformation of glucose into plant material. places where the growth is favorable or unfa- The most important of these factors, many of vorable will also change from year to year, as which are interdependent, are listed in table will the exact growth at the most favorable 15. For example, light saturation can be influ- area in each year. In the absence of long-range enced by the CO2 concentration, which is affected by other things. The key factors are environmental changes, however, such as weather changes or soil deterioration, the aver- light saturation, soil productivity (its ability to age growth over a very large area and for many hold water, supply oxygen, release nutrients, years will remain relatively constant. and allow easy root development), weather (amount and timing of rainfall, absence of se- Some of the environmental factors can be vere storms, length of growing season, temper- controlled, while others cannot within the ature and insolation during the growing sea- present state of knowledge. Managing a plant son), and plant type (leaf canopy structure, and soil system consists of artificially main- longevity of the photosynthetic system, sensi- taining some of the environmental factors, tivity to various environmental stresses, etc.). such as nutrients or water, at a more favorable level than would occur naturally. The many re-

‘V C Coedheer and J W Klelnen Hammans, Nature, VOI 256, maining factors, however, are still left to p 333, 1975 chance. Furthermore, there is a limit to how Ch. 3—Agriculture . 53 much plant growth (or other characteristics concluded that 60 percent of the increase on such as grain yield) can be influenced by these plots was due to genetic improvements changes in given environmental factors. Once while 40 percent was attributed to improved some of the factors have been optimized for management. The management tends to re- plant growth (or, e.g., grain yield) the plant’s duce the environmental stresses, while hybrids performance will not be improved by further were developed that are less sensitive to ad- changes of these factors. Too much water or verse environmental factors and more respon- fertilizer, for example, could actually inhibit sive to the factors that can be controlled (e. g., growth rather than increase it. fertilizers, weed control, insect control, etc.). Because plants vary in their sensitivity to growth-limiting factors yields can be improved Historical Yield Trends by selecting or breeding for plants that are relatively insensitive to environmental factors Past yield trends can be used as a guide for that cannot be controlled and/or that respond projection of future yields. A period of at least well to factors that can. 15 years should be considered because of A dramatic example of the success of breed- weather fluctuations since the desired value is ing and management is corn. The history of the yield trend if weather remained constant. A U.S. average corn yields from 1948-78 is shown period of dry years from 1973 to 1976 tended in figure 7. While the national average yields to exert some influence on data variability. have not been analyzed in detail, Duvick has During the 1948-78 period, corn yields in- analyzed the changes in yields from hybrids creased an average of 2 bu/acre-yr giving a grown in various Midwestern locations.2 He 1978 yield just over 100 bu/acre. Similarly soybean yields showed an increase of about Figure 7.—U.S. Average Corn Yield 0.4 bu/acre-yr and a 1978 yield of 29.2 bu/acre. National wheat yields are somewhat more variable since wheat is grown primarily in areas that are more affected by drought than is corn. Nonetheless, the yield trend indicates an increase of 0.5 bu/acre-yr and a 1978 average yield of 31.6 bu/acre. The U.S. Department . . A of Agriculture (USDA) has calculated a summary of all crop yields per acre using a relative value of 100 for the 1967 yield. The trend for increase over this period was 1.4 units per year, but the uncertainty in this number is large (see figure 8).

Yield increases in the future as in the past will come from a combination of improved crop varieties and improved cultural practices. Since current fertilization practices have reached near optimum rates, increases in yield due to increases in fertilization rates will be less than for the past 20 years. As yield potentials of varieties are increased, however, in- 1940 1950 1960 1970 1980 1990 2000 creased rates of fertilization will be needed to Year keep pace with the increased yields. Since SOURCE. ~gr~cultura~ S~at~stKs (Washington, D C: U S Department of Agri- yield increases result from a combination of culture. 1978) practices, it is difficult to attribute yield in- ‘D N Duvick, Mayd~ca XXII, p 187,1977 creases to any one practice. 54 . Vol. II—Energy From Biological processes

Figure 8.—U.S. Average Crop Output average yield in Illinois in 1979 was 128 bu/acre and in Iowa in 1979 it was 127 bu/ 120 acre. * If county averages within a State are examined, average yields are found to approach 140 bu/acre. If individual farms of 500 acres of corn are considered, yields of 175 bu/acre have 110 occurred. And on selected areas of 2 or 3 acres

II yields of 345 bu/acre have been noted.

100 Future Yields Over the past 30 years corn yields have increased at an average rate of about 2 bu/acre- yr. One would not expect this rate of increase 90 to rise, and it may decline. Therefore, in pro- jecting corn yields in the year 2000, 140 bu/ acre would be optimistic. A less optimistic projection, based on annual average yields in- o 1965 1970 1975 1980 creasing at one-half the rate that they have in the recent past, is 120 bu/acre in 2000. A study Year by the National Defense University in 1978 SOURCE: Agricultural Statistics (Washington, D. C.: U.S. Department of Agri- culture, 1978). gave a projected corn yield for 2000 of 132 bu/acre. Future increases in the yields of other crops are also expected, but each crop together with Record Yields the cropland on which it will be grown must be considered separately. Dramatic increases, Projection of yields and determination of such as a doubling in crop yields by 2000, how- where yield increases will diminish may be ever, are not expected for conventional crops. judged on the basis of yields that have been obtained. For example, the average U.S. corn yield has reached over 100 bu/acre, but the *The weather in 1979 was ideal for corn growing

Land Availability

Cropland is land used for the production of The second is a broader classification that in- adapted crops for harvest, alone or in a rota- cludes some land not currently cropped that is tion with grasses and legumes, and includes rotated into cropland but may now be in pas- row crops (e. g., corn), small grain crops (e. g., ture or other use. The percentage of the total wheat), hay crops, nursery crops, orchard land area of each State that is cropland is crops, and other similar specialty crops. Crop- shown in figure 10 and the cropland used in Iand is generally categorized into the agricul- 1977 is shown in table 16. Both of these are tural production regions shown in figure 9. based on the more restrictive SCS classification of cropland. Of the U.S. total land area of 2.3 billion acres, 413 million or 467 million acres are currently classified as cropland depending on Cropland, however, is not a static category. whether one uses the Soil Conservation Service The location of cropland may shift even (SCS) or the other USDA classification system. though the quantity of cropland remains rela- Ch. 3—Agriculture . 55

Figure 9.— Farm Production Regions of the United States

SOURCE SO iI Conservation Service, U S Department of Agriculture tively constant. Over time there are both addi- tistic is enumerated. For example, up to 1964 tions and deletions to the cropland inventory. the agricultural census was personally enumerated and in 1969 it was done by mail. Ac- The quality of land and cropping systems cording to the broader USDA classification, may shift as well. One such change has been cropland pasture increased by over 30 million the increase in irrigation, in areas like the acres between these surveys, and the suspicion Texas high plains, from 1.2 million acres in is that the farmer applied a less strict defini- 1948 to 6.4 million acres in 1976. This repre- tion to cropland pasture which resulted in the sents both a trend in improving the productivi- inclusion of 30 million acres of pastureland ty of existing cropland and a trend towards and grassland into the cropland pasture cate- opening new, marginal land that is only pro- gory even though the actual usage had not ductive and economic with irrigation. To some changed. extent, the United States has been replacing rainfed arable land that is lost to agriculture One strong influence on the land inventory with irrigated land in dry areas. This trend, has been the Government’s agricultural pro- however, is likely to change due to increasing grams. The land retirement programs of the energy costs and depletion of some Western 1960’s reduced planted cropland and had the ground water. net effect of moving less productive land out Over time, the content of a land inventory of crop production temporarily or even perma- can be influenced by the way that a given sta- nently in the case of very low-quality land. As 56 . VOI. II—Energy From Biological Processes

Figure 10.—Cropland as a Percentage of Total Land Area by Farm Production Region

Table 16.-Cropland Use in 1977 (thousand acres)

Occasionally Row crops Close-grown crops Rotation hay improved hayland Region Irrigated Nonirrigated Irrigated Nonirrigated Irrigated Nonirrigated Irrigated Non irrigated Northeast ., ...... 254 6,771 1,033 9 3,339 5 3,286 Appalachian...... 349 14,445 33 543 7 1,248 10 3,151 Southeast ...... 1,681 12,108 23 342 24 74 8 604 Delta States...... 2,294 15,358 1,489 285 170 299 0 397 Corn Belt...... 1,035 70,291 31 7,228 8 5,987 4 4,116 Lake States...... 799 20,930 80 9,140 36 7,753 2,802 Northern Plains ...... 8,641 18,062 920 40,007 690 5,387 325 3,885 Southern Plains ...... 5,935 13,908 2,354 15,784 33 802 286 725 Mountain...... 4,117 962 3,316 14,021 2,269 380 3,821 1,276 Pacific...... , ., . . . . . 4,477 132 2,757 5,556 1,124 502 1,092 295 Total a ...... 29,750 173,493 11,025 93,865 4,398 25,818 5,559 20,839

Native hay Orchards, etc. All cropland Region Irrigated Nonirrigated Summer fallow Irrigated Nonirrigated Other Irrigated Nonirrigated Northeast ...... 0 789 24 84 508 493 372 16,534 Appalachian, ...... 0 86 112 0 149 690 406 20,339 Southeast ...... 0 0 54 699 661 1,324 2,449 15,053 Delta States...... , 0 84 179 6 136 489 3,979 17,207 Corn Belt...... 0 117 749 22 72 876 1,115 88,739 Lake States...... 0 719 466 42 291 1,073 972 43,167 Northern Plains ...... 33 2,493 13,825 0 15 268 10,790 83,733 Southern Plains ...... 0 405 1,061 33 153 755 9,011 33,223 Mountain...... 1,460 320 9,449 0 641 17,208 26,111 Pacific...... 261 208 4,082 2,242 183 277 12,261 10,927 Total a...... 1,754 5,221 28,319 3,295 2,225 6,806 57,647 355,520 aAlsO Includm Canbbe.an and Hawall SOURCE SoIl Conservation Serwce, U S Oeparlment of Agriculture Ch. 3—Agriculture ● 57

programs were terminated in the early 1970’s, high or medium potential for conversion to some of these acres came back into crop pro- cropland. Of the total potential cropland in duction. 1977, 40 million acres have a high probability to be converted and another 95 million acres USDA’S SCS surveyed non-Federal lands in are classified as having medium probability. * 1977 and identified the land that potentially could become cropland. 3 The survey classified The breakdown of the potential cropland potential croplands according to whether the into high and medium potential for conversion land was judged to have a high, medium, low, is an attempt to define a crude cost curve for or zero potential for conversion. Figure 11 the availability of new cropland. It was judged summarizes the quantities of land that have a by SCS that the land with a high potential will enter agriculture as a matter of course, if price relationships are somewhat more favorable to Figure 11 .—Present Use of Land With High and Medium Potential for Conversion to Cropland the farmer than the 1976 prices on which the by Farm Production Region survey was based. The medium potential, however, is a category involving lands with a wide variety of problems but which cannot be cate- Northeast gorically excluded from conversion if farmland prices increase sufficiently. Lake States The SCS survey, however, does not include a quantitative measure of the price increases Corn Belt necessary to bring potential cropland under cultivation. A conservative approach would be Northern to assume that only land with high potential Plains can be included in the cropland base without

Appalachian excessive inflation in food prices above that which would occur normally due to increased demand for food. A more optimistic approach Southeast would be to include those types of medium-potential land that probably will be considered Delta States high potential in the future, as increased demand for food raises cropland prices. This is Southern the approach that was taken. Plains Two major factors, mentioned above, that Mountain affect crop productivity are water availability and soil quality. Therefore, land was included

Pacific from the medium-potential category that has greater than 28 inches of annual rainfall and

Q 5 10 15 20 potentially has good productivity (capability (millions of acres) classes 1 and 2 of the eight agricultural land capability classes). These land types are the Forest Pasture and native lands most Iikely to be brought into cultivation if de- Rangeland mand exceeds the high-potential category. Other With these assumptions, the potential crop- SOURCE: 1977 National ErosIon Inventory Preliminary Estimates, Soil Conser. vation Service, U.S. Department of Agriculture, Apr!l 1979. I and is shown in table 17. This together with existing cropland provides the cropland base, to-

319775011 Conservation Service National Erosion Inventory Es- timate (Washington, D C SoIl Conservation Service, U S De- *As of November 1979, these numbers were 36 million and 91 partment of Agriculture, December 1978) million acres, respectively

67-968 o - 80 - 5 58 ● Vol. n-Energy From Biological Processes

Table 17.-Potential Cropland (thousand acres)

Source Region Forest Pasture Rangeland Other High potential with over 28 inches rainfall Lake States. , ...... 868 1,206 22 443 Delta ...... 1,482 1,781 0 129 Corn Belt ...... 664 4,451 23 368 Northeast ..,...... 268 562 0 371 Southeast ...... 2,111 2,926 133 151 Appalachian ...... 1,981 2,974 0 183 7,374 13,900 178 1,645 Total of forest pasture, and rangeland: 21,452 Medium potential, class l&2 land only with over 28 inches rainfall Lake States...... , ...... 1,463 723 0 315 Delta ...... 986 1,423 0 Corn Belt ...... 974 2,590 0 356 Northeast ...... 656 433 0 306 Southeast ...... 2,082 1,256 0 17 Appalachian ...... 2,012 1,157 0 119 8,137 7,582 0 1,113 Total of forest, pasture, and rangeland: 15,755 High potential with less than 28 inches rainfall Arid regions ...... 312 5,503 9,549 — Total of forest, pasture, and rangeland: 15,364

SOURCE: Otto C. Doermg lll, “Cropland Availability for Biomass Production,” contractor report to OTA, August 1979. taling 520 million acres. (This is based on the By examining the detailed demand forvari- broader cropland classification as used in ous crops from NIRAP and the land available, USDA’s Agricultural Statistics and all Econom- Doering has derived baseline estimates for the ics, Statistics, and Cooperatives Service pub- quantity of cropland that could be available libations) Although it is impossible to predict for bioenergy production, which are shown in exactly how the cropland base will develop in table 19.5 Doering, also derived high and low the future, one plausible scenario is shown in food demand scenarios for 1984 based on ex- table 18 based on continuation of past trends trapolation of trends in the recent past and in- to 1984 and on USDA’s National Interregional creased this demand range proportionately to Agricultural Projections System (NIRAP) for the increase in baseline crop demand from the 1990 and 2000.4 NIRAP projections for 1990 and 2000. Finally

‘L Quance, A Smith, K. Liu, and L. Yao-Ch~ “AdjustmentPo- these demand ranges were combined with tentials m US Agriculture,” Vol. I–hiational lnterregional Pro- jections System (Washington, DC,: Economics, Statistics, and Cooperatives Service, US Department of Agriculture, May 50ttoC Doering Ill, “CroplandA vailability for Biomass Pro- 1979) diction;’ contractor reportto OTA, August1979,

Table 18.-Cropland Balance Sheet (million acres)

Year 1977a 1979 1984 1990 2000 Cropland (except cropland pasture and hayland) .... 393 395 404 407 439 Cropland pasture and hayland ...... 74 72 65 80 60 Noncropland pasture ...... 27 27 22 10 2 Other potential cropland ...... 26 24 22 13 4 Total ...... 520 518 514 510 505 Total land lost to other uses to date ...... 0 2 6 10 15 Total ...... , ...... , . . . . . 520 520 520 520 520

With scs acreage Courltmg system, the 1977 acreages (in miltions of acres) would be as follows, cropland except cropland pasture and hayland, 343; cropland pastaure and hayland, 63, noncropland pasture which is potential cropland or IS pemdically rotated into cropland, 88; and other potential cropland, 26 The malor differences are that the cropland categories total 406 million acres with the SCS Classiflcatlon rather than 467 million acres and the noncropland categories are increased accordingly. In both classification schemes there are additional noncropland pature categories which are neither ptential cropland nor periodically rotated into cropland, primarily because the terrain is too rocky or rough to allow mechamzed harvests SOURCE. Oeduced from Otto C Ooermg Ill, “Cropland Availability for Biomass Production, “ contractor report to OTA, August 1979 Ch. 3—Agriculture ● 59

NIRAP high and low productivity (yield/acre) tion. Second an investment is sometimes nec- estimates to determine plausible ranges of de- essary to convert the land to crop production, mand for cropland for food and feed produc- such as installation of drainage tiles or remov- tion and thus the ranges of land available for ing trees occupying the site. These costs can bioenergy production. These estimates are vary from virtually nothing to as much as shown in table 19. $600/acre.’ Third, the land that can be brought into production is generally less productive, on Table 19.–Cropland Available for Biomass Production the average, than cropland currently in production. Finally, this land also typically suffers Million acres from problems of drought or flooding that 1984 make crop yields extremely sensitive to weath- From cropland pasture ...... 10 From high potential ...... 10 er (particularly the rainfall pattern). Conse- From medium potential ...... 10 quently, farming this land involves a larger 30 cost and risk than with average cropland; and, Range of uncertainty, ...... 30-70 from the national perspective, using it will in- 1990 From cropland pasture ...... 25 crease the year-to-year fluctuations in food From high potential ...... 5 supplies and prices. From medium potential ...... 5 35 As a result of these added costs and risks, Range of uncertainty ...... 9-69 farm commodity prices will have to rise before 2000 it will be profitable to bring new land into crop From cropland pasture ...... 10 From high potential ...... NA production. Eventually this raises the cost of From medium potential ...... NA all farmland, the cost of farming, and food 10 prices. The exact price rise needed to increase Range of uncertainty...... 0-65 the cropland in production by a given amount NA=none available is unknown, but some things can be deduced SOURCE” Otfo C Ooerlng Ill, “Cropland AvallabMy for Biomass Production, ” contractor report to from this analysis. During the next few years, OTA, August 1979 bioenergy production from cropland is not It should be emphasized that these are not likely to be constrained by the availability of predictions, but rather plausible estimates cropland. However, the quantity of land that given the current state of knowledge. The can be devoted to energy production without ranges are less than &10 percent of the crop- reducing food production is likely to decrease Iand base, so it is unlikely that more accurate in the future. Furthermore, since the marginal estimates can be made. Furthermore, unex- cost of bringing new cropland into production pected increases in crop productivity, in world increases as the quantity of cropland in pro- food demand, or in demand for cropland for duction expands, the added cost in terms of nonagricultural uses could increase or de- higher food prices needed to keep a given crease the quantity of cropland available for amount of cropland in energy production is bioenergy production beyond the ranges likely to increase with time. In other words, it is shown. Also, since this only refers to cropland likely to be increasingly expensive to produce capable of producing more or less convention- energy crops, even if the energy output re- al crops, the development of unconventional mains constant. crops could open new land categories not con- The above comments are particularly appli- sidered here. cable to grains and sugar crops. Considerable In addition to the physical availability of quantities of land, however, already are de- cropland, one must consider the cost of bring- voted to forage grass production and the yields ing it into production. Four major factors in- can be raised through increased fertilization fluence this cost. First the land is currently be- (see below). Furthermore, grass yields tend to ing used for some purpose that the owner con- be less sensitive to poor soil quality than grains siders to be more valuable. than crop produc- Cl bid 60 ● Vol. II—Energy From Biological Processes and sugar crops. Consequently, the economic Nevertheless, in the long term, there may be barriers to increased grass production are con- Iittle cropland suitable for food/feed produc- siderably lower than for increased grain or tion that can be devoted to energy, and any sugar production, and one would expect the in- energy crops would have to be grown on land direct costs of grass production to be less than totally unsuited to food and feed production. those of grains and sugar crops.

Types of Crops

There are over 300,000 plant species in the Table 20.–Annual Production and Disposition of Corn for Grain world, but less than 100 are grown as crops in in the United States, 1966-75 (million bushels) the United States. Of the various crops, three Domestic basic types are currently of interest for imme- Year Production consumption ExPorts Stocks diate energy production: starch, sugar, and for- 1966 . . . . 4,167 3,697 487 826 age. 1967 . . . . 4,860 3,885 633 1,169 1968 . . . . 4,450 3,966 536 1,118 The major starch crops are corn (for grain) 1969 . . . . 4,687 4,189 612 1,005 1970 . . . . 4,152 3,977 517 667 and wheat, accounting for 21 percent each of 1971 . . . . 5,641 4,387 796 1,126 the total acreage of harvested crops, or about 1972 . . . . 5,573 4,733 1,258 706 70 million acres each. The annual production 1973 . . . . 5,647 4,631 1,243 483 1974 . . . . 4,664 3,641 1,149 359 and disposition of corn and wheat are shown in 1975 . . . . 5,797 4,049 1,711 398 tables 20 and 21. In addition, oats, barley, grain sorghum, and rice are other important SOURCE Agr~cullural Stafrstlcs 1977( WashmgIon, D C U S Department of Agriculture, 1977) starch crops. Table 21.-Annual Production and Disposition of Wheat The main sugar crops currently grown in the in the United States, 1966-75 (million bushels) United States are sugarcane and sugar beets. About 760,000 acres are devoted to sugarcane Domestic Year Production consumption Exports Stocks (in Florida, Louisiana, Texas, and Hawaii) and 1966 . . . . 1,967 683 771 513 sugar beets were grown on 1.2 million acres in 1967 . . . . 2,202 626 765 630 1977. Also, a smaller acreage is devoted to 1968 . . . . 2,188 740 544 904 sweet sorghum production, primarily for sor- 1969 . . . . 2,350 764 603 983 1970 . . . . 2,336 772 741 823 ghum syrup. The sugar yields averaged about 1971 .., . 2,442 848 610 983 3.7 ton/acre for sugarcane (some growing sea- 1972, . . . 2,530 798 1,135 597 sons were 18 to 24 months) and 2.6 ton/acre for 1973, . . . 2,305 748 1,217 340 1974 . . . . 2,140 686 1,019 435 sugar beets. The very limited commercial acre- 1975 . . . . 2,572 735 1,173 664 age of sweet sorghum has yielded about 1.9 ton/acre of sugar, however, the acreage is too SOURCE Agr/cu/tura/Stat/sl(cs 1977( Washington, D C U S Department of Agriculture, 1977) small to accurately reflect the yields that grass. Yields average about 1.5 to 2.5 ton/acre would occur from large-scale production of but could be increased to 4 to 5 ton/acre by rel- this crop. atively straightforward changes in manage- Forage crops are grown for feed and bed- ment practices. ding. Including alfalfa, the area under forage Most crops can be grown in several different 7 crop production is about 60 miIIion acres. For- areas of the country. However, each crop has age crops include orchard grass, brome grass, unique characteristics that enable it to do well tall fescue, alfalfa, clover, and reed canary- under certain combinations of soil type, growing season, rainfall, etc. Since these parame- 7Agricu/tura/ Statistics (Washington, D C.: U S Department of ters vary widely throughout the United States, Agriculture, 1978). it is unlikely that any one crop could prove to Ch. 3—Agriculture ● 61 be the correct energy crop for a given product. The crops mentioned here do not exhaust Rather, the available cropland can best be uti- the possibilities, even for starch, sugar, and lized for energy by growing various different cellulosic products. Other crops may be supe- crops suited to the various soil types, climates, rior to these under certain circumstances. But and other conditions. Nevertheless, there are these crops do serve to illustrate U.S. agricul- some striking differences when national aver- ture’s energy potential, costs, and impacts. age yields are compared. (See “Energy Poten- tial From Conventional Crops” below.)

Current Agricultural Practices and Energy and Economic Costs

As mentioned above, the purpose of manag- or disk and the crop planted. During planting ing a plant system is to provide an artificially some additional fertilizer may be added, an in- favorable environment for plant growth. Since secticide may be applied, and herbicides may increasing the intensity of management costs be broadcast on the soil surface. The crop may money, there is always a tradeoff between the be cultivated for weed control once or twice increased cost and the expected increase in within the first month of growth. No additional yields. As price relationships change, the inten- operations occur until the crop is harvested sity of management will also change. A sum- with a harvestor that separates the grain and mary of some current agricultural practices leaves the residue on the field. If the grain has and their costs and energy usage is given be- a moisture content above that needed for stor- low. age without spoilage, it is dried. The grain may be fed on the farm, stored and sold later, or Aside from weather and soil type, the plant- sold directly to a grain company at harvest. ing date, planting density, weed, disease, and insect control, and tillage practices can all af- Minimum or reduced tillage operations are fect crop yield. Different plants have different used to reduce soil erosion. With their use the sensitivities to these various factors. Practices soil may be chisel-plowed rather than mold- also have to be suited to the climate and soil board-plowed so that much of the residue re- type that is being farmed. Consequently, the mains in the surface. Herbicides may be used direct costs of farming will vary depending on to give complete weed control so that no fur- the crop and region. There can even be signifi- ther cultivation is needed. cant differences for the same crop within a Forage crop management is considerably given region (e. g., erosion control measures, simpler. Since forage crops are usually peren- irrigation, etc.). nials, crop planting is done only once every 4 A “typical” farming operation for annual to 5 years, or longer. Aside from planting, the only management is the application of fer- crops such as corn and soybeans, however, tilizers and the harvesting of the forage crop. might be as follows: After harvest of the crop in the fall the residues are chopped or the soil The estimated costs for producing corn (a is disked to reduce the size of the residues and row crop) and wheat (a close-grown crop) are to level the soil. Phosphate and potassium fer- shown in table 22. These costs are fairly repre- tilizer are broadcast and the residues and fer- sentative of what can be expected for the pro- tilizer are plowed under. In the spring, surface duction costs per acre for annual crop produc- tillage to level the soil is done soon after the tion, with intensive agriculture. Costs will vary soil becomes suitable for tillage. Nitrogen fer- from place to place, but where costs other tilizer — anhydrous (dry) ammonia, etc. — if than land costs are higher and/or yields are needed, is applied to the soil. Five to ten days lower, the land will be worth less and land later the soil is surface tilled with a cultivator costs will be lower. 62 ● VOl. II—Energy From Biological Processes

Table 22.-Estimated per Acre Production Costs Table 24.-Energy Inputs and Outputs for Corn in U.S. Corn Belt in Indiana, 1979 Energy units Production cost item Corn Wheat Nonirrigated a Sprinkler Yield per acre...... 110 bu 50 bu (10’ Btu) (105 Btu) Direct cost per acre output a Fertilizer and lime ...... $32.00 $22.50 Grain ...... 543.7 666.4 Seed and chemicals ...... 20.00 10.00 Residue ...... 543.7 666.4 Machine operating and drying ., ...... 25.50 11.25 Total output. , . . . , ...... 1087.4 1332.8 Interest on operating capital...... 9.00 7,00 Input Total direct costs. ., , ., ...... $86.50 $50.75 Irrigation pumping. , ., ...... – 60.0 Indirect costs per acre Fertilizer. ... , ...... 47.0 57.6 Machinery and equipment ...... $43.00 $18.00 Drying fuel ...... 19.4 23.8 Labor and management...... 31.00 20.00 Equipment fuel ...... 10.0 10.5 Grain storage (bin only)...... 11.00 – Pesticides...... 6.0 6.0 Land costb...... 92.00 92.00 Total input. ... , ...... 82.4 157.9 Total indirect costs ...... — $177.00$130.00 a&atfl yield, tsg bu/acre; residue yield: 7,770 ib/acre. Total costs per acre ...... $263.50 $180.75 bpump Irrigated ts inches water, 100 ft depth Grain yield and residue yield are 170 bulacre and Cost per bushel ...... 2.40 3.62 9,520 lb/acre, respectively. SOURCE Barber, et al , ‘ ‘The Potential ot Producing Energy From Agriculture, ’ Purdue Universi- aNitrWen $0, IZ/IfI for corn and $020 for wheat Phosphorus pentoxide Priced at prices at ty, contractor report to OTA, May 1979. SO.18/lb; potassium monoxide priced at $009 for all crops A corn-soybean rotation is assumed Thus soybeans produce a rutrogen credit for corn production and no msectide IS used b~nd COSIS approximate current cash rental rates least from 1.2 million Btu/ton for oats in lowa SOURCE Barber, et al , “The Potenllal of Producing Energy From Agriculture, ” Purdue Universi- to 6.5 million Btu/ton for grain sorghum in ty, contractor report to OTA, May 1979 Texas. For corn the variation is at least from 2.6 million Btu/ton of grain (Illinois average) to The energy used for farming varies consider- 4.6 million Btu/ton (Nebraska). The U.S. aver- ably. Typical energy inputs per acre for various age for corn is 3.1 million Btu/ton of corn crops are shown in table 23. These energies are grain. for cultivation without pumped irrigation. A comparison of the energy inputs for irrigated These differences reflect not only differ- and nonirrigated corn is shown in table 24. ences in cultivation practices and, yields, but Overall, the energy per ton of grain can vary at also the presence or absence of pumped irriga-

Table 23.-Energy Inputs for Various Crops (105 Btu per acre)

Corn Conventional Minimum No tillage tillage tillage Soybeans Wheat Alfalfa Nitrogen a...... 43.75 43.75 43,75 — 20,00 1.25 Phosphorus pentoxide + potassium monoxideb. . . . . 3.20 3.20 3.20 2.70 3.00 6.56 Drying c ...... 19.35 19,35 19.35 — — Diesel d Ground preparation ...... 7.36 e 5.13 f 2.21g 5.67h 3.15i — Planting...... 1.34 1.34 1.34 1.34 1.34 – Cultivation ...... 1.34 1.34 — 1.34 – 21.07 Harvest ...... 2.15 2.15 2.15 1,69 1.54 – Herbicides ...... 4.20 4.65 6.00 4.80 – – Insecticide...... , 1.80 1.80 1.80 — — 5.60 Total ...... 84.49 82.71 79.80 17.54 29.03 34.48

a25,000 Btu/lb nitrogen. fspread feflilizer, chisel plow, apply anhydrous ammOnla, field cultivate b3,000 Btu/lb phosphorus pentoxide and 2,000 Btu/lb potassium rnonoxlde gSpread fertilizer, spray, apply anhydrous ammonia. c93,500 Btu/gal LP-gas, 3,414 Btu/kW-hr electricity. hspread fedllizer, PIOW, disk, disk ’132 250 Btu/gal diesel fuel ioisk, disk, spread ferhlizer in sprin9 espr~d fe~llizer, pIOW, disk, apply anhydrous ammonia, disk 11211,000 Btu/lb actwe ingredient

SOURCE: Barber, et al , “The Potential of Producing Energy From Agricuffure, ” Purdue University, contractor repotl to OTA, May 1979, Ch. 3—Agriculture . 63 tion and the fuel used for irrigation. Examples required for pumped irrigation is more than of energy-intensive crops range from corn twice that shown in table 24.8 In all, 85 percent grown in Nebraska which is irrigated with of the 58 million acres of irrigated cropland ground water brought to the surface by elec- are in the West (Northern Plains, Southern tric pumps and is dried with liquefied petro- Plains, Mountain, and Pacific farm production leum, to grain sorghum which has relatively regions) and 94 percent of the 0.26 Quad/yr low yields compared to energy inputs through- used for pumped irrigation in the United States out the United States. Other crops, such as in 1974 was in the West.9 On the average, the rice, can be even more energy intensive (7.8 energy needed to pump the equivalent of 22 million Btu/ton, U.S. average). inches of rainfall in the West is 6 million Btu/acre. Consequently, this is a reasonable For most corn cultivation, over half of the average figure for the energy input due to irri- energy input comes from fertilizer, principally gation. nitrogen. However, without nitrogen fertilizers, average corn yields would drop from Another possible type of energy crop is for- about 100 bu/acre to less than 30 bu/acre. In age grass. Currently, little or no fertilizer is the example in table 23, the energy used would used to cultivate forage grass; and yields are increase from 3.0 million to 4.9 million Btu/ton about 2 ton/acre. However, if fertilizers were if nitrogen fertilizers were not used, assuming used and the crops harvested more times per the above yield change. year, additional biomass could be obtained. Table 25 shows the costs of producing grass The other big energy input for some areas—

irrigation —can have the opposite effect. In “D, Dvoskin, K, Nicol, and E. O. Heady, “ Irrigation Energy Re- the example given in table 24, the use of irriga- quirements in the 17 Western States,” Agrku/ture and Energy, W, tion raises the energy input from 2.2 million to Locheretz, ed. (Academic Press, 1977) %. Sloggett, “Energy Used for Pumping Irrigation Water in the 3.4 million Btu/ton. And in some areas (e.g., United States, 1974,” Agrku/ture and Energy, W. Locheretz, ed west Texas and southern Arizona), the energy (Academic Press, 1977).

Table 25.–Estimated Costs of Producing Grass Horbage at Three Yield Levels

Yield level (ton/acre) 2 3 4 Growing costs ($/acre) Fertilizer ...... — 19.45 b - 24.42C 41.59 d - 50.70e Seed and seeding ...... 2.30 2.30 2.30 Interest and miscellaneous. . . . . 0.22 2.07 - 2.54 4.17 - 5.04 Total...... 2.52 23.92 - 29.26 48.06 - 58.04 Harvest costs ($/acre) Machine operating...... 8.00 12.00 16.00 Interest and miscellaneous . . . . . 0.76 1.14 1.52 Machine investment...... 34.06 34.06 34.06 Hay storageg...... 0- 8.72 0 - 13.08 0- 17.44 Labor @ $4/hrh...... 3.68 -11.04 5.52 - 16.56 7.36 - 22.08 Total...... 46.50 -62.58 52.72 - 76.84 58.94 - 91.10 Total non-land costs $/acre ...... 49.02 -65,10 76.54 -106.10 107.00 -149.14 $/ton i...... 27.23 -32.55 28.35 - 35.37 29.72 - 37.29

alncludes cost of apphcatlon b60 lb nitrogen< 20 lb phosphorous pentoxide, 50 lb potassium monoxide/acre. c&I lb nitrogen, 30 lb phosphorous pentoxlde, 90 lb potassium rnonoxideiacre dlso lb nitrogen, 30 It) phosphorous perrtoxtde, 60 lb potassium monoxidelacre e150 lb nitrogen, 50 lb phosphorous pentoxide, 150 lb potassium rnmoxldejacre. ‘9-percent interest = O 5% miscellaneous costs gRange from no cost If iarge hay package stored outside to new barn costs for rectangular bales hHlgh labor values for rectangular bale handled by hand, low labor bales for lar9e hay Pa*a9es IAssumes 10% additional storage loss If hay stored outside (average st0ra9e period). SOURCE. Barber, et al , “The Potential of Producing Energy From Agriculture, “ Purdue University, contractor report to OTA, May 1979 64 . VOI. II—Energy From Biological Processes

herbage at various levels of fertilization and from this resource would require a 50- to 100- grass production. The additional production is percent increase in fertilizer use in agriculture. estimated to cost $28 to $37/dry ton. Note With no fertilization the energy used to pro- particularly that no land charges are included duce the grass is about 0.1 million Btu/ton of in these cost calculations, because the use of grass at the present estimated level of 2 ton/ the land has not changed. Only the output has acre. At 3 and 4 ton/acre, the additional energy been increased. Nevertheless, some farmer use is about 1.9 million Btu for the third and profit in addition to the labor charge may be 2.4 million Btu for the fourth ton. About 0.1 needed to induce farmers to increase produc- million to 0.2 million Btu/ton should be added tion. Furthermore, obtaining the full potential to these energy inputs for a 15-mile transport 1OS. Barber, et al., “The Potential of Producing Energy From of the grass. Agriculture,” Purdue University, contractor report to OTA, May 1979. ‘}Ibid

Energy Potential From Conventional Crops

Aside from crop residues, the two major First, let X represent the number of acres of near- to mid-term sources of bioenergy from average soybean production that can be dis- conventional crops are grains and sugar crops placed by growing 1 acre of average corn pro- for liquid fuels production and increased duction, converting the corn to ethanol, and forage grass production, On the land capable feeding the byproduct to livestock. Assuming of supporting grain and sugar crop production, that the corn yield on marginal cropland (i.e., grasses could also be grown; and a comparison the new cropland that can be brought into pro- of these choices is considered first. duction) is y times as great as on average crop- Iand, then 1 acre of marginal cropland grown The mechanism through which food and with corn for ethanol production results in a fuel production compete is the increase in byproduct that can displace yX acres of soy- farm commodity prices. Since farm commodi- bean production. Planting this yX acres with ty prices must rise in order to make it profit- corn for ethanol and using the distillery by- able for farmers to increase the quantity of product for animal feed displace an additional land under intensive production, it is impor- 2 yX acres of soybeans, etc. In all, the total acre- tant to examine the net quantities of premium age of average soybean production displaced fuels that can be displaced, through liquid by this marginal acre of corn is: fuels production, by each of the alternatives when new cropland is brought into production. 2 3 yx + yx + yx . . . . = Y X (1) (For details of the energy balances, see ch. 11.) 1 -x The calculations for sugar crops and grasses If Nm and Na are the net premium fuels dis- are relatively straightforward, since these feed- placement per acre of marginal and average stocks have very little protein in them and, corn production, then the total net premium consequently, the byproduct probably has lit- fuels displacement attributable to bringing 1 tle value as an animal feed (see “Byproducts” marginal acre into corn production is: in ch. 8). The distillation of grains, in contrast, yX produces a protein concentrate byproduct that N = Nm + Na (2) 1 -x can displace significant quantities of soybean () meal and thus soybean production. Additional The ideal crop switching technique would grains could then be grown on the land former- be where X =1, i.e., where one can simply ly devoted to soybean production. Estimates switch to another crop which produces all of of the effect of this substitution are calculated the products of the first crop and liquid fuels below. as well, without expanding the acreage under Ch. 3—Agriculture ● 65

intensive cultivation. Several imaginative sug- tillery byproduct is fully utilized to displace gestions for crop switching have achieved this soybeans. 1 n this calculation, 2.5 acres of aver- ideal but none are proven. 2 The closest to this age soybean land plus 1 acre of marginal land, ideal that has been demonstrated is the corn- all grown in corn for ethanol production, can soybean switch, in which X = 0.77, based on produce an equivalent amount of animal feed national average yields of the respective protein concentrate as 2.5 average acres grown crops. * 1 3 Nevertheless, even this switch is with soybeans, and provide the ethanol as limited by the quantity of land suitable for well. However, as the utilization (i. e., X in corn production and the fact that the corn dis- equation 2) drops, then grass quickly becomes tillery byproduct is not a perfect substitute for a superior alternative. If, for example, 1 lb of soybeans in all of its uses. As a fuel ethanol in- distillery byproduct displaces 0.5 lb of soybean dustry is first developing, however, these limi- meal instead of the maximum of 0.67 lb (see tations are probably of minimal importance. ch. 8), then grass and corn would be roughly equivalent. Similarly if grass yields increase to Assuming, then, that the distillery byproduct 8 dry ton/acre-yr, then grass would be as good is fully utilized and that marginal cropland or better than corn regardless of the byproduct produces 75 percent of the yield of average utilization. (It should be noted, however, that cropland, OTA has calculated the net premium these calculations do not take the economics fuels displacement per acre of marginal (new) of producing ethanol from grass or the diffi- cropland brought into production for various culties of using methanol as an octane-boost- liquid fuels options. These include ethanol ing additive into consideration. ) from various grains and sugarcane and both methanol and ethanol from grass. The energy Sugarcane appears to be roughly equivalent inputs were assumed to be national average to grass, but sugarcane can be grown on only a energy inputs for the various grains and sugar- limited amount of U.S. cropland and the etha- cane and 1 million Btu/dry ton for grass** and nol produced from it would be considerably the alcohols are assumed to be used as octane- more expensive than corn-derived ethanol. boosting additives to gasoline. The results are Other sugar crops have considerably lower shown in figure 12. Although the exact num- yields than sugarcane. bers cannot be taken too literally because of The exact point where the byproduct utiliza- the various assumptions required to derive tion will drop is unknown. Some analyses have them, the relative values are fairly insensitive put it at 2 billion to 3 billion gal/yr of ethanol to the assumptions chosen, provided the alco- when distillers’ grain is the distillery byprod- hols are used as octane-boosting additives. * uct. ‘4 Producing corn gluten meal could, how- Also, utilization of crop residues does not sub- ever, increase this to as much as 7 billion stantially change the results. gal/yr, based on the total domestic use of soy- Among the grain and sugar crops, corn ap- bean meal for animal feed. ’s As mentioned pears to be the best choice, as long as the dis- above, however, the byproduct is not equivalent to the soybean products, so it is unlikely that one can reach this level with full byprod- ‘2R Carlson, B Commoner, D Freedman, and R Scott, “inter- im Report on Possible Energy Production Alternatives in Crop- uct utilization to displace other crops. For the Llvestock Agriculture, ” Center for the Biology of Natural Sys- purposes of these estimates, it is assumed that tems, Washington University, St. LouIs, Mo , Jan 4, 1979 2 billion to 4 billion gal/yr of ethanol from corn ● The byproduct of 1 bu of corn can displace the meal from about O 25 bu of soybean. See “Byproducts” under “Fermenta- tion “ ‘ ‘Improving SoI/s With Organic Wastes, op. cit “R C. Meekhof, W E. Tyner, and F. D Holland, “Agricultural * *One-half that denved above for Increased grass production, Policy and Gasohol,” Purdue University, May 1979. This refer- because here it is assumed that the entire grass production goes ence reports a 3-billion-gal limit based on a 2-lb substitution of to energy distillers’ grain for 1 lb of soybean meal. Other studies, however, *If the alcohols are used as standalone fuels, the relative val- have put the feed ratio at 1 5:1, which would reduce the limit to ues are similar, but the net displacement is about half that shown 2.25 billion gal/yr In figure 12 “improving Soi/s With Organic Wastes, op cit. 66 ● Vol. n-Energy From Biological Processes

Figure 12.—Net Displacement of Premium Fuel(oiland naturalgas) per Acre of New Croplnd Brought into Production

Crop Alcohol Net premium fuel displacement per acre of marginal cropland brought into production (energy equivalent of barrelaof oil/acre-yr)

Grains and sugar cropsb o 5 10 15 20 Corn Ethanol Grain sorghum Ethanol Spring wheat Ethanol Oats Ethanol Barley Ethanol Sugarcane Ethanol by displacement of other crops with byproduct Otherd Grass or other crops with high dry-matter yields.

(4 ton/acre-y r’) Ethanol (10 ton/acre-yr) Ethanol

(4 ton/acre-yr e) Methanol (10 ton/acre-yr) Methanol

aBased on 59 mllllon Btu/bbl alcohol used as octaneboostlng addltlve to 9asOllne

bAssumes national average energy ,nput~ per acre cultivated and ~lelds (On the nl’rgln’l cropland) Of 750/0 of the national average yields between 1974.77 Yields on average cropland are assumed to be the average of 1974.77 national averages This methodology IS Internally consistent, ralstng the average cropland yield to 1979 yields would not slgnlf!cantly change the relative results ff usable crop residues are converted to ethanol. the lower value (no distillery byproduct utlllzat[on) would be Increased by about 1 2 bbl/acre.yr or less for the grains and 26 bbl/acre.yr or less for sugarcane c&onoml~ and physical Opportunltles for full byproduct utlllzatlon dlmlrrlsh with greater quantities of by Product production duncerta(nty of ~ 3000 for methanol and more for ethanol from grass, since the ethanol processes are not well defined at present Assumes 1 mllllon 13tu/dry ton of grass needed for cultivation harvest and transport of the grass and conversion process yields (after all process steam requirements are satlsffed with waste heat or part of the feeds tock} of 84 gal/dry ton of grass for ethanol and 100 gal/dry ton of grass of methanol

‘Four ton/acre. yr can be ac hleved with current grass varletleS grown On mar91nal crodand SOURCE Off Ice of Technology Assessment yields from USDA Agricultural Statistics. 1978

Data Used in Figure 12

Net premium fuels displacementa(bbl of oil equivalent/acre) National average farming Land that is 75 percent as 3 Average 1974-77 national energy (10 Btu/gal of ethanol) Average land productive as average land average yields (gal of Land that is 75 percent as With byproduct Without byproduct With byproduct Without byproduct Crop ethanol/acre) Average land productive as average credit credit credit credit Corn ...... 220 33.3 44.4 4.4 4.0 3.0 2.7 Grain sorghum. 130 54.5 72.7 2.1 1.9 1.3 1.1 Spring wheat ... 73 23.8 31.7 1.6 1.5 1.1 1.0 Oats ...... 74 24.2 32.3 1.6 1.5 1.1 1,0 Barley ...... 79 29.4 39.2 4.6 1.5 1.1 1.0 Sugarcane . 504 30.3 40.4 NA 9.7 NA 6.4 b Grass ... 400 NA —10 . — NA NA ——.—.—NA . 7.3 A =-none available. aA~sumes gross displacement of 140,000 BtU/g’l of ethanol, byproduct credit of 10,500 Btulgal, and 5.9 million Btulbbl of Oil. For methanol, 117,000 Btulgal 9ross displacement. bASSumeS 4 tonlacre on marginat land and 100 gal methanol PertOn.

0.75X Net premium fuels displacement from 1 acre of marginal land Displacement of soybean productionc1-x (total acres of soybeans displaced by 1 0.75X (average acres of soybeans displaced marginal acre of grain and additional cultivation plus 1-x acres displaced soybean land (bbl oil Crop per average acre of grain = x) of grain on former soybean land equivalent/acre of marginal land) Corn...... 0.77 2.5 13.9 Grain sorghum 0.46 0.64 2.7 Spring wheat. 0.26 0.26 1.5 Oats...... 0.26 0.26 1.5 Barley ...... 0.28 0.29 1.6 Sugarcane, 0 0 6.4 Grass...... 0 0 7.3 — ‘Assumes average soybean yield of 27.1 bulacre, a displacement of 1 lb of soybean meal per 1.5 lb of distillers’ grain, and 48 fb of soybean meal per bushel of soybeans. SOURCE: S. Barber, et al., “The Potential of Producing Energy From Agriculture,” Purdue University, contractor report to OTA; and Agricultural Statwtics, 1978 (Wash- ington, D. C.: U.S. Department of Agriculture, 1978). Ch. 3—Agriculture ● 67

(about 0.2 to 0.4 Quad/yr) can be produced tionary to food prices in the long term per unit while utilizing the byproducts fully. This would of liquid fuel produced. However, if 65 million require about 2 million to 5 million additional acres are available for energy production in acres in intensive crop production and expan- 2000, one could conceivably produce over 15 sion of corn production by over three times billion gal of ethanol from corn* or about 1.3 this acreage. It is not certain that cropland will Quads/yr of liquid fuel. Grass production, on be available for energy production by 2000; the other hand, would yield about 2.5 Quads/ but if it is, it is assumed that any further pro- yr** of liquid fuel from this same cropland and duction above this level will use grass as the with the same or lower inflationary impact. energy crop. Judging when the emphasis should shift In the near to mid-term, increased produc- from corn to grass is likely to be difficult. As a tion of forage grass can be obtained on about fallible rule of thumb, however, any significant 100 million acres of hayland, cropland pasture, increase in corn prices relative to the other and noncropland pasture. Assuming a 1- to 2- grains would be an economic signal to distil- ton/acre-yr increase in yields, this would result lers and/or animal feeders to use grains other in 100 million to 200 million tons of grass or than corn, which would make grass a superior about 1.3 to 2.7 Quads/yr. Deducting the ener- option for energy production. Similarly a sig- gy needed to cultivate and transport the grass nificant drop in the price of distillery byprod- reduces the output to about 1.1 to 2.2 Quads/ uct, relative to the alternatives, would be an yr. economic signal that the distillery byproducts are not being fully utilized and, again, grass By 2000 anywhere from zero to 65 million would be superior. Consequently, if there is a acres could be used for energy crops. Assum- significant rise in corn prices or drop in distil- ing that grasses with average yields of 6 dry lery byproduct prices, relative to the alterna- ton/acre-yr on this cropland have been devel- tives, then these could be indications that the oped, then the energy potential would be zero cropland could better be utilized by producing to 5 Quads/yr. grass. Although adding this to the ethanol yield from corn involves a small amount of double counting, the uncertainty in the actual crop- Iand availability and future grass yields is too great to warrant a detailed separation. Conse- quently, OTA estimates that 1 to 3 Quads/yr can be obtained in the near to mid-term and zero to 5 Quads/yr in the long term from the *Seven billion gal with complete substitution of soybean meal and requiring about 10 million of the 65 million acres. The reproduction of conventional crops for energy. maining 55 million acres, with yields of 65 bu/acre, could produce an additional 9 billion gal/yr of ethanol. The above mix of corn and grass was chosen **5 Quads/yr of grass could yield slightly less than 25 as the one that appears to be the least infla- Quads/yr of methanol

Energy Potential From Crop Residues

Crop residues are the plant material left in of soil organic matter. (See also “Environmen- the field after a crop harvest. Their major func- tal Impacts.”) tion is to protect the soil against wind and water erosion by providing a protective cover, and they have a modest fertilizer value16 and a Barber, et al., have calculated the total soil-conditioning value through maintenance quantities of residues by multiplying the crop “Residues are about O 7 percent nitrogen, O 2 percent phos- yields reported by USDA by residue factors, phorus, and 4 percent potash See Barber, et al., op cit. i.e., the ratio of residue to the yield of tradi- 68 . Vol. Ii—Energy From Biological Processes tional crop for the various types of crops. ’7 addition, a 15-percent storage loss was as- The results of these calculations are shown in sumed. The results of these calculations are table 26. The total quantity of residues gener- shown in table 27 as the usable crop residues, ated is about 400 million ton/yr or about 5 which are about 20 percent of the total crop Quads/yr. residues.

Table 26.-Total Crop Residues in the United States for Table 27.–Total Usable Crop Residue by Crop 10 Major Crops (based on 1975-77 average production) Amount Harvestable acres Average yield Acres Total residue Crops (k tons) (k acres) (ton/acre) k acres k tons Corn...... 37,098 39,122 0.95 Corn...... 69,530 171,084 Small grains . . . . 33,623 36,324 0.93 Wheat ...... 68,789 99,890 Sorghum...... 1,452 4,100 0,35 Soybeans ...... 53,616 67,556 Rice ...... 5,457 2,516 2.17 Sorghum ...... 14,714 21,123 Sugarcane...... 590 331 1.78 Oats. ., ...... 12,831 20,677 Total ... , . . . . 78,220 82,393 0.95 Barley ...... 8,772 13,341 Rice...... 2,515 8,584 SOURCE Barber, et al., “The Potential of Producing Energy From Agriculture, ” Purdue IJnlversl- Cotton ...... 10,990 3,578 ty, contractor reporl to OTA, May 1979 Sugarcane ...... 660 4,700 Rye ...... 715 708 Harvesting crop residues would typically U.S. Total ...... 243,132 411,240 consist of moving the residues into windrows, SOURCE. Barber, etal , “ThePotentialofProducingEnergy From Agriculture,’’ Purdue lJnwersh or long thin piles of residues. The windrows Iy, contractorreporttoOTA, May1979 would then be collected with baling machinery During fall plowing many farmers turn under and the bales dumped at the roadside. The the residues, rendering them useless as a pro- windrows would be collected and transported tection against erosion. These residues could to a place where they would be stored or used. be collected and used for energy without wor- Crop residues typically contain 40- to 60-per- sening the erosion on these lands. However, cent moisture 2 days after the grain harvests. current agriculture policy is to encourage In favorable weather conditions, the residues farming practices that Iimit soil erosion to the dry to about 20-percent moisture in 18 days. ” soil-loss tolerance levels, or the levels of ero- With these moisture contents, bacteria will sion that are believed not to impair the long- gradually consume the residues. If the residue term productivity of the land (see “Environ- bales are compacted too tightly, the heat gen- mental Impacts”). Consequently, a more de- erated from the bacterial action can cause the tailed consideration of crop residues is ap- material to spontaneously combust. However, propriate. with relatively loose bales, the bacterial heat- Using data supplied by Dr W. Larson,’* the ing will dry the material to a moisture content total crop residues were calculated for each of at which the bacteria do not consume the ma- the major land resource areas or subregions of terial. Some loss, however, is inevitable (15- States. Using standard equations for soil ero- percent loss has been assumed in table 27). sion,’ the quantities of residues that could be The extra fertilizers necessary to replace the removed without exceeding standard soil-loss nutrients in the residues removed cost about tolerance values were calculated. These were $7.70/ton of residue removed. then modified to take into consideration the quantities that can be physically collected In addition, one of the main problems with with current harvesting equipment (about 60 harvesting residues is that it delays the fall percent in field trials at Purdue University). In ground preparation. If winter rains come too soon, there may not be sufficient time to col- 171 bid lect the residues and prepare the ground for “W. E. Larson, “Plant Residues–How Can They Be Used Best,” paper No. 10585, Science Journal Series, SEA-AR/USDA, the spring planting. The spring planting is then 1979 ● Universal soil-loss equation and wind erosion equation. “Barber, et al., op. cit. Ch. 3—Agriculture ● 69 delayed and yields for the following year may costs might be as low as $20/dry ton and, in un- suffer. Using computer simulations of farming favorable cases, as high as $60/dry ton or more. operations and the actual weather conditions Crop residues contain about 13 million Btu/ in central Indiana between 1968 and 1974, it ton. The energy costs for harvesting and trans- was found that residue harvests reduced corn 20 porting the residues are about 0.9 million yields by an average of 1.6 bu/acre-yr. If this Btu/ton for a 15-mile transport and 1.8 million cost is attributed to the residues, then it raises Btu/ton for a 50-mile transport. (With inte- the residue costs by $2.70/ton. This factor is grated residue and crop harvests the energy less of a problem with most other grains, how- costs would be slightly less, but this may not ever, since they are less sensitive to the exact be a practical alternative because it delays the planting time. harvest.) In addition, the energy content of the Adding these various costs and assuming a additional fertilizer needed to replace the nu- markup of 20 percent above costs gives the trients lost in the residues is about 0.6 million State average costs for various residues (table Btu/ton. Thus, the total energy use associated 28). Care should be exercised when using this with collecting and transporting the residues is table, however, since the costs within a State about 1.5 million to 2.5 million Btu/ton of resi- can vary considerably. In favorable cases the due.

’01 bid National average crop yields can fluctuate by ± 5 percent or more from year to year and Table 28.–State Average Estimated Usable Group the usable crop residues will fluctuate by Residue Quantities and Costs about ± 10 percent, since an absolute quantity of residue should be left regardless of the crop a Total usable crop Delivered cost yield. On a local basis, usable crop residues residues (million ($/ton) (estimated State tons/yr) uncertainty: 20%)) can vary considerably. Within a county lo- Corn cated in a humid region of the country, the Illionois ...... 8.0 32.16 fluctuation may be ± 20 percent and for crops Indiana ...... 4.6 32.42 that are not irrigated in dry regions, the year- lowa ., ...... 6.9 32.77 Minnesota ...... 4.2 38.67 to-year variations can reach ± 100 percent. Nebraska. ., ...... 1.8 41.68 The areas with the largest fluctuations, how- Ohio ., ...... 2.6 35.18 ever, also have the lowest quantities of usable Small grains residues. California...... 1.8 28.29 Illinois...... 1.0 31.53 In summary, the total crop residue produc- Minnesota ., ...... 6.1 30.54 South Dakota ...... 1.8 33.05 tion in the United States is about 5 Quads/yr, Washington ... . . 3.0 31.01 of which about 3 Quads/yr can be collected Wisconsin ...... 2.0 26.93 with current harvesting equipment. The quanti- Sorghum grain ty that can be collected while maintaining cur- Colorado ...... 0.12 35.60 Kansas . . ., ...... 0.72 57.62 rent soil erosion standards is about 1 Quad/yr. Missouri ...... 0.28 36.87 Considerations of a reliable supply, however, Rice would reduce this to roughly 0.7 Quad/yr* of Arkansas, ...... 1.9 36.32 reliable feedstock, if soil erosion standards are California. ., ...... 1.1 34.82 Texas ...... 1.2 36.08 strictly adhered to. By 2000, increases in crop Sugarcane production could raise this by 20 percent to 0.8 Florida...... 0.53 30.93 to 1.2 Quads/yr. alncludmg 15-rnlle transport, labor at $5/hr, $0 80/gal diesel fuel, y!eld Penalty of $2 70/lon of ● residues for corn, adctmonal ferlhzers for $7 70/ton of residues, and profit of 20 percent ot Calculated by assuming that the total quantity of residue can costs fluctuate by ± 20 percent at the local level; i.e , by subtracting SOURCE Barber, et al , < ‘The Potential of Producing Energy From Agriculture, Purdue Unlversl- 20 percent of the total residues from the usable residues on a ly, contractor report to OTA, May 1979 State-by-State basis 70 . Vol. II—Energy From Biological Processes

Environmental Impacts of Agricultural Biomass Production

Introduction Table 29.-Environmental Impacts of Agriculture American agriculture, with its astonishing Water use (irrigated only) that can conflict with other uses or cause productivity and reliability, bestows critically ground water mining. important benefits on the economy and gener- Leaching of salts and nutrients into surface and ground waters, (and al well being of the United States. Unfortu- runoff into surface waters) which can cause pollution of drinking water supplies for animals and humans, excessive algae growth in streams nately, it also has serious negative environ- and ponds, damage to aquatic habitats, and odors. mental impacts. Any substantial increase in Flow of sediments into surface waters, causing increased turbidity, land cultivation or intensification of present obstruction of streams, filling of reservoirs, destruction of aquatic habitat, increase of flood potential. crop production to produce energy crops— Flow of pesticides into surface and ground waters, potential buildup in biomass–will cause an extension and intensi- food chain causing both aquatic and terrestrial effects such as thinning fication of many of the impacts of the present of egg shells of birds. Thermal pollution of streams caused by land clearing on stream banks, system. loss of shade, and thus greater solar heating. - There are substantial uncertainties in the Air • Dust from decreased cover on land, operation of heavy farm machin- understanding of the consequences of relying ery. on agricultural feedstocks for energy produc- ● Pesticides from aerial spraying or as a component of dust. tion. These uncertainties stem from a lack of ● Changed pollen count, human health effects. ● Exhaust emissions from farm machinery. complete understanding of present impacts, Land the potential for changes in crop production ● Erosion and loss of topsoil from decreased cover, plowing, increased methods in the future, and uncertainty as to water flow because of lower retention; degrading of productivity. ● Displacement of alternative land uses–wilderness, wildlife, esthetics, the pace of development. This section at- etc. tempts to place the potential impacts of large- ● Change in water retention capabilities of land, increased flooding po- scale biomass production from agriculture into tential. ● Buildup of pesticide residues in soil, potential damage to soil microbial perspective by briefly describing what is populations. known of the impacts of food crop production ● Increase in soil salinity (especially from irrigated agriculture), degrad- (the energy feedstock production system ing of soil productivity. ● should resemble the food production system), Depletion of nutrients and organic matter from soil. Other describing how the pace of development may ● Promotion of plant diseases by monoculture cropping practices. intensify impacts, and finally identifying those ● Occupational health and safety problems associated with operation of differences between food and energy feed- heavy machinery, close contact with pesticide residues and involve- ment in spraying operations. stock production that are most critical in determining impacts. SOURCE Office of Technology Assessment

The Environmental Impacts of effect on crop yields are almost certainly the American Agriculture most important) many control techniques are rarely used. Agriculture is a major source of pollution and causes serious environmental impacts. Water pollution and land degradation due Table 29 lists the major environmental impacts to erosion are American agriculture’s primary associated with present forms of large-scale problems, and the two impacts are intimately mechanized agricultural production. Most of linked. The action of wind and water strips the impacts apply to the majority of farming farmland of its productive topsoil cover, and situations (although with varying magnitude), much of this soil ends up in the Nation’s water- but some impacts are negligible or nonexistent ways. Thus, estimates of soil erosion are criti- in certain situations. Also, most of the impacts cal to understanding the effects of agriculture are more or less controllable, but for a variety on both soil productivity and on water ecosys- of reasons (a high perceived cost or negative tems. Ch. 3—Agriculture ● 71

SCS has recently revised downward its esti- ● Although SCS generally considers 5 ton/ mates of cropland erosion. Its 1977 National acre-yr as an (average) annual erosion at Erosion Inventory estimates average annual which long-term productivity on good sheet and rill erosion from all cropland to be soils will not suffer, it is not certain that 4.77 ton/acre-yr (or a total of about 2 billion soil is actually replaced this fast. Authori- ton/y. 21 Previously, it had estimated cropland tative estimates of soil replacement rates sheet and rill erosion at about 9 ton/acre-yr,22 do not exist, but average rates of as low as and other sources had estimated total erosion 1.5 ton/acre-yr have been claimed. z’ How- (including wind erosion) from croplands to be ever, the SCS rates do represent the gener- as high as 12 ton/acre-yr.23 SCS attributes the al consensus of the agronomy community. decrease to the greatly improved data base re- ● Even the new lower erosion rate implies cently made available by the 1977 Inventory. that about a billion or more tons of sedi- Also, the original 9-ton/acre-yr estimate appar- ment from croplands are entering the Na- ently referred only to land in row crops, close- tion’s waterways each year.25 grown crops, and summer fallow, whereas the ● Erosion rates from croplands are many more recent estimate includes lands that are in times higher than those of natural ecosys- less intensive (and less erosive) uses such as tems. Forests typically erode at a rate of rotation hay and pasture, or native hay. less than one-tenth of a ton/acre-yr, and grassland at less than half a ton/acre/yr.26 Data from the 1977 Inventory has only recently begun to be released to the general pub- As a result of the mismatch between erosion lic, and it seems likely to generate contro- and soil replacement, the United States has versy — especialIy because its estimate of aver- lost a considerable portion of its topsoil and, age erosion is under the 5 ton/acre-yr that SCS some have claimed, its production potential. generally considers to be a tolerable level (i. e., Pimentel estimates that U.S. cropland has lost a level that wilI not affect long-term produc- about one-third of its topsoil and 10 to 15 per- tivity) for much U.S. cropland. However, the cent of its production potential over the last lower estimate is not especially comforting for 200 years. z’ Bennett estimates that, during the a number of reasons: period prior to 1935, 100 million acres of crop- I and were lost to erosion and an additional 100 ● National (sheet and rill) erosion rates for million acres were stripped of more than half cropland in intensive production are esti- of their topsoil .28 At best, however, these val- mated by SCS to be 6.26 ton/acre-yr. ues represent extremely rough estimates, and ● The national estimate tends to hide sever- the new SCS erosion inventory may cause their al important food-producing areas with downward revision. uncomfortably high erosion rates (e. g., Missouri averages 11.38 ton/acre-yr; Iowa It appears likely that the process of land averages 9.91 ton/acre-y r). degradation will continue for the immediate ● The estimates do not include wind erosion future. Although USDA has spent nearly $15 and alternative forms of water erosion. billion in its soil conservation programs since Cropland wind erosion in 10 western their inception in 1935,29 only 36 percent of the States averages 5.29 ton/acre-yr. Thus, although Texas cropland has a sheet and rill “Ibid erosion rate of only 3.47 ton/acre-yr, its to- “Environmental Implications of Trends in Agriculture and Silvi- tal erosion rate is greater than 18 ton/acre- cu/ture– Vo/ume 1: Trend Identification and Evo/ution (Washing- ton, D C Environmental Protection Agency, October 1977), yr because of wind erosion. EPA-60013-77-I 21 ‘61 bid * 7D Pimentel, op cit “H H Bennett, So;l Conservation (New York. McGraw-Hill, “1977 SCS Natfona/ Erosion Inventory Estimate, op. cit. 1939), ’21 bld “TO Protect Tomorrow’s Food Supp/y, Soi/ Conservation Needs “D Plmentel, et al , “Land Degradation, Effects on Food and Priority Attention (Washington, D C : General Accounting Of- Energy Resources, ” Science, VOI 194, Oct 8, 1976 fice, Feb 4, 1977), CED-77-30 72 . Vol. II—Energy From Biological Processes

472 million acres of cropland in 1967 were lost to agricultural erosion represents more judged to have adequate conservation treat- than half of the sediment entering the Nation’s ment30 and the programs have been criticized surface waters. 33 34 Sediment causes turbidity, as inadequate by the General Accounting Of- fills reservoirs and lakes, obstructs irrigation fice.31 canals, and destroys aquatic habitats. Yearly material damages have been estimated at over A reason for the inability of USDA conserva- 35 tion programs to satisfy their critics may be the $360 million, not including damage to aqua- difficulty of demonstrating to the farmer (in all tic habitats and other noneconomic costs. Adding the flooding damage caused by the de- but the more severe cases) the benefits of additional conservation measures. Because an inch crease in storage capacity of reservoirs and streams would increase annual costs to over $1 of topsoil weighs about 150 ton/acre, a net loss 36 of 5 ton/acre-yr would result in a loss of 1 inch billion. of soil every 30 years. During that time, farm- The effects on aquatic ecosystems of the ing procedures would be gradually changing, enormous flow of sediments into the Nation’s obscuring the effects of any soil loss. For exam- waterways have never been satisfactorily esti- ple, during the past 30 years, more intensive mated. Research on the impacts of “nonpoint” use of fertilizers, pesticides, and other inputs, sources of water pollution—agriculture, con- better information on future weather and struction, etc. –has not been given a high pri- other critical factors, and improved crop varie- ority within the Environmental Protection ties more than compensated for erosion- Agency (EPA) or USDA, and the result is a scar- caused losses on most lands. Also, the actual city of information from which to draw conclu- effect on productivity may not be large in sions about either present impacts or future some circumstances because the effect of soil impacts associated with the devotion of mil- loss is very sensitive to soil conditions: while lions of additional cropland acres to biomass loss of soil from a very shallow soil over rock in production. Kentucky may cause the land to be withdrawn from production, on some deep Ioess soils of The other major water pollution problems of lowa, the loss of several inches of topsoil may agriculture involve the toxic effects of pesti- have little effect on productivity. Few if any cides and inorganic salts and the nutrient in- agricultural scientists would argue that net soil flux into the Nation’s waterways associated with American agriculture’s increasing use of loss can continue indefinitely without major losses in productivity. However, on many lands fertilizers. the damages of erosion may never become vis- Pesticide use in American agriculture has ible to the farmer; rather they will be perceived grown from 466 million lb in 197137 to 900 by his children or grandchildren. Moreover, million lb in 1977.38 By 1985, American farmers short-term economic constraints may compel a are expected to be using as much as 1.5 billion farmer to discount the future benefits of con- lb.39 Much of this increase can be traced to the servation by much more than he would person- growth in minimum tillage practices40 which ally prefer. substitute increased herbicide use for tillage to Aside from the long-term consequences in control weeds. These practices include leaving land degradation, soil erosion represents a crop residues on the soil surface, and these severe water pollution problem. Not only is residues harbor plant pests and pathogens and soil itself a serious pollutant, it also acts as a generally increase pesticide requirements (al- carrier of other pollutants: phosphorus, pesti- ‘]Ibid. cides, heavy metals, and bacteria.32 The soil “Pimentel, op. cit. 357977 SCS Nationa/ Erosion /nventOry f5t;mafe, 0p. cit. “Ibid. ‘“’’Potential Cropland Study, ” Statistical Bulletin No. 578, Soil “Environmental /mp/ications of Trends, op. cit. Conservation Service, U.S. Department of Agriculture, 1977. ‘S7977 SCS National Erosion Inventory Estimate, op cit. 31 T. protect Tomorrow’s Food SUPp/Y, OP. cit. ‘;l bid, Jl~nvjronmental /mp/ications of Trends, OP. cit. 40[nvironmenta/ Implications of Trends, op. cit. Ch. 3—Agriculture ● 73 though they offer substantial benefits in ero- crease in crop cultivation will be the concur- sion control). Recent growth in the practice of rent increase in pesticide use on the new lands. single- and double-cropping may also account There is a distinct possibility that rising public for some of the increase. Although less than 5 concern over pesticide usage could put a se- percent of the pesticides enter the surface and vere constraint both on the continuing in- ground water systems, ” pesticide use has been crease in this usage and on the expansion of associated with fish kills and other damage to crop production for energy feedstocks. aquatic systems as well as reproductive fail- Salinity increases caused by irrigated agri- ures in birds and acute sickness and death in culture present another substantial impact. lr- animals. Under conditions of high exposure— rigated land produces one-fourth of the total in accidental spills, improper handling by ap- value of U.S. crops, mostly in the 17 western plicators, etc. —pesticides have been associ- States. ’J Increased salinity in streams in these ated with the sickness and death of humans. areas is caused by the salts added to irrigation Recent research has implicated some widely water from upstream farms and by the concen- used pesticides as possible carcinogenic trating effect of the high evaporation rates in agents when ingested or inhaled, and EPA has arid climates (evaporation leaves the salts be- removed certain of these— including Aldrin, hind). The same mechanisms can lead to in- Dieldrin, and Mirex–from the marketplace creasing salt concentrations in the soils of under the Federal Insecticide, Fungicide, and downstream farms unless sufficient water can Rodenticide Act (FIFRA). Amendments to be obtained to periodically flush excess salts FIFRA have considerably tightened the require- out of the soil profile. Damages associated ments for testing and registering pesticides. with increased salinity of soils and irrigation However, the tremendous variety of pesticide water include reduced crop yields, inability to compounds [“1 ,800 biologically active com- grow salt-sensitive crops, increased industrial pounds sold domestically in over 32,000 dif- treatment costs, and adverse effects on wild- ferent formulations”42 ) and the difficulty of life, domestic animals, and aquatic ecosys- detecting damages in human populations and tems. Trends in irrigated agriculture are lead- in the environment will greatly complicate suc- ing to improvements in irrigation efficiency cessful enforcement of the Act. At present, the and decreased salt loadings in streams, but long-term impacts of pesticides on the environ- these trends could be overwhelmed by sub- ment and on man are poorly understood. stantial increases in irrigated acreage either to The problems of pesticide use in agriculture grow crops for energy or to compensate for are becoming particularly visible because of competition between food and biomass pro- a recent rash of instances where pesticides duction in other areas. thought to be safe have been accused of caus- Fertilizer use is of extreme importance in ing severe injuries — including birth defects, calculating the environmental impacts of agri- miscarriages, and other acute physical disor- culture. Large amounts of energy— one-third ders–and death to exposed populations, The of the energy consumed by the agricultural controversy surrounding the use of the her- sector and its suppliers— are needed to pro- bicide 2, 4, 5-T in Oregon and its suspension by duce fertilizer. The Haber-Bosche process for EPA is a widely publicized– but by no means the synthesis of anhydrous ammonia fertilizer unique —example of rising national concern. 3 requires around 21 ft of natural gas to pro- Resolution of the conflicting claims about the duce 1 lb of nitrogen in fertilizer (and more for safety (or lack of it) of these pesticides is well 44 other forms of nitrogen); current U.S. nitro- beyond the scope of this report. Based on cur- gen fertilizer production is 10 million metric rent interest, however, it is likely that a major public concern associated with any large in- 4’1 bid “C. H. Davis and G M Blouin, “Energy Consumption In the U.S. Chemical Fertilizer System From the Ground to the

4’ I bid Ground,” Agriculture and Energy, W. Lockeretz, ed. (New York: 421977 SCS Nat;ona/ Erosion Inventory Estimate, op. cit. Academ~c Press, 1977), pp 315-371

67-968 0 - 80 - 6 74 . Vol. Il—Energy From Biological Processes

tonnes per year consuming 3 percent of total combination of artificially low prices for water U.S. natural gas production. If current trends and the requirement of the appropriation doc- of increased rates of fertilizer applications trine that the holder of a water right must continue and food demands increase by 3 per- maintain that right through use (“use or lose”) cent per year, natural gas requirements for fer- has led to the cultivation of water-intensive 45 tilizer production will triple by 2000. crops in arid climates. This has led to water shortages in many Western basins and to ag- The application of large quantities of chemi- gravation of salinity problems in several major cal fertilizers also represents a water pollution rivers. problem because much of the nutrient value ends up in the Nation’s waterways. Wittwer Several water use trends will affect agricul- estimates that only 50 percent of the nitrogen tural production capabilities in the near fu- and less than 35 percent of the phosphorus and ture. First, large-scale energy development— potassium applied as fertilizer are actually re- especially electrical generating stations and, covered by crops;46 other estimates for nitro- possibly, synthetic fuel plants–will consume gen range from 46 to 85 percent.47 Although a substantial quantities of water and, in some portion of that which is lost is due to volatiliza- cases, compete directly with agricultural inter- tion (and consequent loss to the atmosphere), ests for the limited supply. Second, expanded much is lost to surface and ground waters via acreage for food production will occur, in- runoff, leaching, and erosion processes. The cluding projects on Indian land that may have amount of nitrogen and phosphorus (potassi- priority rights to the limited water supply. On um is not considered to have significant en- the other hand, improvements in irrigation effi- vironmental impacts48) entering the waterways ciency will have some conserving effect on wa- from agricultural lands in the early 1970’s has ter consumption even though this is not a pri- been estimated at 1,500 million to 15,000 mil- mary goal of efficiency increase (the primary lion lb/yr and 120 million to 1,200 million lb/yr, goal is to reduce water withdrawals and ‘return respectively.49 flows and to improve water quality rather than to reduce consumptive use). For example, SCS This nutrient pollution from fertilizers may estimates that irrigated acreage in the critical be toxic to humans and wildlife in high concen- Upper Colorado River Basin could increase trations; nitrate poisoning of wells from con- from 1,370,000 acres in 1975 to 1,442,000 acres taminated ground water is not unusual in some in 2000 while water consumption declines by agricultural areas. The more common impact, 93,000 acre-ft with a concerted program to im- however, is to speed up eutrophication of so prove irrigation practices. Further decreases streams and the problems of oxygen demand in water consumption are possible by “crop and algae growth associated with eutrophica- switching” — shifting to less water-intensive tion. crops where markets are available—and re- The remaining major water-associated im- moving marginal, low-productivity land from 51 pact of agriculture is water use. The appropria- cultivation. Also, substantial potential for tion water rights system in the West offers lit- water conservation exists in energy produc- tle incentive to use water efficiently. * The tion.

45S. H Wittwer, “The Shape of Things to Come, ” l?io/ogy of Much of the agricultural land in the United Crop Productivity, P. Carlson, ed. (New York: Academic Press, States was’ obtained by forest clearing or plow- 1978), *bIbid ing native grasslands, and the consequent re- 47Environmenta/ /mp/ications of Trends, op. cit. placement of natural ecosystems with inten- ‘al bid, sively managed monoculture must be consid- “Ibid. *For an excellent review of Western water law see E, Radose- 50Conservation Needs Inventory (Washington, D. C.: Soil Con- vitch, “Interface of Water Quantity and Quality Laws in the servation Service, U.S. Department of Agriculture, 1976). West, in Proceedings of the Nationa/ Conference Irrigation Re 51S. E Plotkin, H. Gold, and 1. L. White, “Water and Energy in turn F/ow Quantity Management, J. P. Law and G V. Skogerboe, the Western Coal Lands,” Water Resources Bu//etin, vol. 15, No. eds. (Fort Collins, Colo.: Colorado State University, 1977). 1, February 1979. Ch. 3—Agriculture ● 75 ered a major environmental impact of agricul- the potential impact of a surge in production tural production. (This process is not a one-way caused by incentives to grow crops for biomass street. A combination of changing crop pat- energy production. terns, alternative producing areas, increasing average productivity, and, especially in the Of the 8.9 million acres, SCS estimated that South, depletion of soils has led during this 5.1 million acres had inadequate conservation century to the abandonment of considerable treatment and water management, and 4 mil- farmland acreage and, in many cases, rever- lion acres had inadequate erosion control. sion to second-growth forest. Principle areas These problems in land selection and environ- involved in this transformation include the mental planning were soon translated into Piedmont areas of the Southeast, the hillier severe erosion losses. Although poor weather areas of the Northeast, and the upper lake conditions (fall and winter drought in the States. However, farm abandonment no longer southern high plains, spring floods in the north- appears to be a significant force.52) Aside from ern Great Plains, torrential spring rains fol- the loss of esthetic and recreational values, lowed by drought in the Corn Belt) aggravated this replacement represents a substantial de- these losses, most observers appear willing to cline in wildlife diversity, loss of watershed place a major blame on the farmers’ land se- protection, and the loss of the alternative lection and inattention to erosion control prac- wood (or other) resource. At present, this loss tices. involves a bit over 400 million acres of desig- Soil losses on the additional acreage during 53 nated cropland and will probably increase the 1973-74 season averaged over 6 ton/acre unless crop production efficiency can keep over and above expected losses without pro- pace with the rising demand for food. Also, be- duction. Those lands designated as suffering cause millions of acres of cropland are lost from inadequate conservation treatment lost each year to roadbuilding and urban develop- an average of more than 12 ton/acre above ex- ment, merely the maintenance of the status pected losses. First-year erosion losses are ex- quo demands continued clearing of unman- pected to be lighter than subsequent years be- aged and lightly managed lands for crop pro- cause the root structures of the original cover duction. crops are not totally destroyed by tilling and provide some protection to the soil until they 1973=74: A Case Study in decompose; thus, erosion rates would be ex- Increased Cropland Use54 pected to rise still further unless conservation practices were begun. In 1973, USDA told American farmers that they would be free to plant as many acres of The hardest hit of the agricultural regions wheat, corn, and feed grains as they wished were the Corn Belt (390,000 acres, 15 to 100 during the 1973-74 season. In response, 8.9 ton/acre additional soil loss on the new land), million additional acres were planted and western Great Plains— North Dakota, Mon- harvested during that season: tana, Wyoming, eastern Colorado (325,000 acres, 5 to 40 ton/acre), eastern Great Plains — . 3.6 million acres from grassland, Nebraska, Kansas, South Dakota (260,000 • 0.4 million acres from woodland, and acres, 5 to 55 ton/acre). Great Lakes (195,000 ● 4.9 million acres from idle cropland and acres, 5 to 55 ton/acre), and the southern set-aside land. Coastal Plains of Florida, Georgia, Louisiana, The results of this new agricultural produc- Alabama, and Mississippi (210,000 acres, 5 to tion may provide a basis on which to predict 70 ton/acre). In addition, a number of other producing regions experienced high soil losses ‘2M Clawson, “Forests in the Long Sweep of American His- on the additional acreage. tory, ” Science, VOI 204, j une 15, 1979 5JPotentia/ Crop/and Study, op cit “Adapted from K E Grant, Erosion 1973-74: The Record and High as these soil losses were, however, they the Cha//enge, are not unusual when compared to losses suf- 76 . VOI. II—Energy From Biological Processes

fered by land in continuous production. As ● production practices, and noted previously, many areas that are critically ● type of crop grown. important to U.S. grain production routinely lose soil at rates well above the 5-ton/acre-yr maximum recommended by SCS. Assuming Land Quality and Previous Use that much of the converted land was taken out The land available for growing biomass of relatively nonerosive uses (the 4 million crops consists of cropland that is presently not acres of grassland and woodland, nearly half in intensive use— for instance, land used to the total, would have suffered virtually no grow native hay— and land currently in range, erosive losses if left undisturbed), the erosion forest, or other use that can be converted to experienced on the additional acreage was cropland. Table 30 presents SCS estimates of only slightly worse than the average erosion cropland not currently being utilized to its rates on all U.S. cropland. On the other hand, maximum production potential, and land the lands designated as inadequately pro- available for conversion to cropland in 1977. tected did have much higher erosion than aver- (The acreage “not in intensive use” includes age. The conclusion appears to be that a rapid land where the current use meshes with the increase in land under production will not nec- farmers’ desired mix of livestock and crops and essarily cause proportionately more erosion thus is unlikely to be converted to more inten- than our current experience would lead us to sive use; thus, the table may overestimate the expect, but that conservation planning and acreage available for switching to biomass pro- treatment will be required to keep erosion duction.) Table 31 presents SCS estimates of rates from escalating beyond current rates. the rates of erosion on these lands, by Iand use and capability class. Potential Impacts of Production of The data shows that there is a very substantial Biomass for Energy Feedstocks amount of land available for biomass production that could be cultivated with few environmental Most proposals for using the agricultural sys- problems. For example, table 30 shows well tem to produce energy feedstocks do not con- over 3 million acres of the highest quality template growing and harvesting systems that (class I) land with high and medium conversion appear to be radically different from current potential. Over 10 million acres of high-quality large-scale mechanized food-growing systems class I I (for brief definitions, see table 30) land found in the Corn Belt and other centers of requiring some drainage correction is avail- American agriculture. Proposals centering able. However, there currently is no guarantee around gasohol, for example, assume that at that land for biomass production will be selected least the near-term feedstock (after food for its environmental characteristics. Erosion po- wastes and spoiled grains are used up) will be tential, which is of critical environmental im- corn and other conventional starch or sugar portance, is only one of several characteristics crops. Even the more radical systems—for ex- used by farmers to decide whether to put land ample, tree plantations — can be viewed as var- into production. Characteristics such as con- iations of common agricultural systems. tiguity of land, current ownership, and the cost The key to identifying the impacts of imple- of conversion may be the deciding factors. menting the various approaches to energy According to table 30, farmers currently feedstock production is to identify those dif- have biased their choice of land for row crop ferences from today’s systems that are most cultivation somewhat in favor of the less critical to causing differences in the impacts. erosive lands. Over 11 percent of land in row These differences in impacts primarily depend crops is prime class I land with both high pro- on differences in: ductivity and minimal erosion. In contrast, • quality and previous use of the land, other land uses typically have about 4 or 5 per- Ch. 3—Agriculture ● 77

Table 30.–1977 Cropland and Potential Cropland Erosion Potential (in thousand acres, % of total acreage)

Present cropland not in intensive use (rotation hay and pasture, occasionally Present cropland in intensive use improved/native Potential cropland Class Row crops Close-grown crops hayland) High potential Medium potential 1. Excellent capability, few restrictions...... 23,034 4,471 2,389 2,186 1,412 (11.3) (3.4) (4.2) (5.5) (1 .5) Il. Some limitations, require moderate conservation practices Erosive ...... 45,954 23,463 11,718 10,543 13,921 (22.6) (22.4) (20.8) (26.3) (14.8) Other problems ...... 58,657 22,762 9,855 8,278 10,750 (28.9) (21 .7) (1 7.5) (20,7) (11 .3) Ill. Severe limitations, reduced crop choice and/or special conservation practices required Erosive ...... 28,054 26,997 12,561 7,893 25,142 (13.8) (25.7) (22.3) (19.7) (26.7) Other ...... 27,676 10,811 6,557 4,797 12,703 (13.6) (10.3) (1 1.6) (12.0) (13.4) IV. Severe limitations, more restricted than above Erosive ...... 9,159 9,324 5,701 1,896 11,531 (4.5) (8.9) (10.1) (4.7) (12,3) Other ...... 5,436 2,933 3,154 1,601 7,210 (2.7) (2.8) (5.6) (4.0) (7.6) V-VIII. Generally not suited...... 5,728 345 4,479 2,888 12,248 (2.6) (0.3) (7.9) (7.2) (13.0) Total, ...... 203,243 104,890 56,414 40,082 94,917 Percent of land that is erosive ...... 40.9 57.0 53.1 50.7 53.4

SOURCE 1977S011 Corrsewahon Serwce Naflona/Eros~on /rrventory Eshrrale (Washington, D C SoIl Conservation Service, U S Department of Agriculture, DecemtNr 1978)

Table 31 .–Moan National Erosion Rates by Capability Class and Subclass (rates are in ton/acre-yr)

Potential Class/subclass Row crop Close grown Nonintensive High Medium Class 1. Excellent capability, few restrictions ...... 3.46” 1.75 0.66 0.31 0.35 Class Il. Some limitations, require moderate conservation practices/erosive ...... 6.51 3.67 0.96 0.67 0.71 Class n/other ...... 3.46 2.55 0.43 0.30 0.31 Class Ill. Severe limitations, reduced crop choice and/or special conservation practices required/erosive . . . 12.39 6.62 1.51 1.08 1.28 Class Ill/other...... 3.41 2.51 0.51 0.21 0.28 Class IV. Severe limitations, more restricted than Ill/erosive ...... 17.88 12.20 2.93 2.01 2.28 Class IV/other...... 4.52 1.85 0.45 0.46 0.43 Classes V-VIII. Generally not suited/erosive ...... 46.82 19.61 5.42 2.38 4.15 Classes V-Vlll/other...... 14.26 3.27 0.80 1.51 0.38

SOURCE 1977 Sod Conservaflori Serwce Naf~ona/ Erosion Irwentory Estmate (Washington, DC SoIl Conservation Service, U S Oepaflment of Agriculture, December 1978)

cent of their land classified as class 1. In land towards use of less erosive land is not surpris- quality classes I through IV, 43 percent of the ing. row-cropped acreage is erosive, whereas over 50 percent of every other land use category is Close-grown crop cultivation is considerably erosive. Because row crop cultivation is gener- Iess erosive than row cropping. Apparently in ally the most vulnerable to erosion, this bias response to this, farmers have placed close- 78 ● VOl. II—Energy From Biological Processes

grown crops on lands that are more vulnerable Iems will be significant in adding new lands to to erosion; 60 percent of close-grown cropland intensive agricultural production, but it does acreage is erosion-prone. not appear on a national basis that these prob- Iems will be very much worse than those that It is important to look beyond these overall could be predicted by extrapolating from cur- percentages and examine the percentage of rent erosion rates. land in each land use capability class. The erosivity of lands categorized as E (erosive] by It is possible to estimate quantitatively the SCS appears to be a strong function of the general erosion danger from an expansion of capability class. For example, the average 1977 intensive production by utilizing the data in annual sheet and rill erosion rates on erosive tables 30 and 31 and by making the following croplands in intensive use were estimated to simplifying assumptions: be (from table 31 ): ● The 1977 erosion rate for land under inten- Class I ...... 3.18 ton/acre-yr sive production, for each land capability Class IIE...... 5.55 subclass, is representative of the erosion Class IIIE ...... 9.56 that would occur if additional land in that Class IVE ...... 15.02 subclass were to be put into intensive pro- Class V-VIIIE ...... 34.70 duction. Thus, the erosion danger appears to increase ● Given a desire to place additional land markedly as land capability declines. If the into intensive production, farmers will se- erosive portions of the land with future bio- lect land mainly from cropland not now in mass potential (present cropland not in inten- intensive production and “high potential” sive use and land with switching potential) land, and their selection will be random were skewed towards the lower quality classes, (this is probably a “worst case” assump- then an examination of the overall erosive tion but may not be seriously in error judg- potentials would underestimate the erosion ing from the discussion above). danger presented by massive shifts to intensive ● A mix of row and close-grown biomass cultivation. An examination of table 30 in- crops will be grown, with the mix being dicates that the erosive portions of the present about the same as the 1977 food crop mix. cropland not in intensive use and the high- potential land are somewhat skewed towards Under these conditions, the average erosion the lower quality lands when compared with rate on the new land put into intensive produc- present cropland, but the differences do not tion will be about 7.5 ton/acre-yr. For compari- appear to be substantial. For example, whereas son, the 1977 erosion rate on intensively culti- 53 percent of erodible land in intensive use is vated lands was 6.26 ton/acre-yr. In other class I I I E or below, 60 percent of erodible land words, erosion from additional acres devoted with high biomass potential is in this erosivity to growing biomass crops may be about 20 per- range. cent worse than similar acreages of food crops in production today (this assumes no Govern- The surprising implication of the statistics ment action to improve land selection). Given presented in table 30 is that the land available the substantial uncertainties in this estimate, for agricultural biomass production is not radical- the 20-percent differential is well within the ly different in its erosion qualities from land cur- range of possible error. It is, however, consist- rently being utilized for intensive agricultural pro- ent with what is known about agricultural land duction. Although clearly some selection has selection and the quality of available (but un- been made in utilizing the best lands and keep- developed) farmland. ing idle the worst, this selection process appears to have been skewed by other physical Because land quality is affected by net rath- attributes and economic and social factors er than gross erosion — i.e., by the difference that are as important or more important than between erosion and soil replacement–the ef- erosion potential. It appears that erosion prob- fect on the land of relatively small changes in Ch. 3—Agriculture ● 79 erosion rates may be greater than would be ap- no guarantee that forests—which make up parent at first glance. For example, if the aver- about one-quarter of the high- and medium-po- age topsoil replacement rate is 5 ton/acre-y r,* tential cropland55— will not be cleared in sig- the 7.5 ton/acre-yr biomass erosion rate yields nificant quantities if large-scale conversion to a 2.5 ton/acre-yr net soil loss, versus 1.26 biomass crop production occurs. ton/acre-yr net loss from food production. Thus, while a large-scale expansion of acreage Production Practices for biomass production may have effects on A variety of practices are available to con- waterways that are similar in magnitude to the trol the erosion and other impacts of farming. effects of present intensive agriculture, this These range from crop rotation to conserva- acreage may lose its topsoil layer at twice the tion tillage to scouting for pest infestations. rate of current agricultural land. However, it Table 32 provides a partial list of these prac- should be noted that the rate of loss is (on the average) fairly low. Table 32.–Agricultural Production Practices Aside from new biomass cropland’s capabili- That Reduce Environmental Impacts ty to resist erosion and its productive poten- Runoff and erosion control tial, an important factor determining the en- Contour farming or contour stripcropping vironmental impact of the conversion to inten- Terraces and grass waterways sive production is the nature of the previous Minimum tillage and no-till Cover crops land use. For example, the conversion of land Reducing fall plowing in rotation hay and pasture to intensive crop Reducing chemical pollution production would clearly be valued differently Scouting (monitoring for pest problems) from a conversion from forest. Because differ- Disease- and insect-resistant crops Crop rotation ent groups value alternative land uses differ- Integrated pest management ently, it is difficult to place more or less weight Soil analysis for detecting nutrient deficiencies on the conversion of one land use relative to Nitrogen-fixing crops Improved fertilizer and pesticide placement, timing, and amount another. It seems likely, however, that most en- Improving irrigation efficiency-trickle irrigation, etc. vironmentally oriented groups would prefer to Incorporating surface applications into soil

see the conversion of lands that are manmade SOURCE Office of Technology Assessment monoculture (e. g., improved haylands) before more natural and diverse ecosystems were tices. Their future use will play a critical role in converted. determining the environmental impacts of bio- The cost of conversion will play an impor- mass energy production. tant role in determining which lands will be The availability of these controls should not chosen. At the present time, conversion of pas- be confused with the probability that impacts tureland and hayland is likely to be less expen- will not occur. in fact, it is unwise to assume sive than conversion of forest, and land con- that the use of many of the practices listed in versions may be expected to be skewed away table 32 will be widespread. There are a num- from forests. Least expensive of all to convert ber of reasons for this. are lands currently in set-aside, and these are likely to be the first to be taken. The cost of First, the costs of the controls may consid- forest conversion may, however, be lowered erably exceed the farmer’s perceived benefits. significantly if the demand for wood-for-ener- The effects of erosion on water quality are gy rises with the demand for energy crops (be- largely “external” effects; although the farmer cause the value of the now-worthless cull may benefit from the control efforts of others, wood and slash can be traded off against clear- he is unlikely to benefit from any water quality ing and site preparation costs). Thus, there is improvements caused by his own efforts. This

‘This IS almost certainly very optlmlstlc, but SCS guidelines def Ine 5 ton/acre-yr as an acceptable rate for many lands 55 Natjona/ ~rosjon Inventory Estimate, op c It 80 ● Vol. 11—Energy From Biological Processes

problem of “external” benefits is endemic to reduced tillage. The net effect on the environ- American agricultural practices. It is, in fact, ment is not entirely clear because a large merely one aspect of the “tragedy of the com- source of pesticide entry into surface waters — mons” that hinders voluntary environmental adsorption on soil particles and transport in control in virtually all of man’s activities. Also, runoff — is considerably reduced by the con- any success in delaying or preventing produc- trols, but EPA has identified increased pesti- tivity declines from erosion effects may be cide use with conservation tillage as a signifi- masked by improvements in other production cant problem .57 Tables 33 and 34 identify in practices and in any case would be very long greater detail the environmental tradeoffs in- term in nature. The farmer must balance these volved in erosion controls. benefits against very high erosion control Third, some of these controls may appear to costs. SCS has examined the effects on farm be incompatible with the present agricultural production costs of requiring reductions in system and may not be accepted by farmers. current erosion rates on croplands. For exam- For example, the use of nitrogen-fixing crops, ple, requiring a 10-percent reduction in each of cover crops, and crop rotations conflict with 105 producing areas would raise corn produc- today’s large-scale, highly mechanized, chemi- tion costs by $0.07/bu in 1985. Requiring all cal-oriented farming although they were wide- acreage to conform to a maximum allowable ly practiced in the past. Although some scien- erosion rate of 10 ton/acre-yr (twice the “no tists argue that the economic advantages of productivity loss” rate) would cost $0.31/bu or present methods will evaporate (or have al- a 16-percent increase over the projected 1985 ready evaporated) in the face of rising prices cost without controls. Further constraints for energy and energy-intensive agricultural could raise costs astronomically (a 5-ton/acre- chemicals, and that the long-term environmen- yr constraint leads to a $23.70/bu production tal viability of the methods is questionable, the cost) because heroic efforts must be made on relative advantages and disadvantages of the some acreage in order to meet the constraints. 56 Although these estimates are sharply present system and its alternatives are a subject of intense controversy in the agricultural dependent on a number of critical assump- community— with defense of the present sys- tions (e.g., the role of Federal soil conservation tem having the upper hand at present. It ap- assistance is ignored), they demonstrate the pears virtually certain that in the absense of large potential cost (and price) increases that Government intervention the provision of erosion control requirements could cause. feedstocks for energy production will rely pri- Second, there are substantive scientific dis- marily on a mechanized, chemical-oriented agreements about the actual environmental philosophy modified only by any economic benefits achieved by these controls. Some of pressures arising from increases in energy the controls may reduce one environmental prices. Any substantive changes from this phi- impact at the expense of increasing others. A losophy would represent essentially a revolu- primary example of this is the effect of some tion from established practice and would be erosion controls— reductions in fall plowing unlikely because the present system has clear- and conservation tillage —on pesticide use. ly succeeded in providing a reliable supply of These controls leave crop residues on the sur- food at (comparatively) moderate prices. face, and the residues in turn act to break the force of raindrops on the soil and drastically Crop Types decrease erosion and runoff. Because the resi- The environmental impacts of growing and dues harbor plant pathogens and insect pests, harvesting agricultural crops for energy will pesticide requirements will go up sharply. vary strongly with the type of crop grown, Also, increased applications of herbicides are since different crops have different fertilizer used for weed control to compensate for the and pesticide requirements, water needs, soil ’66. English, lowa State University, personal communication, June 15,1979. 57 Environmenta/ /mp/ications of Trends, op. cit. Table 33.-Environmental Pollution Effects of Agricultural Conservation Practices

Pollutant changes m Pollutant changes m media: surface water media: ground water– Pollutant changes m Pollutant changes in Extensiveness Resource use sediment Nutrients Pesticides nutrients–pesticides media: soil media: air Contour farming/contour stripcropping Acreage of crops farmed Fertizer and herbicide Sediment loss can be re- Nutrients associated with Pesticlie reductions will Loss of nutrients and Erosion losses can be re- Pesticide losses through on the contour or strip- use remain constant duced substantially on sediment will be re- be less than that for pesticides through duced up to 50% with volatilization will de- cropped decreased 25% Insecticide use will re- moderate slopes, but duced, but reductions nutrients since a great- ground water will re- average reductions of crease if they are incor- between 1964 and main constant to very much less on steep may be proportional to er amount of pesticide main constant or de- 12Y0 (see conclusions porated into the soil by 1969 and continued to slight increases. slopes, Reductions up the amount of sediment IS lost through surface crease slightly. How- on sediment). mechanical means decrease slightly to to 50% are possible, lost, water than bound to ever the amount of N 1976, Contour farming but average reductions sediment. leached IS small com- is more widely used in will be about 35%. pared to amount that nonirrigated crop pro- Contour stripcropping can be lost in runoff duction than m irrigated can reduce sediment and loss of pesticides to crop production. losses more than con- ground water IS minor tour alone. (Note: re- with proper application search shows substan- rates. tial loss can occur with contour watersheds with some soil types, with long slopes and/or with steep slopes. ) Terraces and grass waterways Terraces and-grass wa- Fertilizer, herbicide, and Substantial reductions in Reductions in nitrates Reduction of pesticide N m ground water may Substantial reductions m No change. terways are not impor- insecticide use is not sediment and runoff and phosphates are ex- residues in surface be reduced, based on erosion can result. tant in irrigated produc- expected to increase can usually be expected with decreased water could be substan- limited research data. tion, but are Important (fertilizer could increase pected. soil loss and surface tial with terrace sys- Leaching of pesticides IS for nonirrigated crops. if production per runoff. Reductions terns, since both sur- not likely to result in However, only 6% of all cropped acre is ex- could be substantial face runoff and soil loss significant loss with acres in 1969 had ter- pected to increase to with some soils and are reduced. normal applications races The acres with compensate for land cropping systems. rates terraces in 1976 could taken out of produc- have increased or de- tion). However, terrace creased slightly. practices will not require more fertilizers. Costs and maintenance increase for terraces,

SOURCE U S Enwronmental Protection Agency, ErWlrOnmenfal /mPllCaf10/ts of Trencfslfr ~gricu/(ure and Si/v~cu/ture, Vo/urne // &wlronrnerrla/Ef feck of Trends, EPA-600/3-78-102, December 1978 Table 33.–Environmental Pollution Effects of Agricultural Conservation Practices-continued

Pollutant changes m Pollutant changes in media: surface water media: ground water— Pollutant changes m Extensiveness Resource use sediment Nutrients Pesticides nutrients—pesticides media: soil Conservation tillage; no-till Approximately 2.6% of Fertilizer and herbicide Sediment reductions of While Iarqe soil loss re- Effect of no-till on pesti- Nititrates in ground water Erosion losses will be de- With some pesticides, in- all cropped land was use increases by 15%, 50 to 90% will result. ductions will tend to re- cide losses is not well will show no change to creased 50 to 90%. increased volatilization no-till in 1977. While insecticide use by 11%. duce nutrient losses, documented. Loss to slight increases. Pesti- Crop residues will in- will occur with surface this Practice is expected “An estimated 5 million fertilizer use will in- surface water is greater cide loss to ground crease which may result applications. The vapor to increase to Iimited acres of land could be crease by 15%. There when the compound is water will not be signifi- in increased N loss to pressure, molecular use in 2010, current shifted to crop produc- will probably still tend surface applied and not cantly changed with no- the soil or available for weight, and other prop- projections (up to 55% tion with no-till and re- to be reductions in total incorporated in the soil, till practices. runoff. Additionally, erties of a pesticide will of crops under no-till in duced-till methods. La- nutrient loss, but re- and 11‘Yo more insecti- residues may provide a determine the extent of 2010) seem high. Ex- bor costs are reduced, duction will not be pro- cides and 15% more hiding place for pests vaporization. tensiveness may only More water will be con- portional to reductions herbicides will be used and increase the in- be 10 to 20% in 2010. served with no-till, as in soil loss. N content for no-till. While reduc- cidence of pests. much as 2 inches per of soil may also in- tions of pesticides in year. crease from weathering surface water could oc- of crop residues. cur, current research does not prove this. In- creased use and surface application, even with reduced soil loss with no-till, could even cause slight increases in pesticide losses. Conservation tillage; reduced tillage In 1977, an estimated Fertilizer use will in- Sediment will be reduced There will probably be Effect of reduced tillage Nitrates in ground water Erosion losses decrease Surface applications of 58.8 million acres (19% crease slightly. Herbi- an average Of 14%. Re- reductions in total nutri- on pesticide loss IS not will show no change to an estimated 14%. some pesticides types of total cropped acres) cide use is up (0.6%) duced tillage is less ef - ent loss to surface wa- well documented. Loss slight increases. Pesti- Wind erosion losses will leads to increased vola- will be reduced tilled. and insecticide use in- fective than no-till in ter, but reduction will to surface water is cide levels in ground also decrease slightly. tilization losses. The An additional 40 million creases by 8.6%. An controlling soil loss, not be proportional to greater when a pesti- water will not be signifi- Crop residues increase, vapor pressure, molecu- acres will be classified estimated 5 million reductions in soil loss cide is surface applied cantly changed with re- which lead to increased lar weight, and other as less tilled. Less till acres of land will be (14%), and total pesticide use duced tillage. N available to the soil chemical properties of a includes chisel plowing, shifted to crop produc- IS 9% greater for refor leaching and runoff. pesticide will determine disking once instead of tion with reduced and duced till. While reduc- Residues on soil also the extent of vaporiza- twice, and planting in no-tillage methods. La- tions of pesticides in increase the incidence tion. rough ground. bor output will de- surface water could oc- of pests. In 2010, a total of 40% crease. Energy to plant cur, there is not of all cropland may be crops decreases, but enough research data to classified as reduced increased energy will be support this. tilled. used in manufacture of increased fertilizers and insecticides. Some soil moisture will be con- served with reduced tillage.

SOURCE U S. Enwronmental ProtectIon Agency, Envlmnmen?a/ /mp/icaf/onso( Trends lnAgncu/ture and Sl/vlcu/ture, Vo/urrre //: Environmenfa/ Eflecfsot Trends, EPA-600/3-78-102, Oecember 1978. Ch. 3—Agriculture ● 83

Table 34. -EcoIogical Effects of Agricultural Conservation Practices

Contour farming/contour stripcropping Extensiveness of contouring in 1985 (over 1976 use) will be low, but will increase by 2010. Beneficial aquatic effects result from decreased turbidity and pesticide residues in surface water. Species diversity will also increase in the aquatic ecosystem. Decreased erosion and retention of soil nutrient cycles will have long-term beneficial terrestrial effects. Since pesticide residues at current levels in drinking water are not known to be a human health hazard, reduction of pesticide residues will have no significant human health effects. However, if pesticide residues are later determined to be dangerous at current levels, then human health effects would be beneficial. Terraces and grass waterways Terraces are more effective than contouring in reducing pollutants, but extensiveness of use is tower for terraces. Aquatic effects are decreased turbidity, increased species diversity, and decreased pesticide residues. Terrestrial effects are beneficial, resulting from increased vegetation on terraces and grass waterways, increased diversity of wildlife, and more pathways for animal populations to travel. Valuable topsoil will also be retained. Based on present knowledge, there is no known human health effect. Decreased sediment in water might result in an unpleasant taste or odor in drinking water. Reduced tillage Reduced tillage (with crop residues remaining) is less effective than no-till in reducing soil loss, but extensiveness of reduced tillage will be greater. There- fore, the intensity of ecological effects are comparable for the two practices. Sediment reductions will reduce turbidity and increase species diversity. However, the potential for increased pesticide residues in surface water could have adverse effects on the aquatic ecosystem. Crop residues remaining on the soil and decreased soil loss are beneficial to the terrestrial system, but increased pesticide use will have adverse effects on nontarget organisms. Human health effects will not be significant. No-till Aquatic and terrestrial effects are both beneficial and adverse. Aquatic systems will benefit from reduced turbidity and increased species diversity. How- ever, pesticide residues in surface water could potentially be increased with no-till and create adverse effects in the aquatic ecosystem. Increased pesticide use can also have adverse effects on nontarget terrestrial life. Retention of crop residues and reductions in erosion will have beneficial terrestrial effects. Human health effects will not be significant since pesticide residue in surface water should still be within safety limits even if they increase slightly with no-till.

SOURCE U S Enwronmental Protection Agency, Hrwromnerrfal lrnpllca(~orrs of Trerrds m Agrlcullwe arrd Sf/wmdLve, I/ohmre // Errvrorrrnerrta/ Hfeck of Trends, EPA-600/3-78-102, Oecember 1978 preparation methods, harvesting times, and tion) would be expected to have an aver- other factors that may potentially affect im- age (sheet and gully) erosion rate of about pact. Some of the more important crop-deter- 7.5 ton/acre-yr compared with about 6.3 mined factors are: ton/acre-yr for food production. If the entire biomass crop were a row crop (e. g., ● Annual or perennial, — Perennial crops corn for large-scale alcohol production), (trees, sugarcane, perennial grasses, etc.) the average erosion rate from the biomass offer a substantive environmental advan- acreage is estimated to be 9.3 ton/acre- tage over annuals because their roots and yr— almost 50 percent higher than the ero- unharvested top growth protect the soil sion rate from food production. from erosion year round, while annuals of- ● Water requirements. — High irrigation wa- fer protection only during the growing ter use means greater competition for wa- season and require seasonal tilling (unless ter among competing uses, greater draw- no-till is used) and planting. down of streams and consequent loss of ● Row or close-grown crops. — Row crop cul- assimilative capacity, potential for entry tivation is generally more erosive than cul- of more salts into surface and ground wa- tivation of close-grown crops. For exam- ters, depletion of aquifers (ground water ple, the average erosion rates of close- mining), and energy use for pumping. grown crops are significantly lower than There are substantial differences in water those of row crops in every land capabili- consumption among different crops. For ty class and subclass shown in table 34. In example, irrigation requirements for crops general, the rates of the close-grown crops 58 in Arizona during a dry year are: appear to be about half those of the row crops. Water use, acre-ft/ton of crop In the previous calculation of the ex- Wheat...... 0.9 pected average erosion rates from new Oats ...... 1.6 Barley...... 1.3 biomass production, a mix of row and Alfalfa ...... 0.7 close-grown biomass crops (in the same proportion as existed in 1977 food produc- sacon~~rvat;on NeecjS Inventory, OP. Cit 84 . Vol. II—Energy From Biological Processes

Most discussions of biomass energy crop requirements for very high levels of assume that irrigation generally will not nitrogen may be an environmental advan- be used in growing feedstocks. However, tage; some high-nitrogen crops are com- an extension of the types of irrigation patible with land disposal of sewage water subsidies now available to Western sludge and effluents and thus can be an farmers, however unlikely, could lead to important component of urban sewage such use. treatment strategy. ● —The ability to utilize Soil requirements. ● Yield. — Because yield per acre determines marginal lands can avoid the problem of the amount of land necessary to produce competition with food production that is a unit of energy, it is one of the most im- a major environmental and social/eco- portant factors determining impact. Meas- nomic issue in evaluating biomass fuels. urements of input requirements (water, As discussed elsewhere in this chapter, fertilizer, pesticides, etc.) and measurable however, the potential for high biomass damages (such as erosion) on a “per acre” yields under marginal soil, temperature, basis are inadequate measures of relative and water conditions has been exagger- environmental impact because of the ated. large variation in biomass yields from ● Pesticide requirements. —The importance crop to crop. For example, corn is widely of reducing pesticide applications is a perceived as an extremely energy- and wa- matter of considerable controversy. How- ter-intensive crop, but its very high yields ever, crops that have low pesticide re- essentially cancel its high “per acre” fer- quirements will be perceived as more en- tilizer, pesticide, and water needs; it is, in vironmentally benign. In some instances, fact, a relatively average crop on an “en- present pesticide use may be a poor in- ergy per ton of product” basis. dicator of future requirements for energy crops because cropping practices and The importance of these factors in determin- land characteristics may be altered signifi- ing environmental impacts is extremely site cantly in going to a crops-for-energy sys- and region specific. For example, water re- tem. For example, regulatory restrictions quirements clearly are more important in the on soil erosion could force virtually uni- arid West than in the wet Southeast, while fac- versal use of conservation tillage and con- tors affecting sheet and rill erosion potential sequent increases in herbicide and (to a are more or less important in the reverse order. lesser extent) insecticide applications. The Much of the data needed to assess the differ- lack of esthetic requirements for biomass ent potential crops are not available, and thus feedstocks might also lead to some de- it is premature to suggest which crops would crease in pesticide requirements, but this be the most environmentally benign in each re- effect may be small because minor insect gion or subregion. There are sufficient data, damage can lead to further damage by however, to draw some rough sketches of some fungal and viral infections (especially dur- of the possible advantages or disadvantages of ing storage). Finally, although pesticide re- several of the suggested biomass crops. quirements for grasslands currently are very low, pest problems conceivably may Corn has been most often mentioned as the accelerate if productivity is pushed by ex- primary candidate for an ethanol feedstock. It panded use of fertilizers. is an annual row crop and thus a major contrib- ● Fertilizer requirements. – In general, high utor to erosion, but much of the land on which fertilizer requirements are an environmen- it is grown is relatively flat, a factor that limits tal cost because of the energy used to pro- the erosion rate. Corn’s high yield rate—cur- duce the fertilizer and the nutrient runoff rently about 100 bu/acre, or about 260 gal/acre that results from applications. However, of ethanol —will minimize the land use impact Ch. 3—Agriculture ● 85 of additional production, although yields on level of displacement of the most productive new lands will not be as high as the current av- ecosystems. erage, and the land displaced would be of high quality. Sugarcane has been suggested as a biomass crop for alcohol production in Hawaii and the Because the protein-rich residue from the Gulf Coast. Because its cellulosic content is fermentation (ethanol producing) process is a high enough to supply all of the heat energy substitute (although not necessarily a perfect necessary to ferment the sugar and distill alco- one) for soybean meal in cattle feed, switching hol from it, no coal or other fossil fuel use existing cropland from soybean to corn pro- would be necessary to power the system. duction may allow large quantities of ethanol Sugarcane requires high-quality land and thus to be produced using far less acreage than may displace particularly valuable alternative would be needed if corn for ethanol produc- land uses. tion were planted only on new acreage. As discussed in the section on “Energy Potential Perennial grasses can be supplied in large From Conventional Crops,” corn’s effective quantities by increasing yields on present acre- yield per acre of new land could grow by over age with more intensive harvesting and fertili- 300 percent (i.e., about three-fourths of the zation; the present average yield is 1½ to 2 corn used for ethanol would be grown on land ton/acre, and this can be increased to 3 to 5 formerly planted in soybeans with no loss in tons. Because perennials provide excellent ero- national food and feed values) as long as the sion control, and because no additional acre- soybean meal market remained unsaturated. age would have to be converted from alterna- Significant uncertainties concerning the corn tive uses, the environmental impact of a grass- residue’s nutritive value, potential corn yields based biomass strategy should be far less than on soybean land, soybean market response, that of a strategy based on annual crops. Envi- and other factors must be overcome, however, ronmental impacts of some significance could before this crop-switching scenario can be ac- occur because of the expanded use of fertilizer cepted as valid. In the absence of the neces- (150 lb N, 30 to 50 lb P2O 5, 80 to 150 lb K2 for sary research, the higher estimate of new land an incremental production of 78 gal of ethanol required for each gallon of ethanol produced on each acre) and pesticides. Recovery of should be used as a pessimistic measure of po- added fertilizer is very high for grasses, how- tential impact. At low levels of production, the ever, so the potential for water pollution will more optimistic, lower acreage requirements be less than for annual crops. Also, there is are likely to be accurate, but the requirements uncertainty about changes in susceptibility to may increase as production increases. Above 2 disease and insect damage because of the in- billion to 7 billion gal of ethanol produced an- tensification of production, and substantial nually, feed markets would be saturated even new use of pesticides conceivably could be re- under the most optimistic assumptions and ad- quired. Finally, the frequent harvesting and ditional ethanol production would require greater use of chemicals may disrupt the popu- cropland conversion at the higher rate. lations of wildlife that now flourish in the less intensively maintained grasslands. Sweet sorghum has been praised as a crop of high biomass potential for fermentation and Trees may be grown plantation-style and har- alcohol production. Although ethanol yields of vested by coppicing to supply significant quan- 260 to 530 gal/acre have been projected, these tities of biomass. A carefully designed tree projections are based on minimal — and clearly plantation should have few problems of ero- inadequate — experience. However, these high sion unless cultivation is practiced (which ap- yields, if confirmed, would limit displacement pears unlikely); however, harvesting may con- of alternative land uses. Sweet sorghum may ceivably create an erosion problem unless low- be more tolerant of marginal growing condi- bearing-pressure machines are used to avoid tions than corn, which could lead to a lower damaging the soil. Tree plantations present 86 . Vol. II—Energy From Biological Proceses basically the same ecological problems as do from weeds– and consequently larger herbi- agricultural monocultures — higher potential cide requirements –but the sheltering effect for disease attack and displacement of alterna- of the tree canopy and the greater ability of tive ecosystems. The spacing necessary for tree some tree species to compete for water may growth may also allow greater competition counterbalance this effect.

Environmental Impacts of Harvesting Agricultural Residues

The residues from agricultural production Table 35.–Environmental impacts of Plant Residue Removal have a number of significant effects –benefi- Water cial or otherwise—when left on the land. Un- ● Increased erosion and flow of sediments into surface waters if restric- derstanding these effects is critical to under- tions on removal are not observed, causing increased turbidity, standing the potential environmental impacts obstruction of streams, filling of reservoirs, destruction of aquatic habitat, increase of flood potential; under circumstances where con- of the collection and use of these residues as servation tillage is encouraged by removal of a portion of the residues, an energy feedstock. erosion and its consequences will decrease, ● Increased use of herbicides and possible increased flow into surface The effects of residues left on the land in- and ground waters if conservation tillage is required for erosion conclude (table 35): trol; in some situations, removal of a portion of the residues would increase herbicide efficiency and greater use may not occur. ● Control of wind and water erosion. – ● Increased flow of nutrients if more runoff results from decreased water retention of soil and greater erosivity of soil; if more fertilizer is applied Retention of residues as a surface cover is to compensate for nutrient loss, flow of nutrients will change but the a major erosion control mechanism on net affect is not certain. erosion-prone lands. For example, residue Air retention on land that is conventionally ● Dust from decreased cover on land, operation of residue harvesting equipment (unless integrated operation). tilled (i.e., plow-disk-harrow) can cut ero- ● Added herbicides from aerial spraying or as a component of dust. 59 sion in half. ● Decreased insecticides, fungicides. ● Retention of plant nutrients. – Residues ● Reduction in pollution from open-burning of residues, where formerly practiced. from the nine leading crops in the United Land States contain about 40, 10, and 80 per- ● Erosion and loss of topsoil, degrading of productivity if restrictions on cent as much nitrogen, phosphorus, and removal are not observed; the opposite, positive effect if conservation potassium, respectively, as in total fertil- tillage is encouraged by residue removal. ● 60 Decrease in water retention capabilities of land, increased flooding izer use in U.S. agriculture. potential if restrictions are not observed. ● Enhanced retention of water by soils and ● Depletion of nutrients and organic matter from soil (nutrients may easi- maintenance of ability of soil surfaces to ly be replaced), Other allow water infiltration. • Reduction in plant diseases and pests (if lowering of soil organic mat- ● Maintenance of organic matter levels (nec- ter does not adversely affect this factor) because residues can harbor essary to maintain soil structure, ion ex- plant pathogens.

change capacity, water retention proper- SOURCE: Office of Technology Assessment. ties) in soils. –Croplands in the United States have lost major portions of their or- ● A number of negative effects depending ganic content. Reductions (in North Cen- on the type of crop and amount of resi- tral and Great Plains soils) of one-half to due–” poor seed germination, stand re- two-thirds of what was present under na- duction, phytotoxic effects, nonuniform tive grassland have been cited.61 Reten- moisture distribution, immobilization of tion of crop residues is a critical factor in nitrogen in a form unavailable to plants, maintaining organic matter levels. and increased insect and weed problems."62 In all cases, the residues harbor “W. E. Larson, et al., “Residues for Soil Conservation, ” paper crop pests; this can be a particularly sig- No. 9818, Science Journal Series, AR S-USDA, 1978. ‘“I bid. “Ibid. 6z/mprov;ng Soils With Organic Wastes, OP. cit. Ch. 3—Agriculture . 87

nificant problem if single cropping is prac- thus the erosion rate that will maintain produc- ticed (the same crop is grown in consecu- tivity over the very long term, it seems likely tive years). Because the residues shield the that errors in these computations will not soil, they may hinder soil warming and de- cause significant harm as long as SCS main- lay spring planting (causing reduced yields tains its monitoring efforts at the current level. in corn). The key to preventing significant environ- When the problems associated with crop res- mental damage while harvesting large quanti- idues outweigh the benefits, farmers will phys- ties of residues is for the agricultural system to ically remove the residue (this practice is nec- act in accordance with USDA’s knowledge. essary in rice cultivation) or plow it under in The discussion of the impacts of U.S. agricul- the fall (a common practice in the Corn Belt). ture presented previously seems to indicate a The collected residues may be burned, al- willingness among farmers to ignore warnings though they have alternative uses such as live- about using erosive practices or cultivating stock bedding. Where removal is normally fragile land, in order to gain short-term bene- practiced, use of the residues as an energy fits. In the absense of additional constraints, a feedstock is at worst environmentally benign significant number of farmers might be willing and possibly beneficial (if air pollution from to remove their crop residues even when ad- open burning is prevented). Because fall plow- verse erosion effects would occur. (interesting negates much of the residues’ value as an ingly enough, some farmers may ignore USDA erosion control, collection of a portion of the with the opposite effect—they may be reluc- residue is usually considered benign (full tant to remove any residues because of their removal may affect soil organic content, the fear of erosion and other negative conse- importance of which is somewhat in debate). quences). Under these circumstances, the es- Because an excess of residue may inhibit the tablishment of a market for crop residues effectiveness of herbicide treatments — espe- could result in additional erosion from crop- cially preemergence and preplant treatments Iands that cannot afford it and add to the al- — and also leave large numbers of weed seeds ready significant sediment burden on surface near the soil surface, removal of a portion of waters caused by current farming practices. the residues on land where they are in excess Although the negative effects of any in- may promote the use of reduced tillage by crease in erosion are straightforward, other ef- allowing more effective chemical weed con- fects that have been associated with residue trol, and thus be considered environmentally removal are more ambiguous. For example, the beneficial. removal of pIant nutrients in the residues may When residues are normally left on the soil be compensated for by the return of the con- surface as an erosion control, their removal po- version process byproducts or by chemical fer- tentially may be harmful. However, where sub- tilizers (both of which may have some adverse stantial quantities of residue are produced on effects on water quality). The removal of or- flat, nonerosive soils, a portion of these resi- ganic content has been identified as a signifi- dues may be removed without significantly af- cant impact63 and soil scientists have long fecting erosion rates. SCS and the Science and thought that soil organic content is a critical Education Administration –Agricultural Re- variable of the health of the agricultural eco- search have sponsored extensive research de- system (e. g., increasing the organic content of signed to compute the effects of residue re- soils can stimulate the growth and activity of moval practices and other practices on soil soil micro-organisms that compete with pIant erosion. USDA believes that it can identify the pathogens). However, despite a variety of pa- quantity of residues that can be safely re- pers in the agronomy literature that treat yield moved from agricultural lands in all parts of the United States. Although controversy exists over the rate of creation of new topsoil, and ‘) Pimentel, et al., op cit. 88 . Vol. II—Energy From Biological Processes

as a function of soil carbon level, there is insuf- idue will leave the same amount of organic ficient experimental evidence to establish that material as would have occurred 25 years ago any significant effects on crop yields would oc- if all of the residue had been left on the land. cur. Also, the much higher yields of today’s ag- This is an area that clearly deserves further re- riculture means that removal of half of the res- search.

Needs

Considerable research has been and is di- ● Various crop-switching possibilities that rected at improving agriculture for food, feed, involve fuel production should be investi- and materials production. While much of this gated further to determine the extent to research is applicable to energy production, which they can provide fuels and the tra- the specific goal of producing various types of ditional products from agriculture with-. energy crops has not been adequately ad- out expanding the quantity of cropland dressed. Changing the emphasis to energy or cultivated. The extent to which the corn- energy and food production and the environ- soybean switch actually takes place mental concerns with agriculture suggest sev- should be studied, as should novel possi- eral R&D problems. Some examples are listed bilities such as sugar beets used for ani- below. mal fodder. Included in this should be in- vestigations of the effect of substituting ● A wide variety of crops that are not used current feed rations with varying amounts as food or feed crops could, potentially, of forage-distillers’ grain, forage-corn glu- be good bioenergy crops. The promising ten mixtures, and other feeds that may be varieties should be developed. From a the- involved in the crop-switching schemes. oretical point of view, grasses appear to ● Large-scale biomass development will re- be promising candidates for high biomass quire the placement of millions of acres producers and on marginal cropland (see of land — now in low-intensity agriculture ch. 4): and arid Iand and saline tolerant (e.g., pasture), forest, or other uses– into crops may enable the economic use of intensive production, coupled in many lands and water supplies that are other- cases with very high rates of removal of wise unsuited for agriculture. organic matter. Environmental R&D that ● Food and feed crops are usually quite spe- should accompany, and preferably pre- cific as to their use. Corn, for example, is cede, such development includes: not interchangeable with wheat. Many dif- –further investigation of long-term ef- ferent types of crops, however, can pro- fects of reduction in soil organic mat- duce the same or interchangeable bio- ter, mass fuels. Consequently, extensive com- –determination of pesticide require- parative studies between various crops ments for high-yield grasses in intensive are needed to determine the promising production, bioenergy crops for the various soil types — intensification of breeding programs and climates. for insect/disease-resistant strains of ● If both the residues and the grain can be crops with high biomass potential, sold, then the optimum plant may not be –determination of economically opti- the one that produces the most grain. mum strategies for minimization of soil Farming practices and hybrids that can erosion, and change the relative proportions of grain to –development of effective/ programs to residue in the plant while maintaining a improve farmer (environmental) behav- high overall yield should be investigated. ior. Chapter 4 UNCONVENTIONAL BIOMASS PRODUCTION Chapter 4.— UNCONVENTIONAL BIOMASS PRODUCTION

Page Introduction ...... 91 TABLES Genetics ...... 91 Page Crop Yields .*...... 92 36. Optimistic Future Average Crop Yields for Unconventional land-Based Crops...... 95 Plants Under Large-Scale Production ...... 94 Lignocellulose Crops ...... 95 37. incomplete List of Candidate Vegetable Oil and Hydrocarbon Crops . . . . . 96 Unconventional Bioenergy Crops ...... 95 Starch and Sugar Crops ...... 97 38. Optimistic Cost Estimates for General Aspects ...... 97 Unconventional Crops...... 97 Aquiculture ...... 98 Mariculture ● ..********...... *.***** 101 Other Unconventional Approaches ...... 105 Multiple Cropping ...... 105 Chemical Inoculation...... 105 Energy Farms ...... 105 FIGURE Biophotolysis...... 106 Inducing Nitrogen Fixation in Plants ...... 107 Page Greenhouses ...... 108 13. Macrocystis Pyrifera ...... 102 Chapter 4 UNCONVENTIONAL BIOMASS PRODUCTION

Introduction

A number of unconventional approaches to istics, however, can aid in comparing the vari- biomass energy production have been pro- ous possible types of energy crops. posed. Several nontraditional crops that pro- The general aspects of farming, plant duce vegetable oils, hydrocarbons, and other growth, and the efficiency of photosynthesis chemicals or cellulosic material are under in- are considered in chapter 3. Since future crop vestigation. Both freshwater and saltwater yields will depend on these factors and on the plants are being considered, and various other development of hybrids for energy production, approaches to biomass fuel production are the possibilities for genetic improvements are being examined. A common feature to all of considered here. Following this, crop yields these approaches is that the full potential of and various unconventional bioenergy crops individual plants proposed as fuel-producers and approaches to farming them are discussed. cannot be fully assessed without further R&D. A description of some general plant character-

Genetics

There are two major areas of genetics. The characteristics into specialized cells such as first, which plant breeders have used most ef- the ability to produce insulin,1 and 3) improve- fectively to date is the classical Mendelian ap- ments in photosynthetic efficiency, plant proach (introduced by Gregor Mendel in the growth, and crop yields. The complexity of the 19th century). It involves selecting and cross- tasks increases greatly as one goes from 1) to breeding those plants with desired characteris- 3), as described below. tics (e.g., biomass yield, grain yield, pest resistance). The process is continued through The first type involves subjecting single cells each succeeding generation until a hybrid, or to chemicals or radiation that cause mutations in the celIs’ genes. The way these mutations oc- particularly favorable strain, is isolated. cur is not well understood and the effects are Strains with unique and desirable properties are often crossbred to produce hybrids that generally unpredictable. The result is to in- outperform the parents. Hybrid corn is an ex- crease the diversity of cell types over what oc- ample. This technique is limited, however, by curs naturally; and in favorable cases one may produce a cell that performs a particular func- the variability of characteristics that exist tion “better” than naturally occurring cells. naturally in plants or mutations that occur This method has been applied successfully to spontaneously during breeding. One can isolate the best, but one cannot produce better the production of antibiotics, in biological 2 and in increasing the than nature provides. conversion processes, tolerance of plants to certain diseases; but it is The second approach to genetics, molecular generally a “hit and miss” proposition. genetics, is a recent development that involves manipulating the genetic code more or less directly. Three types of potential advances from molecular genetics can be distinguished: 1 ) im- ‘A Elrich, et al , Science, VOI 196, p, 1313, 1977 provements in the efficiency or rate of biologi- ‘For example, see G H E inert and R. Katzen, “Chemicals From Biomass by Improved Enzyme Technology, ” presented at cal conversion processes (e.g., fermentation, the Biomass as a Non-Fossi/ tue/ Source, ACS/CSJ Joint Chemical anaerobic digestion), 2) introducing specific Congress, Honolulu, Hawall, April 1979

91 92 ● Vol. n-Energy From Biological Processes

The second type involves identifying the Some additional near- to mid-term advances genes responsible for a particular function in are likely in the area of biological conversion one cell and transferring these genes to anoth- processes and with gene transfers in the area er cell. This transfer does not always require a of synthesizing high-value chemicals, like in- detailed knowledge of how the gene produces sulin, that would be either impractical or im- the desired characteristic. One can draw from possible to synthesize by other means. The the pool of naturally occurring characteristics, complexity of plant growth and photosynthet- but the conceptual link between the gene and ic efficiency, however, reduces the chances of the characteristic must be relatively direct. improving ‘these characteristics in plants through molecular genetics in the near future. The third type probably would involve alter- Although the possibility cannot be precluded ing a complex set of interdependent processes that a scientist will alter a crucial process in in the plant. Although some plant physiologists plant growth despite the lack of knowledge, believe that some improvements in photosyn- there are few grounds for predicting that this thetic yield can be achieved by suppressing will occur before the fundamental biochemi- processes like photorespiration (a type of plant cal processes involved in plant growth and respiration that occurs only in the presence of photosynthesis and the way that environmen- light), this belief is highly controversial among tal factors limit them are better understood. specialists in the field. It is generally believed There is a great deal of controversy surround- that the processes involved in plant growth ing this subject, but most arguments— both and photosynthesis and their relation to specif- pro and con– are based on intuition rather ic genes are too subtle and poorly understood than demonstrated fact. at present to know what biochemical processes should or can be altered to improve plant growth and crop yields.

Crop Yields

Current knowledge and theories of plant cropland. Furthermore, since cropland that growth do not enable one to predict the crop could be devoted to energy crops is generally of yields that can be achieved with unconven- poorer quality than average cropland, using this tional crops. Nevertheless, because of the im- as a basis for estimates may be overstating the portance of biomass yields in determining the potential. economics of production, it is important to have an idea of the approximate magnitude of A yield of 140 bu/acre for corn corresponds the yields of various options that may be possi- to a photosynthetic efficiency of about 1.2 per- ble. cent over its 120-day growing season. Perennial crops, however, probably will have somewhat To this end, corn – a highly successful exam- lower efficiencies during the cold weather at ple of crop development– is used as the basis the beginning and end of their growing sea- for these estimates. Corn has the highest pho- sons. Consequently, it is assumed that peren- tosynthetic efficiency of any plant cultivated nials can achieve an average photosynthetic over large areas of the United States. As dis- efficiency of 1.0 percent. With these assump- cussed in chapter 3, an optimistic estimate for tions, and the others stated below, the follow- average corn grain yields would be about 140 ing yields may be possible. bu/acre (3.9 tons of grain/acre) by 2000. Many farmers routinely exceed this yield, as do ex- ● Dry matter yield.— With an 8-month growing perimental plot yields. This number, however, season in the Midwest, biomass production is quite optimistic for average yields from cul- could yield 15 ton/acre-yr of dry plant mat- tivation on millions of acres of average U.S. ter. For the Gulf Coast (12-month growing Ch. 4—Unconventional Biomass Production ● 93

season), the yields could reach 21 ton/acre- cent of the seed weight.3 Assuming that yr. plants which are 50-percent seed contain ● Grain yields.– Based on corn yields, average seeds that are 50-percent oil, the oil content grain production from some plants could may reach 25 percent of the total plant yield 3.9 ton/acre-yr. weight. ● Sugar yields.– Good sugar crops are 40- to Assuming the biochemical reaction pro- 45-percent sugar on a dry weight basis (e.g., ducing the oil is 75 percent as efficient as sugarcane, sweet sorghum). In the Midwest, that which produces cellulose, then for 1- sugar crops will probably be annual crops percent photosynthetic efficiency the oil leading to possible yields of 4 tons of production would be 16 bbl/acre-yr for a sugar/acre-yr. Along the Gulf Coast, there is plant that is 25-percent oil. For an oil-pro- a longer season. Current average sugar ducing reaction that is 50 percent as effi- yields are 4 ton/acre-yr. As with corn, the cient as the reaction that results in cellulose, yields could conceivably be increased by 40 the yield would be 12 bbl/acre-yr. percent to 5.6 ton/acre-yr. Plant material stored as hydrocarbons has ● Aquatic plant yields.– Estimates for water- also been proposed as a source of liquid based plants are more difficult to derive, fuels. Eucalyptus trees and milkweed, for ex- since there is considerably less experience ample, contain up to 12-percent hydrocar- and applicable information. Water plants bons. Assuming that this content could be have a continuous supply of water and are doubled, the same yields as for oil crops never water stressed. For maximum produc- would apply. tivity, nutrients and carbon dioxide (CO,) ● Arid land crop yields.– Another important (for submerged plants) would be added to and sometimes limiting factor in biomass the water and could be available continu- production is water. Generally plants will ously at near-optimum levels. The water grow well without irrigation in areas of the would prevent rapid changes in tempera- United States where the rainfall is 20 to 30 ture. All of these factors favor plant growth, inches or more. For high biomass-producing and if other problems with cultivating crops in relatively humid climates (like the aquatic plants can be solved, yields may be Midwest), the minimum water necessary for quite high (see “Aquiculture” and “Maricul- plant growth in open fields is about 200 ture”). Nevertheless the uncertainty is too weights of water for 1 weight of plant great to make a meaningful comparison growth. There has been interest, however, in with the land-based plants. As with other plants that can grow under more arid condi- plants, experimental yields will probably not tions. In desert regions with very low humidi- be representative of commercially achiev- ties, requirements are more typically 1,000 able yields. weights or more of water per 1 weight of ● Yields in greenhouses.– Yields in green- plant growth. (Some plants survive for long houses are also very uncertain, due to a lack periods of time without water, but they do of sufficient data and potential problems not grow. ) Assuming the 1,000:1 figure, the such as fungal attacks on plants, root rot, maximum plant growth that could be ex- and other problems with extremely humid pected in a region with 5 inches of rain and environments. If these and other problems no irrigation is 0.6 ton/acre-yr. Oil yields are solved, then crop yields approaching would be correspondingly low or less than 1 those estimated for the milder climates may bbl/acre-yr. be achieved. ● Natural systems. — In addition to agriculture, ● Vegetable oil or hydrocarbon yields. – In addi- there has also been interest in using biomass tion to solid material, plant biomass in- produced by plants in their natural state. In cludes oils. New seed oil crops typically con- ✎ ‘D. Gilpin, S Schwarzkopf, j. Norlyn, and R. M. Sachs, “Ener- tain 10-to 15-percent vegetable oil, and in gy From Agriculture– Unconventional Crops,” University of Cal- sunflowers the oil comprises up to 50 per- ifornia at Davis, contractor report to OTA, March 1979. 94 ● Vol. ii—Energy From Biological Processes

the natural state, most of the nutrients are Each plant is, to a certain extent, a special returned to the soil as the leaves drop off case. The experience with large-scale cultiva- or the plant dies and decays. Harvesting of tion of crops is limited to a few food, animal some of the biomass removes some of the feed, fiber, and chemical crops. Many plant nutrients, although animal excretions and scientists argue that maximum food produc- the natural breakdown of minerals in the soil tion implies maximum biomass production. provide new nutrients. The rate of replenish- However, few genetic and development pro- ment varies considerably from area to area, grams have been specifically aimed at max- however, and this determines the rate that imizing biomss output for crops suitable to biomass could be removed from natural sys- large areas of the United States. tems without depleting the soil. These contradictory factors mean that the The potential growth of biomass in contin- potential for biomass production is uncertain. uously harvested natural systems has appar- And the uncertainty of the estimated yields is ently not been studied. (Forestlands are an judged to be at least 50 percent. Consequently, exception, although the emphasis there has the yields could easily vary anywhere from 0.5 been on the production of commercial tim- to 1.5 times the numbers reported. ber rather than on total biomass.) It has been estimated, however, that some natural wet- It is highly unlikely, however, that average lands produce more than 5 ton/acre-yr of U.S. yields for corn will exceed 140 bu/acre growth, and that 11.4 million acres of range- before 2000, and perhaps not after then. And Iand produce more than 2.5 ton/acre-yr. 4 corn is one of the best biomass producers for While no estimates for the production of the U.S. climate known to man. Consequently, natural systems can be given, they will cer- the numbers reported represent reasonable tainly permit less harvestable growth than limits in terms of what is known today. Any intensively managed systems on comparable large-scale production of biomass that signifi- soil. cantly exceeds these yields would represent a In evaluating the possible yields for biomass major breakthrough. Estimates that are based production, all of the yield estimates here on projected yields significantly exceeding should be treated with extreme caution. None those in table 36 either: 1) are limited to the of these yields has been achieved under large- relatively small acreage of the best U.S. soils, scale cultivation (i. e., mill ions of acres) and the Table 36.–Optimistic Future Average Crop Yields for Plants estimates for oil-producing plants are particu- Under Large-Scale Productiona larly uncertain. Experimental plot yields, on Plausible average yieldb the other hand, exceed these yields for many Region Product (ton/acre-yr) plants. Midwest Dry plant matter 15 Moreover, several factors operate to prevent Gulf Coast Dry plant matter 21 Midwest Sugar 4 average yields from reaching these estimates Gulf Coast Sugar 5.6 for large-scale production of biomass. The Midwest Grain 3.9 most important are the less than ideal soils of Midwest Vegetable oil or 1.7- 2.2 hydrocarbons (12 -16 bbl) most potential cropland and the fundamental Area with 5 inches rainfall limitations of plant genetics with current per year and no knowledge. On the other hand, management irrigation...... Dry plant matter 0.6 Area with 5 inches rainfall practices improve with time and increased per year and no costs for farm products may eventually justify irrigation...... Vegetable oil or 0.1 more extensive management practices, such as hydrocarbons (0.7 bbl) additional fertilizers, extensive soil treatment, aln thl~ ~Ont~xt, lar~~.~~al~ @UC(iOfl means cultivation On mllllotls 01 acres of average U S cropland and expanded irrigation. * bEstlmaled Uricerfamly 250 percent *“An Assessment of the Forest and Range Land Situation in the SOURCE Otflceot Technology Assessm@ United States, ” Forest Service, U S Department of Agriculture, review draft, 1979 energy production or whether the necessary water WIII be avail- ● It is unclear whether Irrigation WIII be socially acceptable for able Ch. 4—Unconventional Biomass Production ● 95

2) rely on technologies that do not now exist that are not likely to be cost effective unless and are not anticipated in the near future, or there are dramatic increases in the prices for 3) require extensive management practices farm commodities,

Unconventional Land= Based Crops

A large number of plants not now grown Table 37.–incomplete List of Candidate a commercially in the United States are poten- Unconventional Bioenergy Crops tially energy crop candidates. Some are rela- Lignocellulose crops tively high biomass producers and others American sycamore Red alder could provide a source of a variety of chemi- Bermuda grass Russian thistle Big bluestem Salt cedar cals that could be used as fuel or as chemical Gum tree (eucalyptus) Sudan grass feedstocks. Unlike conventional crops, these Kenaf Switchgrass crops could be considered primarily for their Napier grass Tamarix Poplar Tall fescue value as fuel. (However, see also ch. 10.) Reed canarygrass Assessing and comparing potential yields for Vegetable oil and hydrocarbon crops Crambe Milkweed the unconventional crops from literature re- Guayule Mole plant (euphorbia) ports are extremely difficult, since these re- Gum tree (eucalyptus) Safflower ports often do not give dry yields, the plants Jojoba Turnip rape often are grown on unspecified soils and in dif- Starch and sugar crops ferent climates, and the water and nutrient in- Buffalo gourd Kudzu vine Chicory Sweet potatoes puts often are not given. Furthermore, it is a Fodderbeets Sweet sorghum well-known fact that experimental plot yields Jerusalem artichoke are larger than those achieved with large-scale asome of these crops are produced commercially today on a hmlted scale, but fIOt for their ener9Y commercial cultivation. For these reasons, the value SOURCE D Gdpm, S Schwarzkopf, J Norlyn, and R M Sachs, “Energy From Agrlculture– yields reported below should be treated with Unconventional Crops, ” Unwersity of Cahforrua at Davis, contractor report to OTA. extreme skepticism. Comparative cultivation March 1979, S Barber, et al , “The Potential of Producing Energy From Agriculture, ” Purdue Unwerslty, contractor report to OTA, and J S Bethel, et al , “Energy From experiments and crop development will be Wood, ” UnwersNy of Washington, contractor report to OTA needed in the various regions and soil types in since the perennials develop their leaf cover order to establish which crops are, in fact, sooner in the spring and do not need to gener- suitable or superior for bioenergy production. ate a complete root system each year. One In broad terms, the categories include: 1 ) ligno- would also expect grasses to be superior bio- cellulose, 2) vegetable oil and hydrocarbon, mass producers to trees because of their larger and 3) starch and sugar crops. Each group is leaf area per acre of ground, * but considerable considered briefly below, and an incomplete attention has been focused on trees, since the list of candidate bioenergy crops is shown in technologies for using wood are more ad- table 37. vanced. Experimental plot yields for short-rota- 5 Lignocellulose Crops tion trees are 5 to 20 ton/acre-yr. Yields of as much as 10 to 15 ton/acre-yr may be achieved Various species of hardwood trees (e.g., red for large-scale cultivation of some of these alder, hybrid poplar) and grasses (e. g., kenaf, crops in good soil (see “Crop Yields” section) Bermuda grass, Sudan grass, big bluestem) are but are likely to be 6 to 10 ton/acre-yr in poorer candidates for crops grown primarily for their soils. Since farming costs could be similar to high dry matter yields (Iignocellulose crops). corn, this could result in biomass for about $20 to $50/ton. Theoretically, one would expect perennial “Thereby reducing Ilght saturation, which lowers photosyn- crops (like trees and some grasses) to be su- thetic efficiency perior biomass producers to annual crops, 5Cllpin, et al , op clt 96 ● Vol. I/—Energy From Biological Processes

The trees would typically be grown for 6 to The theory of hydrocarbon and vegetable oil 10 years before harvest, while the grasses production in plants is not adequate to predict would be harvested several times a year. With possible yields. However, from other consid- fewer harvests for the trees, each harvest could erations (“Crop Yields” section) there may be a be considerably more expensive and consume significant potential for improvement. Further- more energy than grass harvests without unfa- more, some of these crops (e. g., guayule) may vorably affecting the economics or net energy do well on land where there is slightly less wa- balance. However, tree crops would require ter available than would be needed for con- that the land be dedicated to the crop for sev- ventional crops. a Others, such as milkweed, eral years and converting the land to other can be grown with brackish water which would uses would be more expensive, due to the de- be unusable for conventional food crops. ’ veloped root system. Also, if a disease were to Comparative tests under comparable condi- kill the crop, reestablishing a tree crop would tions will be necessary to determine which be more expensive. Both trees and some plants show the most promise for energy pro- grasses are perennial crops and, consequently, duction. would require fewer herbicides and would re- Because of the higher prices that can be duce erosion on erosion-prone land as com- paid for chemicals and natural rubber, the fact pared to annual crops. Grasses, having a more that these products are economic in some complete soil cover, would be more effective cases does not in any way imply that energy in preventing soil erosion. production from vegetable oil and hydrocarbon plants will be economic. Some proponents Vegetable Oil and Hydrocarbon Crops of hydrocarbon plant development have failed to distinguish between these end uses, a fact Vegetable oils and hydrocarbons are chemi- that has led to considerable confusion and cally quite different from petroleum oil. Nev- misunderstanding. ertheless, most vegetable oils and hydrocarbons can be burned and might prove to be a Critics of the development of vegetable oil substitute for fuel oils or, with refining, for and hydrocarbon plants for energy argue that other liquid fuels. However, appropriate meth- the production of these products by plants is ods for extracting the oil from the plant and considerably less efficient than normal chemi- for refining the oil are not well defined at pres- cal synthesis (e.g., to produce methanol or eth- ent. anol from dry plant matter). They also point out that the plant often must be subjected to A number of edible and inedible vegetable stress (drought or cold) to produce hydrocar- 6 oils are currently produced commercially. In bons, and this lowers the photosynthetic effi- addition, unconventional crops such as gum ciency. Consequently, they contend that the tree (eucalyptus), mole pIant (euphorbias), high yields being predicted (e.g., 26 bbl/acre10), guayule, milkweed, and others could be used will not be achieved in the foreseeable future. as vegetable oil and hydrocarbon crops (or for natural rubber). The maximum current yields At present, however, the theory of and ex- of commercial oil plants are in the range of perience with these types of plants is inade- 100 to 200 gal/acre (2.5 to 5 bbl) of vegetable quate to make a meaningful judgment. oil and/or hydrocarbon. Reports of 10 bbl/acre (420 gal) for euphorbia were apparently based on measurements of plants on the edge of a ‘K. E. Foster, et al., “A Sociotechnical Survey of Guayule Rub- field, which were 1.5 times larger than interior ber Commercialization, ” Office of Arid Land Studies, University plants. Also, in some cases, 16 months of of Arizona, Tucson, Ariz., prepared for the National Science growth were used to obtain “annual” yields, ’ Foundation under grant PRA 78-11632, April 1979. ‘W. H, Bollinger, Plant Resources Institute, Salt Lake City, 6Agricu/tura/ Statistics (Washington, D. C.: U.S. Department of Utah, private communication, 198o. Agriculture, 1978). ‘“J D. Johnson and C. W Hinman, “Oils and Rubber From Arid ‘Gilpin, et al., op. cit. Land Plants,” Science, VOI 280, p. 460,1980. Ch. 4—Unconventional Biomass Production ● 97

Starch and Sugar Crops Other plants such as fodderbeets, sweet potatoes, and Kudzu vine are also potential etha- Starch and sugar crops are of interest since nol crops. Comparative studies are necessary they can be used to produce ethanol with com- to determine which crops are best in each soil mercial technology. Current corn grain yields type and region of the United States. As was can be processed into about 260 gal of ethanol emphasized in chapter 3, this comparison per acre cultivated and sugar beets (usually ir- should include the displacement of other rigated) can produce about 350 gal/acre, on crops that can be achieved by the byproducts the average. Irrigated corn, however, would of ethanol production. This factor tends to match the sugar beet yield. Furthermore, ex- favor grains, but other possibilities do exist. ’2 perimental plot yields for corn produce about 430 gal/acre-yr and record yields exceed 850 gal/acre. In addition to the conventional starch General Aspects and sugar crops, several other plants have Intensive cultivation of unconventional been proposed as ethanol feedstocks including crops may cost about the same as corn, or $200 sweet sorghum and Jerusalem artichokes. to $400/acre-yr in the Midwest. These costs, Experimental plot yields for sweet sorghum together with the yield estimates given above, could be processed into about 400 gal of etha- allow an approximate comparison of the costs nol per acre year. Furthermore, this crop pro- for various unconventional land crops, which duces large quantities of residues that are suit- is shown in table 38. Since the exact cultiva- able for use as a distillery boiler fuel. The tion needs have not been established, a more yields for large-scale cultivation, however, are detailed comparison is not warranted at this still unknown, and concern has been expressed time. These costs estimates, however, show that droughts during parts of the growing sea- that unconventional crops may be economic son could reduce sugar yields significantly. energy sources. The ultimate costs will depend to a large extent on the yields that can actually Experimental plot yields for Jerusalem arti- be attained with intensive management and chokes have produced about twice the sugar the success of developing crops that can be yields of sugar beets under the same growing cultivated on land that is poorly suited to food conditions in Canada. Whether this result can production. be applied to other regions is not known. Jeru- salem artichoke, like the sugar beet, is a root The crops that are now grown in U.S. agricul- crop. Harvesting it, therefore, causes extensive ture were selected for properties that are unre- soil disturbance which increases the chances I IR Carl son, B commoner, D Freedman, and R Scott, “Inter- of soiI erosion. Im Report on Possible Energy Production Alternatives In Crop- Livestock Agriculture, ” Center for the Biology of Natural SVs- “Gllpln, et al , op clt tems, Washington University, St Louis, Mo , Jan. 4, 1979

Table 38.-Optimistic Cost Estimates for Unconventional Cropsa

Product Ultimate fuel Yield of ultimate fuel per acre cultivated Contribution of feedstock to fuel costb 6 Dry plant matter Combustible dry matter 15 ton (195 106 Btu) $20/ton ($1 .53/ 10 Btu) 6 Dry plant matter Methanol 1,500 gale (95 106 Btu) $0.20/gal ($3.15/10 Btu) Dry plant matter Ethanol 1,300 gald (107 106 Btu) $0.23/gal ($2.80/10’ Btu) Grain Ethanol 364 gal (31 106 Btu) $0.82/gale ($9.89/10’ Btu) Sugar (Midwest) Ethanol 540 gal (46 106 Btu) $0.56/gal ($6.50/10’ Btu) Vegetable oil or hydrocarbon crop Vegetable oil or hydrocarbon 504-670 gal (63-84 106 Btu) $0.45-$0.60/gal ($3.60-$4.70/10’ Btu) aBased on yields m table 36 bAssumlng $3i)o/acre cuttwatlon and harvest costs, does nOt include COnver510n costs cAssumlng yw,lds of 100 gal/ton of btornass dAssumes yields of 85 galiton of biomass e~s not ,n~lude by Droduct credit for distillers’ grain H byproduct Credits Included, the Sltuatlon becomes more complex as described m ch 3, in the section on “Energy Potential From Conventlonai Crops ‘‘ The bypro&ct credit would reduce the costs by roughly one-thtrd SOURCE Otflce of Technology Assessment 98 ● Vol. Ii—Energy From Biological Processes

Iated to energy. It ‘is likely, therefore, that Root plants (e.g., Jerusalem artichokes, other plants will prove to be superior to con- sugar beets, potatoes, sweet potatoes) will ventional crops for energy production. Beyond cause the most soil erosion. Annual crops will the yields of these crops, properties like insen- be next, and perennial grasses can virtually sitivity to poor soils, multiple products (e.g., eliminate soil erosion. vegetable oil, sugar, and/or starch plus dry Soil structure and climate are dominant plant herbage) displacement of other crops features controlling plant growth and these with crop byproducts (e. g., corn distillery by- can be controlled by man only to a very lim- product), the energy requirements to cultivate ited extent. Plants vary as to their sensitivity to the crop, the energy needed to convert it into a these factors and to the presence of nutrient form that can be stored (especially for sugar solubilizing mycorrhizae in the soil, ’3 but and starch crops), tolerance to adverse weath- yields will decrease on going to poor soils and er conditions, ease of harvesting and conver- climates. Crops grown in arid climates without sion to fuels, and the environmental impacts irrigation or an underground supply of water of growing the crop are all factors that should will give low yields; and social resistance to be considered when choosing energy crops. In using water for energy production in the West short, analyses of the net premium fuels could preclude irrigated energy crops, al- displacement per new acre cultivated (as was though some people maintain that this resist- done for various conventional crops in ch. 3), ance will not extend to crops. the cost, and the environmental impacts are needed to compare the options. Due to the di- Finally, any crop that grows very well in an versity of U.S. soils and climates, different area without inputs from man is likely to crops will no doubt prove to be superior in dif- spread and become a weed problem. ferent regions. Many of the possible unconventional crops appear promising, but the ultimate decisions will have to come from experiment and experience. (Typically it requires 10 to 20 years to develop a new crop.) Neverthe- “J. M. Trappe and R. D. Fogel, “Ecosystematic Functions of Mycorrhizae,” reproduced from Range Sci. Dep Sci., series No, less, some general aspects of plants can be ex- 26, Colorado State University, Fort Collins, by U S. Department pected to hold for the unconventional crops. of Agriculture.

Aquiculture

Aquatic plants comprise a diversity of types, systems for energy production. ’4 The general from the single-celled microalgae to the large conclusions were that the production of aqua- marsh plants such as cattails and even some tic biomass has near-term potential in conjunc- trees such as mangroves. Considerable interest tion with wastewater treatment and high-value exists in the cultivation of many different chemicals production. However, the develop- aquatic plants as energy sources. Examples are ment of large-scale “energy farms” based on the production of cattails in the extensive aquatic plants is less promising at present, marshes of Minnesota, the cultivation of water from both an economic or a resource potential hyacinths on wastewaters in Mississippi or viewpoint. Nevertheless, aquatic plants have Florida, and the establishment in the South- certain unique attributes, the key one being west of large-scale brackish water ponds for high achievable biomass production rates microalgal production of chemical feedstocks. - “J. Benemann, “Energy From Aquiculture Biomass Systems: OTA has prepared a detailed review of the po- Fresh and Brackish Water Aquatic Plants, ” Ecoenergetics, Inc , tential of fresh and brackish water aquiculture Vacaville, Cal if , contractor report to OTA, April 1979 Ch. 4—Unconventional Biomass Production ● 99

which justify continued research on a variety United States and the large fraction of biomass of approaches to the development of aquacul- present in the root system which may be diffi- ture energy systems. cult to recover. The submerged aquatic plants such as the notorious weed Hydrilla, are also Higher aquatic plants growing in or on not remarkable for their biomass productivity. water are not, as a rule, water limited — a com- Adaptation to the light-poor environment fre- mon and natural state of land plants. Thus, quently encountered below the water surface they are capable of higher rates of photosyn- has made these plants poor performers at the thesis by keeping their stomata (plant pores) high light intensities that would be the norm in open longer than land plants* thereby, increas- a biomass production system.

ing C02 absorption. Thus, plants such as the water hyacinth and cattails exhibit very high Finally, the case of the microalgae must be rates of biomass production, often exceeding considered. Being completely submerged they 20 ton/acre-yr. This high productivity is also are subject to significant light losses by achieved, however, by evaporation of large reflection from the water surface (at low solar amounts of water, exceeding by a factor of angles) and scattering of light. More important- two to four that transpired by land plants. ly, in a mixed algal pond, the cells near the sur- Thus, cultivation of water plants can only be face tend to absorb more light than they can considered where ample supplies of water ex- use in photosynthesis, resulting in a significant ist or where the systems are covered, such as in waste of solar energy. However, if a microalgal greenhouse structures. production system is designed to enhance mixing, then rapid adjustment by the algae occurs, Some aquatic plants, however, do not exhib- thus overcoming, to some extent, the handicap it very high biomass production rates. For ex- inherent in inefficient sunlight absorption by ample, the common duckweed (Lemna sp.) microalgal cultures. Therefore, microalgal cul- covers a water surface very rapidly; however, tures could be considered in a biomass produc- once this is achieved, further growth in the ver- tion system. A review of the rather sparse pro- tical direction is minimal. Thus, the productivi- ductivity data available, together with consid- ty of such plants is relatively low when com- eration of the basic photosynthetic processes pared to plants such as water hyacinths and involved, suggest that green algae and diatoms marsh plants which extend their shoots up to are promising candidates for mass cultivation, several feet into the air. Indeed, the high leaf probably with achievable production rates of area index (the ratio of the total leaf area to at least 20 ton/acre-yr, with blue-green algae, the ground area), sometimes exceeding 10, of particularly the nitrogen-fixing species, con- these plants, accounts, along with high trans- siderably less productive. piration rates, for their high productivity. It must be noted that the available data on Another type of aquatic plant that exhibits aquatic plant productivity are too limited to relatively low productivity is the salt marsh allow confident extrapolations to large-scale plant Spartina, which does not produce as systems. Most available data are based on much biomass as its freshwater analogues such natural systems where nutrient limitations may as Typha (cattails) or Phragmites (bullrush). The have depressed productivity or small-scale, high salt concentration tolerated by Spartina short-term experimental systems where edge also results in a decrease of transpiration and effects and other errors may have increased productivity. Even among the freshwater productivity. Actual yearly biomass produc- marsh plants, biomass productivities are lim- tion rates in sufficiently large-scale managed ited by both the seasonal growth patterns of systems must be considered uncertain for any the plants in the temperate climate of the aquatic plant, particularly if factors such as stand establishment, pest control, optimal fer- ‘Somata are closed to conserve water, but this also prevents carbon dioxide from entering the leaf tilizer supply, and harvesting strategy are con- ‘ ‘Ibid cerned. Thus, to a considerable extent, assess- 100 ● Vol. n-Energy From Biological Processes

ing the potential of aquatic plants in energy ponds have been used for several decades in farming, like that of other unconventional many wastewater treatment systems through- crops, involves more uncertainty than specific out the United States. More stringent water detailed knowledge. quality standards are resulting in a need for better microalgal harvesting technology and Among the uncertainties are the economics presenting an opportunity for fuel recovery of the production system, including the har- from the harvested microalgae. Several proj- vesting of the plants. Detailed economic anal- ects throughout the United States have demon- yses are not available; those that have been strated the beneficial effects of marsh plants carried out are based on too many optimistic in wastewater. treatment. In all cases, waste- assumptions to be credible or useful. Of water aquiculture appears more economical course each type of plant will require a dif- and less energy intensive than conventional ferent cultivation and harvesting system. How- technologies. 16 However, the total potential ever, in all cases, these appear to be signifi- impact of wastewater aquiculture on U.S. en- cantly more expensive per acre in both capital ergy supplies, even when making favorable and operations than the costs of terrestrial market penetration assumptions, is minimal — plants. This increased cost per acre can only be about 0.05 to 0.10 Quad/yr.17 justified by an increased biomass production rate or a specific, higher valued product. Be- For aquatic plants to make a more signifi- cause the productivity and economics of aqua- cant contribution to U.S. energy resources, tic plants are, to a large degree, unknown, the other types of aquatic biomass energy systems potential for aquatic plant biomass energy must be developed. One alternative is the con- farming is in doubt. version to fuel of aquatic plants already harvested from natural, unmanaged stands. Exam- One approach to improve the economics of ples are water hyacinth weeds removed by me- such systems is to combine the biomass energy chanical harvesters from channels in Florida system with a wastewater treatment function. and other southern States and cattails or bull- As aquatic plants are in intimate contact with rushes cut periodically in natural marshes in water, they can perform a number of very im- Minnesota or South Carolina to improve portant waste treatment functions—oxygen wildfowl habitats. However, the infrequent production (by microalgae) which allows bac- occurrence of such harvests, the small bio- terial breakdown of wastes, settling and filtra- mass quantities involved, and transportation tion of suspended solids, uptake of organics difficulties make energy recovery from such and heavy metals, and, perhaps most impor- sources essentially impractical. The conver- tantly, uptake of the key nutrients that cause sion, if practiced, of natural marsh systems to pollution. The relatively high concentrations large-scale managed (planted, fertilized, har- of nitrogen and phosphorus in aquatic plants vested) plantations will present significant eco- (e.g., about 10 percent nitrogen and 1 percent logical problems and, even if these are ameli- phosphorus in microalgae and 3 percent nitro- orated or overcome, opposition by environ- gen and 0.3 percent phosphorus in water hya- mental groups. Nonetheless, large areas of cinths), makes these plants particularly useful marshes do exist in the United States and they, in nutrient removal from wastewaters. Re- in the long term, may become resources that search in wastewater aquiculture is well ad- could be exploited on a multipurpose and sus- vanced, although some critical problems retained yield basis like the national forests. In main to be elucidated, and several large dem- the near term, however, the technology for onstration projects are being initiated through- aquatic plant biomass energy systems must be out the United States. For example, water hya- developed with presently unused or “margin- cinths are being used in wastewater treatment plants in Coral Gables and Walt Disney World, “’Ibid Fla., in projects which involve fuel recovery by “Ibid anaerobic digestion of the biomass. Microalgal “Ibid. Ch. 4—Unconventional Biomass Production ● 101

al” land and water resources. In addition, Microalgae are known to produce a variety relatively high-value biomass energy products, of useful chemicals. However, the develop- specifically chemicals and liquid fuels, should ment of such production technology is only be produced by such systems. Examples of just now beginning, and the potential resource such systems include the production of alco- base (land, water, nutrients) available for such hol fuels from cattails (either by hydrolysis of systems is not yet quantified. Thus, the future the areal parts or directly from the starches contribution to U.S. fuel supplies of aquatic stored in the roots) or the production of hydro- plant biomass energy systems cannot be pre- carbon fuels and specialty chemicals from mi- dicted. However, sufficient possibilities and croalgae. promise exist to warrant further R&D efforts.

Mariculture

This section describes problems and oppor- tivation techniques. A small test farm has been tunities associated with developing future installed along the California coast. 20 ocean farms which might use the giant kelp Large ocean kelp farms could theoretically (macrocysts) as a future biomass energy supply significant quantities of natural gas source. Other macroalgae have also been pro- (methane). Linked to a methane production posed as potential marine biomass crops. By system, for example, and assuming serious examining the possibilities of kelp and also technical problems are solved, a l-million-acre noting other proposals, OTA hopes to illustrate kelp farm could produce enough gas to supply the status of this technology in general, its 1 percent of current U.S. gas needs. future potential, the problems involved, and the Federal role in this segment of alternative It would be no easy matter to farm such vast energy research. tracts of ocean. Much still needs to be learned about macroalgae cultivation. But serious re Macroalgae are harvested around the world. search is reducing the areas of ignorance and About 2 million wet metric tonnes are now cut seaweed may some day become a biomass pro- annually, and estimates are that the total po- ducer. tential worldwide crop is 10 times this much— about 20 miIlion wet tonnes. 9 Algae are among the simplest and most primitive of plants. The larger macroscopic In recent decades seaweed cultivation has algae are commonly referred to as seaweeds or rapidly become more successful and has sub- macroalgae. Large seaweeds are the dominant stantially added to annual harvest figures. For plant in most shallow coastal waters including example, as of 1970 there were 130,000 acres those off California and Mexico, where they at- of sea surface under cultivation in Japan, tach themselves to rocks or some other hard about 25,000 acres in The Peoples Republic of substrate under water. China, and additional acreage in Taiwan, Korea, the Philippines, and elsewhere. None of To date, the seaweeds apparently most the current annual world harvest is being used adaptable to human cultivation are the red for energy production. and the brown algae. People have eaten red algae varieties for thousands of years, espe- In the United States, where wild seaweed cially in countries such as Japan and China. beds have been harvested for many years, the possibility is beginning to be studied of increasing production through ocean farm cul- 2“A. Flowers, statement before the House of Representatives Committee on Merchant Marines and Fisheries, Subcommittee ‘9G Michanek, Seaweed Resources of the Ocean, U.N, Food on Oceanography, p. 18, committee report serial No, 95-4, June and Agriculture Organization, Rome, 1975, 7,1978. 102 ● Vol. /l—Energy From Biological Processes

The brown algae group includes the giant Kelp may grow in length as much as 2 ft/day kelp Macrocystis (figure 13), already harvested or increase its weight by 5 percent per day under optimum conditions. The plants form natu- Figure 13.—Macrocystis Pyrifera ral beds up to 3 miles wide and several miles long in southern California. This kelp is now harvested and put to a variety of uses, principally in the food-processing industry. Fuels have never been produced from kelp except in minute quantities as part of research testing.22

Unfortunately, there is no consensus among the experts who have made projections as to the potential of ocean energy farms. Their estimated costs vary widely and are based on such very sparse data that they cannot be used to either support or reject ocean farm proposals. Estimates of production rates vary by factors of as much as 100. Better experimental data and more complete biological engineering tests will allow for better estimates in the future. The estimates used here lie approximately in the middle of responsible optimistic and pessimistic projections for a 400,000-hec- tare (1 million acre) ocean kelp farm:

● average productivity = 20 dry ash free (DAF) tons per acre per year, and ● average annual energy produced = 0.2 Quad (1 percent of U.S. gas consumption of 20 Quads/y r).

Such a system, if built, would provide the equivalent in energy supply of one large LNG- importing plant such as the one located at Cove Point, Md. It would, of course, be a domestic rather than an imported fuel, however.

(A: 1/64 natural size; B: 1/4 natural size.) The Giant Kelp is Experiments are underway into the best shown in the left part of the plate in a natural pose with the laboratory-reared seaweed farms. Eventually, long leafy stipes rising to the sea surface from the massive some researchers hope to produce a “pedi- holdfast. On the right is one of the leaf-like fronds showing greed” kelp bred specifically for high methane the gas-filled float bladder at its base and the distinctive teeth along the margin (Anon. 1954). production, fast growth, and hardiness. A key problem faced by potential ocean SOURCE: Velco, Inc. kelp farmers is to deliver enough nutrients to the plants to fertilize them. This is because, in the United States from wild and semiculti- while the deep waters of the ocean contain vated beds and considered at present as the many necessary nutrients, surface water is best candidate for intensive cultivation off often as devoid of nourishment as a desert is California and as a possible fuel producer. z’ ZZM. Neushal, “The C)cmnesticatlorl of the C iant Kelp, ~acro z ‘Neushal, et al., “Biomass Production Through the Cultiva- cystis as a Marine Plant Biomass Producer, ” presented at the Ma- tion of Macroalgae in the Sea, ” p 100, Neushul Mariculture, rine Biomass of the Pacific Northwest Coast Symposium, Oregon Inc , for OTA, Oct 6, 1978 State University, Mar. 3,1977 Ch. 4—Unconventional Biomass Production ● 103

devoid of water. One fertilizing technique be- age corn harvest is 6 DAF ton/acre-yr and Ha- ing tried is artificial upwelling of seawater, waiian sugarcane averages 14 DAF ton/acre-yr. which involves pumping nutrient-rich, deep OTA estimates that if about 1 million acres ocean water to the surface to benefit the kelp were ever farmed, the gross energy production plants. 23 could amount to 0.2 Quad. This is equal to ap- Current research on marine plants can be di- proximately 1 percent of current U.S. natural vided into two categories. gas consumption. These production estimates should be treated with caution since there are The first category, funded by several Federal no ocean kelp energy farms and nobody has agencies to a total of about $1 million in 1979, ever planted and harvested a macroalgae generally includes research projects aimed at a energy crop. better understanding of marine plants, their cultivation, and potential new uses of the Actual gross energy production from such a plants. huge hypothetical ocean kelp plantation has been projected by other researchers to range The other “category” is actually just one from 10 times OTA’s estimate to only one- project: the Marine Biomass Research Program tenth that figure. The entire project might jointly funded by the Gas Research Institute simply prove impossible, others caution. Years (GRI) and the Department of Energy (DOE), of experiments will be necessary before any which has funded over $9 million of directed projections can be confirmed. research as of 1979. There is even less data to draw on in esti- This ongoing marine biomass project in- mating net energy possibilities. In a report cludes a test farm off California. The farm prepared for DOE by the Dynatech R&D Corp., began artificial upwelling experiments late in net energy outputs were estimated to range 1978, but was forced to suspend operations in anywhere from a negative number to about 70 early 1979 due to storm damage. This proto- percent of crop energy .24 type is meant to provide biological information and research clues needed to operate Much of the technology to construct pres- much larger culture farms. It also aims at ex- ent concepts of open ocean farms is already perimental work into cultivation of giant kelp well known. Similar platforms, structures, and on moored structures in the open ocean. The moorings have been built for the offshore oil test farm, may lead to the actual operation of industry and the existing seaweed industry uses a full-scale ocean farm. mechanical harvesting techniques.

There is considerable difficulty at this time Less certain areas of ocean farm engineering in evaluating the appropriateness of the at present include nutrient distribution, disper- Marine Biomass Research Program because lit- sion characteristics of upwelled water, and tle has been produced. It is important that specific configuration of the structure to research results on the cultivation of kelp on which kelp plants will be attached. A major ocean farms be reported in a comprehensive problem for cultivated kelp beds may be to way and subjected to critical review if a future supply an ocean farm with proper nutrients in large program is to be justified. correct quantities. The extreme difficulties of noting the delicate balance of nutrients found Kelp and other seaweeds are potentially a in a natural environment and reproducing this highly productive source of biomass for fuels. in a cultivated one are well known to research- Estimates can vary drastically as to what may ers. be possible for future large ocean farms, but OTA’s evaluation of a hypothetical ocean kelp farm indicates productiveness could range 24 from a low value of 6 DAF ton/acre-yr to a high Dynatech R&D Co , “Cost Analysis of Aquatic Biomass Sys- terns, ” prepared for the Department of Energy, contract No value of 30. I n comparison, this country’s aver- HCP/ET-400-78/I 104 . Vol. II —Energy From Biological Processes

Test farms will upwell deep ocean water to is also needed on plant diseases, preda- supply kelp with proper nutrients. Reservations tors, and water movement and quality); about this procedure are twofold among skep- ● the possibility of developing shallow wa- tics. They worry that deep water could become ter kelp farms either in areas of natural stagnant under the farms, or that, once up- upwelling or in conjunction with other fer- welled, the water would dilute too rapidly and tilizing techniques (see ch. 10); sink again. ● hard data on net energy expectations is lacking; and As previously mentioned, this country’s ma- ● no plans are being readied at present as jor ocean biomass project is jointly supported alternatives to fertilization by upwelling. by CR I and DOE. The project may become the most heavily funded biomass program of the Since plans for future ocean biomass farms 1980’S, with grants projected to grow to over call for the use of millions of acres of ocean $50 million yearly by 1983. Plans for this proj- surface, there will be conflicts with other tradi- ect have been developed mainly by GRI, al- tional users. The dedication of large areas of though regular DOE approval for phases of the open ocean surface for a single commercial project is mandatory.25 purpose such as this is unprecedented. It would require complex, special regulation GRI estimates imply that ocean kelp farming after review of current local, national, and in- could be a commercially viable project for this ternational laws. country. The Institute’s fuel production cost estimates for methane generation from kelp Even though the ocean space within the 200- range from $3 to $6/million Btu. mile zone surrounding the United States is 1½ billion acres, conflicts can be expected with The previously mentioned Dynatech report such traditional ocean users as commercial on fuels from marine biomass comes to a dif- shipping, the navy, commercial and sport fish- ferent conclusion. Its estimates range from ing, offshore oil and gas operation, and recrea- $7/million Btu up to several hundreds of dol- tional boating. To date, no detailed investiga- lars per million Btu, should productivity prove tion of legal or institutional approaches to re- low and design costs high. solving conflicts has been accomplished. This Some critics of the GRI marine biomass pro- issue will need analysis prior to any large-scale gram contend that there is not enough data initiative in ocean farming, and will have a ma- available to justify the level of expenditures jor impact on feasibility, productivity, and cost for the biological test farm. of marine biomass in the future. Analyses of specific sites and siting problems will be cru- Critics have stated that the open ocean test cial to the ocean question. farm is an inappropriate and perhaps premature step in a long, logical process of devel- OTA has found that Federal research pro- oping future deep sea operations. Considera- grams directed toward energy problems have tions which may be overlooked by this test not been adequately coordinated with similar farm approach include: research directed toward production of food, chemicals, or other products. the need for better information on kelp growth and productivity and limiting fac- Much research is needed to develop any tors in natural beds; suitable marine pIant culture regardless of the need for additional basic research into whether the end product is food or fuel. Such nutrients and productivity (much research basic research could be better supported and coordinated by all interested agencies. Pro- grams supported by Sea Grant and the Nation- al Science Foundation have tended to focus on ‘sGeneral Electric Co., briefing, “Energy From Marine Biomass Project, Program Review, ” for the Gas Research Institute, New- basic biological efforts or food production port Beach, Cal if., March 1978. goals while DOE programs are focused on pri- Ch. 4—Unconventional Biomass Production ● 105

mary fuel production. Since DOE now has the in deepwater, open ocean farms. This possibili- major funds available for seaweed research, ty would mean coordination of several existing the tendency has been to create programs fo- research efforts; expanding some, developing cused on fuel production. The encouragement some new ones, and generally integrating of further diversity in existing seaweed re- many efforts focused on basic biological ques- search efforts is essential to a long-term im- tions and food production as well as energy provement in the knowledge and capability of production. developing future marine plant culture programs.

One approach to conducting a systematic program for developing ocean farms would be to expand research in natural seaweed beds and shallow water farms prior to experiments

Other Unconventional Approaches

There are several other unconventional ap- captured by the plants as effectively as it proaches to biomass production. Because of could. Consequently, this approach would also the complexity of plant growth, it is likely that be expected to reduce the total biomass yields. many approaches will be tried and fail. How- ever, this complexity also gives rise to significant possibilities. While all unconventional ap- Chemical Inoculation proaches cannot be covered here, a few are By subjecting some plants to herbicides like discussed below. paraquat or 2,4-D, hydrocarbon or vegetable oil production can sometimes be increased. Multiple Cropping These chemicals block certain biochemical pathways, thus promoting greater production MultipIe cropping consists of growing two or of other products. Preliminary results with more crops on the same acreage in a year. guayule, for example, indicate that 2,4-D may Growing winter wheat on land that produced a cause a doubling of the natural rubber content summer crop is one example. The winter wheat of this plant.26 While it is too soon to assess can delay spring planting, so its use is ap- this approach, it may prove to be an effective plicable only for land where certain summer way of improving yields of these products. crops are to be grown. However, this is basically a conventional approach. Energy Farms The unconventional multiple cropping con- 27 sists of growing more than one summer crop Energy farms have been proposed as a on an acreage by harvesting the first crop means of providing a reliable supply of large before it matures or developing species that quantities of biomass for large conversion mature rapidly. Since starches, sugars, vegeta- facilities located on or near the farm. The ble oils, and hydrocarbons are generally pro- basic idea is to have a large tract of land (tens duced in the greatest quantities in mature ‘6Gilpin, et al,, op. cit. plants, this approach would probably reduce “See, e.g., G, Szego, “Design, Operation, and Economics of the overall yields of these products. Also, the the Energy Plantation, ” Proceedings Conference on Capturing the time between the harvest of the first crop and Sun Through 6ioconversion (Washington Center for Metropoli- tan Studies, 1976); G C. Szego and C C Kemp, “Energy Forests the development of a full leaf cover in the sec- and Fuel Plantations,” Chemtech, p. 275, May 1973; and Si/vicu/- ond crop will be a time when sunlight is not ture Biomass farms (McLean, Va.: MITRE Corp., 1977).

67-968 0 - 80 - 8 106 ● Vol. Il—Energy From Biological Processes

of thousands of acres) dedicated to growing eral years. A pest infestation, however, the biomass feedstock for a nearby conversion could destroy the entire crop in which an facility. Although this is technically possible, a average of 3 to 5 years’ cultivation had number of practical and economic considera- been invested, and this could be financial- tions probably will limit investment in energy ly disastrous. If grass is the energy crop, farms. Moreover, this approach ignores the ef- or the time between tree harvests is re- fect that bioenergy production has on related duced, the loss from a pest infestation sectors. Some of the more important of these would be considerably less, but the yields points are: would fluctuate more from year to year, making it necessary to rely on outside ● Land. – The land available for energy sources of biomass in years with low har- farms has often been estimated to be sev- 2829 vests or to sell surpluses in years with eral hundred million acres. OTA’s bumper harvests. analysis, however, indicates that consider- ● Competition with other uses for the land. ably less land is available for biomass pro- – Because of the uncertainty about fu- duction (see ch. 3). Furthermore, there ture cropland needs for food production, would be practical difficulties with buy- it would be unwise to assume that tens of ing large contiguous tracts of the size millions of acres could be devoted to a needed for large conversion facilities (tens conversion facility for 30 years without af- to hundreds of thousands of acres). fecting the price of farmland and thus If cultivation on very poor or arid land food. proves to be feasible or if irrigation for ● Preclusion of nonenergy benefits. -OTA’s energy production is socially acceptable analysis indicates that bioenergy harvests, and the water is available, then these lim- if properly integrated into nonenergy sec- itations could be somewhat less severe tors, can provide benefits beyond the en- than they appear to be at present. ergy, such as increased growth of timber ● Estimates of future yields Crop yields.– suitable for paper and lumber. Attempting from short-rotation tree farms have been to isolate bioenergy production from as high as 30 ton/acre-yr,30 which OTA these other sectors would preclude some considers to be highly unrealistic. Yields of the potential benefits. of 6 to 10 ton/acre-yr are more realistic for the poorer soil that could be available for Although none of these factors is insur- energy farms. mountable, taken together they make energy ● Initial investment.– If short-rotation trees farms appear considerably less attractive than are used as the energy crop, there would numerous other bioenergy options. Particular- be a 6- to 10-year Ieadtime before the first ly because of the risk and the initial invest- harvest could be made. This would be ment, it is more likely that bioenergy crops will prohibitively long for many investors. be grown as one of the many crop choices Grasses, however, would reduce the lead- available to farmers, rather than on large time to a fraction of a year. In either case, tracts dedicated solely to energy production. the cost of acquiring the land would in- There is, however, no technical reason why en- crease the initial investment substantially. ergy farms cannot be constructed. ● Risk.— Using short-rotation trees as the energy crop would give yields that are less Biophotolysis sensitive to weather than grass because the growth would be averaged over sev- Biophotolysis is generally defined as the process by which certain microscopic algae ‘%zego, op. cit. can produce hydrogen (and oxygen) from wa- 29s;/v;cu/ture /3;c3fnass Farms, op. cit. ter and sunlight. Two distinct mechanisms are ‘OJ. A Allich, Jr., and R E Inman, “Effective Utilization of Solar Energy to Produce Clear Fuel, ” Stanford Research insti- known by which microalgae can carry out bio- tute, final report No NSF/RANN/SE/Cl 38723/FR/2 , 1974 photolysis: 1) through a “hydrogenase” en- Ch. 4—Unconventional Biomass Production . 107

zyme (biological catalyst) which is activated or due to the high energy consumption of the ni- induced by keeping the microalgae in the dark trogenase reaction. Also, the mixture of hydro- without oxygen for a period of time; or 2) gen and oxygen generated by such a system through the “nitrogenase” (nitrogen-fixing) en- may be expensive to separate. zyme which normally allows some types of mi- Whichever biophotolysis mechanisms or croalgae (the “blue-green” algae) to fix atmos- processes are eventually demonstrated to be pheric nitrogen to ammonia but which also capable of efficient and sustained solar energy can be used to produce hydrogen by keeping conversion to hydrogen fuel from water, they the algae under an inert atmosphere such as must take place in a very low-cost conversion argon gas. system. The development of an engineered In the case of biophotolysis with hydroge- biophotolysis conversion unit must meet strin- nase the key problem is that when simultane- gent capital and operational cost goals. As bio- ous oxygen production occurs, the hydroge- photolysis will be limited by the basic proc- nase enzyme reaction is strongly inhibited and esses of photosynthesis— probably no more the enzyme itself inactivated. Although it was than 3 to 4 percent of total solar energy con- recently demonstrated that simultaneous pro- version to hydrogen fuel —this sets an upper duction of oxygen and hydrogen does occur in limit to the allowable costs of the conversion such algae,31 it is uncertain whether it will be unit. In principle, the algal culture—the cata- possible to sustain such a reaction in a prac- lyst which converts sunlight and water to hy- tical system. This difficulty has led to propos- drogen and oxygen–can be produced very als for separation of the reactions either by de- cheaply; however, the required “hardware” to veloping an algal system which alternates oxy- contain the algal culture and trap the hydrogen and hydrogen production, (possibly on a gen produced may be relatively expensive. day-night cycle) or by developing a two-stage Biophotolysis is still in the early stages of process. Such systems are still at the concep- development. No particular mechanism, con- tual stage, although considerable knowledge verter design, or algal strain appears to be in- exists about the basic mechanisms involved. herently superior at this stage. Claims that Somewhat better developed is a biophotoly- near-term practical applications are possible, sis process based on nitrogen-fixing blue-green that genetic engineering or strain selection can algae. I n these algae the oxygen-evolving reac- result in a “super” algae, or that biophotolysis tions of photosynthesis are separated from the is inherently more promising than other bio- oxygen-sensitive nitrogenase reaction by their mass energy options are presently not war- segregation into two cell types — the photosyn- ranted. A relatively long-term (10 to 20 years) thetic vegetative cells and the nitrogen-fixing basic and applied research effort will be re- heterocysts. Heterocysts receive the chemicals quired before the practical possibilities of bio- necessary to produce hydrogen from vegeta- photolysis are established. tive cells but are protected from oxygen by their heavy cell wall and active respiration. Using cultures of such algae from which nitro- Inducing Nitrogen Fixation in Plants gen gas was removed, a sustained biophotoly- The biological process of nitrogen fixation, sis reaction was demonstrated: about 0.2 to 0.5 the conversion of nitrogen gas (not a fertilizer) percent of incident solar energy was converted to ammonia (a fertilizer) has only been found to hydrogen gas over a l-month period. How- to occur in bacteria and the related blue-green ever, significant problems stil exist in the de- algae. These primitive organisms maintain the velopment of a practical system —1 O times ecological nitrogen cycle by replacing nitrogen higher conversion efficiencies must be lost through various natural processes. The achieved, a goal which may not be reached capability for nitrogen fixation expressed by many pIants (soybeans, alfalfa, peas) is due “ E Greenbaum, Bioengineering Biotechnology Symposium, VOI 9, In press solely to their ability to Iive in a symbiotic 108 ● Vol. Ii—Energy From Biological Processes

association with certain bacteria (of the genus rich environment of a plant leaf. In principle, Rhizobia), which form the characteristic “root there would only be a relatively minor advan- nodules.” A certain fraction of the photosyn- tage for a plant to directly fix nitrogen rather thetic products of these plants are transferred than do so symbiotically. Much more basic to the roots where they are used (as “fuel”) by knowledge in many areas of plant physiology, the bacteria to fix nitrogen to ammonia which genetics, biochemistry, etc., as well as devel- is then sent (in bound form) to the protein syn- opments in genetic engineering and plant tis- thesizing parts of the plant. sue culture will be required before the potential for practical applications of such concepts This process is, in principle, energy inten- can be evaluated. sive, with each nitrogen atom (fixed) reducing the biomass production by several carbon atoms (about 2 to 3).32 In practice, significant Greenhouses inefficiencies in the process are often noted, most particularly the recent discovery that It is well known that increasing the C02 con- some Rhizobia bacteria in root nodules waste centration in the air results in significantly im- a large fraction of the “fuel” supplied by the proved plant growth for some plants. Depend- plant in the form of hydrogen gas.33 By using ing on the specific plant and the specific con- Rhizobia strains that can effectively recycle ditions of the experiments, a 50-to 200-percent the hydrogen gas, this loss may be overcome. increase in biomass production has been noted. Greenhouses have the additional ad- Although biological nitrogen fixation can vantages of providing a “controlled environ- substitute, to a large extent, for the fossil-fuel- ment” where pest control, water supply, and derived nitrogen fertilizers currently used in fertilization can be better managed, resulting agriculture, the tradeoff may be an overall in potentially high yields. The higher tempera- reduction in biomass yields. In an era of de- ture in greenhouses allows extended growing creasing fossil fuel availability, such a tradeoff seasons in temperate climates. Greenhouse may be desirable, particularly as the price of agriculture is rapidly expanding thoughout the commercial fertilizers is a limiting economic world to meet the demands of affluent coun- factor in many biomass production proposals. tries for out-of-season vegetables and horticul- However, nutrient recycling could be prefer- tural products. However, the high cost of able to de novo production, as it probably greenhouse agriculture and its high energy would be less costly and energy intensive. consumption make production of staple crops Alternatively to biological nitrogen fixation, unfeasible and proposals for biomass energy thermochemical conversions of biomass to production unrealistic at present. Although synthesis gas and their catalytic conversion to significantly lower cost greenhouse technology ammonia are feasible. Whether this is more fa- is feasible in principle, biomass production vorable both in terms of economics and energy costs in Arizona, for example, would still be 10 efficiency is uncertain. times as expensive as open-field biomass crops 34 A number of scientists have proposed that, grown in the Midwest. A significant inflation through genetic engineering, they could trans- in farm commodity, farmland, and water fer the nitrogen-fixing genes directly to the prices could make greenhouse systems more plant. However, such proposals face technical attractive. At present and in the foreseeable barriers. For example, the nitrogen-fixing reac- future, however, greenhouses do not appear tion is extremely oxygen sensitive and is unlike- economically feasible for bioenergy produc- ly to be able to take place in the highly oxygen- tion.

‘2K. T Shanmugan, F. O’Gara, K. Andersen, and P C. Valen- “L. H. de Bivort, T. B. Taylor, and M, Fontes, “An Assessment tine, “Biological Nitrogen Fixation,” Ann. Rev. Plant Physio/. 29, of Controlled Environment Agriculture Technology, ” report by p. 263,1978. the International Research and Technology Corp. to the National “Ibid. Science Foundation, February 1978. Chapter 5 BIOMASS PROCESSING WASTES Chapter 5.— BIOMASS PROCESSING WASTES

Page Page Introduction ...... , ...... 111 40. Energy Potential From Animal Manure on Wood-Processing and Paper-Pulping Wastes .. ..111 Confined Animal Operations ...... 114 Agricultural Wastes ...... 112 Animal Manure ...... 114 FIGURES

Page TABLES 14. Material Flow Diagram for Forest Page Products Industry...... 111 39. The Ten Major Agricultural Wastes With 15. Total Energy Available From Manure by Potential to Produce Energy...... 113 Farm Size ...... 114 Chapter 5 BIOMASS PROCESSING WASTES

Introduction

There are a number of byproducts associ- wastes of the forest products industries, and ated with growing biomass and processing it the wastes associated with the processing of into finished products. The byproducts that are agricultural products and animal manures. not generally collected in one place, such as Wastepaper, cardboard, and urban wood logging or crop residues, are termed residues wastes are not considered in this report, since and are dealt with in chapters 2 and 3. The by- they fall into the category of municipal solid products that are collected in one place are wastes, which is the subject of a previous OTA termed processing wastes for the purposes of report. this report and are considered in this chapter. The three main types of wastes considered are IMateria/s and Energy From Municipal Waste [Washington, the primary and secondary manufacturing D C.: Office of Technology Assessment, July 1979), OTA-M-93

Wood-Processing and Paper= Pulping Wastes

Based on published surveys and discussions Figure 14 shows an approximate materials with people familiar with the forest products flow diagram for the harvested wood proc- industries, the fraction of wood feedstock that essed by the forest products industry. This is a appears as residue was estimated for the vari- national average diagram. There are, however, ous types of processes and regions of the country. These fractions and the U.S. Department Figure 14.—Material Flow Diagram for Forest of Agriculture’s Forest Statistics* were used to Products Industry (in energy units, Quads/yr) estimate the quantities of residues generated by wood-processing and paper-pulping indus- Forest products tries. There is, however, some uncertainty in industry harvest these figures, since published data usually are 3.1 Pulpwood harvests reported in board feet or cubic feet (rather than dry tons) and often the bark is not counted. Furthermore, moisture loss during drying must be accounted for, Every effort was made to avoid these potential problems and Primary and adjust for the shrinkage. secondary manufacturing 0.7 Paper and pulp 2.0 1.8 Current data on the use of the manufacturing residues are not complete. In some cases Residues of primary data are available for only a few States or for and secondary manufacturing some of the industries. In other published data, regional surveys are extrapolated to the entire country. The estimates presented here are Unused Energy Energy 0.14 0.3 1.0 based on several surveys,3 but are nevertheless based on incomplete data. Total wastes used as energy: 1.2 to 1.3 Quads/yr ‘Forest Statistics for the United 5tates, 1977 (Washington, D C : Unused waste 0.1 to 0.2 Quad/yr Forest Service, U.S. Department of Agriculture, 1978) Energy and unused 1.4 Quads/yr ‘J S. Bethel, et al., “Energy From Wood,” contractor report to OTA, April 1979. SOURCE: Office of Technology Assessment.

111 112 ● Vol. /l—Energy From Biological Processes

significant variations between the regions, manufacturing wastes and 40 percent (4 mil- with the unused fraction being about twice as lion dry ton/yr) of the secondary manufactur- large in the East as in the West. ing wastes go to paper pulping. Another 20 percent of each of these industries’ residues goes The largest user of biomass energy in the to particle board and various other uses. About United States is the pulp and paper industry. 20 million dry ton/yr (0.3 Quad/yr) of wood are This industry is currently 45- to 55-percent used for energy; 9 million dry ton/yr (about energy self-sufficient, up from 37 percent in 0.14 Quad/yr) are unused. 1967.4 A major reason for the use of wood energy in the forest products industry is that the The main reasons that the unused portion is process used to recover the paper-pulping not used appear to be the very low quality of chemicals in most of the pulping processes in- these wastes and a geographical mismatch be- volves burning the spent pulping liquor. This tween the source and potential users of the accounts for about 0.8 Quad/yr. The remaining waste. However, either a strong wood energy 0.2 Quad/yr of bioenergy used in the pulp and market or cooperative agreements with elec- paper industry comes from the bark of the har- tric utilities for cogeneration could bring these vested wood and reject woodchips. wastes into energy use. The primary manufacturing industry pro- There are alternative uses for some of the duces lumber, plywood, poles, etc. The sec- wastes other than for energy. If the demand for ondary industry produces furniture, prefabri- forest products increases and other fuels are cated housing, etc. These industries are 20- to available, then more of the primary and sec- 40-percent energy self-sufficient. s About 50 ondary manufacturing byproduct may be di- percent (40 million dry ton/yr) of the primary verted from energy use to particle board and paper and pulp production. In addition, a

‘E, P Gyftopoulos, L J Lazarides, and T. F Wldmer, Potential small fraction of the spent pulping liquor Eue/ Effectiveness in Industry (Cambridge, Mass : Balllnger Pub- could be used to produce ethanol and Iignin lications). products (as one Georgia Pacific Corp. plant 5S. H Spurr, Renewable Resources for Energy and Industrial Materials (Austin, Tex LBJ School of Public Affairs, University does) instead of simply burning the spent liq- of Texas, 1978) uor to recover the pulping chemicals.

Agricultural Wastes

With the exception of orchard prunings, are produced. Table 39 shows the 10 major agricultural waste byproducts are generally types of agricultural wastes and the energy not collected at the place where the crops are potential of each. These 10 wastes represent grown. Rather, the wastes usually occur as over 95 percent of all agricultural wastes byproducts to the agricultural product-proc- available for energy. Of these 10, about 90 per- essing industries. About 50 to 70 percent of cent are materials relatively low in moisture, these byproducts are sold as animal feed or for and suitable for thermal conversion (combus- chemical production at prices that prohibit tion or gasification). The remaining 10 percent their use for energy. ’ The waste byproducts not appear to be acceptable for anaerobic diges- being used for other purposes are considered tion or possible fermentation to ethanol in the in this section. case of fruit and vegetable wastes and cheese The various agricultural product-processing whey. 7 industries were surveyed to determine the In addition, there is an unknown quantity of quantities and types of waste byproducts that spoiled and substandard grain. One source8 es-

‘R. Hodam, “Agricultural Wastes,” Hodam Associates, Sacra- mento, Calif., contractor report to OTA. OM. T Danz iger, M, P, Steinberg, and A. I Nelson, “Storage of ‘Ibid. High Moisture Field Corn,” Illinois Research, fall 1971 Ch. 5—Biomass Processing Wastes ● 113

Table 39.–The Ten Major Agricultural Wastes With Cotton gin trash is another potential source Potential to Produce Energy of energy. Texas cotton gins produce about Wastes Btu/yr x 1012 five times as much energy in gin trash as they Orchard pruningsa ...... 30-61 consume (mostly electricity). The major prob- Cotton gin trasha ...... 20-31 lems with using the trash for energy seem to be a Sugarcane bagasse ...... 4-8 the difficulty of handling the trash, the season- Cheese wheyb ...... 4-Bc Tobacco (burley)a...... 2.3 al nature of the ginning operations, and the dif- Rice hullsa...... 2.2 ficulty in establishing cooperative ventures b Tomato pumice . ., ., ...... 1.3-1.8 with the electric utilities. In addition, in the Potato peel and pulpb ...... 1.0-1.1 Walnut shella...... , ...... 0.9 areas where the cotton plants are killed with Citrus rag and peelb ...... 0.3-1.0 arsenic acid prior to harvest, such as in much Total ...... , ...... 66-117 of Texas, special precautions will be necessary asultable for combustion or gaslflcatlon to burn the trash in an environmentally accept- bsultable for anaerobtc dtgestlon or fermentation able way. c~sed on starch content of milk and the volume of cheese production frOrTl ~rJrlCUhJh3/ .%?tEflCS (Washmgfon, D C U S Department of Agriculture, 1978) Sugarcane bagasse is widely used in Hawaii SOURCE Off Ice of Technology Assessment, and R Hodam “Agricultural Wastes, ” Hodam Assoclales, contractor report to OTA, 1979 as a source of energy. The sugar refineries have long-standing cooperative agreements with the electric utilities. Cogeneration is used to gener- timated corn spoilage from mold at 250 million ate and export electricity to the utiIities and to bu/yr, but this number should be viewed as produce the process steam used by the sugar speculative. Furthermore, much of the spoiled refineries. The electric generating facilities grain may be accessible only as a supplement range in size from 1.5- to 33-MW electric. Most to existing distillery feedstocks because its oc- of the Hawaiian sugar refineries are 99- to 100- currence is dispersed and unpredictable. percent energy self-sufficient. The four major sources of agricultural The New England and Southern sugar refin- wastes are orchard prunings, cotton gin trash, eries should be analyzed in detail for the po- sugarcane bagasse, and cheese whey. Most tential to duplicate the Hawaii experience, in- States have fruit or nut orchards, with the cluding the potential to purchase orchard largest crops occurring in Arizona, California, prunings or wood wastes which are found in Florida, Texas, New York, and Washington. the same area in some cases. Cotton gin trash is generally localized to the OTA’s analysis indicates that cheese whey is southern third of the United States and Califor- the largest source of food-processing waste nia. Sugarcane is processed primarily along the suitable for conversion to ethanol, although Gulf Coast, in Hawaii, and in New England. other studies have indicated that citrus wastes The majority of cheese whey is produced in are a larger source.9 Based on total cheese pro- Wisconsin, Minnesota, New York, lowa, and duction, 10 OTA estimates that 50 million to 100 California, but 30 States have some cheese million gal/yr of ethanol could be produced production. from cheese whey. Current production from Orchard prunings are generally collected this source is about 5 million gal/yr. and burned onsite. A few growers disk whole ‘The Report of the A/coho/ Fue/s Po/Icy Review (Washington, prunings into the soil, although this is not a D C : Department of Energy, June 1979), GPO stock No 061- preferred practice for growers. With a strong (XX3-OO31 3-4, energy market, much of this could be used for ‘“The Out/ook for Timber in the U.S. (Washington, D C : Forest Service, U S Department of Agriculture, 1974), report No. 24; energy. The major expense is transporting the and Agricu/tura/ Statistics (Washington, D C U S Department of prunings to the place they are used. Agriculture, 1978) 114 ● Vol. II—Energy From Biological Processes

Animal Manure

The major sources of animal manures suit- 1,000 animal units and that about 45 percent able for energy are from dairy cows, cattle on of the potential is on farms with less than 100 feed, swine, chickens (broilers and layers), and animal units. Large feedlots (greater than turkeys. Only animals in confined animal oper- 10,000 animal units) only account for about 15 ations are considered. However, it has been percent of the total. Consequently, any tech- estimated that 48 percent of all manure voided nology development that is aimed at fully uti- from livestock (primarily sheep and cattle), is lizing the potential for energy from animal ma- on open range. ” This open range manure nure will have to concentrate on relatively would require collection and, therefore, will small-scale conversion units. not be economic in the foreseeable future. Figure 15.—Total Energy Available From Manure by The inventory of onfarm confined animals Farm Size (confined animal operations) was derived from inventory numbers for ani- 50 ”/0 mals that remain onfarm for more than a year Percent of energy 16.2% 10,000 + animal units and from sales numbers and the average time I I the animal spends on the farm for animals on 6.6% 40% farm for less than a year. ’z These inventory numbers were converted to the common basis of the number of animal units, or the equiv- 30.9% 100-999 30% alent of a 1,000-lb animal (defined in figure 15). The quantities of manure were calculated and, assuming that the manure is anaerobically di- 20% gested to produce biogas (60 percent methane, 40 percent carbon dioxide), the energy equiv- 41.7% 10-99 alent was derived. 10% Table 40 shows the energy potential from each type of animal operation, and figure 15 4.6% 0.1-9.0 shows the percent of this energy potential that 1 animal unit = is present on confined animal operations of 250 chickens various sizes (expressed in animal units). Cur- 250 broilers 50 turkeys rently most of this manure is used as nitrogen Farm size in animal units fertilizer and soil conditioner or is unused. 1.25 cattle on feed 0.83 dairy cow The total energy potential from manure pro- 6.25 swine duced in confined animal operations is about SOURCE: K D, Smith, J. Philbin, L. Kulik, and D. Inman, “Energy From Agri- 0.3 Quad/yr. From one-third to one-half of this culture: Animal Wastes,” contractor report to OTA, March 1979. manure is currently allowed to wash away with rain or is allowed to dry which makes it unsuit- Table 40.–Enorgy Potential From Animal Manure able for anaerobic digestion. However, if it be- on Confined Animal Operations comes economically attractive to digest the Total energy potential Percent manure, then most of these operations can Type animal Btu x 1012/yr or total change their manure-handling techniques to Dairy cattle...... 90 33 accommodate anaerobic digestion. Cattle on feed ...... 80 30 Swine ...... 32 12 Figure 15 shows that over 75 percent of the Chicken (broilers) ...... 30 11 energy potential occurs on farms with less than Chicken (layers) ...... 25 9 Turkeys...... 18 6 1‘D. Van Dyne and C, C ilbertson, Estimating U.S. Livestock and Total energy potential from all Poultry Manure and Nutrient Production (Washington, DC.: U.S. manures ...... 274 100 Department of Agriculture, 1978), ESCS-12. “1974 Census of Agriculture (Washington, D C : Bureau of the SOURCE K D. Smtth, J Phdbm, L Kuhk, and D. Innran, “Energy From Agriculture. Animal Census, U S. Department of Commerce), vol. 1-50 Wastes, contractor report to OTA, March 1979. Part II.

Conversion Technologies and End Use Chapter 6 INTRODUCTION AND SUMMARY Chapter 6 INTRODUCTION AND SUMMARY

Most bioenergy currently comes from direct Anaerobic digestion is a biological process, combustion of solid biomass for space heating, which produces a gas containing methane (the process steam, and a small amount of electric principal component of natural gas) and car- generation. As chemically stored solar energy, bon dioxide. Suitable feedstocks include many biomass can be converted to a number of gas- wet forms of biomass, such as animal manure eous and liquid fuels, which can be used for a and some aquatic plants. For the near to mid variety of energy purposes not suited to direct term, digesters for onfarm production of gas combustion. from animal manure appear to hold the greatest promise. Not only can this technology serve Thermochemical conversion, or chemical as a waste disposal process, but it also could processes induced by heat, is currently the make most confined animal operations energy most suitable process for the major biomass self-sufficient. There is a need to demonstrate feedstocks–wood and plant herbage. Aside a variety of digesters using different feed- from direct combustion, these processes in- stocks to gain operating experience. Because clude gasification and liquid fuels production. the major cost is the initial investment, pol- Various types of gasifiers are being developed icies designed to lower capital charges will in- which could be used for process heat and ret- crease market penetration of the technology. rofits of oil- and natural-gas-fired boilers. Suit- able gasifiers may be commercially available The alcohols most easily produced from bio- in less than 5 years with adequate develop- mass—ethanol and methanol —are not totally ment support. Methanol synthesis is the near- compatible with the existing liquid fuels sys- term option for liquid fuels production. Wood- tem and automobile fleet. These alcohols can to-methanol plants can be constructed imme- be used in gasoline blends or as standalone diately, while herbage-to-methanol processes fuels, but methanol blends will have more need to be demonstrated. Various other proc- problems than ethanol blends unless suitable esses are being developed, and thermochemi- additives are included with the methanol. All cal conversion of biomass offers considerable of the problems regarding the alcohols’ incom- promise for improved processes and new appli- patibility with the existing system have multi- cations for fuel and chemical syntheses. ple solutions, but it is unclear which strategies will prove to be the most cost effective. Fermentation is the biological process used to convert grains and sugar crops to ethanol – The energy balance for ethanol from grains currently the only liquid fuel from biomass and sugar crops has been the subject of consid- used in the United States. The byproduct of erable controversy, because the farming and distillery grains can be used as an animal feed, processing energy consumption together are thereby reducing the competition between approximately the same as the energy con- food and fuel uses for the grain. Some farmers tained in the ethanol. A net displacement of are producing ethanol onfarm, but with cur- premium fuels—oil and natural gas—can be rent technology the processes are not eco- assured with ethanol, however, if: 1 ) distilleries nomic unless they are heavily subsidized or the do not use premium fuel for their boilers and onfarm production leads to increased grain 2) the ethanol is used as an octane-boosting ad- prices—thereby enabling the farmer to earn ditive to gasoline. Failure to fulfill either of more on the crops he/she sells for feed. How- these criteria could lead to ethanol production ever, process development could decrease the and use increasing the U.S. consumption of costs. Several processes for producing ethanol premium fuels, although there would be a from wood and herbage are being developed, small net displacement of premium fuels in but the costs are highly uncertain. most cases. Failure to comply with both crite-

119 120 ● Vol. n-Energy From Biological processes

ria would almost certainly be counterproduc- biomass chemicals is a widely discussed sub- tive in terms of premium fuels displacement. ject. Numerous plants produce potentially useful chemicals for industrial synthesis and as Methanol and ethanol can be produced a source of natural rubber, mutant cells can from wood and plant herbage, although etha- produce highly specialized chemicals, and nol production is considerably more expensive chemical synthesis from wood and plant herb- with current technology. In each case, how- age is developing or could be developed in a ever, the biomass might be burned or gasified number of potentially very interesting direc- as a substitute for oil or natural gas. Liquid tions. Because of the higher value of chemi- fuels production is considerably less efficient cals, as compared to fuel, the economic limita- than combustion or gasification if the liquid is tions on chemical production from biomass used as a standalone fuel. Using the liquid as are considerably less severe than for energy an octane-boosting additive to gasoline, how- production. ever, makes the options more comparable in terms of premium fuels displacement per ton These topics and related aspects of conver- of biomass. Future developments in refinery sion technologies and end use for bioenergy technology could change this conclusion. are presented in the following chapters. Biomass already supplies substantial quantities of chemicals, and an expanded use of Chapter 7 THERMOCHEMICAL CONVERSION Chapter 7.—THERMOCHEMICAL CONVERSION

Page Page Introduction ...... 123 47 Summary of Cost Estimates for Various Generic Aspects of Biomass Thermochemistry .. .123 Liquid Fuels from Wood via Reactor Type...... 127 Thermochemical Processes ...... 139 Optimum Size for Thermochemical Conversion 48 1979 Cost of Methanol From Wood Using Facilities ...... 129 Oxygen Gasification...... 140 Biomass Densification ...... 130 49 Emission Factors for Residential Wood Direct Combustion of Biomass ...... 131 Combustion Processes ...... 143 Cocombustion of Biomass ...... 131 50 Estimates of Total B(a)P Emissions, . . . . .144 Combustion of Biomass ...... 131 51 “Source-to-Power” Air Emissions for Coal, Wood Stoves and Fireplaces ...... 133 Oil, and Wood Fuel Systems...... 146 Gasification of Biomass ...... 134 B-1 . Electricity From Wood by Direct Airblown Casifier Types...... 135 Combustion...... 152 Efficiency of Airblown Gasifiers ...... 135 B-2. Steam From Wood by Direct Combustion .152 Airblown Gasifier Costs ...... 137 B-3. Electricity and Steam From Wood by Gasifiers for lnternal Combustion Direct Combustion ...... 153 Engines...... 138 B-4. Medium-Btu Gas From Wood in a Dual Liquid Fuels From Thermal Processes ...... 139 Fluidized-Bed Field Erected Gasifier...... 153 Methanol ...... 139 B-5. Medium-Btu Gas From Manure in a Pyrolytic Oil...... 141 Fluidized-Bed Gasifier ...... , . . . .. 153 Ethanol...... 142 B-6. Methanol From Wood Through Environmental lmpacts of Wood and Gasification in a Dual Fluidized-Bed Wood Waste Combustion...... 143 Gasifier ...... 153 Small-Scale Burning...... 143 B-7. Pyrolysis Oil From Wood by Catalytic Utility and Industrial Boilers ...... 144 Direct Liquefaction...... 153 Environmental lmpacts of Cofiring Agricultural B-8. Ethanol From Wood via Gasification in a and Forest Residues With Coal...... 147 Dual Fluidized-Bed Gasifier ...... 153 Environmental impacts of Gasification ...... 148 Research, Development, and Demonstration Needs ...... 149 FIGURES Appendix A.-Optimum Size for a Wood-Fired Electric Powerplant...... 151 Page Appendix B.-Analysis of Break-Even Transport 16. Effect of Moisture on the Heat Content of Distance for Pellitized Wood and Wood ...... 125 Miscellaneous Cost Calculations ...... 152 17. Effect of Feedstock Moisture Content on Appendix C.-Survey of Gasifier Research, Boiler Efficiency...... 126 Development, and Manufacture ...... 154 18. A Comparison of Pyrolytic Weight Loss v. Temperature for Coal and Cellulose ...... 127 19. Differential Loss of Weight Curves...... 128 20. Cost of Electricity From Wood for Various Wood Costs ...... 132 TABLES 21. Cost of Process Steam From Wood for Page Various Wood Costs...... 132 41 . Proximate Analysis Data for Selected Solid 22. Schematic Representation of Various Fuels and Biomass Materials ...... 124 Gasifier Types ...... 136 42. Ultimate Analysis Data for Selected Solid 23. Boiler Efficiency as a Function of the Btu Fuels and Biomass Materials ...... ,124 Content of the Fuel Gas ...... 137 43 Cost of Pelletized Wood...... 130 24. Comparison of Oil/Gas Package Boiler 44 Cost of Electric Generation, Cogeneration, With Airblown Gasifier Costs ...... 137 and Steam Production From Wood . . . . . 132 25. Block Flow Diagram of Major Process 45 Small-Scale Heating Device Efficiency .. ..133 Units...... , ...... 140 46 Cost Estimate for Fuel Gas From Wood B-1. Two Ways to Produce 106 Btu Steam Using a Mass-Produced Airblown Gasifier .. 138 From Wood ...... 152 Chapter 7 THERMOCHEMICAL CONVERSION

Introduction

During the 1980’s, the conversion processes producing methanol from plant herbage can with the greatest potential in terms of both the probably be demonstrated fairly rapidly. More- gross energy use and the largest possible dis- over, there are good theoretical reasons for be- placement of oil and natural gas are the ther- lieving that the flexibility, efficiency, and use- mochemical processes, or processes involving fulness of thermochemical processes can be heat-induced chemical reactions. Currently significantly improved through basic and ap- about 1.5 Quads/yr of biomass are combusted plied research into the thermochemistry of bio- directly for process steam, electric generation mass. (mostly cogeneration), and space heat. lnter- mediate-Btu gasifiers currently under develop- Some generic aspects of biomass thermo- ment will be useful in retrofitting oil- and gas- chemistry and generic reactor types are given fired boilers to biomass fuels and for crop dry- first, followed by a discussion of the optimum ing and other process heat needs. Develop- size of some thermochemical conversion facil- ment of medium-Btu gasifiers is also underway ities and a more detailed consideration of and various processes for producing alcohols select processes including densification, direct and other liquid fuels can be or are being de- combustion, gasification, and direct and in- veloped. Also, methanol synthesis from wood direct liquefaction. Finally, the environmental can probably be accomplished with commer- impacts and research, development, and dem- cially available technology, while processes onstration (RD&D) needs are presented.

Generic Aspects of Biomass Thermochemistry

Possible feedstocks for the thermochemical analyses are shown in tables 41 and 42. As a conversion processes include any relatively fuller understanding of biomass thermochem- dry plant matter such as wood, grasses, and istry is developed, however, new classification crop residues. Some conversion process de- schemes and methods of analysis are likely to signs accept a wide range of feedstocks, while be necessary. others will be more suited to a specific feed- Despite the differences in feedstocks, the stock. Although this is sometimes dependent generic thermochemical process consists of on the chemical properties of the feedstock the following steps: (e.g., manure), it more often depends on the physical properties of the material, such as its ● moisture removal; tendency to clog or bridge the reactor, the ● heating the material to and through the ease with which it can be reduced to a small temperature where it decomposes (about particle size, and the materials’ density. 4000 to 8000 F); ● decomposition to form gases, liquids, and Classification systems that provide informa- sol ids; and tion for assisting the designer of conversion ● secondary gas phase reactions. equipment are not presently available for biomass feedstocks. Standard methods for bio- The drying process absorbs the heat neces- mass analysis or assays do not exist, although sary to evaporate the water. This results in a it is customary to use coal analyses (ultimate decrease in the net usable heat from the feed- and proximate) for biomass. Some of the prop- stock as shown in figure 16. I n this figure, the erties of some biomass materials using coal net heat content per pound of dry wood is

123 124 . Vol. I/—Energy From Biological Processes

Table 41 .–Proximate Analysis Data for Selected Solid Fuels and Biomass Materials (dry basis, weight percent) Volatile matter Fixed carbon Ash Reference Coals Pittsburgh seam coal...... 33.9 55.8 10.3 Bituminous Coal Research 1974 Wyoming Elkol coal, ...... 44.4 51.4 4.2 Bituminous Coal Research 1974 Lignite ...... 43.0 46.6 10.4 Bifumirtous Coal Research 1974 Oven dry woods Western hemlock ...... 84.8 15.0 0.2 Howlett and Gamache 1977 Douglas fir...... 86.2 13.7 0.1 Howlett and Gamache 1977 White fir ...... 84.4 15.1 0.5 Howlett and Gamache 1977 Ponderosa pine...... 87.0 12.8 0.2 Howlett and Gamache 1977 Redwood ...... 83.5 16.1 0.4 Howlett and Gamache 1977 Cedar ...... 77.0 21.0 2.0 Howlett and Gamache 1977 Oven dry barks Western hemlock ...... 74.3 24.0 1.7 Howlett and Gamache 1977 Douglas fir...... 70.6 27.2 2.2 Hewlett and Gamache 1977 White fir ...... , ...... 73.4 24.0 2.6 Hewlett and Gamache 1977 Ponderosa pine...... 73.4 25.9 0.7 Hewlett and Gamache 1977 Redwood ...... 71.3 27.9 0.8 Howlett and Gamache 1977 Cedar ...... 86.7 13.1 0.2 Howlett and Gamache 1977 Mill woodwaste samples -4 mesh redwood shavings ...... 76.2 23.5 0.3 Boley and Landers 1969 -4 mesh Alabama oakchips ...... 74.7 21.9 3.3 Boley and Landers 1969 Municipal rufuse and major components National average waste ...... 65.9 9.1 25.0 Klass and Ghosh 1973 Newspaper(9.4% of average waste) ...... 86.3 12.2 1.5 Klass and Ghosh 1973 Paperboxes (23.4%) ...... 81.7 12.9 5.4 Klass and Ghosh 1973 Magazine paper (6.8%) ...... 69.2 7.3 23.4 Klass and Ghosh 1973 Brownpaper (5.6%)...... 89.1 9.8 1.1 Klass and Ghosh 1973 Pyrolysis chars Redwood (790°to 1,020 F) ...... 30.0 67.7 2.3 Howlett and Gamache 1977 Redwood (800°to 1,725 F) ...... 23.9 72.0 4.1 Howlett and Gamache 1977 Oak (820°to 1,185 F) ...... 25.8 59.3 14.9 Howlett and Gamache 1977 Oak (1,060°F)...... 27.1 55.6 17.3 Howlett and Gamache 1977

SOURCE M Graboskland R Barn, ’’Propertiesof Biomass Relevant lo Gasification, “ in A Swveyo/8/orrrass Gas/ficaf/on (vol. 11: Golden, (Mo, Solar Energy Research Institute, July 1979), TR-33-239

Table 42.–Ultimate Analysis Data for Selected Solid Fuels and Biomass Materials (dry basis, weight percent) Higher heating Material CHNSO Ash value (Btu/lb) Reference Pittsburgh seam coal ., ...... , , , , , . 75.5 5.0 1,2 3.1 4.9 10.3 13,650 Tillman 1978 West Kentucky No. 11 coal ...... 74.4 5.1 1.5 3.8 7.9 7.3 13,460 Bituminous Coal Research 1974 Utah coal ...... 77.9 6.0 1.5 0.6 9.9 4.1 14,170 Tillman 1978 Wyoming Elkol coal ...... 71.5 5.3 1.2 0.9 16.9 4.2 12,710 Bituminous Coal Research 1974 Lignite ...... 64.0 4.2 0.9 1.3 19.2 10.4 10,712 Bituminous Coal Research 1974 Charcoal. , ...... 80.3 3.1 0.2 0.0 11.3 3.4 13,370 Tillman 1978 Douglas fir ...... 52.3 6.3 0.1 0.0 40.5 0.8 9,050 Tillman 1978 Douglas fir bark...... 56.2 5.9 0.0 0.0 36.7 1.2 9,500 Tillman 1978 Pine bark ., ., ...... 52,3 5.8 0.2 0.0 38.8 2.9 8,780 Tillman 1978 Western hemlock...... 50.4 5.8 0.1 0.1 41.4 2.2 8,620 Tillman 1978 Redwood. ... , ...... , . . 53,5 5.9 0.1 0.0 40.3 0,2 9,040 Tillman 1978 Beech...... 51.6 6.3 0.0 0.0 41.5 0.6 8,760 Tillman 1978 Hickory...... 49.7 6.5 0.0 0.0 43.1 0.7 8,670 Tillman 1978 Maple...... 50.6 6.0 0.3 0.0 41,7 1.4 8,580 Tillman 1978 Poplar...... 51.6 6.3 0.0 0.0 41.5 0.6 8,920 Tillman 1978 Rice hulls ...... 38.5 5.7 0.5 0.0 39.8 15.5 6,610 Tiliman 1978 Rice straw...... 39.2 5.1 0,6 0.1 35.8 19.2 6,540 Tillman 1978 Sawdust pellets...... 47.2 6.5 0.0 0.0 45.4 1.0 8,814 Wen et al. 1974 Paper...... 43.4 5.8 0.3 0.2 44.3 6.0 7,572 Bowerman 1969 Redwood wastewood ...... 53.4 6.0 0.1 39.9 0.1 0.6 9,163 Boley and Landers 1969 Alabama oak woodwaste...... , . . . . 49.5 5.7 0.2 0.0 41.3 3,3 8,266 Boley and Landers 1969 Animal waste...... , ...... 42.7 5.5 2.4 0.3 31.3 17.8 7,380 Tillman 1978 Municipal solid waste ... , . . . . 47.6 6.0 1.2 0.3 32.9 12.0 8,546 Sanner et al. 1970 C = carbon H = hydrogen N = nitrogen S = sulfur O = oxygen SOURCE M Graboskl and R Barn, “Properties of Biomass Relevant to Gastf!catlon, ” m A .%rwyof Llornass Gas/f/caf/err, (VOI 11, Golden, Colo Solar Energy Research Institute, July 1979), TR-33-239 Ch. 7—Thermochemical Conversion ● 125

Figure 16.— Effect of Moisture on the out a drop in efficiency, but in practice the ef- Heat Content of Wood ficiency is likely to drop.

9,000 A theoretical example of how the boiler effi- 8,000 ciency drops with feedstock moisture content 7,000 is shown in figure 17. Care should be exercised in applying these results to any given situation, 6,000 since some factors which would vary with 5,000 moisture content (e. g., excess air) are held con- 4,000 stant in the calculations, but it does illustrate 3,000 the point.

2,000 With gasification, the situation is slightly different. In this case the feedstock is decom- o 10 20 30 40 50 60 posed into a fuel gas before combustion. The Moisture content energy needed to vaporize the feedstock mois- (% of fresh weight) ture is still lost, but the fuel gas can easily be

SOURCE: Off Ice of Technology Assessment, mixed with the combustion air, so that excess air is not required, and the feedstock moisture is already vaporized, so the flame temperature shown for various moisture contents. The heat can be high. Consequently, it may be possible content per pound of moist material, however, to maintain the efficiency of gasification-com- decreases much more rapidly with moisture bustion processes over a variety of feedstock content, due to the fact that part of each moisture contents better than with direct compound is water and not combustible material. bustion. Depending on reactor design, how- (Nevertheless, the price of the moist feedstock ever, it may be necessary to limit the feedstock will vary with the moisture content, so that moisture in order to produce a flammable gas, $1 5/ton material at 50-percent moisture con- and this point needs further investigation. tent is roughly equal to $30/ton of dry material. In this report, the feedstock costs are generally The rate that the biomass is heated to and expressed as dollars per ton of dry material, so through its decomposition temperature is a that variations in cost and heat content per ton critical factor in determining the products. are kept at a mini mum.) Many reactor designs are being developed to achieve high heating rates, as described below. There is also a secondary effect of the mois- (The heating rate is also determined by the par- ture content of the feedstock. If moist feed- ticle size— small particles heating faster— and stocks are combusted to produce steam, the moisture content. ) Depending on the products boiler efficiency will usually drop if the feed- desired, however, one may want this heating stock moisture content is not that for which rate to be slower. the boiler was designed. Aside from the heat lost in evaporating the water in the feedstock, The details of biomass decomposition are high-moisture feedstocks have a lower flame not well understood, but one can surmise the temperature in direct combustion, which can following. As the material is heated, the large result in particulate and creosote emissions biomass molecules (cellulose, hemicellulose, (which escape without being completely com- and Iignin) begin to break down into intermedi- busted, if considerable excess combustion air ate-sized molecules. If the material stays in the is not used). (1 n poorly designed wood stoves or heating zone long enough, the intermediate- boilers, simply feeding excess air may not be size molecules decompose into still smaller sufficient to suppress these emission s.) In prin- molecules, such as hydrogen, methane, carbon ciple, a reactor can be designed to accommo- monoxide (CO), ethane, ethylene, acetylene, date this excess combustion air, vaporized and other chemicals. If the heating rate is too moisture, and lower flame temperature with- slow relative to the time the material is in the 126 ● Vol. I/—Energy From Biological Processes

Figure 17.—Effect of Feedstock Moisture Content on Boiler Efficiency

0 10 20 30 40 50 60

Moisture content (o/o)

SOURCE: R. A. Arola, “Wood Fuels – How Do They Stack Up?” Energy and the Wood Products Industry, Forests Products Research Society, Proceeding No. 76-14, NOV. 15-17, 1976. heating zone, the intermediate-sized mole- the pressure, and other variables not fully cules will escape and later condense as oils understood, the resultant gas can vary from and tar. (This may also involve some intermedi- almost pure carbon dioxide (C02) and water to ate reactions that are not well understood at gases with relatively high contents of materials present. ) It also appears that a slow heating such as hydrogen, methane, or ethylene (see rate encourages the formation of char. Thus, a ch. 12), or the gas can contain considerable slow heating rate (either by design or due to ex- quantities of particulate, various hydrocar- cess moisture in the feedstock) will lead to the bons, CO, and other pollutants. formation of varying amounts of char, tar, oil, and gas. With rapid heating, however, virtually Depending on the conditions chosen and the the entire biomass goes to a gas with only the design of the reactor, the product(s) can be ash remaining. heat as in direct combustion, an intermediate- Finally, the gases and vaporized tars and oils or medium-Btu gas suitable for oil- or gas-fired can react in the gas phase to form a new or boilers and process heat, a gas suitable for modified set of products. Very Iittle is under- chemical synthesis, oils, and/or char. But con- stood about these secondary gas phase reac- siderable research into the thermochemistry of tions, but they are of considerable importance biomass will be needed, before engineers will in thermochemical processes. Depending on have the necessary information to design reac- the oxygen and moisture content, the rate the tors that can achieve the full potential for the biomass was decomposed, the temperature, thermochemical conversion of biomass. Ch. 7—Thermochemical Conversion ● 127

Reactor Typ- e

Most commercial biomass reactors used for al’s temperature a given amount)’ coal gasifi- direct combustion or gasification are modification in advanced reactors will ultimately be cations of coal technology. The reactors pro- limited by the rate that oxygen, CO2, steam, posed for direct liquefaction and densifica- etc., can diffuse to and into the surface of coal tion, however, do not fall into this category particles. Biomass gasification and decomposi- and are considered in the sections dealing with tion, on the other hand, do not require the these topics. reaction of two or more separate species. Con- sequently, biomass gasification probably will Although the technology for coal combus- be limited by the rate that heat can be trans- tion and gasification is considerably more ad- ferred to the biomass. vanced than for biomass, it is generally agreed that grasses, wood, and crop residues are more In balance, these differences point to the readily gasified than coal or char. The biomass conclusion that there is the potential for build- gasifies at a lower temperature and over a nar- ing biomass reactors that have considerably rower temperature range than does coal, as il- higher rates of throughput and thus lower costs lustrated in figures 18 and 19, Both of these than will be achieved with coal or has been properties favor rapid gasification. While achieved for either material so far. On the these advantages of biomass over coal are par- other hand, the most rapid heat transfer occurs tially offset by biomass’ higher heat capacity when the feedstock particles are pulverized or (the amount of heat needed to raise the materi- of relatively small size. Most coals can readily be pulverized, but the fibrous nature of many types of biomass makes it difficult to reduce Figure 18.—A Comparison of Pyrolytic Weight Loss the particle size. Biomass densification (see (on a mass fraction basis) v. Temperature below) makes it fairly easy to pulverize the for Coal and Cellulose biomass, but this and other pretreatments add 1.0 to the costs. At present it is impossible to predict whether the difficulty and expense of 0.9 reducing the biomass particle size or the inherent limitations in the rate that coal reacts 0.8 will dominate the economic differences between the two types of fuel reactors. Neverthe- less, it is clear that dramatic improvements in

biomass reactors are possible and that achieving this full potential will require RD&D specifically aimed at addressing and exploiting the unique features of biomass. Furthermore, since biomass char is more like coal than 0.4 wood, grasses, or crop residues, achieving this potential advantage of biomass will involve Cellulose 0.3 reactors that produce Iittle or no char. Generally, the biomass reactors are classi- 0.2 fied according to the way the feedstock is fed into them. Although there are numerous varia- 0.1 tions, the major types are moving grate, mov- 100 300 500 700 900 Temperature, “C ‘M Graboskl and R Baln, “Properties of Biomass Relevant to Caslflcatton, ” m A Survey of L?iorTra55 Gas//ication (VOI 11, SOURCE: M J Antal, Biomass Energy Enhancement—A Report to the Presl. dent’s Counc// on Errv/ronrnental Oua//ty (Princeton, N.J Princeton Golden, Colo Solar Energy Research Institute, July 1979), Unlverslty, July 1978) TR- 13-239 128 ● Vol. n-Energy From Biological Processes

● These curves represent the derivative of curves similar to those given in figure 18. They were obtained by heating a small sample of solid material at a given rate and recording fractional weight loss v. temperature. The peak of each differential weight-loss curve (i.e., for cellulose the value is 15% per 10°C at 315 “C) is indicative of the individual material’s pyrolysis kinetics — a higher heating rate would displace all the curves to higher temperatures and would “sharpen” each peak. Thus the position of each peak is not related to “optimum” operating conditions. The curves simply show that biomass materials pyrolyze much more rapidly at much lower temperatures than coal.

SOURCE: H. H. Lowry, Chemistry of Coal Ufi/ization Supplementary Vo/ume (New York: John Wiley and Sons, Inc. 1963) ing bed, fluidized bed, and entrained flow. The A slightly faster rate of heat transfer is rate of heat transfer generally follows the achieved with moving-bed reactors. In these order given, with the moving-grate reactors be- the bed, or clump of biomass, moves in a ver- ing the slowest. (There are, however, other tical direction as it is decomposed. Additional classification schemes which can be useful as biomass is added at the top, which then grad- well. ) ually works its way down the reactor. Two types of moving-bed reactors exist: updraft Moving-grate reactors consist of a grate that and downdraft. carries or moves the biomass through the zone where it is heated and decomposes. The heat transfer is relatively inefficient and slow, so an The updraft moving-bed reactors have a excess of heat must be generated to sustain the stream of air moving up through the bed of reaction. Therefore this type of reactor is gen- biomass. The hottest part of the bed is at the erally best suited to direct combustion where bottom. As the hot gases move through the the biomass is completely reacted and releases bed, however, they cause relatively large virtually all of its heat in the decomposition amounts of tars and oils to form, which can zone. condense causing maintenance problems and Ch. 7—Thermochemical Conversion ● 129

which may make it more difficult to burn the the advantage, however, of helping to retain resultant gas without forming particulate. heat in the bed, thereby increasing the rate that new pieces of fuel heat up in the bed. The downdraft moving-bed reactors have a stream of air moving downward through the Fluidized-bed reactors have a considerably bed of biomass. Tars and oils are formed near faster heating rate than moving-bed or travel- the middle of the bed (where the air is injected) ing-grate reactors. The churning in the bed, and subsequently move through a relatively however, enables material at all stages of de- large hot zone which gives them time to fur- composition to be found throughout the bed. ther decompose. The net result is a fuel gas Consequently, there may be a tendency for oils with fewer tars and oils, thereby making gas and tars to escape from the heating zone cleanup easier and reducing the amount of before they can be fully decomposed. particulate that form when the gas is burned.

Another type is the fluidized-bed reactor. In The last type of reactor considered here is this case, gas is blown through the bed of solid the suspension or entrained-flow reactor. In fuel so rapidly that the bed of biomass levi- this type, small particles of feedstock are tates and churns as if it were fluid. In coal-fed suspended in a stream of gas which moves rap- fluidized-bed reactors, the fluid bed may con- idly into and through the decomposition zone. tain limestone particles to react with and re- This type has the most rapid rate of heating, move sulfur from the coal. Since biomass usu- but the feedstock particles must be reduced to ally does not contain significant levels of a relatively small size. As mentioned above, sulfur, sand can be used as a fluidizing medi- this would add to the total conversion costs um or one can rely solely on the biomass itself, and the details of this economic tradeoff are with no separate fluidizing medium. Sand has still uncertain.

Optimum Size for Thermochemical Conversion Facilities

Electric generating plants fueled with nu- facilities to the equivalent of 10- to 60-MW clear power or fossil fuels are generally quite electric or less. large in order to take advantage of economies The economy of scale, however, is often of scale. The same is true of most proposals for matched by the cost savings associated with synthetic fuel plants. The optimum size of a mass producing a large number of small units. biomass-fueled electric powerplant or synthet- Furthermore, in many industrial applications ic fuels plant, on the other hand, is determined (e. g., process heat or steam boilers) the size is by a tradeoff between this economy of scale determined by the needs of that industrial and the cost of transporting the feedstock to plant rather than a potential economy of scale the conversion plant. Under favorable circum- for the boiler or heat needs. stances this optimum size could be several hundred megawatts electric (see app. A), and Large-scale facilities are technically feasible some paper-pulping mills do have wood inputs under some circumstances, particularly where that would be sufficient for facilities of this the biomass arises as a waste byproduct in a size. z Under more common circumstances, large manufacturing plant. The number of sites however, the local availability of feedstock where large quantities of biomass are avaiIable may Iimit the size of biomass conversion to a single plant on a continuing basis, however, may be limited. Consequently, the fullest utiIization of the biomass resource for thermo-

2KIp Hewlett, Georg!a Pactflc Corp , private communlcatlon, chemical conversion will require the develop- 1979 ment of small-scale, mass-produced units. 130 ● Vol. n-Energy From Biological Processes

Biomass Densification

Freshly harvested biomass usually contains chips with heat escaping out the burner’s chim- considerable moisture, has a relatively large ney. The exact numbers will vary, however, volume per unit of energy (making it expensive depending on the specific boiler being con- to transport), and is fibrous (making it often sidered. If the boiler is designed to accept difficult to reduce the particle size). These dif- high-moisture woodchips, then there may be ficulties can be partially overcome by densify- no efficiency improvement with wood pellets. ing the biomass. Wood pellet costs are shown in table 43 for There are several types of densification various feedstock costs. While the costs are processes including pelletizing, cubing, bri- quetting, extrusion, and rolling-compressing. Table 43.–Cost of Pelletized Wood Pelletizing typifies the advantages and disad- Wood feedstock cost (dollars/green ton) vantages of densification processes and is con- $6.50 $10.00 $20,00 sidered in more detail here. Dollars/ton of pellets sold Pelletization consists of drying the biomass, Wood. ... ., ...... $14.39 $22.13 $44.26 Operation and heating it until the Iignin melts, and compress- maintenance . . . . . , . . 7.95 7.95 7.95 ing the material into pellets. The pellets are Capital charges...... 5.14 5.14 5.14 denser than the biomass, more easily ground, (30% Of total investment per year) and easier to handle and feed into reactors. Total ...... $27.48 $35.22 Due to their lower moisture content, pellets $57.35 Dollars/ IO” Btu. ... , . . . $1.72 $2.20 $3.58 usually burn more efficiently in boilers than does green biomass. Input: 540 ton/d of wood (50% moisture) Output: 244 ton/d of pellets (10% moisture) for sale and 56 ton/d of At present there are only commercial pelleti- pellets used to fuel the plant. Load 330 operating days per year zation processes for wood. The Iignin content in wood is generally high enough to bind the SOURCE OffIce of Technology Assessment, and T B Reed, et al, “Technology and Economics of Close-Coupled Gaslflers for Relrof!ttmg Gas/Od Combustion Uruts to Biomass Feed- pellets so that no additional adhesives are re- stocks , In Retrof/f ’79, Proceedings 0/a Workshop on M Gas#mal/on, sponsored by quired. Densified crop residues or grasses, the Solar Energy Research Instltule, Seattle, Wash , Feb 2, t979 however, may require the addition of adhe- considerably higher than those for woodchips, sives to achieve the necessary binding strength the pellets’ higher energy density allows them to prevent the pellets from disintegrating to a to be transported at a lower cost than green powder; and the costs for this are uncertain. woodchips. This cost savings in the transporta- The wood pelletization process has an ener- tion pays for the pelletization process if the gy efficiency* of about 90 percent if one starts fuel is to be transported more than 50 to 150 with wood having 50-percent moisture con- miles depending on the transport and wood tent. Furthermore, wood pellets would burn in feedstock costs and the initial moisture con- the boiler depicted in figure 17 to produce tent of the wood (see app. B for details of the steam with an efficiency of about 83 percent calculation). However, this calculation does as compared with an efficiency of 65 percent not include the added cost of transporting very for woodchips with 50-percent moisture. Thus bulky material such as plant herbage where the overall efficiency (50-percent moisture the volume rather than the weight of the woodchips to steam) is increased from 65 per- material determines the transport cost. cent to perhaps 75 percent by including a pel- The most common and least expensive use Ietization process. This efficiency increase of fuelwood, however, is likely to be in the could also be achieved by predrying the wood- region where it is harvested. Consequently, the use of densification processes may be limited.

*E fficlency is defined here as the lower heating value of the On the other hand, the increased ease of han- product divided by the lower heating value of the feedstock dling and burning pellets may make them at- Ch. 7—Thermochemical Conversion . 131

tractive in applications where the process has will have to decide whether the higher fuel to be extremely automated such as in very cost is justified in terms of the labor savings. small industrial applications or where the feed- For the remainder of this chapter, it is assumed stock is particularly unwieldy such as with that raw biomass rather than pellets are being plant herbage. In each application, the user used.

Direct Combustion of Biomass

Biomass can be burned together with coal likely to be higher. Higher coal prices, how- (termed cocombustion) to produce process ever, will make residue cocombustion more at- steam or electricity. Currently however, the tractive largest amounts of energy produced from biomass come from the combustion of wood and Cocombustion can also be used to lower sul- food-processing wastes such as sugarcane ba- fur emissions somewhat. Since the biomass gasse by themselves. Another important use of generally contains negligible amounts of sul- direct combustion is in home heating. Each of fur, the quantity of sulfur being released in the these applications is considered below. combustion (per million Btu of heat) will decrease with the percentage of biomass, typically 20 to 30 percent. The economic savings asso- Cocombustion of Biomass ciated with this will be highly site specific. The Currently, outdated — and therefore unusa- most advantageous situation would be where ble–seed corn is being cocombusted with coal-fired boiler emissions are only marginally coal by the Logansport, lnd., Municipal Utility. above the emissions standards without the use Cocombustion of wood with coal has also of sulfur removal equipment. Since the biomass costs, air pollution benefits, and feed- been successfully demonstrated by the Grand Haven, Mich., Board of Light and Power. ’ And stock availability are site specific, the economics of cocombustion wiII have to be deter- several assessments of the cocombustion of mined through site-specific economic analy- crop residues with coal concluded that it is technically feasible.4 ses. The principal determinant, however, will probably be the availability of a reliable sup- Abdullah has estimated that the added costs ply of low-cost biomass feedstock. at an electric powerplant needed to modify the boilers and handle the crop residues is $0.20 to $0.50/million Btu. 5 Consequently, for coal cost- Combustion of Biomass ing $1.50 to $2.00/miIlion Btu ($30 to $45/ton), crop residues costing $13 to $24/ton would be Direct combustion of biomass for produc- economically cocombusted. Some crop resi- tion of electricity or steam or for cogeneration dues may be available for these prices, but (simultaneous production of steam and either generally delivered crop residue prices are electricity or mechanical shaft power) has commercially ready technology for wood, sugarcane bagasse, and many other feedstocks. IPlerre Heroux, Supplemental Wood Fue/ Experiment, report to Crand Haven Board of Light and Power (Crant Haven, Mich J There are also commercially available suspen- B Sims Generating Stat Ion, 1978) sion burner retrofits for oil-fired boilers of 4.5 ‘See, e g , Wesley t3uechele, D/rect Combustion of Crop /?es- million Btu/hr or larger. The latter retrofitted /dues In Furnace Bo//ers (Ames, Iowa Agriculture and Home E co- nomlcs Experiment Stat Ion), paper No J8791 boilers can return to oil if the biomass feed- ‘Mohammed Abdul Iah, “ E conomles of Corn Stover as a Coa I stock is temporarily unavailable, but they re- Supplement In Steam Electrlc Power Plants in the North Central quire a biomass feedstock that is quite dry (less United States, ” ph D thesis, Agricultural Economics Depart- ment, Ohio State Unlverslty, Columbus, Ohio, 1978 than 15-percent moisture) and relatively small 132 . Vol. //—Energy From Biological Processes

in size (less than 1/8” x 1/2’’ -3/4’’). s A few types Figure 20.—Cost of Electricity From Wood for Various Wood Costs (field-erected of biomass, however, involve special problems generating station) (e.g., the high silica content in rice hulls and residues) and boilers for these are not available.

In many applications today, feedstocks with 40- to 50-percent moisture content are used, resulting in boiler efficiencies of 65 to 70 percent. (The retrofit unit mentioned above, which is restricted to low-moisture feedstock, achieves an estimated 75-percent efficiency). 7 1 There has been little incentive to dry the feed- Low stock in most current applications, since they I usually involve relatively inexpensive waste 20 products. As the use of biomass for direct com- 10 t I bustion becomes more widespread and the average feedstock costs increase, however, pre- 0 $5 $10 $15 $20 $25 $30 drying of the feedstock is likely to be more common. Wood cost (dollars/green ton at 500/. moisture

As with cocombustion, the feedstock cost SOURCE: Office of Technology Assessment. and availability of a reliable supply of the feedstock are major determinants of the eco- Obviously, where the feedstocks can be ob- nomics of using biomass as a fuel. While these tained inexpensively enough, biomass is com- costs vary considerably from site to site, an av- petitive with coal for generating electricity erage feedstock cost of $30/dry ton ($1 5/green and with oil for process steam. 1 n the case of ton) results in the costs of electric generation, electricity, the investment costs are about the cogeneration, and steam production shown in same as for coal-fired powerplants; but wood- table 44. (More detailed cost calculations are fired boilers cost about three times that of oil given in app. B.) The costs for producing only or natural gas boilers.8 electricity or only steam are also shown for various feedstock costs in figures 20 and 21. Wood Stoves and Fireplaces ‘Peabody, Cordon-Piatt, Inc , Wlnfleld, Kan , e g , offers suspension burner retrofits to oll-fired boilers ranging from 45 mll- Wood stoves and fireplaces have long been Ilon Btu/hr and up. The retrofit cost IS slightly higher than for gas- used as a means of space heating in residences, Iflers, but where dry, small particle feedstock (e g , sawdust) IS available at low prices, the system is competitive with fuel 011 but fireplaces are more often used today for Prwate communication with Delvln Holdeman, Solld Fuels Mar- keting Dlvlslon, Peabody, Gordon-Platt, November 1979 ‘A Survey of Biomass Gas/f/cat/on (VOI 1, Golden, Colo Solar ‘Ibid Energy Research Institute, July 1979)

Table 44.–Cost of Electric Generation, Cogenerarion, and Steam Production From Wooda

Wood cost (dollars/ green ton, delivered Product Plant size at 50% moisture) Product cost Electricity 60 MW (field erected) 15 50-70 mill/kWh Steam 50,000 lb/hr (package boiler) 15 $3.50-$6.00/1 ,000 lb Steam and electricity 390,000 lb/hr 15 $4-$6/1 ,000 Ibb 21,4 MW (field erected) 109-30 mills/kWhb

aSee delafls In app B bAs the steam COSI increases, the electric cost decreases SOURCE Ofhceot Technology Assessment Ch. 7—Thermochemical Conversion . 133

Figure 21 .—Cost of Process Steam From Wood for place, through tubes being heated by the fire, a Various Wood Costs (package boiler) or by drawing the combustion air in from outside through tubes that are heated by the fire. Depending on the complexity of the arrange- L ment, the cost can range from as Iittle as $10 to $30 to over $1,000. Wood stoves generally have a higher efficiency than most fireplaces, due to the greater degree of air circulation around and the radiation of heat from the hot stove. In the better wood stoves, the combustion efficiency (amount of heat liberated per pound of wood) is higher than in a fireplace. Often, however, wood stoves do not completely burn the wood gases, leading to deposits of creosote in the flue. The creosote deposits can present a fire o $10 $15 $20 $25 $30 $5 hazard and, at best, need to be regularly Wood cost (dollars/green ton at 500\. moisture) cleaned from the flue. There is no fundamental a reason, however, why these problems cannot .! Ngnlllcant vartatlons (n Installation costs can occur be solved; and research into thermochemistry SOURCE: Off Ice of Technology Assessment. and development of advanced wood stoves are likely to lead to higher efficiencies, greater their recreational or esthetic value. Also many flexibility of operation, and fewer safety prob- fireplaces are inefficient because excess air lems. goes into the fireplace and up the chimney and this air often is drawn into the house through Wood furnaces for centralized heating of a home also have significantly better efficien- cracks in windows and doors. Consequently, while fireplaces do produce some local heat- cies than many fireplaces. Efficiencies as high ing, the overall effect may be a net cooling of as 80 percent have been reported under certain circumstances. 9 The possibility also exists of the house. using wood furnaces as a backup to solar- The efficiency of fireplaces can be improved heated houses. In this case, the heat storage (see table 45) through various methods of cir- system of the active solar heating system culating room air past hot parts of the fire- could be recharged in a few hours and thereby provide space heating for several days with Table 45.–Small-Scale Heating Device Efficiency low solar insolation. Hill has estimated that a wood furnace (300,000 Btu/hr) with hot water Net efficiency storage (500 gal) would cost about $3,000 in- Heat unit (percent) stalled. This, however, should be treated as a Fireplace Masonry – 10- 10% rough estimate and additional work will be Metal prefab, noncirculating -10-10 necessary to establish a more exact cost. Insert or retrofit, circulating ., 40-50 Metal prefab, circulating 10-30 Metal. freestanding 40 Stoves ‘Laatukattlla Oy, Inc , Satamakatu 4, 33201 Tampere 20, Fln- Franklin or fireplace stove 25-45 Iand, sells a YR-60 furnace capable of burning either light fuel 011 Cast iron airtight 50-65 or wood The Fln n Ish Government Cent re for Techn tca I Rewa rch Metal airtight 50-65 (Valtlon Teknllllnen Tutklmuskeskus) has rated this furnace at Box 25-45 793- and 78 8-percent eff Iclency at two-thirds and f Ive-sixths full Circulator, controlled airinlet 40-55 load, respect Ively, when u~lng relatively dry blrchwood as ii fuel, Furnace, convertor or adder 40-60 according to In forrnatlon supplied to OTA by l.aatukattlla Oy, October 1979 SOURCE Auburn Umverslfy Improwng the Ef!lclency Safety and Uhhly of Wood Burning ‘“R C H III, Unlverslty of Maine, Orono, Maine, private com- Umls DOE con!ract reporl DE AS05-77ET 11288 1979 rnunlcatlon, Oct 26, 1979 134 ● Vol. ii—Energy From Biological Processes

In general wood stove and furnace heating ing exclusively, however, is likely to be limited require more labor than oil or natural gas heat. to those people who consider this type of ac- The fuel requires more handling, ashes must be tivity enjoyable or wish to use wood to achieve removed, and the systems must be regularly some degree of energy self-sufficiency. A larg- serviced to maintain efficient and safe opera- er number of people are likely to purchase tion. This is less of a problem if wood pellets or wood stoves as insurance against oil or natural well-dried wood is used or if the wood is used gas shortages or as a supplement to more con- only as a solar backup. The use of wood heat- ventional systems.

Gasification of Biomass

Gasification is the process of turning solid Close-coupled airblown gasifier systems biomass into a gas suitable for use as a fuel or have the potential for higher efficiencies than for chemical synthesis. There are several types direct combustion when a variety of feed- of thermal gasification processes, or gasifica- stocks with different moisture contents are tion induced by heat. Gases produced in blast used (see “Generic Aspects of Biomass Ther- furnaces or by the water gas process are low- mochemistry”), and can be used for process Btu gases (80 to 180 Btu/stdft3). Other gasifiers heat. Moreover, they are likely to be more effi- use pure oxygen and partial combustion of the cient and less expensive, in most applications, feedstock to produce a medium-Btu gas (300 than oxygen-blown or pyrolysis gasifiers (due to 500 Btu/stdft3 ) suitable for regional in- to the energy loss and cost associated with the dustrial pipelines or chemical synthesis. Still added equipment needed to produce oxygen others (pyrolysis gasifiers) provide an external or the external supply of heat). Nevertheless, source of heat to produce a medium-Btu gas for methanol synthesis, these gasifiers would (e.g., dual fluidized bed gasifier described in be necessary. Moreover, there may be some the next section). circumstances where regional industrial natural gas pipelines could be converted wholly or The gasifiers discussed in detail in this sec- partially to gasified biomass. Consequently, tion are the airblown gasifiers. This type blows cost calculations for two medium-Btu gasifiers air through the feedstock to partially combust are included in appendix B. it. The heat generated is used to gasify the remaining material. The resultant gas from up- Airblown (and other) gasifiers have the flex- draft and downdraft airblown gasifiers (termed ibility of being able to be used together with or intermediate-Btu gas) has a lower heat content as a substitute for oil or natural gas in indus- (120 to 250 Btu/stdft3) than with oxygen or py- trial boilers for crop drying, and for process rolysis gasification, due to the dilution effect heat. This means that even where biomass of the nitrogen contained in the air. (Air is feedstocks are not available in large quan- about 78 percent nitrogen and 21 percent oxy- tities, those that are available can be used to gen.) This lower heat content makes the gas un- displace oil and natural gas to the extent of suitable for regional pipeline distribution, but their availability; and (barring regulations pro- it is not a disadvantage if the gasifier is at- hibiting it) the users could return to oil or tached directly to the boiler being fired (so- natural gas if the biomass is temporarily un- called close-coupled gasifier) or used directly available or in short supply. (It should be noted for process or space heat. Gases with heat con- that the suspension burner retrofit mentioned tents of 250 to 400 Btu/stdft3, however, have in the last section also has this advantage but been produced from an experimental fluid- the types of feedstocks it will accept are more ized-bed airblown gasifier, but the gas con- restricted than for gasifies. ) Furthermore, in tains considerable tar and oil. properly designed close-coupled gasifiers, the

‘ ‘Steven R Beck, Department of Chemical Engtneerlng, Texas fuel gas needs only minor cleanup (cyclone Tech Unlverslty, Lubbock, Tex , private communlcatlon, 1979 precipitator and perhaps fiberglass filter). This Ch. 7—Thermochemical Conversion . 135

together with the fact that the volume of fuel struct. Updraft gasifiers tend to produce more gas that needs to be cleaned is less than the ash and tar in the fuel gas than with downdraft volume of flue gas (from direct combustion) reactors, but their construction is the simplest means that the gas cleanup is likely to be less of all gasifiers. Both types require relatively expensive for gasifiers. large feedstock particles so that the gas can flow freely through the bed of biomass. Gasifiers, however, need further development to improve their reliability (particularly The ideal gasifier would be simple to con- with respect to materials clogging), and, in struct and operate, produce no ash in the fuel some cases, to lower the tar and char pro- gas, completely gasify the feedstock (produc- duced. Furthermore, improvements in gasifier ing no char or tar), accept a wide range of efficiency and throughput rates can lower the feedstock sizes and moisture contents, and effective feedstock costs and capital invest- gasify the feedstock rapidly. The downdraft ment, respectively. The types of airblown gasi- and fluidized-bed gasifiers appear to be the fiers, their efficiency, and the costs are dis- most favorable types, but further development cussed below. Finally, gasifiers for internal of all types is required before an unambiguous combustion engines (ICES) are considered choice can be made. In the end it may well be briefly. found that different gasifier types are superior for different feedstocks and applications. A partial list of gasifiers currently under develop- Airblown Gasifier Types ment is given in appendix C. The types of reactors suitable for gasification include updraft, downdraft, fluidized-bed, Efficiency of Airblown Gasifiers and entrained-flow reactors. Each of these types is depicted schematically in figure 22. The heat content of the fuel gas is an impor- The entrained-flow reactor is the fastest of tant consideration in determining the overall these four. It has the disadvantages, however, efficiency of using a gasifier. The Electric that it requires a finely ground feedstock and Power Research Institute has determined the the fuel gas contains considerable ash. If the efficiency of a boiler using gases with various ash is cleaned from the gas by wet scrubbing, heat values, as shown in figure 23. Both the then the wastewater may contain toxic com- sensible heat (gas temperature) and the fuel pounds (e. g., phenol). 2 value of the gas can contribute to this heating value.16 Typical gas values range from 120 to Fluidized-bed reactors can take a wide range 3 200 Btu/stdft from airblown updraft and of particle sizes. In addition the material downdraft gasifiers. Some researchers claim throughput is more than three times as rapid as that the energy content of the gas is increased with the updraft and downdraft gasifiers 13 and and its burning characteristics are improved by the particle stays in the gasifier only minutes14 the presence of pyrolytic oils (incompletely or fractions of a second’ 5 rather than hours decomposed biomass), ” but these oils tend to with the slower gasifiers. Fluidized-bed reac- condense in fuel lines, clog valves, and in some tors release some ash into the gas stream, cases may cause excessive particulate forma- which must be cleaned from it. Tars in the fuel tion when combusted (thereby requiring flue gas can also be a problem. gas cleanup and reducing the combustion effi- The updraft and downdraft gasifiers are the ciency and applicability for process heat). De- slowest, but they also are the simplest to con- termining the optimum gas composition and

“Ralph Overend, “Gaslflcatlon – An Overview, ” In Retrofit “T B Reed, et al , “Technology and Economics of Close- ’79, proceedings of a LVorkshop on AIr Gas/f lcat/on, sponsored by Coupled Gaslflers for Retroflttlng Gas/Oil Combustion Units to the Solar Energy Research Institute, Seattle, Wash , Feb 2, 1979 Biomass Feedstocks, ” In Retrof/t ’79, proceedings 01 a workshop 1 ‘Ibid on A/r Caslffcatjon, sponsored by the Solar Energy Research ln- “lbld stltute, Seattle, Wash , Feb 2, 1979 “Beck, OP clt 1‘Ibid 136 . Vol. n-Energy From Biological Processes

Figure 22.—Schematic Representation of Various Gasifier Types

aNote that other schemes such as moving grate gasifier also exist

SOURCE: From R. Overend, “Gasifiation An Overview, ” Retrofit 79, Proceedings of a Workshop on Air Gasification, Seattle, Wash., SER1/TP-49-183, Feb. 2, 1979. Ch. 7—Thermochemical Conversion ● 137

Figure 23.— Boiler Efficiency as a Function of the Airblown Gasifier Costs Btu Content of the Fuel Gas It has been estimated that oil- or gas-fired boilers can be retrofitted with mass-produced airblown biomass gasifiers for $4,000 to $9,000/ million Btu/hr ($5 to $1 2/lb of steam/h r), with gasifiers ranging from 14 million to 85 million Btu/hr.21 Retrofit costs, however, can vary con-

siderably depending on the difficulty of accessing the boiler and the possible need for an additional building, to house the gasifier. Voss, for example, has estimated the cost at $20,000/ million Btu/hr when new buildings and founda- tions are needed .22 Gas fuel Btu/ft3 The favorable case cost estimates are com-

SOURCE: Fuels From Municipal Refuse for Utilities: Technical Assessment pared with the costs of new oil/gas- and wood- (Electric Power Research Institute, March 1975), EPRI report 261-1, fired package boilers in figure 24 (similar prob- prepared by Bechtel Corp. how to obtain it requires further experimenta- Figure 24.—Comparison of Oil/Gas Package Boiler tion and a better understanding of biomass With Airblown Gasifier Costs combustion chemistry. Nevertheless, some downdraft gasifiers have produced gases approaching 200 Btu/stdft3 from wood with little oil formation, 18 and there appears to be no fun- o Air gasifiers o Oil/gas package boilers ‘\ damental reason why the optimum energy content (see figure 23) with low tars cannot be ---Forest product laboratory summary reached with additional gasifier development, Field-erected The other factor determining the overall ef- wood-fired boilers ficiency of gasifier-boiler systems is the effi- Package wood-fired boilers ciency of the gasifier itself. Since both the sensible heat* and the chemical energy in the gas can be utilized with a close-coupled gasifier, the only gasification losses are the heat radi- ated from the gasifier, that lost during fuel gas cleanup, and the fuel value lost in condensed tars, oils, or char. Gasifiers have achieved efficiencies of 85 to 90 percentl920 and well-insu- lated gasifiers designed to minimize char, oil, and tar formation should be able to reach effi- 10 20 30 40 50 60 80 100 ciencies of 90 percent or better. This would Boiler size raise the overall efficiency of feedstock to (1,000 lb of steam/hr) steam to 85 percent or higher and provide high SOURCE: T B. Reed, D E Jantzen, W P Corcoran, and R Wltholder, “Technol- efficiencies for process heat needs. ogy and Economics of Close. Coupled Gastffers for Retrofitting Gas/Oil Combustion Units 10 Biomass Feedstocks, ” Retrofit ’79, Pro- ceedings of a Workshop orI A/r Gas/~/cat/on, sponsored by the Solar ‘‘j R Gos~, “The Downdraft Gasifler, ” Retrofit ’79, Proceed- Energy Research institute, Seattle, Wash , Feb 2.1979 /rigs of a W’orkshop on A/r Ca\/f/cat/on, sponsored by the Solar Energy Research Institute, Seattle, Wash , Feb 2,1979 ‘Sensible heat IS the energy contained In the gas by virtue of Its be I ng hot, I e , It If the heat that can be sensed or felt directly “Cited In Ibid “Goss, Op Clt “G D Voss, American Fyr-Feeder Engineers, Des Plalnes, Ill , “’Reed, op clt private communication, 1979 138 . Vol. II—Energy From Biological Processes

Iems with installation can occur with these Field-erected gasifiers are considerably boilers as well). It can be seen that the capital more expensive (see app. B). They may be eco- investment for a gasifier retrofit is roughly nomic, however, in cases where very large twice that for a new oil/gas-fired boiler, but quantities of a low-cost feedstock are avail- only two-thirds of that for a new wood-fired able. Alternatively, package gasifiers of sev- boiler. From these preliminary estimates, it ap- eral hundred million Btu/hr could be devel- pears that a new gasifier-oil/gas boiler combi- oped, which, together with smaller gasifiers, nation costs roughly the same as a new wood would cover most situations involving biomass package boiler but more refined data on gasifi- feedstocks. ers are needed before accurate comparisons can be made. Gasifiers for Internal Combustion Engines With costs of $4,000 to $9,000/million Btu/hr Wood and charcoal gasifiers were used dur- and wood fuel at $30/dry ton ($21/air dry ton, ing the 1930’s and 1940’s in Europe to fuel 30-percent moisture) the resultant gas is esti- automobile and truck engines. After some de- mated to cost about $2.70 to $2.90/million Btu velopment, the gasifiers operated satisfactori- (see table 46). In the unfavorable case of ly, but even under favorable circumstances, $20,000/million Btu/hr, the cost could be $3.35/ operation and maintenance required an esti- million Btu with this feedstock cost. mated 1 hour per day of operation .23 Because of this and the 30-percent power loss associated with switching to the gas,24 it is unlikely Table 46.–Cost Estimate for Fuel Gas From Wood Using there would be a large market for gasifiers a Mass-Produced Airblown Gasifier used in automobiles, except under cases of ex- $4,000-$9,000 per 106 treme shortages of gasoline. Gasifiers could, Fixed investment Btu/hr of capacity however, be used to fuel remote ICES for irriga- a Dollars/lO Btu tion water pumping or electric generation. Wood ($21 /ton, 30% moisture, i.e., $30/dry ton). ., ...... $2.38 The principal difference between gasifying Labor, electricity ...... 0.20 Capital charge (30% of fixed investment for close-coupled boiler operation and process per year) 0.150-0.34 heat and for ICES is that the latter application Total ., ...... $2,73-$2,92 requires that the gas be cooled before entering Estimated range ($20-$60/dry ton wood) . $2-$6 the engine and requires particularly low tar Input: 38 to 230 tons of air-dried wood (30% moisture) per day and ash content. The cooling is required to en- Output: 14 to 85 106 Btu/hr of intermediate-Btu gas able sufficient gas to be sucked into the cylin- Load: 330 operating days per year der to fuel the engine and to prevent misfiring.

SOURCE: Office of Technology Assessment The careful gas cleanup is required to prevent fouling or excessive wear in the engine.

These problems were alleviated for charcoal Obviously from table 46, the dominant cost and low-moisture wood by using downdraft is the feedstock cost. If waste byproducts are gasifiers and various gas cooling and cleanup used to fuel the gasifier, the gas could cost less schemes in Europe before and during World than $1/million Btu. For the larger quantities of War 11.25 (Charcoal tended to form more ash, wood, grasses, and residues costing $20 to $60/ while wood more tar, so somewhat different dry ton, the gas price is estimated to range systems were required.) The applicability of from $2 to $6/million Btu. These costs are com- these gasifiers to other feedstocks, however, is petitive with fuel oil at $6.50/million Btu uncertain. ($0.90/gal), but less so with natural gas at about *] Swedish Academy of Engineering, Generator Gas– The $3.50/million Btu. To achieve the full potential Swedish Experience from 7939-1945, Genera lstabens Litograf iska of gasifiers, however, units in the range of 0.1 Anstalts Forlag, Stockholm, 1950, translated by the Solar Energy Research Institute, Golden, Colo , 1979 million to 10 million Btu/hr should also be de- “Ibid veloped. ZSlbld Ch. 7—Thermochemical Conversion • 139

Gasifiers could be used as the sole fuel for petitive with electric irrigation, gasoline, diesel spark ignition (e. g., gasoline) engines or togeth- fuel, and, probably, natural gas. With crop resi- er with reduced quantities of diesel fuel in die- dues costing $30/dry ton, the gas cost with con- sel engines (by fumigation, i.e., replacing the ventional technology is likely to be over $7/ air intake with an air-fuel gas mixture). The en- million Btu, which is competitive with electric ergy lost in cooling the gas and removing the irrigation and will soon be competitive with tar and the added cost of the cooling equip- gasoline and diesel fuel, but is more expensive ment are likely to more than double the gas than natural gas at present. costs over that for close-coupled gasifiers. (This is based on calculations by Reed, z’ in which it is estimated that about half of the Gasifiers suitable for ICES could probably be close-coupled gas energy is sensible heat. The manufactured immediately, but improvements actual value, however, wilI vary with the gas- in the gasifier efficiency and reliability could ifier. ) improve the applicability of gasifiers to ICES for crop irrigation and other uses. The develop- With waste byproducts having no value or ment could parallel the development of other giving a disposal credit, the gas would be com- gasifiers, and improved units could probably

*’Reed, OP clt be available in 2 to 5 years.

Liquid Fuels From Thermal Processes

Numerous liquid fuels can be made from ture. The gas composition is then adjusted to biomass through thermal processes and chemi- the correct ratio of these components and the cal synthesis. The Iiquid fuels considered here resultant gas is pressurized in the presence of a are methanol, pyrolytic oil, and ethanol. Cost catalyst to produce methanol. Finally, the estimates for the production of these fuels are crude methanol may be distilled to produce shown in table 47, with further details given in pure methanol. appendix B. Each of the processes is discussed Methanol can be produced from biomass by below. gasifying the biomass with oxygen or through pyrolytic gasification to produce the CO- hydrogen mixture, with the remainder of the Methanol process being identical to the processes which use natural gas. The oxygen-blown gasifier sys- Methanol (“wood alcohol”) was first pro- tems can be built today, whereas pyrolysis gas- duced from biomass as a minor byproduct of ifiers require further development. charcoal manufacturing. This process for methanol synthesis, however, is no longer eco- Cost estimates for an oxygen-blown gasifier nomic. Most methanol today is produced from used to produce methanol are given in table 48 natural gas. The natural gas is reacted with and a flow diagram of the process is shown in steam and CO2 to produce a CO-hydrogen mix- figure 25. The cost is estimated at $0.75 to

Table 47.–Summary of Cost Estimates for Various Liquid Fuels From Wood via Thermochemical Processes Commercial facilities could Fuel $/bbl $/gal $/millIon Btu be available by. Methanol ., $28-$56 $0.67-$1.33 $10.50-$20.90 Now Pyrolysis oil 30“ 50 0.70-1.20 7 “ 12 Mid to late 1980’s Ethanol ., 23-68 0.55-1.62 6.50-19.10 1990’s

SOURCE Off Ice of Technology Assessment 140 ● Vol. I/—Energy From Biological Processes

Table 48.–1979 Cost of Methanol From Wood $1 .08/gal from $30/dry ton wood, and the capi- Using Oxygen Gasification tal investment is about $2.00 for each gallon per year of capacity, which is somewhat more Fixed investment (field erected) . . ., ., ., $80 million Working capital (10% of fixed investment) ., 8 million expensive than grain ethanol distilleries. Total investment . . , . . $88 million $/bbl Comparable cost calculations are given for a Wood ($15/green ton) ...... 10.35 dual fluidized-bed pyrolysis gasifier in appen- Labor, water, chemicals. ., ...... 1.10 dix B. In this gasifier, the fluidizing medium is Electricity (3.8 kWh/gal, $0.04 kWh) 6.40 Capital charges (15-30% of total heated in one fluidized-bed reactor which investment per year). ., ., ...... $13.80-$27.60 burns biomass and it is transferred to another Total, ., ...... , ., ., ... ., $31.65-$45.45 fluidized bed where it gasifies biomass in the ($0.75-$1 .08/gal) absence of air or oxygen, Although dual fluid- Estimated range:. ., ...... , ., . $28-$ 56/bbl ($10-$30/green ton wood) ($0.67-$1 .33/gal) ized-bed gasifiers are not fully developed, the ($10.50 -$20.90/10 6 Btu) calculations in appendix B indicate that this Input: 2,000 green ton/d of wood (50% moisture) method may produce methanol at somewhat Output: 2,900 bbl/d methanol (40 million gal/yr) lower costs than using oxygen-blown gasifiers, Load: 330 operating days per year principally because it eliminates the equip-

SOURCE Off Ice of Technology Assessment and based on J H Rooker, Davy McKee, Inc., ment needed to produce oxygen. A more accu- Cleveland Ohio private commumcahon May 1980 A E Hokanson and R M Rowell, rate comparison, however, must await devel- Methanol From Wood Waste A Techmcal and Economic Study Forest Products Lab. oratory Forest Serwce, U S Oeparfment of Agriculture general technical report opment and demonstration of dual fluidized- FPL12 June 1977 and E E Badey manager Coal and Biomass Conversion, Oavy McKee Corp Cleveland Ohio, private communication, 1979 bed and other pyrolysis gasifiers.

Figure 25.—Block Flow Diagram of Major Process Units

Wood Wood hand. Wood gasification logs and prep.

oil I Light ends Methanol Crude Methanol Acid gas plant distillation synthesis removal to MeOH synthesis fuel gas

Product Purge gas storage to fuel

SOURCE: J H. Rooker, &fethano/ V/a Wood Gas/ficat/on (Cleveland, Ohio: Davy Mckee, Inc , 1979) Ch. 7—Thermochemical Conversion ● 141

The only part of methanol synthesis, for This could increase the plant cost above that which there is any uncertainty is the operation resulting from the normal diseconomy of of and yield from the gasifier. Oxygen-blown scale, but engineering details and costs are wood gasification can probably be accom- uncertain at present. plished with commercial fixed-bed gasifiers,27 There is little doubt that methanol can be but a large part of the gasifier cost would be synthesized from wood with existing technol- associated with cleaning tars, oils, and other ogy. Since the only uncertainty is with the gasi- compounds from the gas. Consequently, the fier, the cost estimates are probably accurate costs would be reduced somewhat by develop- to within 20 percent. This would put the cost ing advanced oxygen gasifiers that maximize per Btu of methanol from wood at about the the CO-hydrogen yields and reduce the tar and same level as ethanol from grain. However, oil formation. both alcohols are likely to be more expensive With plant herbage as the feedstock, addi- than methanol from coal, due primarily to the tional problems may arise from the handling of economy of scale that can be achieved by this material and possible clogging of the gasi- building very large coal conversion facilities. fier. These problems probably can be solved with a relatively straightforward development of suitable gasifiers. Pyrolytic Oil Methanol yields from wood would vary depending on the type of wood, but have been Pyrolytic oil can be produced by slowly estimated at 120 gal/dry ton in a plant that pur- heating biomass under pressure and in the chases its electricity.28 If the electricity is presence of a catalyst. The pressure suppresses cogenerated onsite the yield would be about gas formation and the catalyst aids the forma- 100 gal/dry ton. ” These yields correspond to tion of the oil. Other possibilities, however, conversion efficiencies of 48 and 40 percent, such as rapid heating and cooling can also pro- respectively. Yields from plant herbage are not duce pyrolytic oils. available, but based on the above efficiencies, The process involving slow heating is cur- they may be 100 or 80 gal/ton depending on rently under development and a pilot plant in whether the electricity is purchased or gener- Albany, Oreg., has produced a small quantity ated onsite. In neither case would additional of oil, following earlier difficulties. The oil is boiler fuel be needed. In theory, however, about 30 percent lower in heat content (per these yields can be increased significantly. gallon) than petroleum fuel oil and it may be Accessing a large part of the potential bio- corrosive but it contains negligible sulfur. The mass resource would be aided by the develop- oil is said to be roughly equivalent to a low- ment of small, inexpensive package methanol grade fuel oil, but further testing is necessary plants. However, because small centrifugal to determine how well the oil stores and what compressors cannot achieve the pressures modifications in boilers may be necessary to needed for methanol synthesis, plants smaller use this oil as a boiler fuel. than about 3 million to 10 million gal/yr of methanol would require a different type of Since the pyrolytic oil is made from feed- compressor, e.g., reciprocal compressor.30 31 stocks that could be used in close-coupled, air gasifiers and would have some of the same ‘7J H Rooker, A4ethano/ Via Wood Gasification [Cleveland, Ohio Davy McKee, Inc , 1979) uses as the gasifier fuel gas, pyrolytic oil pro- I~E E Bailey, Manager, Coal and Biomass conversion, Davy duction should be compared to close-coupled McKee Corp , Cleveland, Ohio, private communication gasifiers. The pyrolytic oil is less expensive to 29A E Hokanson and R M Rowell, “Methanol From Wood Waste A Technical and Economic Study,” Forest Products Labo- transport than raw biomass and it is probably ratory, Forest Service, U S Department of Agriculture, general well suited to fulIy automatic boiler operation. technical report FPL 12, June 1977 It may also be possible to refine the oil to \OBalley, op c it “J H Rooker, Davy McKee, lnc , Cleveland, OhJo, private higher grade liquid fuels. At present, never- communlcatlon, May 1980 theless, the costs appear to be high in relation 142 ● Vol. I/—Energy From Biological Processes

to air gasifiers and the efficiency of using the Consequently, OTA has not analyzed the mul- biomass feedstock in this way is considerably tiproduct liquefaction systems in detail. lower than with gasification, but the oil may be Still another type of liquefaction process comparable in cost to some other synthetic would subject medium-Btu gas to pressure in fuels. Consequently, if gasifiers become widely the presence of a catalyst (the biomass analog available, markets for the pyrolytic oils may be of the South African SASOL process for pro- limited to those users who are willing to pay ducing gasoline from coal). The capital invest- for complete automation of their boilers. ment, however, appears to be quite high,35 and Various other thermal processes are possible further development will be needed to lower for the production of oils from biomass (see these costs. app. C), including processes which do not try to minimize oil production during gasification and collect the oil as one of the products. Ethanol These latter types produce gas, oil, and char Conceptually, ethanol can be produced products. from biomass through rapid gasification to The multi product systems, while being tech- produce ethylene. The ethylene is then sepa- nically easier to develop, have decreased oil rated from the other gases and converted to yields (since part of the biomass is not con- ethanol using commercial technology. verted to oil) and the management and eco- The critical factor in determining the economics are more complicated due to the need nomics is the ethylene yield from rapid gasi- to sell each of the various products. A tech- fication. Present experimental yields have nical solution to these problems being studied reached 6 percent (by weight) from biomass, 36 is to slurry the char with the oil. Although the but some researchers’ believe that yields as char contains ash and the oil is corrosive and high as 30 percent (by weight) may be possible. may deteriorate under storage, the Depart- If so, then this process could produce fuel eth- ment of Energy (DOE) is funding a feasibility anol at prices considerably below those for the study for burning this slurry of oil and char in fermentation of Iignocellulosic materials and gas turbines. j’ Since conventional turbines at costs (per million Btu) comparable to those may not be able to tolerate gases with sodium projected for methanol from coal, or roughly and potassium the project proposes to use tur- $0.65/gal of ethanol. bine combustion technology developed from miIitary programs. 34 The process, however, needs considerable research to determine if and how such ethyl- It would seem, however, to be more tech- ene yields can be achieved. Even under favor- nically and economically sound to develop able circumstances, it is unlikely that commer- conversion processes which produce little or cial processes could be available before the no char and which produce only as much gas 1 990’s. as can be utilized by the conversion facility.

‘]] W Blrkeland and C 13endersky, “Status of Biomass Waste “DOW Ctremlcal, U S A , Freeport, “Technical, Economic, and and Residue Fuels for Use In Directly Fired Heat Engines, ” pre- Environmental Feaslblllty Study of China Lake Pyrolysts Sys- sented at the Conference on Advanced Materia/s for A hernate tern, ” report to the Environmental Protection Agency, 1978 Fuel Capable Direct/y Fired Heat Engines, sponsored by the Elec- “S Prahacs, H C Barclay, and S P Bhada, “A Study of the tric Power Research Institute and the Department of Energy, Possiblllties of Producing Synthetic Tonnage Chemicals From Maine Marltlme Academy, Castine, Maine, August 1979 Llgnocellulosld Residues,” Pulp and Paper Magazine of Canada, J iTeledyne CAE, Toledo, OhIO~ “Gas Turbine Demonstration VOI 72, p 69, 1971 of Pyrolysls-Derived Fuels, ” Department of Energy contract J 7See e ~, M j Anta 1, Biomass Energy Enhancement — A ‘e- E778-C-03-1839 port to the President’s Council on Environmental Qua/it y (Prince- “Btrkeland and Bendersky, op clt ton, N J Princeton Unwerslty, July 1978) Ch. 7—Thermochemical Conversion • 143

Environmental Impacts of Wood and Wood Waste Combustion

The major environmental impacts of wood cent in coal, 38) leads to levels of nitrogen oxide combustion, aside from any impacts from (NOX) emissions well below those of fossil boil- growing and harvesting the wood fuel, arise ers. (Old Environmental Protection Agency from the generation of air pollution in the (EPA) emission factors from “AP-42” showed combustion units. A variety of other impacts, NOX emissions to be as high as those from coal including safety problems with smalI units, boilers, but these factors have been demon- water polIution from wood storage and ash dis- strated to be inaccurate.) Wood sulfur levels posal, and air pollution from wood fuel distri- are equal to or less than 0.05 percent, 39 and bution may be of lesser importance, although sulfur oxide (SOx) emissions consequently are wood appliance safety couId easily become an very low. important public concern. Because the magni- Particulate are an especially worrisome tude of the impacts, even on a “per ton of component of emissions from residential wood wood burned” basis, is quite dependent on the combustion. Areas with high concentrations of size of the operation, this discussion treats wood stoves are known to have particulate residential and other small-scale use separate- pollution problems, especially during winter ly from utility and industrial wood boilers. inversion conditions. Rapid deployment of wood stoves could have significant effects on Small-Scale Burning air quality in New England and the North- west. 40 Residential use of wood as a heating fuel is Condensable organics make up about two- usually a low combustion efficiency, low-tem- thirds of the particulate matter emitted by perature process compared to larger industrial residential wood combustion units .4’ Poly- fossil-fueled or wood boilers. The low combus- cyclic organic matter (POM), species of which tion efficiency is reflected in relatively high are known animal carcinogens, makes up as emissions of CO and unburned hydrocarbons much as 4 or 5 percent of these organics and (see table 49). The low temperature, coupled may be the most dangerous component.42 with extremely low fuel-bound nitrogen in Based on available emission data, POM emis- wood (about 0.1 percent compared to 1.5 per- sions from wood stoves are likely to be far greater (on a “per Btu” basis) than emissions Table 49.–Emission Factors for Residential from the systems they would replace–fossil- Wood Combustion Processes fueled powerplants and residential oil or gas Pollutant g/kg a lb/cord a furnaces. PartlculateC 5-19 20-72 POM is emitted by all combustion sources Carbon monoxide” 60-130 240-520 and is spread throughout the environment, al- Hydrocarbons 2-9 8-40 though usually in low concentrations. Table so So x : 02 0.8

NO X . : : 0.3 12 shows the major sources of benzo(a)pyrene Formaldehyde ., 1.6 6.4 (B(a) P), which is often used as an indicator Acetaldehyde ...... 0.7 3 Phenols ., 1 4 species of POM. Aerosols containing B(a)P and Acetic acid ., ., . . . . 6.4 26 other species of POM can survive long enough Polycyclic organic matter 0.3-4 6% of total particulate Elemental metals . . . . . 7 30 “Comparison of Wood and FossI/ FLJe/s (Washington, D C En- vlronmental Protect Ion Agency, March 1976), E PA-60012-76-056 aunll~ are ~ram~ of Species eml(fed per kilogram of wood burned Wood mols!ure [s not sPeclfled “R H Perry and C H Chlldton, eds , Chemical Eng/neer’s In the references clfed bAlternate units are pounds of species emitted per cord of wood burned One cord IS assumed to Handbook, 5t/I Edition (McGraw HIII, 1973) equal 4000 lb ‘“M. D Yokell, et al , Environment/ Benefits and costs of So/ar cpadlculate Includes lnorganlc ash condensable orgamcs and carbon char Note that other en” Energy, VOI I [draft), Solar Energy Research lnstltute report tries In the table e g polycyclIc orgamc matter and elemental metals, are somewhat redundant SE R1/TR-52-074, September 1979 [n fhat they are subcomponenls of particulate matter and not separate species “j O. Milllken, Environmental Protection Agency, Research SOURCE J O Mllrken Airborne Emlsslons From Wood Combushon Environmental Protec. hon Agency /Research Triangle Park N C Feb 20 1979 with rewslons based on Triangle Park, N C , private communlcatlon, Oct 26, 1979 private commumcahon wNh Mllhken “lbld 144 • Vol. Il—Energy From Biological Processes

Table 50.–Estimates of Total B(a)P Emissions Because POM and other organic emissions, (metric tons/year) as well as CO, are the products of incomplete Major sources Minimum Maximum combustion, the new airtight stoves, which are Burning coal refuse banks...... 280 310 beginning to take an increasing market share, Residential fireplaces ...... 52 110 will have to be evaluated carefully for their Forest fires ...... , ...... , . 9.5 127 Coal-fired residential furnaces ...... 0.85 740 emission characteristics, especially under im- Coke production...... 0.05 300 proper operation. Airtight stoves achieve a

SOURCE Energy and Environmental Analysts, Inc ‘‘Prehmmary Assessment of the Sources, higher overall heating (but not necessarily Control and Population Exposure to Airborne Polycyclic Organic Matter (POM) as indi- combustion) efficiency by slowing down com- cated by benzo(a)pyrene [B(a) P], November 1978 bustion, transferring more of the heat pro- to travel 60 miles (100 km) or farther from their duced into the room rather than up the flue, source .4] However, sources that are far from and avoiding the establishment of an airflow population centers are less dangerous than ur- from the room into the stove and up the flue. ban sources both because of the dispersion The reduction of excess air allowed into the that occurs with distance and because POMs combustion zone increases the emissions of eventually can be degraded to less harmful CO and unburned hydrocarbons. Ideally, these forms by photo-oxidative processes. ” pollutants will be burned in a secondary com- POMs are dangerous for a number of rea- bustion zone fed with preheated air (air that is sons. First, because of their physical nature, first routed through the primary combustion they are more likely than most substances to chamber). However, if the air fed into this zone reach vulnerable human tissues. They are is too cool, secondary combustion will not oc- formed in combustion as vapors and then concur; under these circumstances, airtight stoves dense onto particles in the flue gas. The small- would be substantially more polIuting than or- er particles adsorb a proportionately high dinary stoves. Also, the lower airflow and amount because they have large surface/ cooler exit gases of these stoves cause them to weight ratios. These smaller particles are both deposit more of their organic emissions– in less likely to be captured by particulate con- the form of creosote—on the interior of their trol equipment and more likely to penetrate chimneys. Deposits of creosote from wood deep into the lungs if breathed in. Second, stoves and fireplaces have always been a fire several of the POM compounds produced by hazard; this hazard will be increased by great- combustion are “the same compounds that, in er use of airtight stoves. An added safety prob- pure form, are known to be potent animal car- lem associated with airtight stoves is the cinogens. “45 POM is suspected as a cofactor potential for “back-puffing” —surge back of (contributor) to the added lung cancer risk ap- flames–when the stove is opened. Both of parently run by urban residents. ” Finally, POM these safety hazards are controllable by, re- is suspected of causing or contributing to spectively, having the flue cleaned regularly added incidence of chronic emphysema and and increasing the intake airflow before open- asthma. 47 ing the stove. “C Lunde and A Blorjeth, “PolycyclIc Aromatic Hydrocar- bons In Long-Range Transported Aerosols, ” Nature, 268, 1977, Pp 518-519 Utility and Industrial Boilers “M J Svess, “The Environmental Load and Cycle of Polycy- CIIC Aromatic Hydrocarbons, ” The Science of the Tota/ Environ- Large wood-fired boilers should be more ef- ment, 6, 239, 1979 ficient energy converters than small units and 4’J O Milliken, “ Airborne Emissions From Wood Combus- therefore should have less problems with CO tion, ” presented at the Wood Heating Seminar IV, Portland, Oreg , sponsored by the Wood Energy Institute, Mar 22-24,1979 and unburned hydrocarbons. However, the po- “J O Mllllken and E G Bobaleck, Po/ycyc/ic Organic Matter: tential exists to generate significant quantities Review and Analysis (Research Triangle Park, N C Special of these pollutants, and some existing large Studies Staff, Industrial Environmental Research Laboratory, En- vironmental ProtectIon Agency, 1979) boilers are fairly inefficient and thus fulfill this “K L Stemmer, “Cllnical Problems Induced by PAH,” In Car- potential. (For example, emissions of CO from cinogenesis, Volume 1. Polynuclear Aromatic Hydrocarbons: industrial boilers range from 1 to 30 g/kg of Chem~stry, Metabolism, and Carcinogenests (New York. Raven Press, 1976) wood, compared to 60 to 130 g/kg from small Ch. 7—Thermochemical Conversion • 145

wood stoves. )48 Inefficient boilers will generate current use suffer from a severe drop in effi- the same dangerous organic compounds— in- ciency in controlling the finer, more dangerous cluding species of POM — as do small residen- particles. Baghouses appear to be the only fea- tial stoves and fireplaces. These organics are sible control devices currently available that mostly “low-molecular-weight hydrocarbons are capable of collecting particles below a few and alcohols, acetone, simple aromatic com- microns in size with 99-percent efficiency or pounds, and several short-chain unsaturated greater. It appears quite probable that emis- compounds such as olefins.49 Some of these sion standards for the finer particulate even- emissions are photochemically reactive, al- tually will be promulgated; these standards though the amounts in question should not would almost certainly lead to extensive use of contribute significantly to smog problems. As baghouse controls. the price of wood and wood waste increases, Current EPA emission factors show NOX strong incentives for greater combustion effi- emissions from wood combustion to be com- ciency should work to minimize the organic parable to emissions from coal combustion.52 emission problem. If these factors were correct, large boilers subject to Federal new source performance stand- Sulfur dioxide (SO ) emissions should be 2 ards would require NO reductions of 40 per- minimal because of wood’s low sulfur content. X An exception to this is the combustion of black cent. This would pose a problem in the short term, because there is virtually no experience liquor in the pulp and paper industry; some of

in reducing NOX emissions from wood-fired these boilers should require SOX scrubbing under Federal regulations. 50 boilers. Techniques used for fossil fuel boilers that may be applicable to wood are: Although particulate emissions generated by ● low excess air firing, wood-f i red boiIers can be high (6 g/kg, or about . staged combustion, and as high per unit of energy as a wood stove), ● flue gas recirculation. efficient controls are available for the larger Recent measurements conducted by Oregon units. Available devices or combinations of de- State University” and TRW54 show actual NO vices include multi cyclones coupled with low- X emissions from test boilers to be one-third or energy wet scrubbers, dry scrubbers, electro- less than those predicted by using the current static precipitators (ESPs), or baghouses (fabric emission factors. These measurements are filters). Although ESPs are the most widely much more in Iine with the lower combustion used control mechanism for utility boilers, temperatures in wood boilers and wood’s low they have been said to be less practical for nitrogen content, EPA and DOE researchers wood-f i red boilers because of the very low re- are convinced that the current emission fac- sistivity of both the flyash and unburned car- tors are in error55 56 and it appears likely that bon particles from wood combustion .5’ How- the factors wilI soon be revised. ever, ESPs have been successfully used on In the past, wood boilers have never at- some wood-fired boilers, and the problem of tained the size normally associated with large low resistivity apparently can be handled with coal-fired boilers, Whereas coal-fired utility appropriate precipitator design. boilers are typically a few hundred megawatts Current regulations for emission control 52 from pollution sources do not distinguish par- Cornpl/ation of Air Po//utant Em/ss/on factors Rev/seal (Wash- ington, D c Office of Alr Programs, E nvlronmental Protect Ion ticulates by their size. Most control devices in Agency, February 1972), publication No AP-42 “Memorandum from Paul A Boys, Alr Surveillance and lnves- tlgatlon Section to George Hofer, Chief, Support and Special 4* Mllllken, “Airborne E mlsslons From Wood Gaslflcatlon, ” op Projects Sect[on, U S Environmental, ProtectIon Agency, “Com- c It parison of Emlsflons Between Oil Fired Boilers and Woodwaste “M D Yokell, op clt Eloilers, ” November 3, 1978 Document, 50~n “Iron menta / ~eadjness Wood COmmercia/iza- “J O Mllllken, Envlronrnental Protect Ion Agency, Research rIorI (Department of Energy, 1979), draft Triangle Park, N C , private communlcatlon, June 6, 1979

5 ’ Wood Combu~t/on Systems An Assessment of Env/ronmenta/ “Mllllken, Oct 26, 1979, op clt Concerns (Mlttelhauser Corp , July 1979), draft, contractor report “J Harkness, Argonne National Laboratory, private communl- to Argonne National Laboratory catlon, Oct 26, 1979 146 . Vol. II—Energy From Biological Processes

in generating capacity and range up to 1,000 diffusion of the emissions from the plants; a MW, a 25- or 50-MW wood-fired boiler would higher percentage of the pollution will fall out be considered extremely large. near the plants than would normally be expected for large generating facilities or in- Higher capacities would require using wood dustrial boilers. Also, the high water content of suspension firing, analogous to firing with pul- wood leads to higher concentrations of water verized coal, or fluidized-bed combustion. Pul- vapor in the stack gases and greater visibility verizing wood to extreme fineness for suspen- of the plumes. Although not harmful except in sion firing may be costly enough to offset an esthetic sense, this increased visibility may other economic advantages of going to larger lead to added local objections to wood-fired size plants, so future increases in wood boiler boilers. size may depend on further development of fluidized-bed combustion. The expense of In general, emissions from other portions of transporting wood considerable distances has the fuel cycle are quite low compared to emis- also been a constraint on boiler size in the sions from combustion. The single exception is past, but rising costs for alternative fuels may CO, which is produced in substantial quanti- make longer distance transport of wood more ties by harvesting, chipping, and transport attractive, increasing the effective radius of equipment. Table 51 presents a comparison of supply and the maximum practical size of the the emissions at all stages of the fuel cycle for boiler. coal, oil, and wood boilers. As noted above, The local impacts of utility or industrial CO and organic emissions from wood boilers wood-fired boilers will be moderated by their are far higher than emissions from coal. Note comparatively small size. However, the effects that the emissions of SO2 and particulate are of low stacks (compared to the stacks on large dependent on the level of control, and can be coal-fired utility boilers) will be to allow less reduced significantly if required.

Table 51 .–”Source-to-Power” Air Emissions for Coal, Oil, and Wood Fuel Systems

Emissions ton/yr (basis 50-MW plant) SO a Fuel/energy system 2 co Particulate Total organic Low-sulfur Western coal Surface mining ...... – — 113.1 — Rail transport (1,800 miles), ...... 20.2 218.1 21.8 22.2 Power generation...... , ., . . 2,664.8 87.2 113.1 25.8 Total ...... , ...... 2,685.0 305.3 248,0 28.0 Crude oil Domestic oil pipeline...... 25.8 0.0 3.2 0.5 New Jersey refined with desulfurization . . . . 193,8 4.8 3.2 40.4 Rail transport (300 miles) ...... 1.5 16.3 1.6 1.7 Power generation . . ., ., ...... 854.3 – 80.8 16.2 Total ...... , 1,075.4 21.1 88.8 58.8 Wood Wood recovery...... 6.5 48.5 3.2 8.1 Process chipping ...... 14.5 116.3 6.5 19,4 Truck transport (60 miles)...... 4.4 36.3 2.1 6.0 Power generation...... 119.5 398.9 339.2 398.9 Total ...... 144.9 600.2 351.0 432.4

NOTE NOX levels may be slgmflcant for wood fuel There IS Inadequate data on NOX emfsslon levels There are also production tradeoffs for various con. version systems asoj emlsslons from coal-fired powerplant assume no scrubbers 90% control required by new source performance standards would lower emissions from 2,6648 tons 10266 Ions SOURCE E H Hall, et al , Corrrpar/sorr 0( FossI/ arrd Woud fuels (Washington, O C Enwronmentai ProtectIon Agency, March 1976), EPA-600/2- 76-056 Ch. 7— Thermochemical Conversion ● 147

Environmental Impacts of Cofiring Agricultural and Forest Residues With Coal

Cofiring of coal and agricultural or forest cific factors. However, there appears to be residues has been proposed both as a means of some potential for increased particulate emis- expanding energy supply and as an economical sions under certain conditions. Although bio- way to lower sulfur emissions (from burning mass residues generally have lower inorganic local high-sulfur coal) without importing low- ash contents than the coal they would replace, sulfur coal. Since wood and most crop residues they tend to generate more organics in particu- have very low sulfur contents (cotton gin trash late form. The ability of the boiler to maintain is one exception), total SO2 emissions can be nearly complete combustion conditions will significantly lowered if the residues can re- thus strongly affect particulate emissions. In place a large fraction of the coal normally large facilities with ESPs, the lower resistivity burned in the boilers. Two situations where of the particles generated from combustion of cofiring would appear to be attractive are: the residues may allow a higher percentage to escape control. If the biomass is fed moist into ● reducing SO emissions from existing 2 the boiler, the steam generated during com- coal-fired powerplants that are marginally bustion will increase the flow of hot combus- out of compliance with their State imple- tion gases and conceivably may lead to more mentation plans, and entrainment of bottom ash and higher particu- ● allowing very high-sulfur coals to be late emissions. On the other hand, if the bio- used with scrubbers in new powerplants mass is first artificially dried, particulate emis- (achievement of the current 1.2 lb/million sions from the dryer could be high unless they Btu SO, standard may be difficult with are carefully controlled. The significance of some very high-sulfur coals) any of these effects is uncertain at the present There are few examples of cofiring experi- time. ments in the literature and these examples The importance of these emission changes generally do not examine emission changes depends on the original quality of the coal, the caused by the addition of crop and wood nature of the residues added, the percentage wastes to the coal fuel. Because SO is the only 2 fuel mixture, the type of pollution controls on pollutant whose formation generally does not the boiler, and its operating conditions. All of vary with combustion conditions (except that these factors vary considerably from site to sulfur may be captured in the char from a site. However, it seems Iikely that emission in- pyrolytic reaction), it is probably the only creases will be small except in cases where the pollutant that can be predicted reliably at this cofiring seriously degrades the operating char- time. However, general emission trends for acteristics of the boiIer (it is unIikely that cofir- some pollutants can be predicted. For exam- ing would continue under such conditions un- ple, hydrocarbon and CO emissions may in- less noneconomic pressures– such as the pos- crease slightly, because combustion tempera- sibility of adverse publicity and/or embarrass- tures are lowered and complete combustion is ment of company management— prevented more difficult to achieve when residues are cessation of operations). In addition, emissions added to the boiler fuel. The lower combustion changes will be limited by constraints on the temperature and low fuel-bound nitrogen in amount of biomass that can be mixed with the the residues should cause NO emissions to be X coal. Logging residues and high-moisture crop lowered. If dryers are used for high-moisture- residues have a considerably lower energy con- content residues, their emissions must be tent per unit volume than coal, Because boiler added to those of the boiIer. systems are sized to allow a certain volumetric Particulate emissions are difficult to predict flow rate of fuel feed, a high percentage of bio- because they are affected by several site-spe- mass volume in the feed wiII limit boiIer out- 148 ● Vol. Il—Energy From Biological Processes

put capacity. An additional limit may be pre- content of the biomass may cause condensa- sented by the additional volume of combustion problems in the stack unless the biomass tion gases that would be generated if the bio- content is limited, stack temperatures are in- mass is fed moist into the boiler. These con- creased (by removing less energy from the gas straints do not apply when the energy output and thus lowering system efficiency), or the required is much lower than the boiler’s rated biomass is first dried (which may also lower capacity, or when the system is specifically system efficiency). designed for cofiring. Also, the high-moisture

Environmental Impacts of Gasification

Gasification technologies have a number of hydrogen sulfide (H2S), and hydrogen cyanide potential air and water impacts. Because few (HCN). Other products of the gasification proc- such gasifiers are in operation, quantification ess include carbonyl sulfide (COS) and carbon of these impacts is premature. The low concen- disulfide (CS2) as well as phenols and poly- trations of trace metals and sulfur in the bio- nuclear aromatic (PNA) compounds.58 Gasifi- mass feedstocks and the lack of extreme tem- cation processes that are closer in their nature perature and pressure conditions imply that to pyrolysis and that produce considerable by- impacts should be substantially less than those product char will have lower nitrogen and sul- associated with coal gasification. However, fur-derived emissions; about half of the origi- scientists working for DOE’s Fuels from Bio- nal sulfur and nitrogen in the biomass should mass Branch profess to be unsure as to remain in the char. 59 whether this supposed biomass “advantage” The gas produced will either be burned on- actually exists, especially in the water effluent site (producer gas) or cleaned and upgraded to stream; although the hydrocarbons present in pipeline gas. Either process should eliminate or biomass gasification wastewater should be reduce most of the more toxic pollutants, with more amenable to biological treatment than the onsite burning oxidizing them to CO , SO , coal gasification hydrocarbons (they are more 2 2 NO , and water. Recent tests of a close-cou- oxygenated), they may be produced in greater 2 pled gasifier/boiler combination using wood- quantities and have a higher biological oxygen chips for fuel showed emissions of CO, particu- demand than those of a coal system.57 Also, Iates, and hydrocarbons–which are of major the potential for proliferation of small-scale concern in wood combustion — to be well be- biomass gasifiers may present monitoring and low emissions expected from a direct-fired enforcement problems that would not exist wood boiler, although a fuel oil boiler re- with a few large coal gasifiers. Therefore, bio- placed by such a gasifier would have had con- mass gasification may require as much atten- siderably lower particulate and hydrocarbon tion and concern as coal gasification. emissions. NOX emissions from the gasifier/ The quantity and mix of air pollutants pro- boiler combination were lower than those ex- duced by biomass gasification plants will de- pected from either oil- or wood-fired boilers. 60 pend in large part on the combustion/gasification conditions maintained as well as the environmental controls and the chemical make- “Solar Program Assessment: Environmental Factors, Fuels From up of the feedstock. For example, the concen- B/ornass (Washington, D C Energy Research and Development tration of hydrogen in the reaction chamber Administration, March 1977), ERDA 77-47/7 “lbld and of suIfur and nitrogen in the feedstock wilI “’Ca[ifornla Alr Resources Board, “source Test Report NO C-G-

influence the formation of ammonia (NH3), O(I2-C, Source Test of Exhaust Gas From a Boiler Fired by Produc- tion Gas Generated From an Experimental Gaslfler Unit Using ‘7 Richard Doctor, Science Appllcatlons, Inc , prwate commu- Wood Chips for Fuel, ” Stationary Source Control Dlvlslon, nlcatlon, November 1979 March 1978 Ch. 7—Thermochemical Conversion ● 149

Wood Oil trace elements, ash, and sulfur that are sub- Gasifier boiler boiler stantially lower than concentrations found in Carbon monoxide (lb/hr). 0 1.8-54 0.33 coal, combustion of the char should emit Particulates (lb/hr) 070 45-135 013 Hydrocarbons (lb/hr) 0.90 63 007 lower concentrations of related pollutants Nitrogen oxides (lb/hr) 039 9 1.46 than would coal combustion. Depending on the farming and harvesting techniques, how- These results cannot be readily extrapolated ever, the feedstock may be somewhat contami- to other situations, but they imply that the use nated with pesticides, fertilizers, and soil, of gasifiers may offer a less polIuting alter- which should add to combustion pollutants, native to direct combustion of biomass when a Also, some forms of biomass—for example, shift to renewable (from oil) is being con- cotton trash, with 1.7 percent— have sulfur templated. levels comparable to levels in coal. Leaks of raw product gas represent a poten- Aside from water impacts caused by con- tial for significant impacts, especially on those struction activities and leaching from biomass in the immediate vicinity of the gasifier. The storage piles, gasification facilities will have to probability of such leakage is not known. Al- control potential impacts from disposal and though impact analyses of high-pressure coal storage of process wastes and byproducts. gasification technologies have identified fugi- Water initially present in the feedstock and tive hydrocarbon emissions as a likely prob- that formed during the combustion accompa- lem, it is not clear that similar problems would nying gasification should provide significant occur with (lower pressure) biomass gasifiers. amounts of effluent requiring disposal (al- The combustible char produced by the gasi- though in close-coupled systems, the moist fication process is another potential source of low-Btu gas may be fed directly into the boil- air pollution. It may be used as a fuel source er). Air polIutants identified above may appear elsewhere or else used to heat the bed in a flu- also as water contaminants: NH3 (as ammoni- idized-bed gasifier. In either case, its combus- um hydroxide), HCN and its ionized form, phe- tion will produce NOX, flyash, and SOx as well nols, and trace elements found in the ash. as trace metals either adsorbed on the flyash Leaching from byproduct chars may be a prob- (potassium, magnesium, sodium, iron, boron, lem if the char is (incompletely carbonized) barium, cadmium, chromium, copper, lead, brown char although (carbonized) black char strontium, and zinc) or in gaseous form (beryl- shouId be similar to charcoal and far less likely lium, arsenic compounds, fluorides).61 62 to be polluting. Finally, the tars produced by gasification may well be carcinogenic; as yet Because most biomass feedstocks used in no data confirm this potential. These water gasification processes have concentrations of contaminants present a potential occupational as well as ecological and public health con-

6’So/ar Program Assessment, op clt cern, because plant operators may be exposed “Doctor, op clt unless stringent “housekeeping” is enforced.

Research, Development, and Demonstration Needs

Thermochemical conversion includes the ● Thermochemistry of biomass. — Basic and least expensive, near-term processes for using applied research into the thermochemis- the major biomass resources — wood and plant try of biomass, including secondary gas herbage. Moreover, R&D is likely to lead to in- phase reactions, is needed to better define teresting new possibilities for the production the possibilities for fuel synthesis and to of fuels and chemicals from biomass. Some of aid engineers in designing advanced reac- the more important areas are: tors. The research should include studies 150 ● Vol. n-Energy From Biological Processes

of the effects of the various operating pa- to ensure the development of appropriate con- rameters on the nature and composition trol technologies and incorporation of environ- of products and ways to maximize the mental considerations in system design, siting, yields of various desirable products such and operation. In general, the larger scale tech- as CO and hydrogen, ethylene, methane, nologies are Iikely to be assessed as part of and other Iight hydrocarbons. normal EPA and DOE environmental pro- ● Gasifier development and demonstration. – grams. The smaller technologies generally will Gasifiers should be developed further and not come under Federal new source perform- demonstrated, so as to improve their relia- ance standards (specifications of allowable bility, efficiency, and flexibility with re- emissions), but there is growing recognition in spect to feedstock type and moisture con- EPA and DOE of the potential environmental tent. This should include airblown gasifi- dangers of small-scale technologies such as ers for process heat, boiler retrofits, and wood stoves. ICES, and oxygen-blown and pyrolysis gas- Key environmental R&D areas in thermo- ifiers for methanol synthesis. It should chemical conversion are: also include the demonstration of gasifiers suitable for converting pIant herbage ● development of wood stove designs (or to methanol and should investigate the controls) that achieve complete combus- tradeoff between densifying herbage be- tion and minimize emissions of unburned fore gasification versus gasification of hydrocarbons; herbage directly. Each of the uses for gasi- ● development of combustion controls that fiers will have unique requirements, which will allow efficient — and pollutant mini- probably will dictate separate develop- mizing— thermochemical reactions re- ment and demonstration efforts. gardless of feedstock characteristics; ● Compressor development.– One of the ● assessment of the potential health effects major costs of producing methanol in of emissions from wood stoves and other small plants is the relativley high price of biomass conversion technologies, with a small compressors. The cost of methanol focus on particulate with a high un- synthesis from biomass would be lowered burned hydrocarbon component; substantially if small, inexpensive com- ● evaluation of toxicity and carcinogenicity pressors suited to the process are devel- of biomass gasifier/pyrolysis tars and oils; oped. and Each of the new biomass conversion tech- ● design of controls for gasifier/pyrolysis ef- nologies will require environmental assesment fIuent streams. Ch. 7—Thermochemical Conversion . 151

Appendix A. —Optimum Size for a Wood-Fired Electric Powerplant

The annual cost, C, of producing electricity in a Taking the derivative of C with respect to S and set- wood-fired electric powerplant can be expressed ting it to zero (in order to find the minimum cost as: per kilowatthour) yields (1) c = cc + Cf + ct (9)

Where CC represents the capital and other fixed charges, C represents the fuel and other variable f This represents the optimum size for the power- costs, and C the cost of transporting the fuel. t plant. Letting S represent a dimensionless scaling pa- Evaluating the parameters for the base case rameter o 0.7 given in appendix B results in: c = C c S (2) c S = 11.7 Q06)5 (lo) O Here CC represents the fixed charges for a base where Q is in dry tons per acre year, the base case case and it is assumed that these charges scale with corresponds to a 62-MW powerplant, and the trans- a 0.7 scaling factor. Furthermore, the variable costs port costs are assumed to be $0.20/dry ton-mile are: o ($0.10/green ton-mile). Cf = Cf S [3) As expected, the higher the density of biomass O Where Cf represents the base case. availability, e, the larger is the optimum-sized pow- Assuming the fuel is collected from a circular erplant. IronicalIy, however, as Q increases, the av- area surrounding the powerplant, the transport erage transport distance decreases. I n other words, costs can be expressed as: it is more economic to keep the powerplant size O C t = C t Q o Sr (4) smaller than to transport large quantities of wood O Where Ct is the transport cost per ton-mile, Qo is for greater distances. the annual quantity of wood transported in the If one assumes that Q = 0.5 dry ton/acre-yr, then base case, and r is the average transport distance. the optimum powerplant size is over 500 MW and For a given scaling parameter S, the quantity of the radius of the collection circle is about 50 miles. wood transported is: With Q = 0.05 dry ton/acre-yr, the optimum size is 2 (5) 110 MW and the circle radius is 75 miles. If the Q = Q O S = enr where e is the average availability of fuel wood col- transport charges double, then for these values of lected in dry tons per square mile year and r is the e, the optimum sizes are reduced to 200 and 50 radius of the circle from which wood is collected. MW, respectively, with collection radii of 30 and 50 If one assumes that the actual transport distance miles, respectively, from a harvest site to the powerplant is J2 times In principle, then, large-scale biomass conver- the direct Iine distance, sion facilities are not unrealistic. The values assumed for Q are probably less than what can be achieved in a region where the infrastructure for fuelwood harvests is fully developed. In practice, however, it is likely to be difficult to develop a mature harvest-supply infrastructure devoted to a single conversion facility. As the infrastructure is being developed, many small users are likely to compete for the fuelwood and the resultant availability to a single user may never reach the hundreds of thousands or milIions of dry tons per year necessary for the larger faciIities. Clearly biomass farms dedicated to a single conversion facility would overcome these problems of obtaining a large feedstock source, It is unlikely, however, that these farms will be developed as described in the section on “Unconventional Biomass Product ion.” 152 . Vol. II—Energy From Biological Processes

Appendix B.—Analysis of Break-Even Transport Distance for Pellitized Wood and Miscellaneous Cost Calculations

Two ways for producing 1 million Btu of steam Setting these two costs equal to one another and from woodchips containing 50-percent moisture using the weights of wood and pellets as above, are presented in figure B-1. In the first case 389 lb one finds that: of greenwood are transported to the boiler and d = (0.65 CP – 1.65 CW) / Ct (3) burned directly; in the second 339 lb of greenwood With transport costs of $0.10/ton mile and the are pelletized first and then transported to the wood and pellet costs given in the text, the break- boiler for burning. even transport distance varies from 43 to 71 miles. In the first case the cost of the fuel needed to If, however, the original wood is 40-percent mois- produce this steam is: ture, only 324 lb are needed in the boiler (with the

C = CW W W + Ct W W d (1) same efficiency) and the break-even transport dis- where CW is the cost per ton of wood at a central tance becomes 123 to 134 miles. If the transport yard, WW is the weight of wood to be transported, charges are lower, the break-even distance will in-

Ct is the transport cost per ton-mile, and d is the dis- crease. Conversely, where transport is more expen- tance from the yard to the boiler. sive, the break-even distance wilI be less. Numer- I n the second case: ous other local variables can also change the results. C = CP W P + Ct W P d (2) when C is the cost of the pellets at the pellet mill, Miscellaneous Cost Calculations W Pis the weight of pellets to be transported and the other symbols are as before (assuming the pel- Following are estimates for the costs of various let mill is located at the wood yard). thermochemical conversion processes.

Figure B-1.-Two Ways to Produce 106 Btu Steam From Wood

I 1

● Pellet mill Wood boiler 339 lb wood 153 lb Case 2 1 x 106 Btu steam (50% moisture) pellets 9(3% efficiency 83% efficiency

Table B-1 .-Electricity From Wood by Direct Combustion Table B-2.-Steam From Wood by Direct Combustion (package boiler) Input 2,000 green ton/d of wood (50% moisture) output 62-MW electricity Input: 270 green ton/d of wood (50% moisture) Load 300 operating days per year output, 50,000 lba of steam/hr

Fixed investment (field erected) $50 million Load 330 operating days per year Working capital (10% of fixed Fixed investment (package boiler) $600,000 Investment) 5 million $1,000 lb steam Total Investment $55 million Wood ($1 5/green ton) 338 Mills/kWh Million $/yr Labor($75,000/yr) 019 Capital charges(15-30’%0 of fixed Wood ($15/green ton) 20 90 Investment per year) O 23-0.46 Labor and water 9 40 Capital charges (15’%o of total Total 380-403 Investment per year) 19 825 Estimated range: $350-$6 00/ 1,000 lb steam Total 48 213 ($2 ‘O-$4 80/106 Btu) Estimated range 45-70 mills/kWh ‘1 ,O@I lb of steam = 1 25 mllllon Btu of steam SOURCE OTA from A Survey of B/omam Gas/f (car/on (Golden, Colo Solar Energy Research Institute, July 1979) Ch. 7—Thermochemical Conversion ● 153

Table B-3.–Electricity and Steam From Wood Table B-6.-Methanol From Wood Through Gasification by Direct Combustion in a Dual Fluidized-Bed Gasifier

Input 2,000 green ton/d of wood (50%. moisture) Input, 2,000 green ton/d of wood (50%. moisture) output 21.4-MW electricity and 390,000 lb steama/hr output 3,150 bbl methanol/d (44 million gal/yr) Load 300 operating days per year Load 330 operating days per year Fixed Investment (field erected) $40 million Fixed Investment (field erected) $64 million Working capital (10% of fixed Working capital (10% of fixed Investment) 4 million Investment) 64 million Total Investment $44 million Total Investment $704 million Million $/yr $/bbl Wood ($15/green ton) 109 Wood ($1 S/green ton) 952 Labor and water 4 Labor, water, and chemical 491 Capital charges (15’XO of total Capital charges (15-30~0 of total Investment per year) 66-132 Investment per year) 10.16-2032 Total 215-281 Total $2459-$3475 Product costs Steam EIectricity ($0 58-$0 83/gal) (assumed cost) (derived) Estimated range $22-$40/bbl $/1,000 Ib mills/kWh ($0 52-$0 95/gal) 4 67-109 ($8 20-14 96/106 Btu) 5 48-91 SOURCE OTA trom Ste\en R Beck, Department of Chemlc al E nglneermg. Tcwa$ 6 30-73 Tech Unlkers(tv Lubbock Tex prlk ate communication, 1979

‘1 000 lb ot steam = 1 25 mll I Ion Iltu ot steam Table B-7.-Pyrolysis Oil From Wood by SOURCE OTA trom Ste\ en R Beck Department ot ( hem,{ al t nglneerlng Texas TIX h Unl\ersttv I ubbrx k Tex prlk ate c ommunlt atmn 1979 Catalytic Direct Liquefaction Input 2,000 green ton/d of wood (50%. moisture) Table B-4.–Medium-Btu Gas From Wood in a Dual output 2,500 bbl/d of pyrolytic oil (4 2 106 Btu/bbl) Fluidized-Bed Field Erected Gasifier Load 330 operating days per year Input 2,000 green ton/d of wood (50%. moisture) Fixed Investment (field erected) $50 million output 460 106 Btu/hr medium-Btu gas Working capital (10% of fixed Load 330 operating days per year Investment) 5 million Fixed investment (field erected) $43 million Total Investment $55 million Working capital (10% of fixed $/bbl Investment) 43 million Wood ($1 S/green ton) 1200 Total Investment $473 million Labor, water, and chemicals 7.27 $106 Btu gas Capital charges (15-30% of total Wood ($15/green ton) 326 Investment per year) 1000-20.00 Labor and water O 82 Total $29.27-$39.37 Capital charges(15% of total Estimated range $30-$50/bbl Investment per year) $195-$390 (.$7-$1 2/106 Btu) Total $6.03-$7.98 SOURCE OTA trom Steven R Be{ k Department of Chemical Engineering, Texas Estimated range $550-$9.00/10 6 Btu Tech Uni\er\ttk Lubbock Tex prltate ( ommunl[ atlon, 1979 SOU R( E OTA trnm Ste\ en R B(>( k Department ot ( hemlc al F nglneer!ng lexa~ Table B-8.-Ethanol From Wood via Gasification Te( h Unl\erslty 1 ubb{,( k Tex prlk ~tc { {,mmunl( dt!on 1979 in a Dual Fluidized-Bed Gasifier Table B-5.–Medium-Btu Gas From Manure Input 2,000 green ton/d of wood (50% moisture) in a Fluidized-Bed Gasifier output 1,620 bbl/d of ethanol (assuming 14 wt. %. yield of ethylene from dry wood) Input 1,000 dry tend of manure Load. 330 operating days per year output 400 10’ Btu/hr medium-Btu gas Fixed Investment (field erected) $60 million Load 330 operating days per year Working capital (10% of fixed Fixed investment (field erected) $36 million Investment) 6 million Working capital (10% of fixed Total Investment $66 million Investment) 36 million $/bbl Total Investment $396 million Wood ($15/green ton) 1827 $/10 6Btu of gas Labor, water, chemicals, and electricity 1253 Manure ($3/dry ton) 031 Capital charges (15-30% of total Labor, water, chemicals, ash disposal, Investment per year) 1852-3704 electricity 164 Total $49.32-$67.84 Capital charges (15Y0 of total ($1 17-$1 62/gal) Investment per year) $188-$375 ($13.90-$19.20/10 6Btu) Total $383-570 With ethylene yield of 30 wt % $23-$32/bbl Estimated range $350-$7 00/10’ Btu ($0 55-$0 76/gal) (.%6 50-$9 00/106 Btu) SOLJ K( E () T A trom St(,\ en R t~~[ k [lepart ment ot C hem I{ a I E ngfneerlng Texas SOURCE OTA from Stewen R 13ec k Department of ( hemtcal f ngineerlcrg, Texas TVI h EJnl\erslt~ I ubb[)r k Tex I)riv ate ( fjmmunlc atlon 1979 Tech Unl\ ersltv I ubboc k T ex prtvate ( ommunic atlon 1979

~ -- q L ~ -1 - 80 - 1: . ..

154 ● Vol. I/—Energy From Biological Processes

Appendix C. —Survey of Gasifier Research, Development, and Manufacture

Gasifier type Size Contact Operating Organization lnput mode Fuel products units Btu/hr Air gasification of biomass Alberta Industrial Dev. , Edmonton, Alb , Can A FI LEG 1 30 M Applied Engineering Co , Orangeburge, S C 29115 A u LEG 1 5M Battelle-Northwest, Richland, Wash 99352. , ., A u LEG I-D — Century Research, Inc., Cardena, Calif. 90247 A u LEG 1 80 M Davy Powergas, Inc., Houston, Tex. 77036 A u LEG-Syngas 20 — Deere & Co., Moline, Ill. 61265 A D LEG 1 100kw Eco-Research Ltd. , Willodale, Ont. N2N 558 A FI LEG 1 16 M Forest Fuels, Inc., Keene, N H 03431. A u LEG 4 15-3.0 M Foster Wheeler Energy Corp., Livingston, N H 07309 A u LEG 1 — Fuel Conversion Project, Yuba City, Calif. 95991 A D LEG 1 2M Halcyon Assoc. Inc., East Andover, N.Y. 03231 A u LEG 4 6-50 M Industrial Development & Procurement, Inc., Carie Place, N.Y. 11514 ., ., A D LEG Many 100-750 k W Pulp & Paper Research Inst.,b Pointe Claire, Quebec H9R 3J9 A D LEG — — Agricultural Engr. Dept., Purdue University, W. Lafayette, Ind. 47907 A D LEG 1 O 25 M Dept. of Chem. Engr., Texas Tech University, Lubbock, Tex. 79409. A FI LEG 1 0.4M Dept. of Chem. Engr., Texas Tech University, Lubbock, Tex, 79409 A u LEG 1 — Vermont Wood Energy Corp., Stowe, Vt. 05672 A D LEG 1 008 M Dept. of Ag. Engr., Univ. of Calif., Davis, Calif. 95616 ., A D LEG 1 64,000 Dept. of Ag. Engr., Univ. of Calif., Davis, Calif. 95616 A D LEG 1 6M Westwood Polygas (Moore) A u LEG 1 Bio-Solar Researc & Development Corp., Eugene, Oreg. 97401 ., A u LEG 1 — Oxygen gasification of biomass Environmental En Eng., Morgantown, W V o D MEG 1P 05 IGT-Renugas 0, s FI MEG Pyrolysis gasification of biomass Wright-Malta, Ballston Spa, N.Y.c PG o MEG (C) 1 R, 1 P 4 Coors/U. of MO. P FI 1P U. of Arkansas P o MEG (C) 1R A & G Corp., Jonesboro, Ark, ., ... P 0 MEG (C) 1C ERCO, Cambridge, Mass, P FI PO,C 1P, (1C) 16, (20) EN ERCO, Langham, Pa. P MEG, PO, C 1P, 1 c Garrett Energy Research ... MH MEG 1P Tech Air Corp., Atlanta, Ga. 30341. ... P u MEG, PO, C 4P, 1C 33 M. Antal, Princeton Univ., NS ., ., PG o MEG, C 1R — M Rensfeit, Sweden . PG 0 MEG, C 1R Texas Tech, Lubbock, Tex.. PG FI MEG 1P Battelle-Columbus, Columbus, Ohio Air gasification solid muncipal waste (SMW) Andco-Torrax, a Buffalo, N. Y...... , ...... A u LEG 4C 100 M Battelle-Northwest, Richland, Wash. 99352,

Table Notation (by columns) Input A = alr gaslfier, O = oxygen gaslfler, P = pyrolysts process, PC = pyrolysis gaslfter, S = steam, C = char combustion Contact mode U = updraft, D = downdraft, O = other (sloping bed, movmg grate), FI = fluidlzed bed, S = suspended flow, MS = molten salt, MH = multiple hearth Fuel products LEG = low energy gas ( -150-200 Btu/SCF) produced m air gaslflcation, MEG = medium energy gas produced In oxygen and pvrolysls gaslflcatlon (350500 Btu/SCF, PO = pyrolysls 011, typically 12,000 fltu/lb, C = char, typically 12,(x)0 Btu/lb Operating units R = research, P = pilot, C = commercial size, Cl =commerclal Installation, D =demonstratlon Size Caslflers are rated m a variety of units Listed here are Btu/h derived from feedstock throughput on the basis of biomass containing 16 MBtu/ton or 8,000 Btu/lb, SMW with 9 MBtu/ton ( ) mdtcate planned or under construction ‘Unless noted otherwise, the gaslfwrs Ilsted here produce drv ash (T > 1,100” C) and operate at 1 atm pressure (Coal gaslfwrs and future biomass gastfiers may operate at much higher pressures )

b operate s at 1.3 atm Pressure cOperates at 10 atm pressure dThese ~aslflers produce slagglng (T > 1,3fM0 C) Instead of clw ash

SOURCE A Surieyof B/omass Gas/f /cdtlon (Golden, Colo Solar Energy Reseqrch Instttute, JUIV 1979) Ch. 7—Thermochemical Conversion ● 155

Gasifier type Size Contact Operating Organization Input mode Fuel products units Btu/hr

Oxygen gasification of SMW Union Carbide (Linde), Tonowanda, N.Y.d o u MEG 1 100 M Catorican, Murray HiIIs, NS 0 u 9M Pyrolysis gasification of SMW Monsanto, Landgard, Enviro-chem P, c K LEG, O, C ID 20 (375) Envirotech, Concord, Calif. P MH LEG 1P Occidental Res. Corp., El Cajon, Calif. P FI PO, C, MEG 1C Garrett En. Res. & Eng., Hanford, Calif. P MH MEG 1P Michigan Tech., Houghton, Mich. P ML MEG U. of W. Va. -Wheelebrator, Morgantown, W. Va. P, G, c FI MEG 1P Pyrex, Japan P, c, c FI MEG 1C Nichols Engineerlng P MEG, C ERCO, Cambridge, Mass P FI MEG 1P 16 Rockwell International, Canoga Park, Calif. P MS MEG, C 1P 16 M.J. Antal, Princeton, NS P o MEG, C 2R —

Table Notation (by columns) Input A = alr ga$lfter, O = oxygen gaslfler, P = pyrolysis process, PG = pvrolysls gaslfler, S = steam, C = char combustmn Contact mode U = updraft, D = downdraft O = other (sloping bed, moving grate), FI = fluldlzed bed, S = suspended flow, MS = molten salt, MH = multlple hearth Fuel products I F(; = low energy gas ( -150-200 FIIuSCF) produced !n alr gaslf!catton, MEG = medium energy gas produced In oxvgen and pkrolv$ls gaslflcatmn ( 150-500 f3tu/SCF, PO = pvrolvsls oil, typically 12,000 fltu lb, C = char, tvplcally 12,0CM) Btu/lb Operating untt$ R = research, P = pilot, C = commerc Ial SIZe C I = commercial m$tallatlon, D = demonstration Slrf’ Caslflers are rated (n a varletv ot ucr!ts I !sted here are Btu h derwed trczm feedstock throughput on the basis of bmmass containing 16 MFltu ton or t3,0cK) Fltu, lb, SMW with 9 MBtu ton ( ) Incflc ate planned or under crm$tructlon ‘Unless noted otherwise, the ga$tfler$ I Isted here produce riry ash [ T 1 1 ()()0 C) and operate at 1 atm pressure [Coal gaslfler$ and future biomass ga$lfters mav operate at mu{ h Chapter 8 FERMENTATION Chapter 8.– FERMENTATION

Page Introduction ...... 159 TABLES Ethanol From Starch and Sugar Feedstocks ...... 160 Page Energy Consumption ...... , . . . 160 52 Energy Consumption in a Distillery Process Byproducts ...... 162 Producing Fuel Ethanol From Corn ...... 161 Ethanol Production Costs...... 164 53 Early 1980 Production Costs for Ethanol Onfarm Distillation ...... 166 From Grain and Sugar Crops...... 165 Cellulosic Feedstocks...... 167 54 Cost of Ethanol From Various Sources. ....165 Generic Aspects and Historical Problems 55 Plausible Cost Calculation for Future With Pretreatment ...... 169 Production of Ethanol From Wood, Processes Currently Under Development . . . 169 Classes, or Crop Residues ...... 173 Generic Economics of Lignocellulosic Materials to Ethanol...... 172 FIGURES Environmental lmpact of Ethanol Production. . . . . 173 Page Process Innovations...... 175 26. Process Diagram for the Production of Crain and Sugar Processing ..., ...... 175 Fuel Ethanol From Grain ...... 160 Fermentation ...... 175 27. Process Diagram for the Production of Distillation. , ...... 176 Fuel Ethanol From Sugarcane or Sweet Producing Drv Ethanol...... 176 Sorghum...... 160 Chapter 8 FERMENTATION

Introduction

Ethanol, or “grain alcohol,” is a versatile mented solution by a distillation* process and commercially important liquid which has which yields a solution of ethanol and water been used for a variety of purposes for centu- that cannot exceed 95.6 percent ethanol (at ries. Ethanol is the intoxicant in alcoholic bev- normal pressures) due to the physical proper- erages and, prior to the industrial age, society’s ties of the ethanol-water mixture; and 4) in the most common contact with ethanol was as an final step, the water is removed to produce dry ingredient of beer, wine, or liquor. ethanol. This is accomplished by distilling once again in the presence of another chemical. Beverage alcohol is a major item of commerce and a source of substantial tax reve- The main distinctions among the processes nues. In addition, ethanol is also a key industri- using different feedstocks are the differences al chemical and is used as a solvent or reactant in the pretreatment steps. Sugar crops such as in the manufacture of organic chemicals, plas- sugarcane, sweet sorghum, and sugar beets tics, and fibers. Ethanol has a long history as a yield sugar directly, but the sugar often must combustible fuel for transportation vehicles be concentrated to a syrup or otherwise and space heating. Except under unusual cir- treated for storage or the sugar will be de- cumstances (e. g., wartime Europe), ethanol has stroyed “by bacteria. Starch feedstocks such as been little used for these purposes in the 20th corn and other grains require a rather mild century, having been largely displaced by pe- treatment with enzymes (biological catalysts) troleum-based motor and boiler fuels. or acid to reduce the starch to sugar. And. cel- Iulosic (cellulose containing) feedstocks such as crop residues, grasses, wood, and municipal Beverage alcohol is usually produced by fer- wastepaper require more extensive treatment mentation processes, but the processes are de- to reduce the more inert cellulose to sugar. signed to achieve various qualities of taste and Processes utilizing each of the ethanol feed- aroma which are irrelevant to fuel alcohol pro- stock types are considered below. I n addition, duction. Most industrial ethanol is produced the environmental effects of ethanol distill- from ethylene, a gas derived from petroleum eries are discussed as are various process or natural gas liquids. Rising oil prices have changes that could lower costs. Although etha- made biomass-derived ethanol competitive nol is emphasized in this chapter, it should be with ethanol derived from petroleum but it is remembered that other alcohols (e. g., butanol) unclear whether the chemical industry will and chemicals could be produced from the turn to biomass or coal for its supply of etha- sugar solutions, but technical and economic nol. uncertainties are too great to include a detailed consideration of these alternatives at All processes for the production of ethanol present. through fermentation consist of four basic steps: 1 ) first the feedstock is treated to produce a sugar solution; 2) the sugar is then con- *Distillation consists of heating the ethanol-water solution verted to ethanol and carbon dioxide (CO2) by and passing the vapor through a column in which the vapor con- yeast or bacteria in a process called fermenta- densed and revaporized numerous times, a process that succes- tion; 3) the ethanol is removed from the fer- sively concentrates the ethanol and removes the water

159 160 Ž Vol. II—Energy From Biological Processes

Ethanol From Starch and Sugar Feedstocks

Ethanol can be produced from starch and Figure 27.—Process Diagram for the Production of sugar feedstocks with commercially available Fuel Ethanol From Sugarcane or Sweet Sorghum technology. Starch feedstocks are primarily grain crops such as corn, wheat, grain sorghum, oats, etc., but also include various root pIants such as potatoes. The sugar feedstocks are plants such as sugarcane, sweet sorghum, sugar beets, and Jerusalem artichokes. Since these feedstocks are all crops grown on agricultural lands under intensive cultivation and can be converted with commercial technology, they are considered together. The processes for producing ethanol from starch and sugar feedstocks are shown schematically in figures 26 and 27. The energy consumption of these processes is discussed next, followed by a description of process byproducts, cost calculations, and onfarm processes.

Figure 26.—Process Diagram for the Production of Fuel Ethanol From Grain

Energy Consumption

Most ethanol distilleries in the United States today were designed for beverage alcohol production, with little emphasis on energy usage. A fuel ethanol distillery can take advantage of newer technology and the low purity requirements of fuel ethanol to reduce its energy consumption. Nevertheless, both the type of fuel used and the amount of energy consumed at the distillery will continue to be important determinants of the efficacy of fuel ethanol production in displacing imported fuels.

In the plant currently producing most of the

SOURCE Office of Technology Assessment fuel ethanol today, the germ (protein) in the Ch. 8—Fermentation . 161

corn feedstock is removed in a separate feed and uses “vapor recompression” evaporation processing plant. Consequently, the distillery for drying the DG. The distillery is coal-fired receives a more or less pure starch from the and consumes 42,000 Btu of coal and 13,000 grain processing plant and the waste still age Btu of purchased electricity* to produce 1 gal (material left in the fermentation broth after of ethanol which has a lower heating value of the ethanol has been removed) is fed into a 76,000 Btu. ** The energy breakdown for the municipal sewage system, so that the energy Katzen design is shown in table 52. needed to pretreat the corn and to process the waste stream is not included in the distillery Table 52.–Energy Consumption in a Distillery energy usage. Nevertheless, the distillery con- Producing Fuel Ethanol From Corn sumes 30,000 Btu/gal of ethanol and 96.5 per- 1 Thousand Btu of cent of this is in the form of natural gas. If all a Process step coal/gal of ethanol processing energy inputs are included, fuel Receiving, storage, and milling, ...... 0.8 consumption is about 65,000 to 75,000 Btu/gal Conversion to sugar (including enzyme production). 16.0 of ethanol (exclusive of the energy needed for Fermentation ...... 0.6 waste stream treatment). z Furthermore, the Distillation ...... , ...... 24.8 Distillers’ grain recovery ...... 6.2 economics of this process are predicated on in- Miscellaneous ...... 6.6 come from process byproducts, such as corn Total ...... 55.0 oil, for which the markets are uncertain if large aA55ume5 10,000 Btu of cod per kllowatthour of electricity volumes are produced. SOURCE Raphael Katzen Assoclales, Gram Molor Fuel Alcohol, Technical and Econorrvc Assess. rrml Study (Washington, D C Assistant Secretary for Policy Evaluation, Department OTA’s analysis indicates that the fuel used of Energy, June 1979), GPO stock No 061-000 .00308-8 at the distillery cannot soon be reduced to an insignificant fraction of the energy contained At first thought, one might expect the energy in the ethanol. Thus, if the displacement of im- demand of a distillery using sugar plant feed- ported fuels (oil and natural gas) is to be max- stocks to be less than that for starch feed- imized, fuel ethanol distiIIeries should be re- stocks, since the energy needed to reduce the quired to use abundant or renewable domestic starch to sugar is no longer required. The situ- fuels such as coal or solar energy (including ation is, in fact, quite the opposite. The proc- biomass). esses for extracting the sugar from the feedstock and concentrating it to a syrup (highly A distillery that might be more common in a concentrated sugar solution) are quite energy large-scale ethanol program has been designed intensive. The average energy usage for a sugar by Raphael Katzen Associates. ’ This distillery feedstock (based on sugarcane) would be would produce a dry animal feed byproduct, about 85,000 Btu of coal per gal Ion of ethanol known as distillers’ grain (DC) (see next section produced on the average, ’ or slightly more on byproducts). Although the distillery uses than the energy content in the ethanol. If the some equipment to dry the DC which is not in bagasse, i.e., plant matter left over after the common use in ethanol distilleries, all of the sugar is extracted, is used to fuel the boiler, equipment is commercially available. The de- then 110,000 Btu of bagasse would be needed sign reduces the number of distillation col- to produce 1 gal of ethanol. (This assumes a 70- umns to the minimum using conventional tech- percent boiler efficiency for bagasse, as op- nology (two columns: one to produce 95 per- posed to 90 percent for coal.) cent ethanol and one to produce dry ethanol) For both the grain and sugar feedstocks, ‘R Strasma, “Domestic Crude 011 Entitlements, Application for Petroleum Substitutes, E RA-03° submitted to the Department crop residues could be used to fuel the dis- of Energy by Archer Danlels Midland, Co , Decatur, Ill , May 17, tilleries. In both cases there is sufficient resi- 1979 update * I b}d *1O,OOO Btu/kilowatthour ‘Raphael Katzen Associates, Grain Motor Fuel Alcohol, Tech- * *Lower heating value is measured when water vapor iS the n/ca/ and Econorn(c Assessment Study (Washington, D C Assist- product of combustion The higher heat value, when liquid water ant Secretary for Pol Icv Evaluation, Department of Energy, June iS the product, iS 84,000 Btu/gal 1979), CPO stock No 061-000-00308-9 ‘Ibid 162 . Vol. Il—Energy From Biological Processes

due produced together with the starch or sugar Process Byproducts to fuel an energy-efficient distillery, although the quantity may be only marginally adequate All of the material in the feedstock, except for sugar feedstocks.5 If one requires that suffi- for the sugar or starch (most of which is con- cient residues be left on the land to provide verted to alcohol), become byproducts of dis- adequate soil erosion protection, then the tillation. in addition, the excess yeast or bacte- available residues are not adequate in most ria grown in the fermentation step can also cases. 6* Crop residues gathered from adjacent serve as a byproduct. The grain feedstocks are croplands where the crops are not used for eth- high in protein and, consequently, the byprod- anol production could easily supplement the uct credits will be larger than with sugar feed- shortfall, however. stocks.

Since the sugar feedstocks are generally de- The grain protein can be removed as “glu- livered to the distillery with much of the resi- ten” before distillation and oil, such as corn due, which subsequently arises as a waste by- oil, can be extracted. As mentioned above, product of the sugar extraction step, it is more however, the oil market is uncertain and the re- likely that residues will be used to fuel these quired selling price for such oil is too high for distilleries, although it is technically feasible in it to be considered as a fuel. both cases. The grain processes considered most likely If crop residues are used to fuel distilleries, for large-scale fuel ethanol production would then the fossil fuel usage at the distillery will ferment a mash (crushed, cooked, and treated be negligible. The fossil energy used to collect grain plus water) that still contains all the non- and transport residues and replace their nutri- starch components of the grain. The material ent value to the soil would have to be in- left after the ethanol has been removed, called cluded. OTA estimates this energy to be about “still age,” has in it protein, dead yeast, and 10,000 Btu/gal of ethanol for grain feedstocks bacteria as well as various other materials con- and about 3,000 Btu/gal for sugar feedstocks. tained in the grain. This stillage can be fed to (These estimates assume that no grain residues animals directly or can be dried (to produce are normally harvested with the grain and that DG) for transport and, again, used as an animal the entire sugar plant is harvested and trans- feed. The wet stillage, however, spoils in 1 to 2 ported to the distillery. Therefore, the grain- days, so care must be exercised when feeding fed distillery needs 10.3 lb of residue per gal- the still age wet. 7 lon of ethanol and the sugar-fed distillery needs a supplement of 3 lb of residue per gal- The high protein content makes DG a suita- lon of ethanol.) ble protein supplement to animal feed, although its high fiber content limits the quanti-

‘R A Nathan, “Fuels From Sugar Crops, ” published by Techni- ty that can be fed and the types of animals that cal Information Center, Department of Energy, TID-22781, July can consume it. Although DG contains about 1978. half the protein per pound of material as does ‘Ibid soybean meal, a common protein supplement, ● As an example, the national average available crop residues for corn are about 7,3 lb/gal of ethanol (see ch 3) With a 70-per- the types of protein in DG are such that the cent boiler efficiency, this would provide 70 percent of the ener- cattle use it more effectively and experiments gy needed at the distillery (assuming 6,500 Btu/lb) For sugarcane and sweet sorghum (syrup variety), the total indicate that 1.5 lb of DG can substitute for 1 crop residues are about 11 lb of combustible matter per gallon of ethanol The residues required to protect against soil erosion vary greatly If all of the residue IS used, one gets about 80 to 85 percent of the distillery energy requirement (assuming 30-percent leaves with 6,500 Btu/lb and 70-percent cane with 9,000 Btu/lb and 70-percent boiler efficiency) And in areas where residues are needed to protect the soil from erosion, the available residues ‘E W Kienholz, et al , “Craln Alcohol Fermentation Byprod- might be only the cane, which would be about 60 percent of the ucts for Feeding In Colorado, ” Department of Animal Sciences, distillery energy requirement Colorado State Unlverslty, Fort Colllns, Colo , 1979 Ch. 8—Fermentation ● 163

lb of soybean meal. * 89 Consequently, the byproduct (assuming the byproduct of ferment- product of distilling 1 bu of corn can displace ing 1 bu of corn displaces the soybean meal the meal from about 0.25 bu of soybeans.** from 0.25 bu of soybeans). The byproduct, however, is not a perfect substitute for soy- Other experiments have indicated that DC bean meal and the actual level at which the causes the cattle to digest more of the starch animal feed market becomes saturated is prob- in their feed than would be digested without 10 ably considerably lower than this. DG thereby giving DC an enhanced feed value, since less total corn could be fed to ani- Other uses for DC are possible. Brewers’ mals if part of the corn were converted to etha- yeast is used as a B vitamin source by some nol and the resulting DC fed in place of the people and the protein could possibly be used corn. These results, however, occur only when as a human protein source. It is not clear, how- the animal is fed a starch-rich and protein-poor ever, whether this source of protein will gain diet. Feed rations commonly used today have consumer acceptance. The distiller byproduct a more nearly optimum protein-starch bal- could also be exported as an animal feed sup- ance, so this effect would not occur, and the plement, but if it competes with indigenous feed value of DC is only as a replacement for soybean meal producers (such as in Europe), other protein concentrates used in animal ra- import tariffs or quotas may be imposed. tions. While there are numerous possibilities, most The quantity of DC that can be fed to cattle proposals are vague and involve some obvious has been estimated to correspond to an etha- problems. Consequently, byproduct credits nol production level of 2 billion to 3 billion could drop or disappear in a large-scale etha- gal/yr. 1 2 As mentioned above, the protein in nol program based largely on grain feedstocks. the grains could be removed before fermenta- If the protein in grains is removed in the tion, and this protein feed (“gluten”) would be pretreatment or sugar feedstocks are used, the suitable for a larger variety of animals. Theo- still age consists primarily of yeast or bacteria, retically, if the byproduct replaces all domes- and has smaller feed value than DC. (The dis- tic consumption of crushed soybeans used for tillery producing most of the fuel ethanol used animal feed,13 a production level of 7 billion today removes the protein in the pretreatment gal/yr could be achieved before all crushed and returns the still age to sewage treatment.) soybeans had been replaced with distillery by- Although there is a limited market for this stil- ‘Cattle break ciown some proteins In the rumen ancf later use Iage, it is likely that it will either be dried and the re~ultant ammonia tn the Intestines to s~ntheslze new pro- used as a fuel or subjected to anaerobic diges- teins Other proteins pa~s through the rumen anci are absorbeci cflrectly In the intestine Depencllng on the relatlve proportions of tion with the resulting biogas used as a fuel. the two classes of proteins, the effective qwantltv of usable pro- Drying and burning the byproduct result in tein WIII vdry slightly more energy— an estimated 8,000 Btu/ ‘T K Iopfenstein, Department of Animal Sclence$, Unlver$lty of Nebraska, Lincoln, Nebr , private communlcdtion, 1979 gal of ethanol. * ‘M 1 Poo\ and T Klopfensteln, “Nutrltiondl Value of BVpro~- uc t~ ot Alcohol Produ( t Ion for L lve~tock Eeecls, Cooperat Ive E x- Other possible byproducts of fermentation tenslon Servl( e, Unlver$itV of Nebrdska, I Incoln, Nebr , An!mal include oils, vitamins, other alcohols, various Sctence Publlcatlon No 79-4, 1979 organic acids (e. g., vinegar), fusel oil (a mixture ‘ ‘One bujhel of Cflstllled corn Vlelc!s about 18 lb of DC One bu~hel of soybeans produces about 48 lb of ~oybean meal of alcohols), and other chemicals. The proc- ‘“W P Carrigu~, Unlverslty of Kentu[ ky, Proceedjng$ of Ioth esses, however, are generally controlled so D/\t///ws’ Feed Conference, Clnclnatti, Ohio, Mar }, 1955 that the major chemical byproduct is fusel oil. ‘‘Klopfen$tein, op clt ‘‘R L Meekhof, W E TVner, and F D Holland, “Agricultural This would probably be combined with the Policy and Cd$ohol, ” purdue (_Jn!versltV, west LdfdVette, I nd , Mav 1979, contractor report to OTA These authors assume a 21 subst itut Ion of D~ for soybean mea I a nci ) bl j ! Ion ga I of ethanol ‘If the rnaterlal IS cirleci, 11,000 Btu (2 lb) of material result per per Vear as the saturation point Using 1 51 as the ratio, how- gallon of ethanol The cfrytng however, requires an estlmateci ever, reduces this to 225 bllllon gal/yr 1,000 tltu additional Input energy Anaerobic dlgestlon woulcf ‘ ‘ARr/cu/tura/ $t.?(I\tIc$, / <)79 (Wa~hlngton, D C IJ S Depart- prociuce about 5,0oo Htu of blogas (assum[ng 4 ft blogds/lb sol- ment of Agriculture, 1979), GPO stoc-k No 001- 000-04069-1 Icis) wlt h the process requ Irlng about 1,000 13tu 164 . Vol. II—Energy From Biological Processes

fuel alcohol, resulting in a 0.7-percent increase million in 1980, assuming the feedstock is in the quantity of fuel produced. available for half of the year and half year’s syrup storage is required. These assumptions CO is also a byproduct of fermentation 2 about the length of time that the feedstock which is used in carbonated beverages, dry ice, will be available may be somewhat optimistic and, to a small extent, in chemical processes. for Midwestern grown sweet sorghum, how- Moreover, CO has many interesting properties 2 ever, and the cost could be higher. If the raw that are currently being researched and recov- feedstock is available for only 3 months per ery may eventually become more widespread year, OTA estimates the distillery would cost and profitable. about $140 million in 1978 dollars. Although it might be possible to avoid con- Ethanol Production Costs centrating the extracted sugar solution to a Raphael Katzen Associates has performed a syrup by using antibiotics or various chemi- detailed cost calculation on a 50-million-gal/yr cals, a major cost of the pretreatment is the coal-fired distillery that purchases its electrici- equipment needed to remove the sugar solu- ty from an electric utility. 14 Including coal-han- tion from the raw plant material. Furthermore, dling and pollution control equipment and al- storage of large quantities of dilute sugar solu- lowing the production of dried DC, the total tion would be expensive. Consequently, im- distillery would cost an estimated $53 million. provements in the economics of using sugar feedstock will require methods for storing the Inflating this to early 1980 dollars (20 per- raw sugar feedstocks inexpensively and in a cent) results in a distillery investment of $64 way that the sugar need not be removed and million. (These figures do not include engineer- concentrated. Possibilities include pretreating fees which could be small if a large number ment with chlorine gas, ammonia, or sulfur of distilleries are built, but which are esti- dioxide (to change the acidity and provide a mated at $6 million, in 1978 for a single dis- toxic environment for bacteria). OTA is un- tillery.) aware, however, of any work in this area that A distillery designed solely for sugar crop would serve as a basis for cost calculation. feedstocks would cost considerably more. As An alternate approach is to build a distillery mentioned above, the sugar has to be concen- capable of handling either starch or sugar trated to a syrup for storage, since the feed- feedstocks. Katzen has calculated that this 50- stock is available for only part of the year, dur- million-gal/yr distillery would cost $93 million ing and somewhat after the harvesting season. in 1978 dolIars.17 Hence, the pretreatment equipment has to be able to handle a larger capacity than the distil- The ethanol costs are influenced by the cap- lery for part of the year, while standing idle for ital investment in and financing of the distil- part of the year. In addition storage tanks are lery, the distillery operating costs, and the needed for the syrup. if the bagasse and crop byproduct credits. For a coal-fired 50-mill ion- residues are used as fuel, however, then some gal/yr distillery using starch feedstock, the cap- of the pollution control equipment needed to ital charges are about $0.21 to $0.42/gal of eth- remove sulfur emissions can be eliminated, anol, depending on the financing arrange- due to the very low sulfur content of the ments. These charges, however, can vary sig- biomass. In all, a 50-miIlion-gal/yr distillery nificantly with interest rates, depreciation al- for sugarcane or sweet sorgham would cost an lowances, tax credits, and other economic in- estimated $100 mill ion in 197815 16 or $120 centives. The major operating expense is the feed- ‘“Raphael Katzen Associates, op clt ‘Slbid stock cost less the byproduct credit. For corn “F C Schaffer, Inc , in E S Llplnsky, et al , Sugar Crops as a at $2.50/bu, the feedstock costs $0.96/gal of Source of Fuels, Vol // Processing and Conversion Research, final report to Department of Energy, Aug 31, 1978 1‘Raphael Katzen Associates, op clt Ch. 8—Fermentation ● 165

ethanol and the byproduct credit is about Tables 53 and 54 show the cost of ethanol $0.38/gal ($110/ton of DG), resulting in a net produced from various feedstocks. Although feedstock cost of $0.58/gal. Because farm com- the costs will vary depending on the size of the modity prices are extremely volatile, the net distillery, ethanol can be produced from corn feedstock and resultant ethanol cost could be ($2.50/bu) in a coal-fired 50-million-gal/yr dis- quite variable. A $0.50/bu increase in corn tillery for $0.95 to $1 .20/gal. About $0.10 to grain prices (and a proportionate increase in $0.30/gal should be added to these costs for the byproduct credit), for example, would raise deliveries of up to 1,000 miles from the distil- the ethanol cost by $0.1 2/gal. lery. (Most ethanol is currently delivered in

Table 53.–Early 1980 Production Costs for Ethanol From Grain and Sugar Crops (in a 50-million-gal/yr distillery)

Graln a Sugarb Fixed capital. ., ., ., ., ... ., ., ., ., ., ., ...... $64 million $120 million Working capital (10% of fixed capital) ... ., . . . . 6.4 million 12 million Total Investment ., ., ., ., ., ... ., ., ., ., ., ... ., $70.4 million $132 million

$ per gallon of 99.6% ethanol Operating costs Labor ...... $0.08 $0.09 Chemicals ...... 0.01 0.01 Water . . ., . . . . ., ...... , ...... 0.01 0.01 Fuel (coal at $30/ton for grain feedstock and crop residues at $30/ton for sugar feedstock) ...... 0.08 0.04c Subtotal ...... $0.18 $0.15 Capital charges 15 to 30% of total investment per yeard ...... 0.21-0.42 0.40-0.79 Total ...... $0.37-$0.60 $0.55-$0.94

alncludes drying Of distillers gram bln~lude~ eqU,pmen( for extracting the sugar from the feedsfock Concentratlflg If to a sYrUP for stora9e c~gasse.fueled dlsllllery appropriate for sweef sorghum and sugarcane supplemental fuel requirement Is 3 lb of residue Per 9allon of ethanol

dThere are M any oflen complex formulae [ocOmpufe ac[ualcapl[a(cOs[s Economic factors considered mcludedebl/equity ratio dePrecfatlon schedule In” come lax credit rate of Inflallon terms 01 debt repay menl, Operating capital requirements and Investment hfet!me However, a reahstlc range of posslbll. thes for annual capital costs would he between 15 and 30% of total capital mves[ment The upper extreme of 30% may be obtained assuming 100% equity finance and a 13% aftertax rale of return on Investment The lower extreme of 15% may be obtained assuming 100VO debl flnanclng a! a 9% rate of Interest Both calculations assume constant dollars, a Zo.year protect Mehme, and include a charge for local taxes and Insurance equal to 3% of fixed cap!fal costs For a more detaded treatment of capital costs see OTA, Apphcarlon 01 Solar Tecfrno/ogy /0 Today’s Energy Needs VOI II ch 1 SOURCE Off Ice of Technology Assessment and Raphael Kalzen Assoclales Gram Moror Fue/A/coho/, Techrvcalarrd Ecorromfc Assessmen/ Study (Wash mgfon D C Asslstanl Secretary for Policy Evaluation Department of Energy, June 1979), GPO stock No 061.000 -00308-9

Table 54,–Cost of Ethanol From Various Sources

Net feedstock costb Ethanol cost YieldsC (gallons of Feedstock Price a ($/gal ethanol) ($/gal) ethanol per acre) Corn ...... $2.44/bu $0.57 $0.94-$1.17 220 Wheat ., ., ., ., ., 3.07-4.04/bud 0.73-1.08 d 1,10-1.68 86 Grain sorghum 2.23/bu 0.49 0.86-1.09 130 Oats...... 1.42/bu 0.59 0,96-1.19 75 Sweet sorghum, ., ., 15.00/tone 0.79 1.34-1.73 380e Sugarcane . . . . . 17.03/tonf 1.26 1.81-2.20 f 520

aAverage of 1974-77 seasonal average vices hhe feedslock cost less the byproduct credit The dltference In feedslock costs might not hold over the longer term due to equdlbrahon of prices through large-scale ethanol production cAverage of 1974.77 national avera9e yields dRange due tO d! fferent prices for different tYPeS of Meat eA55umlng 20 fre5h We!ghf tons/acre yteld $300/acre ProductIon costs ‘Excludes 1974 data due 10 the anomalously high sugar Prices that Year SOURCE Agr/cu/fura/ Slaf/sOcs 1978 (Washmglon O C U S Oeparfmerl of Agriculture), and Office of Technology Assessment 166 . Vol. Il—Energy From Biological Processes

tank trucks, but as production volume grows tion prefabricated distilleries for producing 90 other forms of transportation, such as barge to 95 percent ethanol are available both at the shipments, rail tank cars, and petroleum prod- farm size (15,000 gal/yr)19 and coop size (sev- uct pipelines, * could decrease the transporta- eral hundred thousand gallons per year)20 for a tion cost to as low as $0.03 to $0.05/gaI under cost of about $1 for each gallon per year of ca- favorable circumstances.) pacity, but there is insufficient onfarm operating experience to establish the reliability or ex- As shown in tables 53 and 54 the major tradeoff between starch and sugar feedstocks pected operating life of these distilleries. OTA is that the starch-fed distilleries require con- is not aware of smaller distilleries, but there is siderably less investment than the sugar-fed no fundamental reason why they cannot be ones, but the ethanol yield per acre cultivated built. There will, however, be a tradeoff be- may be larger with the sugar feedstocks. As tween the cost of small distilleries and the noted in chapter 3, however, these yield figures amount of labor required to operate them. are highly unreliable for sweet sorghum, and A farmer must consider a number of site- sugarcane cannot be grown on most cropland specific factors before deciding to invest in an potentially available for energy crop produc- onfarm skill. Some of the more important of tion. If comparative studies of potential eth- these are: anol feedstocks grown under comparable con- ● ditions show that certain sugar crops produce Investment. – How much does the still and more ethanol per acre than the starch crops, related equipment cost? ● then there may be a tendency to turn to sugar Use of the ethanol. –Will the ethanol be feedstocks as farmland prices rise. Moreover, used onfarm or sold? What equipment if the grain byproducts are difficult to sell, modifications are necessary? Will the then economics could favor sugar crop feed- farmer be dependent on a single buyer, stocks. For now, however, the lower capital in- such as a large distillery that will upgrade vestment required for grain-fed distilleries 95 percent ethanol to dry ethanol? ● gives them an advantge over sugar-fed distil- Labor. – Does the farmer have access to leries. cheap, qualified labor, or is it better to make a larger investment for an automatic distillery? Onfarm Distillation ● Skill.– Although ethanol can be produced easily, the process yield—and thus the Apart from commercial distilleries, consider- cost — as well as the safety of the opera- able interest has been expressed in individual tion can depend critically on the skill of farmers or farm coops producing ethanol. A the operator. number of factors, however, could limit the ● Equipment lifetime.– Less expensive distil- prospects of such production. leries may be constructed of materials Technology for producing 90 to 95 percent that are destroyed by rust after a few ethanol (5 to 10 percent water) is relatively sim- years’ operation. ple. Several farmers are or have constructed ● Fuel.– Does the farmer have access to their own distilleries for this purpose. In addi- wood, grass, or crop residues and combustion equipment that can use these fuels? ‘Various strategies can be used to eliminate potential prob- Can reliable, inexpensive solar stills be lems with the water sometimes found in petroleum pipelines. If ethanol is being transported, the total volume of ethanol in the constructed for the distiIlation step? batch can be kept large enough so that the percentage of water If oil or natural gas is used in the distil- in the delivered ethanol iS within tolerable limits If gasohol is Iery, would it be less expensive to use this transported, it can be preceded by a few hundred barrels of ethanol which will absorb any water found in the pipeline, thereby keeping the gasohol dry. Other strategies also exist or can be de- “Paul Harback, United International, Buena Vista, Ga , pri- 18 veloped vate communication, October 1979 ‘8L J Barbe, Jr , Manager of Otl Movements, EXXON Plpellne ZORobert Chambers, president, ACR Process Corp , Urbana, Ill , Co , Houston, Tex , private communication, August 1979 private communication, September 1979 Ch. 8—Fermentation ● 167

fuel directly as a diesel fuel supplement in Meeting this price goal for automatic, on- a retrofitted diesel engine? farm, dry ethanol production facilities will ● Byproduct. – Can the farmer use the wet probably require process innovations, particu- byproduct on his/her farm? Will this un- larly in the ethanol-drying step, and could well duly complicate the feeding operations or involve the use of small, inexpensive comput- make the animal operation dependent on ers (microprocessors) for monitoring the proc- an unreliable still? What will drying equip- ess. A major constraint, however, could be the ment cost and how much energy will it cost of sensors, automatic valves, etc. that consume? would be required. ● Water. – Does the farmer have access to For some farmers, however, the cost or labor sufficient water for the distiIIery? required to produce ethanol may be of second- Under favorable circumstances, it might be ary importance. The value of some degree of possible to produce 95 percent ethanol for as liquid fuel self-sufficiency and the ability to little as $1/gal* plus labor with a labor-inten- divert limited amounts of corn and other sive distillery. If the ethanol is used in a diesel grains when the market price is low may out- tractor, the ethanol would be equivalent to weigh the inconvenience and/or costs. 1 n other diesel fuel costing $1.70/gal, or about twice the words, farmers may consider the technology to current diesel fuel prices, Under unfavorable be an insurance against diesel shortages and circumstances, the cost could be several times hope that it will raise grain prices. Although as great. Due to a lack of experience with on- insurance against diesel shortages certainly can farm distilleries, however, these cost estimates be achieved by purchasing large diesel storage may be low. tanks at a cost below an ethanol distillation and storage system, increased grain prices for the en- Onfarm or coop production of dry ethanol tire crop would make the economics consider- could become competitive with commercially ably more favorable to farmers but would be a distilled ethanol, however, if relatively auto- very expensive way for the nonfarm sector to matic, mass-produced distilleries capable of provide fuel to farmers. As evidence of the inter- using fuels found onfarm and producing dry est, the Bureau of Alcohol, Tobacco, and Fire- ethanol and dry DC could be sold for about $1 arms had received over 2,800 applications for for each gallon per year of capacity and if onfarm distillation permits by mid-1979 and farmers charge little for their labor. OTA is not they expected 5,000 by the end of the year. 2 aware of any package distiIIeries for producing ’ As a profitable venture in the absence of large dry ethanol that are available at this price. subsidies or grain price increases, however, on- *Assuming equipment costs of $1 for each gallon per year of farm production of ethanol is, at best, margin- capacity, the costs per gallon of ethanol are: $0.58 for net feedstock cost, $0.20 for equipment costs (operated at 75 percent of al with current technology. capacity), $020 for fuel (assuming $3/m million Btu and 67,000 1’Wllllam Davis, Bureau of Alcohol, Tobacco, and Firearms, Btu/gallon), and $0.05 for enzymes and chemicals, resulting in U S Treasury, Washington, D C , private communlcatlon, July $1 03/gal of ethanol or $0.98/gal of 95 percent ethanol 1979

Cellulosic Feedstocks

The feedstocks with the largest potential for Wood, grasses, and crop residues contain ethanol production –both in terms of the ab- cellulose, hemicellulose, and lignin. The cellu- solute quantity of ethanol and in terms of the lose can be reduced, or hydrolyzed, to sugars quantity of ethanol per acre of cultivated that can be fermented to alcohol. The hemicel- land– are the cellulosic, or cellulose contain- Iulose can also be reduced to sugars capable ing, feedstocks. These include wood, crop resi- of being converted to ethanol with other types dues, and grasses, as well as the paper fraction of bacteria. The Iignin, however, does not con- of municipal solid waste. vert to alcohol and can be used as a source of 168 . Vol. n-Energy From Biological Processes

chemicals or dried and used as a fuel. General- al other countries have expressed interest in ly, paper is primarily cellulose with varying developing the technology, and one plant in amounts of partially broken Iignin. Switzerland is again being used for pilot studies. 24 The removal of hemicellulose from wood, grass, or crop residues and its reduction to Clearly it is technically possible to produce sugar are relatively straightforward. In fact, ethanol from Iignocellulosic feedstocks today. hemicellulose from biomass is the prinicpal The failure of these processes to remain eco- source of the chemical feedstock furfural. nomically viable except under special circum- Although hemicellu lose is not now used as a stances has been due, in large part, to the rela- source of ethanol, the fermentation step can tively low costs of petrochemicals and ethyl- probably be developed without excessive dif- ene-derived ethanol. With oil prices rising, the ficulty. primary competitor is likely to be grain- and sugar-derived ethanol. There are, however, im- The cellulose, on the other hand, is em- provements and developments in the lignocel- bedded with Iignin, which protects it from bio- Iulose processes which can make them com- logical, but to a much lesser extent chemical, petitive with the current costs of ethanol from attack. Thus, the reduction of cellulose in- these other feedstocks. Alternatively, large volves treating the IignocelIulose material with rises in farm commodity prices could make the acid or pretreating the material either chem- cellulosic processes competitive without tech- ically or mechanically to make it susceptible nical developments. to biological reduction with enzymes. While there are processes whose economics What was apparently the first acid hydroly- rely on large byproduct credits or special fi- sis of wood was described in a German patent nancing that could be in commercial operation issued in 1880.22 Modifications of this process before 1985, the key to achieving economic were used to produce animal fodder in several competitiveness without these conditions is to countries (mostly for the sugar) during World develop processes which: War 1. At the end of the war, the economic basis became obsolete. Between World Wars 1 ● produce high yields of ethanol per ton of and 11, however, other acid hydrolysis proc- biomass, esses were used mostly in Germany to produce ● do not require expensive equipment, sugar and alcohol, partly because of materials ● allow nearly complete recovery of any ex- shortages but partly in an attempt at self-suffi- pensive process chemicals, and ciency. 23 Other plants were also built in Swit- ● do not produce toxic wastes. zerland and Korea. No processes currently in existence fully satis- During World War 1, pilot plants were built fy all of these criteria, although there are proc- in the United States for producing ethanol esses that satisfy two and sometimes three of from wood wastes. Acid hydrolysis processes the criteria. Nevertheless, R&D currently un- underwent a series of modifications during derway could yield significant results in 3 to 5 World War Il. Following World War II, how- years. With a normal scaleup of 5 years, one or ever, virtually all of the wood-ethanol plants more processes satisfying these criteria could were closed for economic reasons. * Today become commercial by the late 1980’s. commercial wood sugar plants are in opera- The generic aspects and historical problems tion only in the U.S.S.R. and in Japan but sever- with producing sugars from Iignocellulosic Z~H F J wenz I ~~e Chemica / Tec/?nO/ogy of wood, trans- feedstocks are now discussed, followed by a lated by F E Braun’s (New York: Academic Press, 1970). ~ ‘Ibid Slightly more detailed description of various *One ethanol plant that uses the sugar-containing waste processes currently under investigation. Final- stream of a sulfite paper-pulping plant IS still in operation. It is, however, primarily a waste treatment plant and less than 10 per- 24) L Zerbe, Program Manager, Forest Service Energy Re- cent of the paper-pulping processes used in the United States search, U S Department of Agriculture, Forest Products Labora- produce a suitable waste stream tory, Madison, WIS , private communication, 1980 Ch. 8—Fermentation ● 169

Iy, a generic economic analysis is presented for The historical processes have generally used a hypothetical advanced distillery for produc- acid hydrolysis. The dilute acid methods (Mod- ing ethanol from IignocelIulosic feedstocks. ified Rheinau, Scholler-Tornesch, Madison, Tennessee Valley Authority, and Russian Modi- fication of Percolation processes) all suffer Generic Aspects and Historical from a similar ailment,26 The high tempera- Problems With Pretreatment tures and acidic conditions needed in the processes cause the resultant sugars to decompose, As mentioned above, Iignocellulosic materi- thereby lowering the overall ethanol yield. The als consist of cellulose, hemicellulose, and lig- concentrated acid processes (Rheinau-Bergius nin. Typically, such material would first be and Hokkaido), on the other hand, have re- treated with dilute acid to remove the hemicel- sulted in good product yields. The economics, Iulose, which then would be fermented in a however, have historically suffered due to the separate step to ethanol. The remaining lignin- loss of large quantities of acid in the processes, cellulose combination would be treated with Nevertheless, one of the oldest concentrated concentrated acid at low temperatures (per- acid processes (Rheinau-Bergius) is currently haps 100° to 110° F) or dilute acid at high tem- being reexamined to see if this economic con- peratures (300 0 to 4000 F) to either dissolve the clusion necessarily pertains today (see below). cellulose from the Iignin or to cause the material to swell, thereby exposing the cellulose for Publications over the past 20 years in the So- hydrolysis. Alternatively, the material can be viet Union have reported good experimental exposed to a number of different chemical or results with impregnating wood with acid fol- mechanical pretreatments which render the lowed by mechanical grinding. The details for cellulose susceptible to hydrolysis. The hydrol- an assessment of the commercial viability of ysis is then accomplished by further exposure this process, however, are not available. On to acid or by the action of enzymes (biological the other hand, a mechanical pretreatment is catalysts). also involved in the Emert (formerly Gulf Oil Chemicals) process discussed below. Histori- The relative amounts of cellulose, hemicel- cally, the mechanical pretreatments needed Iulose, and Iignin can vary considerably among have been quite expensive, but the researchers the various Iignocellulosic materials. If pure indicate that this is not a problem with the cellulose is converted completely to ethanol, Emert process. 27 Finally, a variety of other however, the theoretical maximum yield is 170 processes or combinations of processes aimed gal of ethanol per ton of cellulose. The yields at exposing the cellulose to hydrolysis are cur- per ton of hemicellulose are similar. Conse- rently being researched. The most important of quently for a Iignocellulosic material that is so these are considered below. percent cellulose, 20 percent hemicellulose, and 25 percent Iignin, the theoretical yield is about 120 gal/dry ton of biomass fermented. A Processes Currently Under yield of 85 to 90 percent of this is a reasonable Development practical goal, which would result in yields of 100 to 110 gal of ethanol per ton of biomass Emert Process fermented. The expected yield, however, will The development of this process started in vary with the exact composition of the feed- 1971 under Gulf Oil Chemicals Corp., but was stock. For municipal solid waste (29 percent transferred to the University of Arkansas Foun- paper and 21 percent yard wastes and wood packaging 25), the average yield could be about 1~1 Gold5teln D~partm~nt of Wood and Paper Science, North 60 gal of ethanol per ton assuming a 90-per- Carollna State Unwerstty, Ralefgh, N C , private communlcatlon, 1979 cent overall conversion efficiency. “G H Emert and R Katzen, “Chemicals From Biomass by lm- (VOI 1, wash lng- ~ ~~a ~erla /~ a n~ f~efgy ~rorn MU njclpa / Waste proved Enzyme Technology, ” presented In the fymposlum f3iG ton, D C Office of Technology Assessment, 1979), GPO stock mass as a Non-Fue/ Source, sponfored by the AC S/CSJ Joint No 052-001-00692-8 Chemical Congress, Honolulu, Hawal}, Apr 1-6, 1979

67-968 0 - 80 - 12 170 ● Vol. I/—Energy From Biological Processes

dation for scaleup (the transfer reportedly oc- The cost estimates also assume a large by- curred because Gulf had made a management product credit for dried fermentation yeast decision to concentrate its efforts on fossil and hydrolysis bacteria ($0.40/gal ethanol). fuels). This process is the most advanced of the Most of this comes from the hydrolysis bacte- enzymatic hydrolysis methods and, with prop- ria and an animal feed value for this material er financing, can probably be brought to com- has not been established. In addition, large- mercial-scale operation by 1983-85. scale production could lead to a saturated animal feed market similar to that with grain The method consists of a pretreatment de- distillation and subsequent loss of the byprod- veloped for this process which involves grind- uct credit. ing and heating the feedstock followed by hydrolysis with a mutant bacterium also devel- Furthermore, problems encountered with oped for this purpose. A unique feature is that scaling up a process virtually always lead to the hydrolysis and fermentation are performed cost increases above those estimated. Conse- simultaneously in the same vessel, thereby re- quently, these cost estimates could be too low ducing the time requirements for a separate by $0.20 to $0.70 or more per gallon of ethanol. hydrolysis step, reducing the costs and increas- Nevertheless, with municipal bond financing, ing the yield (since a sugar buildup during hy- this process could well be competitive with drolysis could slow the hydrolysis and de- ethanol produced from corn in a privately fi- crease the overall yield). Also the process does nanced distillery by 1983. (Assuming 7-percent not use acids, which would increase equip- annual inflation as apparently was done in ment costs. The sugar yields from the cellulose Emert’s calculations, $1 .10/gal ethanol in 1979 are about 80 percent of what is theoretically would sell for about $1 .45/gal in 1983). achievable, 28 but the small amount of hemicel- While no cost estimates are available for Iulose in the sawdust is not being converted. this process using woodchips, grasses, or crop The process has been brought to the pilot residues as feedstocks, Emert reports that ex- plant stage and funds are currently being periments have shown that modifications in sought for a demonstration (1 million gaI/yr) fa- the thermal-mechanical pretreatment enables cility as part of the scaleup process. Based on ethanol yields of 70 to 75 gal/ton of feed- the pilot plant experience, Emert estimates the stock .30 The increased costs for these feed- selling price for the ethanol to be $1 .49/gal stocks ($40 to $50/ton in 1983 up from $30 to (1983 dollars, 100-percent private equity fi- $40/ton in 1979) would add $0.30 to $().45/gal nancing, and 10-year amortization).29 With 80- to the ethanol price. Consequently, it is less percent municipal bond financing, he esti- likely that this process using these feedstocks mates the selling price to be $1.01/gal (1983 would be competitive with corn-derived etha- dollars, 20-year amortization). nol, unless corn and other grain prices rise more rapidly than general inflation. These cost estimates are based on a feedstock of So-percent “air classified” municipal In sum, it appears that this process could be solid waste (i. e., the paper and plastic fraction) competitive with grain-derived ethanol if mu- at $14/ton, 25-percent saw mill waste at $21/ nicipal wastepaper is used as a feedstock and ton, and 25-percent pulp mill waste at $14/ton. the distillery receives special financing. A re- These costs are all on the low end of estimates liable determination of the competitive posi- for 1978-79 prices and consequently represent tion of other feedstocks and financing arrange- optimistic estimates. Furthermore, by 1983, in- ments are less certain and probably cannot be flation would increase these costs. More realis- determined until a full-scale plant has been tic 1983 feedstock costs (50 to 100 percent built. higher than those cited) would raise the ethanol cost by about $0.10 to $0.20/gal. “lbtd “lbld ‘“G H Emert, private ~ommunlcation, October 1979 Ch. 8—Fermentation ● 171

Reexamination of type should be thoroughly investigated, it has Rheinau-Bergius Process not yet been satisfactorily demonstrated that the process proposed would be economically Much of the detailed information on the competitive as a source of fuel ethanol. Rheinau-Bergius process has been lost. Since the acid hydrolysis of wood involves subtle University of Pennsylvania— chemical processes which can change dramat- General Electric Process ically with small changes in the process conditions, the detailed process chemistry of hydrol- In this process, woodchips are heated in an ysis with concentrated hydrochloric acid is be- alkaline solution containing water, sodium car- ing reexamined at North Carolina State Univer- bonate, and butanol (a higher alcohol). Since sity. The research should provide a basis for butanol is only partly soluble in water, the reevaluating the process as a source of ethanol solution consists of two phases (similar to oil and chemicals and determining whether suffi- floating on water). The hemicellulose goes to cient quantities of the acid can be recovered the water phase, the Iignin dissolves in the bu- to make the process economic at today’s tanol, and the cellulose remains undissolved. prices. Following removal of the cellulose, and cleaning to remove traces of butanol, it can be hy- Tsao Process drolyzed either with acid or enzymes and the hemicellulose can be converted to ethanol This process is being developed at Purdue without removing it from solution. The butanol University with the major emphasis on crop is then cooled, which causes the Iignin to pre- residues as a feedstock and is currently pro- cipitate from solution, the solution is filtered, gressing to the pilot plant stage. Although and the butanol recycled to the process. there have been numerous changes in the process as the research has proceeded, in the cur- Clearly the process economics will depend rently preferred process hemicelIulose is re- heavily on the cost of producing the process moved first with dilute acid and then, the cel- butanol and the quantity of butanol lost to the lulose and Iignin are dissolved in concentrated waste stream. On the other hand, the butanol- (70 percent) sulfuric acid. The acid is recov- water sodium carbonate solution is consider- ered by precipitating the celIulose-lignin from ably less corrosive than other chemicals used the acid through the addition of methanol, to remove Iignin and therefore could result in then the methanol is removed from the acid by lower equipment costs. At this stage, however, distillation. Following this pretreatment, en- the processes are not well enough defined to zymes hydrolyze the cellulose. provide a meaningful cost calculation.

The use of methanol to aid in recovering the U.S. Army —Natick Laboratories acid is a novel aspect of this process. As the recovery has been proposed, however, the Work done at this laboratory has contrib- methanol is likely to react to form toxic by- uted substantially to the basic knowledge products such as dimethyl sulfate, dimeth- about the enzyme system that converts cellu- yl ether, dimethyl sulfoxide, and other com- lose to sugar. ’3 These researchers first identi- pounds. The loss of process methanol as well fied the three-enzyme system involved in the as the disposal of these toxic wastes would in- hydrolysis and have developed fungus mutants crease the costs. I n addition, there are several with improved enzyme productivities. Not places in the process where more expensive only is this research applicable to ethanol pro- equipment will be needed than has been in- duction, but it also provides information for cluded in most cost calculations due primarily those interested in retarding cellulose degrada- to the corrosive effects of the acid.31 32 Al- tion such as that which occurs with jungle rot. though novel acid recovery processes of this “E T Reese, “History of the Cellulose Program at the U S ‘‘I?aphael Katzen Associates, op clt Army Natlck Development Center, ” Biotechnology and B/oener- ’11 Goldsteln, op clt gy Syf71POS/Uf?l, No 6, p 9, 1976 172 ● Vol. n-Energy From Biological Processes

The system developed at Natick, however, Nevertheless, the process at the most ad- requires relatively pure cellulose (such as in vanced stage of development (of those being paper); it has not been effective on lignin- developed) appears to be the Emert process. containing materials such as grasses, crop resi- But as this process now stands and with a suc- dues, and wood. Recently, attention has been cessful scaleup, the ethanol could sell for directed at a mechanical process (ball milling) $0.30 to $0.60/gal more than corn-derived eth- for reducing raw materials to extremely fine anol and the price difference could be greater particles in order to use the Natick fungus, but if woodchips rather than sawdust are used as a this pretreatment is expensive and would prob- feedstock. As mentioned above, however, spe- ably make the process uneconomic, although cial financing of the distillery (and an inexpen- detailed economic analyses are not available sive feedstock source) could lower the selling from the current pilot plant operation. price to a level competitive with the corn-derived ethanol from distilleries not specially fi- University of California at Berkeley (Wilke) nanced. (Because of the larger investment, special financing lowers the price more than it Wilke has concentrated on changing the pre- wouId for corn distiIleries. ) treatment step of the Natick process by using acid and hammer milling of the wastepaper Alternatively, distilleries based on the older and field residues feedstocks. Nevertheless, a acid hydrolysis methods can be built to pro- critical step involving the recycling of enzymes duce ethanol and chemical feedstocks. Katzen has not yet been demonstrated. Associates, for example, has reevaluated the Madison process* on this basis and found that Iotech Process the ethanol could be sold at about $1 .50/gal without byproduct credits (1978 dollars) .34 The This process is proprietary and the subject of economics, however, depend on the byproduct patent applications in the name of the Cana- credits for the chemical feedstocks, but the dian Research and Development Corp. Appar- chemical industry is unlikely to make the com- ently, the novel aspect of the process is the mitment necessary to support a large fuel etha- pretreatment of the material before hydrolysis. nol industry until more information is avail- In this process woodchips are exposed to high- able on the relative merits of biomass- and -pressure steam for several seconds, followed coal-derived chemical feedstocks. by explosive decompression. The product is said to be highly susceptible to hydrolysis. As suggested earlier, the key to producing ethanol from Iignocellulosic materials at a price competitive with corn-derived ethanol Generic Economics of Lignocellulosic without relying on special financing or large Materials to Ethanol byproduct credits is the R&D currently aimed at reducing equipment costs, increasing over- The processes described above represent a all yields, and ensuring a good recovery of sampling of the possible approaches to etha- process chemicals without the production of nol production from lignocellulosic materials. toxic wastes. The descriptions were necessarily brief and could not include all of the ramifications or aspects of the various research groups’ efforts.

The chemistry and physics of Iignocellulosic materials are complex, and there are few pre- dictive theories that enable one to evaluate unambiguously the various approaches. Fur- *Dilute acid hydrolysis process Products are ethanol, furfural, thermore, the competition between research and phenol “Raphael Katzen Associates In The Feasib///ty of Uti/izing For- groups is enormous and details are often pro- est Res/dues for frtergy and Chernica/s (Madison, WIS Forest prietary. Products Laboratories, March 1976), report NO PB-258-630 Ch. 8—Fermentation . 173

R&D currently underway could fulfill these Table 55.-Plausible Cost Calculation for Future Production criteria. If so, the production costs might look of Ethanol From Wood, Grasses, or Crop Residues (in a 50-million-gal/yr distillery, early 1980 dollars) something like those in table 55. These costs represent plausible cost goals for the produc- Dollars tion of ethanol from IignocelIulosic materials. Fixed investment ... , ... ., ., ... ., ... $120 million Working capital ...... , ., . . ., . . ., 12 million Distilleries can and may be built before Total investment ...... , ., ...... $132 million these criteria are fulfilled, but the economics will depend on favorable financing and atypi- $/gallon calIy low feedstock costs or in securing a mar- Labor, chemicals, fuel ., ., ., . . . ., . . $0.30 ket for chemical byproducts. Some distilleries Feedstock ($30/ton, 110 gal/ton) ., ., ., . . . 0.27 Capital charges (15 to 30% of total investment) . . . . . 0.36-0.72 based on these circumstances are likely to be Total ...... , ., ...... , ., ... $0.93-$1.29 built before the late 1980’s. It is unlikely, however, that such circumstances will sustain a SOURCE : Office of Technology Assessment large fuel ethanol industry.

Environmental Impact of Ethanol Production

The major potential causes of environmen- likely fuels for these plants will be coal or tal impacts from ethanol production are the biomass (crop residues and wood), however, emissions associated with its substantial and thus the most likely source of problems energy requirements, wastes from the distilla- will be their particulate emissions. Coal and tion process, and hazards associated with the biomass combustion sources of the size re- use of toxic chemicals (especially in small quired for distilleries —especially distilleries plants). A variety of controls and design alter- designed to serve small local markets–must natives are available to reduce or eliminate be carefully designed and operated to avoid adverse effects, however, so actual impacts high emission levels of unburned particulate will depend more on design and operation of hydrocarbons (including polycyclic organic the plants than on any inevitable problems matter). Fortunately, most distilleries will be with the production process. located in rural areas; this will reduce total population exposure to any harmful pollut- New large energy-efficient ethanol plants ants. Particulate control equipment with effi- probably will require at least 50,000 Btu/gal of ciencies of 99 percent and greater are avail- ethanol produced to power corn milling, dis- able, especially for the larger plants. If all tilling, still age drying, and other operations energy requirements are provided by a single (see “Energy Consumption” discussion). Small boiler, high efficiency control would be easier plants will be less efficient. Individual distill- to provide. This is also true for any sulfur oxide eries of 50-milIion-gal/yr capacity wilI use (SO ) controls (scrubbers) that may be required slightly more fuel than a 30-MW powerplant; * x if the faciIity is fueled with high-sulfur coal. a 10-billion-gal/yr ethanol industry (the approximate requirement for a 10 percent alco- Other air emissions associated with ethanol hol blend in all autos) will use about the same production include fugitive dust from raw ma- amount of fuel needed to supply 6,000 to 7,000 terial and product handling; emissions of or- MW of electric power capacity. ganic vapors from the distillation process (as New source performance standards have not much as 1 percent of the ethanol, as well as other volatile organics, may be lost in the proc- been formulated for industrial combustion fa- ess); and odors from the fermentation tanks. cilities, and the degree of control and subse- These emissions may be tightly controlled by quent emissions are not predictable. The most water scrubbing (for odors and organics) and * Assuming 1 (),0()() Btu k Ilowatthour cyclones (for dust). 174 ● Vol. I/—Energy From Biological Processes

The “still age” — the waste product from the smaller units. Because smaller units are unlike- first still (or “beer still’’)–will be extremely Iy to have highly efficient particulate controls,

high in organic material with high biological this problem will be aggravated. Also, SOX and chemical oxygen demand, and will also scrubbers are impractical for small boilers, and

contain inorganic salts, and possibly heavy effective SOX control may be achieved only metals and other pollutants. When corn is the with clean fuels or else forgone. Because local biomass feedstock, the stillage is the source of coals in the Midwest tend to have high sulfur dried DC, which is a valuable cattle feed contents (5 percent sulfur content is not unusu- whose byproduct value is essential to the eco- al), small distiIIeries in this region may have ob-

nomics of the process. Thus, it will be recov- jectionably high SOX emission rates. Finally, ered as an integral part of the plant operation small plants will be less efficient than large and does not represent an environmental plants and will use more fuel to produce each hazard. On the other hand, sugarcane stillage gallon of alcohol. has far lower economic potential as a byprod- The decentralization of energy processing uct; its recovery is unlikely except as a re- and conversion facilities as a rule has been sponse to regulation. viewed favorably by consumer and environ- The stillage and other wastes from all etha- mental interests. Unfortunately, a proliferation nol plants have severe potential for damaging of many small ethanol plants may not provide aquatic ecosystems if they are mishandled. a favorable setting for careful monitoring of The high biological and chemical oxygen de- environmental conditions and enforcement of mand levels in the stillage, which would result environmental protection requirements. Regu- in oxygen depletion in any receiving waters, latory authorities may expect to have prob- will be the major problem. 35 Control tech- lems with these facilities similar to those they niques are available for reducing impacts from run into with other small pollution sources. For these wastes. Biological treatment methods example, the attempts of the owners of late- (activated sludge, biological filters, anaerobic model automobiles to circumvent pollution digestion, etc.] and land disposal techniques control systems conceivably may provide an used in the brewing industry are suitable for analog to the kinds of problems that might be ethanol production, but controls for stillages expected from small distilleries if their con- from some crop materials will require further trols prove expensive and/or inconvenient to development and demonstration. operate.

Because fermentation and distillation tech- The same may be true for considerations of nologies are available in a wide range of sizes, occupational safety. The current technology small-scale onfarm alcohol production may for the final distillation step, to produce play a role in a national gasohol program. The anhydrous (water-free) alcohol, uses reagents scale of such operations may simplify water ef- such as cyclohexane and/or ether that could fluent control by allowing land disposal of pose severe occupational danger (these chemi- wastes. On the other hand, environmental con- cals are toxic and highly flammable) at inade- trol may in some cases be more expensive be- quately operated or maintained distilleries. cause of the loss of scale advantages. Current Similar problems may exist because of the use experience with combustion sources indicates of pressurized steam in the distillation process. that high emissions of unburned particulate Although alternative (and safer) dehydrating hydrocarbons, including polycyclic organic technologies may be developed and automatic matter, are a more common problem with pressure/leak controls may eventually be made available (at an attractive cost) for small plants, in the meantime special care will have JsCarjbbean Rum Study,. Effects of Distillery Wastes on (he Na- rine Environment (Washington, D C Off Ice of Research and De- to be taken to ensure proper design, operation, velopment, Environmental Protect Ion Agency, April 1979) and maintenance of these smaller plants. Ch. 8—Fermentation ● 175

Process Innovations

The processes for producing ethanol from that the fermentation proceeds rapidly—there- sugar and grains are well established, but the by reducing the size and number of fermenta- traditional concern of the industries who oper- tion vessels required. The two ways of doing ate them has been the flavor (or, in some cases, this are through continuous fermentation or chemical purity) of the product. With the pro- through recycling of the yeast. duction of fuel ethanol, on the other hand, the Continuous fermentation processes have principal concerns are cost and energy effi- been tested in full-scale operation. Due to the ciency. There are several possible process possibility of infection of the mash (resulting in improvements — at various stages of develop- the production of products other than etha- ment — which can result in modest reductions nol), the processes have two complete fermen- in the processing cost and energy usage. Ex- tation systems to allow periodic switchover cept for improvements in grain and sugar proc- and sterilization. The added cost for this essing, the R&D could also be applicable to the equipment effectively nullifies the cost advan- production of ethanol from cellulosic materi- tage of continuous fermentation. 38 Improved als. Some possible improvements in grain proc- handling techniques, which can assure sterile essing, fermentation, and alcohol recovery are operation, may obviate the necessity for this mentioned below. redundancy in equipment. Grain and Sugar Processing One type of continuous fermentation that is under R&D uses a vacuum over the fermenta- Developments in the last 20 years have led tion mash. The ethanol is drawn off by the to more or less continuous grain preprocessing vacuum as it is produced, with the necessary techniques which have lowered the costs over heat for the evaporation of the alcohol being the traditional batch processes. Novel meth- supplied by the fermentation process itself. ods have been proposed, however, such as This would reduce the need for cooling water heating the mash with electrical current rather as well as accelerate the fermentation (which than process steam. This allows production of is slowed by high ethanol concentrations). a more concentrated sugar solution, thereby While added equipment costs might reduce or reducing the load on evaporators at later nullify the potential savings, the question of stages in the operation. While this is a more whether this will be the case has not been re- energy-intensive pretreatment, it could lower solved. the overall processing energy. 36 Another way of maintaining a high yeast The principal problem with sugar feed- concentration is by recycling the yeast (after it stocks, as noted, is the necessity of processing is separated from any grain solids that are to large quantities of feedstock to a syrup for be sold as a byproduct). A hybrid of yeast re- storage. At Ieast one research group is studying cycling and continuous fermentation involves ways to store the sugar crops without reduc- a device called a countercurrent flow fermen- tion to syrup,37 but the details are proprietary. tation tower,39 in which the yeast flows one way (counter to the current) whiIe the sugars to Fermentation be fermented flow in the opposite direction. The high yeast concentrations require addi- The key to cost reductions in fermentation is tional cooling of the mash, which increases the the use of methods for maintaining a high cooling equipment costs somewhat, but re- yeast or bacteria concentration in the mash, so search in this area can probably result in some overall cost savings. “Raphael Katzen Associates, Crain Motor Fue/ A/coho/, Tech- nical and Econom(c A~sessment Study, op c[t “Raphael Katzen Associates, Grain Motor Fuel Alcohol, Tech- “E Llplnsky, Battelle Columbus Laboratories, Columbus, nica/ and Economic Assessment Study, op. clt Ohio, private communtcatlon, 197!I “lbld 176 • Vol. Il—Energy From Biological Processes

Distillation moved from the mash first, so that they do not interfere with the ethanol concentration step The distillation process in the corn-to-fuel- (e.g., by clogging the membrane). Little re- ethanol distillery considered above consumes search has been done in producing a clarified nearly half of the energy used at the distillery. solution from the mash, hence, the costs for Lowering the energy requirements for separat- these methods are highly uncertain. ing the ethanol from the mash is desirable for a fuel ethanol facility in any case, but the in- Numerous other suggestions exist, and re- creased equipment costs for advanced ethanol search in these areas may eventually produce separation techniques could counter part or all usable results. One example is the use of supercritical CO . When gases are subject to high of the potential cost savings from lower fuel 2 use and smaller boiler and fuel-handling re- pressures at suitable temperatures, they form a quirements. Consequently, R&D into this area fluid which is neither gas nor liquid, but is must address both the energy use and the called a supercritical fluid. The properties of equipment cost. supercritical fluids are largely unresearched, but there are proprietary claims that super- One way to lower the energy requirements critical CO2 could be suitable for extracting of distillation is to produce a mash with an ethanol from the mash. The pressure would ethanol concentration higher than the usual 10 then be lowered, the CO2 would become a gas, percent. This would require development of and the ethanol would Iiquefy. yeast or bacteria that are tolerant of the high alcohol concentrations. Since it would be ex- Another possibility is the use of phase sep- pected that any yeast or bacteria producing arating salts. Salts, when dissolved in a liquid ethanol would produce it more slowly at the change the liquid’s structure and properties. It higher ethanol concentrations, this might re- has been suggested that there may be salts quire longer fermentation times with a conse- which would attract the water (or ethanol) so quent increase in the cost of fermentation vigorously and selectively that the ethanol- equipment. It may be possible, however, to water mixture wouId separate into two phases, combine this with advanced fermentation with one being predominantly water and the methods to provide an overall savings. other predominantly ethanol.

Several methods have been suggested for re- These novel approaches should be investi- moving the ethanol from the water. These in- gated, but it is not possible to predict when or clude: if results applicable to commercial fuel ethanol production will emerge. ● membranes using reverse osmosis (something like a super filter that allows the Producing Dry Ethanol water or ethanol to pass through the membrane while preventing the other compo- In a large, commercial distillery, the produc- nent from doing so); tion of dry ethanol only costs $0.01 to $0.03/gal ● absorption agents (solids which selective- (of ethanol) more than the production of 95 ly absorb the ethanol are then separated percent ethanol .40 (The difference in the selling from the solution, with the ethanol finally price per gallon of 99.5 percent ethanol and 95 being removed from the solid); and percent ethanol is due primarily to the fact ● liquid-liquid extraction (extracting the eth- that the latter contains 4.5 percent less ethanol anol into a liquid that is not soluble in per gallon of product.) Furthermore, with mod- water, physically separating the liquids, ern heat recovery systems, the production of and removing the ethanol from the other dry ethanol requires very little additional ener- liquid). gy. Consequently, little economic or energy savings are available here. All of these processes, however, are likely to require that the yeast and grain sol ids be re- ‘“I bid Ch. 8—Fermentation ● 177

On the other hand, the additional cost of move the water from 95 percent ethanol. vari- equipment for producing dry ethanol automat- ous chemicals are known to do this and recent ically onfarm with conventional technology research indicates that corn stover or corn may be prohibitive. If the distillery is of the grain may even be suitable.41 It is not known, labor-intensive type, however, the additional however, how much ethanol would be lost in equipment cost would be small since the same the process or, if grain is used, whether the ab- still could be used to produce 95 percent etha- sorbed ethanol would inhibit the production of nol and then later used to distilI to dry ethanol. sugar from the starch. While the processes are undoubtedly technicalIy possible, the econom- Drying agents or desiccants, however, may ics are still highly uncertain. be a suitable substitute for the conventional process. These materials would selectively re- 4’M. R Ladlsch and K Dyck, Science, vol 205, p 898, 1979 Chapter 9 ANAEROBIC DIGESTION Chapter 9.– ANAEROBIC DIGESTION

Page Page Introduction ...... 181 57. Anaerobic Digester Systems...... 185 Generic Aspects of Anaerobic Digestion. . .181 58. investment Cost for Various Anaerobic Basic Process ...... 181 Digester System Options ...... 192 Feedstocks ...... 182 59. Anaerobic Digestion: Energy Balance, or Byproducts...... 182 the Energy Production Potential and Reactor Types...... 184 Energy Consumption Potential of Single-Tank Plug Flow ...... 184 individual Farms in Major Producing States. 193 Multitank Batch System...... 184 60. Cost of Various Digesters With Electric Single-Tank Complete Mix...... 184 Generating Capabilities ...... 194 Anaerobic Contact, ...... 184 61. Annual Costs and Returns From Digester Two or Three Phase ...... 187 Energy Only ...... 194 Packed Bed ...... 188 62. Economic Feasibility of Anaerobic Expanded Bed ...... 189 Digestion , ...... 195 Variable Feed...... 189 Existing Digester Systems ...... 191 Economic Analysis...... 191 Environmental lmpacts of Biogas Production: Anaerobic Digestion of Manure...... ,195 FIGURES Research, Development, and Demonstration Needs ...... 198 Page Microbiology . . . . .198 28. Chinese Design of a Biogas Plant...... 185 Engineering ...... 198 29. Diagram of a Gobar Gas Plant ...... 186 Agriculture...... 98 30. Plug Flow Digestion System . , ...... 186 31, Single-Tank Complete Mixed Digester .. ...187 32. Two-Stage Digester ...... 187 TABLES 33. Two-Phase Digestion of Cellulosic Feed. ...188 Page 34. Packed-Bed Digester ...... 189 56. Suitability of Various Substrates for 35. Expanded-Bed Reactor...... 190 Anaerobic Digestion...... 183 36. Variable Feed Systems ...... 190 Chapter 9 ANAEROBIC DIGESTION

Introduction

Anaerobic (“without air”) digestion is the ple, inexpensive, existing technology. CO2 can process that occurs when various kinds of bac- be removed by a somewhat more complex and teria consume plant or animal material in an expensive technology, which would need to be airtight container called a digester. Tempera- employed if the gas is to be fed into a natural tures between 950 and 140°F favor bacteria gas pipeline. that release biogas (50 to 70 percent methane The anaerobic digestion process is especial- — essentially natural gas — with most of the rely well adapted to slurry-type wastes and has mainder as carbon dioxide — CO ). The bacte- 2 environmental benefits in the form of treating ria may be present in the original material wastes to reduce pollution hazards and to re- when charged (as is the case with cattle ma- duce odor nuisances. Furthermore, the residual nure) or may be placed in the digester when it from the process can be returned to land, ei- is initially charged. The gas has the heat value ther directly or through animal refeeding tech- of its methane component, 500 to 700 Btu/ nologies, and thus retain nitrogen and organic stdft3, and can be used directly as a heat fuel levels of soil. Most other biomass energy con- or in internal combustion engines. I n some version processes more nearly totally destroy cases there is enough hydrogen sulfide (H S) 2 the input material. present to cause corrosion problems, particularly in engines. H2S can be removed by a sim-

Generic Aspects of Anaerobic Digestion

The anaerobic digestion process involves a ganic acids, and 3) conversion of the acids to number of different bacteria and a digester’s methane. Accomplishing these steps involves performance depends on a large number of at least two different types of bacteria. variables. The basic process is considered first and then the feedstocks and byproducts of the The rate at which the biogas forms will de- process. pend on the temperature (higher temperature usually gives a faster rate) and the nature of the substrate to be digested. Cellulosic mate- Basic Process rials, such as crop residues and municipal solid waste, produce biogas more slowly than sew- Not all of the bacteria involved in anaerobic age sludge and animal manure. Disturbances digestion have been identified and the exact of the digester system, changes in temperature, biochemical processes are not fully under- feedstock composition, toxins, etc., can lead stood. Basically, however, the process consists to a buildup of acids that inhibit the methane- of three steps: 2 1 ) decomposition (hydrolysis) producing bacteria. Generally, anaerobic di- of the plant or animal matter to break it down gestion systems work best when a constant to usable-sized molecules such as sugar, 2) temperature and a uniform feedstock are conversion of the decomposed matter to or- maintained.

1 j j WOIIS, Arnerlcan /ourna/ 01 C /inlcal Nutrition, 27 (11), p When a digester is started, the bacterial 1120, 1974 composition is seldom at the optimum. But if ‘E C Clausen and I L Caddy, “Stagewlse Fermentation of f310mass to Methane, r’ Department of Chemical E nglneerlng, the feedstock and operating conditions are Unlverslty ot Ml~~ourl, Rolla, Mo , 1977 held constant a process of natural selection

181 182 ● Vol. II—Energy From Biological Processes

takes place until the bacteria best able to manure, various aquatic plants, and wet food- metabolize the feedstock (and thus grow) dom- processing wastes such as those that occur in inate. Biogas production begins within a day or the cheese, potato, tomato, and fruit-process- so, but complete stabilization sometimes takes ing industries. See table 56 for a summary of months. the suitability of various feedstocks for digestion. Numerous sources for good anaerobic bacteria have been tried, though the process is basically one of hit and miss. The potential for Byproducts improvement cannot be assessed at this time. Future developments could produce superior The digester effluent contains bacteria as genetic strains of bacteria, but too little is well as most of the undigested material in the known about the process to judge if or when feedstock (mostly Iignocellulose) and the solu- this can be accomplished. It is quite possible bilized nutrients. The process has the potential that if such strains are to be effective, the in- for killing most disease-causing bacteria, but put material may first require pasteurization. volatile losses of ammonia may increase with anaerobic digestion. 5 Biogas yields vary considerably with feedstock and operating conditions. Operating a The most generally accepted technology for digester at high temperatures usually increases disposal of the effluent is to use it as a soil con- the rate at which the biogas is formed, but rais- ditioner (low-grade fertilizer). Animal manure ing the temperature can actually decrease the is already used widely for this but there is some net fuel yield as more energy is required to controversy over whether the digester effluent heat the digester. The optimum conditions for is a better source of nitrogen than the undi- biogas yields have to be determined separately gested manure. * The actual added value (if for each feedstock or combination of feed- any) as a fertilizer, however, will have to be stocks. determined experimentally and is likely to be highly feedstock specific. The effluent may Feedstocks also be used as fertilizer for aquatic plant systems. In one case the effluent is dewatered and G A wide range of plant and animal matter can used as animal bedding in place of sawdust. be anaerobically digested. Both the gas yields Another potential use of the effluent is as an and rates of digestion vary, Generally materi- animal feed. It has been claimed that the pro- als that are higher in Iignin (e. g., wood and tein mix in the cake obtained from dewatering 3 crop residues ) are poor feedstocks because the effluent is superior to that of undigested the Iignin protects the cellulose from bacterial manure. 7 Biogas of Colorado has concluded a attack. Pretreatment could increase their sus- successful animal feeding trial of digester cake ceptibility to digestion. ’ However, even then and Hamilton Standard has also done feeding digestion energy efficiencies generally do not trials.8 exceed 50 to 75 percent. Thus, more usable energy can generally be obtained through combustion or thermal gasification of these feed- ‘J A. Moore, et al , “Ammonia Volatilization From Animal Ma- nures, ” in Biomass Uti/lzation in Minnesota, Perry Black shear, stocks (see ch. 5). ed , National Technlcai Information Service ● The nitrogen IS more concentrated In the effluent, but It IS The best feedstocks for anaerobic digestion also more volatile usually are wet biomass such as fresh animal ‘John Mart[n, Scheaffer and Roland, Inc , Chicago, Ill, private communication, 1980 ‘See also, J T Pfeffer, “131010 glcal Conversion of Crop Resi- ‘B G Hashlmoto, et al , “Thermophllllc Anaerobic Fermenta- dues to Methane, ” In Proceedings of the Second A nnual S ympo~i- tion of Beef Cattle Residues, ” In Symposium on Energy From Bio- urn on Fue/\ for Bioma\s, Troy, N Y , June 20-22, 1978 mass and Wastes, Institute of Gas Technology, Washington, 4P L McC~rty, et al , “Heat Treatment of Biomass for increas- D C , Aug 14-18, 1978 ing Blodegradabllity, ” In Proceeding~ of the Third Annual Bio- ‘D J Llzdas, et al , “Methane Generation From Cattle Resi- mass Energy Systems Conference, sponsored by the Solar Energy dues at a Dirt Feedlot, ” DOE report COO-2952-20, September Research Institute, Golden, Colo , June 1979, TP-33-285 1979 Ch. 9—Anaerobic Digestion ● 183

Table 56.-Suitability of Various Substrates for Anaerobic Digestion

Feedstock Availability Suitability for digestion Special problems Animal wastes Dairy ...... Small- to medium-sized farms, 30-150 head Excellent No major problems, some systems operating. Beef cattle ., ., ., Feedlots, up to 1,000-100,000 cattle Excellent Rocks and grit in the feed require degritting, some systems operating. Swine ...... 100-1,000 per farm Excellent Lincomycin in the swine feed will inhibit digestion–full-scale systems on university farms. Chicken ...... 10,000-1,000,000 per farm Excellent Degritting necessary, broiler operations need special design due to aged manure, tendency to sour. Turkey...... 30,000-500,000 per farm Excellent Bedding can be a problem, manure is generally aged, no commercial systems operating. Municipal wastes Sewage sludge . . . . . All towns and cities Excellent Vast experience. Solid wastes ...... All towns and cities Better suited to direct Designed landfill best option combustion Crop residues Wheat straw . . Same cropland Poor, better suited to Particle size reduction necessary, low digestibility, no direct combustion commercial systems. Corn stover...... Same cropland Poor, better suited to No commercial systems, no data available, particle size direct combustion reduction necessary. Grasses Kentucky blue. Individual home lawns Very good Distribution of feedstock disperse, no commercial systems. Orchard grass. Midwest Fair No commercial systems, no data on sustainability of yields. Aquatic plants Water hyacinth . . . . . Southern climates, very high reproduction Very good No commercial operations, needs pregrinding. rates Algae...... , . . . Warm or controlled climates Good Longer reaction time than for animal wastes. Ocean kelp ...... West coast, Pacific Ocean, large-scale kelp Very good Full-scale operations not proven, no present value for farms effluent. Various woods. . . . Total United States Poor, better for direct Will not digest. combustion or pyrolysis Kraft paper...... Limited Excellent, need to evaluate Premixing watering necessary. recycle potential and other conversion

SOURCE Tom Abeles and David Ellsworth, ‘‘Blologlcal Production of Gas, contractor report to OTA by I E Associates, Inc., Minneapolis, Minn. , 1979

Although most of the disease-causing bacte- its use as a feed or feed ingredient.9 The use of ria are killed by digestion of the manure, sev- digester effluents as an animal feed, however, eral questions about refeeding of digester ef- would greatly improve the economics of ma- fluents need to be resolved. Buildup of toxic nure digestion. Consequently, the value, use, materials, development of resistance to antibi- and restrictions on using digester effluents as otics by organisms in the cake, permissible animal feeds should be thoroughly investi- quantities of cake in the diet, storage, and gated. Moreover, the animal feed value of ef- product quality are all issues that have been fluents from the digestion of feedstocks other raised. There is no firm evidence that these will than manure should also be investigated. present significant problems, however.

To avoid some of these problems, the Food 9T P Abeles, “Design and Englneerlng Conslderatlons In Plug and Drug Administration has generally favored Flow Farm Digesters, ” In Symposium on C/can Fue/s From Bio- cross-species feeding, but has not sanctioned mass, Institute of Gas Technology, 1977 184 . Vol. Ii—Energy From Biological Processes

Reactor Types

There are numerous possible designs for an- in the digester and the feed or digester con- aerobic digesters, depending on the feedstock, tents can be heated as needed. Depending on the availability of cheap labor, and the pur- the placement of the heating pipes, some con- pose of the digestion. The most complex and vective mixing can also occur. expensive systems are for municipal sewage sludge digestion, but the primary purpose of Multitank Batch System these has been to stabilize the sludge and not to produce biogas. This system consists of a series of tanks or chambers which are filled sequentially with Digester processes have been classified into biomass and sealed. As each unit completes three types, depending on the operating temperature: 1 ) psychrophilic (under 680 F), 2) mes- the digestion process, it is emptied and recharged. This type of reactor is best suited to ophilic (680 to 1130 F), and 3) thermophilic operations where the feedstock arrives in (11 30 to 150° F). The cost, complexity, and energy use of the systems increase with the tem- batches, for example, grass or crop residues that are collected only at certain times of the perature, as does the rate of gas production. year or turkey or broiler operations that are The amount of gas produced per pound of cleaned only when the flocks are changed. feedstock, however, can either increase or de- This digester system, however, is relatively crease with temperature. Retention time is also labor intensive. an important consideration, wherein maximum gas production per pound of feedstock is sacri- ficed for reduced size and cost of the digester. Single-Tank Complete Mix Anaerobic digesters in the mesophilic and thermophilic ranges have used agricultural wastes, The single-tank complete mix system (figure residues, and grasses, to produce biogas. The 31) has a single rigid digester tank which is optimum temperature appears to be both site heated and mixed several times a day. It has and feedstock specific. There are still unre- been argued that mixing enhances the contact solved technical questions about the tradeoffs of bacteria with the feedstock and inhibits between mesophilic and thermophilic digest- scum formation, which can interfere with di- ers, but most onfarm systems have been meso- gester operation. Theoretical calculations, ’” philic. however, indicate that the mixing does not improve bacterial contact, and these calculations Other design parameters include continuous have been confirmed experimentally in one versus batch processes, mixed versus unmixed case. Single-tank complete-mixed digesters reactors, and other features. Some of the ma- are used to treat municipal sewage sludge and jor types are summarized in table 57 and dis- have been used in the larger anaerobic digester cussed briefIy. systems (exclusive of landfills).

Single-Tank Plug Flow Anaerobic Contact

This system is the simplest adaptation of The single-tank complete-mix system efflu- Asian anaerobic digester technology (figures ent can be transferred to a second unmixed 28-30). The feedstock is pumped or allowed to flow into one end of a digester tank and re- ‘(’P C Augensteln, “Technical Prlnclples of Anaerobic Diges- tion, ” Dynatech R&D CO , presented at course fl/otechno/ogy for moved at the other. Biogas is drawn off from Ut//izat/on of organic Wastes, Unlversldad Autonoma Metropoll- the top of the digester tank. The feed rate is tana, Iztapalapa, Mexico, 1978 “K D Smith, et al , “Destgn and First Year Operation of a chosen to maintain the proper residence time* 50,000 Gallon Anaerobic D[gester at the State Honor Farm Dairy, Monroe, Washington, ” Department of Energy contract EG- ● The time the feedstock remains in the digester 77-C-06-1016, ECOTOPE Group, Seattle, Wash , 1978 Ch. 9—Anaerobic Digestion • 185

Table 57.–Anaerobic Digester Systems

Type of system Application and inputs Scale a Stage of development Advantages Disadvantages

Landfill Low cost, tanks not re- Gas generation may Iast only municipal solid wastes, waste and up landfills, controlled land quired, high loading rates 10 years in “as is” land- sewage sludge, warm (28-acre filling in pilot stages possible, no moving parts fills, gas usage onsite may climates Iandfill) present problems Single-tank plug All types of organics, farm Small to large Commercial Low cost, simple design Low solids wastes may flow and feedlot operations can run high solids stratify wastes, can have gravity feed and discharge Multitank batch Can accept all types of Small to large Commercial, in Asia Simple, low maintenance, Gas generation not system wastes, limited application low cost, complete diges- continuous, labor-intensive crop residues, grasses, tion of materials feed and discharge, low gas chicken broilers, turkeys production per day Single-tank All types of organics sewage Small to large Commercial Proven reliability, works Greater input energy to run complete mix treatment, farm and feedlot, well on all types of mixers, higher cost than municipal solid wastes wastes plug flow Anaerobic contact Sewage sludge and other Medium to Commercial, for sewage Smaller tank sizes, Two tanks necessary organics, limited apphcation large treatment operation not overly (see variable feed) critical Two or three phase Celluloslc feedstocks Medium to Pilot scale Allows more complete Feed rates vary with large decomposition, greater feedstocks, have not been gas yields, greater load- attempted full scale, require ing rates, lower retention tight controls and manage- times ment of the operation Packed bed Dilute organics–sewage, Medium to Commercial, as waste High loading rates Tends to clog with organic food-processing wastes, large treatment technology possible, short retention particles, Iimited to dilute very dilute animal wastes— times wastes Industrial and commercial Expanded bed Dilute organics-sewage, Undetermined Laboratory High loading rates, low Not developed, high energy food-processing wastes, temperature digestion, input to operate pumps, no very dilute animal wastes high quality gas, short operating data retention times Mixed bed Sewage sludge, animal Small to large Pilot scale Fast throughput, high Tends to clog, high pumping wastes, food-processing loading rates, higher energy input no operating wastes–fairly dilute solids input than packed data mixtures beds Variable feed All types of organics, farms Small to Conceptual–combines Allows seasonal peaking of Feed-discharge may require and feedlots medium plug flow with anaerobic gas production, pre- extra pump contact serves nutrient value of material, low cost ascaje def[ned as small–o to 30,000 gal med!um–30 000 to 80000 gal. andlarge–over 80,000 gal SOURCE Tom Abeles and David Ellsworth Blologlcal Produchon of Gas ‘‘ contractor report to OTA by I E Associates Inc M!nneapohs, Mlnn 1979

Figure 28.—Chinese Design of a Biogas Plant

Gas vent Inlet pipe Outlet Cover Gas outlet pipe

Sludge Over- chamber flow

Base

SOURCE: K. C. Khandeleval, “Dome-Shaped Biogas Plan,” Compost Science, March/Apr!l 1978. 186 ● Vol. /l—Energy From Biological Processes

Figure 29.—Diagram of a Gobar Gas Plant (Indian)

rry pit

All dimensions in meters

SOURCE: R. B. Singh, “Biogas Plant,” Gobar Gas Research Station, Ajitmal, Etaweh (V.P.), India, 1971

Figure 30.-Plug Flow Digestion System

Cross section

Gas utilization system

SOURCE: W. J. Jewell, et al., “Low Cost Methane Generation on Small Farms,” presented at Third Annual Symposium on Biomass Energy Systems, Solar Energy Research Institute, Golden, Colo., June 1979 Ch. 9—Anaerobic Digestion . 187

Figure 31. —Single-Tank Complete Mixed Digester Two or Three Phase

As mentioned previously under “Generic As- pects of Anaerobic Digestion, ” the basic proc- Digester gas ess consists of a series of biochemical steps in- t volving different bacteria. The idea behind the multitank systems is to have a series of digester tanks (figure 33) each of which is separately Raw Active optimized for one of the successive digestion sludge 4 zone steps. The rationale behind such system is the exchanger hypothesis that they: 1 ) can accept higher feedstock concentrations without inhibiting the reactions in successive stages, 2) have greater process stability, 3) produce higher methane Digested sludge concentrations in the biogas, and 4) require lower retention times in the digester than with most single-phase digesters. The majority of SOURCE Environmental Protection Agency, “Process Design Manual, Sludge the work on this approach has been on munici- Treatment and Disposal,” EPA 625/1-29-001, September 1979 pal sewage sludge, although the Institute for and unheated storage tank. Here the biomass Gas Technology hopes to eventually transfer undergoes further digestion and solids settle the technique to kelp digestion. The need for out (figure 32). In other words, by adding a sec- uniform feed rates and controls may limit the ond, inexpensive digester tank gas yields can use of two or muItiphased systems to larger or extremely well-managed operations; but this be improved. These systems have been used extensively in sewage treatment and may re- type of reactor should be carefully examined ceive wide application where preservation of for other anaerobic digestion applications be- the effluents nutrient value requires covered cause of its potentialIy high efficiencies. lagoons or in short throughput systems located in warm climates.

Figure 32.—Two-Stage Digester

Digester gas

Transfer Supernatant sludge

1 Digested sludge - Primary digester Secondary digester

SOURCE Environmental Protect Ion Agency, “Process Design Manual, Sludge Treatment and Disposal,” EPA 625/29-001, September 1979 188 ● Vol. n-Energy From Biological Processes

Figure 33.—Two-Phase Digestion of Cellulosic Feed

Gas

SOURCE S Ghos and D. L Klass, “Two Phase Anaerobic Digestion,” Symposium on Clean Fuels from Biomass (Institute of Gas Technology, 1977)

Packed Bed (Anaerobic Filter) As it now exists, however, it is not well suited to energy production. In this system a dilute stream of feedstock is fed up through a verticle column packed with Like the packed and expanded bed, the small stones, plastic balls, ceramic chips, or mixed-bed systems are intended to provide an other inert materials (figure 34). Because the inert substance to which the bacteria can at- bacteria attach themselves to the inert mate- tach, thereby preventing them from being rial, it is possible to pass large quantities of flushed out with the effluent. The digester feedstock through it while maintaining a high maintains a higher bacteria population. Vari- 13 bacterial concentration in the digester. The ous designs include netting, strips of plastic, system is best suited to municipal sewage (and and rough porous digester walls. In all cases, other dilute feedstocks). More concentrated the inert substance increases the resistance to feedstocks tend to clog the column. flow and thus the energy needed for pumping increases too, but it decreases the necessary Analyses of bench-scale laboratory results reactor size. Sufficient data are not yet avail- on the AN FLOW system indicate that the sys- able for a detailed analysis of this tradeoff. tem could produce enough energy to make this sewage treatment step energy self-sufficient. 12

‘JR K Genung and C D Scott, “An Aneroblc Bloreactor (AN F1.OW) for Wastewater Treatment and Process Appl(ca- ‘ ‘S A St’rfl Ing and C Alsten, “An I nt~gratd, Controlled E nvl- tlons, ” briefing presented to the Subcommittee on E nergv and ronment Aquiculture Lagoon Proce\\ tor Secondary or Ad- Power, HOLJJ(J I nter$tate and Foreign Commerce Committee, vanced Wastt>water Treatment ,“ Sol ar Aqua\ y\tem\, I nc , En- Nov 1, 1979 clnltas, Cal If , 1978 Ch. 9—Anaerobic Digestion ● 189

Figure 34.—Packed-Bed Digester

System gas Process gas pressure chromatography Gas pump

gas meter Magnetic Flow 5 ft dia control valve flow x 10 ft high packed section pH and pH and temperature temperature probes probes Weir box composite composite flow meter samplers samples

Bar screens

SOURCE R K Genung. W W Pitt, Jr , G M Davis, and J H Koon, “Energy Conservation and Scale-Up Studies for a Wastewater Treatment System Based on a Fixed- Film, Anaerobic Bioreactor,” presented at 2nd Symposium on Biotechnology in Energy Production, Gatlinburg, Term Oct 2-5, 1979

Expanded Bed ever, that the process would not be a net energy producer due to the energy required to ex- A variation on the packed-bed concept is the pand the bed. expanded-bed reactor (figure 35). I n this case the column packing is sand or other very small particles. The feedstock slurry is fed up Variable Feed through the column and the bed of inert mate- The idea behind variable feed systems (fig- rial expands to allow the material to pass ure 36) is to store undigested manure in times through. A semifluidized state results, reduc- of low gas demand for use during periods of ing the potential for clogging when relatively high demands. The key is to be able to store concentrated material is fed into the reactor, the manure for long periods (e. g., 6 months) The process has been found to be quite stable without excessive deterioration. The effect of with high organic inputs, short residence times long-term storage is being investigated,15 but in the digester, and relatively low temperatures 14 the systems may be limited to areas with cool (50° to 70° F). The study did indicate, how- summers or to operations in which the gas is used to generate electricity for export during the peak electric demand periods in summer.

‘“W J Jewel 1, (.orn[~l I Un Iveriitp, I t ha( ,1, N } , {)rlv,ltt> c onlnjLl- nlc,l tlon, 1978 190 ● Vol. n-Energy From Biological Processes

Figure 35.—Expanded Bed Reactor

Gas I

Feed

Pump

I Recycle AAFEB II pump Effluent

SOURCE: M. S., Switzenbaum and W. J. Jewell, “Anaerobic Attached Film Expanded Bed Reactor Treatment of Dilute Organics,” presented at 51st Annual Water Pollution Control Federation Conference, Anaheim, Calif., 1978.

Figure 36.—Variable Feed Systems

H 2O

Manure

SOURCE Tom Abeles and Oawd Ellsworth, “Blologlcal Produchon of Gas, contractor report to OTA by I E Associates, Inc , Mmneapolm, Mmn 1979 Ch. 9—Anaerobic Digestion ● 191

Existing Digester Systems

Fourteen experimental and prototype digest- of cattle and has been functioning since late ers of animal manure were identified as opera- fall 1979. The biogas is fed into a dual-fuel tional in 1978. ’6 The capacity of these plants diesel engine and supplies about 90 percent of varies from less than 1,500 gal to 4 million gal. the engine’s fuel needs. The engine drives a Two of the prototypes are owned by individual 125-kW generator (for peak electric demands) farmers and are sized for farm use. The 12 and the generator has an average output of 45 others are owned by private firms, universities, kW. The system supplies essentially all of the or the Federal Government. operation’s direct energy needs.

Since then, however, the field of anaerobic Other systems are operational or are likely digestion has been advancing rapidly, and any to become operational soon. Nevertheless, op- list of existing operations would be quickly erating experience is limited and suitable di- outdated. Several companies currently design gesters for all types of manures and combined and sell digesters and the support equipment. animal operations are currently not available. Most systems are currently designed for cattle Consequently, commercialization of the tech- manure. One example of an apparently suc- nology could be helped by demonstrating a cessful digester system is on a dairy farm in wide range of digester systems in a variety of Pennsylvania. The digester is fed by 700 head confined animal operations so as to provide operating experience and increase the number “D L Klass, “Energy From Biomass and Wastes, ” In Syrnposi- urn on Energy from L?lomass and Waste, I nstltute of Gas Technol- of operations for which suitable digester sys- ogy, Washington, D C , Aug 14-18, 1978 tems exist.

Economic Analysis

Aside from the paper and other digestable less than 15 percent of the resource or less matter in municipal solid waste (which is not than 0.04 Quad/yr (see ch. 5 in pt. l). Since it is included in this report), the best feedstocks for relatively expensive to transport manure long anaerobic digestion are animal manure, some distances, this economic analysis concentrates types of grasses, aquatic plants, and various on digesters appropriate for onfarm use. processing wastes. The supply of aquatic The system analyzed consists of a plug flow plants is likely to be small in the next 10 years digester operating at 700 to 90° F, with a feed- and little information is available on the stock pond and effluent residue storage pit. digester requirements for grasses. Furthermore, (See top schematic, figure 36.) After removal of with grass at $30/dry ton, the feedstock cost the hydrogen sulfide (H S), the biogas is alone would be $4.50/milIion Btu. More energy 2 burned in an internal combustion engine to at a lower cost can usually be produced from generate electricity. The electricity is used on grass by thermal gasification or combustion. the farm (replacing retail electricity) and the Hence, animal manure and some processing excess is sold wholesale to the electric utility. wastes are the most promising near-term The waste heat from the generator engine is sources of biogas by far. The larger of these also used onfarm, but any excess heat goes to two sources is animal manure. waste. On the average,15 percent of the ener- More than 75 percent of the animal manure gy produced is used to heat the manure enter- resource is located on confined animal opera- ing the digester and for the other energy needs tions that h ave less than 800 dairy cows or the of the digester (e. g., pumping). (Other systems equivalent weight of other animals (e. g., vary from 10 to 40 percent, depending on the 250,000 chi ckens). Large feedlots account for type of digester and the operating conditions.) 192 ● Vol. n-Energy From Biological Processes

There is sufficient gas storage capacity to Table 59 gives the energy that could be pro- limit electric generation to those times of the duced with onfarm digesters. It also shows the day when the utility or the farmer has peak quantities that could be used onfarm and ex- electric demands; and the feedstock storage ported for various animal operations in some allows for seasonal variations in the average of the major producing States if farm energy daily energy production. Therefore, this sys- use stays at 1974-75 levels or if it decreases 25 tem can be used either as a peakload or base- percent due to energy conservation. In most Ioad electric generating system. In a mature cases, the digester energy output is sufficient system, the electric utility would be able to to meet the energy needs of the Iivestock oper- call for more or less electric generation from ation and in more than half of the cases con- onfarm units by sending coded signals along sidered, it also fills the farmer’s home energy the electric powerlines. needs and enables a net export of electricity. With conservation, the situation is even more Other systems are possible, including one in favorable with respect to energy exports. The which the water in the digester effluent is lower revenue that the farmer receives for largely removed (dewatered) and the resultant surplus energy as opposed to the replacement material sold as a fertilizer or animal feed. of retail energy, however, makes conservation Table 58 shows the cost of various systems; the less economically attractive unless the farmer basic digester cost represents the sum of the is not energy self-sufficient without conserva- costs for digester, pumps, pipes, hot water tion. In other words, it is more economically boilers, H2S scrubber, low-pressure gas com- attractive to replace retail electricity than to pressor, heat exchangers, and housing. The generate surplus electricity for sales at whole- cost of the manure premixing equipment is sale rates. also included with tanks larger than 40,000 gal. The digester size, capital investment, and However, these costs should be viewed as preliminary and approximate. operating costs for anaerobic digester-electric generation systems for these various operations are shown in table 60. Assuming the farm- Removal of the CO2 and sales of the gas to natural gas pipelines were assumed not to be ers can displace retail electricity costing 50 feasible in small operations because: 1) the gas mill/kWh, sell wholesale electricity for 25 pipelines often are not readily accessible, 2) milI/kWh, and displace heating oil used onfarm costing $6/million Btu, the returns from the cost of CO2 removal equipment is high, and 3) revenues from the gas sales would prob- the digester system are shown in table 61. Also ably be relatively low. In very large systems, shown are the farmer’s costs for two assumed though, production of pipeline quality gas may capital charges: 1 ) where the annual capital be feasible. charges are 10.8 percent 01 the investment, i.e.,

Table 58.–investment Cost for Various Anaerobic Digester System Options

Median capital costs ($1 ,000) Options Tank size (gallons) Basic digester Dewatering Electric generator Feedstock Iagoonb Residue pitc 10,000 ...... $ 19.6 $34.0 $4.0 $ 6.7 $0.4 20,000 ...... 33.7 34.0 5.0 13.4 0.8 40,000 ...... 48.1 34.0 6.0 26.9 1.6 80,000 ...... 65.6 45.0 12.0 53.7 3.2 200,000 ...... 98.8 60.0 45.0 133.0 7,5 400,000 ...... 143.6 90.0 70.0 268.3 15.2 aGeRerator IS In operation 12 hours each day bStorage for 6 months cStorage for 9 months. SOURCE Tom Abeles and Dawd Ellsworlh, ‘Biological ProductIon of Gas, ” contractor repofl 10 OTA by I E Associates, Inc , Mmneapolls, Mmn , 1979 Ch. 9—Anaerobic Digestion ● 193

Table 59.–Anaerobic Digestion: Energy Balance, or the Energy Production Potential and Energy Consumption Potential of Individual Farms in Major Producing States

Methane energy Demands of livestock Household use Excess energy Excess energy options operation (1975 levels) (1974 levels) (25% conservation Direct Electricity 1974 levels 25% conserv. Direct Electricity Direct Electricity use Waste use Waste use Waste only heat only heat only heat 1974 livestock average number sold 1,000 1,000 1,000 1,000 1,000 1,000 (inventory )/farm Btu 106 kWh Btu 106Btu 106 kWh Btu 106 kWh Btu 106 kWh Btu 106 kWh Btu 106Btu 106 kWh Btu 106 Turkeys Minnesota (124,000)...... , 14,914 877 9,545 8,754 248 6,559 186 326 16 5,835 613 465 8,029 675 2,660 California (98,000) ...... 11,787 693 7,544 4,714 147 3,536 110 180 16 6,893 530 2,650 8,071 567 3,828 North Carolina (89,000 ...... 10,705 630 6,851 1,931 160 1,448 120 128 22 8,646 448 4,792 9,129 488 5,275 Broilers Minnesota (52,000), ...... 522 31 334 357 14 268 11 163 8 29 -186 91 12 -97 (198,000)...... 1,986 117 1,271 1,366 55 1,024 41 163 8 457 54 -258 799 68 84 California (56,000) ...... 562 33 360 350 10 262 8 90 8 122 15 -80 210 17 8 (1,377,000)...... 13,812 812 8,840 8,606 244 6,454 183 180 16 5,026 552 54 7,178 613 2,206 Arkansas (63,200) ., ., ., . . 634 37 406 314 9 236 7 107 9 213 19 -15 291 21 63 (186,000) . . . . 1,866 110 1,194 937 27 703 20 107 9 822 74 150 1.056 81 384 Swine lowa ., ., ... ., ... ., 325a 19 208 147 28 110 21 155 8 23 -17 -94 60 –lo -57 Missouri . . 325a 19 208 69 22 52 16 136 8 120 -11 3 137 -5 20 North Carolina (500) ., . . . . . 325a 19 208 29 13 22 10 64 11 232 -5 115 239 -2 123 Dairy cows Wisconsin (36) ., ., ...... , ., 261 15 167 11 16 8 12 161 8 89 -9 -5 92 -5 -2 New York (71) ., ., ., 521 31 333 25 33 19 25 126 8 370 -lo 182 376 -2 188 California (337) ...... 2,459 145 1,574 370 133 278 100 270 24 1,819 -12 934 1,911 21 1,026 Laying hens Minnesota (13,000). ., . . . ., 846 50 541 213 34 160 26 163 8 470 8 165 523 16 218 (41,000) ...... 2,670 157 1,709 672 106 504 80 163 8 1,835 43 874 2,003 69 1,042 Georgia (14,000) ...... 912 54 584 140 34 105 26 72 10 700 10 372 735 18 ’407 (41,000) ...... 2,670 157 1,709 410 98 308 74 72 10 2,188 49 1,227 2,290 73 1,329 California (14,000) . . . . 912 54 584 224 55 168 41 90 8 598 -9 270 654 5 326 (105,000) ...... 6,838 402 4,376 1,680 414 1,260 310 180 16 4,978 –28 2,516 5,398 76 2,936 Beef(500) ...... 1,527 90 977 – — — — 160 10 1.427 80 817 – — —

alncludes breeding stock Assumptions 15% of blogasused lorun digester system electrrc generational 20% effmency,and80% of the engmewaste heatcanbe recaptured SOURCE Tom Abelesand David Ellsworth “Bloloqlcal ProducNon of Gas contractor reporl to OTA by I E Associates Inc , MmneaDohs, Mlnn 1979, Enerayafld L/ S Awrcu/ture, 1974 Da/a Base “. (Washington DC Energy Research and_Development AdmmlstraNon, 1974)

9-percent interest loan with 20-year amoritiza- with 198,000 birds in Minnesota is more eco- tion, and 2) where the annual capital charges nomically attractive than an equivalent oper- are 15 percent of the investment. ation with only 52,000 birds (see table 61). On the other hand, the 52,000-bird broiler farm The principal cost factor in anaerobic diges- (equivalent to 250 head of cattle in terms of tion is the capital charge, or the cost of the di- the quantity and quality of manure) is more gester itself—thus, favorable financing is the economically attractive than a 500-head cattle most effective way of reducing the cost to the feedlot, because the poultry operation con- farmer. sumes considerably more energy, and thus Financing aside, the anaerobic digestion op- could better utilize the digester output within erations that are most economically attractive its own enterprise. are relatively large poultry, dairy, beef, or swine operations (enabling an economy of Based on 1978 fuel prices, the feasibility of scale) which are also relatively energy inten- anaerobic digestion was assessed for various sive (enabling the displacement of relatively types of farm animal operations in the various Iarge quantities of energy at retail prices) For regions of the country. It was found that it example, anaerobic digestion on a broiler farm would be feasible to digest so percent of the 194 . Vol. II—Energy From Biological Processes

Table 60.–Cost of Various Digesters With Table 61 .–Annual Costsand Returns From Digester Energy Only Electric Generating Capabilities Return from 1974 livestock Capital Annual energy average number sold Digester size investment operating costs displacement Digester costs (inventory )/farm (1,000 gal) ($1 ,000) ($1 ,000) and sales of (operating + capital) ($1,000) Turkeys 1974 livestock electricity 10.8% annual 15% annual Minnesota (124,000). 300 220 4.6 (inventory )/farm ($1,000) capital charge capital charge California (98,000) . . . . 250 195 3.9 Turkeys North Carolina (89,000). . 225 182 3.7 Minnesota (124,000). . . . 81 28 38 Broilers California (98,000) . . . . . 51 25 33 Minnesota (52,000) . . . . 12 25 0.9 North Carolina (89,000).. 36 12 31 (198,000). . . . 45 62 2.6 Broilers California (56,000) . . . . . 13 27 0.9 Minnesota (52,000) . . . . 3,3 3.6 4.7 (1 ,377,000). . . 300 220 4.4 (198,000) . . . 12 9.3 12 Arkansas (63,200) . . . . . 14 28 1.0 California (56,000) . . . . . 3.4 3.8 5.0 (186,000) . . . . 40 59 2.6 (1,377,000). . . 80 28 37 Swine Arkansas (63,200) . . . . . 3.8 4.0 5.2 Iowa (500)...... 10 27 0.8 (186,000) . . . . 9.9 9.0 11 Missouri (500) ...... 10 27 0.8 Swine North Carolina (500) ... , 10 27 0.8 lowa (500)...... 2.2 3.7 4.9 Dairy cows Missouri (500) ...... 2.2 3.7 4.9 Wisconsin (32)...... 10 24 0.6 North Carolina (500) . . . . 1.5 3.7 4,9 New York (64) ...... 10 30 0.8 Dairy cows California (337)...... 80 92 1.6 Wisconsin ...... 1.8 3.2 4.2 Laying hens NewYork (64) ...... 2.5 4.0 5.3 Minnesota (13,000) . . . . 34 50 1.6 California (337) ...... 11 12 15 (41,000) . . . . 100 103 1.7 Laying hens Georgia (14,000) ...... 34 50 1.6 Minnesota (13,000) . . . . 4.6 7 9.1 (41,000) ...... 100 103 1.7 (41,000) . . . . 12 13 17 California (14,000) . . . . . 35 51 1.6 Georgia (14,000) ...... 3.7 7. 9.1 (105,000) . . . . 250 169 3.0 (41,000) ...... 9.5 13 17 Beef(500)...... 20 43 0.7 California (14,000) . . . . . 4.6 7.1 9.3 (105,000) . . . . 31 21 28 SOURCE SOURCE Tom Abeles and David Ellsworth, ’’Biological Production of Gas,’’ contractor 600/(500) ...... report to OTA by l. E Associates, lnc ,Minneapolis, Minn. ,1979 3.2 5.3 7.2

SOURCE SOURCE Tom Abeles and David Ellsworth, “Biological Production of Gas, ” contractor animal manure to produce electricity and on- report to OTA by I E Associates, Inc. , Minneapolis, Minn., 1979. site heat if the effective annual capital charges are the quantities of manure that would be fea- were 6.6 percent of the investment. (Digestion sible if the digester effluent were dewatered was deemed feasible if the returns from dis- and sold as a fertilizer at $1 O/dry ton over the placing onsite energy use and wholesaling ex- revenues available from sales of the raw ma- cess electricity were greater than the capital nure as a fertilizer. Furthermore, if the dewa- and operating costs for the anaerobic diges- tered effluent could be sold for feed, higher tion energy system.) This effective capital credits may be possible based on the protein charge could be achieved by a 9-percent in- content of the effIuent.17 Although the feasibil- terest, 20-year loan with 4.2-percent annual tax ity of these higher credits is unproven as yet, writeoff. Other possible credits could be avail- the selling of digester effluent as feed or fertil- able through combinations of Agricultural izer would substantially expand the quantity Stabilization and Conservation Service pol- of manure that could be digested economi- lution abatement cost sharing, soil conserva- cally. tion district cost sharing, energy credits, and other incentives, although these are not in- Turkey farms tend to be the most economic cluded in the feasibility calculations. In table because of their rather large average size and 62, the percent of the manure resource that 17 would be feasible for energy production with Blogas of Colorado, “Energy Potential Through Bloconver- sion of Agricultural Wastes, ” Phase //, Fina/ Report to the Four the 6.6-percent capital charge is shown for Corners Regiona/ Commission, demonstration pro]ect FCRC No various manure types and regions. Also shown 672-366,002, Arvada, Colo , 1977 Ch. 9—Anaerobic Digestion ● 195

Table 62.–Economic Feasibility of Anaerobic Digestion (percent of total manure resource that can be utilized economicallya)

Total usable Region manure problem Layers Broilers Turkeys Cattle on feed Dairy Swine Northeast ...... 52% 68% 82% 98% 6% 43% — Southeast ...... 65 54 75 90 45 80 — Appalachia ...... 47 51 85 89 8 26 — Corn Belt ...... 17 38 67 89 9 19 8% Lake States ...... 30 50 90 98 6 23 8 North Plains ...... 39 37 46 96 55 12 — Delta ...... 69 68 82 86 21 37 — South Plains ...... 82 76 87 94 90 49 — Mountain...... 75 82 44 95 83 49 — Pacific ...... 88 78 97 98 90 89 — Alaska...... 69 0 0 0 0 82 — Hawaii. ..., ...... 70 35 82 0 89 99 — National totals 49+ 20b 59 81 94 61 41 5 With fertilizer enhancement assumption of $10/dry ton . . . . . 69 + 30b 85 95 99 72 60 35

Investment Also assumes 1978 energy costs as follows home healing $3 80/mdhon Btu, farming heat $5 40/mlHlon Blu, retadelectrlclty accordlngto DOE, Typca/Hectm Ms. .larruary 1978, October 1978, and wholesale elecrrlclty 25 mdl/kWh bEstlmated uncertainty These correspond to weighted average percentages SOURCE Tom Abeles and Dawd Ellsworfh, ‘Blologlcal Produchon of Gas, ‘‘ contractor report to OTA by I E Associates, Inc , Mmneapolls, Mmn , 1979 the relatively large amount of thermal energy is, however, the possibility of expanding the consumed by them. Swine operations, how- operation to use the excess heat, for example, ever, are usually too small to be economically by building greenhouses. This could improve attractive, for the energy alone, but because of the economics, but it would require major ad- odor problems these may also be attractive. justments in the farmer’s operation. In the end, site-specific economics and the inclination of If 50 percent of the animal manure on con- the individual farmer will determine whether fined animal operations in the United States is such options are adopted. converted to electricity and heat, about 7 billion kWh of electricity per year (equivalent to Care should be exercised when using these about 1,200 MW of electric generating capaci- data. They are based on a number of approxi- ty) and about 0.08 Quad/yr of heat would be mations and they cannot be taken too Iiterally. produced by 0.12 Quad/yr of biogas. At 70- per- They do, though, indicate the general trends as cent utilization, electricity equivalent to about to economic feasibility and they show that a 1,600 MW of electric generating capacity substantial quantity of the manure produced would be produced along with 0.11 Quad/yr of on livestock operations could be used econom- heat. ically to produce energy, if the effective capi- In either case some of the heat would be tal charges are reduced through various eco- wasted in the systems described above. There nomic incentives.

Environmental Impacts of Biogas Production: Anaerobic Digestion of Manure

Anaerobic digestion of feedlot manure is nure (which often represents a substantial dis- considered to be an environmentally benefi- posal problem) into a more benign sludge cial technology because it is an adaptation of waste. Where the manure was used as a fertil- a pollution control process. The energy prod- izer and soil amendment, the digestion wastes uct–biogas–is basically a byproduct of the can substitute for the manure while eliminat- control process, which converts the raw ma- ing some of its drawbacks. 196 ● Vol. n-Energy From Biological Processes

The environmental benefits associated with concentrations of inorganic salts, some con-

reducing feedlot pollution are extremely im- centrations of H2S and NH3, and variable portant. The runoff from cattle feedlots is a amounts of potentially toxic metals such as source of high concentrations of bacteria, sus- boron, copper, and iron. For feed lot operations pended and dissolved solids, and chemical and where the manure is collected only intermit- biological oxygen demand (COD/BOD). This tently, small concentrations of pesticides used type of runoff has been associated with: large for fly control may be contained in the manure and extensive fish kills because of oxygen and passed through to the waste stream. depletion of receiving waters; high nitrogen A variety of disposal options for the liquid concentrations in ground and surface waters, and sludge wastes exist. Generally, wastes will which can contribute to the aging of streams be ponded to allow settling to occur. The liq- and to nitrate poisoning of infants and live- uid, which is high in organic content, can be stock; transmission of infectious disease orga- pumped into tank trucks (or, for very large nisms (including salmonella, leptospirosis, and operations, piped directly to fields) to be used coliform and enterococci bacteria) to man, for irrigation and fertilization. The high salt livestock, and wildlife; and coloring of content and the small concentrations of met- streams. 8 Other problems associated with als in the fluid make it necessary to rotate land feedlots include attraction of flies and obnox- used for this type of disposal. Large operations ious odors. may conceivably treat the water and recycle it, Because anaerobic digestion is a relatively but treatment cost may prove to be prohibi- simple process not requiring extreme operating tive. Other disposal methods include evapora- conditions or exotic controls, biogas facilities tion (in arid climates), discharge into water- may be designed for very small (10 cow) oper- ways (although larger operations are likely to ations as well as large feedlots. The environ- be subject to zero discharge requirements by mental impacts will vary accordingly. For ex- the Environmental Protection Agency), and dis- ample, recycling of wastewater may be possi- charge into public sewage treatment plants. ble for the larger operations; it is not likely to In all cases, infiltration of wastewater into the be possible for the small onfarm digesters be- ground water system is a possibility where soils cause of high water treatment costs. The prod- are porous and unable to purify the effluent uct gas from the smaller units is likely to be through natural processes. As with virtually all used onsite and, depending on its use, may or disposal problems of this nature, this is a de- may not be scrubbed of its HIS and ammonia sign and enforcement problem rather than a (N H,) content; the product from very large technological one; if necessary, ponds can be units may be upgraded to pipeline quality by lined with clay or other substances for ground removing these pollutants as well as the 30 to water protection. 40 percent of the CO, fraction in the biogas. The organic content of the effluent, which The major problem associated with the di- varies according to the efficiency of the digest- gestion process is waste disposal and the asso- er, will represent a BOD problem if allowed to ciated water polIution impacts that could re- enter surface waters that cannot dilute the ef- sult. As noted above, anaerobic digestion is fluent sufficiently. Similar problems can occur basically a waste treatment technology, but al- with organics leached from manure storage though it reduces the organic pollution con- piles. However, this problem exists in more tent of manure it does not eliminate it. The severe form in the original feedlot operation. combination of liquid and solid effluent from The sludge product can be disposed of in a the digester contains organic solids, fairly high landfill, but it appears that the sludge has

1‘Environmental Impllcation$ ot [rends In Agriculture and S/lvi- culture, Volume I I En vlronmenta I t ffect~ of Trencfj (Wa shlngton, ‘vSolar Program A~se$$ment Environmental factors, Fue/s From D C Environmental Protection Agency, December 1978), E PA- Biomass (Washington, D C Energy Research and Development 600/ 1-78-102 Acimlnlstratlon, March 1977), ERDA 77-477 Ch. 9—Anaerobic Digestion ● 197

value either as a fertilizer or cattle feed if the of 0.1 lb/million Btu,24 compared to the Feder- heavy metals content is not too great. Success- al requirement of 1.2 lb/million Btu for coal ful experience with anaerobically digested mu- combustion in large utility boilers. nicipal sludges, which clearly have higher con- The major air pollution problem of anaer- centrations of heavy metals, indicates that use obic digestion, therefore, is not from combus- of the feedlot-derived sludge as fertilizer tion of the product gas, but from leaks of raw should present no metals problem. 20 I n numer- gas from the system. For a manure sulfur con- ous applications overseas, the sludge is consid- tent of 0.2 percent and digester pH of 7.2, the ered a substantial improvement over the previ- raw biogas can contain H S in concentrations ously used manure fertilizer. I n areas where 2 of nearly 2,000 parts per million (ppm).25 Al- chemical fertiIizers are not available or are too though exposure to this full concentration expensive, the retention of the manure’s fer- seems extremely unlikely, concentrations of tilizer value is a particularly critical benefit of 500 ppm can lead to unconsciousness and the biogas process. death within 30 minutes to 1 hour, and concen-

Although the H2S (and related compounds) trations of 100 ppm to respiratory problems of content of the effluent may present some odor gradually increasing severity over the course problems, this problem, as well as that of the of a few hours; the Occupational Safety and very small pesticide content, should be negligi- Health Administration’s standard is a max- ble.21 imum permissible exposure level of 20 ppm.26

The gas produced by the digester will con- Because of rapid diffusion of the gas, health 22 tain small (less than l-percent each) concen- problems associated with H2S exposures are trations of H2S and NH3. If the gas is burned likely to be confined to these occupational ex- onsite without scrubbing out these pollutants, posures. However, venting of raw gas can combustion will oxidize these contaminants to cause severe odor problems to the genera I sulfur and nitrogen oxides. Because the H2S public. In this case, odor problems associated will form mostly sulfurous and sulfuric acids, with gas venting should be compared to the which are extremely corrosive to metal, the similar (but more certain) odor problems asso- biogas has limited use if it is not scrubbed. For ciated with the sometimes haphazard treat- example, scrubbing is a requirement if the gas ment of manure that the biogas operation re- is to be used in an internal combustion engine. places. Simple and inexpensive scrubbing methods are Because methane is explosive when mixed available, using an “iron sponge” of ferric ox- with air, strong precautions must be taken to ide and wood shavings to react the gas with the 23 avoid biogas leakage into confined areas and iron to form ferric sulfide. However, even if to prevent any possibility of the gas coming the gas were not scrubbed, the pollutant con- into contact with sparks or flames. Although centrations caused by biogas combustion this will be a universal problem with biogas fa- should be of little consequence to public cilities, it is particularly worrisome if small health as long as the combustion did not take units proliferate. place in a confined area. For example, combustion of biogas produced from fresh cow ma- If normal operating conditions hold biogas nure might generate sulfur oxides on the order leakage to near-zero levels, the powerful odor

of the H2S contaminant would serve as an OM ‘ ethane Gerrerat/on From Human, An/ma/, and Agricultural early warning of a leak. Because low concen- Waste$ (Washington, D C National Research Council, National

Academy ot Sciences, 1977), Library ot Congress catalog No trations of H2S will deactivate the sense of 77-92794 smell, the acceptance of small leaks as “stand- ) ‘M C T Kuo ancl J L Jones, “ E nvlronmental and F nerg,y Out- put Ana]ysls for the Conver\lon of Agrlcult ura I Re\Idues to Met h- ane, ” Sympos/um on Energy From Bioma$$ and W a$te, I n~t Itute ot Gas Technology, Washington, D C , Aug 14-18, 1978 “lbld J ‘ So/ar Program A ~jej$men[, op ( It “$o/ar Program As\e$$ment, op c It ‘‘Kuo ancl jone~, op clt “)1 blci 198 ● Vol. n-Energy From Biological Processes

ard operating practice” would eliminate this monitored, and may be operated and main- safety factor. tained by untrained (and/or part-time) personnel. Some of the potential safety and health The institutional problems associated with problems probably will respond to improved assuring that there is adequate control of di- system designs if small onfarm systems be- gester impacts are very similar to those of eth- come popular and the size of the market justi- anol plants: there is an attraction towards fies increased design efforts on the part of the smaller size (“on farm”) plants which may have manufacturers. The ease of building home- some advantages (mainly ease of locating sites made systems, however, coupled with farmers’ for waste disposal and smaller scale local im- traditional independence should provide po- pacts) but which cannot afford sophisticated tent competition to the sale of manufactured waste treatment, are unlikely to be closely (and presumably safer) systems.

Research, Development, and Demonstration Needs

Below, the more important research, devel- digester systems. There are also numerous de- opment, and demonstration needs for anaero- sign alternatives that could lower the digester bic digestion are divided into the general areas costs, and these should be thoroughly exam- of microbiology, engineering, and agriculture. ined. Electric power generation and the related feedstock pumpin is a weak area in digester Microbiology g systems, particularly with respect to reliabiIity, The whole range of studies related to how maintenance, and efficiency of the engine-gen- anaerobic digestion works should be ad- erator units. Development work for small en- dressed. This includes identifying the bacteria gines intended to use biogas and the related and enzymes involved, studying the bacteria’s pumps could lead to improvements in these nutrient requirements (including trace ele- areas, and fuel cells capable of using biogas ments), identifying optimum conditions for the should be developed. Development work is various conversion stages, and investigating also needed into the best ways the farm gener- why some feedstocks are superior to others. ator can supply the electric utility grid during Much of this is in the realm of basic research periods of high demand without undue incon- needed to understand the processes involved venience for the Iivestock operation. so that the yields, rates, control, and flexibility of anaerobic digestion can be improved. Agriculture

Engineering More needs to be known about the difference between digested and undigested ma- A large number of different digester types nure. The digested manure should be investi- need to be demonstrated to aid in optimizing gated in order to determine its value as a fertil- the safety and reliability of digester systems izer, animal feed, and nutrient source for aqua- while reducing the cost for onfarm use or for tic plants. High-value uses for the digester ef- large-scale systems. The unique problems and fluent, proved through thorough testing, could opportunities of various types of animal opera- significantly improve the economics of anaer- tions should also be addressed by the various obic digestion. Chapter 10 USE OF ALCOHOL FUELS Chapter 10.– USE OF ALCOHOL FUELS

Page Page Introduction ...... 201 Gas Turbines...... 215 Spark Ignition Engines–Ef ects From Alcohols Air Pollution Effects of Alcohol Fuel Use in and Blends ...... 201 Gas Turbines...... 215 Gasohol ...... 203 Ethanol...... , ...... 206 TABLES Methanol-Gasoline B ends. . . . 207 Page Methanol ...... 208 63. Various Estimates of Refinery Energy Summary ...... 208 Diesel Engines...... 209 Savings From Use of Ethanol as Octane- Environmental impacts of Automotive Use of Boosting Additive...... 205 Biomass Fuels...... 210 64. Emission Changes From Use of Alcohol Fuels and Blends...... 211 Air Pollution— Spark Ignition Engines...... 210 Air Pollution—Diesel Engines . . 213 Occupational Exposures to Fuels and FIGURE Emissions . . . . . 213 Safety Hazards...... 214 Page Environmental Effects of Spills ...... 214 37 Efficiency, Power, and Emissions as a Hazards to the Public...... 214 Function of Equivalence Ratio...... 202 Chapter 10 USE OF ALCOHOL FUELS

Introduction

Liquid fuels have some unique advantages thermore, methanol requires more heat to va- over sol ids and gases that make them impor- porize it. Both alcohols, as contrasted to gas- tant fuels in some applications. They contain a oline, contain oxygen and conduct electricity. large amount of energy per unit volume (as These properties are important when consider- compared to gases) and their combustion can ing the use of alcohol fuels. easily be controlled (as compared to solids). However, there are substantial differences While oil and hydrocarbon (HC) crops may among Iiquid fuels. At one end of the spectrum some day produce fuels for transportation, is residual fuel oil, which can produce consid- their costs and yields are highly uncertain at erable emissions when burned and is generally this time. For the near to mid-term, the most best suited as a boiler fuel (an application also likely biomass substitutes for gasoline, diesel, open to solid fuels). At the other end are light and light fuel oil are the alcohol fuels. distillate oils, gasolines, and alcohol fuels. The Biomass is the only solar technology for pro- oils and gasolines are superior to alcohols with ducing liquid fuels. The biomass can be con- respect to their energy content per unit weight verted to methanol (“wood alcohol”) through (ethanol has two-thirds and methanol one-half thermochemical conversion or to ethanol of the energy per gallon of gasoline), which (“grain alcohol”) through fermentation or, makes them better suited for aviation. The al- possibly, thermochemical conversion. Ethanol cohols are superior to oils with respect to their lower particulate emissions and higher octane production from sugars and starches is currently commercial technology. Wood-to-methanol values. These properties make them particularly useful for marine and ground transporta- plants can be built with existing technology, although none currently exists, and plant-herb- tion where energy density is not critical and for gas turbines used for peak load electric genera- age-to-methanol technology needs to be dem- tion, the applications considered here. onstrated.

While both ethanol and methanol are alco- Most cost calculations indicate that metha- hols, they have different physical and chemi- nol from coal will be less expensive than either cal properties. Of the two, methanol is less sol- alcohol from biomass. Until and unless a double in gasoline, separates easier, and can mestic liquid fuels surplus develops, however, more easily damage certain plastics, rubbers, this cost difference is not likely to exclude the and metals used in current automobiles. Fur- biomass alcohols from the market.

Spark Ignition Engines— Effects From Alcohols and Blends

Alcohols make excellent fuels for spark- proper fuel distribution among the cylinders, ignited engines which are designed for their and a fourth is cold-starting ability. use. However, when considering alcohol or alcohol-gasoline blend use in gasoline engines, Some materials in some automobiles are in- there are four specific factors that are of over- compatible with alcohols. Contact with alco- riding importance. One factor is the material hols can damage some gaskets, fuel pump dia- from which the engine is constructed. Another phragms, and other plastic and rubber parts. If is the ratio of air to fuel (A/F ratio) in the mix- these parts are adversely affected or fail, the ture that is burned in the engine. A third is engine is likely to malfunction. Furthermore,

201

67-968 0 - 80 - 14 202 . Vol. II—Energy From Biological Processes

some electric fuel pumps are mounted in the Figure 37.—Efficiency, Power, and Emissions as a fuel tank. Electric currents induced in the Function of Equivalence Ratio alcohol fuels by these motors may cause the protective terneplate coating on fuel tanks to dissolve and leave the tank susceptible to corrosion. There may also be a fire hazard associated with electrical shorting. Under certain circumstances, not totally understood at present, the alcohols or blends can also chemically remove the terneplate coating.123 FinalIy, alcohols can cause some deposits in the fuel tanks and lines to loosen and dislodge, leading to a blockage in the fuel filter or carburetor. Three major classes of automobiles are in use today: pre-1 975 cars, oxidation (two-way) catalyst cars (most post-1 975 cars in States other than California), and California three- way catalyst cars. The range of A/F ratios* intended for each class of cars is shown in figure 37 together with the effect of this ratio on the engine power, efficiency, and emissions. If the A/F ratio extends beyond the ranges of this figure, most engines will hesitate or stall. (Strati- fied-charge engines like the Honda CVCC and Ford Proco have somewhat wider ranges.) 1 1 0.85 0.90 0.95 1.0 1.05 Since the pre-1975 cars and oxidation cata- Stoichiometric Rich lyst cars usually have carburetors with fixed Equivalence ratio (+) fuel metering passageways, the alcohol blend fuels, which require less air per volume of fuel The equivalence ratio is the ratio of the A/F ratio which is exactly sufficient to completely burn all of the fuel to the actual A/F ratio than gasoline, will make the effective fuel mix- used in the car. Leaner mixtures have an excess of air, while richer ture leaner (i. e., move the effective A/F ratio to ones have an excess of fuel. less fuel and more air). California three-way SOURCE: H Adelman, et al., “End Use of Fluids From Biomass as Energy Re- sources in Both Transportation and Nontransportation Sectors, ” Uni- catalyst cars, however, have a sensor that ad- versity of Santa Clara, Santa Clara, Cal If., contractor report to OTA, 1979 justs the fuel delivery system to the A/F value intended by the manufacturer. Nevertheless, 4 exhausts from alcohol fuels “fool” this sensor plete. Consequently, while the difference be- somewhat, so the compensation is not com- tween the A/F ratio for alcohol fuels and pure gasoline is less for California three-way cata- ‘K R Stamper, “50,000 Mile Methanol/Gasollne Blend Fleet lyst cars than for other cars, the A/F ratio is still Study– A Progress Report, ” in Proceeding of the Alcohol Fuels somewhat leaner with alcohol fuels compared Technology Third /nterr-tationa/ Syrrrposlum, Asllomar, Callf (Bar- tlesvllle, Okla Bartlesvllle Energy Technology Center, Depart- to gasoline. ment of Energy, May 1979) ‘S. Gratch, Director of Chemical Science Lab, Ford Motor Co , For pure alcohols the fuel metering rate Dearborn, Mlch , private communlcatlon, 1979 must be increased significantly relative to gas- ‘j L Keller, G M Nakaguckl, and j C Ware, “Methanol Fuel 01 inc. This increased rate can result in stream- Modlflcatlon for Highway Vehicle Use, ” Union 011 Co of California, Brea, Callf , final report to the Department of Energy, ing flow rather than well-disbursed droplets as contract No FY 76-04-3683, j uly 1978 the fuel enters the air stream when carburetors ● The ftgure shows the equivalence ratio which is found by di- are retrofitted for alcohol fuel. This change viding the stolchlometnc A/F rat to which IS exactly sufficient to completely burn all of the fuel by the actual AIF ratio used In the car Leaner mixtures are to the left and richer to the right ‘Gratch, private communlcatlon, op clt Ch. 10—Use of Alcohol Fuels ● 203

can seriously aggravate the variation in the A/F Gasohol does not appear to significantly af- ratio among the cylinders which in turn can re- fect engine wear as compared to gasoline but duce performance and economy and increase more experience with gasohol is needed before emissions. Proper design of the fuel delivery a definitive statement can be made. However, system and intake manifold can avoid this an unknown fraction of existing automobiles penalty. have specific components that are not compatible with gasohol, which can result in some The alcohols do not inherently provide good fuel system failures. cold engine starting. Below 400 F, special attention must be paid to avoid cold-starting As older cars are replaced with newer ones problems with alcohol. Aids such as electric warranted for gasohol use and as experience heating in the intake manifold, blending develops in handling ethanol-gasoline blends, agents such as gasoline, butane, or pentane these problems should gradually disappear. added to the alcohols, or auxiliary cold-start fuel are providing solutions to this problem in Thermal Efficiency the alcohol vehicle fleets now in operation. The leaning effect of gasohol relative to gasoline wiII affect the three classes of cars some- Gasohol what differently. For precatalyst and oxidation catalyst cars, the thermal efficiency can either Materials Compatibility increase or decrease with gasohol depending on the original A/F setting (see figure 37). In Gasohol is a mixture of 10 percent ethanol general, automobiles that operate rich will in- and 90 percent unleaded gasoline. The ethanol crease in efficiency, while those that operate blended in gasohol must be dry (anhydrous) or lean will decrease in thermal efficiency with the blend will separate into two phases or lay- gasohol. ers under certain conditions. Typically, gasohol can hold more water than gasoline, but it The Nebraska “two million mile” test can contain no more than about 0.3 percent if showed a large average mileage increase (7 separation is to be avoided down to –400 F. percent) with gasohol. ’ However, the spread of Various additives have been tried to improve data points is so large that the uncertainty in the water tolerance of gasohol, but none have this difference is greater than the difference proved satisfactory to dates itself. * This is a generic problem in trying to deduce small differences in mileage with road Although the use of anhydrous ethanol tests. should minimize water tolerance problems with gasohol, phase separation has been ob- Laboratory tests, however, indicate an in- served to occur in four service stations in crease in thermal efficiency of 1 to 2 percent Iowa. ’ This phase-separated blend was sold to some customers and caused their vehicles to stall. Both the vehicle tanks and the service ‘W A Scheller, Nebraska 2 Million Mile Gasohol Road Test station tanks were drained and up to 0.3 per- Program, Sixth Progress Report (Lincoln, Nebr Unlversit~ of Ne- cent by volume water was found in the mix- braska, January 1977) *Taking data from figure 4 of Scheller, OTA has analyzed the ture, but the origin of the water contamination uncertainty in the average mileage difference Using a standard is not known. statistical test (“t” test) reveals that the spread In data points (standard deviation) IS so large that the mileage difference between gasohol and regular unleaded would have to be more than 30 percent (two times the standard deviation) before OTA would consider that the test had demonstrated a difference In mileage

5 H Adelman, et al , “End Use of Flu Ids From Biomass as Ener- While more sophisticated statistical tests might Indicate that the gy Resources In Both Transportation and Non-Transportation measured difference in mileage is meaningful, the validity of Sectors, ” University of Santa Clara, Santa Clara, Callf , contrac- these statistical methods IS predicated on all the errors being tor report to OTA, 1979, strictly random, and the assumption of random errors IS suspect ‘Douglas Snyder, lowa Development Commission, private unless the number of vehicles in the test fleet is orders of magni- communication, 1979 tude larger than any tests conducted to date 204 ● Vol. II—Energy From Biological Processes

with gasohol in precatalyst and oxidation cata- drivability problems are expected with gaso- lyst cars, which is within the measurement hol, as long as the mixture does not go beyond errors. 8 The changes in thermal efficiency with the capability of the compensation mechanism three-way catalyst cars will be less and there- in these cars. fore negligible. In most cars there may also be some minor Since gasohol contains 3.5 percent less ener- problems with vapor lock, if the vapor pressure gy per gallon than gasoline, precatalyst and OX- of the blend is not adjusted properly by remov- idation catalyst cars are expected to experi- ing some butanes from the gasoline. ence about a zero- to 4-percent decrease in miles per gallon. Three-way catalyst cars are Octane expected to experience about a 3- to 4-percent Addition of ethanol to gasoline increases the decrease in mileage. Probably neither of these octane* for the mixture over that of the gaso- decreases would be noticeable by motorists line. The exact increase will depend on the gas- and, as stated above, will depend on the A/F oline, of which there is a great variety. On the setting of the automobile for all but the three- average, for the range of 5 to 30 percent etha- way catalyst cars. nol, each percent of ethanol added to one base These conclusions are in complete agree- unleaded regular gasoline (88 octane) raised ment with the results of an American Petrole- the octane number by 0.3. However, the oc- um Institute study released in the spring of tane increases per unit alcohol are larger for 1980,9 in which all available data on gasohol lower percentages of ethanol and lower octane mileage were compared, averaged, and treated gasolines and level off at higher alcohol per- statistically to determine the significance of centages and gasoline octane. A 10-percent the results. blend would raise the octane by about 2 to 4 using an “average gasoline. ” Drivability The octane-boosting properties of ethanol Post-1970 noncatalyst and oxidation catalyst can be exploited in either of two ways to save cars that are set at lean A/F ratios on gasoline energy: 1 ) by reducing the oil refinery energy can experience drivability problems such as by producing a lower octane gasoline or 2) by stumbling, surging, hesitation, and stalling increasing the octane of all motor fuels so that with gasohol due to further mixture leaning. automobile manufacturers can increase the While no drivability problems have been recompression ratios and thus the efficiency of ported for precatalyst cars, laboratory tests on new cars. 1978 and 1979 oxidation catatlyst cars sug- There is considerable uncertainty and varia- gested slight deterioration in drivability .’” If bility in the amount of premium fuel that can the percentage of ethanol is increased beyond be saved at refineries by using ethanol as an 10 percent, more and more cars are expected octane-boosting additive. As shown in table 63, to experience drivability problems due to the reported or derived values vary from nearly leaning effect. zero to more than 60,000 Btu/gal of ethanol, Since three-way catalyst cars largely com- depending on the average octane of the refin- pensate for the leaning effect of gasohol, no ery gasoline pool, the octane boost assumed from ethanol, the type of gasoline and the ‘R K Pefley, et al., “Characterization and Research investiga- ratio of gasoline to middle distillates produced tion of Methanol and Methyl Fuels, ” University of Santa Clara, Santa Clara, Calif , contractor report to the Department of by the refinery, the refinery technology used, Energy, contract No FY 76-5-02-1258, 1979 and other specifics. ‘Mueller Associates, Inc , “A Comparative Assessment of Cur- rent Gasohol Fuel Economy Data, ” commissioned by the Amer- ican Petroleum Institute, 1980 ‘“R Lawrence, “Gasohol Test Program, ” Technology Assess- *Octane here refers to the average of research octane and ment and Evaluation Branch, Environmental Protection Agency, motor octane. Ann Arbor, Mlch , December 1978 ‘ ‘Keller, Nakagucki, and Ware, op cit Ch. 10—Use of Alcohol Fuels ● 205

Table 63.–Various Estimates of Refinery Energy Savings From an 8-percent reduction in the gasoline-to-distil- Use of Ethanol as Octane-Boosting Additive Iate ratio from the ethanol, the refinery energy

103 Btu saved/gal of savings from using ethanol as an octane-boost- ethanol blended ing additive are about 40,000 to 45,000 Btu/gal Source 10% in gasoline Conditions of ethanol.15 16 This corresponds to about 0.3 Energy Research to 0.4 gal of gasoline equivalent per gallon of Advisory Boarda . . . . . 8 Unknown b ethanol. Kozinski ...... 16 86 pool octane, reduction in gasoline to distillate ratio, The refinery energy savings are nonlinear 3 octane number boost by ethanol with the pool octane and the greatest savings OTA C ...... 40-45 91 pool octane, reduction in occur with the first increment of ethanol used. gasoline to distillate ratio, 3 octane number boost Consequently, since the supply of ethanol will by ethanol likely be limited to less than a universal 10- Adelman d ., ...... 53 Pool of 91 and 96 research percent blend, 0.4 gal of gasoline equivalent octane, 5 research octane boost by ethanol per gallon of ethanol is used in the calcula- Office of Alcohol Fuelse 63 Unknown tions. aE Research Adwsory Board. nergy “Gasohol. ” Oeparfment of Energy, Apr 29, 1980 If the energy savings from ethanol represent bA A Kozlnskl, Amoco 011 CO , Naperwlle, HI private commumcatlon, 1980 from data In O K Lawrence, et al , ‘‘Automotwe Fuels–Refinery Energy and Economics, ’ Amoco 011 Co , SAE the major economic incentive to the refiner, techmcal paper scrles No 800225, 1980 COTA from data In Lawrence, Op Clt , and from figure 5 In G W Mlchalskt and G H Unzelman ! then refineries with the highest potential for ‘‘Effeclwe Use of Armknocks During the 1980’ s.’ American Petroleum Institute preprint No energy savings would be the most likely to use 22.79, from 44th Refinery Midyear Meeting, May 16, 1979 ‘H Adelman, et al , “End Use of FluIds From Biomass as Energy Resources m Both Transporfa- it and savings would be maximized. Some re- tlon and Nontransporfatlon Sectors, ’ Unwerslty of Santa Clara, Santa Clara, Cahf contractor report to OTA, 1979 fineries, however, may have additional incen- eofflce Of AicOhOl Fuels, Department of Energy, “Commenls by the DOE Off Ice of Alcohol Fuels on tives for using ethanol, including capital sav- the Energy Research Advisory Board, Aprd 29, 1980, Gasohol Study Group, Repod. June 3, 1980 ings and greater gasoline yields from reduced SOURCE Off Ice of Technology Assessment reforming requirements, and access to stronger markets with gasohol. These incentives may Based on published computer simulations of not coincide with maximum energy savings. an oil refinery,12 Kozinski has estimated a sav- Moreover, the widespread use of technically ings of about 16,000 Btu/gal of ethanol on the advanced refining methods could reduce the basis of an average gasoline pool octane of 86, potential for energy savings through ethanol and appropriate reduction in the gasoline-to- use. Clearly, there are numerous factors which distillate ratio, which is appropriate for the can lower the actual savings below that which current situation where the octane of about is technically possible for the refineries mod- half the gasoline is raised with tetraethyl eled previously.l718 Consequently, although lead. ” In the future, if most of the gasoline 0.4 gal of gasoline equivalent per gallon of eth- produced is unleaded, then the pool will have anol is used as the refinery energy savings, it to increase to at least 89 octane, which is the should be viewed as a potential savings, which current average after lead has been added. probably will not be achieved in practice for Moreover, the octane requirements of new all cases. However, too many assumptions cars is increasing, which will induce refiners about future refinery operations are required to increase the pool octane further. in calculating the energy savings to be able to determine a single, correct value; and the ac- Assuming an average pool octane require- tual savings achieved will be very site specific. ment of 91, which can be reduced to 88 by the addition of 10 percent ethanol, and assuming

“D K Lawrence, et al , “Automotive Fuels– Refinery Energy “Based on Lawrence, et al , op clt and Economic s,” Amoco 011 Co , SAE technical Series NO “Based on figure 5 of G W Mlchalskl and C H Unzelman, 800225, 1980 “Effective Use of Antiknocks During the 1980’ s,” American Pe- 11A A Kozinskl, Amoco 011 Co , Napervllle, Ill , private com- troleum Institute preprint No 22-79, from 44th Refinery Midyear munication, 1980 Meeting, May 16,1979 “Bob Tlppee, “U S Refiners Adjusting to Changing Require- “Lawrence, et al , op clt merits, ” 0// and Gas /ourna/, june 23, 1980 18 Mlchalskl and Unzelman, op clt 206 ● Vol. /l—Energy From Biological Processes

The other possibility is to use the ethanol na- If the gasoline retailer blends the gasohol, tionwide to gradually increase the octane of the value of the ethanol is somewhat different. motor fuels. Auto manufacturers could use the Gasoline retailers bought regular unleaded increased octane to improve engine efficien- gasoline for about $0.70/gal in July 1979 and cies by increasing the compression ratios in sold gasohol for a rough average of $().()3/gal automobile engines. The energy savings per more than regular unleaded. (The difference gallon of ethanol would be comparable to that between this and the retail price of gasoline is calculated above, but there would be Iittle sav- due to taxes and service station markup, which ings before higher octane fuels were readily total about $0.29/gal.) One-tenth gallon of eth- available and older automobiles had been re- anol displaces $0.07 worth of gasoline and the placed with the newer, more efficient engines. mixture sold for $0.03/gal more. Therefore, 0.1 gal of ethanol was valued at $0.10 or $1 .00/ Some stratified-charge engines (e.g., Ford gal. This is 2.5 times the July 1979 average Proco) do not require high-octane fuels and the crude oil price of $0.40/gal. benefits from a high-octane fuel would be substantially less than for conventional engines. If Both of these estimates are approximate, these or other such engines are used extensive- and changing price relations between crude oil ly, the octane-boosting properties of ethanol and gasoline couId affect them. Moreover, sev- would be of little value, but it is likely that eral other factors can change the estimated large numbers of automobiles will need high- value of ethanol. If a special, low-octane, low octane fuels well into the 1990’s. vapor pressure gasoline is sold for blending with ethanol, at low sales volumes the whole- Value of Ethanol in Gasohol saler might assign a larger overhead charge per gallon sold. Also, the refinery removes relative- The value of ethanol or the price at which it ly inexpensive gasoline components in order to becomes competitive as a fuel additive can be lower the vapor pressure* of the gasoline, and calculated in several ways. Two alternatives this increases its cost. On the other hand, in are presented here. areas where gasohol is popular, the large sales At the oil refinery, each gallon of ethanol volumes lower service station overhead per used as an octane booster saves the refinery gallon thus raising ethanol’s value. These fac- the equivalent of 0.4 gal of gasoline by allow- tors can change the value of ethanol by as ing the production of a lower octane gasoline much as $0.40/gal in either direction (i. e., (see section on octane above). in addition, 1 $0.04/gal of gasohol) and the pricing policies of gal of ethanol will displace about 0.8 gal of oil refiners and distributors will, to a large ex- gasoline when used as a gasohol blend (i.e., tent, determine whether ethanol is economi- gasohol mileage is assumed to be 2 percent calIy attractive as an octane-boosting additive. less than gasoline mileage). At the refinery gate, unleaded regular costs about 1.6 times the crude oil price. Assuming that the fuels Ethanol saved by the octane boost, which are of lower If pure ethanol is used, carburetors have to value than gasoline, cost about the same as be modified to accommodate this fuel. New crude oil, the ethanol is valued at about (gaso- engines designed for ethanol could have line saved) x (gasoline price) + (refinery fuel saved) x fuel price = 0.8 X 1.6 + 0.4 X 1.0 = *The more’ volatile components of gasoline (e.g., butanes) may 1.7 times the crude oil price. * be removed to decrease evaporative emissions and reduce the possibility of vapor lock Although these components can be used as fuel, removing them decreases the quantity of gasoline ‘This IS in agreement with the value of 16 to 1 8 times the and the octane boost achieved by the ethanol Consequently, the crude oil price that can be calculated using Bonner and advantages of having a less volatile gasoline must be weighed Moore’s” estimates based on $12 25/bbl crude 011 against the resultant decrease in the gasoline quantity and the “A Formu/a for Estimating Refinery Cost Changes Associated value of the ethanol This IS a matter of business economics in With Motor Fue/ Reformation (Houston, Tex Bonner and Moore each refinery and there are no simple rules which would be uni- Associates, Inc., Jan 13, 1978) versally applicable Ch. 10—Use of Alcohol Fuels ● 207

higher compression ratios (due to the higher The octane boost that can be achieved with octane of ethanol) and burn leaner fuel mix- methanol is comparable to that for ethanol, or tures which would improve engine efficiency; about 3 octane numbers for a 10-percent and laboratory tests indicate the improvement blend .23 24 could be 10 to 20 percent. 20 21 Furthermore, the For cars adjusted lean on gasoline, drivabili- increased compression ratio provides more ty problems will occur with a 10 percent meth- power, so engine sizes could also be reduced, anol blend due to additional leaning. However, thereby increasing the efficiency stilI further. at lower percentages of methanol, these prob- In cold climates there can be problems start- lems decrease. Indeed methanol is used as a ing and during warmup of engines fueled by de-icer at concentrations of about 0.5 percent, pure ethanol, due to its low volatility and high with no apparent impairment of drivability. 25 heat of vaporization. Consequently, special The principal problems with methanol equipment will be necessary to enable cold blends are the large increase in vapor pressure starting in vehicles fueled with straight etha- when methanol is added and the poor water nol. Alternatively, it may be possible to blend tolerance of the blends. The higher vapor pres- small quantities of light hydrocarbons in the sure can lead to increased evaporative emis- alcohol to alleviate the cold-start problem, or sions and possibly vapor lock. The decreased one couId use a combination of these strate- water tolerance can lead to fuel separation gies. (into layers), which can lead to poorer drivability in automobiles.

Methanol= Gasoline Blends The vapor pressure of the blend can be decreased by reducing the gasoline vapor pres- In general the effects of adding methanol to sure, but this significantly reduces the volume gasoline are similar to those for ethanol addi- of gasoline blending stock and can result in tion, but more extreme. Thus methanol sepa- less total automotive fuel.26 Newer cars, how- rates from gasoline at lower moisture levels ever, are fitted with charcoal to trap evapora- and damages alcohol-susceptible parts and 27 22 tive emissions from fuel tanks, but these some paints more quickly or extensively. filters may have to be replaced yearly in order Therefore, there would be some added cost to maintain their effectiveness.28 associated with using metals, plastic and rub- Evaporative emissions from carburetor boiloff increase ber parts, or paints that are tolerant to metha- with alcohol blends. However, charcoal air fil- nol over using those tolerant to ethanol, but ters are being used on some 1980 model vehi- the added cost is probably quite small. cles to trap the evaporative emissions from the As with ethanol, the change in thermal effi- carburetor and may reduce blend evaporative ciency for methanol-gasoline blends depends emissions. Moreover, fuel injection systems on the original gasoline A/F ratio, but would have less fuel losses. Vapor lock may also be a generally be in the range of zero to 4 percent problem in some cases,29 but studies indicate for a 10 percent methanol blend, leading to an estimated 1- to 5-percent drop in mileage (miles per gallon).

‘(’H Menrad, “Recent Progre$s In Automotive Alcohol Fuel Ap- ‘lAdelman, et al , op clt pl Icatlon, ” In Proceedings of the Fourth /nternationa/ Symposium “F W Cos, “The Physical Properties of Casollne/Alcohol Au- on Automotive Propulsion Systems, held on Apr 18-22, 1977, tomotive Fuels, ” In Proceedings of the Third /nternationa/ S ym- (NATO Committee on the Challenges of Modern Society, Febru- posium on A/coho/ Fue/ Technology, vo/, //, Asllomar, Callf May ary 1978) 28-31, 1979 ‘‘Wlnfrled Berhardt, “Posslbllltles for Co~t-Effective Use of Al- 2’Adelman, et al , op clt cohol Fuels In Otto E nglne-Powered Vehicles, ” In Proceed/rigs of “Keller, Nakaguckl, and Ware, op clt the international Symposium on Alcohol Fuel Technology, “lbld A4ethano/ and Ethano/, West Germany, Nov 21-23, 1973, engllsh ‘“K R Stamper, Bartlesvllle Energy Technology Center, De- translation by the Department of Energy partment of Energy, Bartlesvllle, Okla , 1979 c llKel jer, Nak aguck I, and Ware, oP ‘t “Keller, Nakaguckl, and Ware, op clt 208 ● Vol. n-Energy From Biological Processes

that proper vehicle design can also eliminate Another possible approach with methanol is this problem .30 to decompose the alcohol into carbon monoxide (CO) and hydrogen in the carburetor. This The water tolerance problem may require gas mixture is then used to fuel the engine. Ex- some sort of cosolvent, or additive, which haust heat from the engine is used to fuel the helps to retain methanol in gasoline. * One decomposition; and the CO-hydrogen mixture such cosolvent, t-butanol (another alcohol), is contains 20 percent more energy than the currently being test marketed in t-butanol- methanol from which it came. This offers the 32 methanol-gasoline blends by Sun Oil. The possibility of improving the engines thermal ef- energy cost of producing the t-butanol, how- ficiency by 20 percent, but it is too soon to ever, is not known. Alternatively, hexanol (still know whether this potential increase can be another alcohol) has been used successfully to achieved in practice. recombine the phases in a separated methanol- gasoline blend .33 Methanol can cause gasoline fuel injection pumps to fail, due to its low lubricity. ” Other Each of the problems with methanol blends tests indicate that methanol combustion cor- has numerous solutions, but it is unclear at rodes cast iron piston rings and may affect nor- present which will be the most effective at in- mal lubricating oils, particularly in very cold creasing motor fuel supplies at the least cost weather starting. ” However, in actual engine to consumers. Additional work is needed to and vehicle tests in warm weather conditions, clarify this matter. methanol has not been found to cause premature engine wear. 38 Methanol Below about 400 F, methanol-fueled engines can experience starting probIems, due to the In order to use methanol, carburetors suit- same factors that affect ethanol-fueled vehi- able for methanol have to be installed on the cles. As with ethanol, special equipment and/ engine or old ones modified. New engines de- or blending of volatile hydrocarbons in the signed for methanol could have higher com- fuel will be needed to enable cold starting. pression ratios and burn leaner fuel mixtures, leading to a potential 20-percent improvement in thermal efficiency .34 35 As with ethanol, Summary slightly smaller engines could be used because of the greater power associated with the higher Ethanol-gasoline blends are currently being compression ratio, which could provide still marked commercially, and methanol blends greater efficiency improvements. (with a cosolvent) are being test marketed. In addition, automobiles fueled with straight ethanol are being used in Brazil and extensive tests with methanol-fueled vehicles are under- ‘“A W. Crowley, et al , “Methanol-Gasoline Blends Perform- ance in Laboratory Tests and Vehicles, ” Inter Industry Emission way. Nevertheless, because the alcohols are Contro/ Progran?-2, Progress Report No. 1 (Society of Automotive not fully compatible with the existing liquid Engineers, 1975) fuels delivery system and automobile fleet, ● Nevertheless, one test of automobiles operating on a phase- separated blend showed fewer drivability problems than would some initial difficulties in using alcohol fuels have been expected ‘1 “Stamper, “50,000 Mile Methanol/Gasoline Blend Fleet Study,” OP. cit. “B C Davis and W H Douthut, “The Use of Alcohol Mix- “j C Ingamells and R H Llndqulst, “Methanol as a Motor tures as Gasoline Additwes, ” Suntech, Inc , Marcus Hook, Pa , Fuel or a Gasoline Blendlng Component, ” SAE paper No 750123, presented at 1980 National Petroleum Refiners Association An- Automotwe Englneerlng Congress and Exposltton, Detroit, Mich , nual Meeting, March 1980. February 1975 “Stamper, private communication, op cit “E, C, Owens, “Methanol Effects on Lubrication and Engine “W j Most and j P Longwell, “Single Cylinder Engine Eval- Wear,” in Proceedings of the International Symposium on AICO uation of Methanol-Improved Energy Economy and Reduced ho/ Fue/ Teclmo/ogy, Methane/ and Ethano/, Wolfsburg, West NOX,” SAE paper No 750119, February 1975 Germany, Nov. 21-23,1977. I%pef ley, et a I , oP Cit ]apef Icy, et al , oP Clt Ch. 10—Use of Alcohol Fuels ● 209

are to be expected. These problems should dis- cause of the energy saved by allowing refiners appear with time, however, as more experience to produce a lower octane gasoline. The situa- is gained at handling and using the alcohol tion with methanol is less clear because of the fuels and as older automobiles are replaced greater difficulties associated with methanol with vehicles designed for use with these fuels. blends and the possible need for cosolvents. The use of methanol both in blends and as a For ethanol the preferred use probably is as straight fuel is currently being pursued. an octane-boosting additive to gasoline be-

Diesel Engines

Alcohols have only limited volubility in since the engine will run normally without the diesel fuel, making diesel-alcohol blends im- alcohol present, but the higher power at full practical at present. * If “ignition acceler- power can cause additional engine wear if the ators”* * are dissolved in the alcohols to engine is not designed for this power. Alter- enable them to ignite in diesel engines, they natively, the diesel injection can be modified can be used as a replacement for diesel fuel, to allow less fuel to be injected, but it would but the fuel metering system would have to be have to be returned to its original state when modified to provide the fuII range of power for alcohol is not being used. Dual injection sys- which the engine was designed and some pro- tems also can be designed to run with or with- visions made for the decreased lubricity of the out alcohol, but the injection controls would alcohols. probably be more complicated.

Alternatively, the alcohols can be used in If the fuel systems are separate, alcohol con- dual fuel systems, i.e., where the alcohol and taining up to 20 percent water probably can be diesel fuel are kept in separate fuel tanks and used. However, the diesel engines must be separate fuel metering systems are used. The modified to accommodate the alcohols. The two main possibilities are fumigation and dual modifications for alcohol fumigation are rela- injection. In a fumigation system, the alcohol tively simple and can be performed for an esti- is passed through a carburetor or injected into mated $150 if, for example, a farmer does it the air intake stream and the alcohol-air mix- himself and uses mostly spare parts. ” Costs ture replaces the intake air. In dual injection, could range up to $500 to $1,500 if installed by each fuel is injected separately into the com- a mechanic using stainless steel fuel tanks and bustion chamber. all new parts.40 Cost estimates for modifying engines for dual injection are not available, Most diesel engines are speed governed, i.e., however, but would be more expensive. In ei- more or less diesel fuel is injected automatical- ther case the long-term effects, such as possi- ly to maintain a constant engine speed for a bly increased engine wear, are unknown at the given accelerator setting. If alcohol is fumi- present time. gated into the cylinder, the diesel injection decreases automatically when the engine is Fumigation systems will generally be limited not at full power to compensate for the addi- to 30 to 45 percent ethanol or about 20 percent tional power from the alcohol. At full power, methanol, because evaporation of the alcohol the alcohol will give the engine additional cools the combustion air and the cooling from power. Consequently, once a fumigation sys- higher concentrations is sufficient to prevent tem is installed, alcohol usage is optional, the diesel fuel from igniting. However, consid-

*E mulslons are possible, but still In the R&D phase ● ● Although alkyl nitrates have generally been used as Ignltlon accelerates, sunflower 5CW! 011 and other vegetable OIIS have “Pefley, et al , op cit btwn suggested for biomass-derived ethanol ‘(’lbld 210 ● Vol. II—Energy From Biological Processes

erably higher proportions of alcohol can be biguities exist for dual injection systems.44 used with dual injection systems.41 These differences in efficiency are due to differences in engine tuning and design. An accu- In fumigation systems, some tests have rate determination is not available at present, shown thermal efficiency increases of up to 30 but it is unlikely that there will be significant percent for certain combinations of alcohol differences in the thermal efficiencies of en- and diesel fuel.42 Other tests43 showed slight gines optimized for the respective fuels. increases (5 percent) in thermal efficiency when the engine is at two-thirds to full load, Considerable uncertainty exists about the while there are large decreases (25 percent) in thermal efficiencies that can be obtained in efficiency at one-third full load. Similar am- practice if, for example, tractors are converted to alcohol use. Assuming, however, that the thermal efficiency does not change, 1 gal of 41 F, F plschlnger and C Haven ith, “A New Way of Direct In- jectton of Methanol In a Diesel E nglne, ” in Proceedings of the ethanol would replace 0.61 gal of diesel fuel, Third International Symposium on Alcohol Fuels Technology, and 1 gal of methanol would replace 0.45 gal vo/. //, Asilomar, Callf , May 1979 of diesel fuel. “K Bro and P S Pederson, SAE paper No. 770794, September 1977 44~ Holmer, “Methanol as a Substitute Fuel in the Diesel En- 4’K D Barnes, D B Kittleson, and T E Murphy, SAE paper gine, “ In Proceed~ngs of the International Symposium on Alcohol No 750469, Automotive Engineering Congress and Exposltlon, Fue/ Techrto/ogy,’ Methane/ and .Ethano/, West Germany, Novem- Society of Automotwe Engineers, Detroit, Mlch , February 1975. ber 1977

Environmental Impacts of Automotive Use of Biomass Fuels

The use of alcohol fuels and gasoline-alco- ● In some important tests, methodological hol blends in automobiles will have a number problems may seriously weaken the de- of environmental impacts associated with rived conclusions. For example, the Envi- changes in automotive emissions as well as dif- ronmental Protection Agency’s (EPA) 1978 ferences in the toxicity and handling character- tests of “gasohol” (10 percent ethanol istics of the fuel alternatives. The potential blend) included some vehicles that either changes in automotive emissions have been operated too “fuel-rich” initially (four identified as the impact of major concern and vehicles) or exceeded the nitrogen’ oxide are treated in the greatest detail in this discus- (NOx) standard on indolene (two vehi- sion. cles).45 If these noncompliance vehicles are dropped from the test sample, the changes caused by using gasohol are less Air Pollution—Spark Ignition Engines than test-to-test variability in exhaust emissions for the same vehicle.46 Predictions of emissions changes can be ● Test results have generally been obtained based on a combination of theoretical consid- from laboratory engines or, in testing erations, laboratory tests, and field measure- alcohol blends, from relatively unmodi- ments. Unfortunately, the results of the emis- fied automobile engines. A strategy that sions tests that have been completed to date provided reliable and plentiful supplies of are varied and confusing. Difficulties with alcohol fuels would presumably be ac- using these results for predicting emissions 45 changes include: Characterization Report: Analyses of Gasohol Fleet Data to Characterize the Impact of Gasohol on Tailpipe and Evaporative Emissions (Washington, D C Technical Support Branch, Mobile . Tests are rarely comparable because of Source Enforcement Division, Environmental Protection Agency, different base fuels (gasolines), fuel mix- December 1978) “’Wiplore K juneja, et al , “A Treatise on Exhaust Emission tures, automobiles, state of “tune,” driv- Test Varlabillty,” Society of Automotive Engineers, VOI 86, paper ing cycle, etc. 770136, 1977 Ch. 10—Use of Alcohol Fuels ● 211

companied by design changes that would For out-of-tune automobiles, which usually take advantage of the different properties operate in a “fuel rich” mode, CO and HC may

of these fuels. Thus, extrapolations from be expected to decrease while NOX increases. current test data may be overly pessimis- Vehicles equipped with three-way catalysts tic, at least for the long run. have feedback-controlled systems that operate to maintain a predetermined value of @ and Aside from test results, emission changes thus should be less affected by the blend lean- can be explained in great part by the depending effect of the alcohol fuels. However, this ence of emissions on the operating conditions feedback system is usually overridden during of the engine. Emissions of CO, HC, NO , and X cold starting to deliver a fuel-rich mixture; dur- aldehydes are strongly influenced by the ing this time period, HC and CO are more like- “equivalence ratio @“ (stoichiometric A/F ra- ly to decrease and NO to increase with tio/actual A/F ratio) and the emission control X alcohol fuels. Also, catalysts with oxygen sen- system (none, oxidation catalyst, etc.). Figure sors can be fooled into adjusting to leaner 37 shows how CO, HC, NO , and aldehydes are X operation because the exhaust emissions from likely to vary with @. alcohol blends oxidize faster than gasoline- Both methanol (6.4:1 ) and ethanol (9:1) have based exhausts and drive down the oxygen lev- lower stoichiometric A/F ratios than gasoline el in the exhaust stream (giving the appearance (14. s: I). Thus, blends of either alcohol fuel re- of overly fuel-rich operation);47 this should also sult in lower equivalence ratios (“leaner” oper- tend to decrease HC and CO and increase NOX ation) if no changes are made in the fuel meter- emissions when alcohol blends are used in ve- ing devices. Examining figure 37, emissions hicles equipped with such catalysts. changes can be predicted qualitatively by ob- Table 64 provides a summary of the type of serving that adding alcohol pushes the equiva- emissions changes that may be expected by lence ratio to the left. For an automobile nor- combining knowledge of test results and the mally operating “lean,” CO may be expected to remain about the same, HC remain the

47 same or increase slightly, and NOX decrease. Cratch, op cit

Table 64.–Emission Changes (compared to gasoline) From Use of Alcohol Fuels and Blends

Pollutant/fuel Methanol Methanol/gasoline Ethanol Ethanol/gasoline ‘‘gasohol’ Hydrocarbon About the same or slightly higher, May go up or down in unmodified Not very much data, should be May go up or down in unmodified or unburned but much less photochemically vehicles, unchanged when @ about the same or higher but less vehicles, about the same when + fuels reactive, and virtual elimination remains constant. Composition reactive. Expected reduction in remains constant; composition of PNAs; can be catalytically changes, the, and PNAs go PNA may change, expected reduction controlled down. Can be controlled. Higher in PNAs. Evaporative emissions evaporative emissions up Carbon About the same, slightly less for Essentially unchanged if@ About the same, can be controlled ;Decrease in unmodified vehicles monoxide rich mixtures; can be catalytically remains constant, lower if primarily a function of * (i.e., leaning occurs), about the controlled; primarily a function - leaning is allowed to occur same when @ remains constant of @ Nitrogen 1/3 to 2/3 less at same A/F ratio, Mixed; decreases when is held Lower, but not as Iow as with Slight effect, small decrease when oxides can be lowered further by going constant, but may increase from methanol; can be controlled 4 is held constant, but may very lean; can be controlled fuel ‘‘leaning’ effect in unmodi- increase or decrease further from fied vehicles fuel ‘‘leaning’ effect in unmodified vehicles Oxygenated Much higher aldehydes, particu- Aldehydes increase somewhat, Much higher aldehydes, particu- Aldehydes increase, most signifi- compounds larly significant with precatalyst most significant in Iarly significant with precatalyst cant in precatalyst vehicles vehicles precatalyst vehicles vehicles Particulate Virtually none Little data Expected to be near zero Little data, no significant change expected Other No sulfur compounds, no HCN or Unknown No sulfur compounds No data ammonia

SOURCE Office of Technology Assessment 212 ● Vol. Il—Energy From Biological Processes

theoretical model discussed above. The most the most abundant aldehyde in automo- significant changes, and their environmental tive emissions–formaldehyde – is not de- implications, are: tectable with conventional HC measuring instrumentation. Aldehyde increases may ● Substantial reductions in reactive HC and somewhat negate the positive effects of NO exhaust emissions with 100 percent X reductions in emissions of HC as well as in (neat) methanol and, to a lesser extent, eth- NO emissions from alcohol use. The mag- anol.— Although HC emissions are ex- X nitude of the potential impacts, however, pected to remain approximately the same is not well understood. with alcohol fuel use at the same *,4849 ● Substantial. reductions in particulate emis- the reactivity of these emissions is much sions if neat alcohol fuels are used. — Use of lower than that of gasoline-based HC neat alcohol fuels may reduce particulate emissions. Reductions in reactive HC and emissions virtually to zero. This is particu- NO should reduce photochemical smog X larly significant when the fuel substituted formation, although predictions of the for is leaded gasoline; particulate emis- magnitude of these effects are difficult. sions from autos using leaded gasoline are ● Increase in aldehyde emissions with neat al- on the order of 0.6 g/mile on the Federal cohols and blends.— Use of pure alcohol emission test cycle, and most of the par- fuels yields several-fold increases in alde- ticles are toxic (mostly lead by weight, hydes,50 51 whereas blends increase alde- with polynuclear aromatic (PNA) com- hyde emissions to a lesser extent. Because pounds adsorbed on their surfaces) and in catalytic converters are effective in re- the inhalable size range (whereas particu- moving aldehydes, catalyst-equipped ve- late emissions from autos using unleaded hicles tend to have low aldehyde emis- gasoline are on the order of 0.2 g/mile on sions whether or not alcohol is used;52 53 the same test cycle, are about 90-percent the major problem lies with cars not controlIable with catalytic converters, equipped with catalysts. and are composed mainly of carbon par- Aldehydes cause eye and respiratory ir- ticles.55 ritations and are photochemically reac- ● Substantial reductions in PNA compounds tive. Despite this, aldehydes are not spe- with neat alcohols and blends. — PNA com- cifically regulated in automobiles, and pounds emitted in small quantities in au-

“aDavid L Hllden and Fred B Parks, “A Single Cyllnder Engine tomobile exhausts are toxic and carcino- Study of Methanol Fuel – Emphasis on Organic Emlsslons, ” Soci- genic. 56 Methanol and methanol blends ety of Automotive E nglneers paper No 760378, presented at the appear to provide substantial reductions Automotive Engineering Congress, Dearborn, Mich , Feb 23-27, in these emissions (methanol exhaust con- 1976 “Samuel O Lowry and R S Devoto, “Exhaust Emissions From tains only about 1 percent of the PNA a Single-Cyl Inder Engine Fueled With Gasoline, Methanol, and compounds observed in gasoline ex- and Technology, Ethanol, ” CornbustiorI Science VOI 12, Nos 4, 5, haust),57 which may be of some signifi- and 6, 1976, pp 177-82 “W Lee and W Geffers, “Engine Performance and Exhaust cance in reducing the cancer hazards of Emission Characteristics of Spark Ignited Engines Burning Meth- urban air pollution. Ethanol and ethanol anol and Methanol-Gasoline Mixtures, ” Volkswagen Research blends may be expected to provide similar and Development Division, Wolfsburg, West Germany, presented at AICLE meeting, Boston, Mass , September 1975 effects, but this has not yet been verified. “Comparative Automotive Eng\ne Operation When Fueled With Ethano/ and Methane/ (Washington, D C Alcohol Fuels “R E Sampson and G S Sprtnger, “Effects of Exhaust Gas Program, Alternative Fuels Utlllzatlon Program, Department of Temperature and Fuel Composltlon on Particulate Emlsslon Energy, May 1978) From Spark Ignited Engines, ” Environmental Science and Tech- “j R Altsup, “Experimental Results Using Methanol and nology, VOI 7, No 1, january 1973 Methanol/G asollne Blends as Automotive Engine Fuel, ” Bartles- 551 bid vllle Energy Technology Center, No B9RC/Rl-76/l 5, january “American Petroleum Institute, “API Toxlcologlcal Review 1977 Gasoline, ” 1967 “j R Allsup and D El Eccleston, “Ethanol/Gasollne Blends as “On the Trai/ of New Fue/s – Alternative Fue/s for Motor Vehi- Automotive Fuels, ” Bartlesvllle Energy Technology Center, draft c/es (Bonn, West Germany Federal Ministry for Research and No 4 Technology, 1974), translated by Addls Translation International Ch. 10—Use of Alcohol Fuels ● 213

Air Pollution—Diesel Engines ficulty, ignition accelerating agents containing nitrates can be added to the fuel to provide an Very little data is available to allow the ignition source for the alcohol. There appears prediction of emission changes from the use of to be some potential for the formation of hy- alcohol fuels and blends in diesel engines. drogen cyanide or ammonia in the combustion Predictions of some limited reliability may be process when these additives are used. Labora- made from the small number of tests, extrap- tory testing will be necessary to verify the ex- olation from spark ignition tests, and knowl- istence of this effect. edge of diesel characteristics. Emission characteristics of mixed fuel oper- The major environmental reason why alco- ation with alcohol and diesel fuels depend on hol fuels appear to be attractive for diesels is the method of introducing the alcohol into the their ability to burn without producing par- combustion chamber. ticulate emissions. Domestic manufacturers are having problems meeting the proposed When the alcohol is mixed with the intake EPA particulate standard of 0.6 g/mile. Par- air (fumigation), the following changes have 61 ticulate emissions from diesel engines are 50 to been observed to occur: 58 100 times those from gasoline engines and ● increase in HC emissions and aldehydes, may contain more PNA; particulate reductions . little change in CO, thus appear to be especially attractive environ- . increase or decrease in NOx, and mentally. ● decrease in particulate. HC emissions from diesels are more photo- The emissions effects of other fuel systems chemicalIy reactive than automobile HC emis- are poorly understood. sions. Although a switch to alcohol fuels by itself will have an uncertain effect on uncon- trolled emissions, the elimination of particu- Occupational Exposures to Fuels late emissions may allow the use of oxidation and Emissions catalysts to improve HC control (because particulates otherwise would plug up the cata- In general, the effects of gasoline and gas- lyst). 59 oline-based emissions are more acute, in an occupational setting, than those of methanol If alcohol fuels behave in diesels in a man- and ethanol. For example: ner similar to their behavior in spark ignition . Short-term exposure to gasoline is consid- engines, they should cause NOX emissions to decrease and aldehydes to increase. CO levels ered more poisonous, tissue disruptive, have been observed in tests to double their and irritative than methanol when effects originally low values when shifting from diesel of eye contact, inhalation, skin penetra- fuel to alcohol;60 however, this is thought to be tion, skin irritation, or ingestion are con- 62 a correctable problem with the fuel injection sidered. Effects of the more severe (in- systems. gestion) exposures to methanol are gener- alIy reversible, although in some extreme Alcohol fuels have poor ignition capabilities cases there can be irreversible effects on when injected into the compressed and heated the central nervous system, optic nerve air in a diesel engine. To counteract this dif- end, and heart. 63

“K j Springer and T M Ra Ines, “E mlsslons From Diesel Ver- sions of Prociuct Ion Passenger Cars, $oclery of A utornotlve ErtgI- “ B S Murthy and L G Pless, “Et fectlvenes$ of Fuel Cetane neer$ Tran/at/ens, VOI 86, 5PC 4, paper No 770818, 1977 Number for Combustion Control In BI-FLJel Diesel E ng,lne, ” /our- “M Amano, et al , “Approaches to Low Emission Levels for nal of the Institution 01 Engineers [India), VOI 45, No 7, pt ME 4, Light-Duty Diesel Vehicles, ” Soc[ety of Automotive Engineers March 1965 paper No 760211, February 1976 “N V Steer, ed , Handbook of Laboratory Sa/etv (Cleveland, ‘“W F Marshall, Experlrnents With Nove/ FLIP/5 for D/ese/ En- Ohio Chemical Rubber Co , 1971) gines ( Bartlesvll Ie, ok la 13artlesvll Ie Energy Technology Center, ‘‘M N Gleason, et al , C/frt)ca/ Toxicology of Cornrnercla/ PoI- Department of Energy, February 1978), II ERCITPR-7718 Jor15, W i I I lams and W I I I la ms 214 . Vol. n-Energy From Biological Processes

● The effects of both acute and chronic ex- the blends have higher evaporation rates than posure to ethanol are considered to be the pure gasoline. Diesel fuels and diesel much less disruptive than methanol and, alcohol emulsions are considered to be safer therefore, gasoline. than gasoline or alcohol fuels. 66 ● The automotive exhaust emissions that are most dangerous in an enclosed space — such as a garage without adequate ven- Environmental Effects of Spills tilation —are CO from gasoline and CO To the extent that domestic alcohol produc- and formaldehyde from methanol. If a tion substitutes for significant quantities of im- fleet of methanol-powered cars is com- ported oil, a reduction in fuel transportation pared directly to a fleet of gasoline- and a consequent reduction in spills can be ex- powered cars, CO will be the most danger- pected. If alcohol is shipped by coastal tanker, ous pollutant (and equally dangerous for the possibility of large alcohol spills is a both fleets, because methanol should not realistic one, and the effects of such spills substantially change CO emissions)—so should be compared to the effects of oilspilIs. long as three-way catalysts are used. With- out catalysts, formaldehyde emissions Alcohol fuels appear to be less toxic than oil could be more dangerous than CO in the in the initial acute phase of the spill and have methanol-powered fleet. fewer long-term effects. Except in areas where It is interesting to observe that, for the alcohol concentrations reach or exceed 1 per- catalyst-equipped methanol fleet, formal- cent, the immediate effects of a spill should be dehyde will act as a “tracer” for CO; if eye minimal. For example, a concentration of and respiratory irritation from formalde- about 1 percent methanol in seawater is toler- hyde becomes acute, this will be an al- ated by many common components of inter- most sure sign that CO is at dangerous lev- tidal, mud-flat, and estuarine ecosystems un- els. less the alcohol is contaminated with heavy metals. b’ I n contrast, crude oil contains several Safety Hazards highly toxic water soluble components that The risks of fire and explosion appear to be can be damaging at low concentrations. Fur- lower with alcohol fuels than with gasoline, thermore, alcohols are extremely biodegrad- although evidence is mixed: able— toxic effects may be eliminated in hours —whereas the effects of heavy fuel oils can ● gasoline has a lower flash point and igni- last for years. tion temperature and is more flammable and explosive in open air than either ethanol or methanol, 64 Hazards to the Public ● alcohols are the greater hazard in closed The widespread distribution and use of alco- areas, 65 hol fuels will result in the public facing the ● higher electrical conductivity of alcohol same potential dangers as exist in the occupa- lessens danger of spark ignition, tional environment. A true assessment of etha- ● high volubility in water makes alcohol nol and methanol public health risks must in- fires easier to fight than gasoline fires, and corporate an analysis of probable exposure, ● alcohol fires are virtually invisible, adding however, and such an analysis is likely to show to their danger (but addition of trace ma- that both alcohol fuels may have considerably terials could overcome this drawback). Alcohol blends will be similar to gasoline “M E LePera, “Fine Safe Diesel Emulsions, ’’Conference orI Trarrsportation Synfue/s, sponsored by the Department of Energy, but they may be more ignitable in open spaces San Antonio, Tex , November 1978 and less ignitable in closed containers when ‘7P N D’E Iiscu, “Biological Effects of Methanol Spills Into Marine, Estuarine, and Freshwater Habitats, ” in Proceedings of “CRC Handbook of Laboratory Safety, op clt the International Symposium on Alcohol Fuel Technology, Meth- “lbd anol and Ethanol, Wolfsburg, West Germany, Nov 21-23, 1977 Ch. 10—Use of Alcohol Fuels ● 215

greater potential than gasoline to harm the fore these denaturants are added. The prob- public. For example, although methanol and ability of such diversion will be especially high gasoline are comparably toxic upon ingestion, if small, onfarm ethanol stilIs are widely used. methanol has a long history of improper ingestion and gasoline does not. Ethanol is even more Iikely to be improperly used, and the eth- A careful risk assessment of ethanol and anol used for motor fuel blending will be con- methanol fuels and blends could identify and taminated with dangerous toxic chemicals. Al- quantify these types of risks and would be in- though vile tasting and smelling denaturants valuable both in setting priorities for research may be added to fuel ethanol to discourage and in devising risk mitigation strategies that improper use, enterprising individuals are Iike- must accompany promotion of alcohol fuel Iy to try to filter out these additives. Also, fuel use. Such an assessment has not as yet been ethanol may be diverted to consumption be- conducted by DOE.

Gas Turbines

Alcohol fuels can be used readily in gas tur- ethanol would replace 0.67 gal of light oil and bine generators used to generate peakload 1 gal of methanol would replace 0.48 gal of electric power. The fuel metering system has light fuel oil. to be modified to meter the larger volumes of Currently about 0.25 Quad/yr of oil and 0.2 alcohols necessary to maintain the same pow- Quad/yr of natural gas are consumed for peak- er output and to accommodate the lower lu- Ioad electric generation .70 This represents bricity of alcohols relative to light fuel oil. about 6 percent of the electricity generated in These modifications are minor. In some cases, the United States. Use of alcohol fuels here however, the alcohols may attack the turbine would save about 0.2 trillion ft3 of natural gas blades or other metal parts and the modifica- per year and about 130,000 bbl/d of light distil- tions needed to use alcohol fuels would be late oil. considerably more expensive. If alcohol fuels are used, care must be taken to ensure that no salts are dissolved in the Air Pollution Effects of Alcohol Fuel alcohols by, for example, contamination with Use in Gas Turbines seawater during barge transport. The salts could greatly reduce the Iife of the turbines. Although alcohol fuels have been tested in gas turbines and the resulting emissions levels The thermal efficiency of a gas turbine is de- have been measured, there is some doubt as to termined by the ratio of the pressures at the whether those levels represent true indicators turbine inlet and outlet. This ratio is limited by of emissions to be expected from an optimized the combustion temperature. The alcohols system. For example, methanol use in an auto- have slightly lower combustion temperatures motive gas turbine produced a tenfold in- and should allow higher efficiencies than with 71 crease in HC emissions in one test, but it is light fuel oil in redesigned turbines. In unmod- quite possible that more optimal design of the ified turbines, the thermal efficiency of the fuel injection nozzles could lower these values alcohol fuels is about the same (within ± 2 per- considerable y. cent) as for light fuel oils. 6869 Thus, 1 gal of

““l W Huellmantt’lf 5 C Te(i(jle{ dncl K) C Hammond, J r , ( ombujt Ion ot Methanol tn an Automotive (;a5 Turbine, ” In Fu- (UV? A u(orno~IL e f lie/\. j IM (; OILICC I and N E Gal Idpoulou$, t% 7“Adelman, et al , op clt (NPW York Plenum f)rt~$$, 1977) “C W Ldpolnte and W L S{ hultz, “C’omparlson of E mlfslon ““P M J arvis, ‘‘Meth~nol as Gas Turbine Fuel, ” presenteci at ln~lce$ Wlthln a Turbine Combustor Operated on Diesel FLIPI or the 1974 E ngl neerlng Fou nda t Ion Conference, Methane/ a \ an Methanol, ” Society of Automotive Engineer$ paper No 710669, ~1 /ternate Fue/, Henneker, N H , j u Iy 1974 June 1971 216 ● Vol. II—Energy From Biological Processes

The most significant emission change should distillate fuels. ’z Ethanol, which has a combus- be a substantial drop in NOx emissions, which tion temperature intermediate between metha- are typically quite high in gas turbines. Metha- nol and distillates, should achieve somewhat nol has achieved 76-percent reductions in NOX smalIer reductions. emissions in large turbines because it has a significantly lower combustion temperature than ‘2JdrVlS, op clt Chapter ENERGY BALANCES FOR ALCOHOL FUELS Chapter 11.-ENERGY BALANCES FOR ALCOHOL FUELS

Page Page Ethanol From Grains and Sugar Crops...... 219 66. Net Displacement of Premium Fuels From Methanol and Ethanol From Wood, Grasses, and Various Feedstocks and Two End Uses. . . . .221 crop Residues ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 222 67. Net Displacement of Premium Fuels With General Considerations ...... 223 Alcohol Production From Various Feedstocks and Two End Uses ...... 222 68. Net Displacement of Premium Fuel per TABLES Dry Ton of Wood for Various Uses...... 222 Page 65. Energy Balance of Gasohol From Corn: Oil and Natural Gas Used and Displaced...... 220

I I Chapter 11 ENERGY BALANCES FOR ALCOHOL FUELS

The energy objective of using alcohol fuels crop residues. Ethanol from grains and sugar from biomass is the displacement of foreign oil crops is considered first, including a compar- and gas with domestic synthetic fuels. The ef- ison of various feedstocks and end uses. Meth- fectiveness of a fuel alcohol program depends anol and ethanol from the other feedstocks are on the energy consumed in growing and har- then considered, and the use of these feed- vesting the feedstock and converting it into stocks directly as fuels is compared with the alcohol, the type of fuel used in the conversion production of alcohols from them. Finally, process, and the use of the alcohol. some general considerations about the energy balance of these fuels are given. The major sources of biomass alcohol fuels are grains, sugar crops, wood, grasses, and

Ethanol From Grains and Sugar Crops

Corn is currently a principal feedstock for nol. Even under the most favorable circum- ethanol production, but other grains and sugar stances, distillery energy consumption is sig- crops could also be used. The energy balance nificant. The distillery producing most of the for gasohol from corn is discussed in detail fuel ethanol used today reportedly consumes below, followed by a summary of the energy 0.25 gal of gasoline equivalent (0.24 in the balance for various possible feedstocks and form of natural gas) per gallon of ethanol pro- for use of the ethanol either as an octane duced. 1 This number, however, involves some booster or as a standalone fuel. arbitrary decisions about what energy inputs shouId be attributed to the faciIity’s food-proc- For each gallon of ethanol derived from essing operations. Total processing energy in- corn, farming and grain drying consume, on puts in this plant amount to about 0.55 gal of the average, the energy equivalent of 0.29 gal gasoline equivalent per gallon of ethanol (see of gasoline’ in the form of oil (for fuel and ch. 7). petrochemicals) and natural gas (for nitrogen fertilizers). (See ch. 3 in pt. l.) The exact Energy-efficient standalone fuel ethanol dis- amount will vary with farming practices (e. g., tilleries would consume the equivalent of irrigation) and yields. I n general, however, the about 0.45 gal of gasoline per gal Ion of ethanol farming energy input per gallon of ethanol pro- produced (see ch. 7). Because the energy con- duced will increase when the farmland is of sumption of distilleries is not Iikely to be insig- poorer quality (e. g., setaside acreage) and/or in nificant in relation to the alcohol produced in dryer or colder climates (i.e., most of the the foreseeable future, it is essential that dis- western half of the country, excluding Hawaii). tilleries use abundant or renewable domestic energy sources such as coal, biomass, and/or The type of fuel used in the distillation proc- solar heat or obtain their heat from sources ess is perhaps the most important factor in de- that would otherwise be wasted. Reliance on termining the displacement potential of etha- these fuels would reduce the total use of oil and gas at the distillery to insignificant levels. ● Some authors have Included the energy used to manufacture farming equipment and the materials from which they are made as part of the farm energy inputs However, for consistency one should aIso Include, as a credit, the energy used in manufacturing the goods that would have been exported to pay for import- ‘Archer Dantels Mtdland Co , Decatur, III , “Update of Domes- ing the 011 displaced by the ethanol Because of the uncertainty tic Crude C)ll Entitlements, Application for Petroleum Substi- in these factors, and the fact that they are relatively small, they tutes, ” ERA-O 1, submitted to the Department of Energy, May 17, are not incIuded in the energy balance caIcuIations 1979

219 220 ● Vol. I/—Energy From Biological Processes

The amount of petroleum displaced by etha- and 3) the same as (2) except that the octane of nol fuel also depends on the manner in which the gasoline is lowered to exactly compensate it is used. As a standalone fuel, each gallon of for ethanol’s octane-boosting properties. They ethanol displaces about 0.65 gal of gasoline result in net displacements of: 1 ) from zero to equivalent. As an additive in gasohol, each gal- one-third gal, 2) about one-half gal, and 3) lon of ethanol displaces about 0.8 (± 0.2) gal slightly less than 1 gal of gasoline and natural of gasoline. * (See ch. 10. ) If the oil refinery pro- gas equivalent per gallon of ethanol used. duces a lower grade of gasoline to take advan- In all, the total displacement of premium tage of the octane-boosting properties of etha- fuels (oil and natural gas) achieved per gallon nol, up to 0.4 gal of gasoline energy equivalent of ethanol can be nearly 1 gal of gasoline can be saved in refinery processing energy (see equivalent per gallon of ethanol if petroleum ch. 10) for each gallon of ethanol used. and natural gas are not used to fuel ethanol Additional energy savings are achieved by distilleries and 2) lower octane gasoline is used using the byproduct distillers’ grain as an in gasohol blends. Failure to take these steps, animal feed. To the extent that crop produc- however, can result in the fuel cycle consum- tion is displaced by this animal feed substitute, ing slightly more oil and natural gas than it the energy required to grow the feed crop is displaces leading to a net increase in oil and displaced. gas consumption with ethanol production and use. This is the situation that is alluded to in Table 65 summarizes the oil and natural gas most debates over gasohol’s energy balance, used and displaced for the entire gasohol fuel but it is a situation that can be avoided with cycle. The energy is expressed as gallons of appropriate legislation. gasoline energy equivalent for each gallon of ethanol produced and used in gasohol (i. e., 1.0 Nevertheless in the most favorable case in the table represents 117,000 Btu/gal of eth- (case 3) and with an energy-efficient distillery, anol, 0.5 represents 58,500 Btu/gal of ethanol, the ratio of total energy displaced to total etc.) The three cases presented correspond to: energy consumed is 1.5 ( ± 0.4), i.e., the energy 1) two ways to calculate the present situation, balance is positive (a ratio greater than 1). And 2) future production of ethanol from the less if the feedstocks are derived from more pro- productive land that can be brought into crop ductive farmland, or local conditions allow production and using coal as a distillery fuel, energy savings at the distillery, e.g., not having *The greater displacement results from the alcohol’s Ieaning to dry the distillers’ grain, then the balance is effect even more favorable. Alternatively, an energy

Table 65.–Energy Balance of Gasohol From Corn: Oil and Natural Gas Used(+) and Displaced ( - ) (in gallons of gasoline equivalent per gallon of ethanol produced and useda)

Set-aside and potential cropland Coal-fired distillery and Present lowering of gasoline Entire plant Ethanol only Coal-fired distillery octane Uncertainty C Farming...... 0.3b 0.3b 0.4c 0.4 ?0.15 e e Distillery ...... 0.55 0.24 0 O — Distillery byproduct ...... -0.09 d o -0.09d –0.09 d ±0.03 Automobile...... -0.8 -0.8 -0.8 -0.8 ±0.2 Oil refinery...... — — — -0.4 ± 0.2 Total...... -0.0 -0.3 -0.5 -0.9 *0.3 aLower heat content Of gasoline and ethanol taken to be 117,000 Btu/gal and 76,000 Btu/gal, resPectlve& bo 16 as nitrogen ferflhzer (from natural gas) and 0.13 tTI05tly as Petroleum Products cEQlmated “nce~alnly of tO 15, assumes 75% of the yield achievable on avera9e cropland dBa5~ on soy~an ‘Ultlvatlon and ~rushlng energy The byproduct of 1 gal of ethanol from corn displaces 12 lb of crushed soybeans, whtch requires 009 gal Of gasoline eqUIValenf tO produce private COM- murrlcatlon with R Thomas, Van Arsdall, National Council of Farmer Cooperates e 55,000 BIU Ot COaI per gallon of ethanol SOURCE Offtce of Technology Assessment Ch. 11-Energy Balances for Alcohol Fuels ● 221

credit could be taken for the crop residues, Table 66.–Net Displacement of Premium Fuels (oil and which would also improve the calculated bal- natural gas) From Various Feedstocks and Two End Uses (energy expressed as gallons of gasoline equivalent per gallon ance. This general approach to the energy bal- a of ethanol produced and used ) ance, however, does not consider the different values of Iiquid versus solid fuels. Ethanol used as an octane- boosting additive to Ethanol used as a The uncertainty factor in table 65 of ± 0.3 Feedstock gasolineb standalone fuelc gal of gasoline per gallon of ethanol is due Corn ...... 0.9 0.4 primarily to inherent differences in farming Grain sorghum . . . 0.7 0.1 Spring wheat ...... 1.0 0.5 practices and yields, errors in fuel efficiency Oats. ., 1.0 0.5 measurements, uncertainties in oil refinery Barley ... ., ., . 1,0 0.4 savings, and the magnifying effect on these Sugarcane 0.9 0.3 errors of the low (10 percent) ethanol content %ssut-mflg lower fl@ content of gasoline and ethanol to be 117000 Btu/gal and 76000 Btu/gal respectwely, crops grown on margmal cropland wlfh yields of 75 percent of average cropland of gasohol. These factors make more precise yields, distillers’ gram energy credts as In table 65 for all grams and no credit for sugars, dlst[l- estimates unlikely in the near term. Iers fueled with nonpremmm fuels, nahonal average energy Inputs S Barber et al The Po- tential of Producing Energy From Agriculture. contractor report 10 OTA bUncerfamty * O 3 Not only does the farming energy used for CUncertamfy t O 2 grain or sugar crop production vary consider- SOURCE Off Ice of Technology Assessment ably from State to State, but also the average energy usage displays some differences between the various feedstocks. A more signifi- grains, ethanol produced from grains and used cant difference arises, however, between use as a standalone fuel (e. g., onfarm as a diesel of the ethanol as an octane-boosting additive fuel substitute) may actually lead to an in- to gasoline and as a standalone fuel, e.g., in creased use of premium fuels, even if nonpre- diesel tractors or for grain drying. As an oc- mium fuels are used in the distilIery. Conse- tane-boosting additive, each gallon of ethanol quently, caution should be exercised if onfarm displaces up to 1.2 gal of gasoline energy ethanol production and use are encouraged as equivalent in the automobile and at the re- a means of reducing the U.S. dependence on finery (see table 65). * As a standalone fuel, imported fuels. however, the displacement at the end use is Furthermore, the agricultural system is so only 0.65 to 0.8 gal of gasoline energy equiv- complex and interconnected that it is virtually alent per gallon of ethanol.** impossible to ensure that large levels of grain Table 66 summarizes the net displacement production for standalone ethanol fuel would of premium fuels (oil and natural gas) for not lead to a net increase in premium fuel con- various feedstocks and the two end uses. In sumption. Two examples illustrate this point. If each case it was assumed that the feedstocks grain sorghum from Nebraska is used as an eth- would be grown on marginal cropland with anol feedstock (to produce a standalone fuel), yields that are 75 percent of those obtained on the net displacement of premium fuels per gal- average U.S. cropland. lon of ethanol is similar to the national average for corn. A secondary effect of this, how- The striking feature displayed in table 66 is ever, could be an increase in grain sorghum that use of ethanol as a standalone fuel is con- production on marginal cropland in Texas, and siderably less efficient in displacing premium the increased energy required to grow this fuels than use of it as an octane-boosting ad- grain sorghum could more than negate the fuel ditive. In some cases, e.g., with grain sorghum displaced by the Nebraska sorghum. Similarly, and in areas with poor yields of the other ethanol production from corn could raise corn *O 4 gal of gasoline equivalent IS due to the octam-boosting prices and lead to some shift from corn to properties of ethanol and () 15 gal 15 due to the leaning effect of grain sorghum as an animal feed. Depending the a Icohol on where the shifts occurred, U.S. premium ● * Used as a standalone fuel in spark-ignition engines, alcohol- fueled engines can have a 20 percent higher thermal efflclencv fuel consumption could either increase or de- than their gasollne-fuelwf counterparts (see ch 10) crease as a resuIt. 222 . Vol. It—Energy From Biological Processes

The crucial point is that the energy usage in In order to avoid this situation, care should be agriculture is an important consideration in taken to ensure that ethanol derived from determining the effectiveness of a fuel ethanol grains and sugar crops be used in the most en- program. Because of this, there can be situa- ergy-efficient manner possible, i.e., as an tions where energy from agriculture does not octane-boosting additive. result in a net displacement of premium fuels.

Methanol and Ethanol From Wood, Grasses, and Crop Residues

Methanol, like ethanol, can be used as an Table 67.–Net Displacement of Premium Fuels (oil and natural octane-boosting additive and the oil refinery gas) With Alcohol Production From Various Feedstocks and TWO End Uses (energy expressed as gallons of gasoline equivalent energy saved per gallon of methanol is roughly per gallon of alcohol produced and useda) equivalent to that of ethanol. The lower energy content (per gal Ion) of methanol, however, Used as an octane- leads to a smaller displacement of gasoline in boosting additive Used as a Feedstock Fuel to gasoline standalone fuel the automobile per gallon of alcohol (0.6 gal of C Wood ...... Methanol 0.9b 0.4 gasoline equivalent per gallon of methanol Grasses or crop b residues. . Methanol 0.8 0.3C versus 0.8 for ethanol; see ch. 10). On the other d d 0.6 hand, the energy used to grow, collect, and Wood...... Ethanol 1.1 Grasses or crop d transport wood and plant herbage for metha- residues. . . . . Ethanol 1.0 0.5d nol production is less than for ethanol feed- aA55ume5 I ) [ower healing values of 57,000, 76,000, and 117,000 i3w/9al !Or Methanol. @tha” stocks such as grain and sugars. There are, nol, and gasoline, respeclwely, 2) culhvatton (grasses) collection and transport (all Ieedslocks) energy of O 75 mlll!on Btu/dry Ion for wood and 2 mdhon Btu/dry ton for grasses and crop resi- however, considerable local variations and dues (mcludmg fertilizers for grasses and fertdlzer replacements needed when crop res!dues are where, for example, crop residues are col- collected), 3) methanol ytelds of 120 gal/ton for wood and 100 gal/ton for grasses and crop resl. dues (50V0 energy conversion efflclency), 4) ethanol yields are 100 gal/ton of Ieedstock fer- lected on lands with poor yields, the energy mented, but addlhonal feedstock amountmg to 25,000 Btu/gal of ethanol IS required for dlshllery energy over and above that obtained from burmng the byproduct hgnm (based on G H Emert consumed in collection could be comparable and R Katz@n, “Chemicals From Biomass by Improved Enzyme Technology, ” presented at the SyrrrposIurn orr tlorrrass as a Non-Foss// Fuel Source, ACWCST Joint Chemical Congress, Hono- to that needed to produce some grains and lulu, Hawall, Apr 1-6, t979), resulting m net yields of 86 and 84 gal/ton of feedstock for wood sugar crops. and grasses/crop residues, respechvely, 5) methanol and ethanol displace 1 0 and 1 2 gal, r@- spectlvely, of gasollne energy equivalent (per gallon of alcohol) at the refinery and m the automobile when used as octane-boosting addlwes 10 gasohne, 6) they replace O 48 and O 65 gal of gasollne equwalent (per gallon of alcohol) al the end use when used as standalone fuels bUncertamty *O 3 Table 67 presents a summary of the net dis- cUncertamty *O 1 placement per gallon of alcohol for the vari- dunce~alnly large, since future processes for producing ethanol from these f@edstocks are not fully defined (see footnote a) ous Iignocellulosic feedstocks and two end SOURCE Offlceof Technology Assessment uses. The net displacement per gal Ion of methanol is comparable to that obtained for etha- Table 68.-Net Displacement of Premium Fuel (oil and nol from grains and sugar crops, because the natural gas) per Dry Ton of Wood for Various Uses lower energy content of methanol (as com- Net displacement of premium fuel pared to ethanol) is largely compensated for (106 Btu/dry ton (% of feedstock by the lower energy required to obtain metha- Use of feedstock) energy content) ab nol feedstocks. Direct combustion (68% efficiency). 12 75 Air gasification and combustion of fuel ab gas (85% overall efficiency) . . . . . 15 95 Another aspect of the energy balance for the Methanol (used as octane-boosting 13C Iignocellulosic feedstocks is the net displace- additive)...... 80 Ethanol (used as octane-boosting C ment of premium fuels per ton of feedstock. In additive)...... 11 70 C table 68, direct combustion, airblown gasifica- Methanol (standalone fuel)...... 6 40 6C tion, and alcohol fuels production are com- Ethanol (standalone fuel) ...... 40 pared with wood as the feedstock. Similar re- aASSUrnln9 16 rllllllofl Btuld~ 10rl, 075 mllhon Btu/ton requwed for collechon and Iransporf bAssumlng It replaces 011 burned with 85% @fflcl@ncy sults can also be derived for crop residues and cBased on table 67 grasses. SOURCE Off Ice of Technology Assessment Ch. 11-Energy Balances for Alcohol Fuels ● 223

Care should be exercised when interpreting boosting additives can be nearly as efficient in table 68. The ethanol yields (per ton of wood) displacing premium fuels as the direct com- and the energy that will be required by wood- bustion or airblown gasification of wood. On to-ethanol distilleries are still highly uncertain. the other hand, if the alcohols are used as Nevertheless, this table does display the gener- standalone fuels, the premium fuels displace- al feature that alcohol fuels used as octane- ment is considerably smaller.

General Considerations

The results presented in tables 65 through 68 grain or sugar feedstocks are used for the pro- are based on OTA’s estimates of average val- duction of standalone fuel ethanol. ues for the energy consumed and displaced by Another important factor in the energy bal- the various feedstocks. These figures, however, ance is the end use of the alcohol fuel. As has cannot be taken too literally since local vari- been emphasized above, there is a significant ations and changing circumstances can influ- increase in the displacement of premium fuels ence the results. Two of the more important when the alcohol is used as an octane-boosting factors which influence the results–the ener- additive. In the 1980’s, however, there could be gy required to obtain the feedstock and the an increased use of automobile engines that end use of the fuel — are discussed below. do not require high-octane fuels and that have automatic carburetor adjustment to maintain The energy needed to grow, harvest, and the proper air to fuel ratio (see ch. 10). With transport the feedstocks varies considerably, these engines, the octane-boosting properties depending on a number of site-specific factors of the alcohols are essentially irrelevant. Con- such as quantity of available biomass per acre, sequently, if the automobile fleet is gradually terrain, soil productivity, plant type, harvest- converted in this way, there will be a gradual ing techniques, etc. Generally, however, fac- reduction in the fuel displacement per gallon tors that increase the energy requirements also of alcohol, until the energy balances derived increase the costs. For example, where the for standalone fuels pertain. The same conclu- quantity of collectable crop residues per acre sion would hold if oil refineries convert to is small both the energy used and the cost (per more energy-efficient processes for producing ton of residue) will be higher than the average. high-octane gasoline. The economics will therefore usually dictate that— locally, at least—the more energy-effi- Another consequence of these possible cient source of a given feedstock be used. changes would be to increase the importance of the energy required to obtain the feedstock. As the use of bioenergy increases, however, For example, if ethanol only displaces as much the tendency will be to move to less energy- premium fuel as indicated when used as a efficient sources of feedstocks, and large Gov- standalone fuel, then, as mentioned above, ernment incentives could lead to the use of cultivating and harvesting the grains or sugar bioenergy that actually increases domestic crops used as feedstocks may require more consumption of premium fuels. The danger of premium fuel than is displaced by the ethanol. this is minimal with wood, but somewhat great- The danger of this is considerably less for er for grasses and crop residues due to the grasses and crop residues and virtually nonex- larger amount of energy needed to grow and/or istent for forest wood used as feedstock for collect them. The danger is even greater when alcohol production. Chapter 12 CHEMICALS FROM BIOMASS Chapter 12.-CHEMICALS FROM BIOMASS

Page Page 227 74. 1974 Production of Plastics, Synthetic

~~btib•• ● ● ● ● ● ● ● ● .•• 227 Fibers, and Rubber, and Estimated Chemical Synthesis From Lignocellulose...... 229 Lignocellulose Raw Material Base Required 231 75. Product Results in Fast Pyrolysis of Biomass and Its Constituents...... 233 TABLES Page 69. Species With Long-Chain Fatty Acids in Seed Oil ...... 228 FIGURES 70. Species With Hydroxy and Keto Fatty Page Acids ... , ...: ...... 228 38. Synthesis Routes for Converting 71. Potential Sources of Epoxy Fatty Acids ... 228 Lignocellulose Into Select Chemical 72. Sources of Conjugated Unsaturates...... 228 Feedstocks ...... 230 73. Major Petrochemicals That Can Be 39. Chemical Synthesis Involving Thermal Synthesized From Lignocellulose...... 231 Processes and Microwaves...... 232 Chapter 12 CHEMICALS FROM BIOMASS

Introduction

Biomass is used as a source of several indus- involve consideration of various alternative trial chemicals, including dimethyl sulphoxide, synthesis routes in most cases. At present, how- rayons, vanillin, tall oil, paint solvents, tannins, ever, too little information is available about and specialty chemicals such as alkaloids and the relative merits of biomass-versus coal-de- essential oils. Biomass is also the source of fur- rived chemicals to expect widespread, new in- fural which is used to produce resins and adhe- dustrial commitments to biomass chemicals in sives and can be used in the production of ny- the near future. This uncertainty depends as ion. ’ Aside from paper production, biomass much on uncertainties surrounding the costs currently is the source of cellulose acetates and possibilities of coal syntheses as on those and nitrates and other cellulose derivatives (4 surrounding biomass chemicals. Continued re- billion lb annually). Other chemicals include search into both options is needed to resolve tall oil resin and fat acid, Iignosulfonate chem- the problem and it is likely (as has been the icals, Kraft Iignin, bark chemicals, various seed case in the past) that a mix of feedstocks will oils, and many more. Every petroleum-derived result. chemical currently being used could be pro- Biomass-derived chemicals can be divided duced from biomass and nonpetroleum miner- into two major areas: 1 ) those in which the als, but some (e. g., carbon disulphide) would plant has performed a major part of the syn- require rather circuitous synthesis routes. thesis and 2) those in which chemical industry In the future biomass-derived chemicals feedstocks are derived by chemical synthesis could play an increasing role in the petro- from the more abundant biomass resources chemical industries. The economic decisions such as wood, grasses, and crop residues. Some to use or not to use biomass will be based on examples of each type are given below. The an assessment of the overall process from possibilities are so enormous, only an incom- feedstock to end product and it will probably plete sampling can be given here. A thorough ‘ I S Goldstein, “Potential for Converting Wood Into Plastics, ” analysis of the options for chemicals from S( /ence, vol 189, p 847, 1975 biomass is beyond the scope of this study.

Chemicals Synthesized by Plants

Several plant species produce relatively have properly placed chemical groups which large quantities of chemicals that can be used are susceptible to chemical attack so that they to produce plastics, plasticizers, lubricants, can be readily converted to the needed indus- coating products (e. g., paints), and various trial chemicals. There is also the possibility of chemicals that can serve as intermediates in using biologicalIy derived chemicals to pro- the syntheses used for numerous industrial duce products (such as plastics) which would products. 2 be expected to have similar properties to the products currently produced. For example, ny- The biologically synthesized chemicals that lon could be made from acids and amines are most easily used in the chemical industries other than the six carbon acids and amines cur- are those that are either: 1 ) identical to existing rently used for nylon synthesis. feedstock or intermediate chemicals, or 2) ‘1 H Prlnc en, ‘ Potential Wealth Is New Crops. Research ancl llevf~l(jpm[~nt ,“ ( rop l?e~ource} (New York Academic Pres$, Some plant species producing various 1’977) classes of chemicals are shown in tables 69

227 228 ● Vol. II-Energy From Biological Processes

through 72. (Note that these lists are incom- Table 72.–Sources of Conjugated Unsaturates plete and used only to illustrate some possi- Common name Species Type of saturation bilities.) Included are the following types: Common valeriana...... Valeriana officinalis 40% 9,11,13 ● long-chain fatty acids which might be Potmarigold colendula . . . Calendula officinalis 55% 8,10,12 Spurvalerian centrathus. .Centranthus macrosiphon 65% 9,11,13 used for the production of polymers, lu- Snapweed ...... Impatiens edgeworthii 60% 9,11,13,15 bricants, and plasticizers; Blueeyed Capemarigold . Dimorphotheca sinuata 60% 10,12 ● hydroxy fatty acids which could displace ( + hydroxy) Myrtle Coriaria ...... Coriaria myrtifolia 65% 9,11 the imported castor oil currently used as a (+ hydroxy) supply of these fatty acids; SOURCE L H Prmcen, ‘‘Potential Wealth m New Crops” Research and Development, Crop l?e- ● epoxy fatty acids which may be useful in sources (New York Academic Press, Inc ), 1977 plastics and coating materials; and ● conjugated unsaturates potentially useful Table 69.–Species With Long-Chain Fatty Acids in Seed Oil as intermediates in the synthesis of various industrial products. (These can also be Component in obtained from structural modification of Common name Species triglyceride oil soybean and linseed fatty acids.3).

Crambe...... Crambe abyssinica 60% C22

Money plant . . . . Lunaria annua 40% C22, 20% C24 Other possible new sources of chemicals Meadowfoam. Limnanthes alba 95% C + C 22 20 and materials include natural rubber from gua- Selenia ...... Selenia grandis 58% C — 22 4 Leavenworthia alabamica 50 % C22 yule (Parthenium argentaturn), and possibly jo-

Marshallia...... Marshallia caespitosa 44% C22 joba (Sirnmondia chinensis) and paper pulp 5 SOURCE : L. H. Prlncen, ‘‘Potenllal Wealth in New Crops Research and Development, Crop Re- from kenaf (Hibiscus cannabinus). Some of sources (New York Academic Press, Inc. ), 1977 these plants have also received considerable attention because it may be possible to grow Table 70.–Species With Hydroxy and Keto Fatty Acids them on marginal lands or land where the irri- Component in gation water is insufficient to support conven- Common name Species triglyceride oil tional crops (see ch. 4). It is risky, however, to

Bladderpod . Lesquerella gracilis 14-OH-C20 (70%) extrapolate unambiguous conclusions about Consessi their economic viability from incomplete data Holarrhena. Holarrhena antidysenterica 9-OH-C, (70%) 8 on the cultivation. In many cases they would Bittercress ., Cardamine impatiens Dihydroxy C22 and C24 (23%) also compete with food production for the Thistle ...... Chamaepeuce afra Trihydroxy C18 (35%) available farmland. Nevertheless, continued Bladderpod . . . . Lesquerella densipila 12-OH-C 18 diene (50%) Blueeyed screening of plant species together with cul-

Capemarigold. . Dimorphotheca sinuata 9-OH-C 18 conj. diene tivation tests should provide numerous addi- (67%) tional options for the cultivation of crops Myrtle Coriaria. . . Coriaria myrtifolia 13-OH-C 18 conj. diene (65%) yielding chemicals for industrial use. Cuspedaria pterocarpa Keto acids (25%) Another type of chemical synthesis involves SOURCE L H Prmcen, ‘‘Polenhal Wealth m New Crops Research and Oevelopmenl, Crop /?e. the use of specific bacteria, molds, or yeasts to sources (New York Academic Press, Inc ), 1977 synthesize the desired chemicals or substance. Table 71 .–Potential Sources of Epoxy Fatty Acids Commercial production of alcohol beverages by fermentation is one example. Mutant bac- Epoxy acid teria designed to produce insulin or other Common name Species content,% Kinkaoil ironweed. . . . Vernonia anthelmintica 68-75% ‘W j De Jarlals, L E Cast, and j C Cowan, / Am, Oi/ Chem, Euphorbia Euphorbia Iagascae 60-70 SOC., VOI 50, P 18, 1973 75 Stokesia Stokesia Iaevis ‘K E Foster, “A Sociotechnlcal Survey of Eluyule Rubber Com- — ...... Cephalocroton pueschellii 67 mercla Iizat Ion, ” report to the National Science Foundation, — Erlangea tomentosa 50 Dlvlslon of Policy Research and Analysis, grant No PRA Hartleaf Christmasbush Alchornea cordifolia 50 (c ) 20 78-11632, April 1979 — Schlectendalia Iuzulaefolia 45 ‘M O f3agby, “Kenaf A Practical Fiber Resource, ” TAPP/ Press Report Non-Wood Plant Fiber Pulping Process Report, No SOURCE L H Prmcen, ‘‘Potenllal Wealth In New Crops Research and Development, Crop Re- sources (New York Academic Press, Inc ), 1977 8, p 175, Atlanta, Ca , 1977 Ch. 12—Chemicals From Biomass . 229

drugs is another. ’ 7 Furthermore, other basic Further study into the details of photosyn- biochemical processes such as the reduction thesis, the biochemistry of plants, and molecu- of nitrates to ammonia may also be used even- lar genetics could lead to the development of tually.8 other plants or micro-organisms that could syn- “Pe~r( e Wright, 1 Ime for Bug Valley, ” New Sc/ent/sL p 27, thesize specific, predetermined chemicals. The Julv 5, 1979 options seem enormous at this stage of devel- “’Where Genctlc Englneerlng Will Change Industry, ” f3usJness Wee&, p 160, Oct 22, 1979 opment, but considerable additional R&D is ‘P Candan, C Manzano, and M Losada, “Bloconverslon of needed before the full potential of this ap- Light Into Chemical Energy Through Reduction With Water of proach can be evaluated. Nitrate to Ammonia, ” Nature, VOI 262, p 715, 1976

Chemical Synthesis From Lignocellulose

The second major area of chemicals from These estimates were derived by Goldstein’ biomass involves using the abundant biomass using optimistic assumptions about the yields resources of wood, grasses, and crop residues of the sugars- and benzene-based chemicals (Iignocellulosic material) to synthesize large- from wood. Obtaining these sugars- and ben- volume chemical feedstocks, which are con- zene-based compounds from wood is currently verted in the chemical industry to a wide varie- the subject of considerable R&D. (See ch. 8.) ty of more complex chemicals and materials. The yields for the other chemical reactions The large (polymer) molecules in lignocellulo- were based on established experimental and sic materials are converted to the desired industrial data. chemical feedstocks either: 1 ) by chemical About 95 percent of these synthetic poly- means or 2) with heat or microwaves. The dis- mers (plastics, synthetic fibers, and synthetic tinction between these two approaches, how- rubbers) can be derived from wood or other ever, is not always clear cut. Iignocellulosic materials, although the circuitous synthesis route required for some of The chemical means include treatment with them might make such processes uneconomic acids, alkaline chemicals, and various bacteri- at this time. I n al 1, SIightly less than 60 million al processes. Pretreatments also often involve dry tons (about 1 Quad) of wood per year some heating and mechanical grinding (see ch. could supply 95 percent of these synthetic pol- 8). The three basic polymers– Iignin, cellulose, ymer needs; and the ratio of cellulose to Iignin and hemicellulose — are reduced to sugars and required (2:1) would be about the same as their various benzene-based (so called aromatic) natural abundance in wood. This quantity of chemicals, which can be used to synthesize the wood is relatively modest in comparison to chemical feedstocks by rather direct and effi- OTA estimates of the quantities that can be cient chemical synthesis or fermentation (see made available, and in all cases it serves as a figure 38). Some of the major petrochemical direct substitute for chemicals derived from feedstocks that can be produced in this way fossil fuels (mostly oil and natural gas). About are shown in table 73, together with the quan- three to five times as much wood would be tities of these chemicals (derived mostly from needed to supply all petrochemical needs petroleum) which were used by the chemical using more or less established chemical syn- industries in 1974. thesis routes,10 and again there appears to be no technical barrier to supplying these quan- The quantities of wood that would be re- tities of wood. In both cases, however, addi- quired to satisfy the 1974 U.S. demand for plastics, synthetic fibers, and synthetic rubber ‘Goldsteln, op clt 1 “1 S Goldstein, Department of Wood and Paper Sc Ien( e, from the above chemical feedstocks are shown North Carolln~ State Unlverslty, Raleigh, N C , prlvdte communi- in table 74 for the various types of products. cdtlon, 1980 230 . Vol. I/—Energy From Biological Processes

Figure 38.—Synthesis Routes for Converting Lignocellulose Into Select Chemical Feedstocks

Hydrolysis C H Hydrogenation Phenolic C 6H 50H 6 6 Lignin Pyrolysis Benzene

SOURCE: 1. S. Goldstein, “Chemicals From Lignocellulose,” Biotechnol. and Bioen. Symp. No. 6, (New York: John Wiley and Sons, Inc., 1976), p. 293. Ch. 12—Chemicals From Biomass . 231

Table 73.–Major Petrochemicals That Can Be Synthesized Table 74.–1974 Production of Plastics, Synthetic Fibers, From Lignocellulose and Rubber, and Estimated Lignocellulose Raw Material Base Required 1974 U.S. production (in billions of pounds) Lignocellulose a Total Iignocellulose Production required 3 3 Ammonia ., ...... , 31.4 Material (10 tons) (10 tons) Methanol ., ., ...... 6.9 Plastics Hemicellulose Thermosetting resins Ethanol. ., ., ., ... ., ., ., . . 2.0 Epoxies ...... 125 355 (L) Polyesters ., ., ., 455 1,220 (L) cellulose Urea ...... 420 — Ethanol, ...... , ., ., ... ., 2.0 Melamine...... 80 — Ethylene ...... 23,5 Phenolic and other tar-acid resins 670 1,915 (L) Butadiene ., . . 3.7 Thermoplastic resins Lignin Polyamide ...... 100 285 (L) Phenol ...... , ., ... 2.3 Polyethylene Benzene . . 11.1 Low density...... 2,985 11,940 (c) High density . . . . 1,420 5,680 (C) SOURCE: From l. S. Goldstein Chemicals From Lignocellulose in Biotechnology and Bioener- Polypropylene and copolymers 1,125 4,500 (c) gy Symposium No 6( John Wiley&Sons Inc. p 293 1976) Styrene and copolymers, ., 2,505 7,445 (L) Polyvinyl chloride ., 2,425 4,225 (C) tional energy would be needed to provide heat Other vinyl resins ...... 175 440 (c) for the syntheses; and, in some cases, this is Total plastics. ., 12,485 more than the energy content of the chemical Synthetic fibers feedstock.11 As of 1976, the petrochemical in- Cellulosic Rayon ...... 410 — dustry consumed 1.2 Quads/yr of oil and natu- Acetate ...... , 190 — ral gas for fuel and 2.3 Quads/yr for feed- Noncellulosic stocks. 12 Nylon...... 1,065 3,045 (L) Acrylic, ...... 320 640 (C) Another approach to chemicals from ligno- Polyester ...... 1,500 4,020 (L) Olefin ...... 230 920 (C) cellulose involves heat, partial combustion, or Total noncellulosic fibers, 3,115 the use of microwave radiation to break the Synthetic rubber natural polymers into smaller molecules suit- Styrene-butadiene ...... 1,615 5,700 (c) able for the synthesis. Biogas derived from the 1,920 (L) anaerobic digestion of biomass could also be Butyl...... 180 1,060 (C) Nitrile ...... 95 190 (c) used in some of these processes, but the yields Polybutadiene ...... 360 2,120 (c) are Iikely to be lower than for the more direct Polyisoprene 100 — processes. Some possible synthetic routes are Ethylene-propylene. ., 140 825 (C) Neoprene and others...... 280 — shown in figure 39. Total synthetic rubber. . . . 2,770 Total plastics, noncellulosic fibers, and rubber. . . . . 18,370 Obtainable from Iignocellulose 17,490 58,445 Cellulose drived (C). ., . . . . . 38,240 (C) Lignin derived (L) ...... 20,205 (L) ‘1“ BIg Future for SVnthetlcs, ” Science, VOI 208, p 576, May 9, 1980 aEstlmated from Opllmlstlc approximate yields of monomers obtainable (C) cellulose derived (L) 12G B Hegeman, Report to the Petrochemical Energy Group hgmn derwed

on 1976 Petrochem/ca/ /ndustry Prof//e (Cambridge, Mass Ar- SOURCE I S Goldstein. “Potential for Converrlng Wood Into Plasttcs, Science VOI 189, P thur D Little, Inc , June 28, 1977) 847, 1975 232 ● Vol. n-Energy From Biological Processes

Figure 39.—Chemical Synthesis Involving Thermal Processes and Microwaves

SOURCE: Office of Technology Assessment.

The production of ammonia’ 3 and methanol types of biomass are shown in table 75. Pre- from wood can be accomplished with commer- sumably by learning more about pyrolysis, the cial technology (see ch. 7 for further details of yields of select chemicals would be increased the methanol synthesis). The Fischer Tropsch to a level where it could be economical to ex- process is commercial in South Africa (al- tract that chemical from the gas. An example though the source of the synthesis gas is coal might be the conceptual equation: rather than biomss). The economics of the C H O - 3CO processes other than methanol synthesis have 10 14 6 2 + 3.5 C2H4 not been assessed by OTA for this report. Wood Carbon dioxide Ethylene (solid) (gas) (gas) The other processes yielding various chemicals are considerably less developed. The where 43 weight percent of the dry wood is yields of some chemicals that have been pro- converted to ethylene (which is by far the Iarg- duced in laboratory experiments using rapid est volume petrochemical used for chemical heating and gasification (pyrolysis) of various synthesis in the world). If it becomes practical to achieve relatively high (e. g., 30 weight per- 1‘R W Rutherford and K Ruschln, “Production of Ammonia cent) yields of ethylene, then this process Synthesis Ca$ From Wood Euel In Incfla, ” presented at a meeting of the Institute of Chem Ica I E nglneers, London, Oct 11, 1949 could be competitive with petroleum-derived Ch. 12—Chemicals From Biomass ● 233

I I

I 234 ● Vol. II—Energy From Biological Processes

References to Table 75

Antal, M. J., W. E. Edwards, H. C. Friedman, and F. E. Rogers, ‘‘A Study of the Steam Gasification of Organic Wastes, Environmental Protection Agency University grant No. R 804836010, final report, 1979. Berkowitz-Mattuck, J, B. and T. Noguchi, “Pyrolysis of Untreated and APO-THPC Treated Cotton Cellulose During l-See. Exposure to Radiation Flux Lev- els of 5-25 cal/mc2 sec.1, J. App/ied Po/ymer Science, vol. 7, p. 709, 1963. Brink, D. 1.., ‘ ‘Pyrolysis –Gasification–Combustion: A Process for Utilization of Plant Material, Applied Polymer Symposium IVO. 28, p. 1377, 1976. Brink, D. L, and M. S. Massoudi, ‘‘A Flow Reactor Technique for the Study of Wood Pyrolysis. 1. Experimental, J. Fire and flammability T, p. 176, 1978. Diebold, J. P. and G. D. Smith, ‘‘Noncatalytic Conversion of Biomass to Gasoline, “ ASME paper No. 70-Sol-29, 1979. Hileman, F. D., L. H. Wojeik, J. H. Futrell, and 1. N. Einhorn, ‘(Comparison of the Thermal Degradation Products of Cellulose and Douglas Fir Under Inert and Oxidative Environments, Therma/ Uses and Properties of Carbohydrates and Lignins Symposium, Shafizadek, Sarkenen, and Tillman, ed. (Aca- demic Press, 1976), p. 49. Lewellen, P. C., W. A. Peters, and J. B. Howard, “Cellulose Pyrolysis Kinetics and Char Formation Mechanism, Sixteenth Symposium (International on Combustion (The Combustion Institute, 1976), p. 1471. Lincoln, K. A., ‘‘Flash Vaporization of Solid Materials for Mass Spectrometry by Intense Thermal Radiation,Ana/ytica/ Chemistry, VOI 37, p. 541, 1965. Martin, S., “Diffusion-Controled Ignition of Cellulosic Materials by Intense Radiant Energy, ’ Tenth Symposium (/niernationa/) on Combustion (The Com- bustion Institute, 1965), p. 877. Prahacs, S., “Pyrolytic Gasification of Na-, Ca-, and Mg- Base Spent Pulping Liquors in an AST Reactor, ’ Advances in Chemistry Series, vol. 69, p. 230, 1967. Prahacs, S., H. G, Barclay, and S. P. Bhaba, ‘‘A Study of the Possibilities of Producing Synthetic Tonnage Chemicals From Lignocellulosic Residues, Pu/p and Paper Magazine of Canada, vol. 72, p. 69, 1971. Rensfelt, E., G. Blomkvist, C. Ekstrom, S. Engstrom, B. G. Espenas, and L. Liinanki, ‘‘Basic Gasification Studies for Development of Biomass Medium-Btu Gasification Processes. Stern, E. W., A. S. Logindice, and H. Heinerman, ‘‘Approach to Direct Gasification of Cellulosics, Industrial Engineering Chemistry Process Design and Development, vol. 4, p. 171, 1965. ethylene. The ethylene could be converted to netic radiation) could possibly be used to ethanol, and overall processing costs (wood to break specific predetermined chemical bonds ethanol) may be considerably lower than those in order to guide and control the decomposi- projected for fermentation processes (see chs. tion of the biomass. 7 and 8). Rapid pyrolysis, cracking, and microwave The liquefaction process for producing a py- processes are still at the research stage and rolytic oil (see ch. 7) might also be carried fur- considerable work is required to determine ther by cracking the oil in a way that is analo- their feasibility. Efforts along these directions, gous to current oil refinery technology. In ad- might lead to significant advances in the use of dition, microwave energy (or other electromag- biomass for chemicals and fuels.

U . S . GOVERNMENT PRINTING OFFICE : 1980 0 - 67-968 : Q L 3