ECONOMIC ANALYSIS OF BIOFUELS PRODUCTION IN ARID REGIONS by Helen Ann Kassander Ruskin

ECONOMIC ANALYSIS OF BIOFUELS PRODUCTION IN ARID REGIONS by Helen Ann Kassander Ruskin

ECONOMIC ANALYSIS OF BIOFUELS PRODUCTION

IN ARID REGIONS by

Helen Ann Kassander Ruskin

A Dissertation Submitted to the Faculty of the

COMMITTEE ON ARID LANDS RESOURCE SCIENCES

In Partial Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

In the Graduate College

THE UNIVERSITY OF ARIZONA

1983

THE UNIVERSITY OF ARIZONA

GRADUATE COLLEGE

As members of the Final Examination Committee, we certify that we have read the dissertation prepared by

Helen Ann Kassander Ruskin entitled

Economic Analysis of Biofuels Production in Arid Regions and recommend that it be accepted as fulfilling the dissertation requirement for the Degree of

Doctor of Philosophy

Datê

3---

Date

Date

CI

/

0 t3

Date

Date

Final approval and acceptance of this dissertation is contingent upon the candidate's submission of the final copy of the dissertation to the Graduate

College.

I hereby certify that I have read this dissertation prepared under my direction and recommend that it be accepted as fulfilling the dissertation requirement

Di

-

S

i

tation Director

dl

Dat

STATEMENT BY AUTHOR

This dissertation has been submitted in partial fulfillment of requirements for an advanced degree at The University of Arizona and is deposited in the University Library to be made available to borrowers under rules of the Library.

Brief quotations from this dissertation are allowable without special permission, provided that accurate acknowledgment of source is made. Requests for permission for extended quotation from or reproduction of this manuscript in whole or in part may be granted by the head of the major department or the Dean of the Graduate College when in his judgment the proposed use of the material is in the interests of scholarship. In all other instances, however, permission must be obtained from the author.

SIGNED:

ACKNOWLEDGMENTS

I am deeply grateful to the many people whose assistance and support made this work possible. Special recognition is due Dr. David

Pingry for invaluable information, editorial guidance, and unfaltering wisdom and patience. No expression of thanks could adequately convey my debt or gratitude for his help.

Dr. J. D. Johnson and the Office of Arid Lands Studies provided intellectual and financial support throughout the course of my academic career. He and Dr. Helmut Frank provided the impetus for this study.

Their comments, along with those of Drs. Daniel Evans and Dennis Cory, contributed greatly to the manuscript.

Dr. Timothy Shaftel graciously provided the solution algorithm and computer code used to solve the model. Mercy Valencia and Vicki Lee

Thomas helped me cope with the program execution. N. Gene Wright has been more than generous with his computations and experience in agricultural economics. Dr. Steve McLaughlin, Barbara Kingslover, and the staff of the Hydrocarbon project supplied considerable data.

Erika Louie produced many drafts and the final manuscript. Her aid in format and editorial matters was as valuable as her expert typing.

Kathleen Crow verified the calculations.

To all these, to my exceptional family, and to many more who have been an inspiration, I offer my profoundest thanks.

TABLE OF CONTENTS

LIST OF TABLES

LIST OF ILLUSTRATIONS

ABSTRACT

SUMMARY

OF NOTATION

ABBREVIATIONS

CHAPTER

1.

INTRODUCTION

The

Energy Problem

Possible Alternative

Energy

Supplies

Biomass

Objectives/Methodology

2.

BASELINE

PRODUCTION

PROCESS

Euphorbia lathyris

Plant Propagation

A.

Land

Preparation

and

Planting

B.

Growth

and

Cultivation

C.

Harvest

and Package

Yields

Total

Cropping Costs:

A

+ B + C

Transportation

to Extractor

(D)

Processing

E.

Storage

F. Extraction

Transportation

to Refinery

(G)

Refining

(H)

Systems Summary

3.

DESCRIPTION OF

MODEL

Notation

Indices

iv

Page

vii viii ix

xi

xii

10

12

13

20

20

22

22

14

16

16

18

23

27

27

28

30

33

33

1

6

8

2

3

TABLE OF CONTENTS -- Continued

Parameters

Variables

Relationships

Objective Function

Constraints and Assumptions

Farm

Extraction

Refinery

Solution Program

4. ARIZONA CASE STUDY

Biofuels and Arizona Agriculture

Delineation of Study Area and Model Input

Study Area Mapping

Assumptions and Baseline Conditions

Program Input

Simulations

Simulation 1: Extractor Layout

Simulation 2: Cropping Costs

Simulation 3: Yield

Simulation 4: Harvest Strategy,

Reflected in Storage Cost

Simulation 5: Extraction Costs

Simulation 6: Transportation

Comparative Energy Pricing

Summary of Results

5 SUMMARY AND CONCLUSIONS

Model Capability

Arizona Case Study Results

Policy Implications

Energy Supply

Agricultural Policy

Environmental Policy

Regional Growth and Development

Conclusion

APPENDIX A: CROP PLANTING AND GROWTH DATA

A. Land Preparation and Planting

Bed Preparation

Page

44

44

47

47

52

54

61

62

66

68

70

72

75

75

79

34

35

35

35

37

38

38

39

41

83

90

91

92

92

83

85

89

89

94

95

95

TABLE OF CONTENTS

-- Continued

Seed

Growth

and

Cultivation

Fertilizer

Weed

and Pest

Control

Additional Growth

Stimulants

Irrigation

Salinity

Additional Cultivation

Yields

APPENDIX B: PROCESSING

DATA

E.

Storage

F. Extraction

Conveyers

Grinding

Solvent

Filtration

Recovery

Products

APPENDIX C: MAPPING

OF POSSIBLE

E.

LATHYRIS

GROWTH AREAS

IN ARIZONA

Natural

Habitat

Temperature

and Insolation

Soil Quality

Availability

of Land

Infrastructure

Water Needs

and

Availability

APPENDIX

D:

ENERGY

BALANCE

REFERENCES vi

Page

96

98

98

102

102

103

108

111

111

112

113

113

113

113

117

117

117

119

123

124

124

130

133

139

139

156

164

LIST OF TABLES

Table

15.

Equalization levels for "competitive" scenarios at higher energy input price levels

Page

1.

Estimated crop budget (A), preparation and planting

($/acre)

15

16

2.

Plant density and productivity

3.

Estimated crop budget (B), growth ($/ac)

4.

5.

6.

7.

8.

17

Estimated crop budget (C), harvest ($/ac)

19

Estimated total crop budgets

21

Estimated processing capital budget (F) (million $) . .

25

Estimated processing operation budget (F) (million $) .

26

Farm area acreage of baseline network

49

9.

Objective function cost components as percent of input costs

56

10.

Simulation 1: Costs corresponding to differing extraction facility distribution

11.

Simulation 2: Response of total cost to changes in cropping cost

12.

Simulation 3: Response of total cost to changes in farm and extraction yield

13.

Simulation 4: Variation in total cost with changes in harvest timing and storage

14.

Simulation 5: Response of total cost to extraction cost changes

63

67

69

73

74

78

82

16.

Summary of simulations

17. Results of simulations ranked in order of input on delivered cost

86 vii

LIST OF ILLUSTRATIONS

Figure

1.

Generalized flow diagram of Euphorbia lathyris production

2.

Solvent extraction of Euphorbia lathyris by Arizona

Method

3.

Biofuels production model overview

4.

Constraint matrix

5.

Estimate of acreage suitable for E. lathyris cultivation

6.

Baseline network of farm sites and transportation routes used in all simulations

7.

Annual capital costs as a function of tons processed

.

8.

Input matrices for sample problem

9.

Output matrix

10.

Optimum industry network from Simulation I

11.

Optimum four-extractor network

Page

11

50

60

65

71

51

55

58

24

29

40 viii

ABSTRACT

The objective of this study is to develop a model to evaluate the economic feasibility of

biofuels

production, and in particular to isolate the variables crucial to feasibility. The model constructed to define these variables is unique in its ability to accommodate a variety of plants and to integrate all portions of the production process; it was tested on a case study of a Euphorbia lathyris industry.

The model minimizes costs of production to determine the best configuration for the industry. Total cost equals the sum of costs incurred in each segment of the process: growth, harvest, transport, and extraction. The solution is determined through a non-linear transportation-transshipment algorithm which describes production as a series of nodes and links.

Specific application of the model was analysis of E. lathyris

biofuel

production in Arizona. Simulations were run examining the sensitivity of

biocrude

cost to changes in input parameters. Conclusions are summarized as follows.

• No change in any single element can reduce final cost sufficiently to enable competitive production in the near future.

• The major factor necessary to bring cost into range is improvement in biological yield. Two components of equal importance are tonnage produced per acre and percentage

extractables

recovered in processing.

i x

• Lowering cropping costs provided the most effective improvements of economic inputs. Perennial crops significantly reduced farm costs.

• Transportation costs outweighed economies of scale in extraction; extractor location close to crops is more efficient than centralized.

The cost minimization model was successfully used to isolate the critical factors for an E.

lathyris

industry in an arid region. Results determine that this industry would not be competitive in Arizona without dramatic improvements in yields and moderate changes in a combination of input costs. Viability is critically dependent on improvements in tonnage yield produced per acre and percent extractables recovered.

SUMMARY OF NOTATION a

= capital scale factor a.

= acreage at farm i

C a

= cost per ton of farming

C k

= extractor capital cost

C o

= extractor annual operating cost

C s

= storage cost

C t

= shipping cost

C' = post-extraction shipping cost

I

= biomass supply nodes

J = biomass demand nodes m..

= miles from i to j ij m r

= refinery capacity

N = number of nodes n a

= number of farm sites n e

= number of extraction sites

R = extraction rate

S.

= percent of shipment stored t a

= tons/acre yield t ij

= tons shipped from i to j

Y i

. = yield at farm i x i

ABBREVIATIONS

ETOH ft ft

2 gal gal/min gm ha hr ac ac-ft acre acre-foot (feet) bl bl/ac barrel(s) barrels per acre

Btu British thermal unit

British thermal units per pound

Btu/lb

CAP Central Arizona Project cyclohexane CH cm centimeter(s) carbon dioxide

CO

2 d day

ETA U.S. Department of Energy, Washington, D.C., Energy

Information Administration

HP ethanol foot (feet) square feet gallon(s) gallons per minute gram(s) hectare(s) hour(s) horsepower xii

lb m

2 mi

mM

Na

NaC1

NO

3

N-P-K

0

2

OALS

OPEC

OVPR

PGR

PO

4

PPm psig

RPM

SRI

T/ac

T/d t-mi

T/yr pound(s)

square

meter(s) mile(s) millimole(s)

sodium sodium chloride nitrate nitrogen-phosphorus-potassium oxygen

University of Arizona, Tucson, Office of Arid Lands

Studies

Organization for Petroleum Exporting Countries

University of Arizona, Tucson, Office of the Vice

President for Research plant growth

regulator(s)

phosphate parts per million pounds per square inch revolutions per minute

SRI, International

ton(s)

tons per acre tons per day ton-mile tons per year

US

w/m

2

WUE yr

$

%

/

° C, ° F

United States watts per square

meter water

use

efficiency year(s) inch(es)

percent per

degrees

Centigrade, Fahrenheit dollars xiv

CHAPTER 1

INTRODUCTION

"Energy is involved in all aspects of society; not only does our quality of life depend on the availability of energy, so does our national security, economic strength, and international stability" (University of Arizona, Office of the Vice President for Research (OVPR), 1980).

In recent years the United States domestic demand has been greater than domestic supply at historic prices. Several additional energy sources have been considered in an attempt to increase production in the United States and decrease dependency on foreign imports. One of these alternative sources is energy derived from plant products. This study will examine biofuels, liquid plant extracts, as potential producers in arid regions. Although some research has been done on portions of a biofuels industry, there has been no examination of the integrated system of production, and in particular of the economic feasibility of such a system.

The objective of this study is to evaluate the possibility of industrial biofuels production, and to assess the conditions necessary to create a viable industry. The hypothesis is that there are critical variables which will determine industry feasibility. The model constructed to find these variables joins the various sectors of the

1

2 production process in a transportation model which is used to evaluate the costs involved in the industry as a whole.

Chapter 1 will lay the foundation for the examination

of

biofuels. It will outline the energy problem, examples of possible supplies, this study's focus on biofuels, and model objectives.

The Energy Problem

During the first half of the 20th century, availability and relatively low market prices promoted economic dependency on increasing energy inputs, in particular liquid fuels. World consumption doubled each decade (Cheney, 1974). The high fuel costs and supply dislocations of the 1970s brought new realization that conventional sources are unreliable and exhaustible. With this disruption came concern that economic productivity would be curtailed by continued reliance on finite resources with uncertain supplies and escalating prices.

The United States (U.S.) depends heavily on energy inputs, and is especially vulnerable to disturbances in world supplies. Oil, gas, and coal provide 95% of U.S. energy (Cheney, 1974; Calvin, 1977b;

1978b). All three sources are fossilized photosynthetic carbon. Since

1967 more oil and gas have been used than have been discovered in new reservoirs. U.S. production, which peaked in 1970, will probably continue decreasing for the next decade (Mohs, 1981). New discoveries of fossil fuels are declining, while exploration costs increase. Deeper drilling and lower oil qualities dictate greater effort and cost to obtain oil from proven reserves. New fields will require heavy capital investment, as well as long development time.

As

U.S. petroleum production leveled off, imports rose--from under 20% of use in 1960 (Hale, 1976) to nearly 50% by the end of the

1970s (OVPR, 1980). Foreign oil payments reached $80 billion in 1980

(Mohs, 1981), disrupting capital formation and stimulating inflation.

The Organization of Petroleum Exporting Countries (OPEC) gained virtual control of supply to dictate a ten-fold price increase in under ten years

(Weaver, 1981). Foreign influence and the gradual depletion of reserves will in the long run continue to push traditional fuel prices upward.

Fuel prices are not the only obstacle to continued use of fossil fuels. OPEC's emergence as a world power accentuates the importance of reliable supplies. Dependence on foreign sources makes the U.S. economy subject to political instability in producing regions. War, internal uprisings, or policy disagreements have previously and could again cut off vital shipments. It may no longer be prudent to rely on one principal fuel source. Alternative sources are already increasing diversification of U.S. energy supplies.

3

Possible Alternative

Energy

Supplies

Current additions to oil and natural gas include coal, synthetic fuels derived from coal, nuclear and geothermal energy, and solar power which includes biocollection. Although exploitation is technically possible for many sources, each has associated production constraints; uncertainties remain about their costs, safety, and environmental effects.

Coal, at present use rates, remains a relatively abundant resource even though exhaustible. Its use has waned since World War I due

4 to production and environmental problems. Open pit mining is destructive of the environment, produces unsightly pond and tailings dumps, and can be dangerous if steep-sloping pit sides collapse. Underground mining is extremely hazardous. Labor strikes and increasing manpower costs add further expense to a capital-intensive industry. Bulky material is difficult to handle and transport. Among the most serious concerns is environmental degradation due to coal slag disposal, and sulfur, heavy metal, and nitrogen oxide emissions from combustion. Emissions can result in acid rain, while increased carbon dioxide has potential to modify climate.

Synthetic fuels are expensive, and embody the ecological and safety hazards of coal production. Shale oil requires capital investments and mining operations with corresponding habitat disruption.

Mining and processing will release dangerous chemicals and particulates.

The need for large quantities of water for shale oil recovery poses a problem for Western states with large shale deposits. Along with large capital investment and risks in production and pollution similar to those for coal, liquefaction involves release of carcinogenic agents dangerous to workers. Crude oil from tar sands, heavy oil, and residuals from conventional extraction also are costly to recover.

Nuclear power has vast potential; enriched fuel contains over a million times the energy in the same weight of coal (National Geographic Society,

1981).

Corresponding capital investments are enormous.

Extensive regulations and costly precautions are necessary, yet plant safety and waste disposal still represent serious political concerns.

Possibilities of accidental release of radioactive materials to air or water supplies, even if unlikely, pose such serious and extensive threats that nuclear development must be approached with caution. Used fuel still contains high levels of radioactivity, and must be stored or disposed of in isolation which can be secured for thousands of years. Views

5 on risks of nuclear energy vary widely; but whatever the actuality, opponents to nuclear power have gained political prominence and have forced delays in further development.

Geothermal power is generally not yet economically competitive.

Most energy supplied is low to medium heat, and must be used on-site since the power cannot be transported. Environmental impacts vary considerably, but are again a major concern. Gases such as mercury and amonia and volatile compounds such as barium and arsenic are emitted, damaging air quality. Liquid wastes contain dissolved gases and solids so that disposal methods must protect acquifers and land use. Removal of large amounts of fluid from underlying rock can lead to subsidence and other instabilities.

The sun is the ultimate source of energy and the only resource which can be used indefinitely. Collection of solar energy through absorbers, windmills, or hydropower generators is one method of using solar radiation. All are subject to large capital expenditures and changes in weather conditions. Chemical compounds used in collectors and the need for disposal of system fluids pose some environmental threats.

Large expanses of land are necessary for significant power generation.

6

Institutional constraints at this time include building codes and lack of defined legal rights to sunlight.

Biological conversion represents a second option for solar power utilization, since green plants are solar energy collection devices.

With the exception of wood burning and alcohol, biomass fuel production is in the early development stage. Although subject to drawbacks, in particular the environmental impacts of opening large areas to new farming, biofuels offer a potential renewable contribution to energy supplies and warrant examination.

Relative to the other options discussed, little research has been done on biomass energy production. Most of the plants under consideration have not been domesticated. New methods will be needed to extract fuel products. Analysis of the economic viability of the industry is needed before committing resources to the development of a biofuel system.

Biomass

Biomass is the renewable resource generated by biological sources. It includes living material such as trees, field crops, and microorganisms, and products such as plant and animal wastes. Benemann

(1980) classifies four types of usable biomass:

1. Wastes: organic materials accumulated at specific locations, with associated disposal costs.

2.

Residues: plant materials remaining in fields after crops or timber have been harvested.

3.

Energy crops: plants specifically cultivated for their fuel content. Examples include short-rotation tree farms (energy plantations), ocean kelp farms, land-based aquatic plant systems, or grain or cane alcohol plants.

4.

Integrated biomass systems: multiple types of plants grown together, interacting in complementary ways with the resource base.

7

This analysis deals specifically with energy crops. Biomass farming is advantageous because it provides a relatively clean, renewable feedstock with a large measure of social acceptability. Production relies on photosynthetic conversion of solar energy into biomass. Hydrogen captured from water and carbon dioxide from the atmosphere are converted into sugar or cellulose. Although many plant carbohydrates can be used to generate energy, the prime candidates for fuel production are those such as Euphorbia lathyris which store further reduced carbon in the form of low-molecular-weight hydrocarbon.

Biomass resources are processed through combustion, pyrolysis, gasification, liquifaction, chemical extraction, or biological (anaerobic digestion, fermentation, biophytolysis) processes. Resulting bioenergy products include solid fuels, alcohols, methane or other gasses, electricity, and directly combustible or biocrude oils. Rubber, rosins, or other specialty chemicals are additional extracts. Depending on future energy price levels, the higher valued specialty products may be the primary crop extract, with energy as a by-product.

8

Objectives/Methodology

The growth of previously uncultivated energy crops in marginal conditions is essentially unexplored both in terms of cultivation and economics. As stated by the California Energy Commission (1979), "Development of these plants as possible energy farm crops is in its infancy.

These plants have been cultivated on a very limited scale, and as a result, little is known if

its

(sic) value for energy production." In trying to domesticate a wild plant for commercial cropping, factors such as resource requirements, productivity levels, end uses, and processing steps remain relatively unknown. Economic questions include the cost of planting, harvesting, transportation, and processing into fuel and other products. The processing of the harvested crop contributes a significant cost increment to the overall product cost. Economies of scale in processing may dictate a minimum crop production size. The agricultural complex and the fuel processing plant must therefore be treated as a system in conducting economic analysis.

The objective of this study therefore is to evaluate the possibility of industrial biofuels or biomass production, and to assess the conditions necessary to create a viable industry. Total utilization of the crop is assumed. The feasibility model constructed is generalized, to be capable of accommodating a variety of plants, and wholistic, to integrate the portions of the production process. Two principal topics are being joined: plant production and resource (especially water) utilization. The interaction of the two is critical in domestication of

a new crop in arid regions, where there must be a realistic assessment of

9 water availability before cropping is undertaken.

The model is created by first defining the component parts of the production process. Each portion is simulated, and this series of sub-models combined in a transportation-transshipment format to define an integral whole. The model is tested through a case study of

Euphorbia lathyris growth in the state of Arizona. Several simulations are analyzed to determine the ranges of parameters which result in a feasible industry, and to evaluate the industry sensitivity to factor price and mix changes.

CHAPTER 2

BASELINE PRODUCTION PROCESS

In order to construct an industry model, it is necessary to understand the underlying production process. This model is based on growth and extraction of Euphorbia lathyris. E. lathyris is chosen as a candidate biofuel plant because of possible adaptation to dry regions, high biomass production potential, and high content of hydrocarbons similar to petroleum.

The cultivation and processing of E. lathyris is divided into several stages. Each stage is examined individually and acts as a submodel, or node, within the overall model framework. The general characteristics of E. lathyris will be stated first, followed by a description of each element of the production process. These are:

A.

Planting

B.

Cultivation

C.

Harvest and Package

D.

Transport to Extractor

E.

Storage

F.

Extraction

G. Transport to Refinery

Figure 1 shows the generalized flowchart of these stages. Each process is keyedby letter to the corresponding text paragraphs and tables.

10

Inputs:

Land, Labor,

Capital, Water,

Energy, Etc.

4-

Plant Propagation

Plant (A)

Cultivate

(B)

Harvest, Field

Dry, Package

(C)

Processing

Storage

(E)

Extraction

(F)

Refine and

Market

(H)

=

Transportation; D

=

Extraction Plant; G

=

To Refinery

Figure

1. Generalized flow diagram of Euphorbia lathyris production.

11

Euphorbia lathyris

1

Euphorbia lathyris is a member of the subgenus Esula in the large and complex genus Euphorbiaceae. Common names are caper spurge,

12 since fruits resemble capers, and gopher plant, mole plant, or gopher weed, due to reputed rodent-repellant qualities.

Smith and Tutkin (1968) suggest its origins are restricted to

Mediterranean regions, but Prokhanov (1949) sighted it occurring wild in the mountains of western China. It may have been cultivated for seed oil in the Soviet Union, China, and Japan. E. lathyris was probably introduced to Europe during the Middle Ages. It was imported from the

Mediterranean to California and the Southeastern U.S. where it now grows wild. It is distributed widely in temperate areas, particularly along the coast and in disturbed areas of mesic habitats. It has not encroached on true arid regions.

E. lathyris is a biennial herb but also may be propagated as an annual or perennial. The growing plant has a single stalk. Mature plants have a radial symmetry and are approximately 12 inches (") in diameter. Height ranges from 20-80". Irregular branching initiates from lower nodes of the main stem when it is 12-20" tall.

The dark green leaves grow thickly alternate and opposite along the main stalk. This decussate pattern is unique to lathyris within the subgenus. Leaves are simple, lanceolate, approximately 4" long, with a major center vein.

1. Hinman et al. (1980); Sachs and Mock (1980); and Peoples (1981a) .

13

Plants flower in midsummer. Flowers appear in terminal clusters on mature branches. Seed ripen in late summer or early fall. They are approximately 0.1" in diameter, borne in 3-lobed fruit each 1/2" in diameter. Capsules are large, spongy, and indehiscent--a trait which separates lathyris from the other members of Esula, and may indicate previous cultivation due to ease in harvesting. Seed, although poisonous, have been reported used as substitutes for capers and coffee

(Nakao, 1976), as an emetic (Schauenberg and Paris, 1977) and as a treatment for dropsy (Chopra, Nayar and Chopra, 1956).

Interrupted lactifers exude an emulsion of hydrocarbons in water. Initial estimates (Calvin, 1977d) indicate a content of 5-8% latex.

Preliminary analysis of Euphorbia species shows a hydrocarbon with molecular weight of approximately 50,000. Chemical extraction from dried plant material isolates products ranging from high-grade fuel of approximately 17,500 British Thermal Units per pound (Btu/lb), to bagasse of

5400 Btu/lb.

Plant Propagation

Prices for baseline production estimates are based on current agricultural and chemical prices. Real 1979-1980 factor prices, rounded to the nearest dollar, are used throughout. Differences in relative prices over the time horizon are accounted for by varying factor ratios in the model manipulation. Predictions of price changes and inflation are not attempted.

14

All estimates used here are on an annual per-acre (ac) basis.

They do not depend on any optimal-size farm or any scale economies. The assumption is made that farmers are aware of constraints on machinery

(particularly of time limits in harvesting), and will purchase and utilize machinery efficiently. According to Wright (1982), scale economies are not usually significant factors in cropping costs.

Estimates of growing costs vary widely. Wright's (1981) calculations of variable costs range from a total of $100/acre (ac) (8" irrigation and 90 lb fertilizer) to $619/ac (112" water and 500 lb fertilizer). Total costs on the same scenarios vary from $206 to $867.

Crop budgets for this report are prepared by attempting to estimate appropriate growing conditions for E. lathyris as a baseline for future development. Arizona and inland southern California are used as examples of an arid growth environment. Implications of cultivation in other regions are discussed in the concluding chapter. Values chosen are not intended to be accurate depictions of development situations, but to serve as a reasonable base for examining variations in costs.

Cropping budgets are presented in three segments: (A) land preparation and planting costs; (B) growth and cultivation costs; and (C) harvest and packaging costs.

A. Land Preparation and Planting

Estimated crop budgets for land acquisition and planting activities are shown in Table 1. Planting density and productivity relations are represented in Table 2. Full data sets and a review of factors involved are presented in Appendix A.

15

Table 1.

Estimated crop budget (A), preparation and planting ($/acre).

Activity/

Material

Land

Planting

Seed

Plow

Disc

Total

Harrow

Plane

List

Buck Rows

Disk Ends

Total

Plant

Mendel et al.

(1979)

100

4

8

2

14

60

4

178

Doanel Alexander Wright

Peoples

(1981)

125

9

4

4

.4

.2

18

40

4

187

(1981)

1. Doane Agricultural Service, Inc.

Note: -- indicates information not available.

50

15

65

(1981)

50

16

26

2

8

70

65 (incl. seed)

141

(1981b)

16

Table 2. Plant density and productivity.

Density

(1000 plants/ac)

91

183

371

741

1483

(Kingslover, 1982,

P.

25)

Biomass

(dry lb/ac)

6,700

15,300

12,200

18,300

36,000

Total CH and ETOH

(bl/ac)

4.9

11.6

9.2

13.4

26.9

B.

Growth and Cultivation

A summary crop budget of activities during the growth period appears in Table 3. Fertilizer and insecticides, as well as irrigation, cultivation, and management are included. Determination of input amounts and prices, particularly irrigation, are discussed in AppendixA.

Not covered in crop budgets are repairs and maintenance, general supplies, and facilities which might not exist on previously untilled property. A contingency figure of 10% of total costs should be added.

C.

Harvest and Package

Before processing, biomass must be dried and ground into a coarse powder. Mechanical drying at 70

°

C (158

°

F) for 24 hours would accomplish adequate dessication (Sachs and Mock, 1980). The alternative is to field-dry the crop. Two possible harvesting scenarios are suggested (Mendel, Schooley, and Dickenson, 1979):

17

Table 3.

Estimated crop budget (B), growth ($/ac).

Activity/

Material

Fertilize

Nitrogen

Phosphorus

Application

Total

Insecticides

Herbicides

Application

Total

Irrigation

Water

Application

Pump

Repair

Fixed

Total

Other

Cultivation

Equipment

Labor

Management

Interest

Total

Total

Mendel et al.

(1979)

37.5

12

55

5.5

8.5

5

7

20.5

170

38

38

283.5

Doane

(1981)

12

58

65

3.5

73.5

14

14

8

73

7

16

48

16

87

247.5

Alexander

(1981)

180

26.5

5

10

15

221.5

Wright

(1981)

50

12

36

9

65

3

15

9

24

151

1.

Green-chop: cut, chop and load in one serial process. A drying

18 shed would be needed on farm premises. Transportation costs would increase, as would the chance of biomass rotting or fermenting.

2.

Dry-chop: mow and windrow to field-dry, then chop, bale or cube, load and transport. A sickle-bar type mower with moving knife blades is used.

Harvesting is usually done five to seven months after sowing, in

Arizona in late April or early May. Nemethy et al. (1978) recommend a

12-month cycle; but to avoid pathogens, harvest should not be delayed beyond June. Dry-chop harvesting is more efficient and is therefore the chosen method for this study. Costs for dry-chop harvesting appear in

Table 4.

Clearing or waste disposal must be added, along with storage, loading, possibly added drying, maintenance, and contingency. Mendel et al. (1979) estimate $42/ac total for harvest and drying; Alexander

(1981) raises the figure to $100.

Yields

The amount of biomass produced per acre is an important factor in the feasibility of bioenergy projects. It is often noted that

Euphorbia lathyris is an unimproved crop, being grown in less than ideal conditions. Kingslover (1982) notes that E. lathyris is a plant with little genetic variability, so that there is little prospect for increasing the hydrocarbon yield through selective breeding. The plant does

19

Table 4. Estimated crop budget (C), harvest ($/ac).

Activity/

Material

Harvest

Rake

Mow

Bale

Wire

Stack

Total

Fixed

Total

Cropping

Foster

Mendel and Fox and et al. Doane Alexander Wright Brooks Jorgensen

(1979) (1981) (1981) (1981) (1981)

(1980)

42

503

11

15

4

30

464

100

386

5

8

29

.1

4

46

60

398 650 392

20 have relatively large chromosomes that can be identified individually, providing the possibility for cytogenic improvement (University of

Arizona, Office of Arid Lands Studies (OALS), 1979). Wright (pers.

comm., 1981) reports initial yield data for Southern ecotype E. lathyris planted at 240,000 plants/ac, receiving 20" of water, 150 pounds of

16-24-0 fertilizer, no pesticides, and normal cultivation. Plants were sickle-bar harvested and dried to 10% moisture, resulting in 4.4 dry tons(T)/ac. At 30% extractables, 9 barrels(b1)/ac could be produced.

At 20% extractables, 4 bl/ac would result.

Total Cropping Costs: A + B + C

Cost estimates range from $400/ac or $50/T for 8 T/ac, to

$650/ac (Foster and Brooks, 1981) or $81/T. Wright (pers. comm., 1981) calculated the most consistent and complete estimates. The data set which most closely approximates optimal cropping conditions as discussed here is presented as the Total Estimated Crop Budget (Table 5). It is used as the baseline from which variations are calculated.

Transportation to Extractor (D)

Transportation costs will vary widely depending upon dry weight percentages, fuel costs, labor and maintenance, and packing procedures.

The most important factor will be mileage covered, which will in part be determined by extraction locations and transport routing calculated within the model. Although calculations of shipping costs are unavailable for the distances covered in this study, a charge of $0.10 per tonmile transported is extrapolated from data on agricultural trucking charges (iathorn and Armstrong, 1981).

21

o t*

-

n

1-1

C./)

't4

0

• •••

00

Cra

(.) v. ru

0 .c.

C cCs h cr: N C

,

..0

. . .

C. c r-- rsi

NG c C 1-.3

CV

•C

CI

C 07:

C.

..

nn

C

1 nn

• SC •-• .

n

C

CC

1

,.

.

.

N N

,. 4.: n

n t

,

I.

cC V

:

v. t*".

CV

,,,,,

4 A

0 c'l 0 if. C.; h

CV

C

C

.--.

5.

5. .5 — r,

ΠC.

C. C.

0".

C C

C

C"

COCC C

N

C cC c

C

,..

c

CL))

NN

4S- 4

,

:.a

-, ti-

.

CL .-...

C.

•C•

4 S:

,

'J

.7

4S-

.1.,

,..

o

..,

C

C

4e

C

COL

Ci))

L- if)

LI i

C

CC

C) C) v

.

:

..r,

4-1

4

-

.....

U

CC

7

0

0

0

OA

N.,

IA

4-

4-

I

Li

=

C.':

V;

1,-,

CV

Ili

CV hi

V.

1---

h

0

c .

.

i.n.

N

.-, h

C:

C,C.

0

.I.J

t....

.= CO

-0-

=

.C.

4 .,

4-

4-

0

C...

h

CJ

5

-7..

C

..S.

0

L.

.v..

h

N

C tri

-... v.

.-

C

C

1.0.

C)

c c c c cc c c c ...:

V: II; LC; v 14 V c v c c c c c 0 c

C.-

tl, LA

if: in

N

CV

.-•

..--

-

L...

œ

Cv

....-

C.V:.

......

7

.

c c c r 1

......

..:

7-

."

0-

Y.'

U) r‘s c

C

Li

c

Li)

. I

,

: CC ni

, r,

'. tr.

OC .—

C. LC CC r, •

-

•.:. rv

LP: . tC c

I

,

SC c c

IA

•n r- c c

V. G

n

.- h g

.i.,

CA

VI

.

=

I L. sC., CV NI 1

-

,.

LC p.

-

. OC

1_,: t''''. C CV

....

1,-. r•-•

VI r4

.-. v. CV C

.-1 CV

,,-.

...

C.

1,.

0,

C

C

11:

= " , ,

7 ..1

CV

C U .. V, LC

VI CV r-- .-.

X cd

.....,

Ott

C U

8 a

*4 c

C)

.....

1,-;

L'

LU v.

:i

4.4

3

-

:..,

1.1)

L

CL

I

4, CA

0

V .

-.

I.. e)

C .- ...,

'3. v•

._.

,....

.,

.

, '8

E

..

•-••

.

0 -

....., C-

G.

SC.

...: t

5 ._

C.

, c ni C

... - csi

.."

= c

C

_ 3

C-P-- .7 r:1 c c c

C G a u

..--..

Q

4) ct.

OC

.-

r.

'0:. a cr,

.1-, v.

{..) o

C)

4-

c.

(.

n

7; '

-

I/. g.

.7.

C

C.

0

1- c <7. !„?....

G) c4.

.---.

.g n

C.

L.

L.

0,

.-g

0

L. L.

,

:-..

L.

.--

C.

I

Li

1

0

L.

C.

Cd

4-.

C

0 1.

, .= tr.

I..,

L.

.__ s

cc.,

CZ

.1-.

V,

0

sr.

0)

m 1

C.1 s..-

:-.:,

0

0

7

0

'0

0 0 '-'g

Lii ... .-i CC.

Ej

..

.I

4,-, 7.;,

tn

4

-

C

1

tn

I)

.,

,-, 0 C 0

0 /- ,- Li

-.E L''' '-'-' .3.

L.

V)

.

i

,

›-

$.•

0

0

L.

a:

.4-

-

<E5

0<

0 r. . . .

= .:.

.7.• "'

'C

n:::

0 x

•-. i

0 C1

,

, -

; l

:

I

1

.

i

[

22

Processing

After the initial biomass is cut, dried, and ground, raw materials transported to the processing facilities are treated with solvents to extract crude biofuel. Extracted oil is then transported to a refinery where it is cracked with a zeolite catylist as is crude petroleum

(Calvin, 1980). Processing will be described here in terms of storage before extraction (E), extraction itself (F), transportation to refinery

(G), and refining (H). Full data are included in Appendix B.

E. Storage

If Euphorbia lathyris is harvested only once annually and an extraction plant is operated 11 months (328 days = 90% capacity), on-site storage will be needed. The amount of storage will depend upon the number of acres served, the harvest schedule, the cropping pattern, and the transportation program. The 30-day supply for a large extraction plant would be 110,000 T. This would occupy 1 million square feet (ft

2

). For a concrete floor and supported roof warehouse, the installed cost is

$15/ft

2

, or $27 million for 30-days operational storage capacity. The cost would be $81 million for 90-days capacity (SRI, International (SRI),

1979). Added mechanical services may be required. Ventilation, refrigeration, fire and seepage protection, as well as fork lift and labor, would increase costs.

The figures for 30 and 90 days capacity are used to calculate per-ton costs, which in turn are used to estimate storage costs for longer time periods.

23

F. Extraction

The Arizona Method of solvent extraction produces three fractions (Figure 2). Steps and equipment involved are outlined in Appendix

B. The following data and assumptions for solvent extraction plants are based on those for guayule or synthetic rubber (Taylor, pers. comm.,

1982).

The minimum feasible size is 1,000 T/day. Plants must operate for 292 days per year. Efficiency continues to improve as extraction facilities grow larger, up to 3,000 1/day maximum. Beyond this there is no further advantage to size.

SRI (1979) bases extraction plant cost estimates on a 3000 T/day extraction plant. Assuming 15-20% moisture content in field-dried biomass, 3700 T/day of input will result in 700 T of water output and 3000

T of biomass. At hydrocarbon content of 9% and oil-free residue of 91%, there will be 2739 T of residue output and 261 T of oil. Estimated capital costs for a plant of this size are listed in Table 6, and operating costs in Table 7.

Buchanan and Otey (1978; SRI, 1979) calculated costs for an extraction plant at 60% of SRI scale, or 607,800 dry T/year (yr) feedstock as opposed to 985,000. Processing consisted of whole plant flaking, single-stage solvent extraction, and residue drying. Plant cost is estimated at $25 million, or $13,600/daily T feedstock, as compared to

$23,600 for base and $14,900

for

optimistic case of SRI (1979). Differences exist primarily in SRI's on-site power plant.

Dry Plant

CYCLOHEXANE (CH)

EXTRACTION

CH

Extract

=

17,000 BTU/lb

high-quality

fuel

4-5% of dry

weight,

polycyclic triterpenoids

(Bagby et

al.,

1981)

Marc

4,

ETOH

extract

=

10,000 BTU/lb

ETHANOL

(ETOH)

EXTRACTION

4,

Bagasse =

5400 BTU/lb

low-quality

fuel

Other

possible uses:

cattle

feed,

pulp

fibre,

chemical feedstock

(Bagby et

al.,

1981)

Figure 2.

Solvent

extraction of Euphorbia lathyris

by

Arizona

Method

(Peoples

and Johnson, 1980).

24

Table 6. Estimated processing capital budget (F)

(million $) (SRI, 1979).

Activity/Material

Storage

Capital (30 days)

Capital Investment

Receiving and Handling

Chopping

Refining

Mixing and Filtration

Drying

Flashing

Utilities

General Service

Total Plant

Land

Organization

Interest

Working Capital

Total

27.0

2.4

3.1

4.8

10.9

8.3

5.1

26.9

9.2

70.7

0.2

3.5

5.4

1.0

80.0

25

Table 7. Estimated processing operation budget (F)

(million $) (SRI, 1979).

Activity/Materials

Materials

Field-dry E. lathyris

Chemicals

Maintenance

Total

Labor

Operating

Supervision

Maintenance

Administrative

Payroll

Total

Water

Fixed Costs

General and Administrative

Taxes and Insurance

Depreciation

Total

Total

$/year

16

.1

1.1

17.2

2.3

.5

.6

.1

.7

.3

.6

1.4

4

1.8

7.2

27.2

26

27

Not included in processing cost calculations are the adaptations necessary for correcting environmental or health effects. Environmental emissions cannot yet be estimated, but should be controllable by baghouses or wet scrubbing (SRI,

1979).

No aqueous wastes are produced since water is returned to the field. Storage and handling precautions may be needed to minimize contact with possible carcinogens. E. lathyris latex is only slightly irritating. Skin studies show the extract to be more irritating than the whole plant material and bagasse. E. lathyris is rated moderate for eye irritation and bagasse is rated mild

(OALS,

1979).

Packaging and storage of the finished extractable would also need to be added to cost estimates.

Transportation to Refinery (G)

Crude would be transported from extraction plants to the refinery. Distances and costs are calculated as discussed in Section D.

Only one refinery is considered in the Chapter

4 examples, due to a minimum requirement of

50,000 barrels/day.

Refining (H)

The actual refining process and marketing are not discussed. This analysis provides a price comparable with that for crude petroleum delivered to the refinery. It is assumed that the refining process is similar for fossil and biomass oil, and that products derived will have comparable values. Products from cracking are shown in Appendix B.

Calvin

(1977b) recommends using hydrocarbon as it comes from the plant as crude oil: refining it, rescuing sterols contained, cracking

28 the rest of the compounds, and then reconstructing the desired chemicals from those products. E. lathyris breaks down into 12.7% crude protein,

7.6% polyphenol fraction, 9.9% oil fraction, and 0.4% polymeric hydrocarbons (Bagby, Buchanan, and Otey, 1981).

In addition to chemical products and marketed fuels, it would be possible to use E. lathyris crude as a blending stock to upgrade petroleum products as their crude quality declines (Calvin, 1978a). Energy feedbacks to cropping, production, and extraction might also be considered, along with the possibility of supplying energy directly to small nearby communities, industry, or power plants.

Although bagasse does not appear attractive as a paper pulp feedstock, burning the material directly as a solid fuel for production of steam and cogeneration of electricity seems attractive economically

(OALS, 1979). It can also be converted to high-quality diesel fuel or gasified into high-energy gas.

Systems Summary

The unified system flowchart for the biofuels industry is presented in Figure 3.

<z

<

,

I

<

= f

4, 2 /

r

nn•P'

. .

29

CHAPTER 3

DESCRIPTION OF MODEL

The goal in modeling Euphorbia lathyris production is to determine the best configuration for the industry to minimize costs of the production process. Cost minimization, rather than maximization of profit or net benefits, is selected as the criterion for optimization for two main reasons. First, the industry as initially described involves excessive costs which must be reduced before production could be considered feasible. One of the primary interests of analysis is in determining the sensitivity of final cost to changes in inputs in order to direct further research activities. Second, benefit and profit modeling require detailed information on revenues and the response of sales to changes in price. Since commercial E. lathyris production would probably not be initiated within the next decade, demand and revenue data would be speculative at best. Biocrude would most likely compete in the mutable petroleum market, and it is felt that market and pricing analysis is beyond the scope of this study. Advantages and drawbacks of this approach are discussed in the concluding chapter.

Total cost is equal to the sum of costs incurred in each segment of production: growing costs, plus harvesting costs, plus costs of transporting and processing. The flow diagram for a biofuels production system is shown in Figures 1 and 3 (Chapter 2). Minimization within this system

30

31 is accomplished with a non-linear transportation-transshipment model which describes the flows of materials in a series of nodes and links. Nodes represent farm sites, extraction facilities, and the refinery. Material flows from one node to another through transshipment links.

Nodes consist of two types: biomass supply nodes and biomass demand nodes. Farm sites are initial sources of biomass, and act only as supply nodes. The refinery constitutes a sink within the model, and acts only as a demand node. Extraction facilities act as demand nodes in receiving raw materials from the farms, and act also as supply nodes in providing extracted biocrude to the refinery. The model is constrained by the necessity to preserve mass; material entering a node or link must equal material leaving it. Handling and processing losses are assumed to be negligible. The flow process is described below.

Farm sites, the initial supply nodes, use agricultural inputs to produce biomass. E. lathyris is planted as would be a traditional crop.

It is cultivated, irrigated, and harvested after a fall-to-spring growing season (as described in Chapter 2). Yields are limited by the amount of acreage planted and the tonnage produced per acre.

After harvesting, field drying, and packaging, the raw material supplied by the farm nodes is transported to extraction plants, which act as demand nodes for raw material feedstock (biomass) and as supply nodes for extracted liquid and bagasse. Economies of scale and plant location determine minimum and maximum size of each node, and the amount of feedstock received and processed by each. The total amount farm-harvested, minus water lost in field drying, must be accounted for in the shipping process and must be received by the processing (extraction) plants.

32

In the extraction process the biomass is separated into solid refuse, or bagasse, and biocrude or the equivalent of crude oil. The amounts of the two different products must sum to equal the same amount harvested, transported, and accepted by the extractors. The materials flow is unchanged, although the two products have differing transportation cost functions from this point on. Bagasse may be recycled in the system to offset some costs of power generation. Biocrude is transported from the extractors to the refinery, a biomass demand node. Although refineries act as biofuel suppliers to the outside market, in this model the refinery is treated as a sink since it is at the point of entry to the refinery that price comparisons are made. The amount of biocrude produced by extraction facilities must equal the amount transported from extractors to refinery, which in turn must equal the amount received by the refinery. Refineries will also require a minimum sustainable level of input and will probably restrict biocrude accepted to a maximum amount, depending upon availability of other crudes and total output demand.

The model determines the cost of biofuel as delivered to the refinery. It is designed to minimize the costs of transportation and extraction by selecting appropriate size(s) and location(s) of extraction facilities, and routing farm produce to each appropriate facility. Objective function parameters representing yields and prices and amounts of inputs are used to determine costs for nodes. Differing scenarios are analyzed to allow for variability in these parameters.

The model is described in three segments. Notation is defined first, in categories of indices, parameters, and variables. The equations

33

and relationships represented by them follow, in sections on objective function and constraints. A description of the computer algorithm used to solve the model cost minimization concludes the chapter.

Notation

Indices

I

=

biomass supply nodes

= {1

to

N

}

N=n

a

+n e

n

a

=

total number of farm sites (initial biomass sources)

n

e

=

total number of extraction plants, acting as both biomass demanders and suppliers

J =

biomass demand nodes

= {n

a

+ 1

to

N + r n

r

=

total number of refineries

Demand and supply nodes may overlap; when flow is entering a given node, that node is considered a demander; and when flow is leaving the same node it is considered as a supplier.

Sources (farms)

=

I s

= {1

to

n

a

I

Extraction plants

=

I

e

= J

e

= {n

a

+ 1

to

N);

I e for flow from the facility

J

e for flow to the facility

Refinery

=

J r

= {N + 1

to

N + nr1

34

Parameters a.

=

acreage at farm

i

(in acres)

t

a

=

tons per acre yield expected (in tons) y.=yieldatfarrni=a.*t a

(in tons)

C

a

=

cost per ton of farming (in dollars per ton)

==

miles from

i

to

j

over standard truck transport route (in miles)

C

t

=

shipping cost rate (in dollars per ton-mile)

R =

extraction rate: percent weight of field-dried tons recovered as hydrocarbon product (in percent)

C

t

= c

t

* R =

cost of shipping post-extraction. Since the extraction process changes the nature and weight of the shipped substance, the cost of shipping is adjusted to reflect the extraction rates after the product leaves the facility (in dollars per ton-mile).

C

s

=

base yearly capital cost of storage facility. Costs are assumed to be linear, since square footage is added in direct proportion to tonnage stored and there are not apparent significant economies of scale. Storage is assumed to occur at the extraction site because of a more likely availability there of capital and facilities (in dollar per ton).

S.

=

percent of shipment tonnage which is stored for each extraction facility. The tonnage stored depends greatly on the distribution of farming geographically and over time, and on the resulting harvest times. If E. lathyris

35 is harvested only once each year, eleven months of storage and preservation may be needed. If harvest times are staggered or double-cropping is accomplished, storage needs are greatly reduced (in percent).

C k

=

generalized capital costs of extraction facilities (in dollars per ton) a

=

scale economy factor for capital costs. For many operations, costs of facilities may be calculated on the basis of costs for existing facilities of different size but similar configuration. Cost of the new facility is calculated as a constant

(C k )

multiplied by cost of the old facility raised to a power factor for plant-capacity scale (a).

C o

=

annual operating costs for extraction facilities (in dollars per ton). For all of the above,

iEI • jEJ s' e

M r

=

refinery capacity

Variables

t.. =

tons shipped from

i

to

j,

for all

iGI

and

jEJ.

lj

Relationships

Objective Function

The objective function, or total cost function, represents the summation of all costs incurred. These are the costs which are to be

minimized, subject to the constraints stated in the following section.

36

The total cost function takes the form:

1.

Cost of farming and harvest (A,

B, C)

+ 2.

Cost of transportation to extraction facility (D)

+ 3.

Cost of storage

(E)

+ 4.

Cost of extraction

(F)

+ 5.

Cost of transportation to refinery (G).

Except for Equation

(4), which contains scale factors for extraction facilities, all portions of the objective function are linear.

Amortization calculations are done prior to model solution, and thus are included in the linear costs.

In component parts, the cost equations are as follows:

Minimize TC (total costs)

=

(1) C E E t..

a jeje iels

13

(2)

(3)

(Cost of farming and harvesting

= farm cost per ton times tons produced.)

+ C E t .EJ .

3 e

E

(m..t..)

13 13

'GI

s

(Cost of transport = sum of each transport route cost

= cost per ton-mile times miles traveled, times tons shipped.)

+C

s

E jeJ e iGI s

(S. t..)

3

13

(4a)

(Cost of storage

= sum of cost per ton times tons stored, expressed as a percent of tons shipped. The annual payment is calculated as a standard accounting annuity within

C

e

.)

37

+ E (C

k

( E t

ii

)

a

) j6je iEI

s

(Capital costs

= the sum of capital costs at each facility, which are determined by facility capacity (tons processed) and economies of scale. Annual payments are determined as annuity within

C

k

.)

(4b)

+C

E o

E t.,

13 j&J e iGI

s

(5)

(Annual operating costs = the sum of costs per ton, or cost per ton times tons processed.)

+ C

t

E iGI e

(m.

t..) jGJ r

(Transportation from extractor to refiner

= sum of transport cost times tons shipped times miles traveled.)

Constraints and Assumptions

Constraints represent basic physical conditions which must be satisfied in order for the model to simulate real conditions. Constraints are listed here according to the process to which they apply: farm, extraction, or refinery. Each constraint is accompanied by a description of the relationship it represents.

38

Farm. Production at a given farm site is equal to the number of acres at that site multiplied by the yield per acre. Production cannot be negative. Production also must equal tons shipped from each site.

y. = a.t > 0 a —

y.

1

E t.

jej

13 e

Total production of the study area is the sum of production of all individual farms. Total production cannot be zero or negative, and will be defined for each scenario.

Extraction. Maximum operating time or capacity in one year is

365 days (d). Facilities are assumed to be operating at near capacity.

A minimum feasible size for facilities has been determined to be 1000 T/d for 292 d, or 292,000 T/yr (see Chapter 2). Scale factors determine size efficiency beyond that, up to 3000 T/d or 1,200,000 T/yr. At that point equipment would be duplicated rather than enlarged, counteracting any benefits which would occur from increased size. One million tons per year is therefore the maximum size allowed for any extractor. Extraction facility capacity is determined for each site within the scenario specifications, as is total extraction facility intake, which equals the sum of intake of all individual facilities.

Total input of all extraction facilities must equal the sum of output of all extraction facilities. Output actually emerges in two forms, as liquid fuel feedstock and as residual solid bagasse. The two will have differing transport cost functions. Extraction plant operating

39

cost calculations are made with the assumption that bagasse is recycled for interior power generation; therefore, costs of transportation and returns from sales are not calculated. These factors could be added to the model through scenario changes if necessary.

E t.. =

E t.

iCI

13

kCJ s

r

3-

jCI

e

= J

e

The above equation ensures that extraction input equals extraction output by assuring that tons shipped to each facility equal tons shipped from that facility, and that total tons to extractors equal total tons from extractors.

Refinery. Total extraction facility product constitutes the total of material transported to the refinery, and must equal refinery input.

iCI

E t..

=M.

e

=J

e

13

3`jr

When taken together, the constraints insure that total farm output equals the amount shipped to extractors, the amount shipped from extractors, and the amount shipped to and received by refineries.

The constraints may be summarized in matrix form (see Figure

4).

All rows and columns must total. This assures that all farm produce is shipped (sum of row

S

1

= y

l

)

and that all produce reaches an extractor

(sum of column E l

=

e l ). Produce shipped to each extractor must equal produce shipped from that extractor (sum of column

E

1

=

sum of row

E

1

=

e l ). The sum of extractor capacities in the total row equals total

E

1

E

2

S

1

S

2 t

11 t

21

E

1

[t31] t t t

12

22

32

E

2 t

41

[ t42]

Total e

1

e

2

R1 t

13 t

23 t

33 t

43

T

Total

Y1

Y2

} e

1 e

2

}

X

Figure 4. Constraint matrix.

40

41 produce, as does amount received by refineries (sum of column R

1

).

Bracketed figures represent self-shipment.

Solution Program

The algorithm and code used to derive a model solution were developed by Rao and Shaftel (1980). The program is a combination of the primal transportation method (Charnes and Cooper, 1961; Srinivasan and Thompson, 1969) and the convex simplex method (Zangwill, 1969).

The standard transportation-transshipment problem is one of programming commodity movements to incur as little transport cost as possible; in other words, to find a schedule of shipments (flows) from a set of in supply nodes (farms and extraction facilities) to a set of n demand nodes (extraction and refining facilities) which will minimize cost. Minimization must occur within a series of constraints, such as nonnegativity and the mass balance conservation conditions listed earlier.

The transportation-transshipment formulation is generally solved through linear programming. In this case the need to account for economies of scale in extraction capital costs requires extension of the linear model. The non-linear transportation-transshipment problem as formulated by Rao and Shaftel (1980) is as follows.

Minimize the objective function: m n

E E c. x.

i=l i=l

in n

+ E d

7 7

(X..) t .

13 t=1

.

1=1 3=1 ht..

(linear)

(polynomial)

42

in

n

a.

+ E D.[ E

K..

x•• ] 1

1.1

1

j=1

13 i

(3.

+ E S.[ E

1..

x..]

3

j=1

3

i=1

13 13

1

Subject to the constraints:

Z x.. =

a. for j

= 1 . . .

j=1

13

E x.. =

b. for

j = 1 . . . n i=1

13

3 x..

>0 for all

— i +

i

E

a.

= E

b.

1

j

(supply and demand interdependencies)

, and all a.,

13.,

K..,

1..,

and

3 13 13

> 0.

-13 —

When the objective function for our problem is restated to fit the above format, there is no corresponding polynomial. The linear portion is reduced to

E e

Et.

s

[C +CEM.. +C+CE

a

t

ej 13 0

ES.] s 4ej iej

e e s

+cE

t

jEJ

r

Em..t..

iEI

13 13

e

It is possible to include the nonlinear annuity calculation because the total cost varies linearly with changes in

shippage.

The nonlinear term becomes

43

E

(C

k

jeJ

e

( E t..) ij iei

s a

)

For a detailed description of the solution technique and computer algorithm, see Rao and Shaftel (1980). An example of program input and output as used in this case study is presented in Chapter 4.

CHAPTER4

ARIZONA CASE STUDY

The model developed in Chapter

3 is used to evaluate a

Euphorbia lathyris industry in the state of Arizona. The primary goal is to identify combinations of parameters which will result in a feasible industry. A generalized mapping of factors affecting E. lathyris growth is used to estimate distribution of acreage, and along with the budgets presented in Chapter

2 provides the foundation scenario for industry analysis. A series of simulations allows examination of changes in industry feasibility with variation in yields, transportation network, costs of inputs, and other parameters. This case study is presented in the following manner: background of the Arizona agriculture problem, baseline conditions and model input, the various simulations, and summary of results.

Biofuels and Arizona Agriculture

The growth of fuel crops in arid or semiarid regions is suggested for several reasons. First, plants' ability to produce hydrocarbons may increase with levels of solar radiation (Johnson and Hinman, 1980;

Hinman et al.,

1980).

Insolation is highest in the desert regions of

Africa, Mexico, and the Southwest United States. Arizona, New Mexico,

Texas, and Southern California lend themselves especially well to solar

44

45 energy collection, including solar farms. Here, insolation averages

250 watts/square meter (w/m

2

), as opposed to 200 w/m 2 in Florida (Farm

Chemicals, 1979). The combination of solar energy and long growing season in Arizona gives rise to average agricultural productivity exceeding that of the nation in general, and double for many crops (Valley National Bank of Arizona (VNB), 1981; McLaughlin and Hoffman, 1980).

Yet much land remains uncultivated due to extreme temperatures and lack of water. Of Arizona's 74 million acres (30 million hectares), only 1 million acres (1/2 million hectares) are currently under cultivation (Johnson and Hinman, 1980). This combination of high solar radiation and large areas of unused land is potentially advantageous for energy farming. As discussed in Chapter 5, extensive acreage will be needed to make appreciable amounts of fuel. If energy production could be developed on unused acreage, it would avoid driving up the prices of arable land and other agricultural resources.

Water for agricultural use is scarce in Arizona. In desert regions precipitation is slight and extremely variable. Rates of evaporation are high; available moisture is low and often of poor quality.

Air and soil temperatures reach extremes and vary widely, both diurnally and seasonally. Soil may contain an excess of salts and little humus.

The regions of highest solar activity are therefore not those of highest natural vegetable productivity; large amounts of land are available because conventional agriculture is limited. The most important restraining factors are the finite supply of fresh irrigation water and the problem of disposing of saline water to avoid high soil sodium

46 concentrations. Thus, both botanic and economic viability of farming in

Arizona will depend on plants with the ability to produce economic benefits using brackish groundwater or agricultural or industrial effluent.

Plants also must be developed for high water-use efficiency and general tolerance of environmental stress.

There is inadequate surface water to supply Arizona's needs, so groundwater pumpage is a critical water source. The state's current consumption exceeds replenishment by over two million acre-feet annually.

Water tables in many areas are dropping dramatically. Agriculture uses

90% of groundwater pumped in Arizona. Although the Central Arizona Project (CAP) may temporarily alleviate the problem for urban users and some existing farms, it is not designed to support new irrigation. It is estimated that total overdraft will be cut initially by only 50%, and less as demand grows (VNB, 1981). Agriculture, as the largest user and the one least able to tolerate higher prices, will be most affected by rising costs. Even small increases in water costs will reduce or even eliminate cropping in some agricultural areas, while other users will be curtailed only slightly (Kelso, Martin and Mack, 1973). Limits on future agricultural production are likely to stem from continuing increases in irrigation and energy costs, competition from urban and industrial uses, and institutional controls on the amount of water allocated to agricultural use. Declines in cropped acreage are projected through the year 2020 (Foster, Rawles and Karpiscak, 1980).

Curtailment of high water-users and replacement by higher-value low-water uses may lead to a restructuring of agriculture. The steady

47 expansion of population and accompanying land, water, and energy demands will lead to increased land values and costs, possibly resulting in extension of cropping to less suitable areas. Costs are likely to rise further in this case, and productivity decline. An important benefit of biofuels production, if irrigation and fresh water needs are lower and plants are more adapted to arid regions, is that it may be possible to farm in harsh conditions or on acreage which otherwise would be abandoned because of prohibitive costs. Concern that ploughing would aggravate erosion or other land degradation could be counteracted by an increase in foliage that would stabilize denuded regions and mediate the microclimatic effects of plant removal due to cities, roads, or mining.

Added plant materials promote rebalance in the albedo and the carbon dioxide (CO

2

) and oxygen (0

2

) content of the atmosphere. CO

2 is absorbed by plants during photosynthesis, and 0 2 is discharged. Although burning of biofuels subsequently releases CO 2 , it is recycled from the atmosphere rather than being introduced from fossil material. An energy cropping industry thus might provide both economic and environmental benefits to the state of Arizona.

Delineation of Study Area and Model Input

Study Area Mapping

To delineate areas in which farming operations might be located, possible biofuels growth areas in Arizona are mapped (see Appendix C).

The purpose is to determine whether the arid and semiarid regions of

Arizona would provide biologically suitable or economically feasible conditions for an E.

lathyris

industry.

48

Several criteria are considered in choosing possible cultivation sites.

1.

Plant's natural habitat: geography, climate, rainfall, and soil quality.

2.

Air and soil temperature ranges, and insolation received.

3.

Soil quality factors: nutrients, salinity, texture, slope, and drainage.

4.

Land use and ownership.

5.

Availability of and access to support facilities: transportation, processing plants, refineries, storage.

6.

Water availability: amounts and variability of precipitation, and proximity and quality of irrigation.

Each set of criteria is outlined on maps of Arizona. Discussion of the factors included in each category and maps of areas outlined as acceptable are presented in Appendix C. Resulting maps are overlain to determine regions most suitable for cultivation. Areas under consideration meet all of the following conditions:

• 500 langleys sunshine per year, with 80% of possible annual sunshine received (Criteria 1 and

2).

• 180 freeze-free days, and average temperature greater than

32

°

F

(0

°

C) (Criteria 1 and

2).

• Fall (September and October) germination soil temperatures between 16 and 26

°

C (61 and 79

°

F)

(Criterion

3).

• Soils without excess nitrogen, and with good drainage, slope less than

17

°

, and tillable composition (Criteria

1 and

3).

In addition, land availability is constrained by public owner-

49 ship and existing use. Whenever possible, acreage near current cropping or on idle farms is considered for existing infrastructure (Criteria 4 and 5). Since no areas received enough rain to dry-farm, irrigation facilities and water availability, especially saline water, are important considerations (Criterion 6). Possible regions are correlated with more detailed use maps (see Arizona Water Commission, 1977), grossly outlined, and available acreage within those curves estimated as a baseline farming industry configuration for use in the simulation model. Figure 5 is an overlay of all the conditions listed above.

Figure 6 shows the network of crop areas and transportation linkages used for the industry baseline. Acres farmed are calculated by county, as shown in Table 8. Individual farms are not delineated since costs are assumed to be linear with regard to farm size. Each designated area is also associated with a nearby population area in order to facilitate distance calculations and extractor locations (see Table 8).

Table 8. Farm area acreage of baseline network.

Farm Designation

4

5

6

2

1

3

Total

County

Cochise 1

Cochise 2

Santa Cruz

Pima/Pinal

Maricopa

Yuma

Acres

80,000

20,000

64,000

20,000

10,000

100,000

294,000

Site

Wilcox

Bisbee

Nogales

Tucson

Phoenix

Yuma

Figure 5. Estimate of acreage suitable for E. lathyris cultivation.

50

6

Yuma 60 mi.

Mobile

(Refinery)

PHOENIX

k SO mi.

Gila ohawk mi. Bend

60 mi.

Casa Grande isbee

Nogales

Figure 6. Baseline network of farm sites and transportation routes used in all simulations. Numbers refer to farm designations in Table 8.

51

52

Assumptions and Baseline Conditions

Unless stated otherwise, the following assumptions hold throughout the simulations in this chapter: one crop is produced annually in a fall-to-spring season. A single harvest occurs per farm per year, although farm timing may vary. Farm facilities and irrigation system are existing, and equipment was purchased previously. Either relatively fresh irrigation water is used, or saline/effluent water requires no major additional capital investment. Both biomass and biocrude are transported by truck on existing roads. The Arizona Method (seeFigure2,

Chapter 2) of solvent extraction is used. Extraction facilities are built to the tonnage capacity specified by the model solution, with minimal slack. Capital expenditures are amortized over 20 years. The interest rate is 16%. A single refinery located at Mobile accepts all extractor output, on a competitive basis with crude petroleum, without modification to its facility. These assumptions are not necessarily the most realistic ones, but represent those upon which most available data are based.

Parameters defining costs for each portion of the Objective

Function are changed for each simulation; however, all calculations are based on the initial conditions defined in Chapter 2. A summary of the baseline costs from Chapter 2 and Appendices A and B follows.

C a

= farming costs = $450/acre, yielding 4T/ac = $112.50

C t

= transport costs = $.10/ton-mile (t-mi)

C' t

= post-extraction transport costs = $.10/t-mi * .07

extraction rate

= $.007 t/mi

C o

= extractor operating costs = $34.50/T (based on

$34,000,000 for 985,000 T).

C s

= storage costs, at 25% of extractor capacity (approximately

3 months), $27,000,000 amortized over 20 years at 16% =

15.00/T

53

Summing these costs produces total linear costs per ton shipped.

Rounded to the nearest dollar total linear cost = $162. + .11 m...

C k

= extractor capital costs, amortized over 20 years at 16% =

$47.00/T a = scale factor (see Chapter 3) for extractor capital costs =

0.8

The value for a was chosen by examining processes similar to those used in E. lathyris extraction (Peters and Timmerhaus, 1980). A range of reasonable scales was selected as representative. Sensitivity analysis tested these factors and indicates a relatively small change in capital costs and payments for a change in a from 0.7 to 0.9. The solution matrix, or configuration of shipments, is identical for all three scale factors from 0.7 to 0.9. Variations of the scale factor within this range therefore do not alter the optimal node and transportation network. The least cost value of the nonlinear portion of the objective function does vary slightly, reflecting differences in the capital cost of extraction facilities. Total annual costs ranged from $217,900,000 for a = 0.7 to $216,600,000 for a = 0.9. Costs per barrel were $367.00

for a = 0.7, $366.00 for a

= 0.8, and $365.00 for a = 0.9. The differ-

54 ences represent a 0.4 and 0.2 percent variation from scales of 0.7 and

0.9, respectively, to 0.8. Due to this small variation, the median value of

0.8 was used in case study scenarios.

Given the scale factor of a

= 0.8, the base capital cost, C k , is calculated by fitting a curve to the known cost points. SRI (1979) calculates a capital cost of $80.8 million for an extraction facility processing 985,000 T/yr (3000 T/d). Buchanan and Otey (1978) eliminate the on-site power plant and calculate costs for 607,800 T/yr (approximately

2000 T/d).

With adjustments for on-site utilities generation and inflation, capital costs would be nearly $52 million dollars. The resulting base capital cost per ton, adjusted for mortgate amortization, is

$47.25.

Figure 7 shows the resulting curve. Table 9 lists the baseline cost components and their percentages of input cost per ton.

Program Input

The objective function (see p. 43) with these baseline costs inserted, rounded to the nearest dollar, becomes: or

E iCI s jGJ e

(112.50 t..+ .1

13 m. .t. . + 15 t. . + 35 t. .)

13 13 13 13

+ (

E iGI

E 1

.01 m.

.t..

(47 (Et..)

13 13 13 e jEJ r jej e

.8

)

E E iGI jGJ s e

(162.50 t.. + .1 t..m..) + (47 (Et..) .8 ,

13

13 13 13

) j ej e

(

E iGI e

E

).01m..t..

13 13

jEJ r

E iCI s

n

_

NJ

— 0

1

' u

0

!II

! 1/\.!

S

I cf)

Lc) v

- re)

I o

-o

Lr)

—•0 re)

— o

0

55

56

Table

9.

Objective function cost components as percent of input costs.

Parameter

Represents

Cost/T ($)

% of Total

Input Costs

C a

C t

C .

1

Farming costs at

41/ac

Transport to extractor

(0.10/mi * 100 mi)

Transport to refiner

(0.007/mi * 100 mi)

C o

C s

Extractor operating costs

C

K

Extractor storage costs

Subtotal Total linear costs

(162 + mi for 100 miles)

Extractor capital costs

Total Omitting scale factor

112.50

10.00

0.70

34.50

15.00

172.70

47.00

219.70

51

5

< 1

16

7

79

21

100

57

Input for the program is accomplished through a series of matrices. The linear cost coefficients, calculated for each shipment route using the linear portion of the objective function, are stored in

MATRIX I (Figure 8a). Also included are the rim conditions for the constraints. These are the capacities of suppliers and demanders. The linear cost coefficients and rim conditions for a problem of m supply nodes by n demand nodes (m x n) are stored in a (m + 2) x ea

+ 2) matrix.

This allows one row and column each for rim conditions, and one each for dummy nodes. Dummy sources and sinks provide a slack variable, allowing free transshipment from any node to itself.

Input for the nonlinear portion of the objective function is accomplished in two following matrices with different format. Since supply-and-demand interdependency calculations are combined into a single equation, one of these matrices contains this portion of the objective function (Figure 8b) and the other matrix is zero. For each nonlinear term, i

=

1

. . .

m, the coefficients k.., j

1 . . . n, and the ij exponents ai are stored. The data for the m rows and the exponent are stored in a m x en + 1) matrix (m and n are expanded to include the dummy nodes). The constants

C K are stored in an additional column.

Output (Figure 9) consists of the value of terms for each route in the solution; i.e., the tonnage shipped on each segment. The total cost and linear cost are also output.

Figure

8.

Input matrices for sample problem.

a.

Matrix I. Linear costs and rim conditions. Cost of shipment from supply node

3 to demand node

2 is

$168/T.

Where shipment is not allowed, costs are assigned

$99999.

Demand capacity for extraction

(Columns

1 and

2) must equal supply capacity for extraction (rows

7 and

8).

Column

3 is total.

b.

Matrix II. Nonlinear objective function.

Supply

1

Nodes

2

8

9

Demand

Capacity

5

6

7

3

4

Demand

Nodes

1

182

999

O

240

182

180

174

162

179

0 a

.

2

170

174

181

999

170

168

162

936

0

o

3

999

999

999

999

999

999

.29

.71

0

1176

0

0

0

0

X

0

0

0

4

0

0

Supply

Capacity

320

40

400

240

936

0

X

80

256

80

58

Column

1

5

4

2

3

Row

1

1

1

1

1

1

1

1

1

1

1

2

8

9

6

7

1

0

0

1

1

0

0

1

1st

n

columns

are

term

coefficients

(K.. = 1 for

13

all

initial

terms, except

self-

shipment)

1

1

1

1

1

1

1

1

1

3

1

1

1

4

1

1

.8

.8

5

.8

.8

.8

6 7

0

0

0

8

0

0

0

0

9

0

0

0

1

0

0

.8

.8

.8

0

0

0

0

47

47

1

.8

0 0

Exponent Calculated

Con- Constant

(scale terms

and stant

cost

factor)

sums. Not cost

multipliinput

, added

cand

b.

59

Supply

Nodes

1

3

2

Demand

Nodes

1 2

320

80

256

80

5

6

7

8

40

200

200

3

240

936

Figure

9.

Output matrix. -- Matrix III: resulting terms, or solution tonnage shipments.

60

61

Simulations

The case study consisted of a series of simulations to assess the feasibility of an E. lathyris industry under differing cost and production conditions. Using a baseline production cost scenario and a determined optimum extractor layout, values of parameters representing costs, yield, and percent crude recovered are varied, and the industry simulated to evaluate total cost and shipment response. Six simulations, each representing change in a specific parameter, were conducted.

1.

Variations in the size and locations of extraction facilities, to determine the optimum network layout for the industry

(D, F, G).

2.

Changes in cropping costs, C a , to determine industry reponse to reduction in input price or amount (A, B, C).

3. Changes in yield, y i and R (A, B, C, F).

4

-Changesinstorageneeded

,

S-

(E )

3

5.

Changes in extraction costs, to represent variations in input prices or amounts, the cost of capital, or technological change.

Parameters manipulated are Ck , C o , i, and a (F).

6.

Change in transportation costs C . and C't (D, G).

At least one scenario in each simulation represents an alteration in a combination of parameters. Each simulation is summarized, accompanied by tabular presentation of the results. A discussion of the energy balance implications and summarization of results concludes the chapter.

62

Simulation 1:

Extractor Layout

The initial simulation was run to determine the most feasible extractor layout. The first set of scenarios uses exclusively the baseline conditions as defined in the section on assumptions. Farm acreage and the transportation network linking farms, extractors, and refinery are outlined under mapping. The location and sizes of extraction facilities within the network are combined into different permutations to allow identification of the optimum configuration.

Scenarios are postulated for differing combinations of numbers and locations of extraction plants. Production in the initial simulation might support up to three extractors, so scenarios vary from one facility to three. One large facility collecting from the entire area could be located at each node, or additionally at a central location such as Casa Grande. Two facilities would be located at each tail end or in central and end locations. Three facilities are assumed to be fairly equally distributed in the region. When more than one extractor is in operation simultaneously, runs are conducted with the assumption of slack allowable in the system; each facility is able to accept the total product, with excess allotted to self-shipment. In this manner each facility is assigned its optimum allocation. Results were verified by subsequent runs of set extractor size. Locations of extractors and corresponding costs are shown in Table

10.

The rankings of costs listed in Table

10 indicate that in the current simulation transportation costs outweigh economies of scale.

63

Table 10. Simulation 1:

Costs corresponding to differing extraction facility distribution.

Scenario

Number

Number of

Extractors

Locations and

Tonnage/Yr

($ million)

Total Cost

1

2

3

1

1

1

Phoenix

1,176,000

Tucson

1,176,000

Yuma

1,176,000

224.8

217.8

229.4

4 1

CasaGrande1,176,000 221.7

5

6

7

8

9

1

1

2

2

2

Nogales

1,176,000

Bisbee

1,176,000

Tucson

736,000

&

Phoenix

440,000

Tucson

776,000

Yuma

400,000

Bisbee

736,000

Phoenix

440,000

10

(3)

Tucson

776,000

Phoenix

0

Yuma

400,000

10a optimum when forced to

3 facilities

Tucson

576,000

Phoenix

224,000

Yuma

376,000

11 lla optimum for

3 forced facilities

12

13

(3)

3

3

Yuma

400,000

Nogales

736,000

Casa Grande

0

Yuma

400,000

Nogales

500,000

Casa Grande

276,000

Willcox

400,000

Yuma

400,000

Nogales

376,000

Tucson

520,000

Nogales

256,000

Yuma

400,000

224.7

222.6

218.3

212.2

217.1

212.2

215.2

214.1

214.9

210.4

212.

$

Cost/Barrel

434.

421.

443.

433.

438.

430.

422.

410.

420

410.

416.

414.

415.

407.

410.

64

Scenarios with two and three extractors consistently are preferable to scenarios with only one extractor. This correlates with calculations of the most efficient radius of production for a single extractor. Radius computations are based on the assumption of an uninterrupted transportation surface covering the cropping area, so that produce from each point is shipped discretely to a processor. At baseline conditions, the marginal cost of shipping an additional unit to a new facility equals the marginal cost of processing an additional unit at the original facility when extractors are situated at 10-1/2 mile intervals in cropped regions.

In cases where one of three facilities falls below the minimum size, it is better to redistribute produce to only two extractors than to three more equal in size, indicating some scale benefits. Extraction facility location close to cropped areas appears to be more important than location centralized among the differing areas.

The most effective layout, with extractors in Wilcox, Yuma and

Nogales (Figure 10) results in a cost of over S400/barrel. This is obviously not competitive with petroleum.

1

This configuration continues to be used in the following simulations. Since transportation costs are expected to become relatively more important as other costs are decreased, less efficient scenarios are not retested.

1. Products from E. lathyris crude are expected to be substitutes for those of petroleum, so prices for the two are compared. Costs of biocrude are termed "feasible" when in the range of predicted petroleum prices (see p. 81).

65

66

Simulation 2: Cropping Costs

Simulation 2 compares changes in various cropping costs. It should be noted that the calculations could represent either a decrease in the price of an input or a decrease in the amount of the input required, since both actions would have the same effect on the resulting figures entered into the cost matrices. All costs are in real 1979-1980 dollars.

Using the optimum Wilcox-Yuma-Nogales extraction combination from Simulation 1, input costs of farming are varied from the baseline.

Since irrigation is always necessary and is the major cost factor in arid regions, water use estimates vary the most. Both amounts and quality of water are reduced. Other practices curtailed or eliminated include fertilization, tillage, and planting and harvesting. The latter two factors are modified to simulate development of a perennial crop.

Results are summarized in Table 11.

The most significant total cost reduction in the simulation, produced by halving all cropping costs, results in only a 28% reduction in final cost to $293 per barrel. Changes in cropping cost alone, and particularly in one element of cropping such as irrigation, do not have enough of an effect on the total cost to bring production into a feasible range.

Aside from cutting all costs in half, the greatest reduction

(nearly 20%) results from development of a perennial crop. Elimination of post-establishment irrigation reduces cost per barrel less than 10%.

Use of saline water or reduction of water use do not in themselves

• -

Nr) vp co

CV Hi

N.

CD

Ln f---

0)

CO

4:I-

CO

V)

0)

I---•

V)

CV

N-

V)

0-) r--

V)

471-

CO tr)

V)

0-)

CV

1-4

CV

N

Lr)

CD

OA

%.0

CO

‘..0

01 0)

Hi

C n

1

01 k.0

VD

CO 4-I

C)

Ii)

1

-

1

'71'

C)

N-

4-)

0.)

$-1

67

NI

V)

•zzr N.0 N-

CO

C)

CCI

cd

L.r)

LI)

1-1 CNI

produce significant gains. Even in combination, increases in yield would be necessary to reduce costs sufficiently to compete with petroleum in the near future.

68

Simulation 3: Yield

The third simulation is designed to illuminate the response of final cost to changes in yield per acre in farming and to percentage of crude recovered in extraction. Yield is allowed to increase without more inputs to simulate improvements which might be possible through crop domestication, plant breeding programs, spacing corrections, etc.

Scenarios show variation from 8T/ac to 40T/ac, 7% extractables to 50%, and combinations within the range of factors. Results are shown in

Table 12.

Doubling the yield from baseline 4T/ac to 8T/ac reduces cost per barrel to $275. Raising extractables to 20% breaks below $200/barrel. Not until a combined rate of 8T/ac and 50% crude does cost drop below $100/b1 to $81. Sixteen T/ac and 30% extracables results in

$51/b1 total cost, well within a competitive range for projected petroleum crude prices by the year 2000 (United States Department of Energy,

Energy Information Administration (ETA), 1980). By halving cropping costs in addition, cost of a barrel is further reduced to $44.

Changing either the percentage extracted or tons per acre produced significantly impacts final costs. Neither alone reduced cost to feasible levels, but a combination of the two does. Amount of extractables seems to be more important. Eight tons per acre is probably possible at this time, and sixteen tons per acre may result under

Table 12. Simulation 3: Response of total cost to changes in farm and extraction yield.

Scenario

Number Description

2

1

3

4

5

6

7

8T/ac; 7% extractables

8T/ac; 20% extractables

8T/ac; 50%; 3 or 4 extractors

16T/ac;

7%; 4 extractors

16T/ac; 30%; 4 extractors

40T/ac; 7%; 42 extractors

16T/ac; 30%;

1/2 crop costs

$

Total

(10

284.6

289.2

298.9

438.7

453.8

3,503.

387.9

6

)

$/Barrel

275.

195.50

81.

212.

51.

169.

44.

69

70 improved conditions and commercial breeding. Twenty to 30% extractables is not unreasonable. This combination results in more tonnage than can be processed at three facilities; a fourth extractor must be added to scenarios involving more than 3,600,000 T/yr. The facility is located at Tucson, proximate to the largest remaining cropped area and central to the others without contiguous extractors (Figure 11).

Simulation 4: Harvest Strategy,

Reflected in Storage Cost

Scenarios in Simulation 4 represent changes in harvest timing.

If the extractor is to run year-round as assumed, produce would have to be received continuously by the extractor or stored there to fill in during slack harvest periods.

It is always assumed that extraction facilities are constructed so that they are operated to capacity year-round. Changes in the amount of raw material received by the facility at a given time (as opposed to the total annual amount) therefore change storage requirements rather than extractor size and capital requirements. Storage is expressed as a percentage of total amount processed. If a perennial crop is developed, double cropping is accomplished, or harvest staggered throughout the year, a minimum of storage would be needed. If a single crop is harvested once annually, up to 11 months of storage would be necessary. The baseline assumes 90 days of storage (25%), a relatively small amount.

Since storage costs are likely to increase from the baseline, this analysis begins with lower cost and higher yield assumptions. Scenario 4:1 is identical to Scenario 3:7; 16 T/ac, 30% extractables, 1/2

(.1

v) a)

71

72 cropping costs, and 25% storage. Doubling storage to 6 months (50%) increases costs 18% to $52/b1. Nine months of storage further increases costs, to $59.50/b1, and eleven months raises costs to $65/b1 (see

Table 13). Each additional month of storage adds 5% to total costs.

Simulation 5: Extraction Costs

The fifth simulation measures the effects of changes in costs of each portion of the extraction process. Lowering operating costs by

50% results in a 10% decrease in final cost per barrel. In combination, reduction in operating, capital, and interest rates results in a 14% decrease in barrel cost. Changes in interest rate and economies of scale have no significant effect. This is probably due to the long time period over which expenditures are spread, and the small proportion of total costs involved.

With the 14% decrease,the final barrel cost is $350. In itself, modification in extraction would not be sufficient to make E. lathyris crude competitive with petroleum (see Table 14). At higher yield levels, the effect might be more pronounced. With 16T/ac and 20% extractables, halving capital and operating expenses for the extractor and dropping interest rates to 10% result in $45.50/b1 final cost. This is comparable to the rate produced at baseline capital and operating costs with

30% extractables. A technological change could compensate to an extent for lesser improvement in yield.

Table 13. Simulation 4: Variation in total cost with changes in harvest timing and storage.

Scenario

Number

1

3

4

2

Description

16 T/ac; 30% extractables;

1/2 crop costs; 4 extractors;

25% storage

50% (6 months)

75% (9 months)

92% (11 months)

Total

$ (10

457.3

527.8

574.9

6

387.9

)

$/Barrel

44.

52.

59.5

65.

73

74

Table 14.

Simulation 5: response of total cost to extraction cost change.

Scenario

Number Description

5

3

4

1

2

6

7

Baseline as Simulation 1; 1/2 C o

1/2 C

K

; 1/2 C o

1/2 C o ; 10% interest (i)

1/2 C o ; 1/2 C

K

10% i

1/2 C o ; a = .5

1/2 C o ;

1/2 C

K

; 10% i; a = .5

1/2 costs: C a , CK, C o ;

10% i;

16T/ac; 20% extractables

Total

$ (10 6 ) $/Barrel %-1,

190.4

183.2

186.7

181.7

189.9

182.4

268.9

368.

354.

361.

351.

367.

352.5

45.5

13

89

14

10

10

13

11

Simulation 6: Transportation

The final simulation consists of one scenario with the ton-mile

75 transportation charge cut in half, to represent technological change.

With baseline assumptions, total industry costs were $209 million, or

$404./barrel. This is a reduction of less than 1% in cost/bl. In spite of the small change in cost, the lessening of transportation rates does allow a slight modification of the industry layout. Production from the

Maricopa County farm is shipped to Yuma instead of Nogales. This provides a savings in nonlinear costs by capitalizing on economies of scale. As transportation costs decrease, scale economies become more significant; however, the halving of ton-mile charges causes only one rerouting of a relatively small tonnage. With increased yields and greater shipments it is unlikely that lowered transport costs would be outweighed by scale savings.

Comparative Energy Pricing

Euphorbia lathyris development and production is a long-term contributor to energy supplies rather than an immediate solution. Commercial production would not be likely to occur before the year 2000. It would be expected that real energy prices at that time would be higher than those used for input prices in the case study. Costs of energy input will rise in proportion to prices of petroleum and electricity. The final biocrude product cost will in turn rise. To estimate the production cost of biocrude at $50/b1 does not, therefore, imply that it will be competitive when the price of petroleum crude reaches $50/b1.

76

Since energy inputs are only a portion of biocrude costs, the final product cost change will be less than energy input cost changes.

If the proportion of energy costs input is small enough, there is a level at which the price of a barrel of petroleum crude will overtake the cost of a barrel of E. lathyris crude. The model will calculate the total cost per barrel of biocrude at higher energy input costs, but will not automatically vary input values to derive the equalization level.

Calculating the actual price at which biocrude costs would equal those for petroleum crude involves isolating the portions of production costs which are energy-related and quantifying their contribution to the total. To simplify analysis, inputs are considered in petroleum Btu equivalents. These factors would obviously differ with changes in the production process, such as amount of irrigation or distance transported, so that the relationships must be recalculated for each scenario.

Only one example of the calculations is presented, and the results for other scenarios are summarized.

Figures on energy requirements (see Appendix D: McLaughlin and

Hoffman, 1980; Calvin, 1980; SRI, 1979) and corresponding production costs (Wright, pers. comm., 1981; see also Chapter 2) are compared and analyzed to estimate the portion composed of energy costs. The initial analysis is performed with baseline costs and assumptions, but the number of barrels per acre is increased to bring final cost per barrel into the feasible range. At $172/T or $688/ac (4T/ac) approximately 12% of

total cost, or

$85., is attributed to energy. At

10 hi/ac, final cost is

$69.00/b1, with

$8.50/b1 energy cost. Approximately

$60/b1 is the basic cost, which does not vary directly with changes in fuel prices.

The total cost consists of this base cost plus the energy factor, or in this case, the portion dependent on the price of oil.

77

60 + n(P

0

) = 68.80

At the current price of P o

= $35/b1

60 + n(35) = 68.80

n = .24

At the break-even point, the price of petroleum crude equals the price of E. lathyris

biocrude

P o

= P el

60 + .24 P el

= P el

60 = P el

(1 - .24) = .76 P el

P el

= $79./b1 = P o

World oil prices projected to the year

2000

(in constant

1979 dollars) range from a low of

$48/b1 through a midpoint of

$66/b1 to a high of

$87/b1 (EIA, 1981).

Eight of our scenarios result in costs within those boundaries, ranging from

$44/b1 to

$81/b1. The initial scenario costs, energy factor n and final equalization prices are presented in Table

15.

In only one case, scenario 3:3, is the final price

Scenario

Number

3:3

3:5

3:7

4:1 & 2

4:3

4:4

5:7

example

Table

15.

Equalization levels for

"competitive" scenarios at higher energy input price levels.

C

1

a

81.

51.

44.

59.50

65.

45.50

69.

n

.14

.15

.06

.24

.2

.08

.05

P

b

63.

71.

46.

79.

92.50

55.

46.

a.

Cost as initially calculated in the scenario in

$/b1.

b.

Price of crude petroleum oil and E. lathyris

biocrude

at equalization in

$

1

b1.

78

79 pushed beyond the estimated limits. As yields increase, the energy factor becomes less important.

Summary of Results

Testing of sensitivity to changes in input costs shows that no single element will reduce final cost sufficiently from the baseline to enable competitive production in the near future. Increases in yield do have a significant effect, and allow production in the feasible price range. With improved yields, variations in input costs could further contribute to lowering costs, although combinations of input modifications would produce the most effective results.

Transportation costs outweigh economies of scale in extraction, making extractor location close to cropping areas more efficient than centralized location. The one case in which consolidation proves desirable occurs when an extractor does not meet minimum size requirements; distributing its produce to other extractors is preferable to diverting tonnage to fill it. Even the most efficient scenario results in costs of over

$400/b1, not competitive with petroleum crude.

Changes in cropping costs do not reduce this cost sufficiently to bring production into the feasible range. The lowest cost scenario, resulting from halving cropping expenses, reduces total costs by 28%, to

$293/b1.

Other production costs are important contributors to the overall layout, so that improvements in any one area are diluted. If yields increase, smaller changes can have more effect, but the results are further diluted by the per acre division of cropping costs as opposed to extraction costs.

80

Aside from an overall halving of cropping costs, the most effective reduction is obtained by perennial cropping. Final cost is reduced

20% by eliminating planting and modifying harvest. Use of saline rather than fresh water does not in itself adequately reduce cost. Coupling saline water with water use reduction, in particular elimination of postestablishment irrigation, and yield improvements, could result in a feasible production scenario.

Improvements in yield proved to be the major factor necessary to bring costs into the competitive range. The percentage of extractables recovered appears to be as important as the tonnage per acre produced.

A combined improvement, in both extractables and tonnage, is necessary to produce in the feasible price range. Eight T/ac with 50% extract, or

16 T/ac with 30% extract both reduce costs to within $50 to $80/b1 without other cost or input modifications. Additional extraction facilities are necessary to handle the increased tonnage.

Storage costs, reflecting changes in harvest strategy, are more likely to increase from the baseline costs, and so are analyzed from the increased yield scenario. Each subsequent month of storage adds 5% to cost.

Extraction facility cost savings could not sufficiently reduce costs without yield improvements. Changes in interest rate and scale factor have negligible effect. At improved yields, a combination of reduced capital, operating, and interest expenses lowers cost to $45.507

bl. This is comparable to the cost for 10% improvement in extractables.

Reducing production costs could compensate for some lessened yield improvements.

81

Transportation costs, although outweighing the benefits of economies of scale in extraction, do not contribute enough to the overall cost to effect it. Halving the ton-mile fee for transport reduces final cost by less than 1%. A slight shift in the transportation network does occur to allow a minor improvement through scale economies. Even this combination of reduced linear and nonlinear cost does not significantly impact industry feasibility. Table 16 summarizes simulation results.

Energy balance does appear to be positive, with more energy produced than required for the production process (see Appendix D). Burning of bagasse for fuel is recommended to ensure positive net energy production, especially since bagasse has no opportunity cost at this time.

The amount of energy input to the industry does not appear to override cost advantages when production is well within the feasible range.

Table 16.

Summary of simulations.

Simulation

Number

1.

2.

3.

4.

5.

6.

Parameter

Changer Range

% A

Total

Cost Range

($/b1)

%A

Number and

1 to 3 location of facilities extractors

C= crop- a ping cost

$112.50/T to

$56./T

Combination

Linear costs

Yields

Interest

NA

50

Yield and % 41/ac and 7% to extract 16T/ac and 30%, or 517,000 bis to

8,898,000 bis

1600

270

Harvest

25% to 92%, strategy and

$15 to $55/T storage S.

Extraction costs C C

' K' C a, i

C o

K

: $17 to $35/T

: $24 to $47/T i : 10% to 16% a :

.5 to .8

Transportation

$.05 to $.10

per ton/mile

50

50

38

38

50

50

1000

38

407-443

293-407

51-407

44-65

351-407

404-407

45.50

28

48

14

89

1

9

87

82

CHAPTER 5

SUMMARY AND CONCLUSIONS

The objective of this study has been to develop a methodology to evaluate the economic feasibility of biofuels production, and in particular to isolate the variables crucial to attaining feasibility. The model constructed to define these variables is unique in its ability to accommodate a variety of plants and to integrate all portions of the production process. It was tested on a case study of a Euphorbia lathyris industry in the state of Arizona. This chapter summarizes the results of modeling and sets the context for model use. The assumptions, capabilities, and limitations of the model are discussed first, followed by the empirical context of the case study and the resulting critical variables.

Model Capability

The problem to minimize the cost of production by efficiently routing flows from supply nodes to demand nodes is formulated in a nonlinear transportation-transshipment format. Costs of production are based on a postulated Eurphorbia lathyris industry in arid regions. Total cost consists of the sum of costs of farming, harvest, storage, solvent extraction processing, and transportation between stages.

The model is unique to biofuels industry studies in its scale of analysis and integration of various segments of production. Previous

83

84 investigations of biofuels have concentrated on one phase of the process, such as cropping or extraction, but have not joined the parts into a cohesive whole. The transportation-transshipment model provides a vehicle for linking the nodes representing each procedure into a unified system.

The aggregate industry can then be optimized and total costs estimated for comparison with competitive products.

The model has proved effective for this analysis. The Arizona

Case Study demonstrated the model's ability to reach a solution over a wide range of values of the parameters. However, it is subject to the following limitations:

• It does not analyze net benefits or maximize profits. This study aimed, specifically, to target biofuel competitiveness with petroleum crude products, and did not attempt to analyze the oil market. Oil crude price is assumed to be relatively constant at predicted levels

(see Comparative Energy Pricing, Chapter 4).

As long as this is the case, profit will vary directly with cost. Variations in petroleum price will change both profit and feasibility. The model calculates the delivered price of biocrude from the cost side. This approach is especially useful when direct cost data are needed. Since industry development is not yet to the stage at which decisions on planting and construction can be made, the lack of profit analysis is not a serious drawback. It does limit comparisons to other crops and evaluation of effective resource use.

• The minimum feasible size is not included in computer program constraints. Situations in which the optimum is calculated to fall below

85 the minimum occur seldom, and only when excess capacity exists in the system. When they do, a subsequent run is easily made, manually forcing produce out of or into the (in)appropriate extractor by specifying capacity exactly.

• The program may reach an acceptable solution before reaching the optimal solution. This may be due to the existence of local minima in the cost function, in which case using alternative random starting points would force examination of other portions of the function.

• Modeling the entire production process requires depending upon a number of different sources for input data on different portions of the process. Data from one source may not use information or assumptions comparable to other sources. As much as possible, calculations used in the case study have either been chosen for their consistency or adjusted to correspond to the baseline assumptions. The information used is the best available at this time; it must continue to be refined until a production function is developed. In spite of uncertainties, the data are effective in demonstrating the usefulness of the model, and in providing a baseline from which alternatives and sensitivities can be gauged.

Arizona Case Study Results

Case study analysis indicates that a Euphorbia lathyris industry in the state of Arizona will not be feasible unless biomass yields can be quadrupled without increasing demand for inputs. Based on the case study simulations, the following specific observations are made. They are listed in order of importance (see Table

17).

Table

17.

Results of simulations ranked in order of impact on delivered cost.

Rank

Simulation

Number

Parameter Changed

% Cost Reduction

(or Increase)

2

3

4

5

6

7

1

Combination Linear costs

(all),

+ Yields,

+ Interest

3 Yield and

% extractables

4

Storage

2

5

1

6

Cropping cost; perennial crop

Extraction costs, linear and nonlinear

Extractor layout

Transportation cost

89

14

9

1

87

(48)

28

86

87

• By far the most important parameter in reducing final biocrude costs to a competitive level is biomass yield produced per acre farmed.

Increasing farm yield not only decreases costs of farming per ton, but also increases the number of biocrude barrels over which capital costs are spread. Yield can be increased more than costs can be decreased-doubling yield (100% increase) is not impractical, but cutting costs

100% is impossible. Amount of extractables is at least as important as tonnage biomass yield in reducing costs. The two factors together produce a dramatic response that results in a feasible industry.

• Increasing yields and extractables to the extent necessary is not likely. The most obvious opportunities for improving yield exist in selective breeding programs. Variability is present in the wild population. Genetic manipulation will probably extend the natural range. Optimizing other conditions such as planting configuration and combinations of inputs will contribute to yield improvements. In all cases, it is important to consider the possible inverse relationship between biomass tonnage yield and percent extractables recovered, and to concentrate on increasing the ultimate biofuel productivity. This is the crucial factor in the feasibility of a biofuel industry.

• Although E. lathyris is an annual, the possibility of perennial candidate species warrants examination. After first establishment, land preparation and germination irrigation costs were eliminated. Older plants are hardier, so that pesticides could be reduced and costs of increased resource use over time offset. Harvest and storage costs are lessened.

88

• Reductions in individual component costs of farming in themselves are not capable of resulting in competitive delivered cost. The importance of water in an arid region is not adequately reflected in the pricing structure (see environmental policy). Combined changes as part of a package of modifications in production or improved yields will impact industry feasibility. Halving all farm costs simultaneously results in nearly

30% reduction in final cost of biocrude as delivered to the refinery.

• Transportation costs are more important in determining extraction facility location than are economies of scale. Locating extractors close to farmed areas rather than concentrating processing in centralized locations improves total cost up to

10%.

Transportation costs become more important as yields increase and other costs decrease.

• Halving extractor capital and operating expenses combined produced a

14% lowering of total cost. This is again not enough to produce feasibility from baseline conditions. Lowering interest rates or improving economies of scale do not have significant effects.

The Arizona Method of solvent extraction has fewer steps than the process used by SRI

(1979), so it should reduce both capital and operating expenses. Increased storage costs could offset gains from more efficient processing, so development of different storage methods or harvest timing must be coordinated with processing research.

89

Policy Implications

An industry based on Euphorbia lathyris would not be economically feasible in Arizona at predicted prices without dramatic improvements in yields and some cost reductions. Although no single element in the production process can be improved enough to bring about feasible production, a moderate change in several factors combined with biomass increases results in lowering costs to well within a range projected to be competitive with future petroleum prices.

Cost of production is undoubtedly a vital element in projecting whether or not a project will succeed, and cost figures prominently in any profit or net benefit analysis. However, economics does not constitute the only reason for pursuing E. lathyris or other biofuel cultivation. Other policy areas to be considered include energy supply security, agriculture, environment, and regional development.

Energy Supply

Demand for energy in Arizona will continue to grow due to expansion in the state's population and industrial sector. Although substitutes for petroleum are already being used, it remains the main source of Arizona energy. Chapter 1 discussed the disadvantages of dependence on a non-renewable, imported energy supply. A biofuel grown and processed within the state would provide an available and relatively secure additional source. At 16 T/ac and 30% extractables, nearly 9 million bls/yr would be produced on 1.2 million acres; at 6 million Btu

13 per barrel this equals 5.4 x 10

Btu gross. Arizona's 1980 petroleum

demand amounted to 7 x 10

7

bis, or 4 x 10

14

Btu (University of Arizona,

90

1981). Disregarding product mix, approximately 13.5% of Arizona's petroleum demand could be met by E. lathyris biocrude at the above production levels. If 500,000 acres were cropped, 5.6% of demand could be produced.

This would fill total residential petroleum demand (1.8%) and half the commercial (8.6%) or electric utility (9.5%) petroleum demand. Fuel inputs required in the production process are included in the demand figures; net fuel generation for the biofuel industry itself also is positive (see Appendix D).

Agricultural Policy

The scale of production used in the model is large, approximately equalling total agricultural acreage in the state. It is not necessary to assume that all currently cropped areas will be converted to Euphorbia. While some more water-intensive or costly crops may be replaced, much of the area considered for biomass crop planting is marginal for current cultivation. These areas could be tolerated by E. lathyris and would expand available acreage without increasing land prices. At low yields the acreage could be cut to one-third and still support an extraction facility.

Costs will be more favorable for E. lathyris than for current commercial crops since it is more suited to arid conditions. Ability to use less fertile land or lower-quality water provides a cost advantage.

As water becomes more scarce, the relatively low water requirement will further improve this position. At 2 ac/ft/yr. E. lathyris uses less

91 water than most major Arizona crops (Hathorn and Armstrong,

1982).

In areas where Euphorbia replaces higher users, current water supplies may be regarded as adequate. In other areas, saline or otherwise untapped sources would need to be utilized.

Environmental Policy

The environmental impact of agricultural development is related to land and water use. Any new cultivation requiring additional water will have adverse effects on groundwater and surface fresh water supplies.

Several factors might affect the water costs incurred:

• Without price discrimination, additional supplies of fresh water from the CAP are more likely to go to municipal uses than agricultural, especially since the estimated user cost is likely to exceed the price agriculture is willing to pay (Barr and

Pingry, 1977);

• An increase in market pricing to more adequately reflect water scarcity;

• Changes in depth to groundwater, increasing pumping costs and energy usage;

• Lack of infrastructure in now marginal areas, and particularly of facilities for obtaining and transporting previously untapped saline or effluent sources. Water law restricting new irrigation may be a barrier to development;

• As other users such as mines or recreation respond to water shortages, demand for saline water may increase and drive up prices. A

92 use cascade allotting progressively saltier water to succeeding uses might provide a solution to competing demands;

The quantity of water needed may differ if salinity rises.

Crops may require additional water for growth or to leach salts from the soil.

Tilling of new land may have undesirable results unless appropriate land management techniques are employed. Habitat disturbance poses an additional problem. In areas previously farmed but no longer under cultivation, environmental effects are potentially positive. New ground cover may aid in soil stabilization and prevent erosion.

Regional Growth and Development

The advantages of developing a new biomass industry include additional employment, increased personal and business income, and a broadened tax base. Additional basic income would not be generated since the product is intended for in-state use, but the flow of revenues out of the state for imported fuel products would be reduced. Development of coproducts would increase feasibility by increasing revenue and distributing costs. Basic income might be generated.

Conclusion

The cost minimization model was successfully used to isolate the critical factors for an E. lathyris industry in an arid region. Subject to the limitations of data and model capability discussed above, results determine that this industry would not be feasible in Arizona at predicted prices without major improvements in yields and moderate changes in

93

a

combination

of input

costs. Viability is critically dependent

on

improvement

in tonnage

yield produced

per acre and on percent

extractables recovered

in

processing.

APPENDIX

A

CROP PLANTING

AND

GROWTH

DATA

94

95

A. Land

Preparation and

Planting

Available acreage and its location are important because a minimum scale of production is required to support the machinery component, and particularly processing plant; transportation distance to storage and processing may add significant costs if acreage is dispersed. For the Arizona Case Study, rough mapping of possible sites in Arizona is used to determine acreage likely to be available (see Appendix C). Land values or rent charges are difficult to estimate. If significant acreage is committed to biofuels production, an increase in demand for farmland may push prices upward. On the other hand, if production is constrained to regions unsuited for other cropping, the valuation may be significantly lower than usual for agricultural use. Estimates of land rental charges from the literature are as follows:

Doane Agricultural Service, Inc. (Doane, 1981): $125/ac (Pima

County);

Mendel, Schooley and Dickenson (1979): $100/ac (Yuma);

Alexander (1981): $50/ac (California).

Estimated Crop Budgets for land preparation and planting are shown in Table 1 (Chapter 2). Individual cost estimates for planting activities are listed below. All estimates are per acre.

Bed Preparation

Well-drained soil is essential. Nemethy et al. (1978) recommended sand/peat mix.

96

Costs will include disc, plane, plow, and shape beds (raised to insure rapid drainage). Peoples (1981b) deletes beds, to keep ground flat for flood irrigation.

Mendel et al. (1979) estimate cultivation costs for Yuma (per acre) as follows:

Plowing

Tandem discing

Harrowing

$ 8

4

2

$1

4

Doane (1981) estimates for Tucson (per acre):

2 Offset discs

Land plane

List

Buck rows

Disc ends

$ 9

4

4

$1

8.

.4

.2

Alexander (1981) estimates $15/ac.

Seed

Germination percentage and optimum planting density are not certain. 240,000 plants per acre may be optimum (see planting data below).

Sachs and Mock (1981) recommend overseeding by a factor of two and thinning to desired density after sprouting.

Cost estimates for seed/acre:

Peoples (1981b):

Wright (1981):

Doane (1981):

Mendel et al.

(1979):

35 lbs at $2/1b = $70

20 lbs at $4/1b - $80

20 lbs at $2/1b = $40

30 lbs at $2/1b = $60

97

Seed cost initially will be much higher than estimated above, but should be reduced significantly by commercial propagation.

Additional costs may be incurred by the need to remove a germination inhibitor (Fontes and Watson, 1978). Scarification or seed coat removal is recommended. Pre-imbibition on seeds for five days resulted in 30% increase in emergence (OALS, 1979), but scarification did not produce a significant change.

It should also be noted that potential for mechanical transplanting exists, since plants appeared to tolerate frost and transplant shock

(OALS, 1979).

The effects of planting density on biomass and hydrocarbon yields and on water use are still not certain. The United States

Congress Office of Technology Assessment (USOTA, 1980), reported that plants on field edges were 1.5 times larger than interior plants. Hydrocarbon yield changes were not reported. High-density planting has been shown to stimulate vigorous and leafy growth (OALS, 1979). Plants at

300,000 and 600,000 plants/ac tended to be 7-12" taller than in lower densities, and to have more leaves and therefore possibly more extractables. Southern-variety plants responded more to increased population pressure than did Northern and Chico (OALS, 1979), but the increased height exhibited may be correlated with lower hydrocarbon concentration.

Kingslover (1982) reported higher-density plots produced more dry matter and more biocrude/ac (Table 2, Chapter 2). However, competition was clearly shown and the biomass production of individual

98 plants decreased. This tradeoff between individual and overall production needs to be further examined.

Increasing population density to

240,000 plants/ac on

40" beds increased per-acre yields of both dry matter and hydrocarbons. Three

T/ac of dry biomass and

4.4 bls/ac of extractables were produced.

Sandoval

(1981) planted

2 rows/bed, with beds spread one meter

(39 in) apart. One-meter plant rows did not provide complete ground cover. With a seed every

3.5" and

60-65% germination,

60,000-64,000 plants/ac resulted.

Peoples

(1981b) recommends drilling in six-inch rows; Sachs and

Mock

(1980),

Mendel et al.

(1979) and

Nemethy et al.

(1978) all advise one-foot centers on rows

12-15 inches apart. Peoples

(1981a) suggests use of sorghum planters.

Costs are reported as follows:

Mendel et al.

(1979): labor costs

= $4/ac

Doane

(1981): equipment

= $4/ac

Growth and Cultivation

Fertilizer

In greenhouse tests, no significant effects on biomass production resulted from applications of Nitrogen-Phosphorus-Potassium

(N-P-K).

A significant difference was noted between soil types, and it is possible that nutrient response was hindered because of rapid root-binding.

Earlier tests

(OALS, 1979) showed that granular fertilizer produced more vigorous growth and an increase in lateral and height growth. Nitrogen

99 at 10 parts per million (ppm) nitrate (NO 3 ) increased the weight of leaves, stems, and roots significantly, but stunting occurred at 20 ppm

NO

3

(Figure Al). Levels of 40 ppm and above hindered seedling emergence

(Kings lover, 1982; OALS, 1979). Phosphate (P O 4) applications aided biomass production. Best plant height increases (59%) were obtained at

10 lb/ac of 15-30-15 N-P-K. Phosphorus in hydroponic solution also increased plant weight (Figure A2). It should be noted that higher levels of biomass production are often negatively correlated with hydrocarbon extractables, so the effectiveness of fertilizer applications is not completely known.

Other recommended levels of nutrients/ac are as follows:

Mendel et al. (1979): Arizona soils need added nitrogen, but potassium, phosphorus, and lime are not considered necessary. Cost estimates included 150 lb nitrogen (anhydrous ammonia) at $37.50, 60 lb super phosphate at $12, and $5.50/ac application, including labor.

Sachs and Mock (1980): 100 lbs nitrogen/ac (50 lb twice), for

California soils.

Nemethy et al. (1978): Ammonia sulfate (NH 4 ) 2 (SO 4) in doses of

100 lbs and 50 lbs, in California.

Doane (1981): 45 lbs nitrogen, at $.27/1b = $12

200 lbs phosphorus, at $.29/1b = $58 application cost = $3.50/ac total = $73.50/ac

Peoples (1981b): 200 lb 16-20-0/ac

Alexander (1981) estimates $180/ac fertilizer and application costs.

Nitrogen Concentration, Millimoles/litre (mM)

Leaves

— -- Stems

Roots

Figure A.1. Effect of three levels of nitrogen on

E. lathyris dry weight in hydroponic nutrient solutions (Kingslover, 1982).

1 00

Leaves

--- Stems

Roots

101

Phosphorus Concentration, Millimoles/litre

(mm

)

Figure A.2. Effect of three levels of phosphorus on E. lathyris dry weight in hydroponic nutrient solution (Kingslover, 1982).

102

Weed and Pest Control

E. lathyris is particularly susceptible to fungi in the hot summer months. In areas where weather allows, such as Arizona, planting should be done in late September to early October to avoid the hottest period. Pathogens encountered include Pythium aphanideratum, Rhizoctonia solani, Macrophomina phaseolina, Phytophthora, Verticilium, and Fusarium.

Some reduction in germination occurs when fungicides are applied immediately after planting. When applied during early growth, fungicides appeared to drastically lower hydrocarbon content. Some strains may be resistant to chemical applications, and attempts are being made to select these. Terracoat, Terrachlor, and Vitarax do not appear to cause significant damage to mature crops (OALS, 1979). Pre-emergent herbicides suggested for use include Prowl, Tolban and Treflan.

Mendel et al. (1979) calculate chemical costs/ac as follows:

Insecticide (metasystox R),

0.5 gallons (gal) each, at

Application

Herbicide and cultivation

$8.50/gal

7

5

$20.50

Doane (1981) accounts for 1/2 gal/ac of Prowl, costing $14, and

Alexander (1981) calculates costs at $26.50/ac. Peoples (1981b) recommends 1 lb of Prowl per acre.

Additional Growth Stimulants

Pinching at the apical meristem causes branching and increased dry weight by 24%. Cyclohexane extractables increased from 5 to 5.6%,

103 ethanol (ETOH) from 30.4 to 35.1%, and total crude and total extractables from 35 to 40.7%. Hydrocarbons also were increased by cutting to

4" (8" cut off).

Experiments with plant growth regulators (PGRs) were inconclusive. None significantly increased hydrocarbon yields. Final dry weight apparently was stimulated. Surfactants are needed with PGRs; and

Tween 20, a surfactant alone, seemed to stimulate growth and hydrocarbon content in the greenhouse. Mild stress also appears to increase hydrocarbon production (OALS, 1979).

Mycorrhizal fungi established relationships and had marked growth response. Glomus mosseae inoculation produced four times more dry matter and 35-40% more cyclohexane extractables. Pathogenic fungi do not seem to interfere. Phosphorus uptake appears to be aided by mycorrhizae

(OALS, 1979).

It should be noted that plant size (general biomass production) is inversely related to hydrocarbon content (OALS, 1979). Experiments with growth stimulants should therefore concentrate on increasing extractables rather than plant size.

Irrigation

Components of irrigation costs are water, application, labor, equipment, and power. Baseline estimates assume that infrastructure already exists. Pumping costs are likely to be a major factor, and should be considered if energy balance studies are undertaken. Additional factors which will increase costs include availability of fresh and saline

104 water sources, added infrastructure, and special needs for saline application or cascading use. Adequate drainage must be assured if flood irrigation is to be used.

A linear relationship exists between water consumption and cropping cost (McLaughlin, pers. comm., 1981), with water use calculated as contributing up to 65% of the cost of farming. For E. lathyris, Mendel et al. calculate 60% (Table 2). There is considerable potential to reduce feedstock costs by selecting crops with high water use efficiency

(WUE), the ratio of dry matter production to consumptive water use.

Crops most suited to arid regions are those that have a maximum annual water use below 2 acre-feet (ac-ft). At water requirements of 700 to

1,500 tons (.5 to 1 ac-ft) of water per ton of biomass, considered within the empirical range for arid regions, E. lathyris yields vary from 5 to 10 T/ac COALS, 1979).

Water studies with Euphorbia lathyris (Sandoval, 1981) indicated that wetter treatments, i.e., application of more irrigation, encouraged higher consumption and produced higher maximum yield of dry matter.

Higher biocrude yields overall resulted, but percentages of extractables did not increase. Drier treatments produced greater yields than wetter treatments at the same level of water used, but plants in drier treatments were older at that level. For example, a plant which received 15" of water over a four-month period as part of a total application of 30" would produce less biomass at the 15" point than would a mature plant with a total application limited to 15". Peak use occurred in April.

MONTH

80'

OCT I NOV I OEC I JAN I F=3 I mARCH I APRIL I MAY I

80

105

—10

0

50 100

150

Days After Sowing

200

250

O

Figure A.3. Cumulative water use of E. lathyris (Southern ecotype) in four irrigation regimes. -- In treatment A, soil moisture was maintained at or near field capacity (Sandoval, 1981;

Kingslover, 1982).

—60

— 50

— 40

— 30

— 20

106

Table A.1.

Water used, plant height, percent ethyl acetate extractables, and total biocrude produced in two ecotypes of

E. lathyris

(Kingslover, 1982).

Plant

Southern

Northern

Water Used cm (")

Plant Height cm (")

Percent

Ethyl

Acetate

Extract

Biocrude Production bl/ha (bl/ac)

71 (30)

60 (24)

48

(19)

32

(13)

67 (26)

58 (23)

46 (18)

24 (9)

105

(41)

97 (38)

80 (31)

64 (25)

43 (17)

39 (15)

33 (13)

25 (10)

7.2

6.6

7.3

7.5

7.1

7.3

7.2

7.2

7.5

6.7

5.8

(18.5)

(16.5)

(14.3)

4.5 (11.1)

5.1 (12.6)

4.6 (11.4)

3.1 (7.7)

2.7 (6.7)

107

Plants in the highest water use group received 29.5"; least water received was 10.5". In the lowest group, plants showed stress but continued to live beyond what is generally assumed to be the permanent wilting point. At that level, three T/ac of biomass with 7.5% ethyl acetate extractables produced 1.8 barrels of biocrude per acre.

OALS (1979) reported that 27" of water produced 3.3 bl/ac of biocrude. At some point between 31.3" and 45" of combined irrigationrainfall levels,

Euphorbia lathyris

ceased to produce significant added dry matter.

Estimates of the optimal amount of water vary. Sachs and Mock

(1980), without giving exact amounts of water, recommend irrigating to field capacity after sowing, and subsequently watering every 10 days to replace 2/3 pan evaporation. Mendel et al. (1979) recommend 4 ac-ft/ year, in 6 applications; cost for water and application is $170/ac.

Both base their calculations on California conditions.

Nemethy et al. (1978) assume that there will be 5.5" of rainfall during the growing season, and suggests an additional 1.2" every 2 weeks to 19". Combined with rainfall, this totals 24.5". Peoples (1981b) also feels that 24" may be near optimum. Doane (1981) calculates costs for Arizona with only 1.33 ac-ft of water; at $49/ac-ft and with 1.6

hours (hr) of labor at $5ihr, irrigation costs are $73/ac.

Water and pumping prices used in cost estimates and scenario calculations are current. Both are likely to change in the future due to population growth, increasing groundwater scarcity, and the CAP. No attempt is made to

account for these changes in the case study simulations. The implica-

108 tions are discussed in Chapter 5, Conclusions.

Pumping and machine costs must be added to this figure. For the

Avra Valley area, with a 375-foot electric pumping lift, energy is $36, repairs $3, and fixed costs $9, totaling $48.77/ac-ft pumped (Wright, pers. comm., 1981).

Salinity

In greenhouse tests plants were treated twice weekly with water containing 0, 50, 500, 2500, and 5000 parts per million (ppm) sodium chloride (NaC1). The high-salt treatment increased height up to 24%, but no significant difference was found in average dry weight. Up to 2500 ppm, total extractables increased from 30% to 35%, but fell to 26.4% by

5000 ppm (see Table A2). In the field, plants were irrigated with soluble salts up to 19,000 ppm, with soluble soil sodium concentration to

4600 ppm. Height significantly increased with salinity, dry weight was not affected, and cyclohexane extractables decreased slightly. Germination is affected by salinity. At 300 ppm 100% of seeds sprouted, but at

900 ppm only 62% appeared. At 1800 ppm, germination was reduced to 19%

(see Figure A4).

Using saline water may reduce water and pumping costs, but additional problems may result. Irrigation efficiency has previously been defined as not watering below the root zone. Salinity will build up very quickly in this region, especially if fresh water is unavailable for leaching. Unless land is only to be used

for

a short period, salt removal will need to be considered.

109

Table

A.2. Effect

of six

levels

of

soil salinity

on

height, weight,

sodium

level

and

cyclohexane extractables

of

E. lathyris

(Kings lover, 1982).

Irrigation

Water

NaC1 Level

(PPm)

Final

Soil

Soluble Salt

Level

Total

(PPm)

Na

(ppm)

Plant

Height

(cm)

"

Plant

Dry

Weight

(gm)

Extractables

% C

H

0

50

500

2500

4000

5000

1,162

1,421

2,611

10,419

17,017

18,753

166

242

614

2,852

3,758

4,685

(67) 26

(71) 28

(78) 31

(84) 33

(85) 33

(85) 33

31

31

40

37

34

34

9.0

7.5

7.9

8.0

7.3

7.2

0

0

...

In m-

ti)

=

4-) cd

s-4

>, r4241

110 i

O

0

1••• i

0 in

UORBUIUlieD %

111

Additional Cultivation

Doane (1981) accounts for two added cultivations, with equipment cost of $7; three hours of labor add $16. Alexander (1981) calculates

$5/ac for machinery and $10/ac for labor.

Crop budgets summarizing cost calculations appear in Chapter 2.

Yields

Preliminary field plot data indicate a minimum of 3T/ac dry matter and 4.5 bl/ac of biocrude. OALS (1980) uses these figures to suggest that under proper cultivation conditions existing strains could yield 8T/ac dry matter; with 15% extraction rate this produces 9 bl/ac/ yr of biocrude.

Field weight produced per acre, before drying, has been estimated at from 8T/ac (Peoples, 1981a; Calvin, 1977a; 1977c; 1978a;

Foster and Brooks, 1981) to 15T/ac or 17T/ac (Mendel et al., 1979). The 15T/ac yield is translated into 3T/ac dry weight, which at 28% crude extract would yield 6 bl/ac (OALS, 1979). Planting density increases are expected to double yield. Mendel et al. (1979) assume that a 17T/ac yield at

20% dry weight, or 3.4T, would produce 8% extractables or 10 barrels.

Commercial spacing is expected to reduce yield to 3.7 barrels. Eight

T/ac or 8-12% extractables would produce 8-10 bls/ac. Calvin speculates that yield could be increased to 20 barrels (Calvin, 1978; Foster and

Brooks, 1981).

APPENDIX B

PROCESSING DATA

112

113

E. Storage

Field-dried plants twine-tied in bales measure 18 x 24 x 48 in and weigh 150 lb. Six bales placed on 4 x 5 ft pallets will have a total weight of 0.45 tons. Three pallets stacked 15 ft high weigh 1.35 tons.

This pallet can be conveniently worked with a fork lift.

For a 30-day supply for a 3,000 ton/day plant, total weight is

3700 T/D times 30 days = 110,000 T. This requires 82,200 stacks 20 ft 2 , or 1.64 million ft

2

. Add 10% for personnel and equipment movement, for a total of 1.8 million ft

2

. Costs for storage facilities are calculated in Chapter 2.

F. Extraction

A diagram of the proposed extraction facility is presented in

Figure Bl. Steps in the extraction process are described below, and capital costs are listed in Table Bi.

Conveyers

Conveyers transfer plants to chopping facilities or to temporary on-site storage. Temporary storage of 2-3 days is maintained to assure continual operation.

Grinding

The feed rate to choppers averages 3700 field T/day or 3000 dry

T/day. Equipment needed includes 20 operating knife-type shredders with

54 ” throat openings and 1/2-in screens. Chopped plant size is 1/2 x 0. Screw conveyers feed plants from choppers to live bottom hoppers, which in turn feed to single-disc attrition mills (refiners). Milling

Plant

Material

Receiving

Milling

Process

Storage

Tank

Capacity

Scale

Conditioner

mmin

IFFAR ii.4

Meal Preparation

Bin

Feeder

Flaker Rools

114

Oil to

Storage

Meal to Storage

Meal Finishing

Hammermill Grinder

Figure Bi. Proposed facility for Arizona Method extraction (F)

(McLaughlin and Hoffman, 1980).

Table B1. Estimated capital investment for E. lathyris oil conversion (F) (SRI, 1979).

Plant Investment

Million $

E. lathyris receiving and handling

Truck dumping

Conveyers

On-site storage

E. lathyris chopping

Live bottom-feed storage hoppers

Tramp iron magnets

Knife-type shredders

Discharge hoppers

E. lathyris refining

Live bottom hoppers

Tramp iron magnets

Screw conveyers

Refiners

Discharge hoppers

Mixing and filtration

8300-b1 mixing tank

Rotary vacuum filters

Pumps/exchangers

Drying and vapor recovery

Screw conveyers

Rotary calciners

Solids coolers/hoppers

Exchangers

Liquid receiving vessels

Flashing and distillation

Heat exchangers

Vapor-liquid receiving vessels

Pumps

Atmospheric distillation column

(15,800 bl/day)

Utility facilities

Boiler, plant-residue fired

21 Megawatt steam turbine

Solvent storage tank (1 day: 190,000 bl)

Small product oil storage tank

$ 2.4

3.1

4.8

10.9

8.3

5.1

26.9

115

Table 81. -- Continued

Plant Investment

General service facilities

Total plant

Land cost

Organization and start-up

Interest (during construction)

Working capital

Total

Million $

9.2

70.7

0.2

3.5

5.4

1.0

$80.8

116

117 requires 8 Model 240 series (largest) with 52-inch discs, screw feeders, and 2500 horsepower (HP) 1800 revolutions per minute (RPM) direct-drive motors. Rollers may be used instead of refiners to break cell walls containg oils.

Solvent

22,200 T/day or 154,170 bl/day (141,400 recovered). Laboratory extractions typically use 18:1 weight ratio of solvent (acetone, heptane, etc.) to dry plant matter in an 8-hour extraction period. SRI (1979) assumes development will allow a 6:1 ratio and 1-hour residence time, and that solvent will be derived from a selected cut of product oil, analogous to solvent-refined coal (SRC) technology. No external solvent replacement is necessary. This lowers estimated inventory and operating costs. Operating costs are summarized in Table 7, Chapter 2; costs per barrel are delineated in Table B2. 154,170 barrels of solvent daily, with density of 228 lb/b1, are used. 141,400 barrels are recycled.

Filtration

The solvent-plant mixture is pumped to a filtration section.

There rotary vacuum filters remove 20 lb/hr of solids per ft

2 of filter area. Twenty-three operating filters are required for 500 ft

2

. The result is a filter cake containing 60% solids by weight.

Recovery

The filter cake is sent by screw conveyer to 3 rotary drum calciners for liquids recovery. 800-1000 ° F flue gas (combustion of plant

Table B2. Estimated annual operating costs (regulated producer) (F)

(SRI, 1979).

118

Million $/yr $/b1 $/Million BTU (Oil)

Materials

E. lathyris at $13.21/T $16.05 field-dried

Catalysts and chemicals

Maintenance materials

0.1

1.1

(1% of plant facilities costs)

Total

Labor

$17.25

Operating (8.5 men/shift) $ 0.6

Supervision 0.1

Maintenance 0.7

Administrative and support 0.3

Payroll burden 0.6

Total $ 2.3

$30

0.3

2

$32.3

$ 1.1

0.15

1.3

0.5

1.1

$ 4.15

$5.9

0.05

0.4

$ 6.35

$ 0.2

0.03

0.3

0.1

0.2

$ 0.8

Purchased Utilities

Water (1810 gal/min)

Electricity (21,000 kilowatt hrs)

$ 0.5

Fixed Costs

General and administra- $ 1.4 tive expenses

Property taxes and 1.8 insurance

20-year depreciation 4

$ 1

Supplied By Bagasse

$ 0.2

$ 2.65

3.3

$ 0.5

0.65

1.5

7.45

$13.4

Total

Total $ 7.2

27.2

$ 2.65

10

6.9

50.9

12.9 2.5

Return on rate base and income tax

Total

$34.1 $63.8 $12.5

119 residue) heats the internal temperature to 300

°

F to reduce liquid content of plant residue to about 1/2% by weight. Calciner vapors are condensed and sent to solvent storage. Solids are cooled; 2141 1/day are sent to sales and 611 T/day to an on-site residue-fired power plant. Costs of

1000 T/day processing plant include interior electricity generation; onsite bagasse is sufficient to support a 40-megawatt generator (Kingslover,

1982).

Solvent and oil products from the filtration step are pumped to

150-250 pounds per square inch (psig), heated, and flashed in a flash drum. Vapors contain 90% of original solvent, 2% of product oil, and all original 700 T water. Vapors are cooled and condensed, 129,000 barrels of solvent recovered, and water returned to field. 15,800 bl/day of flash liquids are heated and distilled, with 14,100 bl/day solvent recovered.

1625 hi/day of product oil are produced. These are sent to a temporary heated storage tank with 1-day capacity. From temporary storage they are sent to an off-site bulk station. 93-94% of incoming product oil is recovered. Material balance is presented in Table B2. Each pound of hydrocarbon produces 17,000 BTU; a day's production equals 8,874 million

BTU. Residue, at 7300 BTU/lb, produces 39,989 million BTU daily.

Products

Major components of E. lathyris crude apparently are pentacyclic triterpenones. Paraffins and carotinoides are also present. Pentacyclic triterpenones have a typical heat combustion of 17,000 BTU/lb (Nemethy et al., 1978). Up to 20% of dry weight is crude oil fraction, with 40% fermentable carbohydrates.

Table B3.

Material

balance (F).

1000 lb/hr

Input

Dry

E. lathyris

Moisture

Combustion air

Water

250.

58.3

356.6

904.0

1568.9

Output

Oil

Residue

Flue

gas

Evaporative loss

Wastewater returned to

field

Blowdowns

From process

20.3

178.4

407.6

540.0

364.0

58.6

1568.9

15.9

3.7

22.7

57.7

100.0

1.3

11.4

26.0

34.4

23.2

3.7

100.0

120

121

Sachs and Mock (1980) estimate only 12% extractables, with 4% heptanes and 8% acetone. Table B4 shows cracking products from E.

lathyris crude. Based on the value of crude products, lathyris average value for 100 lbs is calculated at $8.75, compared to $6.15 for petroleum crude (OALS, 1979).

122

Table

B4.

Catalytic Cracking of Crude Oil from

Euphorbia lathyris

(Zeolite - 500 C) (Kingslover, 1982)

Product

Percent

Ethylene

Propylene

Toluene

Xylenes

Fuel Oil

C

1

-C

4

Alkanes

C

5

-C

20

Units

Coke

20

15

10

10

20

10

10

5

APPENDIX C

MAPPING OF POSSIBLE E.

LATHYRIS

GROWTH AREAS IN ARIZONA

123

124

Criteria and corresponding areas of possible growth are presented for each category of conditions beneficial for Euphorbia lathyris cultivation.

Natural Habitat

E. lathyris does not occur wild in Arizona. Of the ecotypes known, none is a true desert plant. It prefers moderate climates, such as those in the Mediterranean regions from which it is reported to originate. In the regions of the United States where it has naturalized, mild climates without temperature or rainfall extremes and well-drained soils predominate. No region of Arizona fits the combination of characteristics in which E. lathyris occurs in nature. Of the Thornthwaite climatic regions, the semiarid mesothermal and subhumid mesothermal might provide the best approximation. These areas are essentially the same as those covered in the Koeppen regions of semiarid and humid subtropical (see

Map 1). The northern and central eastern areas of the state are eliminated, as are the far western and south central zones. The Holdridge life zone maps classify the vegetation of this area as subtropical lower montane thorn steppe (see Map 2).

Temperature and Insolation

There are three principal elements of importance in this category: insolation required for hydrocarbon production, avoidance of temperature extremes during the growth period, and proper germination conditions.

Although not directly related to temperature, insolation is most conveniently considered here. As stated in Chapter 4, hydrocarbon produc-

Combination of Koppen Semiarid and

Humid Subtropical and Thornthwaite

Semiarid and Subhumid Mesothermal

Map 1. Suitable climatic regions (from Hecht and Reeves, 1981).

125

ttatt.1=1*Iir"4""

I.

"L"=“""="..4611

toottatoban finib mw

1.41111.111MOOM141111

FLAGSTAF

'

3

"

111. 1=1:140.1

1:INM=113:1=11

n

11.01=

10101.1MIHNIMMUMMII.111.1111111

.

.10MININIINIMOI.110111

agrIhr

,

''

"1===1:::: oPHOENIX

126

Subtropical

DOS011

Subtropical Dow? Scrub

Subtropical Thorn

Woodland

Subtropical Lower Mlonlans

Detail Scrub

Subtropical Lobar Montane

The,

Steppe

Subrrobcol Lbw Montano

(7. Fo..st

Subtropical Montane

Dew, Scrub

Subtropical Montane Sl*PP•

I

Subtropical Montane

Moist Fonts,

Subtropical Subolpin.

Moist Forest

Subtropical Subolpfne

Wet Fora.,

Subtropical Alpine

Rom Tundra

I Subtrorncal nhvol

0 '0 70 SO 40 50

11•111 3=Ma

E

Map 2. Holdridge life zones (from Hecht and Reeves, 1981).

127 tion may be directly related to levels of solar radiation. Nearly all of Arizona receives 500 langleys (240 w/m

2

) of solar radiation annually, and all falls into the 450 langley intensity range (Calvin, 1977;

Johnson and Hinman, 1980). Arizona, in particular the southwestern portion, receives a large percentage of available sunshine--cloud cover is slight and there are few particulates or water vapor to filter out rays

(Hecht and Reeves, 1981). Insolation and percent of sunshine are illustrated on Map 3.

E. lathyris is an annual, warm-season plant in temperate areas.

It requires a growing season of at least 180 days with no freezing each year (Mendel et al. 1979), as shown on Map 4. Some plants may withstand a freeze, but few will tolerate sustained cold. Older plants are hardier. A gradual shift in temperature will cause less damage than will a sharp drop (California Energy Commission, 1979). White, Dyer and

Sloane (1941) report that dry plants may withstand cold better, and suggest that winter temperatures should be as low as possible without freezing; however, Nemethy et al. (1978) note that during low temperature periods growth is slow and may result in low yields. Santa Ana

(California Energy Commission, 1979) field station plots tolerated temperatures to 28

°

F, but exhibited poor growth and produced little biomass and latex. In Arizona 80% of plants established in April 1980 flowered in June of 1981 following average monthly temperatures as low as 33

°

F

(.4

°

C) in February. Mild winters resulted in more rapid growth and a lower water requirement (OALS, 1979).

128

rl

180 Days or Less Frost-Free (Hecht and Reeves,

1981)

Additional Areas of Mean January Temperature Below

35

°

F (Sellers and Hill, 1974; Hecht and Reeves, 1981)

Map 4. Severity of winter.

129

130

Although high insolation is desirable, there may be some danger of sunburn in the summer months (White et al., 1941). High air temperatures in summer, averaging nearly 100

°

F (38.8

°

C) daily maximum, coupled with low relative humidity may have contributed to fungus infestation and crop failure (Hinman et al., 1980; Flug, Sandoval and Fangmeier,

1981).

Each species of

Euphorbia

has a narrow range of temperatures for maximum germination. Mendel et al. (1979) report that soil temperature must be at least 60

°

F (15.5

°

C) for germination and emergence. Optimum planting occurs when average daily soil temperature is between 16

°

C and

26

°

C (80

°

F) for a minimum of 4 days prior to planting; midday soil temperature at 4" depth should be at least 12

°

C (54

°

F). Under these conditions, 70% emergence should occur within 10 to 20 days (Hinman et al.,

1980; Peoples, 1981; Sachs and Mock, 1980; Peoples and Johnson, 1980).

These temperature ranges could correspond to either spring or fall planting, as illustrated in Figure Cl. In Arizona a fall planting in late

September or early October is recommended to avoid summer temperature extremes and soil pathogens.

Soil Quality

Soil quality factors considered include nutrients, salinity, texture, slope, and drainage.

Nitrogen fertilizer trials indicated that high levels of soil nitrogen reduced the rate of emergence. High levels of nitrate in hydroponic solution also significantly decreased dry matter production (OALS,

100

050

8

n

0

0

J

F

A

WI

.

JJ A S 0

i4

D

%

Emergence

- - -

Mean Monthly Temperature

Month

Figure

C.1.

Mean monthly temperatures

and percent field

emergence

of

Euphorbia lathyris (Hinman

et

al.,

1980).

30

131

132

1979).

Since irrigated soils of Arizona are often deficient in nitrogen

(Mendel et al., 1979), this should pose no problem. Response to other nutrients is not certain.

Soil contents of potassium, phosphorus, and lime are probably adequate for Euphorbia production. Peoples and

Johnson

(1980) cite significant response to phosphorus. Soil deficiency could be compensated by fertilizer application, and excess in soils is unlikely to affect planting distribution.

E. lathyris appears to be tolerant of salt and sodium and grows well in marginal soils. High levels of sodium chloride (NaC1) in soil are tolerated with a slight reduction in cyclohexane extractables.

Under greenhouse conditions, plant height increased at higher salinity levels and plant performance remained constant to nearly

15,000 ppm

(OALS, 1979).

Field soil salinity has no apparent affect on seedling emergence, although in the laboratory sodium ion concentration of 1200 ppm reduced seed germination

50%

(Hinman et al.,

1980).

Salt stress appeared at

15,000 to

20,000 ppm in early growth stages

(OALS, 1979).

Sodicity up to

700 ppm generated no apparent adverse effects in field tests.

Greenhouse soil texture tests indicated that fresh biomass weights were higher in soils from Tucson than in soils from

Marana (OALS,

1979). Marana plants appeared stunted, perhaps because the heavier, less porous soils may restrict root development and respiration. Sandy soil not only promotes drainage but also requires minimum tillage and preparation (Mendel et al.,

1979).

In contrast, Foster and Brooks (1981) report that preferred soil types range from silt loam to loam.

133

Since good drainage is vital, some slope in cropland may be desirable to avoid saturation (White et al., 1941). Lands with slope greater than 17

°

or 30% are unsuitable because they are too steep for mechanization. Irrigation pumpage is difficult, also (California Energy

Commission, 1979). Map 5 shows major soil groups. Map 6 indicates areas suitable for irrigation. Criteria used to define suitability fall into three categories: soils, topography, and drainage. Soils covers texture, moisture retention, depth, saline or sodic, permeability, coarse fragments, rock outcrops, and erosion. Topography covers stone removal, slope, and leveling and tree removal. Drainage covers depth to water table, depth to drainage barrier, and surface and air drainage. Map 7 shows major topographic features. Map 8 illustrates surface rock types.

Availability

of Land

Land ownership, current use, or designation for special purpose could interfere with agricultural development activities. It is expected that E. lathyris production not compete with current major food and fibre prOduction activities. Other activities likely to preclude cropping are settlements, either principal or rural subdivisions, government lands, forestry or mining regions, Indian reservations, and natural or recreational areas. Map 9 illustrates public lands, the category which removes the most acreage from consideration.

Some areas covered above, while in themselves unsuitable for crop production, might provide waste water which could be recycled for irrigation use. Locations near settlements or mines might be advantageous for that purpose.

BROWN, REDDISH BROWN AND

LITHOSOL

SOILS OF SEMI-ARID

UPLANDS

LIGHT COLORED SOILS OF ARID REGIONS

ISIEROZEM, LITHOSOL,

RED DESERT, ALLUVIAL/

SWELLING CLAYEY SOILS OF UPLAND PRAIRIES

DEEP SOILS OF THE ALLUVIAL FLOOD PLAINS

IIII

ALPINE MEADOWS

SALINE AND SODIC SOILS

SOURCES US DEPT OF AGRICULTURE,

SOLS OF THE

WESTERN MATED STATES, 1964

SW BURL, SOILS OF ARIZONA: TECHNICAL

BULLETIN Ill, FEB

966

REGOSOLS

AND

LITW:50LS

ON CONSOLIDATED UPLAND

MATERIALS AND MISCELLANEOUS LAND TYPES

Map 5. Major soil groups (Hecht and Reeves,

1981).

134

Soil Suitable for

Irrigation

Already Irrigated

Unsuitable

Map

6.

Suitability of soils for irrigation (Lower Colorado Region

State-Federal Interagency Group, 1971,

Appendix X).

135

•••••• •

1.1

.....

136

SOUTHWEST

-‘• i

.. •

INV .• ..

T•len• IN.

Al

• I

.

In

.{. * ..• .

'.:

n

7.

...........,

1 .‘‘ \ . : :

..,:...:

1,ii:

. . .....

3.

.

-

: : .,

- •"•••

. • •

. .:Z k7ft ," •

/ Nola :

. n•

I - •

''•\ ft1;6.114 • .'••

.....‘..... anI nn

'

...." ...'.- - ' .

.

î.

1

!

I

I

. .. . i,

1.

'''

n

,

...

".

'

,

' • ..

ç'i

Map 7.

Topographical features and section boundaries in Arizona

(Sellers and Hill, 1974). Note:

Hecht and Reeves (1981) classify the northern regions as Colorado Plateau, the central region as transition, and southern as Basin and

Range.

LATE CENOZOIC AND RECENT SEDIMENTS

(LARGELY UNCONSOLIDATED)

-

CRYSTALLXVE METAMORPHIC

AND IGNEOUS COMPLEX

COMPILED BY

MARK A. MELTON

Map

8.

Surface rock types (Hecht and Reeves,

1981).

137

44 P840

'or

FORESI

',ATTARS

Hart FORA

-

5

ANION DC SPELL,

NATO siotourroir

AVALIPAY MDR

ES

MAWR,

RACY

NONLINEM rt r

.e.

: r

MAMAS

FOOT 1001401E

AN RES

NATY. FOREST hi lh ilillilli r1,11

CANI ON

I WTI

romisENr cocorsm0

n in,

massy wart

WROLIFt REPOSE

111

AH

INTL FMCS?' l'-011TVIEGOOTAMr

I )

,

'12*(

N4ri ormuiewr

Sr

Hill

1

Wart

„iwAr

C, ,,,,,,-\ oravreztau

CASTLE l i

,.. A,

AU

a • NArt

.

dartor

SITSREA/ES

1, . il COWER

CAMP' WAVE

RONAN IITS.

ILL

FOREST -1.

Alan

411111

I

COLORADO INVER

INDIAN RES

TROIA* RES

PEIRIFTED

FOREST

*Art ft....

lk

6ILA RENO

MOAN

PCS.

'21

SAL

T NAT*

WOTAN RES.

FORT I.,

41:000FLL

INDIAN RES

I l ai

Mon ismaielvr

GIL•

ZIT;

nomfri

i

Nd

II I

TONTO

sari MISVAIMENT

SAN CARLOS

ma".

0(54

MACON

/AMIN NeS.

CASA MAWR

MUNI

WIZ NOMINENT

(CN

1

MIL

TART

LANDS

NATIONAL

FORESTS

INDIAN

RESERVATIONS

NATIONAL GAME AND WILDLIFE

REFUGES AND RANGES

M

8 NATIONAL WONUITENTS

PARTS AND RECREATION ARE AS

PAPAGO

SAGuAIRO

NAT'L

APONLAMENT

1184,11

SIN AMER

AENAN RCS

r

-

TAIWASACOM MISSION

*Arc NONAINEN re

-

Tart PORES, TCREI

SAGUARO

NAT 'L

mCIPAILIVT mum,

1:

11 er7

1

,

1 ter/

pl.L

)

PORI NOACNVCA

.1'

'ICAO?

,11

-

111

LI 1

P44TtNEMOINAl

0 10 20 30 40

50

AIM

1.11•1

WW1

NIL ES

Sow cos

A

.2

NA Of f

n

.• of P1onnmoRod Dowlopolont

PURL IC LINO OWNERSHIP IN •POZON• f97,

04040 $ .oios Dep.. fnont of .

ARIZONA:

INTN

Map

9.

Public lands

(from Hecht and Reeves,

1981).

138

139

Infrastructure

Availability of support facilities such as irrigation, transportation, processing plants, or refineries would make some areas more desirable for farm location. It is unlikely that E. lathyris growth within the state of Arizona would be enough to support a refinery of the minimum size of 50,000 T/d. Extracted crude must therefore be channeled into existing facilities. A refinery is being constructed at Mobile, approximately 35 miles south of Phoenix, on a graded road between Gila

Bend and Casa Grande. All scenarios assume that extractors supply the

Mobile refinery exclusively.

Processing plants do not exist at this time. A series of alternative placements are postulated, with the resulting industry configuration analyzed by the transportation model. Trucking is assumed to be the primary mode of transportation. Existing roadways are used whenever possible.

In addition to use of available transportation networks, it would be advantageous to locate on farm areas equipped with irrigation systems, storage sheds, and other support facilities. Prime areas for consideration would include land previously cropped but now idled. Unused lands adjacent to cultivated plots also would require less capital investment than isolated regions.

Water Needs and Availability

Optimizing the efficiency of water use will reduce costs and increase net returns. Estimates of water needed and resulting yields vary

140

(see Chapter 2). Mild stress, in terms of both quantity and quality of water, may increase hycrocarbon production. No crop was produced by a fall planting receiving 10" (25 cm) of water. With 20" (50 cm) combined irrigation and precipitation, 2.6 T/ac (5770 kg/ha) was produced.

Less than 20" of water resulted in 4 bl/ac (10 bl/ha) of biocrude over a 6-month growing period. Twenty-four to 30 inches (60 to 75 cm) has been estimated as optimal by several sources (iemethy et al., 1978;

Mendel et al., 1979) with 27.5" (69 cm) producing over 3 bl/ac (8.3 bl/ ha) (OALS, 1979).

Mendel et al. (1979) suggest that 18 to 20" (45 to 50 cm) of a total 26" (65 cm) water should be received by plants during the period of rapid growth. They consider this to be spring and summer; if fall plantings are preferable in Arizona, evaporation and moisture stress might be lessened due to cooler climates, so that water requirements might be reduced.

Availability and quality of water in Arizona is restricted. The

Arizona climatic pattern is a biseasonal regime of precipitation in winter and summer, separated by periods of drought in spring and fall. Presummer drought is most severe, since it is accompanied by high temperatures.

In desert regions precipitation is highly variable. In Arizona, winter precipitation is more variable from year to year, in terms of both amount and time of occurrence, than is summer rainfall. This is an important consideration if fall planting is recommended. Winter

precipitation is characterized by relatively slow air movement, wide-

141 spread cloudiness, and gentler rainfall.

Total precipitation increases on mountain gradients by approximately 5" for each 1000 ft increase in elevation (Lowe, 1964). At more humid elevations yearly variation is less than at lower levels. Map 10 shows average annual precipitation isohyets for Arizona, along with seasonal distribution. Very few areas of the state receive 20 "or more, and most are in the higher areas too cold for Euphorbia growth. Mean annual precipitation of 21" is associated with Ponderosa Pine forest biotic community in Arizona, and 26"with Douglas Fir forest (Lowe,

1964). Higher precipitation areas also fall within the regions that receive maximum rainfall in summer, when E. lathyris utilization would be lowest. Irrigation will be necessary in all Arizona growing locations.

There are two primary sources of irrigation water in Arizona: surface water (rivers, streams, and reservoirs), and groundwater. Since clear water is a scarce commodity in arid regions, and since E. lathyris appears to be relatively salt-tolerant, areas in which water is generally considered too saline for traditional crop use are given first consideration for development. Urban or industrial waste sources may also supply a portion of irrigation.

The major sources of surface water in Arizona are the lower Colorado, Gila, and Salt Rivers. All three are almost totally appropriated

(Hecht and Reeves, 1981). Outside of northern Arizona, only Greenlee

> 20"

Map

10. Average annual precipitation (in inches) in Arizona

(from Sellers and Hill,

1974; Hecht and Reeves, 1981).

142

County does not depend on an overdraft of groundwater for irrigation supplies (Table Cl). Map 11 shows the extent of irrigation in

143 southern Arizona, and Map 12 potential for future surface water developments. Potential dam sites are located primarily in northern Arizona, and Central Arizona Project aqueducts service Phoenix and Tucson rather than croplands.

Water generally considered to be too saline for standard irrigation might be the best source of Euphorbia cropping. Map 13 shows average salt concentrations for streams. Most of the stream water in southern Arizona is high or very high in salinity. Sodicity does not appear to be a problem. The projected Yuma desalination plant will treat approximately 100,000 ac-ft of effluent each year. One third of that will be concentrated brine (University of Arizona, 1981). Seawater from the Gulf of California might also be usable in that area. E.

lathyris response to salinity is discussed in Appendix A. After germination, plants should be able to tolerate concentrations in the C-3 category or better. All but one of the wells in Map 14 would be usable.

Quantity and quality of water are again important factors with regard to groundwater used for irrigation, but a third element, depth to water table, is added. Costs of pumping water to surface crops from low and continually falling levels can prohibit irrigation.

Most current pumpage occurs in the area around Phoenix and midway between Phoenix and Tucson. South of Tucson and near Yuma there has been significantly less removal (U.S. Geological Survey,

Water

Table

Cl.

Annual sources and consumptive use of water

(1970 estimates for "normal" conditions).

Area

Source Consumption

Dependable a Groundwater b

Supply

Overdraft Total

Agriculture

(1000 ac-ft)(1000 ac-ft) (1000 ac-ft) (1000) % Total ac-ft

Northern Arizona

Apache County 17

Coconino County 14

Gila County

Navajo County

Yavapai County

19

44

22

Totals 116

0

0

12

0

0

12

17

14

19

44

34

128

14

75

9

2

26

24

82%

64%

11%

59%

71%

59%

Southeastern Arizona

Greenlee County 33

Graham County

132

Cochise County

85

Totals 250

0

27

268

295

33

159

353

545

17

157

335

509

52%

99%

95%

93%

Southcentral Arizona

Santa Cruz County 5

Pima County

72

Pinal County 254

Maricopa County 971

Totals 1,302 1,797

8

267

620

902

13

339

874

1,873

3,099

11

211

830

1,681

2,733

85%

62%

95%

90%

88%

Western Arizona

Mohave County

Yuma County

Totals

67

1,086

1,153

State Totals

2,821

5

(79)c

(84)c

2,188

72

970

1,042

4,814

23

954

977

4,294

32%

98%

94%

89%

(Source: Arizona Water Commission, 1975; Hecht and Reeves, 1981) a.

Surface diversions plus estimated natural groundwater recharge, less legally required return flows to surface supplies. Dependable supplies in Northern Arizona probably exceed table values by small amounts.

b.

Net values, representing total pumpage less groundwater recharge occurring subsequent to use of either surface or groundwater supplies.

C. Values for overdrafts in Yuma County represent overdrafted area only.

Surplus conditions elsewhere in Yuma County are not included.

COLORADO RAW, INDAN IRRWATiO* PROJECT

000

000

000

0`

DOD

000

DOD

Diversion dom -2-

0*:

Dv=

IDDIDDEsoMMIMEsr emoMsel

n. rtirs ea aan taa.

ewe Mee ,.9t0000lS 100.000 awn fee t)

0 M! 20 50 00 50

MILES

Sorre•• Airar•

••••••n

*I Woes AI...

100 .1969

»as. NW.

Cosa.maiso

...••n

1••

n

•• ,

M.%

•I

A

n

.1•011

1974

Map 11. Croplands and water use

(Dutt and McCreary,

1970;

Hecht and Reeves,

1981).

145

LOAO

Possal

Pos., Dos SIR

-

7.4.

lb Noses,

Gsnosofoop and Rows. • yy

.

Stab.

o /

Merlbsa

Lois

_Cocoons; Dom Soy

(

' )tt

'

.

\ s

'` .f

-'--

'.1

._ ',''''," ' •

"III

2....4...

„ly

----, \

I ' '

1

-. .

Ifflkins

Dom S,).

.. \,,-;-'___.) (..." 1 l ost R•sssys: i rk

\

\

D.

...,

GRANITE

1 -- _,,

1

, -

Ll

) ..

II .\,

.......„

11

VP

...

,

'''1_»..--'1 . -

l

',

-

0

(„\\

PRESCOTT

,,,,,,„ ^‘‘ , ,

_.‘„:,.:._:.f

j-1.

,

0

''

and

I•

/

••• \ 04 .:” _ _ __ .„ j L

L

— T1

' 1 -

Crosby Cross

. .

Dom Silo

'.."--"' ersl lissessob

-- - , Rwmpson

Ranch .- •

,,

: .... -

n

'' „

'S:-..A.

5COTTSOS1_,0%... d.m

.-

. 53',..-, ....„

[I

L--

and Rousso,' ...• _;

'

.. L.,_

...

\.,

Dulles Dom 5,3.

and

Ressrsob slsCP

.

( -

'-

r

\ConsIsbocIr(

OAT,

Sits osS Resets.,

-'r

0 10 20 30 40 50

MMIM:7= n

MILES

CENTRAL ARIZONA PROJECT

(UND

(

R CONSTRUCTION)

®

PUMPING STATION

THERMAL GENERATING

POWERPLANT

SUBSTATION

<7.

DAM SITES AND RESERVOIRS

OTHER PROPOSED

DAM SITES AND

RESERVOIRS

C7

SOILS SUITABLE FOR

IRRIGATION

SOURCES ARIZONA BuREAu Of

MINES, MINERAL AND MATER

RESOURCES GE ARIZONA (MIL

LET1N ISO), 1 9 69, LONER Iona opp

55 1 COLORADO I

, y_Ejt W

-

PRENEMS14 FRAmEV/ORK STLDT,

APPEND°, X. 3 970

Map

12.

Potential surface water developments (Hecht and Reeves,

1981).

146

147

Resources Division, and Arizona Water Commission, (USGS), 1973). Potential well production in southern Arizona varies from 0-10 gallons per minute (gal/min) to more than 2500 gal/min (USGS, 1978).

Areas under 200 ft depth to water are indicated on Map 15. Map

16 indicates changes in groundwater levels that have occurred as a result of pumping. As of 1972, the water table decline in most areas has been less than 50 feet, but substantial acreage near and south of Phoenix has lost over 100 feet. In addition to pumpage costs, possible land subsidence is a deterrent to further water removal in these areas.

Near Yuma, a slight rise in water table has resulted from heavy surfacewater irrigation. Irrigation, coupled with excess water added to leach soluble salts, has resulted in some areas requiring drainage pumping to maintain the water level below the root zone. Removal of saline "tail water" is accomplished through a drainage canal into Mexico (Hecht and

Reeves, 1981). Tailings and effluent in the Yuma area might provide an adequate irrigation supply for E.

lathyris

cropping, as long as obligations to Mexico for Colorado River allocation can be met.

Pennington (pers. comm., 1982) listed four areas where saline water is available and not currently used for irrigation.

1. Wellton/Mohawk constitutes the primary region where water remains unused. There is little incentive to develop new water supplies since the water table is high and drainage pumping is necessary.

2.

West end of the Safford area.

148

3.

Maricopa County near Buckeye, where Phoenix effluent would also be available. The latter is less salty than ground water, and would not be suitable for food crops.

4.

Papago Farms area, with mildly saline water.

The largest source of saline water in the state is Blue Springs, in the northeast (University of Arizona, 1981). A large area of Pima,

Maricopa, and Yuma Counties contain ground water with salinity in the

1000 to 3000 ppm range, with some portions above 3000 ppm. The eastern and southeastern regions contain a number of smaller deposits with concentrations to 3000 ppm. Maps 17 and 18 show distribution of dissolved solids and suitability for traditional irrigation use.

The consolidation of possible areas and resulting acreage estimates are discussed in Chapter 4.

Concentration, in milligrams per liter

Map 13. Average concentrations of dissolved salts in Arizona's streams (from USGS, 1973).

149

'''••

c4-$I

MOH AVE

\

rj

• .

., .--1:- LAKE

MEAD

C3-S

_

1

Lower Main

Stem

r

i

LAKE

1 MOHAVE

%.-........" )

\

J

" * " ....j.‘ "FLAGSTAFF

C2

-

31

-

11\ z

0

NAVAJO

; CO 10pd0

APACHE

A,-S1

447:

II

P

_II

LAKE

HAVASU

1/../I.

o

. A

-.. ,,,

. .....•

ArSCAVOM,

r

C3-S1 C3-S1

. f

"

... ..... i°

..4..

--r-r

f

.• . P .

n

. '

C2-3I

)

!

)

I

YAVAPA

I

*-

,•••(

,?

11

N

C2-38

••

V.

-;- 1

A. '

-

G 1 L

A

•-•-•

_ __t _.t.

7,

. .;

_ •

CI-SI

.....

!it

_____

MARICOPA

3- SL

r

. ROOSEVELT

rn

1

LAKE

'S'eL;

Gila

Ji

SAN CARLOS

C3-St

LAKE

r;.

f

C3-

Sl

\ r/

G

'

'

\

ft

• I

• Ai

150 a.

-**"- Region Boundary

— Subregion Boundary

11111 Irrigated

Land

C4-SI C4-52 C4-

$

3

b.

C2-$I

LOW !MEDIUM

SODIUM (ALKALI) NAZAR°

3

NIGH

NIGH

151

Map

15.

Average depth to groundwater (Hecht and Reeves,

1981;

Lower Colorado Region, 1971, Appendix V).

152

'‘ n

-

-

H

,

)

?; • !::

J

,

---

/

!

‘-,

Ri

SE (in feel/

Generoay less than

10

I

DECLINE Ion feel/

Less thon 25

F373 51 - 10 0

Greater

thon

150

25-50

101-150

EL 1 NO DATA or DATA NOT

ANALYZED

„ ---,

-,--',. r

/.--is-- ----,

I

I

- --- - -

)

1

L, r

-

Map

16.

Change in groundwater levels,

1940-1972

(Hecht and

Reeves, 1981).

153

1000 to

3000 milligrams/liter

> 3000 milligrams/liter

Map

17.

Distribution of dissolved solids in groundwater (Lower

Colorado Region, 1971,

Appendix

XV; Feth et al.,

1965).

154

155

Map

18.

Suitability of groundwater for development of irrigation water supplies (Arizona Water Commission,

1975, phase I).

In areas of known potential for development, the depth to water is less than

1,000 feet, the water contains less than

1,000 mg/1 of dissolved solids, and most wells are capable of yielding

1,000 gpm.

APPEND

IX

D

ENERGY

BALANCE

156

157

The amount of energy input used up in the production of energy has been a concern in some energy analyses. Net energy studies are conducted to determine whether the balance is positive. Net energy production is desirable, but not essential. In some cases, the energy used is in a form usually not productive, and conversion to biofuels results in readily usable fuel. Burning of bagasse to generate electricity falls into this category, but use of petroleum products to run farm machinery or transport bales does not.

Energy production is proportional to the percentage of extractables available in the feedstock. Since this may be reduced as biomass increases, it is possible that net energy may be inversely proportional to biomass yield. E. lathyris has a relatively high biomass yield and a relatively low percentage of extractables (Kingslover, 1982). The energy produced from burning bagasse for electricity may therefore exceed biocrude energy.

The extent to which net energy is produced will depend largely on irrigation requirements; up to 70% of energy used in agriculture in

Arizona is for irrigation (University of Arizona, 1981). The remainder is used in equipment manufacture and operation, fertilizer, and farm upkeep and operation. Distances between farms and extraction facilities and between extractors and refinery, total acreage needed to support production, mechanical drying, and the efficiency of conversion will be additional factors important in energy use.

158

The flow of materials and energy in cropping and processing is presented in Figure Dl. Tables D1 through D4 show gross energy requirements, product, and net balance. Although Figure D1 projects net energy production excluding bagasse, calculations in Table D4 show a necessity for utilizing or selling bagasse for electricity production in order to achieve a positive energy balance. It should also be noted that transportation requirements are not included in any totals.

Direct

Solar

12.7

Solvent Makeup

0.1

Product

H

80T

2.7

Manufacturing

Sector

(Materials, Machinery,

Solvents, etc.)

A

159

Agriculture

A,

B

Energy Trans formation Sector

3.2 (Liquid

Fuels and Elect.)

A

1000T:15.9

Feed

Preparation

0.07

Process Feed

(Solvent

Extraction)

E, F

Bagasse

468T

6.7

Sold

192T

2.8

0.2

Sugars

260T

3.7

Figure Dl.

Materials

(Calvin,

1980). -dry T/day

and

energy flows

for a

biofuel industry

1980;

SRI,

1979; McLaughlin

and Hoffman,

Based

on

10 bl/ac/yr, 33,000

acres,

1,000

;

figures are in

109 BTU.

Table Dl. Energy balance totals for a biofuel industry.

Inputs

Outputs

Net

a

3.6

9.2

5.6

652

2674

2022

4513

5822

1309

a.

From Calvin,

1980;

in

10 9

BTU; excludes solar inputs, recycled bagasse, and transportation.

b.

From SRI,

1979;

in

10

12

joules; excludes solar, bagasse, and transportation.

c. From McLaughlin and Hoffman,

1980;

in

10

12

joules; includes bagasse.

160

161

Table D2. Gross energy requirements for a biofuel industry (McLaughlin and Hoffman, 1980; OALS, 1979). -- Based on 40,600 ac, 9T/ac,

8% extract, producing 5.6 bl/ac, to supply 1000T/day processing facility 365 days.

10

12

Joules

Energy Requirements

Cropping A, B

Farm electricity

Machinery

Irrigation system

Stand establishment (fuel)

Ancillary inputs

Irrigation electricity

Harvest (fuel) C

Transportation D

Subtotal

Processing F

Capital (steel, concrete, machinery, maintenance)

Milling

Flaking (electric)

Extraction

Air systems

Drying systems

Conveying

Receiving/dust control

Cooling water

Process steam

Solvent loss (1 gal/T/day)

Subtotal

Total

78

31

174

11

255

3110

146

5

3810

12

368

27

4

19

46

60

8

41

55

64

704

4514

162

Table

D3.

Extraction

facility energy

balance

(F)

(SRI,

1979).

Million

BTU/hr %

Inputs

Total field-dry Euphorbia

Outputs

Product oil

Plant

residue

and

entrained oil

Heat rejected to cooling

Oil lost to water

phase

Stack losses

Miscellaneous losses

Total

2036

345

1276

359

5

34

17

2036

100

16.9}

thermal

efficiency

ratio

3.7:1

62.7

17.6

0.3

1.7

0.8

100.0

163

Table D4. Total energy budget for a biofuel industry: 1000 T/d in 10 12 joules (McLaughlin and Hoffman, 1980).

Using Bagasse to

Generate Electricity and Steam

Gross Energy Requirements

Agricultural

A, B, C

Processing

E, F

Total

Plant

Energy Yield

Oil

Bagasse

Total

Net

Yield

Oil Only

Total

3810

703

4513

862

4960

5822

-3651

1309

622

30

652

862

1812 (excess)

2674

210

[2022]

REFERENCES

Alexander, Alex G., 1981. "Production of Nonwoody Land Plants as a

Renewable Energy Source," in Klass, Donald L., ed, Biomass as a Nonfossil Fuel Source, American Chemical Society, Washington,

D.C. pp. 49-76.

Arizona Water Commission, Phoenix, 1975. Arizona State Water Plan,

Phase I: Inventory of Resource and Uses. 224 p.

Arizona Water Commission, Phoenix, 1977. Arizona State Water Plan,

Phase II: Alternative Futures. 145 P.

Bagby, M. O., R. A. Buchanan and F. H. Otey, 1981. "Multi-Use Crops and Botano-chemical Production," in Klass, Donald L., ed.,

Biomass as a Nonfossil Fuel Source, American Chemical Society,

Washington, D.C. pp. 125-136.

Barr, James L. and David E. Pingry, 1977. "An Economic Analysis of the Central Arizona Project," Hydrology and Water Resources in

Arizona and the Southwest, proceedings of the Arizona Water

Resources Association, Arizona Section, Vol. 7. 13 p.

Benemann, John R., 1980. "Biomass Energy Economics," The Energy

Journal 1(11:107-131.

Buchanan, R. A. and F. H. Otey, 1978. "Multi-use Oil and Hydrocarbon

Producing Crops in Adaptive Systems for Food, Material, and

Energy Production," presented at International Conference on

Larrea: A Vast Resource of the American Deserts, Saltillo,

Coahuila, Mexico, August 9-11, 1978.

California Energy Commission, Development Division, Synthetic Fuels

Office, 1979. Low Cultivation Energy Crops Grown on Land

Unsuitable for Agricultural Production, Draft Staff Report

167A:09 R2 4/27/79. 44 p.

Calvin, Melvin, 1977a. "Energy and Materials via Photosynthesis," in

Buvet, R., M. J. Allen and J.-P Massuê, eds., Living Systems as Energy Converters, North Holland Publishing Co., New York.

pp. 231-259.

Calvin, Melvin, 1977b. "Hydrocarbons via Photosynthesis," Energy

Research 1:299-327.

164

165

Calvin, Melvin, 1977c. "Materials and Energy from the Sun," Sunworld

5:2-6.

Calvin, Melvin, 1977d. "The Sunny Side of the Future," ChemTech

7(7):

352-363.

Calvin, Melvin, 1978a. "Green Factories," Chemical Engineering News

12:30-36.

Calvin, Melvin, 1978b. Petroleum Plantations, prepared for Agricultural

Chemical Society of Japan, Nagoya. 37 p.

Calvin, Melvin, 1980. Hydrocarbons from Plants: Analytical Methods and

Observations, presented to the American Association for the

Advancement of Science, San Francisco, California, January 4,

1980. Various pagings.

Charries, A. and W. W. Cooper, 1961. Management Models and Industrial

Applications of Linear Programming, Vol. I, John Wiley and Sons,

Inc., New York. 467 p.

Cheney, Eric S., 1974. "U.S. Energy Resources: Limits and Future Outlook," American Scientist 62:14-22.

Chopra, Ram N., S. L. Nayar and I. C. Chopra, 1956. Glossary of Indian

Medicinal Plants, Council of Scientific and Industrial Research,

New Delhi, India. 329 p.

Doane Agricultural Services, Inc., St. Louis, Mo., 1981. Projected

Costs and Returns for Farmers Producing Euphorbia Lathyris, unpublished manuscript for Diamond Shamrock Corporation, Painesville, Ohio. 22 p.

Dutt, Gordon R. and Ted W. McCreary, 1970. The Quality of Arizona's

Domestic, Agricultural, and Industrial Waters, University of

Arizona, Tucson, Agricultural Experiment Station, Report

#256.

83 p.

Farm Chemicals, 1979. "Take a Good Look--These May Be Our Next Major

Crops," Farm Chemicals, February 1979. pp. 76-84.

Feth, J. H. et al., 1965. Preliminary Map of the Coterminous United

States Showing Depth to and Quality of Shallowest Ground Water

Containing More than 1000 Parts per Million Dissolved Solids,

United States Geological Survey, Washington, D.C., Hydrological

Investigation Atlas, HA-199. 31 p. and 2 sheets.

Flug, Marshall, Jorge E. Sandoval-Illescas and Delmar D. Fangmeier,

1981. "Response of Euphorbia lathyris to Soil Moisture," in

University of Arizona, Tucson, Office of Arid Lands Studies,

Liquid Fuels from Green Plants. 25 p.

166

Fontes, Miguel and Carolyn Watson, 1978. Progress Report on Hydrocarbon

Yielding Vegetation: Agronomic/Horticultural Efforts, University of Arizona, Tucson, Environmental Research Laboratory. Various pagings.

Foster, Kennith E. and William H. Brooks, 1981. Assessment of Arid

Lands Plants as Future Energy Crops for the Electric Utility

Industry, Electric Power Research Institute, Palo Alto, California, Engineering and Economic Evaluations Program, Advanced

Power Systems Division, Special Report AP-2172-SR. Various pagings.

Foster, Kennith E., R. Leslie Rawles and Martin M. Karpiscak, 1980.

"Biomass Potential in Arizona," Desert Plants 2(3):197-200.

Fox, Roger and Edward Jorgensen, 1980. Production Cost Estimate of

E. lathyris, University of Arizona, Tucson, Department of

Agricultural Economics. 5 p.

Hale, Ellen M., ed,, 1976. Arizona Energy: A Framework for Decision,

The University of Arizona Press, Tucson, Arizona. 160 P.

Hathorn, Scott, Jr. and James F. Armstrong, 1981. 1981 Arizona Field

Crop Budgets: Pima County, University of Arizona, Tucson,

College of Agriculture, Department of Agricultural Economics.

55 p.

Hathorn, Scott, Jr. and James F. Armstrong, 1982. 1982 Arizona Field

Crop Budgets, University of Arizona, Tucson, College of Agriculture, Department of Agricultural Economics. Various pagings.

Hecht, Melvin E. and Richard W. Reeves, 1981. The Arizona Atlas, University of Arizona, Tucson, Office of Arid Lands Studies. 164 p.

Hinman, C. Wiley, Joseph J. Hoffman, Steven P. McLaughlin and Timothy

R. Peoples, "Hydrocarbon Production from Arid Land Plant

Species," in Franta, Gregory E. and B. H. Glenn, eds., Proceedings of American Section of International Solar Energy Society

1980. pp. 110-114.

Johnson, Jack D. and C. Wiley Hinman, 1980. "Oils and Rubber from Arid

Land Plants," Science 208:460-464.

Kelso, Maurice M., William E. Martin and Lawrence E. Mack, 1973. Water

Supplies and Economic Growth in an Arid Environment: An Arizona

Case Study, University of Arizona Press, Tucson. 327 p.

Kingslover, Barbara E., 1982. A Review of Research on Euphorbia Lathyris,

1979-1981, University of Arizona, Tucson, Office of Arid Lands

Studies, Submitted to Diamond Shamrock Corporation. 89 p.

167

Lowe, Charles H.,

1964.

Arizona's Natural Environment: Landscapes and

Habitats, University of Arizona Press, Tucson. 136 p.

Lower Colorado Region State-Federal Interagency Groups, Water Resources

Work Group, 1971. Lower Colorado Region Comprehensive Framework

Study, Appendix V: Water Resources; Appendix X: Irrigation and

Drainage; Appendix XV: Water Quality, Pollution Control and

Health Factors. Pacific Southwest Interagency Committee, Water

Resources Council. Various pagings.

McLaughlin, Steven P., 1981. Personal communication, Research Associate in Arid Lands, University of Arizona, Tucson, Office of Arid

Lands Studies.

McLaughlin, Steven P. and Joseph J. Hoffman, 1980. Potential of Desert

Plants to Produce Energy and Petrochemicals: An Energy Budget

Analysis, University of Arizona, Tucson, Office of Arid Lands

Studies. 17 p.

Mendel, David A., Frederick A. Schooley and Ronald L. Dickenson, 1979.

A Production Cost Analysis of

Euphorbia Lathyris,

U.S. Department of Energy, Washington, D.C., Division of Distributed Solar

Technology, Fuels from Biomass Systems Branch, SRI project 7877.

26

Mohs, Mayo, 1981. "Energy: Battle for Survival," Discover, March 1981:

96-102.

Nakao, Sasuke, ed., 1976. Tanaka's Cyclopedia of Edible Plants of the

World, Keigaku Publishing Company, Tokyo. 924 p.

National Geographic Society, Washington, D.C., 1981. "An Atlas of

Energy Resources," in Energy National Geographic Special Report,

February 1981: 58-69.

Nemethy, Esther K., John W. Otvos, Melvin Calvin and Roy M. Sachs, 1978.

"Hydrocarbons and Energy from Plants," in Proceedings of Workshop on Biomass Energy and Technology, Santa Clara, November

8-9, 1978, Electric Power Research Institute, Palo Alto, California. 13 p.

Pennington, Dean, 1981. Personal communication, Extension Specialist in Soils, Department of Soils, Water, and Engineering, University of Arizona, Tucson.

Peoples, Timothy R.,

1981a.

Gopher

Plant in Arizona, University of

Arizona, Tucson, Cooperative Extension Service, Arizona Agricultural Experiment Station, Agri-file

#288.20,

and Department of Plant Sciences Information Series

81-11. 2

p.

168

• Peoples, Timothy R., 1981b. Preliminary Production Economics of Two

Hydrocarbon-Yielding Plant Species, University of Arizona,

Tucson, Office of Arid Lands Studies. 5 p.

Peoples, Timothy R. and Jack D. Johnson, 1980. "Crop Alternatives in

Semiarid Regions," National Council for Science and Technology

(CONACT) and National Commission for Arid Lands Studies (CONAZA)

Conference on Natural Resources and Regional Development: The

Case of the Semiarid Region, Cuautla, Morelos, Mexico, October

6-8. 22 p.

Peters, Max S. and Klaus D. Timmerhaus, 1980. Plant Design and Economics for Chemical Engineers, McGraw-Hill Book Company, New York.

973 P.

Prokhanov, Y. I., 1949. "Euphorbia," Flora SSSR 14:233-378.

Rao, Ram C. and Timothy L. Shaftel, 1980. "Computational Experience on an Algorithm for the Transportation Problem with Nonlinear

Objective Functions," Naval Research Logistics Quarterly 27(1):

145-157.

Sachs, Roy M. and Tom Mock, 1980. Euphorbia Lathyris: A Preliminary

Guide to Planting, University of California, Davis, Agricultural

Experiment Station, Agronomy Progress Report No. 103. 2 p.

Sandoval-Illescas, Jorge E., 1981. Hydrocarbon Production and Growth of

Euphorbia Lathyris: A Water Use Study, University of Arizona,

Tucson, M.S. Thesis. 65 p.

Schauenberg, Paul and Ferdinand Paris, 1977. Guide to Medicinal Plants,

Keats Publishing Co., New Canaan, Connecticut. 349 p.

Sellers, William D. and Richard H. Hill, eds., 1974. Arizona Climate:

1931-1972, The University of Arizona Press, Tucson. 616 p.

Smith, A. R. and T. G. Tutkin, 1968. "Euphorbia," Flora Europea 2:

213-226.

SRI International, Menlo Park, California, 1979. Mission Analysis for the Federal Fuels from Biomass Program, Volume VI: Mission

Addendum, U.S. Department of Energy, Washington, D.C., Division of Distributed Solar Technology, Fuels from Biomass Systems

Branch. 74 p.

Srinivasan, V. and G. L. Thompson, 1973. "Benefit-Cost Analysis of

Coding Techniques for the Primal Transportation Algorithm," in

Journal of the Association for Computing Machinery 20(2):194-213.

169

Taylor, Jonathan, 1982. Personal communication, Research Associate in

Arid Lands, University of Arizona, Tucson, Office of Arid Lands

Studies.

U.S. Congress, Washington, D.C., Office of Technology Assessment, 1980.

Energy from Biological Processes, Volumes 1 and 2: Technical and

Environmental Analysis, U.S. Government Printing Office, Washington, D.C., Superintendent of Documents. 195 p., 234 p.

U.S. Department of Energy, Washington, D.C., Energy Information Administration, 1981. Annual Report to Congress, Volume 3: Forecasts,

U.S. Government Printing Office, Washington, D.C., Superintendent of Documents. 348 p.

U.S. Geological Survey,

Water Resources Division, Tucson, Arizona, and

Arizona Water Commission, Phoenix,

1973.

Water Resources Investigations in Arizona

1973. 2

sheets.

U.S. Geological Survey, Tucson, Arizona, and Arizona Water Commission,

Phoenix,

1978.

Annual

Summary of Ground-Water Conditions in

Arizona, Spring 1977 to Spring 1978, U.S.G.S. Water Resources

Investigations 78-144, Open-File Report. 2 sheets.

University of Arizona, Tucson, Office of Arid Lands Studies, 1979-1982.

Liquid Fuels from Green Plants, formerly Hydrocarbon-Yielding

Vegetation; Reports of Project Activities to Diamond Shamrock

Corporation, Painesville, Ohio. Various pagings.

University of Arizona, Tucson, Office of Arid Lands Studies, 1980.

Euphorbia Lathyris. 1 p.

University of Arizona, Tucson, Office of the Vice President for Research,

1980. Energy: Our Heritage and Our Future, A report on research for 1978-1979 at the University of Arizona. 43 p.

University of Arizona, Tucson, 1981. Arizona's Energy Future; Making the Transition To A New Mix, 39th Arizona Town Hall, October

25-28, 1981, Arizona Academy, Phoenix. Various pagings.

Valley National Bank of Arizona, Economic Research Department, Phoenix,

1981. Arizona Progress: The Future of Crop Agriculture in

Arizona. 4 p.

Weaver, Kenneth F., 1981. "Our Energy Predicament," in Energy,

National Geographic Society Special Report. pp. 2-23.

White, Alain C., R. Allen Dyer and Boyd L. Sloane, 1941. "Notes on

Euphorbia Culture," in The Succulent Euphorbieae (Southern

Africa), Volume II, Appendix E, Abbey Garden Press, Pasadena,

California. pp. 983-990.

Wright, N. Gene,

1981-1982.

Personal communication, Agricultural

Economist, University of Arizona, Tucson, Office of Arid Lands

Studies.

Zangwill, Willard I.,

1969.

Non-Linear Programming: Unified Approach,

Prentice-Hall, Inc., Englewood Cliffs, N.J.

356

p.

170

Was this manual useful for you? yes no
Thank you for your participation!

* Your assessment is very important for improving the work of artificial intelligence, which forms the content of this project

Download PDF

advertisement