Production of Jet Fuels From Coal Derived Liquids Forks Energy Research Center

R194 9 PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS V/3
VOLUMIEI MRCET ASSESSN. (U) NORTH DAKOTA UNIV ORON

FORKS ENERGY RESEARCH CENTER J1E SINOR AUG0?
r UNCLSSIFIED AFWAK-TR-8-2042-VOL-i MJPR-FY1435-SE-NW.S? F/O 21/4 M

1.01I It I-8
328 , J
IM 122. 1
LI
IlII
1.5 L -l.
If111W
MICROCOPY RESOLUTION TEST CHART

1963-A
BUREAUM1 STANDARDS
NATIONAL
%: %
1 00
FILE
AD-A 184 84-6
AFWAL-TR-87-2042
Volume I
PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS
VOL I - Market Assessment for Liquid By-Products from the

Great Plains Gasification Plant
J. E. , , ,,.oP
SINOR
D TIC
I
J. E. SINOR CONSULTANTS INC. LL T Eu
7960 NIWOT ROAD SUITE 12
NIWOT CO 80544-0649 SEP 1 61987 U
AUGUST 1987
INTERIM REPORT FOR THE PERIOD 20 January 1987 - 30 April 1987
APPROVED FOR PUBLIC RELEASE; DISTRIBUTION IS UNLIMITED
AERO PROPULSION LABORATORY

AIR FORCE WRIGHT AERONAUTICAL LABORATORIES
AIR FORCE SYSTEMS COMMAND
WRIGHT-PATTERSON AIR FORCE BASE, OHIO 45433-6563
87 I 9
NOTICE
When Government drawings, specifications, or other data are used for any
purpose other than in connection with a definitely Government-related
procurement, the United States Government incurs no responsibility or
any obligation whatsoever. The fact that the Government may have
formulated or in any way supplied the said drawings, specifications,or
other data, is not to be regarded by implication, or otherwise in
any manner construed, as licensing the holder, or any other person or
corporation; or as conveying any rights or permission to manufacture,
use, or sell any patented invention that may in any way be related
thereto.
This report has been reviewed by the Office of Public Affairs

ASD/PA) and is releasable to the National Technical Information Service
NTIS). At NTIS, it will be available to the general public, including
foreign nations.
This technical report has been reviewed and is approved for publication.
WILLIAM E. HARRISON CHARLES L. DELANEY, Actir Chief

Fuels Branch Fuels Branch
Fuels and Lubrication Division Fuels and Lubrication Division
Aero Propulsion Laboratory Aero Propulsion Laboratory
FOR THE COMMANDER
Robert D. SherlChf
Fuels and Lubrication Division
Aero Propulsion Laboratory
"If your address has changed, if you wish to be removed from our
mailing list, or if the addressee is not longer employed by your orga-
nization, please notify AFWAL/POSF, Wright-Patterson AFB, Ohio
45433-6563 to help us maintain a current mailing list."
Copies of this report should not be returned unless return is required

by security considerations, contractual obligations, or notice on a
specific document.
NAM IWw
REPORT DOCUMENTATION PAGE TOMSF8 M. 070O-OM,
Is. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS

IINri AqlFTF____
2a. SECURITY CLASSIFICATION AUTHORITY 3. DISTRIBUTION /AVAILABILITY OF REPORT
LOWNGRADING SCHEDULE Approved for Public Release;
2N.'OECLASSIFICATION Di stri buti on is Unlimited
4. PERFORMING ORGANIZATION REPORT NUMBER(S) S. MONITORING ORGANIZATION REPORT NUMBER(S)
AFWAL-TR-87-2042, Vol. I
6a. NAME OF PERFORMING ORGANIZATION 6b. OFFICE SYMBOL 7a. NAME OF MONITORING ORGANIZATION
JC T (NaPlicabl) Air Force Wright Aeronautical Lab
CONSULTANTS INC.
E SINOR Aero Propulsion Lab (AFWAL/POSF)
6c ADDRESS (City, State, and ZIP Code) 7b. ADDRESS (Cty,State, and ZIP Code)
7960 Niwot Rd Suite 12 Wright-Patterson AFB Ohio

Niwot CO 80544 45433-6563
6.. NAME OF FUNDING /SPONSORING 8b. OFFICE SYMBOL 9. PROCUREMENT INSTRUMENT IDENTIFICATION NUMBER
ORGANIZATION Air Force (If applicable) F1455-86-N0657
Wrlaht Aeronautical Labs AFWAL/POSF __
8c ADDRESS (City, State, and ZIPCode) 10. SOURCE OF FUNDING NUMBERS
PROGRAM PROJECT TASK WORK UNIT
ELEMENT NO. NO. NO CCESSION NO.
AFWAL/POSF
____ht-Ptterson AFB OH_45433-6563 63215F 2480 16 01
11. TITLE (Include Security Classification)
PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS VOL. I - Market Assessment for
liguid By-Products from the Great Plains Gasification Plant
12. PERSONAL AUTHOR(S)
FF STNOR
13a. TYPE OF REPORT i13b. TIME COVERED 14. DATE 07 REPORT (Year, Month, Day) 115. PAGE COUNT
Tnt_.rim I 3
FROM20-J-A&JZ- TO 0AR71 AUGUST 1987
16. SUPPLEMENTARY NOTATION
17. COSATI CODES 18. SUBJECT TERMS (Continue on reverse if necessary and identify by block number)
FIELD GROUP SUB-GROUP Turbine Fuel, JP-4, JP-8, Heavy Crude Oil,
21 107 Great Plains Gasification Plant, tar oil, crude
E4 0 103 ]ohenols. naphtha, Jet fuel. market analysis
BSTRACT (Continue on reverse if necessary and identify by block number)
In September 1986, the Fuels Branch of the Aero Propulsion Laboratory at Wright-Patterson
Air Force Base, Ohio commenced an investigation of the putential of production of jet
fuel from the liquid by-product streams produced by the gasification of lignite at the
Great Plains Gasification Plant in Buelah, North Dakota. Funding was provided to the
Department of Energy (DOE) Pittsburgh Energy Technology Center (PETC) to administer the
experimental portion of this effort. This report details the effort of JE Sinor Consul-
tants, who, as a subcontractor to the DOE contract with the University of North Dakota
Energy Research Center (UNDERC), determined the market potential of these by-product
streams. This report describes the market potential of the by-product streams for the
production of phenols, benzene, cresote, cresylic acids, coal tar pitch, paving
materials, carbon black, naphthalene, rubber processing oils, as well as, jet fuels.
20. DISTRIBUTION / AVAILA{:'ITY OF ABSTRACT 21. ABSTRACT SECURITY CLASSIFICATION
rl UNCLASSIFIEDUNLIMITED 0 SAME AS RPT. 0 DTIC USERS UNCLASSIFIED

22a. NAME OF RESPONSIBLE INDIVIDUAL 22b. TELEPHONE (Include Area Code) 22c OFFICE SYMBOL
W EHARnenm 513-255-660 1 AFWAL/POSF
DO Form 1473, JUN 86 Previous editionsare obsolete. SECURITY CLASSIFICATION OF THIS PAGE
UNCLASSIFIED
.42.
FOREWORD
In September 1986, the Fuels Branch of the Aero Propulsion Laboratory at

Wright-Patterson Air Force Base, Ohio commenced an investigation of the
potential of production of jet fuel from the liquid by-product streams
produced by the gasification of lignite at the Great Plains Gasification
Plant in Buelah, North Dakota. Funding was provided to the Department of
Energy (DOE) Pittsburgh Energy Technology Center (PETC) to administer the
experimental portion of this effort. This report details the effort of
JE Sinor Consultants, who, as a subcontractor to the DOE contract with the
University of North Dakota Energy Research Center (UNDERC), determined the
market potential of these by-product streams. DOE/PETC was funded through
Military Interdepartmental Purchase Request (MIPR) FY1455-86-N0657.
Mr William E. Harrison III was the Air Force project engineer, Mr Gary
Stiegel was the DOE/PETC project engineer.
Accesion For
NTIS CRA&I
ODTIC TAB 0
Uannouwced U. ... .
J,,3tlt~
alv~rl ...
....
....
......
QUALY By ...... ............

,NSi,,CTED D, , :b't~o,,l
A', I,:v C. des
iii
TANA OF OCGrUWS
Section Page
I. IN'TROetWUCrION 1
II. EXECUTIVE SUMMARY 2
III. BASIS FOR THE STUDY 7
IV. MARKET ASSESSMENTS 11
Military Jet Fuels 11
Petroleum Products 22
Crude Oil 22
Crude Oil Balance 25
Refined Products 25
Ref inery Feedstock 57
Creosote 60
Cresols and Cresylic Acids 71
Carbon Black Feedstock 78
Benzene, BTX 84
Phenol 90
Rubber Processing Oil 98
Tar Crude 101
Pitch 104
Paving CoMpounds 112
Naphtha lene 115
Miscellaneous Uses 119
V. TRANSPORTATION ROUTES AND OPTIONS 122
Regional Pipelines 122
Rai lroads 127
Truck Transport 131
Barges 132
VI. NETBACK PRICES 135
Market Prices 135
Transportation 136
Netback Prices 136
VII. PROCESSING OPTIONS 138
Refinery Feedstock 138
Disti llation 139
Pitch 141
Naphtha lene 142
Cresylic Acid 142
Caustic Extraction 142
Phenoraffin Extraction 143
Methanol/Hexane Extraction 145
Distillation Purification 146
Benzene 147
Sulfolane Extraction 147
Hydrodea lky 1at ion 149
Benzene Plus Phenol 149
VIII. PROCESSING ECONOMICS 152
Economic Parameters 152
El Paso Products Case 154
Modified El Paso Case 154
Non-Hydrotreating El Paso Case 161
Cresylic Acids Case 161
Phenols Only Case 161
TAL 0ONWUTSM (Concluded)
Sect ion
Grady/ANG Case 171

Grady/El Paso Case 171
Grady Non-Hydrot teat ing Case 171
Grady/El Paso Non-Hydrotreating Case 180
Benzene Only 180
BTX Only Case 180
Benzene Plus Phenol Case 188
Dynaphen. 11 Case 188
Sumur of Economic Results 192
Sensitivity to Transportation Costs 192
Sensitivity to Product Prices 192
Risk Analysis 198
IX. OUTLOOK '1! 1995 201
World Trends 201
Organic Chemicals 201
Petroleumn 202
Outlook for SIC Industries Using Great Plains Products 205
X. DEFINITION OF OPTINIM PRODUCT' SLATE 210
XI. CONCLUS IONS AND RECOMMdENDATIONS 214
vi
LIST OF ILUW ITINM
Figure Title Page
1 Air Force Bases in Great Plains Region 20

2 Great Plains Region Crude Pipelines 23
3 Great Plains Region Product Pipelines 28
4 Jet Fuel Demand 38
5 Aviation Gasoline Demand 38
6 Motor Gasoline Demnd 39
7 Distillate Fuel Demand 39
8 Kerosene Demnd 40
9 LPG Demand 40
10 Asphalt and Road Oil Demand 41
11 Residual Fuel Demand 41
12 Other Petroleum Products Demand 42
13 Total Petroleum Products Demnd 42
14 Petroleum Addinistration for Defense Districts 56
15 Creosote Production 64
16 Creosote Prices 66
17 Wood Preservers Using Creosote 66
18 Wood Preservative Consunption 69
19 Treated Crosstie Insertions 69
20 Creosote and Oil Prices 70
21 Cresols Production 72
22 Cresylic Acid Prices 72
23 o-Cresol Prices 74
24 p-Cresol Prices 74
25 m-Cresol Prices 75
26 m,p-Cresol Prices 75
27 Carbon Black Oil Prices 82
27a Locations of Carbon Black Producers 82
28 Benzene Production, All Grades 86
29 Benzene Production, Coke Ovens 86
30 Benzene Prices 87
31 Phenol Production 92
32 Phenol Prices 94
33 Ultimate Markets for Phenol 94
34 Crude Coal Tar Production 102
35 Producers of Tar and Tar Crudes 102
36 Pitch of Tar Production 106
37 Coal Tar Pitch Prices 106
38 Coal Tar Pitch Prices 106
39 Aluminum Smelter Locations 109
40 Coal Tar Pitch Producers 109
41 Naphthalene Prices 116
42 Percent of Total United States Value Added by Manufacturing 123
43 Great Plains Region Products Pipelines 123
44 Major Railroads in Great Plains Region 128
45 Major Railroad Routes for Great Plains Products 128
46 Inland Waterway Transportation System 133
47 Unicracking Process 138
48 Atnspheric and Vacuum Tar Distillation 140
49 Tar Acids Extraction With Caustic 144
vii
Kr
LIST OF ILLWSTRATICNS (Concluded)
Figure Title Page
50 Phenoraffin Extraction Process 145

51 Fractionation of Phenols 146
52 Sulfolane Extraction of Aromatics 148
53 LT Unibon Hydrotreating Process 150
54 LIOL Hydrodealkylation Process 150
55 Dynaphen Process 151
56 El Paso By-Products Schene 155
57 El Paso Case Economics 155
58 Modified El Paso By-Products Schem 158
59 Modified El Paso Economics 158
60 Non-Hydrotreating El Paso Scheme 162
61 Non-Hydrotreating El Paso Economics 162
62 Cresylic Acids Scheme 165
63 Cresylic Acids Economics 165
64 Phenols Only Schem 168
65 Phenols Only Economics 168
66 Grady/ANG Schem With Hydrotreating 172
67 Grady/ANG Economics 172
68 Grady/ANG2 Economics 176
69 Grady/ANG Schem Without Hydrotreating 177
70 Non-Hydrotreat ing Grady/ANG Economics 177
71 Non-Hydrotreating Grady/ANG2 Economics 181
72 Benzene Process Options 184
73 LITOL Benzene Economics 187
74 Unibon Plus Sulfolane Economics 187
75 Dynaphen Benzene Plus Phenol 191
76 Dynaphen II Benzene Plus Phenol 191
77 Sensitivity to Product Transport Costs 196
78 Sensitivity to Product Prices 196
79 Risk Analysis Results on DCFROR 199
80 Risk Analysis Results on Net Present Value 200
81 Inflation Corrected Coal and Oil Prices 204
82 The Coal-Oil Differential 204
viii
LIl OF TAM=
Table Title Page
1 Liquid By-Products from the Great Plains Plant 7

2 Typical Analysis for Liquid By-Products 8
3 Component Analysis for Liquid By-Products 9
4 Potential Tar Acids Production 10
5 Primary Military Aviation Fuel Specification 14
6 Typical Hydrocarbon Composition of Jet Fuels 16
7 Properties of C-10 Hydrocarbons 16
8 High Density Jet Fuel Production 18
9 October 1986 Bid Prices for JP-4 to Air Bases 20
10 Production of Crude Oil in Great Plains Region 23
11 Great Plains Region Crude Production 24
12 1986 Great Plains Region Refinery Capacities 26
13 Great Plains Region Crude Oil Balance 27
14 1986 Refined Product Inport/Export Balance 28
15 Refined Product Demand in Great Plains Region 30
16 1985 Great Plains Natural Gas Plant Operations 32
17 Refinery Operating Rates in Great Plains Region 34
18 Annual Yields From a Barrel of Crude Oil 35
19 Overall Refined Product Balance 36
20 Average Jet Fuel Properties 44
21 Average Unleaded Gasoline Properties 46
22 Summry of Requirements for Diesel Fuel 47
23 Typical Properties of Diesel Fuels in Great Plains Region 48
25 Specifications for Residual Fuel Oils 50
26 Average No.6 Fuel Properties in the Great Plains Region 51
27 Sales for Resale Prices of Unleaded Gasoline 53
28 Sales for Resale Prices of Kerosene-type Jet Fuel 53
29 Sales for Resale Prices of Kerosene 54
30 Sales for Resale Prices of No. 1 Distillate 54
31 Sales for Resale Prices of No. 2 Distillate 55
32 Sales for Resale Prices of Residual Fuel 55
33 Catalytic Hydrotreating Capabilities of Refineries in Great Plains
Region 59
34 AWPA Standard for Coal Tar Creosote 61
35 AWPA Standard for Coal Tar Creosote in Marine Applications 62
36 AWPA Standard for Creosote Solutions 63
37 Specific Gravity of AWPA Creosote Grades 64
38 Creosote-Treated Wood Products 67
39 Typical Carbon Black Oil Specifications 79
40 Major United States Carbon Black Producers 81
41 Specifications for Camrcial Grades of Benzene 85
42 Phenol Specifications 91
43 United States Production of Phenolics 92
44 Major Producers of Synthetic Phenol 96
45 United States Producers of Natural Phenol 96
46 Specifications for Rubber Processing Oils 99
47 Purchasing Specifications for Aromatic Processing Oil 100
48 Properties of Pitch Binders Used in Anodes 105
49 Locations of U.S. Aluminum Smelters 108
50 Locations of Coal Tar Pitch Producers 110
ix
LIST OF TABLES (Concluded)
Table Title Page
51 United States Consumption of Aluminum 110

52 Comparison of Road Tar and Petroleum Asphalt 113
53 Comparison of Physical Properties of Road Tar 114
54 Estinated U.S. Naphthalene Consunption 117
55 U.S. Producers of Chemical Naphthalene 118
56 Specifications for Coal Dust Suppressants 120
57 Antiknock Performance of Ethers 121
58 Distance from Beulah by Railroad and Highway 124
59 Williams Product Pipelines in Minnesota 126
60 Key Estirates of Rail Transport Costs 129
61 Costs for Railroad Shipment of Liquid Products 130
62 1987 Netback Product Prices to Great Plains 137
63 Product Slate for El Paso Case 156
64 Capital Investment for El Paso Case 156
65 Operating Costs for El Paso Case 157
66 Product Slate for Modified El Paso Case 159
67 Capital Investment for Modified El Paso Case 159
68 Operating Costs for El Paso Case 160
69 Product Slate for Non-Hydrotreating El Paso Case 163
70 Capital Investment for Non-Hydrotreating El Paso Case 163
71 Operating Costs for Non-Hydrotreating El Paso case 164
72 Product Slate for Cresylic Acids Case 166
73 Capital Investment for Cresylic Acids Case 166
74 Operating Costs for Cresylic Acids Case 167
75 Product Slate for Phenols Only Case 169
76 Capital Investment for Phenols Only Case 169
77 Operating Costs for Phenols Only Case 170
78 Product Slate for Grady/ANG Case 173
79 Capital Investment for Grady/ANG Case 173
80 Operating Costs for Grady/ANG Case 174
81 Product Slate for Grady/El Paso Case 175
82 Capital Investment for Grady/El Paso Case 175
83 Operating Costs for Grady/El Paso Case 176
84 Product Slate for Grady/ANG Non-ydrotreating 178
85 Capital Investment for Grady/ANG Non-Hydrotreating 178
86 Operating Costs for Grady/ANG Non-Ilydrotreating 179
87 Product Slate for Grady/El Paso Non-ydrotreating 181
88 Capital Investment for Grady/El Paso Non-Hydrotreating 182
89 Operating Costs for Grady/El Paso Non-Hydrotreating 183
90 Economic Parameters for Benzene Only Case 185
91 Economic Parameters for BTX Case 186
92 Economic Parameters for Dynaphen Case 189
93 Economic Parameters for Dynaphen II Case 190
94 Sumnry of Economic Results 193
95 Sensitivity to Product Transport Costs 195
96 Sensitivity to Product Prices 197
97 Results of Economic Risk Analysis 198
98 Process Rankings 212
x
ImTriON
The University of North Dakota Energy Research Center requested J. E. Sinor

Consultants Inc. to carry out a marketing analysis for liquid by-products from
the Great Plains coal gasification plant at Beulah, North Dakota. These liq-
uid streans are currently burned as plant fuel within the plant. At the same
time, lignite fines, which cannot be accepted by the gasifiers, must be sold
to the adjoining power plant. If these fines or another inexpensive replace-
ment fuel could be utilized for plant fuel, then the liquid by-product streams
could be processed as necessary and sold into higher-value mnrkets.
The objective of this study was to identify those products which would yield
the highest net return to the facility owners, after paying capital and
operating costs of any new processing units needed. The scope of the study
included the definition of rmrket volumes, product specifications, mrket
locations, prices, transportation costs, and estimated manufacturing costs.
Factors to be considered in the inplementation of a by-products nmnufacturing

and marketing scheme include the Internal transfer cost which must be assigned
to the raw liquids, long distances to some markets, and questions about the
ability of the liquid products to meet product specifications. In addition,
recent large swings in world oil prices have cast doubt on the ability of some
coal-derived products to copete with petroleum-derived products in the near
future. A reconmended course of action is developed in light of constraints
and uncertainties.
Although the funding for this study was derived from an overall investigation
of the feasibility of producing military jet fuels from the Great Plains liq-
uid streams, the scope of the study was not limited to a specific product
slate involving jet fuel. The economic calculations assumed that the entire
liquid streams could be made available for any product of interest.
,.. ,-.-
-- --- . ,€, , ,- ',-'.-...-.. ,.- .%. ., , .,, .-
, - ....-.--. :.-
EXECUJTIVE 51&ARY
FM A TUIICAL STAINPOINT, TWE VARIETY OF PRODUS MIICH COULD

BR U FROM THE GREAT PLAINS LIQUID BY-PRODUCTS STHEA IS ALMST
UIIMITED. The coal tar refining industry has had a long history.
Such products as jet fuel were produced in the 1940's and 1950's.
However, there has never been a large-scale coal refining industry
in the United States based on low-tenperature tar products, such
as the tar oil from the Great Plains coal gasification plant.
THE TAR OILS PRIXNJUH IN THE GREAT PLAINS PLANT HAVE SOW SIG-
NIFICANT DIFFERENCES FOM THE MATEIIALS PRODUCED AS COKE OVEN BY-
PRODUCTS. This creates a handicap for Great Plains products in
certain markets. Creosote produced at Great Plains, for exanple,
should be an effective wood preservative, but it is barred from
major markets because of existing specifications. Penetrating
these markets will require changing the specifications.
TIE GREAT PLAINS REGICN IS BASICALLY SELF-SUFFICIEINT IN REFINED

PKlrTOLUM PRODUCTS. Any new production of materials such as jet
fuel or gasoline would have to displace existing production in the
region. Because the Great Plains output would be less than one
percent of regional demand, any effects on prices will be minimal.
THERE IS A LAE 31(11GH DEAN FOR MILITARY JET FUEL IN THE IN-
M)IATE REGION 70 ABSORB ANY PRODUCTI(O FROM GREAT PLAINS. A con-
tract to produce jet fuel for the Air Force bases in North Dakota
should be a prime objective for the Great Plains plant.
ONLY ONE REFINERY IN TE REGION WILL HAVE THE CAPABILITY TO ACCEPT

GREAT PLAINS TAR OIL AS A FEEDSTOCK. This is the refinery in
Regina, Saskatchewan, which is building a 50,000 barrel per day
heavy oil upgrading facility.
THE ACCURACY OF THE INFORMATION AVAILABLE FOR EVALUATING THE

ECONOMICS OF UPGRADING THE LIQUID BY-PRODUCT STREAMS IS INSUFFI-
CIENT FOR MAKING INVESTMENT DECISIONS AT THIS TIME. This is
partly due to the fact that there has never been a large, modern
low-tenperature tar refinery in operation in this country. As a
result, it is not feasible to give definitive answers about the
economics. Results in this study are order-of-magnitude only.
A MAJOR DEFICIENCY IN THE COST DATA AVAILABLE IS A RELIABLE ES-

TI TE FOR THE COSTS OF EXTRACTING ALKYLPHEIOLS FROM THE CRUDE
PHENOL AM) TAR OIL STREAMS. The calculated return on investment
differs by a factor of two when different cost estimates are used
for the extraction step.
RZDDVERY OF THE PHENOLS IN A PURE STATE IS THE KEY TD A PROFITABLE

BY-PRODUCTS OPERATION. These materials do not have to conpete
directly with large-volume petroleum refinery operations, as would
be the case for gasoline or other fuel products.
2
WE M WILOGOK FR, BU AND PHENOL 18 POSITIVE. Demand for
light aromatics as octane-improvers will be strong because of the
total elimination of tetraethyl lead for this purpose. At the
same time, chemical demand is starting to recover from a several-
year slunp.
$LL INKE VOLUS CONSTITUlE A RISK FR SOM FQINTIAL BY-

PERUCTS. Maximum creosote production from the Great Plains plant
could represent one half of the smeller creosote market expected
in the future. Maximum cresylic acids production could supply a
large fraction of the total natural cresylic acids market in the
United States. A marketing strategy for cresylics will require
penetration of overseas mrkets. Products such as benzene and
phenol carry no such market volume risk.
70Z SIP DROP IN WOR OIL PRICES IN 1986 HAS WDE UNPROFITABLE
iHE MJFACIURE OF P3ODUM'TS COIWNTING DIRBCLY WITH PER1OLEUM.
The table below shows that benzene and the phenols have been
proportionately less affected.
(IARISON OF 1985 AN) 1987 PHOOX1IC PRICES

(Dollars per Pound)
Product 1985 1987
Gasoline 0.14 0.08

Methane 0.54 0.34
Jet Fuel, JP-4 0.12 0.07
Phenol 0.28 0.21
Cresols 0.45 0.38
Xylenols 0.48 0.50
Benzene 0.18 0.16
W1E PEIXCfIN OF REFINED FUEL PROUUXrF IS UNLIKELY 710 BE OWETI-

TIVE WITH PEIRLEUM IN MnE NEAR FUTURE. The economies of scale
which can be achieved with high-volume petroleum refining make it
extremely unlikely that a nmdest-sized by-products operation would
be effective in the fuels area. Products derived from the unique
nature of the gasification plant streams (i.e. high aromaticity,
high phenols) offer a better opportunity for profitable operation.
CCNTlTIONS OF M)ST Il)IVIDUAL CIPJI IN THE TAR OIL STREAN

ARE PROBABLY 7100 S1MA T0 BE ECNOMICALLY IMPORTANT. Coal tar oil
is an extremely diverse material, with hundreds, even thousands of
individual species identifiable. Although some of these would be
valuable as specialty chemicals, the concentrations of mst are
too low to be of economic interest. Recovery and sale of in-
dividual chemical species may provide a fringe benefit, but will
never be a major economic force.
%A'-- ..-
A LGI--,T TREN) ARY nam CWLEX FWIUXHff1OS TO PR SD#=1I'r
CONTINUES. At one time a large number of chemicals were produced
from coal tar. The number has dwindled steadily over the last 20
years due to declining coal tar production combined with increas-
ing competition from cheaper and higher purity synthetics nude
from petroleum. Within the last few years the coal tar refining
industry has greatly diminished. Rather than attempting to fight
this trend, a by-products operation should be structured to
produce what the market wants, if at all possible. Pure benzene
and phenol would be ideal products.
IMESMIXTURES MY
FR IRE A LONG TIM TO DEVELOP. With
pure compounds, marketing is a matter of meeting specifications
then offering a reliable source and price. With conplex mixtures,
specifications are unlikely to be able to shield a user from risk
in using a new product derived from coal gasification by-products.
Therefore, prolonged market development efforts may be required,
with large samples carried through the buyer's manufacturing
process, and the end-product subjected to extensive reliability
and quality tests.
33KJIWIM PU-EJNUFACTURIre NDTICES IRI)U THE TOXIC SUBSTANS ON-

TRM ACr WILL BE A MAJOR DETRE T1D ESTABLISHIIM NEW MARKETS FOR
BY-PRODUCTS. Products sold from the plant should be kept as
generic as possible in order to avoid the need for PMN's for every
possible end-product. In any case, a PMN strategy will be a major
factor in project planning.
TH OFTIEM PRODUCT SLATE WOULD BE MILITARY JET FUEL PFDUCED ROM

TH TAR OIL 1111AM PLUS BENZEN AND PHENOL PR=XZU FROM THE NAPH-
THA AND CUM PEEDOLS S'RE. The production of jet fuel would
require government support, but the benzene and phenol operations
should be profitable whether or not a jet fuel facility is imple-
mented. If no support can be obtained for jet fuel production,
the tar oil may find its best outlet as creosote.
saNaMICs WILL BRE 7ONLY DEPDNDIT ON THE TRANSFER VALUE IICH

MMT BE ASSIGI I0 HK RAN LIqUID STREAM. This will be deter-
mined by the cost of converting the gasification plant from
liquid-fired to lignite-fired operations, and by the alternative
price which could be received for the lignite fines. The table
below highlights the effect that the liquid transfer value will
have on by-product economics.
DCFROR, Percent, DCFROR, Percent,

If Liquids Value If Liquids Value
Product Is $I/MIIBTU Is $3/IMBTU
Benzene 23.3 16.9

BTX 26.8 20.0
Benzene + Phenol 35.9 22.5
Multiple Products 28.9 18.9
4
S IINAiR I N IN)ICATD
E RETURNS T ABOV
40 PERCEIT. The above rates of return are based on constant-
dollar analysis, with 100 percent equity financing. Adding infla-
tion would raise the rate of return by about the rate of inflation
(forecast to be approximately five percent). Leveraged financing
would also raise the rate of return. Borrowing 70 percent of the
capital cost at 12 percent interest would increase the DCFROR by
about another 10 percentage points.
TIUMSIRATION OD'TS ARK A &JOR DETRRENT 70 PRDING LOf-VALUE

PRIODCTS IN TE (REAT PLAINS RE31ON. There are no large markets
within economical reach of truck transport. For shipment by rail
or combined rail and barge, a generic average cost of transporta-
tion will be about $6 per barrel. The table below highlights the
effect that transportation costs can have on the rate of return
(basis is liquids transfer cost of $2 per million BTU).
EMar OF TAiNSPor 00Ms ON RATE OF RZTURN

No-Cost High-Cost
Product Transportation Transportation
Benzene 23.1 17.1

BTX 26.8 20.1
TH P ucr PRICE OUTLOOK IS GOIALLY POSITIVE. The worldwide

decline in comwdity prices which occurred over the last few years
appears to have reached bottom. Many facilities have been closed.
Further closures are unlikely to have a substantial additional ef-
fect. Increased product prices will have a very large effect on
the rate of return for a by-products operation, as illustrated
below (basis is liquids transfer cost of $2 per million BTU).
m OF PROUCT PRICES ON RATE OF RETUR

If Prices Fall If Prices Increase
Product By 25 Percent By 50 Percent
Benzene 12.6 32.1

BTX 15.5 36.2
5
7ME FOIIMIAL PiEFITABILITr OF A !Y-1EhXXrS UPGWWIN13 OPM TION
WA S S OUJCS 1D&lTICQ FMO IWLI
iONS ATICIN REMMZSS OF
11m SUWftW CM BE OBTAINW IMR DMNSf3TINI 7M PWMWICN
OF Jfr FUEL FU COAL LIWJIED. The first priority should be to
obtain inproved estimates for the costs of extracting benzene and
phenols from the naphtha and crude phenols streams.
6
V~ - v .
BASIS FM U STUD!
The basis for the market analyses carried out in this study is information on
flow rates and stream conpositions provided by ANG Coal Gasification Conpany.
There are three by-product liquid stream produced in the gasification plant
with quantities as shown in Table 1.
Current opinion is that the dusty tar stream mst continue to be recycled to
the top of the gasifiers in order to hold down the dust in the gas stream.
Therefore, only the three streams shown in Table 1 were considered as being
available for by-product production. All three of these stream are presently
being burned for fuel within the gasification plant conplex. Typical analyses
and the high and low values observed for these three streams are provided in
Tables 2 and 3.
Some of the potentially mDst valuable conponents in the phenol and tar oil
stream are the tar acids. The total tar acids available are indicated in
Table 4. Average quantities contained in the phenol and tar oil streams are
indicated in the first tow coluams of numbers. It is assumed that these tar
acids could be concentrated into a 70 percent fraction of the crude phenol
stream and a 30 percent fraction of the tar oil stream, amounting to
67.77 million pounds per year and 95.96 million pounds per year, respectively.
The last three columns then give the average, low and high values for total
quantity of each tar acid in the conbined stream.
The agreed basis for the marketing analysis was that only products for which
nanufacturing processes are comnrcially available would be considered.
TABLE 1
LIqUID BY-PRWOUl eM I GEAT PLAINS PLANT

(Based On 91 Percent Onstream Factor, Gas Production
At 137.5 Million SCFD)
Specific
Stream Gravity Flow Rate
Tar Oil 1.016 319.9 Million pounds per year

Crude Phenol 1.059 96.8 Million pounds per year
Rectisol Naphtha 0.825 59.2 Million pounds per year
7
CD
ot 0 CD
40
0 Q
V4 LaoC.
LOi "I tot04VL oua
'J1 10 V1fto0w0"vLaM c o
000
go0 0
CCDD4@e§e t0 @0 -
0n"
la4 "a1
4 w4
CD 01
0D
0a
V- t
Z, 010
WI w
>00Uu0 Z 0
-4@4~@4
0 -co
.0
-4.
Aw:x
C! I Li e O
14
aq co in LAL OW40 I
. t at 4 qa! t a
tco 0 40 0
C*0040C§
t-
C!0 t0 0!4'~@ qt0! -a .C!

CD m 4 0 .l .
w t- C.
.if Co OOO
w' Goo 4 c§c

0"40 O' O DO"4 e
O00 . "4 W
.a . C) CD . *
* 1 l* 0! -4 -94 CM
11 0P-4
0
0 A,
.4 .. V4 1, 4
u 4 0z .1c Z x4c COE-0 06u F
40q ~W# - W4
a- 0 0
1.4
to e.0- "I 00
wcor 0#4 a4 w 0 0o
a Coco- 0. 0 1 'q0 - 0 4 0 e
0!I . 11 1: C I t
V~~~ ~ ~~ CD0L
0 7 - t t c
C'"
o ~L 00' 0 4-t0#0t m 0 w~#
Q 1-~ "q0 - mIV q o a0 m 4 t.#4 b.
0 t t 14 0- 0-. c0 ot 04
0 0~
.4 4)fl0 4 0 #
a* 0 *01 a *
06.
Y
G o oooo
o co Vf to4 cp 0pa0
4)z
4lo
laMuzr -81lmr
MILITARY JKT DS
The end product of primary interest to the Department of Defense is military

jet fuel. Synthetic fuels have been widely proposed as military fuels for
national security reasons. For the most part, jet fuel has been produced from
the straight distillation of light petroleum crude oil. Because of changing
feedstock characteristics in the future, jet fuel will have to be hydrotreated
in order to meet military and commercial specifications. The continuing
depletion of petroleum reserves in the United States has led the Air Force to
evaluate the potential of alternate sources of jet fuels. They are greatly
concerned about security of energy supply.
JEr FUEL FE SAOL OIL
The United States Department of Defense (IXD) developed an aggressive progrpm

known as "Project: Rivet Shale." A research and development phase was com-
pleted including testing of over 1.3 million gallons of shale-derived fuel in
complete engine systems. The next phase, "Operation Validation," will involve
the use of shale-derived JP-4 jet fuel at Hill and Mountain Hot Air Force
Bases. The production from Union Oil Conpany's Parachute Creek Oil Shale
Project (Colorado) is planned to supply 100 percent of the jet fuel needs of
both Air Force bases. Hence, for the near term, the Air Force will have an
adequate supply of shale-derived jet fuel.
After the operational validation phase, which is expected to last two years,
the Air Force plans to be in a position to accept JP-4 that is derived from
shale oil in any percentage, at any location where the fuel is used.
The Department of Defense has played a very significant role in all the large-
scale shale oil refining studies which have been done to date. This includes
sponsoring the refining of 100,000 barrels of Paraho shale oil (Colorado) at
the Sohio refinery in Ohio, refining of 40,000 barrels of Geokinetics oil
(Utah) at the Caribou refinery in Utah, and the future refining of Union's
output at the Gary refinery in Colorado. In addition, DOD has sponsored ex-
tensive laboratory Research and Development studies on western shale oil
upgrading and refining.
In the early 1980s, the United States Air Force sponsored a nunber of bench-
scale and pilot plant studies to ascertain the suitability of western shale
oil to produce acceptable JP-4, meeting the current specifications. These ef-
forts were with Sun Oil Conpany, Ashland Petroleum, UOP Process Division and
Anoco Oil Company. Using different refining approaches it was concluded that
western shale oil produced by a number of different recovery techniques was
an excellent source of heavy naphtha and middle distillate materials. A shale-
derived fuel was produced which exceeded all the requirements of the MIL-T-
5624L specification. Because of the hydrorefining required to produce this
fuel, there is a possibility of some lubricity problems in fuel-wetted close-
tolerance components, i.e., fuel punps, controls, etc. The United States Air
11
Force routinely uses surface-active corrosion inhibiting materials to reduce
this problem and a ball-on-cylinder test is being standardized to test for the
property.
J* I= FEE TAR SAND
In 1983 the Fuels Branch (AFWAL/POSF) of the Aero Propulsion Laboratory at

Wright-Patterson Air Force Base, Ohio, awarded contracts to the Ashland
Petroleum Company and to the Applied Research and Development Division (ARD)
of Sun Oil Company to conduct the research on aviation turbine fuels produced
from Kentucky and Utah bitumens. The goal of the program was a comnercially
viable processing scheme to produce high yields of aviation turbine fuel,
grades JP-4 and JP-8, with a minimum energy efficiency of 70 percent and net
production of coke and residual materials of no more than 10 percent. The
program was completed in 1986.
The Air Force selected a pilot-scale produced Kentucky bitumen, recovered by

Westken Petroleum Company as part of the Kensyntar program. The bitumen was
recovered via an in situ wet conbustion process from the Big Clifty deposit of
Edmonson County, Kentucky. The Air Force also selected a pilot-scale produced
Utah bitumen, recovered by the Chevron/GNC joint venture. The tar sand was
mined from the Sunnyside deposit and the bitumen recovered via a solvent as-
sisted water flotation process.
Each contractor investigated a different processing scheme. Ashland's ap-

proach to processing the Kentucky bitumen was based on established conmercial
processing techniques (Commercial Chemical Desalting, ART, RCC,
Hydrotreating) used at their Cattletsburg Kentucky refinery. Sun chose to
process the bitumens using conventional refining processes for desalting,
hydrovisbreaking, hydrotreating and hydrocracking.
Each contractor investigated three product slates; a product slate which mini-
mized the cost of JP-4; a slate which mximized the production of JP-4; and a
product slate which maximized the production of JP-8.
The processes all produced high quality JP-4 fuels. The processes also
produced high quality JP-8 fuels.
At Ashland, four JP-4 (MIL-T-5624L) samples were produced by varying the

severity of the final hydrotreater. These samples provided a range of
aromtics contents from conventional to extreme, and addressed the Air Force
efforts to evaluate the effects of higher aromatics contents on jet engines.
Increased aromatics levels offer cost reductions and increased fuel energy per
unit of tank volume, but also result in decreased engine combustion liner life
and increased snvke emissions. The JP-4 samples were excellent fuels. In
contrast to conventional JP-4, this fuel showed very naphthenic character,
with a low API gravity, hydrogen content, and K factor. Thermal stability and
freeze point were excellent. Key characteristics of all the JP-4 sanples
varied linearly with aromatics content.
12
N,
A JP-8 fuel (MIL-T-83133) was also prepared. The result was an extremely
naphthenic fuel; existing hydrogen content specifications were met only by
reducing the distillation end point of the fuel by about 70 OF. Thermal
stability freeze point, and water reaction for this fuel were outstanding.
SmDke point, however, was umrginal due to the low hydrogen content. Of par-
ticular interest, this fuel confirmed the high volumetric heating value an-
ticipated for naphthenlc or so-called "high density" fuels.
JEr MM FM COAL LIiUID8
Not as much attention has been given to producing military turbine fuels from
coal liquids as to the use of shale oil and tar sands. One reason is that the
high level of aromatics in coal liquids makes them unsuitable (aromatics cause
smoke) without hydroprocessing. Even a few percent naphthalenes in jet fuel
are a problem. However, the work discussed under "Jet Fuel From Tar Sands"
makes it clear that aromatic compounds can be converted to cycloparaffinic
(naphthenic) compounds, which may be ideal Jet fuels. In this case, coal-tar-
oil-derived liquids might even be a preferred feedstock.
FUE SPECI FIC&TIONS
Aviation gas turbine fuel represents only about 7 percent of the refinery
production in the United States. The 1984 domestic production of naphtha-
type military jet fuel (JP-4) was 221,000 barrels per day while the production
of kerosene-type fuels (Jet A, JP-5, Jet A-i) was reported to be 922,000 bar-
rels per day. The growing scarcity of light sweet petroleum could inpact the
future availability of a minor product such as Jet fuels. For this reason, DOD
has studied alternative refinery feedstocks for the future and evaluated their
potential for producing aviation gas turbine fuels. The potential for mdify-
ing fuel specifications, making them more identifiable with the basic
feedstocks, and improving aircraft performance has also been considered.
Jet fuels are the mDst critical product, in terms of quality, produced by the
refining industry. The naturally occurring hydrocarbons in petroleum can be
essentially fractionated with little additional processing to produce highly
stable jet fuels. There are instances where fuels are mildly hydrotreated to
renvve sulfur but for the mst part, chemical conversion processes have not
been necessary.
In 1951 the United States Air Force standardized aircraft operations on JP-4
Jet fuel in order to optimize performance and availability. Some of the mst
important chemical and physical properties of JP-4 are depicted in Table 5.
This fuel can be characterized as a "wide-cut" fuel with a boiling range be-
tween 150 degrees F and 480 degrees F. This range optimizes the requirement
for a vapor pressure between 2 and 3 psi at 100 degrees F and a maximum
freeze point of -72 degrees F. These properties are required for high alt-
13
TABLE 5
PUMRWY MILITARY AVIATION FM SPCIFIC&TION REWIJI3Ir
JP-4 (Jet B) JP-8 (Jet A-i)

MIL-T-5624L MIL-T-83133A
Specific Gravity, 60 F 0.751-0.802 0.775-0.840

Gravity, API 060 F 45-57 37-51
Distillation, Max Degrees F
10% Recovere d . . .. ..
20% Recovered 293 401
50% Recovered 374 ---
90% Recovered 473 ---
Final Boiling Point 518 572
Freezing Point, Degrees F -72(-58) -54
Viscosity, Centistokes @ -40 F, Max --- 8.0
Aromatics, Vol% Max 25.0 (20.0) 25.0 (20.0)
Olefins, Vol% Max 5.0 5.0
Sulfur, Wt%, Max 0.40 (0.30) 0.40 (0.30)
Net Heat Of Caibustion, BTU/lb, Min 18,400 18,400
Hydrogen Content, Wt%, Min 13.6 13.6
Thermal Stability, JFrOT
Pressure Drop, nm Hg, Max 25 25
Heater Deposit, Max 3 3
Flash Point, Degrees F, Min --- 100
Vapor Pressure, Reid, psi, 2-3 ---
14
itude operations and low ambient temperature conditions at certain air
bases. Chemical composition is controlled by limits on aromatic, olefin, sul-
fur and mercaptan sulfur contents. For operations where JP-4 may not be
available, a kerosene-type fuel, specification JP-8, nay be used. The Navy
fuel specification is JP-5, which is identical to JP-8, except for a minimum
flash point of 140 *F, for safety on aircraft carriers.
Although made up of many different hydrocarbons, jet fuels consist essentially

of four general types of compounds: paraffins (including isoparaffins);
cycloparaffins or naphthenes; aromatics; and olefins. The proportion of each
compound type varies with the different types of crude oils and the process-
ing needed to produce the fuel. A typical fuel will contain hundreds of dif-
ferent conpounds. Except for restrictions on aromatics and olefins, the
proportions of hydrocarbon types are not directly controlled. The require-
ments on density, energy content, freeze point and hydrogen content favor
paraffinic or Isoparaffinic based fuel. The hydrocarbon constituents of most
JP-4 fuels are primarily normal or isoparaffins (50-70 volume percent) and
cycloparaffins (20-40 volume percent).
The hydrocarbon-type composition of a typical petroleum JP-4 is depicted in

Table 6. The refiner produces JP-4 from a blend of 60-70 percent light naph-
tha and 30-40 percent kerosene in terms of crude oil fractions. The comner-
cial specification for the wide-cut fuel is designated as Jet B and is
similar, with a few minor exceptions, to JP-4, as indicated in Table 5.
The basic differences between Jet A-1 and JP-8 are the additives required for
military operations. The "average" boiling range for Jet A-1 is between 350
degrees F and 520 degrees F where volatility is limited by a flashpoint mini-
mum of 100 degrees F. This fuel has a freeze point maximum limit of -54 de-
grees F. The hydrocarbon constituents in this fuel are, in general, denser
than JP-4.
HIGH DEISITY JET FUEL
The possibility of mdifying jet fuel specifications has been the objective of
various research efforts over the years. Most of these efforts were aimed at
reducing fuel costs while minimizing any effect on aircraft performance. One
way of doing this is reducing the hydrogen requirements and increasing
aronmtic content thus eliminating the expense of hydrotreating. There is
also a benefit to be gained by allowing the jet fuels to become more cyclic,
thereby increasing fuel density and allowing additional range. A major in-
crease in density is experienced for each ring formed from a given carbon-
numer species. Methyl or ethyl substitution generally does not lower the
density and major improvements in freezing point depression can be realized.
Volumetric heating values increase with density. Therefore large concentra-
tions of cyclic paraffins (naphthenes) could increase the range of aircraft
and maintain good low temperature capability. Table 7 shows the effect of
cyclization in terms of two representative C-10 hydrocarbons, n-decane and
decalin (decahydronaphthalene). These compounds are components in varying
amunts in jet fuels. The decalin, which is a bi-cyclic saturated hydrocar-
15
TABLE 6
TYPICAL HDaW CKFWOSITION OF JET FUILS

(Modified ASTM Method D-2789)
Hydrocarbon JP-4 Conposition Jet A-1 Coffposition

Type Weight % Weight %
Paraffins 66.0 44.5
Cycloparaffins 16.7 36.2
Alkylbenzenes 15.8 16.5
Indans and Tetralins 1.0 2.3
Naphthalenes 0.5 0.5
Total 100.0 100.0
TABLE 7
PWRTPrIES OF C-10 HYIOCARBDONS
n-Decane Decalin
Formula C1_H_
Molecular Weight 142.29 138.25

Specific Gravity 0.7342 0.9011
Boiling Point, Degrees F 345.5 384.5
Freeze Point, Degrees F -21.4 -45.6
Characterization Factor, K 12.61 10.50
Hydrogen, Wt% 15.59 13.12
Net Heat Of Combustion
BTU/Pound 19,017 18,324
BTU/Gallon 116,410 137,660
16
bon, produces a denser material, with a lower freeze point and an 18 percent
increase in volumetric heating value. However, the drawback to using cyclic
hydrocarbons is the reduction in hydrogen content, which affects the comtbus-
tibility in terms of increased flame radiation and soot formation.
Worldwide there is a variety of crudes that might be characterized by the

dominant hydrocarbon types present in the mixtures. For exanple, paraffinic
crudes are generally waxy, hence, they are selected for the manufacture of
paraffinic lubricants. Naphthenic-type crudes are relatively low in paraffin
content. They are selected for special applications or for the manufacture
of aromatics and gasoline. Aromatic rich crudes may be high In asphaltene
content. They are used for manufacturing gasoline and asphalt. In the United
States, the production of low sulfur naphthenic crude is declining rapidly and
it will soon become a scarce comnwdity. Hence, this raw material cannot be
considered to be a major supply source for manufacturing higher density jet
fuels.
However, modern petroleum refiners utilize a variety of thermal, catalytic and

physical separation processes to manufacture high quality fuels from crude
oils. In the refinery conversion processes where large molecules are cracked
either thermally or catalytically, fractions rich in aromatic rings are
produced. These fractions can be converted to naphthenes by various addi-
tional hydrogenation processes.
The Air Force recently reviewed the availability of hydrorefining and

hydrotreating facilities in domestic refineries. From these specific
refineries and their hydroprocessing capacities, the potential volume of
naphthenic fuel that could be blended from these plants was estimated to be
1,772,700 barrels per day. This potential volume is far in excess of any
demand. However, the resulting quality and characteristics of these fuel com-
ponents would have to be determined.
Tar oil frbm coal gasification contains high levels of both naphthenes and
aromatics. By selectively hydrogenating the aromatics, but not cracking the
naphthenic rings, a high density fuel could be produced.
In order to evaluate a naphthenic fuel produced from an aromatic stream an

effort was carried out under a research contract with Stone & Webster En-
gineering Corporation. A light pyrolysis fuel oil produced from the steam
cracking of a petroleum gas oil fraction was converted into jet fuel. The raw
aromatic fuel oil was converted into naphthenic products by hydrogenation in a
pilot plant operation carried out at the Institute Francais de Petrole at
Rueil-Malmaison, France. Table 8 presents the critical fuel properties and
conpares them with the starting feedstock.
This sample of naphthenic fuel (JP-8X) has the same volatility as JP-8 or Jet
A-i, essentially a kerosene. The high density JP-8X fuel is 93.6 percent
naphthenic (cycloparaffins). When comparing the volumetric heating value of
this JP-8X to conventional JP-4, the energy increase is nearly 12.5 percent.
For certain volume limited aircraft this energy increase could result in
17
TABLE 8
KIM nuesir Jr FUEL aMMIxGicn
Feedstock High Density

Pyrolysis Fuel Oil JP-8 (JP-8X)
Specific Gravity, 60 F 0.978 0.872

Net Heat of Combustion
BTU/Pound 16,881 18,271
BTU/Gallon 138,060 135,425
Hydrogen, Wt% 8.33 13.25
Flash Point, Degrees F 175 134
Boiling Range, Degrees F 392-562 369-500
Freeze Point, Degrees F --- 100
Thermal Stability --- Excellent
Viscosity, Centistoke 0 -40 F 74 21
Chemical Analyses
Naphthenes, Wt% --- 93.6
Aromatics, Wt% 100 6.4
Paraffins, Wt% --- 0
•p
18
comparable range increases. The JP-8X prototype fuel therefore offers
promise as a range extender for aircraft.
Such a fuel could be derived from many of the refinery feedstocks of the fu-
ture, such as heavy oils, tar, bitunen, and coal liquids. The cyclic nature of
all these materials makes then ideal for a higher density product if the
hydrocarbon ring is preserved in the processing.
MILITARY JET FUEL DDWI
Total Jet fuel consumption by the United States Air Force is approximately
220,000 to 240,000 barrels per day. However, within a reasonable marketing
radius of the Great Plains plant, there are only four air bases (Figure 1).
Current demand at these bases is given below:
Grand Forks 2,348 barrels per day

Minot 2,479 barrels per day
Malmstrom 130 barrels per day
Ellsworth 1,300 (currently shut down)
PRICES AND MNRKET STRUCTURE
Military fuel purchases are obtained by open bid, and therefore follow the
prices for commircial jet fuel very closely. Supplies for the air bases are
generally let for bid once a year. The mst recent contracts have a bid date
of October 1986. The results are listed in Table 9. These contracts are for
one year, and have escalation and de-escalation clauses depending on posted
prices for petroleum products in the region.
Because of the escalation clauses, the base prices in the contracts, as listed
in Table 9, actually have very little neaning. The contract is typically
written with two complex escalation clauses. The first clause determines the
delivered price. It is based on published prices in Platts and another publi-
cation. Two weighting factors are used. Seventy percent weight is assigned
to a fornula involving the price of unleaded gasoline and thirty percent
weight is assigned to a formula involving reported contract prices for commer-
cial jet fuel (Jet A).
The second escalation clause is applied after delivery, when the Department of
Energy Petroleum Marketing Monthly is issued. Depending on the factor, the
jet fuel supplier may then either receive an additional payrmnt, or may have
to make a refund to the government.
To illustrate the importance of the escalation clauses, Table 9 indicates that

current (April 1987) delivered prices to Grand Forks, North Dakota should be
in the range of $0.80 per gallon. Actual deliveries in April, 1987 are run-
ning about $0.45 per gallon.
19
-5.
EISM
MT F Ia
NDPNT P
'W • ELSWORTH
FIGURE I
AIR FORCE BASES IN GREAT PLAINS REGION
TABLE 9
OCTOBER 1986 BID PRICES FOR JP-4 DELIVERED 70 AIR FOWZ BASES
(For Deliveries Through September, 1987)
Quantity Price
Base Gallons Vendor $/gallon
Grand Forks, ND 15,000,000 Ashland 0.775

15,000,000 Ashland 0.79
3,320,000 Ashland 0.81
2,680,000 Western Petroleum 0.8335
36,000,000
Minot, ND 12,600,000 Western Petroleum 0.8510

12,600,000 Western Petroleum 0.8510
12,600,000 Farmers 0.8631
200,000 Ashland 0.8635
38,000, 000
Malnmtrorn, MT 2,000,000 Black Eagle 0.91406
Ellsworth, SD Not Operating

20
1
-
Interrediate resellers, such as those listed in Table 9, are able to function
in this market only by having agreements with refineries which exactly paral-
lel the government escalation clauses. They are able to work in this way with
some refineries sinply because the refineries do not want the problems of
dealing directly with the government themselves. The internmediate sellers
deliver the product directly form the refinery to the airbase, to avoid any
chance of the additives or other specification properties deteriorating during
storage.
Because of the limited military jet derrand in the imnediate region, and the
once-a-year bid process, it would be inpractical to build a free-standing
facility targeted to this market. An advance purchase contract such as that
which was provided by the Defense Department for the shale oil from Unocal 's
Parachute, Colorado shale oil plant, would be necessary.
Dm9/ OULOOK
Although the Department of Defense makes five-year projections of military

fuel requirements for internal use, these projections are classified and are
not available for public release. With no rajor change in aircraft engines
planned in the near future, and a stable world political situation, it is as-
suned that military jet fuel denmnd will renin approxirmtely constant.
21
# ~ ~ - - ~ . . .~- - - - -
PErEoL5KU PRONXIYS
The liquid by-product streams from the Great Plains plant could be processed
by petroleum refining techniques to rmake a variety of products conpeting with
conventional refined pet roleum products.
The objective of the next section is to describe the overall petroleum refin-
ing network in which North Dakota and the Great Plains plant operate. This
entails examining the production, consunption and mvements of crude oil and
refined products in the Great Plains region. For purposes of this analysis,
the Great Plains imrketing region is defined as the states of North Dakota,
Minnesota, South Dakota, Wyoming and Montana, and the provinces of Manitoba
and Saskatchewan.
(i OIL
Pipeline Network
The network of nmJor crude oil pipelines in the Great Plains region is shown
in Figure 2. As seen in the figure, there are five pipelines which inport
crude oil into the region. Crude oil from Alberta, Canada enters Montana near
Cut Bank. A pipeline also runs from Alberta into Saskatchewan. This line en-
ters Saskatchewan near Macklin, then runs east through Manitoba and Minnesota
to points east. Two other pipelines that inport crude into the region run
north from Iowa to Minneapolis, Minnesota. The fifth incoming line is the
Frontier line, operated by Amco. It is a new 16-inch line which leads from
the Anschutz Ranch oil field, just inside Utah at the southwest corner of
Wyoming, to Casper. At Casper it connects to the Anoco refinery and to the
Platte and Sinclair pipelines. Ultinate capacity of the line is expected to
be 120,000 barrels per day.
Wyoming is a najor exporting state with five pipelines carrying crude oil out
of the state. Two rmjor pipelines export crude oil from Wyoming to the east.
The 20-Inch Platte (Marathon) pipeline connects Casper, Wyoming to St. Louis,
Missouri. The 20-inch Armco pipeline connects a large crude oil gathering
system in Wyoming to Sterling, Colorado then to Freenmn, Missouri and goes on
to Whiting, Indiana. A Conoco crude oil line originating in Wyoming runs
south to Cheyenne and Denver. It interchanges with the Platte line at
Guernsey station. A four-inch pipe connects from Guernsey station to the
AnDco system at Ft. Laramie station. The Conoco line carries both crude and
asphalt to Denver, with a capacity of 56,000 barrels per day. To the west, an
eight-inch Armco crude oil pipeline runs to Salt Lake City from southern Wyom-
ing. Amco also has two parallel four-inch lines that run from southern Wyom-
ing to Rangely, Colorado.
Production
The 1986 crude oil production for the Great Plains region states and provinces
is presented in Table 10. Wyoming is by far the largest producer of crude oil
in the region. Historical production figures are given in Table 11.
22
me oa itl mr
3 ~~Chec 791t 1 e aCnumr 1Monaer
FiGURSD
GREA PLREINRUD PPLIE
ManiStb 14 20 1.8----
Minnsot 0 hWoigRt 0
2ut ~ ~ oookMontana Rot0.8
n p orth
N Dkorta 12500 eroCnd 16.5amfExn
SakatcFIeURn 2902.
TotalABL 7580010.
Su:State oic andoGas ctmions Prcn

Manitoa 14900 1.
Minnesta 0
Montan 23 74,20*
TABLE 11
GI&? LAINS RGION = PRODUCTION

(Units: Million Barrels Per Year)
1979 1981 1983 1985 1986
Manitoba 3.67 3.42 4.65 5.16 5.19

Minnesota 0.00 0.00 0.00 0.00 0.00
Montana 29.96 30.81 29.67 29.85 27.08
North Dakota 30.91 45.71 47.55 52.63 45.62
Saskatchewan 58.31 46.00 59.38 72.86 73.32
South Dakota 0.85 0.63 1.17 1.58 1.59
Wyoning 129.61 122.21 120.98 130.98 124.00"
Total 253.31 248.78 263.40 293.06 276.80
Estimt imon
Source: State Oil and Gas Cnmnissions
24
CDE OIL B&LANCE
The locations of refineries in the region are also shown in Figure 2. Nominal
crude oil distillation capacities of these refineries are listed in Table 12.
Several small refineries in the region, which have been shut down, are not
listed. Using an average of 85 percent operating capacity for the refineries
in Table 11, a crude oil balance for the region is constructed in Table 13.
Although a reasonable closure accuracy is indicated, it must be emphasized
that no data were available for several of the pipelines. The near closure of
the balance may be only fortuitous.
It is clear from the table, however, that refinery consumption is much less
than crude oil production in the region, and that the region is a net exporter
of crude oil. This is important because it suggests that any new source of
liquids within the region will have to find a market by displacing other liq-
uids to markets outside the region.
In other words, if the Great Plains plant should start using coal for plant
fuel and convert its liquid by-products to fuels for sale within the region,
the effect would probably be to reduce crude runs at existing refineries in
the region. If these refineries are operating at less than full capacity
while crude oil is being shipped out of the region, it indicates that they are
unable to sell more product than they are presently producing. New production
from Great Plains, therefore, would cause them to reduce crude runs. This, in
turn, would push more crude oil into markets outside the region. Since the
region is already a large net exporter, the infrastructure for such a market
shift is in place, and the effect on crude oil prices would not be noticeable.
The end result would rmst likely be a reduction in imports, either from Canada
or from overseas sources shipping to the Gulf Coast. With an apparent net of
exports over inports of 262,000 barrels per day for the region, an additional
5,000 barrels per day would have minimal impact on prices.
REF 1NE PROIUCTS
Pipeline Network
Refined products are produced at refineries in the Great Plains region and
also inported into and exported from the region via products pipelines. A map
of major refined products pipelines is shown in Figure 3. Refined products
enter the region via the Dome, Interprovincial and Petroleum Transmission
pipelines all from Alberta, Canada. The Williams, MAPOXO, Dcme, Amoco, Wyco,
Medicine Bow, and Pioneer pipelines export refined products to many locations
outside the Great Plains region.
Refinery crude capacities in the Great Plains region were listed in

Table 11. Refineries located in Casper, Sinclair, and Cheyenne, Wyoming and in
Billings, Montana serve Montana, Wyoming, Colorado and North Dakota. From
Sinclair and Casper, Wyoming, products can move to southwestern Wyoming and
Salt Lake City, Utah. The refineries located in Regina and Moose Jaw, Saskat-
chewan serve Saskatchewan and Manitoba and send some products to Minnesota and
points south. Minnesota itself has an extensive products pipeline network
connected with North Dakota, South Dakota and regions to the south of Min-
nesota.
25
% N %
Lo. cl- 0 1 ZICZJ* 7;? 5-
TABLE 12
19"6 (REAT MAINS IC REFINRY CAPACITIES
Crude Distillation
Refinery Capacity, BPCD Location
Manitoba
No Refineries
Minnesota
Ashland 67,143 St. Paul Park
Koch 137,000 Rosemount
Subtotal 204,143
Montana
Cenex 40,400 Laurel
Conoco 48,500 Billings
Exxon 42,000 Billings
Kenco 4,700 Wolf Point
Montana 6,300 Great Falls
Subtotal 141,900
North Dakota
Amco 58,000 Mandan
Subtotal 58,000
Saskatchewan
Consuner's 40,000 Regina
Petro-Canada 13j300 Moose Jaw
Subtotal 53,300
South Dakota
No Refineries
Wyoming
Amco 48,000 Casper
Frontier 28,800 Cheyenne
Little Armrica 20,000 Casper
Sinclair 53,000 Sinclair
Wyoming 12,500 Newcastle
Subtotal 161,500
Total 618,843 Barrels Per Calendar Day
Source: Oil and Gas Journal
26
TABLE 13
(R PLAINS REGIGN CRIZE OIL BALANCE

(Units: 1,000 Barrels Per Year)
SUPPLY & IMPORTS
Production: Crude Oil

Manitoba 5,194
Minnesota 0
Montana 27,080#
North Dakota 45,624
Saskatchewan 73,321
South Dakota 1,586
Wyoming 124,000#
Total 276,805
Pipeline Inports Into:
S ask at chevan
Interprovincial Pipeline 307,634
Wyoming
Frontier Pipeline 18,250#
Montana
Conoco Pipeline 17,338
Minnesota
Woodriver Pipeline 34,218"
Minnesota
Williams Pipeline .4,563
Total Imports 382,003

Total Supply and Imports 658,808
OONSLMPTION AND EXPORTS
Refinery Consumption 191,995

Exports From:
Wyning
Conoco Pipeline, South 20,440
Platte Pipeline, East 36,594
Aroco Pipeline, Southwest 3,650
Amoco Pipeline, East 39,694*
Amoco Pipeline, West 14,600
Minnesota
Lakehead Pipeline, East 362,733
Total Exports 477,711

Total Consumptlin & Exports 669,706
Closure Deficit = 669,706 - 658,808

= 10,898 thousand barrels
Estimated capacity, not actual value
# Estimation
27
TABLE 14
1936 3D INN) PIWK PIPSEINK INNUUM/RT

DALAM PM GO= PLAINS INOIQE
Total Products
Net Inports Net Exports
Manitoba 23,649
Minnesota 12,305
Montana 10,950
North Dakota 5,220
South Dakota 1,898
Saskatchewan 1,092
Wyoming __ _38, 124
legimil Totals 42,266 50,972
Source: Pipeline Conpanies

P5
5-
101
NOE 12 MN
4~~s
SMPO1 l~mm aamr onane 12ao
5~ COtnl f
M2i~eSw1 YO tExn mPtoCnd
5 A FIGUREFIERE
GRNEAT PLIN REINPOUC1IEIE
*S~* V5
'~:~---IS*~ ~ & -.- 5- A 0C
---- h K1~nco Sinclair ~ -
A pipeline inport/export balance (Table 14) constructed from the inforrmtion
available shows that the region is a slight net exporter of refined products.
These figures must be viewed with much caution because some pipelines are
missing from the accounting and because both refined products and natural gas
liquids pipelines are included. Nevertheless, it indicates in a general way
that refined products production in the region is nearly in balance with
demand.
Refined Prodnct Demand
Data on refined product demand (consunption) is available from the State

Energy Data Report from the EIA. The most recent data available are for 1984.
Attenpts were rmde to update the EIA data by contacting individual state
agencies, but Information is scarce and some states keep no record of the type
of inforrmtion needed. Demand based on 1984 figures (EIA 1986 report) is sum-
marized in Table 15.
Refined Prodacts Product ion
There is no good source of public infornation on actual production by in-

dividual refineries. The nominal capacities listed in Table 12 are a good
guide because if a refinery falls below about 70 percent capacity it usually
becomes uneconomic and is likely to be shut down. However, actual production
nay be either below or above the design capacities listed in the Oil and Gas
Journal. Numbers listed in Table 12 are barrels per calendar day, or average
daily rates after accounting for all scheduled maintenance shutdowns. On the
other hand, direct contacts with refineries were couched in terms of barrels
per stream day.
North Dakota
The only refinery operating in North Dakota is the Amoco refinery in Mandan.
Throughput is about 50,000 barrels per day. They run only 42 gravity light
crude from North Dakota. Product mix is about 60 percent gasoline, 40 percent
distillate.
Minnesota
Minnesota has the largest refining capacity of any of the states in the region
although there are only two refineries, in the Minneapolis area. Koch Refin-
ing runs about 150,000 barrels per day of Canadian sour crudes. Output is
mostly gasoline. The Ashlane refinery runs 60,000 barrels per day of mostly
North Dakota crude, with some Texas and some Canadian sour. Output is heavily
gasoline, with some asphalt.
Saskatchewan
Saskatchewan has two refineries. However, the Petro-Canada refinery in Moose

Jaw is only a topping operation to produce asphalt from heavy crude. They
operate from April to October at 8,000-9,000 barrels per day. They had con-
sidered Great Plains tar as a blending stock for asphalt, but backed away be-
cause of environmental concerns. Consumers' Cooperative refinery in Regina
runs about 45,000 barrels per day of light Alberta crude. Production is about
50/50 gasoline and diesel fuel.
29
. r r f.-
W. I. .- . r r, r r - 'r.
m Co
00 C) t~o I C11 00
40 Ct tn t- oa mo to LM v o
wo -4 V-4 0 '.4I "

-4
m~ t- m' eq .4 q wo t- U
tt~ eq
a, 0 o o t'- Co Co
00 Cl Mt
g oo co 00 Co Co 4 1-4 Lt-
ca-4
t- 4 vLfl co o 3 m m
oo0 n co co t- cq co
CO) Oo tf- 04 Co
z -4 cq
UCo -4 co Co .4 t- t- m
I:
Co m~ oo U13 cq eq co 00
C ' -t- co m m
0~~. -4l
e eq -I C f cq~C
*A4IaLa Co -4 a~ 0 - ,
o c 00
t- t- v ' 4 Co
t- C
a, Co C -4 Co 0 vY-4 C
'-4
.-m' M 00 q4 0C4 c w 0
Coo
4Q)
oo 9.
0 0
at : a, )
9 ) 0 * - 41 0
30
A.' ~ ~
.N.
Montana
Montana has one simll and three large operating refineries. Montana Refining
runs 7,000 barrels per day of one-half Canadian light and one-half Montana
heavy sour crudes. Output is about 50 percent gasoline, 33 percent diesel,
som #5 fuel oil.
The Conoco refinery swings between 33,000 barrels per day in the winter and
54,000 barrels per day in the sunrmr depending on the rmrket for asphalt.
Feedstock is a mix of Canadian sweet and Wyoming sour crudes. The distillate
product slate is mostly gasoline, with some diesel.
A similar type of operation is carried out at the Exxon refinery. It runs

close to capacity of 50,000 barrels per day in sumner and drops back to
35,000 barrels per day in winter. Input is a mix of sweet and sour crudes,
while distillate output is 50 percent gasoline, 50 percent diesel and jet
fuel.
The Cenex refinery has a listed capacity of 40,400 barrels per day.
Wyoming
Wyoming has the second largest refining capacity after Minnesota, and has the
largest number of operating refineries at five.
The Little America refinery runs 24,000 barrels per day, using both sweet and
sour crudes. They had at one point planned to distill som of the Great
Plains tar oil, shipping som of the product back to Great Plains and using
some of it. They eventually decided against it because of environmental con-
cerns.
Frontier Refining (fornerly RMT Properties) is running about 30,000 barrels

per day of 60/40 sour/sweet Rocky Mountain crudes. They produce about
65 percent gasoline and 35 percent diesel. The Sinclair refinery runs
35,000 barrels per day, all 40 degree Wyoming and Anschutz sweet crudes.
The Mountaineer refinery in western Wyoming is basically a reclaiming opera-

tion, running 500 barrels per day of slop crudes. They usually pay less than
$2.50 per barrel for feedstock.
Wyoming Refining runs 15,000 barrels per day of Wyoming sweet, rmking rnostly
gasoline which is nmrketed In South Dakota.
Natural Gas Liquids Production
The Great Plains region, Wyoming in particular, produces appreciable quan-

tities of natural gas liquids. A breakdown by state and province is given in
Table 16. The total for the region is 58,000 barrels per day or 21 million
barrels per year. Natural gas liquids produced in the region are partially
consuned in the region and partially exported. Natural gasoline is added to
refinery streams.
31
TABLE 16
1985 G T LAIS NtURAL GAS PLANT OMUTIOW
(Units: 1,000 Gallons/Day)
No. Gas Raw NGL,

of Throughput Ethane, Natural
Plants MWcfd Other LPG Gasoline Total
Manitoba 0 0.0 0.0 0.0 0.0 0.0

Minnesota 0 0.0 0.0 0.0 0.0 0.0
Montana 9 38.0 32.6 78.8 8.6 120.0
North Dakota 10 197.8 210.1 244.4 83.1 537.6
Saskatchewan 13 91.4 149.9 - - 149.9
South Dakota 0 0.0 0.0 0.0 0.0 0.0
Wyoming 41 1,188.9 623.3 727.2 277.2 1,627.7
Total 73 1,516.1 1,015.9 1,050.4 368.9 2,435.2
MWcfd = Million standard cubic feet per day
Source: Oil and Gas Journal
32
From Table 15 it is seen that total LPG consunption Is 14.5 million barrels
per year, which is of the same order of magnitude as the production indicated
in Table 16.
Refined Producta Balance
On the basis of the above refinery survey, current refinery throughput in the
Great Plains region is approximately 601,000 barrels per stream day
(Table 17). Multiplying this by an average 94 percent onstream factor gives a
calendar day average of 565,000 barrels per day. Although this is about seven
percent above the value which was calculated earlier by assuming an 85 percent
capacity factor, some of the refineries in the survey reported only capacity
rather than actual rates, so the true value is probably 550,000 barrels per
day (200 million barrels per year) or less.
Without having detailed information on the configuration of each refinery, it

is impossible to calculate product yields on the basis of crude oil input.
However, on a national average, the volume of products produced is ap-
proximately equal to the volume of crude oil processed. This is because most
of the products have a lower density than the crude oil from which they are
made. This gain in volume approximately offsets the amount of material con-
sumed during processing. This is illustrated in Table 18 where the volume
gain is approximately equal to the amount of refinery gas produced. We may
therefore conclude that refinery output of refined products is 200 million
barrels per year.
Table 15 shows total product demand of 221 million barrels per year, which may
be compared to the sum of 200 million barrels per year from refinery produc-
tion plus 21 million barrels per year of natural gas liquids production (221
million barrels per year total). The exact match of these two numbers Is
simply coincidental, because not all of the data are equivalent, and no con-
sideration was given for movement of products into and out of the region.
Nevertheless, it is clear that, on the whole, the Great Plains region is ap-
proximately in balance between production and consumption of refined products.
This reinforces the earlier conclusion that any increased production in the
region would either force a cutback at existing refineries or else force them
to export outside the region. The latter will be difficult to achieve.
Comparison of Tables 15 and 18 shows that the relative demand for different
products in the Great Plains region corresponds fairly well with the national
average, with the exception of appreciably lower demand for residual fuel oil
and jet fuel.
An overall refined products balance for the Great Plains region is given in
Table 19. The indicated closure accuracy of 3.0 percent is much better than
appears warranted by the data and the assumptions used, and therefore should
not be relied upon excessively. The missing quantity is most likely a part of
the pipeline data. Also, mDvements by truck, rail and barge are missing from
the table.
An obvious conclusion to be drawn from the refined products balance is that It

should be easier to market new production from Great Plains to the east rather
than to the west. Although the region as a whole is approximately in balance,
there is little direct interchange between the product-surplus area in Wyoming
33
TABLE 17
REFINERY OI'ATING RAES IN GREAT PLAINS RoIN
Crude Distillation
Refinery Rate, BPSD
Minnesota
Ashland 60,000
Koch 150,000
Subtotal 210,000
Montana
Cenex 40,400
Conoco 44,000
Exxon 42,500
Montana 7,000
Subtotal 133,900
North Dakota
Armco 50,000
Subtotal 50,000
Saskatchewan
Consumers 45,000
Petro-Canada 4,250
Subtotal 49,250
Wyoming
Armco 35,000
Frontier 30,000
Little America 24,000
Sinclair 54,000
Wyoming 15,000
Subtotal 158,000
Total 601,150 Barrels Per

Stream Day
Source: Contacts with refineries
34
34 f
TABLE 18
AiWAL YTIUI MM A D*I3E OF (M= OIL
Gal Ions
Product Per Barrel % Yield
Leaded Gasoline 6.9 16.4

Unleaded Gasoline 12.3 29.4
Distillate Fuel Oil 9.1 21.6
Residual Fuel Oil 3.0 7.1
Jet Fuel 4.0 9.5
Petrochemical Feedstocks 1.2 2.8
Asphalt and Road Oil 1.3 3.2
Still Gas (Refinery Gas) 2.0 4.7
Coke 1.5 3.7
Liquefied Gases 1.3 3.1
Lubricants 0.5 1.2
Kerosene 0.3 0.8
Miscellaneous 0.2 0.4
Special Naphthas 0.2 0.4
Wax 0.1 0.1
Processing Gain -1.9 -4.5
Total 42.0 100.0
Source: American Petroleum Institute
35
TABLE 19
OVmALL mRVN P M uAN&C FOR

TR PLAIWARION
Total
Products
Refinery Production 200,750
Natural Gas Liquids

Production 21,000
Pipeline Imports
From Other States
Or Provinces 2,266
Total Supply & Imports 264,016
Consurption 220,980
Pipeline Exports 50,972
Total Consumption & Exports 271,952
Difference 7,936
Balance Closure, Percent 3.0
Source: DOE-EIA, state agencies, com-

pany contacts
36
.,~ -.- *. .'

and the product-short area in Minnesota. North Dakota, like the region as a
whole, appears to inport only a relatively small fraction of its needs. Thus,
if it were decided to market conventional fuel products from the Great Plains
plant, it would be necessary to displace local production from conventional
refineries.
Individual Products Demand
The 10-year demand history for individual petroleum products in the five-state
region is shown in Figures 4 through 13. Overall petroleum demand (Figure 13)
fell sharply from its peak value in 1978 until 1981, and has shown only small
gains since. The pattern has been basically the same in all states. In-
dividual products, however, may show quite different demand histories.
Conrercial jet fuel, or Jet A, has registered long-term strong growth in

demand in Minnesota (Figure 4). This is of course due to the presence of a
regional international airport at Minneapolis-St. Paul, and no conparable
facilities in the other states. Aviation gasoline, used in non-connercial
aircraft, has shown declining sales in all states (Figure 5).
Motor gasoline is the major product of almDst all refineries in the region,
and its demand pattern (Figure 6) follows closely that for total petroleum
products in all cases. Distillate fuel demand, on the other hand, shows sae
variation in pattern from state to state, and is generally subject to much
wider swings in demand than gasoline.
Kerosene was once an inportant product in the region, but demand has almst
disappeared over the last ten years (Figure 8). Liquefied petroleum gas
(Figure 9) demand has declined from peaks in the 1970's, but in nmst cases the
decline has not been great. The demand for residual fuel oil has declined al-
mDst as drastically as for kerosene (Figure 11). The change in markets for
residual fuel in particular has had a large irpact on refinery operations in
the region. Those refineries unable to convert residual fuel to other
products have suffered from the inability to match their product slate to
market requirements. In some cases, overall production has been constrained
because of an inability to sell the residual oil fraction.
Regional demand for asphalt and road oil has declined somewhat since peaking
in the early 1970's, but has been reasonably stable (Figure 10). rdiscel-
laneous products (Figure 12) have been somewhat less stable.
It is clear from the product histories that continuation of these trends into
the future would make it unattractive to produce any product conpeting with
residual fuel oil. Only gasoline, jet fuel and asphalt appear to be on stable
or rising demand trends.
Speciflcat ios
Definitions
The terminology used for petroleum products is not always consistent and is
sometimes confusingr. Most major distinctions are made on the basis of boiling
point. As generally used, gasoline is the lowest boiling fraction; it will be
10 percent distilled at 122-158 IF and 90 percent distilled at 365-374 °F.
37 _
~.
J,
t-- - ......
• ,..... ....... ]
7
'I 7,4
-6-
SOU&IMOMTA,
-ell
MOWMIA
I I II I I I | I I I I | I J l I - ' , - "
1964 1 ?E8 1972 1976 19E0 1 984-

YEAR
FIGURE 4
JET FUEL DEMAND
800
700 -
1 600
500
400 -
300--\
\-
200 3--4.
ico - \ :--h-- . + i--
0* I I I I I I I I I I I I I ,
1964 ?81972 1 S--73 19E~()

YEAR
FIGURE 5
AVIATION GASOLINE DEMAND
38
v,' ' . .' . ,* *I' .. ~

*, A , "-,, , ",,'.°,, v'
_. ,'' . . .- ,,-
%,, - - -,V,-,-., - " . . - ,, - . - , . ,,. -
60
-
311
20-
~ 0
1 9 6 . 1 !) € I1 9 '7 2 . 1,3 19 E;:]( 84
~FIGURE 6
0 -I I I I I I I
MOTOR GASOLINE DEMAND
30-
28 - /
26 - /
24
22 -
14. '
12 , ",
10 -
S I l I i I I I I I , I .
1 .1 i E 1152 "19E, '84.
FIGURE 7
DISTILLATE FUEL DEMAND
39
3.0-
2.8 -
2.6
2.4A
2.2 -
2,0-
1.8
1.6
1.4 -
1 .2 -,:
.o.,rJ,
06
06.4
0.0 -
1964 1 9EC 1972 1 S-713 1980 1984-
YEARM
FIGURE 8
KEROSENE DEMAND
11-
10-
/ '+t" "Ru
N
r .t
6-
WA.
-, -- " - -"" -
.,_ _ _ _--__._ . ._____..._. __

_.___ _ _ -- s V.
I . . . . .
1 $ ;4.19E8 197' 1 -,5, 19E;(]O') 4
YEAR
FIGURE 9
LPG DEMAND
40
8*
4, .
2-
NORM DAKOTA
SOUTH DARR
1964 1 817 -7618
FIGURE 10
ASPHALT & ROAD OIL DEMAND
6-
5
/0-
IT
on~
- - - - t ,I- ---
C ) I , I I I I I I I 1 I I 1I
1 G64, 1 )E8 1972 1V73 19E;0 984.
FIGURE 11
RESIDUAL FUEL DEMAND
41
I I I 17 II VI'-
" I 1
FIGURE- 217
- - -
1 _ _ - I- -i k
20196 1. 1 E' 197 1 £,'7,3 1 9E0 9 84.
90 - ' , FIGURE YEAR

12
OTHER PETROLEUM PRODUCTS DEMAND
170
10 /
I 70-
"
30-
6 Wf 0 AKNaANA
p b
1~~~ ~ ~ ~ ~97 G9 13EE*4.1L;
YEAR
FIGURE 13
TOTAL PETROLEUM PRODUCTS DEMAND
42
%r.
After gasoline, the next boiling fraction is usually defined as naphtha. In
an alternative definition, the naphtha boiling range includes gasoline, which
is referred to as refined petroleum naphtha. Within a refinery, naphtha is a
process intermediate; it is comnonly fed to a steam reformer to make addi-
tional gasoline blending components. In this usage, naphtha may refer to a
stream boiling above gasoline and up to 500 or even 600 IF endpoint. "Light
naphtha" may have an endpoint of 400 OF. When prepared as a finished product,
"naphtha" denotes a more specific type of product with a narrow boiling range.
Here the terms "solvent" and "naphtha" are often used interchangeably.
Specialty naphthas may have endpoints as low as 220 IF.
Kerosene is usually considered to have a boiling range higher than naphtha,

such as 300-550 °F. Thus a kerosene-type jet fuel may have a 20 percent dis-
tilled point of 400 OF, compared to a 20 percent distilled point of 293 OF for
a naphtha-type jet fuel. The allowable end points are not so far apart; thus
the naphtha-type fuel may be called a "wide-cut" fuel.
The term distillate of course embraces any material which comes overhead in an
atmospheric distillation column. The bottoms from the colunn are residual
fuel. Light distillate may refer to gasoline and light naphtha. Middle dis-
tillate includes kerosene and the distillate fuel oils, whereas gas oil in-
cludes only the fuel oils. The term gas oil descends from when this fraction
was used to make illuminating gas.
The distillate fuel oils include #1, #2, and #4 heating oils, and #1, #2 and
#4 diesel fuel. However, #4 fuel oil may be made by blending residual fuel
with varying proportions of distillate. Thus #4, #5 and #6 are termed "black
oils," while #1 and #2 are "white oils."
Grade #1 fuel oil is designed for use in vaporizing pot-type burners, while
#2 fuel oil is designed for use in atomizing type burners for domestic heating
or rmderate capacity comnercial/industrial burner units. Grade #4 fuel oil is
intended for conercial burner installations not equipped with fuel preheating
facilities. It is used extensively in industrial plants.
Diesel fuels are distillate fuels for use in compression-ignition engines.

Diesel #1 is for use in high-speed diesel engines generally operated under
variations in speed and load. It includes type C-B diesel fuel used for city
buses and similar operations. Diesel #2 is a lower volatility fuel, designed
for high-speed diesel engines which are operated under generally uniform load
and speed. It includes type R-R for railroad locomotive engines and type T-T
for trucks. A #4 diesel is intended for low and mdium-speed diesel engines.
Residual fuel, which includes #5 and #6 fuel oil, and Bunker C fuel oil, is
used for the production of electric power, industrial steam generation, vessel
bunkering, large comiercial heating systems, and various other industrial pur-
poses.
Jet Fuel
Jet fuel specifications were presented earlier. Specifications provide limits

on certain parameters. It is also of interest to know how close the fuels ac-
tually being marketed come to those limits. Table 20 presents the results of
the fuels testing program carried out at the National Institute for Petroleum
43
TABLE 20
AVUAMIGE JET FUEL PROPERTIES IN THE UNITE) STATES
JP-4 JET-A JP-5
Type Naphtha Kerosene Kerosene
Gravity, °API 54.7 42.2 41.0
RVP, psi 2.6 - -
Freeze Point, OF -82 -50 -54
AS= )86 Distillatim

Percent Overhead
10% 195 369 387
50% 277 414 417
90% 378 472 462
Aromatic Content, Vol% 12.3 17.8 18.8
Olefin Content, Vol% 0.8 1.1 0.7
Heat of Combustion, 18,706 18,698 18,504

Net, BTU/lb
Source: Bartlesville Energy Technology Center, 1986
44
and Energy Research (NIPER). The average aromatics content of JP-4 is well
below the specification value of 20 percent.
Gasoline
Motor Gasoline is a conplex mixture of relatively volatile hydrocarbons, with

or without small quantities of additives, that have been blended to form a
fuel suitable for use in spark-ignition engines. Specifications for motor
gasoline, as given in ASTM Specification D439 or Federal Specification VV-G-
1690B, include a boiling range of 122-158 OF at the 10 percent point to
365-374 OF at the 90 percent point and a Reid vapor pressure range from 9 to
15 psi. "Motor gasoline" in the past included finished leaded gasoline,
finished unleaded gasoline, and gasohol. However, leaded gasoline will be
eliminated in the future.
Gasoline properties are changed from sunmer to winter as a function of the

prevailing tenperature. Table 21 shows the sumimr/winter variation for the
district in which the Great Plains plant is located.
Kerosene
Kerosene is defined as a petroleum distillate that boils at a tenrperature be-

tween 300-500 OF, that has a flash point higher than 100 OF by ASTM Method
D56, that has a gravity range from 40-46 degrees API, and that has a burning
point in the range of 150-175 OF. Included are the two classifications recog-
nized by ASTM D3699: #1-K and #2-K, and all grades of kerosene called range
or stove oil which have properties similar to #1 fuel oil, but with a gravity
of about 43 degrees API and a rmximrun end-point of 625 OF. Kerosene is used
in space heaters, cook stoves, and water heaters and is suitable for use as an
illuminant when burned in wick lanps.
Diesel Fuel
Diesel fuel #1 and #2 specifications are given by ASTM Specification D975.

These specifications are listed in Table 22. By conparison, the average
properties of #1 and #2 diesel fuel sold in the Great Plains region are given
in Table 23.
Fuel Oil
The specifications for #1 and #2 fuel oil are basically the same as for #1 and
#2 diesel fuel (ASTM D975). Average properties of #1 and #2 fuel oils sold in
the Great Plains region are given in Table 24.
ASTM D396 specifications for #4, #5 and #6 fuel oils (including #4 diesel) are
listed in Table 25. Very little #4 and #5 fuel oil is sold in the Great
Plains region. Average properties of #6 fuel oil sold in the region are given
in Table 26.
Sunmary
All petroleum products are sold to a set of specifications which have been
developed through a marriage of consurers' needs and refiners' capabilities.
A simple distillation of lignite tar oil would not produce any products
45
TABLE 21
AVR4G WE.EADWD W1VR G&BCLINE ?OERTIES IN GEAT PLAINS RIBIN

(NIPER District 9)
Regular P remiumn
Sumimr/Winter Summr/Winter
(R+M)/2 (R+M)/2
Below 90 Above 90
Gravity, OAPI 60.1/62.9 59.4/54.6

Sulfur, Wt% 0.023/0.029 0.017/0.009
Benzene, Vol%
RON 91.6/91.5 95.9/96.7
MON 83.4/83.6 87.8/88.6
(R + M)/2 87.5/87.6 91.8/92.7
RYP, psi 11.3/14.7 11.1/14.7
hSIU-D66 Distillatiem, Temp, OF
Initial Boiling Point 83/78 81/71

5 % Overhead 99/86 99/79
10% 115/100 123/98
20% 138/120 151/128
30% 161/143 182/161
50% 209/195 228/211
70% 257/249 262/241
90% 333/333 331/307
95% 368/372 362/353
End Point 414/420 415/408
Residue, Vol% 0.9/0.8 1.0/1.0
46
TABLE 22
SUMMU OF ROQJUIRME FM DIESE. FUELS
Grade No. 1-D No. 2-D No. 4-D
Flash Point °F, Min 38 52 55
Carbon Residue On 10%

Residual, Percent, Max 0.15 0.35 -
Ash Wt%, Max 0.01 0.01 0.10
Distillation Tenp To
Give 90% Overhead,
Min - 282 -
Max 288 338 -
Kinematic Viscosity
cSt At 40 OC
Min 1.3 1.9 5.5
Max 2.4 4.1 24.0
Sulfur Wt%, Max 0.50 0.50 2.0
Cetane No., Min 40 40 30
Source: ASTM D-975
47
TABLE 23
TYPICAL PROETYIES OF DIESEL rULS

in THE GRET PIAJS REGION
(NIPER Central Region)
No. 1-D No. 2-D

Low High Low High
Gravity, OAPI 41.0 44.6 29.8 40.8
Flash Point, *F 127 169 137 201
Cloud Point, *F -26 -26 -14 24
Sulfur, Wt% 0.0 0.3 0.100 0.415
Ash, Wt% 0.001 0.008 0.000 0.010
Cetane No. 45.5 47.0 37.8 54.7
Initial Boiling Point 288 393 315 429
10% Overhead 365 412 389 485
50% Overhead 408 452 485 544
90% Overhead 468 510 556 632
End Point 494 545 594 677
48
TABLE 24
AVMW NO. I AM) ND. 2 IUU OIL MEWrIIS

I M ul (7MAT PIW
M I IINI
EIM
No. 1 No. 2
Gravity, °API 42.4 34.1
Viscosity, 100°F, Centistokes 1.64 2.71
Sulfur, Wt% 0.129 0.260
Aniline Point, OF 130.9 127.7
Carbon Residue, Wt% 0.081 0.127
Gross Heat of Combustion, JkTU/Gal 134.3 139.2
AS= Distillatim, OF
Initial Boiling Point 345 367
10 382 427
50% 429 509
9096 489 598
End Point 530 647
49
/
TABLE 25
SPBCIFIC&TIM FOR R38 IIAL FM OILS
Water
and Kinenntic
Flash Pour Sedi- Viscosity
Point, Point, ment, Ash, cSt at
Grade Of OF OF Vol% Weight% 100OF Sul-
Fuel Oil Min Max Max Max Min Max fur %
No. 4 130 20 0.50 0.10 5.8 2 6 .4 A legal
Preheating not
usually required
No. 5 (Light) 130 - 1.00 0.10 26.4 6 5A legal

Preheating nay
be required
No. 5 (Heavy) 130 - 1.00 0.10 65 1 94 A legal

May need pre-
heat ing for
burning; and in
cold clirmtes,
for handling
No. 6 130 B 2.00 - legal
Preheating re-
quired for
burning and
handling
A Where low sulfur oil is required, fuel oil falling in the viscosity
range of a lower numbered grade down to and including No. 4 nay be sup-
plied by agreement between purchaser and supplier.
B Where low sulfur fuel oil is required, Grade 6 fuel oil will be class-
ified as low pour, 15 degrees C (600 F) mrax, or high pour (no nax). Low
pour fuel oil should be used unless all tanks and lines are heated.
50
TABLE 26
AVEAGME ND. 6 FURL OIL PROPERTIES

IN =9 &T PLAINS REGION
Gravity °API 11.6
Sulfur, Wt% 0.86
Ash, Wt% 0.010
Gross Heat Of Conbustion (MBTU/Gal) 153.4
Source: Bartlesville Energy Technology Center,

1986
51
matching existing specifications. However, modern refining processes are
available to mrodify the characteristics of hydrocarbon streams almost at will.
Large mlecules can be cracked and small mlecules can be combined. Un-
saturated compounds can be hydrogenated and saturated compounds can be
dehydrogenated. Ring compounds can be cracked and linear compounds can be
reformed into rings. Thus there is no doubt that, with sufficient processing,
the Great Plains liquids could be formed into specification fuels.
Most fuel specifications are stated in terms of physical properties rather

than chemical composition. Thus there is no inherent barrier to the use of
coal-derived liquids. The major exceptions are limits on aromatics content
for jet fuels and a proposed similar limit for diesel fuels. Although
aromatics in diesel fuel are not currently limited by specification they lower
the cetane nunber and are undesirable for that reason. The high aromatics
content of the Great Plains liquids would have to be nmdified for these ap-
plicat ions.
Prices
Prices paid for petroleum products in the Great Plains region over the last
four years are sunxmrized in Tables 27 through 32. The generally small dif-
ferences in prices between states attests to the efficiency of the United
States petroleum products distribution system and to the keen competition be-
tween refineries. As would be expected, prices are slightly higher in Min-
nesota, which is a product importer, than in Wyoming, which is a product ex-
porter.
In a number of cases there are not enough sales of a particular fuel in some
states of this region to form a statistical base. For this reason, the
Petroleum Administration for Defense District (PADD) averages have been in-
cluded in Tables 27 through 32. PADD definitions are shown in Figure 14. It
is unusual to see a spread of mre than three or four cents per gallon in fuel
prices between states, which corresponds to the cost of mDving fuel about 150
miles by truck.
Market Outlook
It Is unlikely that petroleum products in general will again show the large
year-to-year increases in demand which were common prior to the price shocks
of the 1970's. However, gasoline, the largest volume product, could be poised
to show some gain in volume. The federally mandated increases in Corporate
Average Fuel Econonb' have reached their legislated limits. Average automobile
mileage will continue to increase for some time as new cars replace older
less-efficient cars on the road. The average mileage for new autormbiles,
however, is unlikely to increase from present levels. Gasoline demand should
therefore be fairly stable in the future.
Conmercial jet fuel is expected to show continued gains in demand. The

restructuring of labor rates in the airline industry has dropped fares to the
point where they are stirnulating increased travel. As consumer desires for
goods such as autos, home appliances and electronics becomes sated, there ap-
pears to be an almost unlimited desire for travel.
Kerosene is an obsolete product with little chance of revival.
52
<.-.; <.-;X-'< -:' .-A,

?..<, :< ..; "
TABLE 27
SALES IF( RESALE PRICES

OF IULEADW GASOLINE
(Cents Per Gallon Excluding Taxes)
1983 1984 1985 1986
Padd 11 Average 93.0 85.0 89.6 49.13

Minnesota 95.5 86.8 91.0 50.1
North Dakota 95.6 86.2 91.5 48.9
South Dakota 95.6 84.5 90.8 47.3

Padd IV Average 91.2 86.7 90.3 50.4
Montana 89.3 85.3 89.3 53.0
Wyoming 92.5 90.0 90.4 47.7
TABLE 28
SALES V~ft RESALE PRICKS

OF KH1OE-Ty'K JET FUE
1983 1984 1985 1986

Padd II Average 89.7 86.1 82.6 45.9
Minnesota - 82.9 85.3 47.7

North Dakota -94.8 - -
South Dakota - 95.6 93.5 -
Padd IV Average 90.9 88.2 83.3 50.7

Montana - 87.7 - -
Wyoming -95.0 94.7-
53
F
TABLE 29
SALES RO RESALE PRICES

0MOIM
or
1983 1984 1985 1986
Padd I Average 96.0 91.8 95.7 54.8
Minnesota - 93.8 100.9 55.5
. - -
North Dakota
- - - -
South Dakota
Padd IV Average 90.9 90.9 93.0 54.4
Montana - - - -
Wyoming ... - -
TABLE 30
SALES FOR RESALE PRICES

OF NO. 1 DISTILLATE
1983 1984 1985 1986
Padd II Average 92.1 87.2 94.0 50.6
Minnesota - 89.0 95.7 52.9
North Dakota - 90.6 96.0 53.0
South Dakota - 89.6 95.1 51.9
Padd II Average 91.2 90.1 91.0 50.8
Montana - 91.2 91.6 50.4

90.8 90.9 50.9 Z'
Wyoming
54 4-
*1'.-
TABLE 31
SALUM PR RESALE PRICES

OF ND. 2 DIIILLTE
1983 1984 1985 1986
Padd II Average 86.4 81.9 87.9 44.9
Minnesota 87.9 83.1 89.3 47.9
North Dakota 87.7 - - -
South Dakota 87.6 - - -
Padd IV Average 84.4 83.7 83.9 46.2
Montana 83.9 - - -
Wyoming 84.1 --
TABLE 32
SALES FOR RESALE PRICES

OF RSIUM FURL
1983 1984 1985 1986
Padd 11 Average 62.4 62.4 52.9 28.7
Minnesota - 56.9 - -
North Dakota - --
South Dakota - - - -
Padd IV Average 59.3 57.8 52.6 27.3
Montana---
Wyoming
55
ninin.
P. An= a s W W-1V
Diesel fuel demand is strongly linked to overall economic activity. Trucking,

railroads, and construction all exhibit cyclic demand patterns which will con-
tinue in the future.
Heating oils are in a long-term decline due to loss of markets to natural gas.
This loss is mostly irreversible.
Residual fuel is regaining some markets which were earlier lost to natural gas
and coal. If oil prices rise even moderately, however, these markets will
again be lost. This is because the differential price between crude oil and
residual oil is considerably smaller than it was a few years ago. The dif-
ferential has shrunk because nmre refiners have installed processes for con-
verting residual fuel to higher-value products.
Because of strong conpetition in the petroleum refining industry, all refined

product prices move in relationship to the price of crude oil. Because of the
dominance of OPEC in supplying crude oil, world oil prices are set not by com-
petitive supply and demand, but by OPEC policy. Price projections for oil
products therefore amount to projections of OPEC policy. If oil demand were
to decline below present levels, OPEC would lose its power to dictate prices.
But with even a modest increase in demand, and the drop-off in non-OPEC
production which is occurring at early 1987 price levels, OPEC's position
clearly should strengthen. The forecast for this study, then, is that world
oil prices will recover slowly, regaining 1985 levels (in constant dollars) by
1992, and 1980 levels by the year 2010.
1A VAS
°%
W .... ...
uu -WIS.
7~ 01o4
S,,OAK-
.r NJ.
S UAH : "'.
NOW OF
KY.5
N.C..
d.~~GA
d
FIGURE 14
-5 . C
ARK.
PETROLEUM ADMINISTRATION FOR DEFENSE (PAD) DISTRICTS%

AL
WRINYU FKUMWTU1K
A number of properties of the Great Plains tar oil make it undesirable as a

refinery feedstock. These include high levels of oxygen, nitrogen and par-
ticulates. Conventional refineries would be unable to accept the tar oil as
feedstock unless blended down to very low concentrations. The configurations
of refineries in the Great Plains region were evaluated to determine their
hydrotreating capabilities. Table 33 lists the refineries in the defined
market area, and their hydrotreating capabilities. As used in the table,
hydrocracking includes those processes where 50 percent or more of the feed is
reduced in nmlecular size. In hydrorefining, 10 percent of the feed or less
is reduced in molecular size; and for hydrotreating, essentially no reduction
in mlecular size of feed occurs. None of the refineries in the region have
the capability to hydrotreat the whole crude stream, as would be required for
tar oil. Only the Exxon refinery in Montana lists hydrocracking capability,
and this is for distillate upgrading only.
There is, however, one interesting future possibility for direct sale of the
tar oil stream as refinery feedstock. Consumers Cooperative Refinery in
Regina, Saskatchewan is building a heavy oil upgrading system to allow it to
process the very heavy oils of Saskatchewan and Alberta. The Canadian federal
and provincial governnnts are providing financial aid to the project, which
is being carried out by New Grade Energy, Inc., a partnership of Consumers
Cooperative Refinery Ltd. and the Saskatchewan government. The 50,000 barrel
per day project is expected to cost $650,000,000. Consumers Cooperative is
providing 5 percent of the equity, and the provincial government 15 percent.
The federal and provincial governments are providing loan guarantees for the
remining 80 percent of the cost as debt.
The project is well along in construction and should be finished next year
(1988). Hydrocracking processing technology is provided by Unocal, with their
Unicracking process. Unicracking carries out hydrodesulfurization and mild
hydrocracking using a fixed catalyst bed. It is normally used to upgrade
residual fuel oils and many vacuum resids by remving significant amounts of
metals, sulfur, nitrogen and residual carbon. Feed and hydrogen-rich recycle
gas are preheated, mixed, and fed to a guard chanber, where a smnll bed of
catalyst renmves particulate matter and residual salt. Effluent flows into
one or more main reactors, which contain nmst of the catalyst. Liquid product
is cooled, separated and sent to further processing.
According to Unocal, the major feedstream constraints are on oxygen, nitrogen

and particulates. If the oxygen level is too high, the resulting partial
pressure of water is deleterious. It destroys the catalyst support. The
oxygen level of 2.7 percent in the Great Plains tar oil would be borderline
acceptable. Because the total tar oil stream would amDunt to only about 6
percent of the total feed to the upgrader, however, oxygen content is not seen
to be limiting.
Similarly, the maxirmun allowable nitrogen level to the hydrocracker is about

5,000 ppm. The Great Plains tar oil has a slightly higher level (6,000), but
not enough higher to be a problem for blending.
The limiting factor is likely to be particulates. Because a fixed catalyst

bed is used, Union specifies nnximum particulates in the 2-5 parts per million
57
range. Although good nunbers are not available for the Great Plains tar oil,
one analysis of sediment by extraction gave 0.1 weight percent. If this
material were all particulates by Unocal's definition, it would be 1,000 ppm,
or too high by two orders of rmgnitude. Some kind of filtration would have to
be provided.
Consumers Cooperative currently brings in some crude oil from North Dakota and
Montana, and the trans-border movement of feedstock Is not seen to be a
problem. Because of favorable government financing assistance, the NewGrade
project will likely have a lower break-even point for operating rmrgin than a
conventionally financed hydrocracker. It should therefore be able to conpete
for feedstock on attractive terms. If reasonable transportation costs could
be arranged with the railroads, this might provide an outlet for the Great
Plains tar oil yielding a higher netback than sale as No. 6 fuel oil.
Two possible issues to be addressed are the strong odor associated with the
naphtha and to a lesser extent, the tar oil stream and treatent of the was-
tewater streams. The refinery would want to be assured that those would not
be problems.
58
TABLE 33
CATALYTIC HYDIOTURTING CAPABILITIES Of REFINEIES

IN (REAT PLAINS REGION
(Quantities In Barrels Per Stream Day)
Crude Hydro- Hydro- Hydro-

Refinery Capacity Cracking Refining Treating
Manitoba
No Refineries
Minnesota
Ashland 69,220 --- 23,000 39,700
Koch 160,000 --- 63,500 86,000
Subtotal 229,220 --- 86,500 125,700
Montana
Cenex 42,500 --- 14,000 15,000
Conoco 50,000 ---. 41,500
Exxon 44,000 4,900 --- 41,500
Montana 6,500 ...... 2,200
Subtotal 143,000 4,900 14,000 100,200
North Dakota
Anoco 60,000 --- 15,000
Subtotal 60,000 --- 15,000
Saskatchewan
Consumer's 40,000 --- 9,000
Petro-Canada 13,000 ---
Subtotal 53,000 --- 9,000
South Dakota
No Refineries
Wyoming
Armco 41,000 --- 7,100
Frontier 35,500 --- 4,350 7,000
Little America 24,500 --- 7,200
M ou nta inee r 400 .... ... ..
Sinclair 54,000 --- 10,000 14,000
Wyoming 13,500 .........
Subtotal 170,189 --- 14,350 51,050
Total 655,709 4,900 114,850 300,950
59
i
N
WEWXMU VD17VrrE.%%
Definitimn and specifiatims
Creosote nay be loosely defined as the fraction of coal tar distillate boiling
in the range 400OF to 650*F. It is sometimes referred to as "dead oil," in-
dicating that it is the higher boiling fraction of the oil obtained by dis-
tilling coal tar. Rather than producing a virgin distillate in this range,
creosote is often produced by blending various tar oil fractions.
The oils are blended to give creosotes conforming to certain user specifica-
tions. The major outlet for coal-tar creosote is for the preservation of tim-
ber. Although other fungicides have been increasingly used, coal-tar creosote
has been the most widely used wood preservative. It is particularly effective
in prolonging the life of wood that is used outdoors and in contact with the
ground, or seawater or freshwater. Exanples are telegraph and electricity
transmission poles, railway ties, marine piling, fencing, and farm buildings.
To give these structures service lives of 50 years or rmre, they are inpreg-
nated under pressure with creosote so that the vulnerable sapwood cells arn
either filled with oil (by the Bethell or "full-cell" process generally used
for marine piling and railway ties), or coated with a film of creosote (by the
Rueping or Lowry "enpty-cell" processes more generally used for poles, fenc-
ing, and farm buildings).
Other minor uses for creosotes are as horticultural washes to kill overwinter-
Ing pests on fruit trees, and for making disinfectant eniulsions and sheep
dips.
Specifications for creosote for tintber preservation are published by the

American Wood Preservers' Association (AWPA). Standard Pl-78 is the nost
widely used, general-purpose standard (Table 34). Standard P7-85 gives the
requirements for a crystal-free product to protect cut surfaces of treated
wood, and P13-85 specifies the properties of a marine grade creosote for piles
and tinbers (Table 35). Solutions of coal tar in creosote are covered in
Standard P2-85 (Table 36). These solutions ay be used in applications where
the surface condition of the treated wood is not inportant. A nmarine grade
creosote-coal tar solution, standard P12-85, gives better protection against
marine borers than Standard P2-85. In general, the more severe the service,
the higher the required specific gravity (Table 37).
As seen in Tables 34 through 36 the AWPA standards define creosote as a dis-

tillate of coal tar produced by the carbonization of bituminous coal. Thus
the low-temperature lignite tar produced at the Great Plains plant is excluded
a priori. Also, it probably would not meet the specific gravity specifica-
tions. Nevertheless, there is reason to believe that, from a performance
standpoint, the Great Plains tar should be satisfactory. Low temperature lig-
nite tars have been used successfully for wood preservation In the past. The
existing specifications are wholly enpirical; the requirements evolved through
a conpromise between the consumers' needs in timber preservation and the
producers' processing capabilities and coal tar feedstocks.
60
V -v-~~~ ~ .,".,".'-
~ i
TABLE 34
ANPA P1-78 (EVISUD) STAISARD

FOR (DAL TAR (aOhe=
The creosote shall be a distillate derived
entirely from tar produced by the
carbonization of bituminous coal.
Not Not
Less More
Than Than
Water, Percent By Volun- 1.5
Matter Insoluble In Xylene
Percent By Wt 0.5
Specific Gravity
Whole Creosote 1.050
Fraction 235-315 °C 1.027
Fraction 315-355 °C 1.095
Distillation: Percent By
Wt On A Water-Free
Basis:
Up To 210 °C 2.0
Up To 235 °C 12.0
Up To 270 °C 10.0 35.0
Up To 315 °C 40.0 65.0
Up To 355 °C 60.0 77.0
61
TABLE 35
ANFA P13-85 STANDAE)D FOR AL TAR CRBDSTB 70 BE USED IN

ThE TWZAMI OF NRINK PILES AM) TIMERS
The creosote shall be a distillate derived entirely from

tar produced by the carbonization of bituminous coal.
Not Not
Less More
Than Than
Water, Percent By Volum 1.5

Matter Insoluble In Xylene,
Percent By Weight 0.5
Specific Gravity:
Whole Creosote 1.080
Fraction 235-315 *C 1.090
Fraction 315-355 °C 1.105
Residue Above 355 *C 1.160
Distillation: Percent By
Wt On A Water-Free Basis
Up To 210 °C ---- 2.0
Up To 235 0C ---- 12.0
Up To 270 °C 20.0 40.0
Up To 315 °C 45.0 65.0
Up To 355 °C 65.0 75.0
62
tl4
00
Z 0
4 -.
V
0 z 0 La
04 m q *; W C4
(d%4
C)a t-
000
* 0
~ aa
*003al0
-4-L
0. 41 mefM
U4 A4 L "V4M. P 00
4.0 41 1 d 4t
0: CD inC4C2
10 C.) 0c !0 0
63E-t
TABLE 37
SPECIFIC (RAVITr OF AXPA CREDON GRADES
Specific
Standard Gravity
P1 Coal Tar Creosote 1.05

P2 Creosote Solutions 1.06
P7 Creosote For Field Cuts 1.06
P13 Creosote For Marine Piles 1.08
P12 Creosote Solution For Marine Piles 1.09
120-
210-
19 7 1901- 91 99 1
50 AR
FIUR-1
70O8T PRDCTO
Same users are willing to utilize material which does not meet AWPA specifica-
tions. Concrete evidence of this is provided by previous sales of the Great
Plains material and by a long history of sales of creosote from the Lurgi low-
temperature lignite carbonization plant operated by Husky Industries in Dick-
ensen, North Dakota (this plant was shut down in 1986 because of environmental
problensp but had operated since 1926). The Husky plant reopened January 1987
with a Nichols hearth furnace which does not produce by-product liquids.
The most likely near-future method of utilization would be as a blend with

normal creosote. AWPA Standard P3-67 "Standard for Creosote-Petroleum Oil
Solution" allows for blending up to 50 percent petroleum-derived oil with
creosote for some applications. Som western railroads but not eastern, are
known to use creosote-petroleum oil solutions. The Great Plains creosote
should definitely have better preservative effect than petroleum oils. The
availability of a large, secure supply of lignite creosote from the Great
Plains plant may prouiote evolution of the AWPA specifications, as well as ac-
ceptance by consumers. However, considerable testing by the AWPA and in-
dustrial users would be necessary to demnstrate satisfactory performance.
Product sold to date by Great Plains has been found unsatisfactory because of
the odor.
Product ion
United States production of creosote for the years 1967 to 1985 is shown in
Figure 15. A significant decline on the order of 50 percent has occurred
since 1974. Historically, total consumption has been anywhere from 10 to
30 percent higher than production due to net imports from Europe. The quan-
tity inported varies readily with currency exchange fluctuations. A srmll
amount of creosote has been exported to Canada.
Prices
Historical creosote prices, as quoted in the Chemical Marketing Reporter, are

presented in Figure 16. Prices were stable at $0.24 to $0.27 per gallon from
1960 to 1973. They than began to climb rapidly, reaching a peak of just under
$1.20 per gallon. Although these are prices quoted for bulk quantities, we
find that they do not represent prices which might be expected for large scale
contract sales. Such prices have dropped significantly in the last year, from
approximately $1.00 per gallon to $0.80 per gallon. These prices are for
specification grade creosote delivered to the buyer's plant. The last price
received by the Husky carbonization plant in North Dakota for its non-
specification low-teirperature lignite creosote was $0.60 per gallon.
Comsummrs
The locations of individual creosote treating plants listed by the American

Wood Preservers' Institute are shown in Figure 17. These plants are con-
centrated in the Southeastern United States, perhaps because of the combina-
tion of a large timber industry in the region, plus a humid climate which
makes tinter preservation necessary for long life. Som of the largest pur-
chasers of creosote are Kerr-McGee and Southern Wood Piedmont. It is es-
tirated that these two conpanies each use around eight million gallons per
year, or about 10 pers.nt of the total defend, each. AImost all of this usage
is for railroad ties.
65
Si .20...
1.10
10.70
S. .0
74 1977 1980) 19,3 1986

YEAR
FIGURE 16
CREOSOTE PRICES
TAIKS,WOE89 FAB
:"". I I I I
FIGURE 17
WOOD PRESERVERS USING CREOSOTE
p66 .
On an industry-wide basis, over 65 percent of creosote usage is for railroad
ties (Table 38). The data in Table 38 refer only to cubic feet of wood
treated, without regard to the treatment procedure. Because railroad ties are
always pressure-treated and other nmterials nay not be, the fraction of total
creosote usage accounted for by railroad crossties is higher than suggested in
Table 38.
The mst recent data available (1984) from the Anerican Wood Preservers In-
stitute show that total creosote usage in that year consisted of 41 million
gallons of creosote and 61 million gallons of creosote-coal tar solutions.
Producers
Producers of creosote listed by the International Trade Comnission include:
Allied Corporation
Reilly Tar and Chemical
Koppers Conpany
United States Steel Corporation
Witco Chemical Corporation
Coopers Creek Chemical Corporation
TABLE 38
CRNOwr-TRE&TED If= PRwOr

(1984 Statistics)
Product Cubic Percent

Feet Total
Railroad Crossties 88,720 65

Switch & Bridge Ties 8,198 6
Poles 22,639 16
Piling 6,850 5
Fence Posts 1,675 1
Lumber 3,892 3
Tinters 4,179 3
Other 1,441 1
137,594 100
Source: International Statistics Council, 1986
67
~
(~v 44 ~ ~ ' ,- .~-*~ A,.- -. *. -. *-* - - --..- *
Market structure
The creosote market is dominated by railroad requirements for crosstles.

Railroads who operate their own tie-treating facilities normally purchase
creosote on the basis of delivery anywhere on their rail system. They than
provide transportation to their tie treating plant.
Tie treating plants have encountered serious environmental problems over the
last few years. Some railroads, Burlington Northern for one, will contract
for treated ties rather than operate their own plants.
The largest volumes of creosote are purchased on once-a-year solicitations

issued by the major consumers. Same operators of tie-treating facilities have
their own captive sources of creosote. Some is marketed by middle-men, or in-
dependent brokers at fees of up to 10 percent of the purchase price.
Market Out look
Overall consunption of creosote for wood preservation has declined slowly over
the last 30 years while the use of water-borne preservatives has grown
strongly (Figure 18). Creosote usage has been sustained by the demand for
railroad crossties, for which creosote has been the only accepted preserv-
at ive.
The introduction of concrete crossties has taken away some market share, al-
though not a significant share to date. Because of environmental and health
hazard concerns, it is very unlikely that creosote demand will show appreci-
able growth in the future. The market is vulnerable to replacement by con-
crete crossties or inproved water-borne preservative conpounds. Meanwhile,
the continuing consolidation of railroad trackage systems is reducing the to-
tal market potential for replacement crossties. Past and estimated future
demand for creosote-treated crossties, as prepared by the Railway Tie Associa-
tion, are shown in Figure 19. A significant decline from recent demand levels
is expected. Furthermore, the Railway Tie Association's projections do not
consider increasing use of concrete ties. An Australian-American joint ven-
ture has recently begun to build a facility in Denver to manufacture some 3.5
million concrete ties for the Burlington Northern railroad by 1992. Because
concrete ties cannot be intermingled with wood ties, market penetration will
be limited in the near future.
If uses other than crossties and pilings were eliminated, demand could drop
below 40 million gallons per year. Assuming that lignite-derived creosote
would not be used at over 50 percent concentration, there would be a maxinum
nmrket potential of 20 million gallons or 175 million pounds per year. Poten-
tial creosote production from the Great Plains plant is 128 million pounds per
year. Thus, disposing of all Great Plains heavy oil as creosote would require
capturing 73 percent of the potential long-term market in the United States.
Canadian markets would expand the potential but not greatly. The Canadian
market is estimated at less than 10 percent of the United States market. It
may be concluded that the Great Plains creosote would have to be deeply dis-
counted in order to place it all in the wood preservative market. Thus, op-
tinum marketing strategy would be not to convert the entire Great Plains tar
oil stream to creosote, but to conbine creosote production with coproduction
of rubber processing oils, coal tar asphalt, etc.
68
160
120-
110-
6-
30
20 -
- c - N EDFBO
FIUR 18
26-
15-
214-
FIUR 19
1669
15 k' . *4 b1
Although AWPA specifications state that creosote must be derived from coal
tar, creosote prices are not entirely insulated from petroleum prices. This
is because, first, some uses allow the substitution of up to 50 percent
petroleum oil, and second, alternative uses for the creosote fraction conpete
directly with petroleum. Thus decreases in petroleum prices force down
creosote prices for alternative uses, creating pressure on the prices for wood
preservative. Because the contracting cycle for large-volume creosote users
Is 12 months long, the effects of the early 1986 crash in oil prices is only
now being translated fully into the creosote market.
Figure 20 shows clearly that creosote price changes lag, but eventually follow
oil price changes. Because creosote can substitute for petroleum as a fuel,
upward trending oil prices quickly stirmlate upward trending creosote prices.
Because petroleum cannot substitute for creosote, however, downward trending
oil prices will affect creosote more slowly. The down trend in petroleum
prices in the last few years has been temporarily counter-balanced by
decreases in creosote capacity. The historical relationship in Figure 20 sug-
gests that since 1987 oil prices correspond to about 1979 oil prices, 1987
creosote prices should correspond to about 1979 prices. This would be about
$0.80 per gallon. Contact with creosote buyers indicates that this is indeed
the market level, although not yet reflected in the Chemical Marketing
Reporter data.
$60
CREOSOTE
I %4-0 qo
$Vo
OI
510
1967 1970 197'- 1976 13 193Ie;

YEAR
FIGURE 20
CREOSOTE AND OIL PRICES
70
CRKSOLS AND 43LIC ACID
Definitions and Specifications

,5
The Phenosolvan crude phenol contains nmstly nmnohydric "phenols", which in-
clude phenol, cresols, xylenols, ethylphenols, and other alkylphenols as well
as som polyhydric "phenols", including catechols and resorcinol.
The term "cresols" is used to designate the isolated monomethyl phenols,

either as pure ortho, meta, or para isomrs or as mixtures of isomers. This
report uses the term "phenol" to denote the chemical monohydroxybenzene and
the term "crude phenol" to denote the conplex mixture of mno- and polyhydric
aromat i cs.
Crude phenol can be separated into fractions which can be designated as tar
bases, neutral oil-free phenol, cresylic acids, and phenolic pitch.
"Cresylic acid" refers to various mixtures of tar acids, mstly polyhydric

phenols and monohydric phenols higher than C6. Cresols may be the chief con-
stituents in some mixtures, in others the xylenols (dimethyl phenols) and
higher boiling homelogs of phenol may predominate. The individual conpounds
all have boiling points in the range 180-2300C.
Phenolic resins manufactured with cresols and other constituents of crude

phenols require feedstocks typified by the following:
- Meta-cresol 50-70 percent

- Not nmre than 0.25 percent neutral oil
- Not rmre than 0.17 percent pyridine base
- Distillation range as required by consumerr
Conrpanies that report production data to the United States International Trade
Commission define cresols as phenolic mixtures in which 50 percent of the
material bolls below 204 0 C and cresylic acids as phenolic mixtures in which .5,
50 percent of the material boils above 204 0 C. The category "cresylic acid,
refined" includes primarily cresol/xylenol mixtures, many of which contain a
high percentage of xylenol. It may also include some xylenol mixtures sold
either as is, or used for further separation of the individual xylenol
isomrs. Sam higher boiling phenolic tar acids may also be included in the
refined cresylic acid category. Refined cresylic acid is defined the same as
crude cresylic acid except that the yield of product distilling below 2150C is
more than 75 percent by weight of the original distillate.
Specific end-uses will require limitations on various minor conponents present

in the Great Plains liquids. For instance, guaiacol causes problems for cer-
tain products. Other uses require low pyridine limits. In other cases the
phenol content must be zero.
Product ion
Total cresols production in the United States is shown in Figure 21. The last
year of data available, 1985, shows a dramatic drop in production. Several
producers of synthetic cresols closed down about that time. Current producers
of natural cresylics state that their sales were not widely variable from 1984
71
1 30-
1 20 -
110-
90
00
IS67 197C 1 9-i3 1976 1 ~79 1 M298

YEAR
FIGURE 21
CRESOLS PRODUCTION, TOTAL
so. 70
$0.60
so.50o
50.40-
50.30
sc111 I I I II T I
1967 19 70 "97 19 7612

YEAR
FIGURE 22
CRESYLIC ACID PRICIFS
TANWS, PREIGHT ALLOWED
72
to the present, so the drop in Figure 21 appears to be related only to syn-
thetic cresols.
Primcs
Prices for cresylic acid are shown in Figure 22. Prices for the individual
cresol isomers are shown in Figures 23 through 26. Prices have been on a
stable or rising trend in all cases except for p-cresol. The reason for aber-
rant p-cresol prices is not known. One of the major known uses for p-cresol
is to make the anti-oxidant BHT.
The prices shown in Figures 22 through 26 must be interpreted with caution.

Although useful for showing long-term trends, prices from the Chemical Market-
ing Reporter are known to be unreliable for cresols. The long, flat trend in
Figure 24, for exanple, is sinply due to the fact that no new price postings
have been received since 1981, and is not the result of a truly stable narket.
Actual 1987 market prices for mixed cresylic acids are approximately $0.40 per
pound.
Consumers
Cresylic acid mixtures containing a high fraction of m-cresol can be used for
synthesis of phenol-foraldehyde resins and high-quality engineering plastics,
such as polycarbonates.
Because of the difficulty of controlling the product resin properties with a

mixed cresylic feedstock, the market for such material is smll relative to
the phenol market. The conpound cresol, which is the most useful conponent of
cresylic acid, has ortho, meta, and para isorers which have different comer-
cial values. Meta-cresol Is the preferred isomer for preparing phenol-
formaldehyde mlding conpounds and adhesives from cresylic acid feedstock.
The o-cresol and p-cresol react nmre slowly to yield lower quality resins than
does m-cresol. Consequently, cresylic acid with the highest m-cresol content
is the mst marketable. (Ccpare Figures 23, 24 and 25). Cresylic acid com-
* petes with phenol as a feedqtock for phenol-formaldehyde resins. Because syn-
thetic phenol is pure, the control of resin product properties is easier. The
decline in coal tar production and the difficulty of product quality control
with cresylic acids have discouraged their use in the petrochemical industry.
Cresols were at one time used extensively to make the phosphate esters
tricresyl phosphate and cresyl diphenyl phosphate, but these have been re-
placed by phenolic phosphates. Several different cresols and cresylic aci,
mixtures are used as wire enamel solvents, ore flotation frothers, disinfec
tauts, fiber treatments, tanning agents and degreasing agents. High-purity m-
cresol is used to make thyrol (cough medicines) and pyrethrVid insecticides.
In the electrolytic zinc process, cresylic acid may be used as a frother to

trap gases emitted from the anodes and cathodes. Cresols were once inportant
as paint removers, but have been eliminated because of toxicity. Cresylic
acid i used as an aromatic solvent in the Duo-Sol process for refining
73
S0.60 -
so0.60
50.50 -
10.30 -
sc' 20 -
sc0 10 F I I I I I I I
197 1970 19 7 3 1976 1(979 1982 19E
YEAR
FIGURE 23
o-CRESOL PRICES
98%PtIM DMB, T. TANKS, FOB
S$1.50-
I1S,30
$1.10
19 79 19s(1 1 S'81 1982 1983 19E;4 1 35 1 S863 -387

VIMR
FIGURE 24
ip-CRESOL PRICES
DMB, TI., TANKS, FOB
74
dpS
$1.70
$1.60
S1 .50
11.10
|0.90,
1979 1980 1981 1 e! 1933 19'84 1 995 1 I6 "38"
YMA
FIGURE 25
m-CRESOL PRICES
TAKS, DMB, TL, FOB
$0.90-
$0.80
]
$0.70
$0.60
$0.50-
S0.4-0
$0.30
$0.20
|0.1I0 ,
1967 1970 1973 1976 1937 1982 1985
YEA
FIGURE 26
mp-CRESOL PRICES
mULK, FOB
- 75
lubricants. Ortho-cresol is used in insecticides. An end-use breakdown for
consumption of total cresols and cresylic acid in 1981 is given as follows:
Magnet wire 17
Antioxidants 15
Resins 13
Exports 11
Phosphate esters 8
Cleaning compounds 6
Ore flotation 6
Miscellaneous 25
10096
In 1986, a large demand for low-grade cresylics opened up in China. However,
the prices offered are very low in the range $0.15 to $0.18 per pound.
Producers
Natural cresols and cresylic acid can be produced as by-products of both

petroleum and coal tar refining. Petroletn-based cresols are the result of
catalytic cracking processes and are obtained from the caustic liquor used to
wash the petroleum distillates. Although cresylic acids were formerly ob-
tained from several coal tar sources in the United States, there is currently
no production from this source. The only producers of natural cresylic acids
are PMC (formerly Productol) in Santa Fe Springs, California, Northwest
Petrochemical in Anacortes, Washington and Merichem in Houston, Texas. Al-
though Merichem uses som imported coal-tar-derived feedstock, none of the
three currently use feedstock from domestic coal-based materials. All three
state that they would use such naterials if available, and all expressed an
interest in the Great Plains rterial.
A major factor on the world market scene is cresylic acid from the SASOL coal
conversion plants in South Africa. SASOL has been selling crude cresylics to
others for refining, but it is believed that international trade sanctions
against South Africa have reduced their ability to sell on the world market.
It is thought that SASOL intends to install a hydrocracker to convert the
cresylics to fuel products.
Synthetic cresols are produced from petroleum. One route is the catalyzed
methylation of phenol. High-purity p-cresol can be made by toluene sulfona-
tion. A mp-cresol mixture can be manufactured via cymene, obtained by the
alkylation of toluene. Although there were five United States producers of
synthetic cresol isomers until recently, the only two companies involved in
1987 are General Electric (o-cresol) in Selkirk, New York, and PMC (formerly
Sherwin-Williams) in Chicago (p-cresol).
Market Structure
Great Plains would have the option of either selling its crude cresylics to a
cresylic acid refiner, or refining in a toll processing facility, or install-
ing its own refining facilities and marketing to the manufacturers of resins,
etc. In the former case, there are only three possibilities, at least in the
United States. Of the three, only Merichem has a large enough capacity to
process the entire Great Plains crude streams.
76
Overseas markets probably offer the best opportunity at present because major
European users are attempting, for political reasons, to find sources other
than South Africa's Sasol.
Present market prices for refined cresylic acid are about $0.40 per pound. It
is estimated that refining costs are on the order of $0.25 and that sales
costs are on the order of $0.08 to $0.10 per pound. This leaves a net price
for crude cresylics of $0.05 to $0.07 per pound, or about $17 per barrel.
This corresponds to one refiner's estimate that the current price for crude
cresylics would be roughly equivalent to fuel oil value. This study did not
consider the option of toll processing through other facilities which may be
available. It is possible that lower costs might be obtainable through such
faci lities.
The very small nunber of producers of both synthetic and natural cresols makes
for a closed market situation. It nmy be very difficult for a new producer to
obtain a major share of a business which is held by only two or three
producers. The total potential cresylic acid production from Great Plains is
80 million pounds per year, which is of the same order of magnitude as current
total domestic demand. In other words, it would be necessary to drive all
other natural cresylic acid out of the market entirely if the Great Plains
output were all consumed domestically.
Market Outlook
The outlook for natural cresylic acids acknowledges two broad, long-term nega-
tive factors. One is the general loss of market share by natural products to
higher-purity synthetic materials. The other is the toxic and carcinogenic
nature of coal tar products, which prompts users to find more benign sub-
stitutes. Currently, low petroleum prices act to accentuate the conversion to
petroleuzn-derived substitutes.
None of the uses listed for natural cresylics has been identified as a high-
growth area. Although individual cresol isomers may see increasing demand,
these can be derived in high purity from petroleum components. Cresylics are
often low-performance feedstocks for resins and other uses. In ore flotation,
for instance, cresylic acid frothers are usually much less efficient than
glycols. Also, flotation reagents must contain essentially zero phenol be-
cause of water pollution problems. Because petroleum-derived synthetics can
substitute for all natural cresylic applications, future demand and price
levels will both be constrained by petroleum prices. The outlook therefore is
for no price increases relative to petroleum, and for static or decreasing
demand levels.
A compensating factor in the short term is the price of phenol derived from
benzene. Benzene prices are rising substantially due to demand for polys-
tyrene and for octane enhancement in the gasoline pool. Because cresols com-
pete with phenol in som applications, this will offer an opportunity for
cresol price increases until benzene prices return to equilibrium values.
77
irwiigwwpwwuwwwwwiniumirrigvv
1WVVV W"VWVW'WWVMWNWVVWVVWVWWVWW~uwuvwvwMxwRxwExE,wFN-m1., .P -.
CAiMM BLACK FEED37OCK
Carbon black can be made by the partial conbustion of any carbonaceous liquid
or gaseous fuel with insufficient air. There are basically three modern
methods of making carbon black, known as the channel process, the furnace
process, and the thermal process. Almost all current United States production
is achieved via the furnace process using oil feedstocks. The process is
operated in an insulated furnace chamber. An auxiliary fuel is first burned
to completion to supply the necessary process heat, raising the tenperature of
the gas stream to about 2,500 *F. Oil is then mixed with the hot gas and
pyrolyzed instantly. The carbon black laden gas stream is cooled,and the
carbon black removed by cyclones, electrostatic precipitators or bag houses.
In the United States, carbon blacks are produced almst exclusively from
petroleum derived feedstock, although coal-derived liquids have been used, and
are mre widely used elsewhere in the world. Oil feedstocks must be highly
aromatic and have low ash and asphaltene contents. The United States Bureau
of Mines Correlation Index (BNIE) is used as an index of feedstock aromaticity
by most carbon black manufacturers. The BMCI is given by the formula:
BMCI = 100 (876/A + 670/B - 4.568)
where:
A = 50% boiling point plus 460 °F

B = 131.5 plus degrees API
Typical specifications for a carbon black feedstock would be a mininium BrdI of

130. Low BNCI oils give much lower yields of carbon black. Other specifica-
tions are listed in Table 39. Because the yield of carbon black is relatively
low (2.5 to 5 pounds per gallon) in all cases, low feedstock prices are highly
des i red.
True carbon black is formed by vapor phase cracking and/or condensation of

hydrocarbons. Both the formation of the solid carbon nuclei, and their sub-
sequent growth into high-surface-area particles must occur directly from the
vapor phase. If, instead, cracking occurs in the liquid phase, the result
will be the formation of coke or grit particles in the carbon black. These
will have the same deleterious effect as ash. Thus a primary requirement for
a carbon black feedstock is that it be conpletely vaporizable in a very short
time. Because asphaltenes cannot be vaporized in this way, they are un-
desirable.
The mist inportant requirement for carbon black feedstock is that it be highly
aromatic. Large conpact mlecules with many double bonds and short side
chains have a greater tendency to polymerize and eventually form carbon than
long straight chain molecules which tend to crack into smaller fragments.
Compounds having two or more fused aromatic rings have very high BMNI values
and provide the best feedstock.
The aromaticity of the feedstock influences the properties as well as the

yield of the carbon black product. While a wide range of black particle sizes
78
U'*
,pe -. . '. " - . " . " . " , " - - " , " 'e .,. -" . ,. ""- . . % - ". '."." -
TABLE 39
TIVICAL CAR2 BLACK OIL

"1W!FICLTIGN8
BMCI 130 Minimum

API Gravity 2 Maximum
Viscosity, SSV @ 210 °F 125 Maximum
Sulfur 2.2% Maxim 1
Asphaltenes 5% Maximum
Ash 0.05% Maximum
BS8W 0.2% Maximum
Sodium 10 ppm Maximum
Potassium 1 ppm Maximum
Flash Point 160 OF Minimum
Aniline Number 100 Maxinum
Initial Boiling Point 500 OF
50% Distillation Point 900 OF
can be obtained from almost any oil feedstock by adjusting the furnace condi-
tions, the structure of the carbon black is determined largely by the
aronmticity of the feedstock. "Structure" is the linking together of carbon
black particles into chains. High structure indicates long chains, which are
required for good reinforcement and abrasion resistance in synthetic rubber.
The concentration of aromatics in the fuel determines the rate at which carbon
black nuclei are formed. High rates result in small particle size and high
structure.
Currently, the mst suitable carbon black feedstocks are the high boiling bot-
tom from fluid catalytic cracking processes. Carbon contents of over 90 per-
cent are typical.
Product ion
Current carbon black production figures are not available, but United States
production was essentially static at approximately three billion pounds per
year from 1970 to 1980. It is not believed to have changed significantly
since then. At a yield of 3.75 pounds per gallon, this would indicate a
carbon black oil demand on the order of 50,000 barrels per day. This number
was confirned in informal communications with the industry.
At moast, 50 percent of the Great Plains tar oil output, (the portion boiling
above 500 OF) or 2,500 barrels per day could conceivably qualify as carbon
black feedstock. This would represent five percent of the total United States
demand.
79
Price
Prices for carbon black feedstock are given in Figure 27. Prices are usually
based on Gulf Coast plus transportation. The base price is the price for
3 percent sulfur #6 residual fuel oil. Every 10 points increase in the BMI
value is worth $0.50 per barrel, and every percentage decrease in sulfur con-
tent below three percent is worth one percent increase in price. However, the
mxinum premium allowed over 3 percent sulfur residual fuel prices is almost
never nre than one dollar per barrel.
Residual fuels of the type used for carbon black feedstock are often difficult
to sell in the Great Plains mrketing region because of a low demand for
residual fuels in general and the lack of alternate uses for the rmterial.
Thus FCC recycle oil, a good carbon black feedstock, has recently been selling
for only $8 per barrel at the Amco refinery in Mandan, North Dakota, and only
$6 per barrel at the Frontier refinery in Cheyenne, Wyoming.
Consmers
Major carbon black plants are listed in Table 40 and Figure 27A. A con-
siderable consolidation of the industry occurred in the early 1980's, with al-
mst a billion pounds of production capacity closed down.
Producers
The major source of carbon black oil is fluid catalytic cracker bottom from
oil refineries. Thus all nmjor oil refineries are potential conpetitors for
this nmrket.
,t
80
.*. N N ...
TABLE 40
MNUC WNITED BTATT CARBON BLAKI PROIX2
Carbon Black
Company Location Capacity, lMIb/yr
Ashland Aransas Pass, TX 160

Ashland Belpre, OH 120
Ashland New Iberia, LA 260
Cabot Franklin, LA 230
Cabot Pampa, TX 60
Cabot Villa Platte, LA 305
Cabot Waverly, WV 180
Columbian El Dorado, AR 90
Columbian Mojave, CA 55
Columbian Marshall, WV 155
Columbian North Bend, LA 180
Columbian Hickock, KS 60
Continental Carbon Phenix City, AL 55
Continental Carbon Ponca City, OK 132
Continental Carbon Sunray, TX 98
J. M. Huber Borger, TX 120
J. M. Huber Baytown, TX 175
J. M. Huber Orange, TX 135
Sid Richardson Borger, TX 275
Sid Richardson Addis, LA 123
Sid Richardson Big Spring, TX 108
Total 3,076
811
~ V *~*~i1
$26
$25- -
$24-
-$22
1-
19-79 1980 1E81 I E,. 19:3 3 198.4 19 BE 1 S86 98

YWA
FIGURE 27
CARBON BLACK OIL PRICES

BASED ON OILF WiERIMES
LOCATIONS OF CARBON BLACK PRODUCERS

82
^W WW WN WCJ WrwasImWI5.UN"
Market Structure
Most carbon black feedstock is purchased on the spot market and therefore fol-
lows spot oil prices very closely. Some is purchased under contract. in
either case, a new material, regardless of meeting specifications, would prob-
ably have to be subjected to a test burn before purchase. This might involve
a truck load of material.
Market Out look
Demnd for carbon black oil has declined as a result of inprovements in the
conversion efficiency of the furnace process and as a result of declining
demand for tire production.
There are many different comercial grades and varieties of carbon black. By
far the greatest use, however, is as a reinforcing agent in rubber tires
(90 percent). The finely divided state of carbon black greatly reduces oxida-
tion of the rubber and increases the tensile strength. It is a little-known
fact that carbon black ay account for over half of the weight of a tire.
Recent trends to smaller cars and smaller tires, longer wearing radials and
emrgency-only spares have all acted to retard the demand for autcmotive rub-
ber and thus for carbon black. These trends appear to have bottomed, and
carbon black may be ready to resume a modest growth period. Carbon black
feedstock prices, however, will remain strictly related to petroleum prices.
83
Z-, u
Almost all of the benzene and BTX (benzene, toluene, xylene) aromatics in the
United States are now derived from petroleum. Small amounts are derived from
coke oven light oil) Chemical grade benzene must meet a minimum freezing
point specification of 5.35 'C, corresponding to a product purity of 99.7 per-
cent. Other typical specifications are listed in Table 41. BTX mixtures are
not sold to general specifications.
Product ion
Benzene production data are given in Figure 28. These data represent only the
amount of benzene isolated as such, not the total amount produced by catalytic
reformers and left in the gasoline stream at refineries.
There are four major sources of BTX or benzene liquids: .
1. Catalytic reforming of petroleum naphtha

2. Pyrolysis gasoline from ethylene plants
3. Hydrodealkylation of toluene
4. Recovery from coke oven light oil
Benzene recovery from coke ovens declined drastically from 1979 to 1982
(Figure 29).
The volume of BTX isolated and used for chemicals production is relatively
small conpared to usage in gasoline. Because of high octane values, BTX liq-
uids are valuable gasoline conponents. It is estimated that 50 percent of to-
tal benzene and 90 percent of total toluene and xylenes end up in the gasoline
pool.
Most of the catalytic reforate and some pyrolysis gasoline is used directly
in gasoline without extracting the aromatics for chemical use. If only one
arormtic product is desired and low purity is acceptable, it can be removed
from the stream by extractive distillation with a solvent such as phenol. If
more than one product is desired, and high purity is inportant, then liquid-
liquid extraction is used. The most widely used extraction process is the
Sulfolane process.
Pries-
Because of the large usage in the gasoline market, the amunt of BTX available
for chemicals manufacture fluctuates from time to time, with a corresponding
effect on merchant prices. A benzene price history is given in Figure 30.
The drastic reduction in the amount of lead allowed in gasoline, starting in

1986, is having a predictable effect on benzene prices. As refiners are
forced to keep more BTX in the gasoline stream in order to meet octane
requirements without tetraethyl lead, and as demand has increased for
petrochemicals such as styrene, benzene availability is dropping and prices
have clinbed sharply in early 1987, to levels of around $1.50 per gallon.
84
*1
0Q
CO
00 0
Cd 000 M
0 0
CO00 4-
0C'~ 0 C4 ~
4Ja) 0
00
o. 0
co c GoC0
0 -'4)
0
E 4J 10
-1I x 00.
~0c(q to
hi~. .t.~
00 W
o 04J4 0 '410 *
un-4J 00 4 41j 0 0 '
I. bO 0:
)0 4% -
-4-
Lf~ C.) e
0q 0
>0 a *I ;e 9
~00 1:0~i~ 0
0 L-- 0 inZ L z. I
-4C )
-4 c
L4 1
0 0
0a 0z
85
7 m0 1.7~ __ _ _ _ _ _ _ _ _ _ _ _ _
1 .6
1.5-
1.4.
1979 1990 1961 1932 198.3 E3

YEAR
FIGURE 28
BENZLNE PRODUCTION, ALL GRADES
45.
eo
-
201
11EEO
1960 1
YEAR
FIGURE 29
BENZENE PRODUCTION
BY COKE OVE OPERATOW
86
pUNCLASSIFIE
WW
46
9I-A6 PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS
VOLUME I MARKET RSSESSN. (U) NORTH DAKOTA UNIV GRAN
FORKS ENERGY RESEARCH CENTER J E SINOR AUG 6?
AFIML-TR-87-2042-VOL-1 HIPR-FY1435-66-N@65? F/G 21/4
2/3
mohmmmmhhhhl
smmohhhmsommlm
1.09
MICROCOPY~ RESOLUTION TEST CHART

NA1WONALBURIAL) OF 'ANOARI)', 1961 A
wq W jW .0 V -we~
5 - wr
r .
'V%
U~.00
1 .00
11.80
S1 .70
1 .601
S1.50
I 1.40
11.10-
I1
10.70-
so.c.o
10.4-0
1C.20 W
1 O ? I E,7' 13
I ' 3 16.',V
3 1S'7. 19 1932 19135
YEAR
FROURE SO
BeNZN PRICS
The largest volum uses for BTX aromatics (other than gasoline) are for the
manufacture of plastics and fibers. Major end products of benzene include
styrene mnomer for polystyrene production, cumene for phenol production and
cyclohexane for nylon production. Major end products of toluene Include sol-
vents, and toluene di-isocyanate for production of polyurethane foams.
Xylenes are used in paints, solvents, plastics and polyester fibers. Most o-
xylene is first converted to phthalic acid and most m-xylene is first con-
verted to isophthalic acid. Any excess materials simply find their way back
to the gasoline arket.
One option for Great Plains would be to sell the naphtha stream to a conipany
having excess hydrodealkylation capacity. A nunber of such facilities exist,
and same are being re-comnissioned in light of the 1987 benzene market. Those
being mentioned for re-comiissioning include Standard Oil in Lima, Ohio, Dow
Chemical in Freeport, Texas, Sun Oil in Toledo, Ohio, Coastal Corporation in
Corpus Christi, Texas, and Crown Central in Houston, Texas. These units,
however, probably could not accept the mixed Great Plains naphtha unless
blended in small quantities. Toluene hydrodealkylation units are based on
either catalytic (UOP's Hydeal and Houdry's Detol processes) or thermal
processes (HRI's HDA and Gulf's THD processes) which would not accept an im-
pure stream like the Great Plains naphtha unless blended off to small con-
centrations. Suitable processes for such a stream are Houdry's Pyrotol or
Litol processes. Ethylene plant by-products are generally processed with
Pyrotol and coal tar products with Litol. Litol units are installed at
Polysar/Petrosar in Sarnia, Ontario and at Bethlehem Steel in Baltimore.
87
The other option would be to build a hydrodealkylation unit at the Beulah
facility and sell benzene in the merchant rmrket. Possible consumers would
include any of the nnny rmnufacturers of the aronatic chemicals listed ear-
lier.
Proameers
The primary source of BTX liquids is catalytic reforming in petroleum

refineries. The objective of reforming is to convert paraffinic and naph-
thenic conpounds to aromatics. The feed is usually desulfurized before being
charged to the reforming unit. The usual feed is either a straight run or a
cracked naphtha fraction boiling over the range 66 to 204 0 C if gasoline is
the desired product, or 66 to 132 0C if BTX chemicals are the desired product.
The most comnon process for reforming is the Platforming process, licensed by
UOP. All major petroleum refineries have catalytic reforming units. Total
United States capacity is on the order of 250,000 barrels per day of BTX liq-
uids, about one-fourth of which would be benzene.
A second significant source results from the steam cracking of hydrocarbons to

make ethylene and propylene. An aromatics-rich by-product stream called
pyrolysis gasoline, dripolene, or drip oil, is produced. This rmaterial is a
mixture of hydrocarbons, olefins, diolefins, styrenes and aromatics. When
heavy feedstocks are used, the pyrolysis gasoline nay contain high concentra-
tions of BTX. After hydrogenation to saturate the olefins, the BTX's nmay be
recovered by extraction.
Potential benzene recovery from pyrolysis gasoline Is estimated to be on the

order of 30,000 barrels per day, all from ethylene plants located on the Gulf
Coast.
If additional benzene is needed above that present in BTX streams, it is ob-

tained by the dealkylation of toluene. Processes which acconplish this in-
clude hydrodealkylation and disproportionation (in which two nolecules of
toluene are converted to one of benzene and one of xylene).
Hydrodealkylation processes provide the swing capacity needed to nmtch benzene

supplies to fluctuating demand. These units are started up or shut down
depending on benzene demand and the benzene to toluene price ratio. Thus
capacity is usually far in excess of actual production at any given point in
tine. Because of the difficulty of determining the condition of shut down
plants, it is difficult to derive an effective production capacity for
hydrodealkylation, but it is believed to be on the order of 50,000 barrels per
day, heavily concentrated in the Gulf Coast area.
Although the production of BTX liquids from coke oven by-products was once an
irportant factor it is no longer significant (Figure 29). In 1981 there were
six producers of BTX from coke oven light oil but the 1985 survey by the
United States International Trade Comnission identified only two renmining,
Bethlehem Steel and USS Chemicals. Only USS Chemicals (now Aristech) is
believed to be in the business at this time. Conpetition from this source
should have no effect on the narketability of benzene from Great Plains.
,-,)
88
I. ~ ~ ~ ~ ~ ~ ~ - R7.&
VUA Efz .
JUA W. I^- f. m AK K'm.
J %vm.
Market Structure
Most hydrodealkylation plants are based on captive toluene sources. Although

some of the benzene product is captively consumed, most of it is sold to
others. Thus the benzene market may be characterized as open. Both spot
sales and contract sales may be used. Hydrodealkylation plants are sometimes
operated on a toll basis (processing feed for others, for a fee).
The maxinum Great Plains output of 32 million pounds per year from the lec-
tisol naphtha stream and 24 million pounds per year from the tar oil stream
would anount to less than one percent of United States benzene production
(Figure 28) and should have no effect on market prices.
Market Out look
The need for octane enhancement to replace lead in gasoline should have a
positive effect on the benzene market for several years. The recent sharp
rise in benzene prices, however, is most likely not sustainable. For the long
term, benzene prices will be set at the margin by the costs of toluene
hydrodealkylation. If benzene prices stay high, more dealkylation units will
come onstream until the price drops far enough that some of them will once
again shut down. Based on historical relationships, that price is estimated
to be $1.15 per gallon at early 1987 oil prices.
89
7
-i e.,
Definitions and Specificatioms
Phenol and other constituents of crude phenols are basic building-block

materials for the United States petrochemical Industry. A wide variety of
plastics, chemicals, and drugs are derived from phenol and closely related
conpounds. Some ambiguity surrounds the definition of "phenol" and "phenols"
as the term are applied in the coal tar and petrochemical industries. The
chemical "phenol" is monohydroxybenzene, i.e., a benzene ring with one
hydroxyl (-OH) group attached. It is sonetimes known as carbolic acid. The
term "phenols" is also used industrially to describe a large family of com-
pounds with one or more hydroxyl groups attached to aromtic rings. The crude
phenol available from coke ovens and coal gasification plants contains tar
acids, tar bases (pyridines, picolines), neutral oil, organic acids, and
pitch, recovered by extraction of gas liquor using the Lurgi Phenosolvan
process. The crude phenol contains monohydric "phenols," which include
phenol, cresols, xylenols, and ethylphenols. The crude phenol may also con-
tain polyhydric "phenols," including catechols and resorcinol.
The term "cresols" is used to designate the isolated monomethyl phenols,

either as pure ortho, meta, or para isomers or as mixtures of isomers. This
report uses the term "phenol" to denote specifically the chemical
monohydroxybenzene and the term "crude phenol" or "total phenols" to denote
the complex mixture of aromatics coming from the Phenosolvan process.
Phenol may be synthesized from benzene, toluene, or cumene. Over 90 percent

of synthetic phenol is synthesized from cumiene, for which phenol synthesis is
the only use. Originally, coal was the predominant source of phenol and other
precursor chemicals for resins and plastics. As long ago as 1966, however,
96 percent of the phenol produced in the United States was "synthetic."
The USP specifications for phenol are as follows:
- Phenol content of 98 percent or more

- Clear solubility of one part phenol in 15 parts water
- Fusion tenperature 40.9 degrees C
- Boiling point 181.0 degrees C
- Specific gravity at 45/20 C of 1.0563
- Content of nonvolatiles not more than 0.05 percent
To control the properties and purities of subsequent synthesis products, most

synthetic phenol is used and sold at a purity of over 99.5 percent. Typical
specifications are given in Table 42. Pharmaceutical-quality phenol must pass
the most stringent quality specifications of all uses. Commercial grades of
natural phenol containing 90-92 percent or 80-82 percent phenol have been
marketed in small quantities.
Product ion
In the bulk of its ultimate markets, phenol suffers from cyclical dennd, as
illustrated in Table 43 and Figure 31.
90
TABLE 42
P1UL 81?UlrIVIlT
Property Specification Test Method
Appearance Crystalline solid Visual

Color Of Melt, APHA 10 nmx. ASTM D-1686
Color Of Sodium
Hydroxide Solution 1.5 max. ASTM D-1500
Freeze Point, °C 40.6 min. ASTM D-1493
Turbidity Of Melt,
APHA I max. HC 390A-80
Iron, ppm 0.5 max. Wr-2*
Water, Wt% 0.07 nux. ASTM D-1631
Non-Volatile Residue,
Wt% 0.05 Mnx. ASTM D-1353
Inpurities By OC
Alpha Methyl Styrene,
ppm 100 Ux.
2-Methyl Benzyl Furan,
ppm 25 nmx.
Di-Methyl Benzyl
Alcohol, ppm 100 mvux.
Acetophenone, ppm 5 mx.
Source: Monsanto
91
TABLE 43
!UIiM wI v
PS U8 (N Ovin10.OICS
Year 1,000 tons

1979 1700
1980 1200
1981 1200
1982 1000
1983 1300
1984 1400
1985 1400
3,1
3.0 -
2.6l
27-
2.6
2.5
2.4-
2I3-
21.
2.0-
I A.
1 ,7
1 "-
1 e=
1'G,'7 197C, 1973 1976 ; 1, ,,

um
FIGURE 31
PHNOL PRODUCTON
92
A history of list prices for phenol is given in Figure 32. These prices are
on a freight equalized basis, that is, including freight, assuming average
transportation costs. Prices paid by large-volume consumers are somewhat
lower. Contacts with industry in February, 1987 suggested a current market
price in the range of $0.21 to $0.25 per pound. However, by late April, 1987,
prices had jumped to $0.30 per pound.
Comneirs
Phenolic resins consume about 40 percent of the phenol produced in the United
States. Industrial chemicals, drugs and caprolactam (a nylon precursor) ac-
count for the remainder of phenol consumption.
The ultimate markets for phenol are heavily dominated by durable products,
(Figure 33). Resins and adhesives derived from phenol are used in auto-
mobiles and housing, with adhesives accounting for about 50 percent of
phenolic resin usage.
Large regional markets for phenol are present on the Gulf Coast, East Coast,
West Coast and Micvest. If we assume that the naximum phenol production from
Great Plains would be about 35 million pounds per year, this would represent
less than 10 percent of Gulf Coast nameplate capacity. Thus there should be
no problem with finding a market provided specifications can be met.
The heterogeneity of the Great Plains crude phenol stream makes the reaction
of polycondensation with formaldehyde of little value as a means of utilizing
total phenols. Polycondensation of total phenols with formaldehyde yields
products that contain low-molecular weight conpounds and is therefore un-
suitable for the production of thermosetting resins, varnishes, etc.
Monatomic phenols which consist of phenol hornlogs and also contain naphthol
homologs and heterocyclic phenols are suitable for the production of varnishes
and surface-active agents.
The top five producers of phenolic resins are Borden Chemical, Georgia
Pacific, Monsanto, Owens-Corning Fiberglass, and Reichhold Chemicals.
Producers
Almost all current phenol production is synthetic phenol, and most of that is
derived from cumene. Natural phenol production from coke ovens accounts for
an insignificant percentage of current production. Some natural phenol is
derived from petroleum.
In the curene process, cumene [C6 -ClI(CH 3 )2 ] is prepared by alkylating benzene

with propylene. Purified cumene is oxidized with air to form cumene
hydroperoxide. The hydroperoxide in the presence of sulfuric acid then splits
into acetone and phenol. Although there are several other routes from benzene
to phenol the curene route has been corrpletely dominant. Most non-cuniene-
based phenol plants have gradually closed as new cunene plants canm onstream.
93
50.30-
$0.20
SO5.2I-c
1S67 197 7 -5 S A9 1 ~
YEAR
FIGURE 32
PHENOL PRICES
S"N, TAWKS, PRT. EauALD
PIUNOL-UOUM~~iD8I10M .mwe A£ES

&lm
LASNATE
PLYWOOD
PWROTECTIVEa COATING
PLOONING A PAVMN
1POYCARSONATES APPUMICE
UHTMO a own
PMWNL- TRANSOMTTO
CAMWOACTM PnoNYO- - . TM mao
ALYLFIEOs AKNT)AJIT
SUNACTANIS
NALOPOLS U"mmcrnus
WOOD a I.EATNE
PUSRATVE
IPWAuACuIcALS
(MINY ASPIRIN
FIGURE 33
ULTIMATE MARKETS FOR PHENOL
94
Very little phenol Is Inported. The largest United States producers of syn-
thetic phenol, their plant locations and capacities, are listed in Table 44.
Samr 900 million pounds of capacity was closed between 1981 and 1987, but
Aristecb has recently announced plans for expansion.
Most natural phenol originates from petroleum caustic wash streams. Phenols
are generated as coproducts during catalytic cracking of petroleum. The ef-
fluent stream of the cracker is washed with sodium hydroxide, primarily to
remove mircaptans and other organic sulfur conpounds. The caustic wash opera-
tion simultaneously extracts phenols and cresols as the sodium salts. Natural
phenol producers process the caustic wash streams to regenerate and recover
phenol, cresols, xylenols, and mixtures.
Table 45 lists United States producers of natural phenol and their capacities.
In conparison with Table 44, the total production of natural phenol is an in-
significant market factor.
Market Structure
All of the major phenol producers listed in Table 44 use most of their output
for captive production of materials such as phenolic resins, bisphenol A,
Caprolactam, salicyclic acid, 2,4-D, pentachlorophenol, etc. Most of them
also produce excess material for merchant sale.
The merchant phenol market has been subject to violent price swings, as evi-
dent in Figure 32. In order to protect both sidez against these swings, long-
term contracts between buyers and sellers may be used. Because of the captive
production situation only a small percentage of the total phenol market is ac-
cessible to independent producers. In spite of the small fraction of the
phenol market which is non-captive, the amount of phenol represented by this
fraction is large with respect to the potential production from Great Plains.
Market Out look
As seen in Figure 31, phenol has shown long-term growth interrupted by large
cyclic ups and downs. Phenol's primary end uses are in the manufacture of
durable goods and the housing and construction industry, which are strongly
sensitive to economic cycles. Because the traditional phenol markets are
technologically mature, future demand growth is expected to be nominal, prob-
ably little more than overall GNP growth. Phenolic resins are the oldest
class of plastic materials in comiurcial production, having been first intro-
duced in 1910.
The most valuable property of phenolic resins is their outstanding performance

in adhesive applications. The largest single use is in making exterior grade
plywood. A major market for various pressed-wood-fragrmnt products had been
expected to develop in finished siding for housing. Instead, many problems
and lawsuits occurred as panels have warped in high humidity areas. One
phenol end-use showing strong growth has been the manufacture of bisphenol A,
an intermediate in the production of epoxies and polycarbonates. A growth
rate of 7-8 percent is expected for bisphenol A.
Future phenol prices will be related to both the cost of feedstock and the
cost of manufacturing. A new cumene production process developed by Monsanto
95
-
&Vic( W
TABLE 44
iN&a P20UIR or Sxmurric FIUMM
Company Location Capacity, Mlllb/yr
Allied Chemical Frankford, PA 600

Aristech Haverhill, OH 500
Shell Chemical Deer Park, TX 500
Dow Chemical Oyster Creek, TX 500
General Electric Mount Vernon, IN 400
Georgia Pacific Plaquemine, LA 300
Clark Blue Allen, IL ?
Texaco/Getty Eldorado, KS ?
Kalama Chemical Kalam, WA ?
TABLE 45
UNITE STATES PRUH OF NTlRAL PHMIOL
Capacity,
Company Location MMIb/yr Source
Merichem Houston, TX 20 Petroleum
Diamond Shamrock Tuscaloosa, AL 10 Petroleum
PMG Santa Fe Springs, CA 8 Petroleum
Koppers Follansbee, WV 8 Petroleum,

coal tar
Northwest Petrochemical Anacortes, WA 2 Petroleum
P -S . - .. '- - - '- U - ' . - . - - ' - L -'' -- ~

- '- - *.*..--
- - - - -
96
and Lwmmus Crest in 1984 is said to provide a co-t saving of about a penny per
pound. The cost of benzene feedstock accounts for about half the total cost
of phenol (propylene, the other feedstock component, is essentially
100 percent recovered as LPG). Because of the phaseout of leaded gasoline in
1986, deand for benzene as an octane improver in gasoline is increasing at
the saue tine as the petrochemical denand for benzene is surging. These fac-
tors are causing significant price increases for benzene. This in turn, must
impact the cost of synthetic phenol. For the short term, sone see additional
phenol price increases of $0.10 per pound or rmore as likely. For the longer
term, after benzene prices have stabilized around a new supply/dermnd balance
point, a phenol price of about $0.28 per pound is expected.
97
P!OSIM OIL
Dfimltion ad Speclflctlm
The synthetic rubber industry uses a wide variety of processing oils. There
are twr broad classes of rubber processing oils: aronmtic and naphthenic.
The aromatic oils are used nmstly in tire production and the naphthenics are
used in lubrication.
Four general types of processing oils are defined by ASTM as shown in

Table 46. This table sinply defines general classes. Actual purchase
specifications are prepared by each rubber manufacturer, and ay be very
detailed. Table 47 is an exanple of a purchase specification for an aromatic
processing oil.
On any new material, compounding tests and factory trials would be required.
This procedure could take up to two years and cost $250,000 in the rubber
manufacturer's facility before it would be known if a material is acceptable.
Because each manufacturer establishes his own specifications some parts of
the process might have to be repeated with each client. There would have to
be a large financial incentive for the rubber corpanies to consider such a new
source.
Prodkict,ion
Rubber products shipments are highly dependent on the motor vehicle market.
Approximately 85 percent of its shipments go to passenger cars, trucks, and
buses. Other markets include aircraft, agricultural, industrial, rmtorcycle,
and bicycle uses. Demand for rubber processing oils fluctuates in direct
linkage with rubber production. Estimated current demand is about 68 million
gallons per year (4,400 barrels per day) of naphthenic type oils and
90 million gallons per year(5,800 barrels per day) of aromatic oils. Thus the
total Great Plains tar oil stream would represent a significant percentage of
the total market in this area.
Prices
Because they are sold to detailed specifications and are not fungible com-
modities, rubber processing oils have not fallen in price in direct proportion
to crude petroleum prices. Recent prices (dollars per gallon) are estimated
as follows:
Type 1985 1986 Early 1987
Arormatic 0.50-0.95 0.65 <0.50

Naphthenic 1.00-1.25 1.00-1.05 <0.90 %
Paraffinic 1.15-1.25
Cmairs
The major consumers are tire manufacturers, who are located mstly in the Mid-
west and East. The largest single consumer is probably Goodyear, with plants
in Alabanma, Kansas, North Carolina, Oklahona, Tennessee and Texas.
98
ilket Out look
The rubber industry faces stable or declining demand for new vehicles and
tires.
According to the International Institute of Synthetic Rubber Producers total

1986 consumption fell an estimated one percent from 1985. Flat demand is pre-
dicted for 1987 and the next five years. The Institute's forecast through
1991 shows a conpounded annual growth rate of only 0.3 percent -- considerably
below gross-national-product projections of 2.8 percent per year for the sam
period.
The North Amrican tire industry is expected to show little or no growth be-
cause of increased imports of tires and longer tire life. The industry has
shifted to higher-value, longer-lasting radial tires. More than 20 older tire
plants have closed. Employment in the rubber tire industry dropped form
114,000 in 1977 to 66,500 in 1985.
However, demand for specialty rubbers has grown as markets have developed for
new automotive applications. Performance requirements in automotive and other
applications have forced the substitution of specialty rubbers for general-
purpose rubbers and rubbers have been increasingly alloyed with plastics to
form new materials. Growth areas include roofing and oil additives, plus new
products created from alloys of rubber and plastics.
Because rubber processing oils are derived from petroleum, their prices should
eventually be a function of world oil prices. A near-term decline in prices
should be expected.
TABLE 46
SPBCIFICATICNS FOR RUBBE PROCESSING OILS
Maximum Percent Allowed

ASTM Asphal- Polar
Tye Class tenes Compounds Saturates
Highly Aromatic 101 0.3 25 20
Aromatic 102 0.5 12 35
Naphthenic 103 0.7 6 65
Paraffinic 104 0.1 1 65*
*Mininum percentage
99
TABLE 47
T!PICAL PUZABI1518?UiFICATION FOR

AhGEMC 15351N OIL
ASTM
Test
Limits Method
Appearance Dark colored oil--
Specific Gravity 0.985-1.008 D1298
Saybolt Viscosity 115-150 SUS D88
Volatile Matter 0.8 percent maxint D972
Neutralization No.
(Mg KOII/g 0i1) 1.5 maximium D664
Aniline Point, *C 29.4-51.7 Dell
Flash Point, OC 204.4 mininwin D92
Pour Point, OC +29.4 maximum D97
Refractive Index 1.065-1.074 D2159
Viscosity Gravity Constant 0.930-0.960 D2501
Copper and Manganese Nil
Asphaltenes 0.10 percent nuxinumi D893
Resins (Polar Aromiatics) 8.0-16.0 D2007
Aromiatics 64.0-80.0 D2007
Saturates 10.0-22.0 D2007
100
Detimtious, Specificatious
Tar and tar crudes are basically the miterials which come directly from coke
plants and are marketed in this form. There are few end-users for tar crudes
in their original form; therefore purchasers are nnly tar processors who
fractionate the material into end products. They will be interested in buying
raw tar only if they can make specification grade products from it.
Product ion
The production of crude coal tar In the United States began a long decline in
about 1965 (Figure 34). The data show a nmdest rebound In 1984. Unfor-
tunately the Energy Infornstion Admninistration, United States Department of
Energy, discontinued collecting data on coal tar after 1984. The United
States International Trade Comission began collecting data for the 1986 year,
but at the tinie of this writing, those data were not yet available.
Priem
There are no published prices for tar crudes, perhaps because of the
variability of the product.
Prices for grade RT-12 coal tar, a processed product, have been stable at
$1.05 to $1.25 per gallon over the last three or four years.
Consuinrs
Major purchasers of raw coal tar are distillers who separate the tar into
pitch plus creosote. At present, only the pitch fraction is in demand. The
creosote mrket is over-supplied. Purchasers of raw coal tar are Reilly Tar
and Chemical, Allied Chemical, and Koppers Conpany.
Producers
tunaelythe Admnisratin,
nery Inormtio nitd SttesDepatmet o
Locations EnerT,
clletingdat
of tar
of producers rad crudles are shmvn in Figure 35.194.
iscotined oncoaltarafte Te Uite
Basically, these
Statesa Intenatinal C %.alssion began colctn data.*'
for te 198 yea ,.'"
are metallurgical coke plants for the steel industry. One of the largest
single producers is Allied Chemical, which produces 80-90 million gallons per
year, or about 25 percent of the total rmrket.
Market Structure
Independent coal tar refiners provide part of the market for tar crude. Som
producers of tar crude, such as Allied, also process additional material ob-
tained fromn other producers.
Although a nunber of different chemicals were at one time extracted from coal
tar, that nuaer has steadily dNindled. Even when the industry was strong,
the "chemical" products accounted for only about 10 percent of the tar. The
remaining 85 percent (allowing for liquor and distillation losses) was com-
posed of the tar oils and pitch. Historically, the profitability of the
10 196 bo n.
36 The -A,
d se Un1o
800
700-
1sE65 1936 l97's 19 77 )E 1

YM
FIGURE 34
CRUDE COAL TAR PRODUCTION
PRODUCERS OF TAR AND TAR CRUDES
102
chemicals production was never high enough to pay for the total recovery
operations. Profitable operation depended on the sale of the bulk tar oils
and pitch. The oils could be blended to give creosotes and cut-back oils for
pitch.
Before 1960 there were a large numiber of coal tar distillers in the United
States. By 1980, they had been reduced to a handful, with Koppers, United
States Steel, and Allied Chemical accounting for more than 85 percent of total
production. Since 1980, the production of chemicals from coal tar has vir-
tually ceased. The United States Steel Industrial Chemicals Division, for ex-
ample, shut down its coal tar facilities completely from 1981-1983. When it
was reactivated again in 1984, the operation was greatly simplified. Although
benzene, xylene and toluene are still produced, everything in the light oil
fraction now is simply put into creosote, with the remainder going into
electrode pitch or else being burned for fuel.
The market for raw coal tar now depends mstly on the aluminum industry and
the demand for pitch. No barriers to entry in this market are known, if
material suitable for the market can be provided.
Market Outlook
The market outlook for crude coal tar features both declining supply and
declining demand. Supply is determined by the state of the domestic steel in-
dustry. All signs point to a continuing shrinkage of the industry, due to a
combination of a less steel-intensive econonm and conpetition from overseas.
Demnd is determined by the state of the aluminum industry, which faces strong
conmpetition from overseas. On balance, the steel industry's problems seem
greater than those of the aluminum industry. ,'his leads to a prediction that
coal tar supplies for pitch production will be somewhat short of demand and
prices should remain firm.
103
Sr v r' " " , " - ," - .-.. . .".-".. . .. . . . .. . .

PITCH
Definitions and Speificatimzs
Most organic substances, when pyrolized, yield dark-colored viscous liquids

termed tar and pitch. The differentiation between these terms is not precise.
If the liquid is of relatively low viscosity at anbient temperature, it is
designated as a tar. If it is very viscous, semisolid or solid, it is desig-
nated as a pitch. When crude coal tar is distilled to yield first a chemical
oil fraction, then a creosote fraction, the residue is called coal tar pitch.
Coal tars are extremely complex mixtures, consisting mostly of aromatic

hydrocarbons ranging from benzene and alkylbenzenes to conpounds containing as
many as 20 fused benzene rings. Heterocyclic conpounds containing oxygen,
sulfur and nitrogen are present, usually with only one heteroatom per ring
system. Low temperature tars, such as the Great Plains tars, differ strongly
from coke oven tars. The latter contain relatively small amounts of non-
aronmatic hydrocarbons and relatively small amounts of phenols. The phenols
are present mostly in the form of phenol, cresols, xylenols and naphthols dis-
tilling up to 300 OC. Low temperature tars, on the other hand, contain much
higher proportions of alkylphenols, and the content of unsubstituted ring
structures is relatively small. Coke oven tars, when distilled, yield
55-60 percent pitch, whereas low temperature tars (that is, the total tar, not
just the tar oil fraction) may yield no more than half as much.
Because coal tars and pitches contain several hundred individual conponents,
it is not possible to create specifications based on chemical conposition.
Rather, market specifications are based mostly on physical properties. The
most comn specifications for pitch used as electrode binder are the specific
gravity, the softening point, a minimum C/H ratio, maxima for ash content,
moisture content, volatile matter, quinoline insolubles, toluene insolubles,
and viscosity. Table 48 lists some typical values. Coal tar pitch for other
uses is mostly sold to individual users' specifications. Pitch intended for
roofing and waterproofing is covered by ASTM D450-71, and hot-applied tar-
based pipeline enamels are covered by American Waterworks Association
specification C203-73.
The largest current market for coal tar pitch is as a binder for the
electrodes used in aluminum smelting. However, the specifications for
electrode binder pitch, including high C/H ratio, high coking value and high
beta-resin content, effectively rule out unmodified pitches from low tempera-
ture tars. Experiments have been tried with air blowing of lignite tar to
make specification binder pitch, but these were unsuccessful.
Product ion
The production of crude coal tar in the United States began a long decline in
about 1965 (Figure 34). The data show a modest rebound in 1984. Unfor-
tunately the Energy Information Acinistration, United States Department of
Energy, discontinued collecting data on coal tar after 1984. The United
States International Trade Commission began collecting data for the 1986 year,
but at the time of this wrltirg, those data were not yet available.
104
TABLE 48
TIfICL POpRImTIRS OF PITCH SIM) USED IN MM

IN = AUIIIl INDUSTRY
Prebaked Continuous
* Anodes Anodes
Softening Point, 0 C
Cube-In-Air 105-115
Cube- In-Water 65-90
Specific Gravity, at 25/25 °C 1.28-1.31 1.25-1.29
Quinoline Insolubles, % 5-18 5-15
Toluene Insolubles, % 20-35 15-25
Ash, % (Max) 0.3 0.3
Moisture, % (Max) 0.1 0.1
Sulfur, % (Max) 0.5 0.5
Conradson Carbon (Min) 57 50
Production of tar pitch would be expected to be correlated with overall coal

tar production, as it is not practical to convert pitch into other products.
However, coal tar producers can choose between burning the pitch for fuel, or
selling it as a product. Figure 36 shows recent production history for coal
tar pitch. The reporting basis for these data is inconsistent, soe years
being reported as total pitch, some years as hard pitch only, some years as
sum of soft, nedium and hard pitch, and scoe years sinply as pitch with no
qualifying term.
Prim
Figures 37 and 38 show the pricing history for two categories of coal tar
pitch. The prices for industrial liquid, Figure 37, appear to be close to the
overall average price calculated from United States International Trade Com-
mission data. Industry contacts indicate that prices have been fairly stable
at these levels for 3 or 4 years. However, long flat periods followed by
sharp changes, such as seen in Figure 38, indicate that the data are based on
list prices rather than actual transactions. Such data may be badly out of
date at any point in time. A rising price trend (Figure 38) in the face of
declining production indicates that production (availability) is driving
prices rather than vice versa. Pitch production is declining for reasons
unrelated to prices, with a resulting upward pressure on prices.
105
3.0-
05
I1E79 19 80 I 9E~ 191832 198e.3 1 '4 8 5i8
FIGURE 36
PITCH OF TAR PRODUCTION
5S360
5340-
5320-
$300-
5260 V.
$240- INDUSMAL UQUID ~...
5200-
18c0 C
514.0
1 r579 198(1 1 NJO I I e8l! 1933 19 fE;4 . 9 3t 1 E i8; -
YEAR
FIGURE 37, 38
COAL TAR PITCH PRICES
106
Cosuamrs
The imjor outlet for coal tar pitch in the United States is the binder for
electrodes used in aluminum smelting. Older smelters use Soderberg furnaces
which incorporate continuously forming electrodes from a mixture of about
70 percent petroleum coke or pitch coke and 30 percent of a mdiun-hard coke
oven pitch. The paste is added to the top of the electrode well, and as it
slowly moves downward is first baked into a hard electrode and then consumed
in the electrolytic smelting process. Newer smelters use prebaked electrodes
requiring less pitch (18 percent) for binder.
Table 49 and Figure 39 give the locations of aluminum smelters. Consunption

of binder pitch is estimated to be on the order of 0.12 ton per ton of primary
aluminum. At current aluminum production rates of 3.4 million tons per year,
this translates to a pitch denand of over 400,000 tons per year. This ac-
counts for rmst of the pitch production reported to the International Trade
Conmi ss ion.
Other minor uses for coal tar pitch include binder for electrodes in electric
arc steel furnaces (in aluminum smelters the carbon electrodes participate
chemically in the reduction process and are therefore consumed at much higher
rates than in steel furnaces), binder for briquettes, adhesive for membrane
roofs, feedstock for coke, impregnant for pitch-fiber pipes, binder for
foundry cores, sealant for dry batteries, and raw material for clay pigeons.
Roofing pitch is the second largest use after electrode binder pitch.
Producers
Table 50 and Figure 40 give the locations of conpanies reporting production of

soft medium or hard coal tar pitches in 1985.
Market Structure
Coal tar pitch is produced and sold both by the primary coke oven operators
and by intermediate tar distillers who purchase crude coal tar and sell pitch
to the aluminum smelters and anode manufacturers. The intermediate tar dis-
tillers currently have a shortage of good pitch feedstock and would welcome
new sources.
Market Out look
The electrode pitch market is of course dependent upon the market for
aluminum. The United States is the world's largest consumer on both an ab-
solute and a per capita basis, but per capita consunption has been flat for
10 years (Table 51).
United States aluminum producers have not been faring well of late. After
maintaining almost flat primry aluminum smelting capacity of five million
tons per year from 1974 to 1984, primary aluminum capacity fell in the United
States to 4.4 million tons in 1985. Production of metal followed this trend
and was reduced to 3.5 from 4.1 million tons.
107
0
TABLE 49
LOCATIONS OF U.S. AIUINN S-T1ZS

(As Of January 1, 1986)
CoPany Location Status
Reynolds Listerhill, AL Closed

Revere Scottsboro, AL Closed
Reynolds Arkadelphia, AR Closed
Reynolds Jones Mills, AR Closed
Alcoa Evansville, IN
Alcan Sebree, KY
National Hawesville, KY
Kaiser Chalmette, LA Closed
Eastalco Frederick, MD
Noranda New Madrid, MD
Columbia Falls Colunbia Falls, MT
Alcoa Massena, NY
Reynolds Massena, NY
Alcoa Badin, NC
Ormet Hannibal, OH
Reynolds Troutdale, OR
Martin Marietta The Dalles, OR Closed
Alumax Mt. Holly, NC
Alcoa Alcoa, TN
Consolidated Johnsonville, TN Closed
Alcoa Rockdale, TX
Alcoa Palestine, TX Closed
Reynolds Corpus Christi, TX Closed
Alcoa Vancouver, WA Closed
Alcoa Wenatchee, WA
Reynolds Longview, WA
Kaiser Spokane, WA
Kaiser Tacoma, WA
Coma lco Goldendale, WA
Intalco Bellingham, WA
Kaiser Ravenswood, WV %
Source: United States Bureau of Mines, Primary

Aluminum Plants, Worldwide
108
FIGURE 39
ALUMINUM SMELTER LOCATIONS
FIGURE 40
COAL TAR PITCH PRODUCERS
109
K..A
TABLE 50
LWC&TIGS OF WDAL TAR PI1CH PWJIUX

IN 7M UNRITSD STAT
Company Name Location
Allied Chemical Birmingham, AL

Allied Chemical Detroit, MI
Allied Chemical Ironton, OH
Coopers Creek W. Conshohocken, PA
Koppe rs Chicago, IL
Koppers Follansbee, WV
Koppe rs Houston, TX
Koppe rs Woodward, AL
Reilly Tar and Chemical Cleveland, OH
Reilly Tar and Chemical Granite City, IL
Reilly Tar and Chemical Lone Star, TX
United States Steel Clairton, PA
United States Steel Gary, IN
TABLE 51
(MITE) STAM COMMI[NTION OF AL[EMM

(Pounds Per Person Per Year)
1974 60.4 1979 65.8

1975 44.9 1980 56.8
1976 57.1 1981 55.2
1977 60.3 1982 51.0
1978 65.9 1983 57.7
1984 59.3
Source: The Aluminum Association
110
IL
All of the major producers have revealed their intention to cut back prirmry
production to meet only their own demand or else to diversify into other busi-
ness areas offering higher return. Recent closures are noted in Table 49.
In contrast to the United States picture, overseas producers with either low-
cost electricity or government support continue to add to capacity. Expan-
sions or new facilities are under way in Argentina, Brazil, Canada, Mexico,
Venezuela, Norway, Iceland, Australia, Bahrain, Dubai and others. Low
electric power costs are the driving force for aluminum Industry relocations.
In Australia, the aluminum producers are obtaining new power contracts for
15 mills/kWh, in Brazil for 18 mills/kWh, and in Canada for less than
20 mills. On the other hand, United States producers outside the Pacific
Northwest may be subject to rates of 28-30 mills/kh or mre.
In sunnmry, although aluminum denund is expected to continue to increase

worldwide, firm electricity prices of less than 20 mili 'kWh are required for
an aluminum smelter to conpete in today's worldwide market. United States
aluminum capacity is therefore likely to continue to decrease, and along with
it, the dennd for electrode pitch.
-- --
PAY 111 OOMMQNS
Definition8 and SpecificatiMB
Coal tar and petroleum asphalt are both used as binders in road construction.
There is a great deal of confusion over terminology.
Strictly speaking, bitumen is the residue of crude oils after renval of

volatile conponents, by either natural or artificial processes. Asphalt, on
the other hand, is a mixture of bitumen and mineral matter, produced either in
nature or by artificial means. Casual usage in the United States, however,
applies the term "asphalt" to either asphalt or bitumen. The residues from
coal tar distillation are properly called pitches or tar pitches to distin-
guish their origin. In comnmn usage, tar and pitch are used interchangeably.
Various grades of road tars may be made either by the fractional distillation
of crude coal tar, or by blending different refined coal tars, or by blending
base pitch with a coal-tar derived flux oil.
In the fractional distillation process, after preheating and dehydration of

crude tar, progressively higher boiling range conponents are removed until a
residue of the desired consistency is obtained. In principle, the process is Z_
the same as that used to produce coal-tar pitch, except distillation is nor-
mally carried out at near atmospheric pressures insieed of under vacuum. Dis-
tillation teaperatures and duration times are controlled to yield road tar of
the required viscosity for a particular grade.
Grades of tar have been defined ranging from RT-1 through RT-14. The most
widely used paving grade is RT-12.
Petroleum-derived "asphalt" and coal-derived "road tar" can both be used to

form an asphaltic concrete type of paving rn-terial by mixing with gravel,
crushed rock or other aggregate. Although still used in Europe, coal tar has
been almost conpletely abandoned for road construction in the United States.
Part of the reason is probably due to lack of availability and part is due to
differences in road construction technique. In the United States the base
courses and final course are designed to flex slightly under large loads. As-
phaltic concrete made from petroleum bitumen is able to yield and recover.
Coal tar is more brittle, and requires a rigid under-base.
Tables 52 and 53 illustrate the wide differences in properties. It is not

possible to substitute one for the other, and road tar is not considered a vi-
able product today.
Mobil has recently patented a process for making a road paving nmaterial by
mixing used tires, coal-tar pitch and fluid-catalytic-cracking bottoms.
lowever, the process is not comnercial.
A potential application could lie in asphalt recycling agents. These agents

are used to dissolve and recondition old asphalt paving which is then re- ..
spread on the road. Again, there is no comnercially established product, and ,.:
therefore no further consideration will be given in thi study. , .,
Work with asphalt derived from shale oil has shown that the nitrogen compounds
112
%.J
in shale oil asphalt have superior resistance to nristure danmge and could
therefore be useful as "anti-strip" agents. These agents inprove the bond be-
tween the bitunmen and the aggregate in asphaltic concrete. It may be possible
that the nitrogen conpounds in the Great Plains tar oil could perform the same
function. This is only speculative, and no market evaluation was attenpted.
TABLE 52
O0WAARIS( OF ROAD TAR AND PEROLEUN ASPHALT
Asphalt Road
Material Type: Cement Tar
Grade: AC-10 RT-12
Elemuental Analysis, Percent:
C 85.8 92.2
H 9.7 5.2
N 0.6 1.5
0 0.5 1.0
S 2.8 0.6
C/H Atomic Ratio 0.74 1.49
Molecular Weight, (Number Average) 1,030 420
Carbon Ataom Distribution:

(Percent of Total Carbon)
Aronmt ic Carbon 34 80
Naphthene Carbon 23 15
Paraffin Carbon 43 5
113
L ,,-. , L. '.',' . ... e. .. ..... -. -. ",','v .,." " . .". ' "" .2.'', .%""4. ""V " , ,',
TABLE 53
00HPAZI aON OF PMfSICAL Piiwiri ES oF MmD TAR

AND FEI'ARLUN ASP~kLT
Asphalt Road
Material Type: Cermnt Tar
Grade: AC-10 RT-12
Original Material:
Viscosity
140 OF, Poises 1,057 133
275 OF, Centistokes 341 36
77 OF, (Initial), Megapoise 0.64 0.17
Softening Point (R&B), OF 111. 104
Penetration
77 OF (100g, 5S), 0.1 nrm 116 Too Soft
32 OF, (2 00g, 60s), 0.1 n 22 22
Flash Point (COC), OF 590 380
Solubility In CS2 , Percent 99.9 83.2
Specific Gravity At 77 OF 1.015 1.248
Ductility
77 OF, cm 100+ Too Soft
60 OF, cm 100+ 100+
Thin Film Oven Test Residue
Weight Loss, Percent 0.1 14.6
Viscosity
140 OF, Poises 2,478 465,000
275 OF, cS 49594
Ductility At 77 OF, an 100+ 0.5
114
Deflitms and Specifiteatime
Naphthalene, C1 0 H8 , consists of two fused benzene rings. Pure naphthalene has

a melting point of 80.3 'C and a boiling point of 218 *C.
Naphthalene can be recovered either from coal tar or petroleum by-products.

The coal tar product is of a lower purity and sells at a discount for that
reason. The highest grade of crude coal tar naphthalene has a melting point
of 78 *C. An impure grade melting below 74 OC may be sold. A typical coal
tar distillation plant produces three fractions with increasingly higher boil-
ing points: chemical oil, creosote, and pitch. Naphthalene is recovered from
the chemical oil fraction. The usual procedure is to obtain crude naphthalene
product by batch or continuous fractional distillation of the tar acid-free
chemical oil. If only creosote and pitch fractions are produced, it may be
possible to crystallize naphthalene from the creosote.
The production of naphthalene from petroleum involves two principal steps:

production of an aromatic oil in the naphthalene-alkylnaphthalene boiling
range by hydroaronmtization or cyclization; and dealkylation of such oils
either thermally or catalytically. The naphthalene that is produced is
recovered as a high quality product and usually by fractional distillation.
Pro act lon
Recent data on coal tar naphthalene production have not been published by the
International Trade Conmission due to the small number of companies reporting.
However, production declined steadily over the period 1968 to 1982, from 526
million pounds per year to 233 million pounds per year.
Production has declined for two reasons: a declining availability of coal tar
from the steel industry; and a declining demand for naphthalene to make
phthalic anhydride. The production of phthalic anhydride has been the largest
use for naphthalene, but o-xylene derived from petroleum has supplanted naph-
thalene as the preferred feedstock.
Pries
Although demand for crude naphthalene has been shrinking over the past several
years, this has been offset by shrinking availability, with the result that
prices appear to have held firm (Figure 41). Petroleum naphthalene (phthalic
anhydride grade) sells for about $0.08 per pound more than coal-tar-derived
material. These prices are from the Chemical Marketing Reporter for the years
shown.
There appears to be some doubt about the validity of their prices in recent
years due to the use of posted prices and the small nuner of suppliers in the
market. Direct contact with producers indicates that current prices for high
purity, refined naphthalene are around $0.31 per pound, while phthalic an-
hydride grades are $0.16 to $0.23 per pound. Coal-tar-derived material would
probably sell at the bottom end of this range.
115
~, "0
Naphthalene has a wide variety of uses, from insecticides to leather tanning
agents, but the majority of naphthalene produced is used in manufacturing
phthalic anhydride. Estimated uses of naphthalene in different markets are
shown in Table 54.
As noted above, the phthalic anhydride market has been usurped by o-xylene.
Naphthalene is one of the principal raw materials for the insecticide known
generically as carbaryl, and more comm~nly by the trade name Sevin. Union
Carbide is the United States producer of Sevin.
I 5.1
Naphthalene is also used by Arrerican Cyanamid to make beta-naphthol, an inter-

mediate for dyes, pigments and rubber processing chemicals. American Cyanamid
is the only producer.
Naphthalene -based surfactants derived from naphthalene sulfonic acid are used
as superplasticizers in concrete formuilations made by a nunber of conpanies.
S0.24-
50.23-
50122
10.21
$0.20
50.19
50.17
SSCI. 13 -
$0.112 -
50.1 1-
sc.16 190-2 17 9 i
NAPHIAEN PRCE
1967 9 70R.9E, TAKS
19O/WO1RKS
116
Pro toers
Known United States producers of crude chemical naphthalene in 1985 are

listed in Table 55. It may be noted that the listed production capacities are
much greater than the estimated production rate. Production is constrained by
the availability of raw material. Ashland no longer makes naphthalene, and
Texaco (Getty) is now the only producer of high purity material.
Market Structure
The market for merchant naphthalene is thin because much of the production is
captively consumed. Only Allied and Koppers appear to be major suppliers on
the merchant market. Ashland has recently discontinued production but con-
tinues to market naphthalene. They could be a potential buyer for Great
Plains naphthalene if a high purity material is produced.
There is considerable uncertainty with respect to the amount of naphthalene in

the Great Plains tar oil stream. Various analyses have indicated anywhere
from two percent to 15 percent naphthalene. At five percent naphthalene con-
tent, potential naphthalene production would be 16 million pounds per year,
which would be insignificant with respect to the total market.
Market Out look
Naphthalene for phthalic anhydride will undoubtedly continue to be replaced by

o-xylene as a feedstock. Other uses such as insecticide, tanning agents, moth
repellants, etc., are mature, slow-growth or declining markets. The one
strong growth area for naphthalene is the surface active agents for concrete
acnixtures. On the whole, naphthalene demand will continue to decline. Once
the decline in demand substantially exceeds the decline in production
capacity, it is likely that a large break will occur in the price for merchant
naphtha lene.
TABLE 54
3STI& U. S. NAPHTHALRKN CONISUMPTION
Percent
Use Of Total
Phthalic Anhydride 59
Insecticides (Sevin) 20
2-Naphthol (Dyes, Pigments,
Intermediates) 8
Synthetic Tanning (Leather) 6
Moth Repellants 2
Surfactants 4
Miscel laneous 1
117
". "",..Z """ - . . "'. ."o" , . " , .. ., .° . "- . ". '. -- -"++"". + " + " . "" - """ + . . . -""""" , - - -
TABLE 55
UNITE STATS PRO OF C=ICAL NAPIWHALI

(1985 Data)
Location Capaeity Source

(idlb/yr)
Allied Chemical Ironton, OH 75 Coal tar

Ashland Oil Ashland, KY 90 Petroleum
Getty Oil Delaware City, DE 100 Petroleum
Koppers Co. Follansbee, WV 190 Coal tar
Monsanto Chocolate Bayou, TX 90 Petroleum
USS Chemicals Clairton, PA & 155 Coal tar
Gary, IN
Bethlehem Steel -- Coal tar
LTV Steel -- Coal tar
Indiana Gas & Chemical -- Coal tar
Total 700+
Source: United States International Trade Comnission and Industry

reports
118
..I
MISC. LAND MS USES
Coal Dust Suppressant
Various liquids can be used to control dust emissions during coal handling and
shipping. In the Powder River Federal Coal Region (southern Montana and
northeastern Wyoming), a heavy oil is used at several mines as a dust suppres-
sant. The oil is delivered to the mines by trucks, and is sprayed onto the
coal at the tipple. Other materials have also been tested as dust suppres-
sants, but the heavy oil is preferred because it adds approximately
150,000 BTUs per gallon of oil to the heating value of the coal. .4
It has been indicated that the amount of oil sprayed on the coal is from 1.5
to 4 gallons per ton. The mDst comrnly quoted figure Is 2 gallons of oil per .
ton of coal. The four coal mines known to have installed oiling systems
produce approximately 20 million tons per year. Thus, the demand for heavy "
oil dust suppressants at these four mines alone could be in excess of
2,600 barrels per day. For the entire region, approximately 115 million tons
per year is shipped out of the Powder River.
The use of heavy oil as a dust suppressant in other areas of the country is
not conmon. Unlike other regions, most of the coal produced in the Powder
River region is shipped long distances to markets. Powder River subbituminous
coal is also subject to more decrepitation and dusting problems than most
coals. The specifications for coal dust suppressants are primarily set by the
consuming utilities. The oil functions to control dust during handling and r
shipping, to inhibit moisture loss from or addition to the coal thereby
preventing freezing of the coal, to suppress spontaneous combustion, and to
prevent decrepitation during shipment. Little is known about the chemical and
physical characteristic of oils that contribute to good or poor performance.
Oils that have been tested and rejected include: Bunker C, No. 6 fuel oil,
and crude oil. It is unknown if Great Plains tar oil could meet performance
specifications. Odor would obviously be a problem.
Known specifications for the dust suppressant are given in Table 56. p.
It is likely that health concerns and odor problems would prevent the use of
tar oil in this application. In any event, the selling price is estimated to
be not appreciably higher than the price for residual fuel.
Drilling Md
Oil-based drilling nud is used to solve a variety of problems In drilling oil

wells. The major problem is hold instability caused by the hydration of for-
mation clay when water-based mud is used. This causes material to slough away
from the walls. Although coal tar liquids might have some favorable perfor-
mance characteristics in this use, the environmental problems associated with
disposing of the mud will prohibit their use. Environmiental guidelines for
oil-based mud suggest that the base oil should have total aronatics content of
less than five percent, and no high molecular weight polycyclic aromatics.
Oils which have been used are paraffinic and naphthenic types. Phenols would
not be acceptable.
p.
119
~ ~PP%
~f~d ? 'J . - p .- -..- '...* *p7*
TABLE 56
SPECIFIC&TIGO FR COAL DUST SUPPRESSANTS
API Gravity (Degrees) -5 to +5
Pour Point (F) 40 to 65 mx
BSdW trace to 1% max
Flash Point (F) 230 nin to 275
Viscosity @ 100 F (Cst) 700 mnn to 1,500
Sulfur (Weight Percent) 2.5 to 5 max
Light Ends (% Off @ 500 F) 1 to 5 (10 nmx)
MetWI Aryl gtbers
It has been suggested that methyl aryl ethers manufactured from coal liquids
would be a desirable anti-knock additive for gasoline.
Because phenols are present in coal liquids and are undesirable in fuels, and
because destroying the phenols by hydrotreating is expensive, it was proposed
to convert the phenols to methyl aryl ethers. Methyl aryl ethers are also
knows as anisoles. (Anisole is C 6H 5 OC11 3 ). They can be prepared by direct
reaction of methanol and phenol.
Compared with the alkyl ethers (e.g. methyl tert-butyl ether, MTBE), the aryl
ethers have higher boiling points and densities, and lower octane blending
values (Table 57).
The logic put forth for investigating the methyl aryl ethers was that, if coal
liquefaction and/or gasification becomes comnercially significant, far more
phenolics will be produced than can be absorbed by the combined demnd of all
chemical industries utilizing phenolics. That is certainly not the case with
respect to the Great Plains plant and the current outlook for coal conversion.
The phenols would have to be extracted first, before reacting with methanol to
make the ethers. At the point where phenol is extracted it becomes a saleable
product. There are nmny other products which can be rmde from phenols if it
is desired to extend the nmnufacturing process downstream. No special atten-
tion to methyl aryl ethers is warranted in this study.
120
,,--.'
Briquette Binder
Coal tars are used as briquette binders in various applications. One pos-
sibility in the Great Plains marketing region is the FC formcoke plant at
Kenmerer, Wyoming. This plant makes a metallurgical coke from non-
metallurgical grade coal via the FMC Formeoke process. In this process the
coal is carbonized, the tar is separated and processed to make a good binder,
and then the tar is mixed back with the char and formed into briquettes.
Insufficient pitch binder is obtained from the process itself, so FM buys ad-
ditional binder material. Their total binder pitch requirements are about
20,000 tons per year.
Inportant characteristics for this use are softening point (ring and ball),
coking yield (at least 50 percent), quinoline insolubles (20 percent maximn),
and nutual solubility with other binder coniponents.
At present, FNU is purchasing only a small part of their binder requirements

from the outside, and they are using mlasses because it is cheaper than coal
tar.
Other
Many exanples can be found in the literature of other products which can be
made from coal-derived liquids. Notable among these are the many coal tar
.w bases which have not been discussed. Pyridine derivatives of all types can be
produced. Specialty solvents, such as 2,4,6-trimethyl pyridine were commr-
cially produced at one time. None of these uses which were examined seem to
provide a large enough market opportunity to justify significant development
efforts at this time.
TABLE 57
SANTIKNOCK PERIOlUANCg OF 9IlERS
0
% Boiling Point, Density (20 C), Performance
0
P . Conpounds C g/ml Numbera
Methyl tert-Butyl Ether 54.6 0.7403 143

* Ethyl tert-Butyl Ether 71.9 0.7395 144
Isopropyl tert-Butyl Ether 87.4 0.7413 149
Methyl Phenyl Ether 153.6 0.9939 107
Ethyl Phenyl Ether 169.9 0.9651 111
o-Methyl Anisole 171.8 0.9796 82
m-Methyl Anisole 176.6 0.9716 110
p-Methyl Anisole 176.7 0.9701 112
aASTM performance number of 25 volume percent ether in base fuel of 120
rating and containing 4 ml/gas TEL
4121
TRANSPOfrATION ROUTES AND OPTIONS
A problem to be faced in marketing certain by-products from the Great Plains

plant is the lack of a substantial market in the irnlediate region. This is
illustrated in a general sense by Figure 42, which indicates the fraction of
total United States manufacturing activity which takes place in each state.
The data are based on United States Census data for value added by manufactur-
ing.
Value added by manufacturing is derived by subtracting the cost of raw

materials, parts and conponents, supplies, fuels, electricity, and contract
work flowing into all plants from the total value of shipments out of all
plants. It is considered to be the best value measure available for conparing
the relative economic importance of manufacturing between geographic areas.
Value added by manufacturing ranges from a high of $5,193 per person in Con-
necticut to a low of $791 per person in Wyoming.
The data in the figure indicate the potential Great Plains disadvantage. Long
rail distances would be required to reach a significant fraction of total U.S.
manufacturing capability. A 500 mile radius enconpasses only about two per-
cent of total United States manufacturing activity:
Percent of Value Added

Distance, Miles Activity Accessed
500 2.0
1,000 15.0
1,500 49.0
2,000 96.0
2,500 98.0
Distances by railroad and highway to cities of interest in the inrmedlate
region, and in the nation as a whole are given in Table 58.
RMICJAL PIPELINES
The existing regional product pipelines which serve as transportation links

for petroleum products in the Great Plains region are showm in Figure 43.
Regional refining centers connected to the pipeline system are also shown in .
the figure. Major refineries located at Mandan, North Dakota; St. Paul and
Rosemount, Minnesota; Casper, Sinclair, and Cheyenne Wyoming; Laurel and
Billings, Montana; and Regina, Saskatchewan. Contact with all the major
pipeline operating conpanies in the region suggests that considerable excess
capacity is available. Thus the introduction of up to 5,000 barrels per day
of refined fuels in this region would probably displace other products through
the pipeline system.
A 10-Inch Anrmo products pipeline runs from Plandan, North Dakota through Min- N,
nesota to Minneapolis/St. Paul. The line continues to Dubuque, Iowa where it
122 .,
0.905 .62 3-10 6'3 . .30 a,
0.44~FIUR 4.028.5 .1
OFT~~~.2T L...VA U7DD D B

PERCENT
MANFATUIN IN8EACH STATE.5
S0KACH3A
MANUFACTland IN EAHSATE
SA ii
K TC En Aa M ANo
ITO nBA S nc a
2 Kaeb cCena Litle mer. Ra7
3 enx Ite'pov0f dCooc kMotea
Cooco 0
4 MACO 2 Wilias e onamers1 Muntanee
4
ott. 9Mdcn o 3 YO fExn aPtoCnd
ND 12 M
12p
WY 5 D
TABLE 58
DISTANCE FMNOMBEULAH 70 U. S. AN) CANADA CITIES, BY

RAILMD ND HIGMBY, IN MILES
Distance By Distance By
Railroad Highway
Regional
Billings, MT 503 405

Grand Forks AFB, ND 344 286
Great Falls, MT
(Malmstrom AFB) 738 597
Laurel, MT
(Cenex Refinery) 518 420
Mandan, ND
(Amco Refinery) 73 73
Minot AFB, ND 268 105
Rapid City, SD
(Ellsworth AFB) 631 325
Regina, Saskatchewan
(Consumr's Refinery) 559 413
Winnipeg, Manitoba 503 578
National
Atlanta, GA 1,635 1,664

Calgary, Alberta 951 739
Chicago, IL 899 915
Cincinnati, OH 1,180 1,214
Cleveland, OH 1,250 1,267
Houston, TX 2,273 1,567
Los Angeles, CA 1,914 1,538
Minneapolis, MN 498 509
New Orleans, LA 1,797 1,652
Philadelphia, PA 1,730 1,704
Pittsburgh, PA 1,381 1,396
St. Louis, MO 1,098 1,120
Seattle, WA 1,378 1,145
Winnipeg, Manitoba 578 417
124
then splits into two directions. One 10-inch line connects to Chicago and
Detroit. The other, a 12-inch line, connects to Kansas City. Information olx
flows in this line was not available.
Cenex
An eight-inch Cenex pipeline originates at their refinery in Laurel, Montana

and flows to Minot, North Dakota. Cenex representatives felt that production
of light products by the Great Plains plant would be serious cornpetition for
them, and declined to provide their capacity or average daily flow.
Co0o0
A 10-inch Conoco products pipeline originates in Billings, Montana and flows

to Grant County, Washington. An additional six-inch pipeline connects to this
line in Helena, Montana from Great Falls, Montana. Average daily flow is
20,000 barrels per day from Montana to Washington.
An eight-inch Continental products pipeline originates in Billings, Montana

and flows to Sinclair, Wyoming. A six-inch Medicine Bow Products Pipeline
System pipeline then flows south of Sinclair to Comnerce City, Colorado, while
an eight-inch Pioneer pipeline, also owned by Continental, connects Sinclair
and Salt Lake City.
Flow in the Continental pipeline averages 10,000 barrels per day from Billings
to Sinclair. Flow in the Medicine Bow pipeline averages 19,000 barrels per
day from Sinclair to Comnerce City. Flow in the Pioneer pipeline averages
30,000 barrels per day from Sinclair to Salt Lake City.
Dome Petroleum has a 12-inch propane line that originates in Ft. Saskatchewan,
Alberta and flows southeast through Saskatchewan, North Dakota, Minnesota, and
on to Ontario via Illinois, Indiana and Ohio. The flow nostly passes through
provinces and states in the Great Plains region, with small deliveries of
3,000 barrels per day in Saskatchewan and 7,000 barrels per day in Minnesota.
The flow into Saskatchewan averages 40,000 barrels per day while the flow out
of Minnesota averages 30,000 barrels per day.
Interprovincial
A 16-inch Interprovincial products pipeline originates in Edmonton, Alberta

and flows in a southeas'-erly direction into Saskatchewan, Manitoba, and Min-
nesota. Average daily 1ows are 125,796 barrels per day from Eckmnton to Sas-
katchewan, 115,103 barrels per day from Saskatchewan into Manitoba, and 61,011
from Manitoba to Minnesota. Interprovincial expects a 30 percent increase of
all products rmving through this pipeline in 1987.
A 12-inch MAPO) LPG pipeline runs from southwestern Wyoming to Farmington,

New Mexico, Texas, and then northeast to Janesville, Wisconsin. The flow from
Wyaning south averages 28,000 barrels per day. A second MAPCO line is a two
branch line that rrves propane at 10,000 barrels per day in one branch and NGL
125
- & . bh *
.....-
at 10,000 barrels per day in the other. Both lines are six inches in
diameter. Direction of flow is north into Minnesota from Iowa.
Pet ro-C4mada
The six-inch Petro-Canada line moves products from just west of the
Alberta/Saskatchewan border into Saskatchewan and Manitoba. Average
deliveries to Manitoba are 10,700 barrels per day.
Williams
Williams has an extensive products pipeline network in Minnesota. The infor-

mation is sunnmrized in Table 59. The two six-inch and one 12-inch pipeline
that flow from Minneapolis south connect with Des Moines, Iowa, then Kansas
City, Missouri and Tulsa, Oklahoma.
Wyco
A 10-inch and an eight-inch Wyco products pipelines connect Casper, Wyoming

and Colorado. The eight-inch terminates in Commerce City and the six-Inch
terminates south of Colorado Springs. Average flow from Wyoming to Colorado
is 37,450 barrels per day. Another branch runs from Wyoming to Rapid City,
South Dakota at 5,000 barrels per day. A connection with the Cheyenne
pipeline runs between Wyoming and Nebraska at an unknown flow rate.
TABLE 59
WILLIAM PROU PIPELINES IN MINNESOTA "
Fig- 6
ure # Direction BPD Size Capacity (BPD)
I MN to WS West to East 13,200 8 inch 22,000 ".v-

2 MN to WS South to North 10,500 8 inch 14,000
3 WS to MN North to South 10,500 8 inch 14,000
4 rFti to ND 14,300 8 inch 22,000
5 SD to MN 9,600 8 inch 24,000
6 MN to SD 4,400 6 inch 11,000
7 Mi to South of MN 32,000 6,12 inch 80,000
12%
126/
~ ~~A% jTi-(,'. %-
RAILIKMDS
Figure 44 illustrates the major railroad connections in the Great Plains

region. Figure 45 shows the major transcontinental routes to cities outside
the imediate region. Mileages from Beulah, North Dakota to regional cities,
Air Force bases and refineries were given earlier in Table 58. To obtain
shortest distances possible, some of the mileages given involve switching rail
line corpanies. This might not be feasible due to lack of switching arrange-
ments, or unwillingness by the conpanies to conply.
A significant fact is that the Great Plains region is served by only a single
railroad, the Burlington Northern, with only a single route through that
region. Thus anyone using the railroad in this area becomes a "captive
shipper".
The Staggers Rail Act legislation, passed in 1980, partially deregulated the
railroads. Previously, the Interstate Commerce Conmission (ICC) assigned
routes and authorized rates. Sam irnrdiate results of the Staggers Rail Act
have been:
Long term contracts now possible

Rail rate increases
Increase railroad management responsibilities
Railroads now have increased flexibility in setting rates. Rates can be in-
creased on a 20-day notice, and reduced on a 10-day notice. The ICC can ques-
tion a rate if it exceeds 180 percent of the revenues to variable cost ratio.
Investigations by the ICC cannot be performed unless a shipper charges that

the railroad has "market dominance" and there is no alternate means of
transporting the product. If the shipper can prove excessive rail rates and
no alternative conpetition exists, the "revenue adequacy" is considered, and
the financial position of the carrier is reviewed.
The existing Burlington Northern rail system in Central North Dakota connects
to the east with rail lines in Fargo, North Dakota and Minneapolis, Min-
nesota. To the west, connections are made in Terry, Billings, Helena and
Butte, Montana. Connections from Montana can be made to the West and Gulf
Coasts.
For established rmvermnts of particular products by railroad, rates can be ob-

tained from the carrier in the area. Specific points of origin and destina-
tion are required. Information on contract rates, however, is usually not
available. These rates vary depending on whether the railroad or the shipper
owns the cars, the interchanges required with other rail lines, the condition
of specific sections of track and the allowable speed thereon, the state of
congestion at any terminals involved, etc. Thus, specific estimates for a
contract novement are conplex and difficult to calculate, and are not freely
provided by the railroads.
In general, railroad tariffs per ton-nile vary greatly for short hauls, and
decrease significantly (perhaps by a factor of three) as the length of haul
increases to the range of about 500 to 1,000 miles. Beyond 1,000 miles the
rate per ton-mile usually remains fairly constant.
127
NCP CaainPcfc
SL SooLn
MAJO RALODSIRATPANSRGO
Plantt O
FIUR 45
OUTE FORGREA
MAJORRAILOAD PLANS POUT
N SL8
C&NW
Na NW.
The Burlington Northern provided a current quote for products transport based
on Rule 35. Their rate for Beulah, North Dakota to Minneapolis, Minnesota is
$3 per hundred-weight or about $10.50 per barrel. These rates would clearly
be uneconomical. BN was asked to supply an estimate of rates which could be
provided under long term contract. Although the contract rates they were
willing to suggest on a casual quote are likely to be somewhat different than
could be negotiated in an actual situation, these rates are probably within
20 percent of a realistic value. Estimated rates for shipment to key cities
are shown in Table 60. Using these rates as a basis, and applying the same
ton-mile rate as a function of distance, the costs for mving liquid products
to various major cities in the United States are given in Table 61.
These rates are for shipper-owned cars. Thus, total costs must include an
added arunt for car ownership or leasing. Ownership or leasing from third
parties is generally cheaper than using railroad-owned cars. In the latter
case, charges may be on the order of $0.60 per mile for a 100-ton car, or
0.6 cents per ton-mile.
Rates will also vary with the density of the product. The rates used in
Table 61 assume that 190,000 pounds can be shipped in a standard 23,000 gallon
tank car. If a light liquid is shipped, so that only 140,000 pounds can be
loaded per car, rates would be 20 percent higher.
The railroads have dernstrated an extreme sensitivity to the presence or ab-

sence of conpetitive mdes of transportation. In the absence of shipping al-
ternatives, "captive shippers" have been forced to accept large rate increases
since deregulation. Only the presence of conpetition, either from other rail-
roads or other mdes of transport, provides an effective bargaining tool.
This was illustrated clearly when the railroads substantially cut the price
for rmving Powder River Basin coal to Arkansas and drove the ETSI coal slurry
pipeline out of business. Shipment of liquid products from Great Plains would
be captive to the Burlington Northern. The closest pipeline is the Armco
pipeline that runs east to Minneapolis. The pipeline, of course, would accept
only fully refined products conpatible with the other products being shipped
in the line.
TABLE 60
KEY KSTIA1TKS OF RAIL TRANSPORT COSTS
Destination Miles $/BBL Cents/ton-mile
Minnapolis/St. Paul (All 8N) 500 3.86 4.4

Regina, Sask. (BN, Soo, CP) 560 4.30 4.4
New Orleans (All BN) 1800 6.60 2.1.
New Orleans (ON,UP) 1800 7.92 2.5
129
TABLE 61
zwriIKT Ool FOR RAILROAD SHIPMTiT

OF LIJID PB17 rmF BUJLA, NORM DAKOTA
Route Dollars
Destination Miles Railroads Per barrel
Calgary, Alberta 1,262 BN, CP 6.29

Chicago, IL 1,204 BN, lOG 6.00
Cincinnati, OH 1,694 BN, B&O 6.90
Cleveland, OH 1,553 BN, lOG, CR 6.80
Houston, TX 2,365 BN 6.90
Los Angeles, CA 2,293 BN, UP 8.20
Minneapolis, MN 500 BN 3.86
New Orleans, LA 2,019 BN, MP 6.33
Philadelphia, PA 2,063 BN, CR, ICG 6.47
Pittsburgh, PA 1,710 BN, CR, IOG 7.12
Regina, Saskatchewan 560 BN 4.30
St. Louis, MO 1,384 BN 5.80
Seattle, WA 1,461 BN 5.90
Railroad Key:
BN Burlington Northern
B&O Baltimore and Ohio
CP Canadian Pacific
CR Consolidated Rail
ICG Illinois Central Gulf
MP Missouri Pacific
UP Union Pacific
1.3
130
If the jLotal Great Plains liquid stream were shipped to a single destination,
the volume would be nmrginally large enough to qualify for unit train treat-
rent. A mininum of 60-80 cars at a tire to a single destination would be
required. With a unit train turnaround time of one week, this eans that an
average daily shipment of 4,000 to 5,000 barrels would be necessary.
Unit train costs are strongly influenced by whether more than one railroad is
involved. When crossing from one railroad system to another, the train rey be
stopped, the engines and caboose disconnected, and a new set of engines and
caboose hooked up to continue the journey.
Truck transport is not conpetitive with railroads in mving bulk products over
long distances. Truck transport offers an option for transporting liquids
over relatively short distances where pipelines or railroads either do not ex-
ist or are not appropriate. Trucks could possibly be used for delivery of
products to Minneapolis/St. Paul or Duluth for barge shipping down the Missis-
sippi or across the Great Lakes. Truck transport can be cost-conpetitive in
some situations because the only capital costs are for tractors and trailers,
whereas for rail transport capital expenditures are required for the track.
One of the greatest benefits of truck transport is that smll quantities of

product can be directed in nmny different locations. The mjor drawback to
truck transport is the limited, if not nonexistent, econon-, of scale. Such
economies are large with pipeline transport and substantial with rail, but In-
significant for truck transport.
The costs for truck transport were calculated from infornmtion provided by two
trucking conpanies located in Minneapolis:
Copany Destination $/Barrel $/ton
Dah!ine Minneapolis 10.32 58.98

Quickie Minneapolis 7.73 44.20
Dahline Duluth 10.57 60.38
Quickie Duluth 8.96 51.20
The naxirnvn truck size allowed in Minnesota is 80,000 pounds gross weight or
about 170 barrels net. In North Dakota, double bottoms are allowed, or
120,000 to 140,000 pounds gross weight. These nine-axle units can haul about
90,000 pounds of payload or about 250 barrels of heavy oil.
On a volume tender of several truck loads per day, a negotiated contract at a

lower rate would be possible. A .0 percent reduction of the lowest rate given
would result in $6.20 per barrel to Minneapolis. This is less than the exist-
ing tariff under Rule 35 quoted by the railroad, but considerably higher than
the $3.86 per barrel which could probably be obtained through a negotiated
contract with the railroad.
For shipments within North Dakota, where double trailer rigs can be used,
trucking costs could be competitive with rail rates, with both In the range
$0.01 per barrel per mile.
131
River transport of bulk liquid products is acconplished primarily by towboats
pushing rafts of barges. The towboats can range in size from less than
1,000 horsepower up to 10,500 horsepower. Larger towboat designs in the fu-
ture are unlikely due to limited channel depths and lock sizes. The average
power rating for most towboats on United States waterways is about
5,000 horsepower.
There has been little standardization in the sizes of barges, except that due
to the size of existing locks on the rivers. The size and number of barges
used depends on the size of the locks which will be traversed and the width
and depth of the river. A list of typical barge sizes and their capacity is
given below:
Capacity
Length (ft) Width (ft) Draft (ft) (Barrels)
175 26 9 7,000
195 35 9 10,000
290 54 9 23,000
Actual towing operations involve lashing a flotilla of barges together to form

a single unit. Average tow speeds are on the order of five miles per hour. A
towboat may push a single barge or as many as 45 barges at a time. When pass-
ing through large lock systems, barge groups of four-wide and three-long or
four-long may be used depending on the size of the locks and the size of the
barge. Smaller locks may handle only one or two barges at a time. Integrated
tows provide an efficient method of transporting large tonnages over long dis-
tances.
The size, shape, and capacity of an integrated barge tow is greatly affected
by the physical dimensions of the waterway and the construction of the lock
systems. Figure 46 shows navigable rivers accessible to Great Plains
products, once transported to Minneapolis or Duluth.
Lock chaners of proper size are extremely inportant to the overall economics
of a barge operation. If the tow nust pass through small locks, then breakup
of the tows will be necessary. Breakup and reassembly of the barges, together
with tow loekage operations, can add substantial charges to the total
transportation costs. It is inportant that tows move smoothly in and out of
locks. Traffic surges and delays in and around locks can produce the same
types of problems as are encountered on highways at peak usage times.
The movement between Minneapolis and New Orleans using dedicated boats/barges,
totals 25-30 days round-trip, assuming one day at the origin for loading and
one day at the destination to unload. Movement to Houston or Corpus Christi
uses the Port Allen cut-off from Baton Rouge to Morgan City, Louisiana, and
then westward on the Intracoastal Waterway. The Houston destination adds
8.5 days to the round trip transit time, and Corpus Christi adds 13.5 days.
As with unit trains, operating costs make up a large percentage of the total
costs associated with a tow. Because of this, future barge rates are expected
to rise approximately parallel to the general rate of inflation. Many escala-
132
FIGURE 46
INLAND WATERWAY TRANSPORTATION SYSTEM
cost of wages, fuel, and operating supplies, as reported in the Wholesale

Price Index. If loading times are extended due to crowded docks or slow
pumps, a waiting charge is assessed, at a typical rate of $185 per hour.
The barge industry for the most part is unregulated. Each operator can set
his own rate depending on his particular situation. For instance, if a back-
haul load is available for one operator and not another, the rates quoted may
be vastly different. Rates not only vary widely, but can change rapidly.
Contacts in the barge industry currently consider prices to be depressed and
expect this to continue until the overabundance of equipment on the rivers is
diminished in two to three years.
In order to reach the waterway system, products could be shipped by railroad

to Minneapolis or Duluth and then transferred to barges. Price quotes for a
typical four piece tow from Minnesota to New Orleans (65,000) barrels are
$3.00 per barrel if no heating is necessary. Ingram Barge Company has the
facilities to load directly from rail cars to the barge at Minneapolis, so no
storage would be necessary. If large tows can be put together, the price
could drop to $2.00 per barrel.
Using barges as a means of transportation from Minneapolis is limited to a

shipping season of April to November due to ice on the river. The city far-
thest north on the Mississippi that is typically ice-free year round is
Quincy, Illinois, about mid-way up the state. Similarly, shipment from Duluth
on the Great Lakes is limited to the ice-free season. An alternate route to
the Great Lakes is possible from Minneapolis down the Mississippi to Just
above St. Louis, then up the Illinois River to Joliet where the Chicago ship
133
the Great Lakes Is possible from Minneapolis down the Mhississippi to just
above St. Louis, then up the Illinois River to Joliet where the Chicago ship
canal connects to the Great Lakes. Scrw lce problems can be avoided by spot-
ting with smaller barges fron Minneapolis to St. Louis, but this increases
costs significantly.
Major cities in the East are reached via the Ohio River system.
41
134'
BK PRICES
mum" PRICES
In order to carry out economic evaluations for the various product pos-
sibilities which exist, it was necessary to determine market prices for each
of the products.
Because of great changes in the prices for many products over the last two or
three years, there is considerable uncertainty with regard to the prices which
would exist at the time any by-products actually reach the market. Neverthe-
less, it was assumed that, in most cases, the most recent prices available
would be the mst valid for establishing an economic base case. Wherever pos-
sible, early 1987 prices were used.
Sources of price information are varied. Petroleum products are summarized

and averaged on a regional basis by the Department of Energy's Petroleum
Marketing Monthly. Current quotes for specific products in specific locations
are available from publications such as Platts Oilgram Price Report.
For chemical products, the mst conprehensive and regular source of price
listings is the Chemical Marketing Report (CMR). Sporadic price information
on specific products appears in publications such as Chemical Week. A major
problem with prices quoted in the CMR is that they are often posted prices or
list prices, which may be at significant variance from actual transactions
taking place In the nmrket place. The only way to obtain current information
on actual transactions is by direct contact with buyers and sellers. Although
dozens of phone calls were ade during the course of this study, there were
too many different products involved to be able to construct a conplete pric-
ing picture for every product. Also, many conpanies, for conpetitive reasons,
are unwilling to share price information.
Direct contacts were used as a check against prices in the CMR. If CPAR prices
were found to correspond approximately with prices obtained by direct contact,
then the OM prices were assumed to be valid and were used. If a significant
discrepancy showed up, then further effort was devoted to direct contacts in
order to assess actual market prices. An adjustmrent to the CTIR price was then
made on the basis of the best information available.
Even direct contacts may not be fruitful in cases where there are few transac-
tions taking place, and where products are purchased on long-term contracts or
on once-a-year tenders.
A major difficulty airises when products are sold in various grades, usually as
a function of purity, at significantly different pri levels. It is tenpt-
ing, but unrealistic, to use the highest prices reported. Price differences
between grades must be attributed to the cost of refining or upgrading the
product from one grade to another. For the purposes of a first-stage economic
analysis, it should be assumed that these costs equal the price differences.
Although the prices used in this analysis are conservative in the above
aspect, they are undoubtedly optimistic in many cases sinply because the
processes and costs required to upgrade the Great Plains liquids to a par-
135
.~ .Y - A V.,
ft.
ticular specification are unknown. If the product is one which has to be

mdified to meet a nurrmer of different buyer specifications, the costs will be
even nmre.
In sumnary this study uses the rmst current (early 1987) price information
available to construct an economic base case for each product slate which is
of interest. These prices are usually for the lowest-value grade found in
conmerce. Although this assunption seems safely conservative, It must be em-
phasized that it may be necessary to accept even lower prices until product
performance has been established.
TRANSPORTATION
For products which could be sold in different parts of the country, transpor-
tation costs will ultimately define the imrket locations in which the Great
Plains products will be rmst conpetitive. One approach to the market analysis
would have been to siaply conpute the cost of transportation to the nearest
potential buyer. However, there may be any reasons why that particular buyer
would not end up purchasing all, or even any, of the Great Plains products.
An average transportation cost to a region with potential demand several times
the Great Plains output would be a mere conservative assunption.
Conparison of several different transportation routes by rail and barge showed

roughly conparable costs. All-rail to the Gulf Coast was conparable to rail
to Minneapolis and then barge down the Mississippi. Both were conparable to
rail to Duluth and barge through the Great Lakes. All-rail to Pennsylvania
was coeparable to a combination of rail and barge up the Ohio River. The
first few hundred miles of product shipment are by far the most expensive, on
a cents-per-ton-mile basis. Transportation costs per mile are relatively low
after the first 1,000 miles. The isolation of the Great Plains plant means
that mst markets are over a thousand miles away, and therefore transportation
costs to different markets are roughly of the same order of magnitude. An es-
timated transportation cost of $6 per barrel was found to cover an amazingly
large nuber of cases.
NK1MX PRICES
Subtracting transportation costs from market prices yields the netback value
of products at the Great Plains plant gate. Calculated netback values used in
the economic analysis are listed in Table 62.
136
:..-:
::- .:.- :'--...
. . . -- . : .. .. _-
liliIiil iii~ lllill
lil. ... .. ........ :.: : 1,. ..
; ".:.: :..-.....
"-"""'... "'"
.. ...... ,.,,
"t
II-v- -% o~
TABLE 62
1987 PRMtXT NIR1B PRICKS

70 GU3T PLAIN PLA~f
(Dollars Per Pound)
Benzene 0.157
Carbon Black Feedstock 0.029
Cresols, Cresylic Acid,
Refined Cresylics 0.383
o-Cresol 0.503
m,p-Cresol 0.551
Creosote 0.090
Gasoline 0.081
Lignite Tar 0.108
Methane 0.343
Naphtha lene 0.143
Phenol 0.213
P itch 0.108
Raffinate 0.060
Rubber Processing Oil 0.053
Solvent Oil 0.108
Toluene 0.125
Xylene 0.160
Xylenols 0.503
61 -
.,XVt -
137
PIEXZSSIM OPTIONS
RFINERY FEED3I7X
Examination of the hydrotreating capabilities of the regional refineries sug-

gests that none would be able to process Great Plains tar oil as a crude
feedstock. However, a new heavy oil upgrader under construction in Regina,
Saskatchewan would possibly be able to do so. The upgrader will have a
capacity of 50,000 barrels per day based on the Unicracking technology
provided by Unocal Corporation.
The Unicracking process is illustrated in Figure 47. Feed is combined with

hot hydrogen-rich recycle gas, contacted with a hydrotreating catalyst for
partial desulfurization and denitrogenation, and then contacted with
hydrocracking catalyst. Unicracking catalysts for a specific unit are
selected for optinum attainment of processing objectives, and are based on
molecular-sieve containing supports. All of the catalysts have very long
operating cycles and are readily regenerated. Product from the hydrocracking
reactor is condensed, separated from hydrogen-rich recycle gas and frac-
tionated into desired products. Unconverted oil is recycled to the
hydrocracking catalyst. Sixty one Unicrackers have been licensed to date.
The Great Plains tar oil stream would amount to only about five percent of the
Unicracker's capacity, and thus could possibly be blended into the feedstream.
REACTOR REACTOR SEPARATOR SEPARATOR FRACTIONATION
Fuel gem
C Lighte
Light
naphtha'h
Heavy
naphtha S
Tbne or
7,UR 470.
UNICRACKING PROCESS ,
13ed
138
.- ~•.. .'/'
DISTIIUATICIN
Coal tar distillation was once a widespread industry in the United States,
producing a wide spectrum of products. Today it has shrunk to the point that
the few refineries left operating are for the most part only splitting the tar
to make a creosote product and a pitch product. A multiproduct tar oil refin-
ing facility at the Great Plains plant might resenile the older operations at
the coal tar refineries.
The warning that low tenperature tars are entirely different materials than
coke oven tars applies to the basic process of distillation as well as to any
other processing unit. Established processing techniques for coke oven tars
may not be applicable to the Great Plains materials. There are no large tar
distilleries in the United States operating on a similar feedstream, although
there are installations in Europe.
Two major differences of inportance to distillation are the instability of tile

Great Plains tar oil, and the large amount of tar acids which it contains.
The University of North Dakota Energy Research Center reports that it is im-
possible to perform an ASTM D-86 distillation on the tar oil because it starts
deconposing before the last 10-30 percent distills over. Deconposition and
polymerization is indicated by smoking. The tenperature at which smoking
begins depends on the heating rate used.
The laboratory results indicate that to distill material in the 260 to 300 OC
range, vacuum distillation would be required to avoid smoking. It might be
just barely possible to fractionate off the cresylic acids before this hap-
pens.
It is not clear what the effect of smoking would be on distillation. The ac-
tual corposition of the bottoms does not appear to change significantly, at
least in a short period of time.
Because of the instability problem, a tar oil distillation system would have
to be a conination atmospheric and vacuum unit rather than a straight atmos-

pheric still. Coal tar stills have been designed in the same way, although
for different reasons. These reasons include the much heavier constituency of
coal tar and the need to use vacuum to increase the recovery of distillate
materials. A combination atmospheric and vacuum tar distillation system is
illustrated in Figure 48, showing the coiplex heat exchange arrangements which
may be used to recover the process heat. In this Teerverwertung-Koppers
design, the crude tar is purped through five heat exchangers, in which it ex-
tracts heat from all the oil products and the pitch, and is purped into the
dehydration tower. The light oil and steam vapors from the top of the
dehydrator are cooled, condensed, and separated, and some of the light oil
passed back as reflux. The dehydrated tar is punped through a heating coil in
the tube furnace and injected into the atmospheric pressure fractionating
stage, which consists of two fractionating columns and a splitting column.
The products include residual light oils and water, taken as overheads from
the first colunr and passed back to the dehydrator, phenolic oil as a side
stream from the first colunm, naphthalene oil as the overhead from the second
colunn, and wash oil as the base product from the second column.
139
0
w
z
U z
zz
40 ~doe
0C
'Tr
uw
00
z4
La
00
IC.
1401
The topped tar from the base of the first column is heated in a separate heat-
ing coil in the furnace and injected Into the pitch flash chamber that, like
the final fractionating column, is maintained under a vacuum of 100 rrm Hg.
The residual oil vapors separated from the pitch are divided into a heavy wash
oil, and a light and heavy anthracene oil in the vacuum column.
The detail of the preceding design would not apply directly to the Great
Plains tar oil because of its much lower content of high boiling material.
The high content of tar acids found in the Great Plains tar oil stream may af-
fect distillation econoics because of their corrosive effect. In most tar
distillation systems, the crude tar is doped with a small amount of aqueous
alkali to neutralize aninnium salts and thus prevent corrosion. If this is
not effective, more expensive materials of construction would have to be used
to avoid corrosion problems.
PITCH
Although there are references in the literature to thermal treatment of low

temperature tars to make a binder pitch, it is not known whether the Great
Plains low temperature lignite pitch can be processed to yield anode grade
pitch. In making binder pitch from coke oven tars, the crude tar is distilled
to a medium-soft pitch residue and then hardened by heating for 6-10 hours at
385-400 *C. This treatment increases the toluene-insoluble content with only
a slight increase In the quinoline-insoluble material.
The coke yield from coal tar pitch has been increased by adding an air blowing
or catalytic oxidation step before coking. If an analogous process could be
devised for low temperature lignite tar without increasing the quinoline in-
solubles, perhaps a suitable binder could be produced. Domtar Inc. has
reported (1984) success with a new filtration process for removing quinoline
insolubles. Quinoline insolubles consist of various resins (true QI) plus
particles of dirt, coke, crystals, etc. which agglomerate to form lumps in the
pitch. These lumps are detrimental to the flow and irrpregnating properties of
pitch when mixed with coke to form electrode paste. It is conceivable that
the Domtar proc-ss could be coupled to a pitch rodification step.
Pitch once was stored in solid form at the tar distillery in open bays, from
which it was removed by small explosive charges, and then loaded by mechanical
shovel. Now, pitch is stored in tanks heated by steam or circulating hot-oil
coils and transported in liquid form in insulated rail or road tankers. When
transport as a hot liquid is not feasible because of distance or quantity or
need for long-term storage, the pitch is converted into a dust-free particu-
late form (short rods termed pencils, pastilles, or flakes).
Some plants employ cooled-belt flakers. Molten pitch flows from a tank over a
weir to give a flat thin sheet on a flexible steel belt, which is cooled from
below by water sprays. At the end of the belt, the solid pitch is broken up
by rotating tines.
Direct water cooling is also employed in a pitch-pencilling plant. MoIt en

pitch is punped through nozzles at a rate and pressure to give a series of al-
nrst horizontal jets. These jets project a short distance into horizontal
tubes through which a cocurrent streamline flow of cooling water is rMin-
141
tained. The stream of pitch solidifies into solid rods about 10-20 an thick
which break up into short lengths as they are forced around a bend in the
tube.
In another type of pitch-pencilling plant, circular streams of mDlten pitch

from numerous nozzles fall by gravity into a tank containing water where the
pitch hardens into solid rods.
Naphthalene is a principal conponent of coke oven tars and can be concentrated

to a reasonably high level by primary distillation. A naphthalene oil can
be produced which contains 60-65% naphthalene. In an older procedure, the
naphthalene oil is siaply cooled in pans, and the naphthalene crystallizes
out. The crude crystals are centrifuged and then pressed in a mechanical
press. The lower melting crystals which are expressed as a liquid are
95.5 percent pure and are satisfactory for producing phthaltc anhydride.
With the Lurgi gasifier tar oils, it will be necessary to renDve the tar acids
before naphthalene can be crystallized.
A variety of processes have been developed for upgrading crude naphthalene to

a higher purity grade. In one approach the crude crystals are melted and then
allowed to re-solidify in a tank. The tank is then warmed to just below the
desired melting point, and any oil formed is drained away.
In another approach, crude naphthalene crystals are melted, charged to tall

cylindrical tanks, and partially cooled. Mother liquor is drained off, and
the crystals washed with methanol. Brodie crystallizers developed to separate
o- and p-dichlorobenzene, have also been used for naphthalene refining.
WISILIC ACID
Because phenol, the cresols and the xylenols all boll over a narrow tenlpera-
ture range, a crude tar acid cut can be obtained from the tar oil by distilla-
tion. This cut will also contain neutral oils and tar bases. Because these
other materials boil in the same range, the tar acids cannot be purified by
distillation. Three different methods of separation which have been practiced
eomnrcially are described in the following.
Caustic Extraction
The tar acids can be separated from the neutral oil and tar bases by first ex-
tracting the acids with caustic, then "springing" them from the caustic by
neutralization with carbon dioxide.
Refined cresylic acid rmy then be produced by sieple distillation. Phenol and
o-cresol ,nuy be separated relatively easily, but the other cresol and xylenol
iomers have boiling points too close together to separate easily by distilla-
tion alone.
In n typical pr,),odtir-, t.ir oils are ,nix.d with a slight excess of 10% aqueous
caustic soda in stirred vessls or continuous extraction columns. The extrac-
tion i,§ ,arri,,d out at i:vbiint teiperature or just above the crystallizing
t42
point of the oil. It is best carried out in two stages, using 90 percent of
the alkali to contact the fresh oil and the remainder to conplete the removal
of the phenols. The crude phenate or cresylate solution is separated, and
contains some neutral and basic naterial which must be removed by extracting
with phenol-free light oil and/or treating with live steam.
The purified sodium phenate solution is then deconfposed by passing it down

packed towers at 80-85 0C countercurrent to a gas containing 25-30% carbon
dioxide. This operation is called springing. Since the rate of C02 absorp-
tion depends on the C02 partial pressure, a richer gas reduces the nunber and
size of the springing towers required. Favorable conditions would be present
at the Great Plains plant because of the availability of C02 vent gas. An up-
per layer of crude wet tar acids and a lower layer of sodium carbonate solu-
tion is obtained. The crude wet tar acids contain ca 20-25% sodium carbonate
solution and, to reduce their moisture content to ca 10 percent, the crude wet
tar acids are passed down an after-carbonation tower countercurrent to a
stream of the 00 2 -containing gas. Separation of the layers yields crude tar
acids containing ca 10% liquor and a solution of sodium carbonate and bicar-
bonat e.
Caustic soda can be recovered from the carbonate/bicarbonate solution by

treating with an excess of hard-burned quicklime in a stirred reactor. Dis-
posal of the washed limne nud poses a problem.
In the next stage in the recovery and refining of tar acids, water and pitch
are removed from the crude tar acids in a continuous-vacuum still. The
aqueous phenol overhead distillate is recycled, the stream of once-run tar
acids is refined, and the phenolic pitch bottoms are burned.
The once-run tar acids are fractionated in three continuous-vacuum stills con-
taining 40-50 bubble trays and operating at reflux ratios between 15 and 20:1.
The overhead product from the first column is 90-95% phenol; from the second,
90% o-cresol; and from the third, a 40/60 m,p-cresol mixture. Further frac-
tionation is required to give the pure products.
At other refineries, only two continuous stills in series are used, but these
are of 80-100 plate efficiency and yield pure grades of phenol and o-cresol
and a base mixture of cresols, xylenols, and higher boiling tar acids. The
latter are fractionated batch-wise to various salable grades of cresylic
acids.
A flow sheet of a typical tar acid extraction and refining plant according to
the above description is shown in Figure 49.
Phenoraff in Extractlon
The Phenoraffin process developed by Lurgi uses physical solvent extraction

rather than caustic reaction to extract phenols from tar oils. A flow sheet
is given in Figure 50.
The phenolic oil stream is charged to the middle of a imlti-stage extractor.

Aqueous sodium phenolate solution is adknitted to the one side of the extractor
and light naphtha or toluene to the other side. The aqueous phenolate solu-
tion picks up the phenols, and the light naphtha absorbs the neutral oils.
143
*. ~ ''*,, ;.- .%U

FEI-
4L4
I-
144U
is is i ~
.IGHIT MAPSIINA C0 PHNO FEWh ETON Von ar
EfiLTTAM
Cra N PtENISOL
AaWEOUS PEI0AIEUTO
EXTRACT WAS"
NIURL OIL EPPLUENT TO PWNIOSOLVAN PLANT ARPONIMICPHEIO
FIGURE 50
PHENORAFFIN EXTRACTION PROCESS
Light naphtha and neutral oils are separated by distillation, the light naph-
tha being returned to the extractor.
The phenol-laden phenolate solution contains a small anmunt of naphtha and

neutral oil which are removed in a stripper. The phenol/phenolate solution
leaving the stripper bottom is treated with isopropyl ether (IPE) in an ex-
tractor whereby the phenol dissolves in the IPE. The remaining phenolate
solution is freed of dissolved IPE in a stripper before it is returned
to the%
nialti-stage extractor.
The phenol/IPE extract after remoival of pyridine by sulphuric acid wash, Is

distilled into IPE and refined phenols.
Methanol/Hexane ExtractionV
Pitt-Consol Chemical Cm~any operated a cresylic acid extraction plant at

Newark, New Jersey which produced about 35 to 50 million pounds of cresylic
acids per year. It used a dual-solvent system similar to the Phenoraffin ap-
proach, and an agitated Schiebel-type extractor. The solvents were methanolr
and hexane. Both coal tar and petroleum caustic washes were used as 4
feeds tock.
This sanm system was used in a pilot plant extraction facility for the lignite
tars produced by the Parry carbonizer which was built at Rockdale, Texas. The
sauim approach was planned to be used for the proposed low-tentperature tar
refinery at Cresap, West Virginia. It has also been used by Merichem, the
largest U~nitedl States producer of cresylie ntr'ids.
145
Distillation Purification
After extraction by any of the above methods, the phenols stream can be
separated into phenol and individual cresol isomers and xylenols. A typical
distillation scheme is shown in Figure 51.
The extracted phenols are fractionated under vacuum batchwise in one or two
columns, if the capacity of the plant is small, or continuously in a series of
coluns, if the capacity of the plant is large.
In the first colun, phenol and o-cresol are distilled overhead together with
the low-boilers. This overhead product is fractionated further in a second
column. The second colunn overhead product consists of aqueous low-boilers,
the side-stream fraction is pure phenol, and the bottoms product is an
o-cresol fraction which is supplied to an intermediate storage tank. The
first colufn bottoms product consisting of m- and p-cresol and xylenol is fed
to a second intermediate tank. In a third column, o-cresol, m- and p-cresol
and xylenols are alternately treated batchwise, depending on the specification
of the required product.
Modern cresylic acid processors have developed their own proprietary tech-
niques both for producing isomers other than o-cresol, and for remving cer-
tain impurities to meet specific consumer needs. This processing technology
is neither public knowledge nor offered for license. It would be difficult %
for a new producer to obtain enough processing knowledge-to be able to supply%-
specification grade products to all consumers on startup.
COLUM. BATCH COLUM
AQUEOUS LOW BOILERS
LOW SOILERS
_O-CAROL
"uot c soto
O-CESO"
REFINED
FEEDPHENoOL CONTINOUG YL ML
MP-CftSOL r
PHNENOL, O.- RMSO COLUMN "INTEMEIATE RlUNS
m ~O-C
E SOL
-
MYL.ENOL.
~RE SIOUIE
MW-CBffSft
FIGURE 51
FRACTIONATION OF PHENOLS
146
.I
*L~j N%.N~ . .. d5. I*( .. '-. .%
Great Plains may have a feedstock cost advantage over other natural cresylic
acid producers. In going from cresylic acid to pure cresols, however, Great
Plains would have no advantage, other than econonm of scale, over any other
producer. Thus, if an intermediate marketing arrangement could be worked out
with a producer or producers having existing surplus refining capacity, this
would result in the best overall production economics. New capital investment
would not be required. Merichem in Houston may have enough capacity to
process all the Great Plains material by rearranging feedstocks and flows in
their plant.
BENZEE
Sulfolane Extraction
Two basically different process routes are available to recover benzene from
the naphtha stream. The first is solvent extraction. The process chosen for
evaluation is the Sulfolane process offered by UOP, Inc. A schematic flow
sheet is given in Figure 52.
The process recovers a mixture of BTX aromatic liquids. As typically applied,

the sulfolane separation of the aromatics from a hydrocarbon mixture is ac-
coeplished partly by liquid-liquid extraction and partly by extractive distil-
lation. It is distinguished both from conventional liquid-liquid extraction
and from earlier commercial processes using only extractive distillation.
Fresh feed enters the extractor and flows upward, counter-current to a stream
of lean solvent. As the feed flows through the extractor, aromatics are
selectively dissolved In the solvent, and raffinate of very low aromatics con-
tent is withdrawn from the top of the extractor.
Rich solvent from the extractor enters the extractive stripper, in which par-
tial stripping of hydrocarbon from the rich solvent takes place. The non-
aromatic components having volatilities higher than that of benzene are essen-
tially completely stripped from the solvent and removed in the overhead
stream. This stream is returned to the extractor as reflux for recovery of
aromat ics.
The bottoms stream from the extractive stripper consists of solvent and
aromatic conponents, substantially free of non-aromatics. It enters the
recovery column, in which the aromatic product is separated from the solvent
stream. Because of the large difference in boiling point between sulfolane
and the heaviest desired aromatic product, this separation is acconfplished
readily. Loan solvent from the column bottom is returned to the extractor.
Raffinate from the extractor is contacted with water to remnve dissolved sul-
folane, and the rich water is returned to the extract recovery column as
stripping steam generated via exchange with the hot circulating solvent in the
water stripper reboiler. Dissolved sulfolane accunulates in the water strip-
per bottom and is punfped buck to the recovery column.
Suifol.,tre h~is ben applied to the recovery of petrochemical grade aromatics

frun 'oke )ov#-n ndplthas. The naphtha strewn would have to be hydrotreated to
stabi liz! it before fee ling to the sulfolane extraction unit. The UOP LT
Unibor process (Figure 51) ;vas asstunvd for this purpose.
147
000
0 0
WOF
It-
LPx
148
The UOP LT Unibon process is a two stage hydrotreating process employing
suitable catalysts to allcPN the selective hydrogenation of a wide range of di-
olefins, rnono-olefins, sulfur and nitrogen contaminants, without the formation
of troublesome gum, polymer and coke on heat exchangers, heater tubes, and
catalyst beds. Hydrocracking of feed to light ends is minimal and is confined
to those fragments created by the fracturing of the sulfur and nitrogen con-
taining compounds.
Two LT Unibon unita have been built which operate on coke oven light oils.
Hydrodealky lat ion
A nunber of comnercial processes are available for the hydrodealkylation of

alkyl benzenes to benzene. Both catalytic and non-catalytic processes are
available. These processes are normally used to convert toluene to benzene
whenever benzene is in short supply. The process selected for evaluation is
the Litol process offered by Houdry (Figure 54).
The Litol process was developed for the purification of coke oven light oils.
Benzene yield is significantly greater than benzene present in feed because of
dealkylation of heavier aromatics to benzene, and dehydrogenation of naph-
thenes.
Crude light-oil, together with a hydrogen stream, is heated at above atMs-

pheric pressure and passed over a catalyst. The effluent is cooled by heat
exchange and passed through a gas/liquid separator. The light gases are
recycled. The liquid is stabilized to remove hydrogen sulfide and light
gases. Conventional distillation can then separate the remaining liquid into
benzene, toluene and xylene. The toluene and xylene fractions may be recycled
through the Litol reactors until completely converted, with a corresponding
increase in benzene yield.
BEEINE + PHENOL
As an extension of the [DA thermal hydrodealkylation process to make benzene,

HRI Inc. has developed the Dynaphen process to convert mixed alkylphenols in
coal liquids to phenol and benzene.
The Dynaphen reactor primarily dealkylates cresylic acid molecules, although

some dehydroxylation occurs as well. Dehydroxylation products (toluene and
xylene), plus the unconverted alkylphenols are recycled to the Dynaphen reac-
tor so that the major reactor products are phenol and benzene.
Figure 55 illustrates the basic flow scheme of the Dynaphen Process. The
alkylphenolic feed to the process is mixed with hydrogen and sent to the
Dynaphen preheater and reactor. The reactor effluent is cooled by exchange
with reactor feed (and other process streams) for heat recovery, and is
separated into vapor- and liquid-phase components. Some light gas is produced
in this process, mostly methane and some carbon monoxide.
The liquid reactor effluent may be sent through a stabilizer and a clay tower
before distillation. Benzene and phenol are recovered in two series-connected
149
Rerun Reactor Healer Reactor Separator, SI!.Rper
Column
Pyrolysis Heavy Hydrogenated

Naphtha Hydrocarbons FIGURE 53 Product
LT UNIBON HYDROTREATING PROCESS
C. TOC,
NAC C.RA
PROCSSIN
n STNA
4-7A PCAR-OCEElactofIi
11101nK TOLUEN
Ce RCYCL GASCv RODUT PRDU.
cutd OIN SEPARATO

NFIFER
IdVAPORIZER LITOL AN OTHERAU

cf.AEACIORS HEAT11CHANGE SISMUER
PRAIRACTINATION CIUZENE TOLUENE
TowER TOWER
FIGURE 54
LITOL HYDRODEALKYLATION PROCESS
150
distillation towers. Unconverted aikylphenols are recycled to the Dynaphen
reactor. Dehydroxylation by-products, prinmrily toluene and xylene, are also
recycled, eventually producing benzene. Thus, the net products of the
Dynaphen system are benzene and phenol.
__2_TO__FUEL___GAS__BENZENE
PLANT
fOULA
N.XILPHEMOLS
FEED
EEFFLUENT
FE
DYKAHENDNAPHNEN PROCESS
T 151NO
PH"SSIfG HOOtMaICS
Capital and operating cost data from available sources were updated to 1986
costs and prices. A complete economic analysis was then carried out for
several different product slates, process configurations, and capital cost es-
timates. It must be emphasized that the available cost data are far from
investment-grade. The data are from preliminary feasibility studies which
were not based on plants which had actually been built and operated in the
recent past.
WOONGMIC PARANUE'
Ground rules for performing the economic runs included the following:
1. Basis is 100 percent equity for capital investment.
2. Plant costs are given in 1986 dollars.
3. Construction starts in 1988 and the plant is onstream in 1990.
4. No land purchase necessary.
5. Onstream factor is assumed to be 100 percent.
6. Operating and capital costs, and product prices are assuned to esca-
late according to the Data Resources Inc. United States GNP deflator
forecast.
7. Plant life is 20 years.
8. After calculating depreciation allowances and taxes, the year-by-

year cash flows are then de-escalated to give net present values and
rates of return on a constant dollar basis. This procedure properly
accounts for the devaluation of depreciation deductions by infla-
t ion.
9. Federal taxes correspond to the Tax Reform Act of 1986.
10. At the time of beginning this study, the North Dakota tax structure
allmved federal taxes to be deductible. This is likely to change in
the future, but for this study it was necessary to assume no change.
North Dakota's tax structure contains some complex provisions for special
taxes on coal convr-rsion facilities. The "coal conversion facilities
privilege tax" is imposed on the operator of a coal conversion facility for
the privilege of producing electricity or other products from coal conversion
plants. The coal conversion tax is in lieu of property taxes on the plant it-
self. For coal gasification plants, the tax amounts to 15 cents per
1,000 cubic feet of gas produced for sale, or else 2.5 percent of gross
receipts, whichever is greater. The definition of gross receipts excludes any
financial assistance from the federal govcrnment.
152
A five-year exenption fron the tax may be allowed for any facilities con-
structed after July 1, 1985. This exel)tion is discretionary, and is granted
in consideration of new jobs created by the facility. Privilege tax revenues
are distributed 65 percent to the state and 35 percent to the county. The
county may grant or deny a five-year exeryption to their own portion of the
tax. If an exemption is granted by the county the facility owner can then
apply to the state for a similar exenption.
In this analysis, it was assumed that the new capital investment required for
by-products upgrading would be a free-standing project. It would buy already-
converted liquids from Great Plains and upgrade them for sale. As such, it
would not be a coal conversion plant and would not be subject to the coal con-
version privilege tax. Increased revenues flowing to the Great Plains plant
under this arrangement might be subject to the 2.5 percent gross receipts tax.
However, if the gas were being sold at current conditions, the rate of
15 cents per thousand cubic feet on the gas product would be much greater than
2.5 percent of total revenues. Therefore, there would be effectively no addi-
tional privilege tax to be paid as a result of added by-product revenues.
Although it is likely that a by-products plant would qualify for a five-year

exenption from both ad valorem taxes and income taxes, such exenptions are
discretionary, and this analysis did not assume that such an exenption would
be available. For the purposes of this analysis, where only order-of-
.agnitude estimates are available for capital and operating costs, it was
found that the tax effect would be small in conparison to other uncertainties.
In order to conpute variable operating costs, it was necessary to assign a

value to the three by-product streams coming from the plant. Because these
streams can always be burned in the plant (as in current practice), they rrust
be assigned a value equivalent to the cost of providing replacemnt fuels.
Economic analyses were run by assigning three different values to this liquid
feedstock transfer price. Liquids transfer prices of $1, $2 and $3 per mil-
lion BTU were used to bracket the possible future value of these liquids as
plant fuel. The naximum liquids transfer prices would be that required to
purchase replacement residual fuel oil fro' outside sources. At early 1987
prices, this would be less than $3 per rMBTU. If it proves practical to fire
the plant boilers and other systems with lignite fines, then the mininrum liq-
uids transfer price would be equivalent to the netback now received by sending
the fines to the adjoining electric power plant (estimated to be about $0.58
per NTBTU). This is exclusive of the capital investment required to be able
to fire lignite fines.
The other major unknown in each case, besides the transfer cost for the raw
by-product streams, is the capital cost for the upgrading facilities. There
are no reliable data available. No coal tar refining plants have been built
for many years, and no up-to-date cost estimates were found. A numter of the
processes which might be used have never been demon:strated on a similar
feedstock, and/or have not been built in the size range which would be neces-
sary. This places an extrenely high neasure of uncertainty on the capital
cost estimates. To partially account for this uncertainty, each cast was run
a second time, with the capital cost arbitrarily increased by 100 percent. It
should be noted that it is possible that actual costs wolId be less than the
base case. Verbal cmniiunication with ANG personnel ind ioat es that ti,h 'ap itatl
cost for a cresylic acids process coti ld be -rnt
-0 less than the I ) P;vs,
1II-)
.53
Pe ,.d ¢ ., '' '.2 "'

-. '.' . , ,.' .''2'_,,-'d,, '%L~.
i" o -' -,-'m % V.o
%,,..- '..- '. 2 ..-. ,. -,. .. , '- .. ,
estimates due to utilities and auxiliaries already in place and lesser factors
for overhead, administration and marketing.
E. PASO PiODUCrs CASE
The first case analyzed was based on work carried by El Paso Products Company
for a comnercial coal gasification project in New Mexico. This was a paper
study only, the plant was never built. The El Paso by-products flow sheet is
shown in Figure 56. It is a total utilization scheme, converting all the liq-
uid by-product schemes to saleable products. Flow rates have been adjusted to
correspond to the quantities produced at the Great Plains plant. The product
slate resulting from this configuration is given in
Table 63. Prices are early 1987 prices. r
Capital investment costs corresponding to Figure 56 are summarized in

Table 64. Operating costs are given in Table 65.
The El Paso cost data are considered to be especially applicable because the
plant was based on Lurgi gasifiers running on New Mexico subbituminous coal.
The liquid by-products from such a facility would more nearly resemble the
Great Plains lignite tar liquids than would liquids from coke oven tars (the
liquids for which most tar refining facilities have been designed). %
Results of the economic analysis are given in Figure 57. Discounted cash flow
rate of return varied from 14.3 percent at a liquids transfer cost of
$1.00 per million BTU to 2.6 percent at a transfer cost of $3.00 per million
BTU. Increasing the capital investment by 100 percent dropped the rate of
return by over seven percent.
MWlFIN EL PASO COSE
The El Paso processing scheme includes a hydrodealkylation unit to convert the

Rectisol naphtha and tar oil naphtha to benzene and naphthalene. Co-product
methane is returned to the gasification plant product stream for credit.
Hydrogen for the dealkylation is obtained by cryogenic separation of a
sidestream of synthesis gas from the gasification plant. This scheme was
modified by changing the hydrogen generation step.
Instead of the cryogenic separation it was assumed that hydrogen would be ob-
taimed over-the-fence fron a pressure-swing adsorption unit located at the
gasification plant (Figure 58). k menufactured hydrogen cost of $1.70 per
thousand cubic feet was provided by ANG Coal Gasification Company. This
figurf is based on a value of $3.00 per million BTU for the synthesis gas, and
return of the carbon monoxide to the process. The product slate, capital in-
vestmTnt, and operating costs for this case are given in Tables 66, 67 and 68,
respl ct iv,' I y.
Results of tho r'conomf. ;inalysis are plotted in Figure 59. The rate of return
wa' irnprovtd consid'rnbly, with a mnxirrun return of 20.2 percent for a liquids
transfer cost of $1.00 per million BTU, dropping to 10.7 percent at $3 per
million B111. Doubling tht capital cost had about the same effect as before.
154
"
",,"
.. ,,r
" '.. ",,"
,-"" , ° '"+.,',,, "."
-" ","
-- t"
" .'..
",. .. ' '. '-." ,'- .' - ,.. - ' j . J' " '-.' . ," " "-"". - ". j ,q. *. . " .,,"
WATER A OILI 7 __
ELARTO
B1PODCT
PASO SCHEM
(ADJUSTED~~~~
FOTRATPANSROS
320
PIT
4RC 8
PICC
20DE EL ICHA
-JTSLSYGS6. I0L N ULII W
owTRAI$ PTHCRBNBAKP.9
.. k 15I IPR~ HEO 0
=VPEO 7 VOCR4LCAISI
UE
ELPS0YRDCSSHM
35
25230
20ISTANFR-U~ SMT
FIUE0
ELPSOCS
(TTLPRD.S CEE
is-5
TABLE 63
K
BLO U
M EL AID CAME
Production, Price, Revenue

Product Mil Pounds Per Year S/lb $/Year
Benzene 85.4 0.157 13.41

Nph tha Iene 17.7 0.143 2.53
Creosote 47.1 0.010 4.24
Electrode Pitch $4.4 0.1041 5.87
Carbon B lack Feed 35.9 0.023 1.04
Natural Phenol 30.2 0.213 6.43
Ref Ined Crosy Ii, 27.5 0.383 10.53
44 .05
T44LI b4
QWI1M. inw M I P~ma
Pr'K-.ea Iimz? 1306tomr
Ig- 6. (010f
I *' e .00. WhifD
NO-s I-tumm, .n I I flf
Tor 14, -. YW. 011040.
Total Gmery Linds PSin elf
* . -q ~ oft
T4041 I~u~ItS. P64n0 4101t
1,~ .~~*a94 ,,
M.- S~II~'n
TABLE 65
COM FR ELhD
iOMMIN . CAW
(1986 Dollars Per Year)
Fixed Operating Costs

Labor and Supervision 2,229,000
Operating Supplies 455,000
Plant Overhead 335,000
Laboratory 550,000
Maintenance 2,442,000
Insurance and Taxes 1,462,000
Royalties 534,000
Total Findm C t 8,007,000
Corporate Overhead, R&D, Marketing 4,0S6,000
Toal Imd * Omsg'hind C it 12,062,000
Variable Operating Costs

Ut Iliot .aS 1,366,000
Chemivals/atslysts 260,000
Liquids Transfer Cost a $1JM 75446000
Total Variable Coit * gMM V,194,000
Total Varible Cot 1 62AM 16,742,000
Total Vlable Cas 4 3 24,230,000

MOOIFInEMD22hm
EL ASiBPROUTSSCEM
1LN2FR01
0 T "gets-4
EL PASO CASSE2
rv@ADJIa O GRT hAN O
N W ~ s f okw I
PQS NVIA
a S~ j~~ d~2~:..~:& .
-- ,W U" 'f R 'E "W.NJ "
TABLE 66
FN= VIATS M EDIFINE) m PASO CASS
Production, Price, Revenue,

Product AIN Pounds Per Year S/lb $/Year
Benzene 85.4 0.157 13.41

Naphthalene 17.7 0.143 2.53
Creosote 47.1 0.090 4.24
Electrode Pitch 54.4 0.108 5.87
Carbon Black Feed 35.9 0.029 1.04
Refined Cresylics 27.5 0.383 10.53
Methane 4.0 0.343 1.37
45.42
TABLE 67
CAPITAL INY3WIW -M MIVIW UL PAII) CAMl
Process Unit 1986 Dollars ..
Dehydrat ion 352,000

Distil I gatin 2,776,000
Hydrodeai ky lat io 13,500,000
Residual (i and pitch 7,012,000
Tar Acid Processing 16,369,000
Total Battery Llnta Plant 40,009,000
Offslites 0 504b 20,005,000

Subtotal 60,014,000
Cfontingency t0 25% 15 ,004,000
Total Deprseiable Pllait Cost 75,018,000
Pr.. Pruoductton %t.rtup k nperse 3.,287,000

Working Capital 9 .4.73, 000 "-a.
Total Caltai lnstimmt 87,77,000
I .)p. ."~
"%"
TABLE 68
GMATIW3 WINS IM R. PABD CNIX

Fixed Operating Coats

Laboratory 550,000
Mai ntenance 2,442,000
Royalties 534,000
Ttal Find Cat 8,007,000
Corporate Overhead, R&D, Marketing 4,055,000
Total Fland + Oviebad Cost 12,062,000

Uti litiles 1,366,000
Chemicals/catalysts 210,000
Hydrogen Transfer Cost a 1.70 2,716,000
Liquids Transfer Cost Q $1/~Ih 7,648,000
Total Variable Cat * $190M 11,910,000
Total Variable Cost * $2A J 19,451,000
Total Variable Cat * $3/1J 27,006,000

After observing the major effect that hydrogen costs could exert on overall
economics, the 91 Paso scheme was modified further by eliminating the
hydrodealkylation unit. The naphtha stream were simply returned to plant
fuel and received a BlU credit equal to the liquids transfer charge
(Figure 60).
The product slate, capital investment, and operating costs for this case are
given In Tables 69, 70 and 71, respectively. Results of the economic analysis
are graphed in Figure 61. The maximxun discounted cash flow rate of return
dropped sharply, to 10.8 percent. This result suggests that the benzene
product itream is of major irnportanep to the economics of by-product refining.
UMTLIC SIEi CAS
It had been assumd that the tar acids could be the moat valuabl, constituents
in the by-products. Becaune the Great Plains plant would have the option of
continuing t) burn any of the liquid by-products instead of attempting to
upgrade them, the El Paso flow schenw was modified to process only the tar
tcids. 4ll other liquids were returned to plant fut.l. A co-product of the
phenols extraction pro.ess would be an aronatic raffinate strewn which could
be sold for rr than plant fuel value (Figure 62).
The product slate, capital investmnt, and operating coqts for this (ase are
given in Table 72, 73 and 74 respectively. The re-sults of the economic
analysii are shown in Figure 63. The results were not encouraging, with a
nruxinmaia rate of return of 12.8 percent at a liqui'ta transfer (oat of $1.00 per
million BiIJ, and only 5.0 perent for a liquids transfer c-ost of $3.00 per
million B'). Doubling the ostifmuted plant vost dropped the rate of return to
zero st a Iiquild trantsfr ('ost of $3.00 per million IT.
- 1 (W. CMI
r %jet) Ify the f low oho.wnF. F'rther, and v I aintnat t the .ooqt of tar tol 114.ti-
Sati on, it vano viis teies i ed in whic i al) the'- trutlv: ph.'no I it re',v Wa%
pIro'essed. Al ither itreanw ,ont tinuetl I )4e t a Ii ed iq plant fuel
(Fiur. 4 .
Theo pluvt 'pi tta i nviostepnt , Find topCritting itunl-e for thaso f-ocs.- tire
flt,
giv et in Tsitol,. N,5, 71i atnd 77, rptpoectav,.ly. Resultq of the ,cwova ni" si€n lyot 4
to r,- shown ir FtVI lir,. bi . Net u r or t I nvo s trvu'nt wa , favorable- than for the.
tro-viou% ,'.,., in whic.h phenols were nttroicted friwn t))th the crude, phenol
ntr.atn sndi the tar oil 9 trewn. The' rvurmr i i j it-,iurit.d co.a h fl.m rit#, (if
retUrn w II .11pi nrt for a I iq ris t rtinifr -()4t of $1 .000 per tll II Itfi l1,
fk*a'reat Ing t,) 4.7 perceo t ,1 $3.00 p.r tua I ' in Hr . Increa. ing the. tpi . tal
coqt by 100 percent re,qilt,.d irt a negative rat., -it return when usang the
higifi'v t liqluift- trI nsfo-r t'9 t. -
.%
*5,
*5
Ii~ I
* . . . .
NO-YROEAEAAO 1WAR
IauSE PTON PLWS PLANT FL&)F
11t MAUH 2c.2T

7ho
*MSp
ta C6 A IdLCAFID3
0, I NloW3 of p6 6Po4
SID'
0C ~ 2~' 3.
Smm SA I . 9 * S.. *9 *9oS

TABLE 69
PRGM IMN01-11MRGMA
SLAE EL P'ASO CASK

Product M~ Pounds Per Year $/lb $/Year
Creosote 47.1 0.090 4.24

Electrode Pitch 54.4 0.108 5.87
Carbon Black Feed 35.9 0.029 1.04
Refined Cresylics 27.5 0.383 10.53
28.11
TABLE 70
CAPITAL llVUIIF VU MGM- IlO=&=

IL PS CURE
Process Unit 1986 Dollars
Dehydrat ion 352,000 4
Distillation 2,776,000
Residual Oil and pitch 7,012,000
Total Battery Limits Plant 26,479,000
Offsites a 50%1,3,0
Subtotal 39,718,000
Contingency a~ 25% 9,929,000
Total Dupreclablo Plant Cost 49,648,000

Pre -Production Startup Expense 1,966,000
Working Capital 6,504,000
Total Capital Invostamot 58.138,00t)
163
A.
TABLE 71
OUMF IN1i WISTS FM NON-lT L PASO

Hh *3 ..ASG

Laboratory 440,000
Maintenance 1,954,000
Royalties 427, 000
Total FId Cost 6,406,000
Total Fined + Overhead Cost 9,649,000

Utilities, Chenicals 1,152,000
Liquids Transfer Cost $1/MMTU 5,470,000
Total Variable Cost 0 $1/'f 7,829,000
Total Variable Cost 0 $2/0M 13,298,000
Total Variable Cost 0 S /3il 18,769,000
164
PLOW LUO"iWas TO PLSANTFUEL 62.8 a
FIIRELI
02CAPW 4s
IIAAO .
AE
S.*A ION
CREIVCTA
ACD CIDM
AM
TANO NL 2UI.)
IhINSIL
TABLE 72
PUUJY SLATE FOR (DRMLIC ACIDS CASX

Product MM Pounds Per Year O/lb $/Year
Phenol 34.8 0.213 7.41

Cresols 32.5 0.383 12.45
Mixed Xylenol3 12.4 0.503 6.45
Aroi tic Raffinate 84.2 0.060 5.05
31.36
TABLE 73
CAPITAL IWvWWO KR SmiLCID AS

Cm
Process Uni 1966 DIlIars
Deihydrat i (m 352,000
Di1st illationi 2.776,000
Rib-%idui Oil and piteh 7,012,000
Tiar keld Proves~Ang 22,424,000
Phenol Fratctionatio 2130,0
Total Battery Liddis Plat 34.864.000
Stahltale >2.296.000OJ
towit ingfl' 0 254k0U 00
TOWd DrWcabl* plea Cost~t
Pro- Protjttioei Startlup I.mpen4*. .Ijt

Wwkng t'alital b, 60 O~w
Ai
TABLE 74
G TE3r1116; oom PM CIMLIC A S M


Laboratory 550,000
Mai ntenance 1,942,000
Roys Itaes 376,000
Tota I Fim Corot 6,957,000
Corporate Overhvefm, R&l, arketing ___500U
Total Flod * Oauba"d Cmit 9.811,000

tIhut s ,016.000
Liquids Trunqfer ( t 0 $1 1Tl 4L0,0000
Total Variable Coat 6 !l2lR1 .. 0oo
Total Variablo (~ma 41 11iW,340(

Totl Vrt411 4t 0 10111M 3.0 QU
40I
POCTISOI. NAPHTHA SO NANN TO PLANT F~L
OGSIIR9
R 64"TN
~~IG
TANA
E3TA5 IL 2
NETA G0FCA
PIC
TA CD a
NE PHNL9
20T PTHN
TA A 3PHNO 4.8
CRSTILLC ACIDOL
SCHEM
PHEN(S ONY SCEEI ONLY)
36168
TABLE 75
P2UJLr SLATE FOR IEdOL ONLY CURS

Product MMd Pounds Per Year s/lb $/Year
Phenol 30.2 0.213 6.43

Cresols 21.0 0.383 8.04
Xylenols 6.5 0.503 3.27
17.74
TABLE 76
CAPITAL INYRSI1~II OR PIUMMOS ONLY CASE
Cresol Distiilaticat 2,300,000

Xylenoi Distillation 2,300,000
Offsites C 50% 10,985,00

Subtotal 31,454,000
Contingency Q 25%6 7,863,000
Total D~rociable Plaint Cost 391-.17,000
Pre*-Production Startup Expense 1,573,000

WorkIing Capital 3,840,000
?east CAltaI Investment 44,730,000
ItJ.:
TA&Lt - 7
f11366 Ibollmr% For Veer I
Fixd Operating Costs

Labor and Siqpervisirm l.4fi.5OD
Operating supplies 362.000
Plant Overhead 264,00o
Laborastory 660,0,00
Mai ntenance 1,J91 90
Insurance and Taxes 779.000
Roya It I i 2o16,000
Ttal Fix" Cost 4 ,949,000
Total Find * Oweihai Cost 6,610,000
Variable Operating Coats

Utilities 760,000
Liquids Transfer Cost 0 $I/Th 1,660,000
Total Variable Cost 6 $i/i'ii 2,843,000
Total Variable Cost 0 $2/NM 4,703,000
Total Variable Cost 4 $3j1U@ 6,563,000
170
MO o %oIh Ffi t 490 1 'ad If I I' at I-W~ 440."1 41 -a
, 1'111 J to II.. of
.9.. ' """.
fIg1 V 1'i 46 is 'F1111 I~ IY
*1t - **I a* ft* FI ~. t pas., 41-" av V. .1 t IHV~9 I* .fo t rot 114-01
8h. , 4~*
qsr Iph rt' 01 l t Olt.. 'M '1011* i
us fe'ytam 1oh, 41r .
-. 1111vtV *.-al 1- % .4~4 '140fy ItI.-oo- ! lot iof It%- iti p t o , i, r too
, *I f 'I 1 1 " 1 , , ~ fl I
f3 Ff -o* )1' 0 I ,W.. . o. tf . IIFJ o-%la% rt I of r- rn*
I 4gYjvipoi-foi )f , i#
19, *'jt
a i 'at
to -t i priOi 4"ofi In t he Grad~y M , -ame -ind I t#
Pol%, -no roc.t tt'IIs t he F I Psc vrmae, andI suggpo -
,in +-.rihly rrwuire det qi I in
t i wim 'f h.
ItI *'j~ewnitl t - r-, of -- ino ide roil) I y mpwro- '.ngi ne. r I ng dies I gCh efftort . Fur
t ho.r vwupe rI %tin re.-eeutc 1t1hat t1he nojor di mgrepancy I q I n the *4t Inuted i'oat of
th.. t .4r o i 'd -- %Itrsat li lint . The Grady 'ii uiu. u.se#s ft vc ust it, 4-xt reet In flpro)-
"e..kj r. aoie rest the' 1,.1 P~oe%.a case in bma.'aI 'n Lurgi '% Phenoraff in sol vent ex -
t ut in pr 10 41toOl. r,) i I I i t rate the #-ffo-1*'if this unit tn the eecnzn, the
(Jrady iAM; aaeta was ro-runi by subst itutI ig El Paso's coat for the tar acid ex -
trittt ion unit in plavo. of (ir.'iy 's cost est imate.
The priodut 'ilate, capital coat and operating coat for this case are given in
Tabieps 81, 82 and 83, respectively. Figure 68 shows the results of the
ecoorltifc analysiis. Ths- miaximn return on investment was calculated to be
15.4 percent, very sirmiuur to the 14.3 percent obtained in the El Paso case.
This return drops to 7.8 perce~nt at a $3.00 per maillion B~lJ transfer cost for
the liquids. The high v &Itje declines by about seven percent if the capital
cost for the facility is douaied.
-. The results of thii case clearly illustrate the crucial importance of a reli-
able cost estimte for the tar acids extraction unit. While the Lurgi es-
tiiute provided to the El Paso project riuy have been reasonably valid at that
tine, no Phenoraffin units have been built since 1963. There rmy be manny
questions about such old cost estimates being accurate, especially in today's
envir(mnental setting. The Grady estimaite is not convincing in its depth of
detail. No reliable cost estimte is available.
CRAM NCN-IUDR3RATING CASE
As was done with the El Paso scheme~, the Grady/ANG flow sheet was rearranged
to elfiminate the hydrotreating step (Figure 69). The product slate, capital
costs and operating costs are given in Tables 84, 85 and 86, respectively.
The economic results are shown in Figure 70. The mximum rate of return (at a
liquids transfer cost of $1.00 per million BTU) increases slightly to
32.0 percent frtxn 28.9 percent. This Is In contrast to the El Paso case,
where the return dropped sharply when the hydrogenation step was eliminated.
171
ai sil cdllo,"tlot PU2427
GRADY/ANG
SCEM ~ ~
WiHHDOTET pa
IUR 66
35f
30m7 ILNt
F tot@Olt2To
20Iof0a o
0 MA6
0NA 111,11311
TheID ToAftE PRICEN OIL 1A.MU
FIGUEI6
RAYAGRADY/ANHGYDOTEAIN
25172
TABLE 78
PR KATE IB GRMr/AMI CUE

Product W Pounds Per Year $/lb $/Year
Phenol 34.8 0.213 7.41

O-Cresol 9.8 0.503 4.93
M,P-Cresol 22.7 0.551 12.51
Xylenols 12.4 0.503 6.24
Lignite Tar 53.4 0.108 5.77
Gasoline 150.8 0.081 12.21
Solvent Oil 16.8 0.108 1.81
50.88
TABLE 79
C&PITAL INVIrM=IT FOR WAUDY/M CKR
Depitching Tower 3,250,000

Tar Distillation 2,200,000
Hydrot reat ing 6,790,000
Tar Acid Processing 31040,000
Phenol Fractionation 4,830,000
Offsites @ 35% 7,0409000

Subtotal 27,150,000
Engineering @ 20% 5,430,000
32,580,000
Contingency C 35% 11,400,000
Total Depreciable Plant Cost 43,980,000
Pre-Production Startup Expense 1,759,000

Working Capital 109300t000
Total Capital Investment 56,039,000
173
TABLE 80
OPA ~ ~ I GV3ANM CA5I

Fixed Operating Costs 5,604,000
Total Flxed + Overhead Cost 9,609,000

Utili ties 2,359,000
ChemicaIs/catalyats 938,000
Hydrogen Transfer Cost 814,000
Liquids Transfer Cost 0 $1/IWTU 6,248,000
Total Variable Cost 0 iA rU 10,359,000
Total Variable Cost 0 *j/iN'ii 16,607,000
Total Variable Cost 0 $311iJ 22,855,000
1A
5,
q
174
A -!
TABLE 81
F.M IM5 GIWM/. PAID CASM

Product MM Pounds Per Year S/ lb $/Year
Phenol 34.8 0.213 7.41

0-Cresol 9.8 0.503 4.93
M,P-Cresol 22.7 0.551 12.61
Xylenols 12.4 0.503 6.24
Lignite Tar 53.4 0.108 5.77
Gasoline 150.8 0.081 12.21
Solvent Oil 16.8 0.108 1.81
50.88
TABLE 82
COPITAL INVIUIUIT FOR GRADY/L PABO CASE

Hydrot reat ing 6,790,000
Offsites C 35% 14,631,000

Subtotal 56,434,000
Engineering 0 206 11,287,000
67,721,000
Contingency a 35% 23,702,000

U15
TABLE 8.3
CVATIM COUI= (BD B

GRWI AND CAll
Fixed Operating Costs 9,143,000
Corporate Overhead, R.&D, Marketing 4,055,000
Total VFind + Overhead Cost 13,198,000

Utilities 2,359,000
CheicaIs /cata lys ts 938,000
Hydrogen Transfer Cost 814,000
Liquids Transfer Cost a $1/hWUBJ 6,248,000
Total Variable Cost 6 WON=l 10,359,000
Total Variable Cost Q $2/ 'IU 16,607,000%
Total Variable Cost 0 $3/1M 22,855,000 m
35-
30-
25-7
20-
0
10
15
10
0 FIGURE 684
176
650RA OIL 106 PITCU.
FIGURE 6INTETN9VO
TANOI
ool 25DK I
AT PEO K4CT
OTE156r
0 44665 05
30-
.P%
15 -B'B J ' % % ~
TABLE 84
PJMM ULATR PR WAMIMIO NCN-EMONUATING CURE

Product NM1 Pounds Per Year S/lb S/ear
Phenol 34.8 0.213 7.41

0-Cresol 9.8 0.503 4.93
MP-Cresol 22.7 0.551 12.51
Xylenols 12.4 0.503 6.24
Lignite Tar 53.4 0.108 5.77
36.86
TABLE 85
CAPITAL INVKSWlP FOR (RADY/ANG

tUI-U!DTRATINO CASE

Offsites @ 35% 4,662,000

Subtotal 17,982,000
21,578,000
Contingency @ 35% 7,552,000

Total CapIital Investmnt 38,110,000
178
TABL.E 86
GIPTmN Cum MR GRW/Agmi NW-TIM CNN

(lV66 Dollars Per Year)
fixed Operating Costs 3,811,000
Total Flaud + Oiinrhead Coat 6,864,000

Utilities 1,769,000
Liquids Transfer Cost 0 $/WTU 3,330,000
Total Variable Coat 0 $i/MInn 6,037,000
Total Variable Cost 6 $2/1UM 9,367,000
Total Variable Coat 0 $3dMITU 12,697,000
179
iniL raw n-umm
The previous oase was noditled by again substituting EL Paso's estinmte of the
coOt for the tar acid extraction unit. The product slate, capital cost and
operating coat are given in Tables 87, 86 and 89, respectively.
Figure 71 illustrates the economic results. The maximum discounted cash flow
rate of return is 17.4 percent compared to the previously obtained value of
16.4 percent for the comparable hydrotreated case.
Although the Grady hydrotreated versus non-hydrotreated cases produced seem-

ingly equivocal results, the El Paso cases indicated a strong economie incen-
tive for producing benzene. There are a number of options for producing ben-
zene, of which four were Investigated (Figure 72). These four options are
defined as follows:
Benzene only - Naphtha stream to Litol process
BTX option - Naphtha stream to Unibon hydrotreating fol-

lowed by Sulfolane extraction
Benzene + Phenol - Naphtha and phenol streams to Dynaphen

process
Dynaphen II - Same as preceding case but add tar acids

from tar oil stream
The first case involves processing only the naphtha stream by hydrodealkyla-
tion to produce benzene. Houdry's Litol process was selected.
The product slate, capital cost and operating cost for this case are given in
Table 90. Results of the economic analysis are shown in Figure 73. The nax-
imum rate of return was calculated to be 23.3 percent at a liquids transfer
cost of $1.00 per million BTU, falling to 17.0 percent at $3.00 per million
BTU. Increasing the capital cost by 100 percent decreased the rate of return
from 23.3 to 13.6 percent.
BYTX ONLY CASE
An alternate method of treating the naphtha stream is to use a solvent extrac-

tion process. In this case a mixed benzene, toluene, xylene (BTX) product
stream would result. The raw naphtha stream would have to be subjected to
mild hydrotreatment before it would be acceptable to the solvent extraction
process.
This case is based on UOP's LT Unibon process for hydrotreating, and their
Sulfolane process for extraction. The product slate, capital cost and operat-
ing cost are given in Table 91. Economic results are seen in Figure 74. The
maxinmn rate of return was found to be 26.8 percent at a liquids transfer cost
of $1.00 per million BTU. This is essentially equivalent to the result ob-
tained for the Litol/benzene case. However, the assumed product revenues are
overly optimistic because prices for the pure components were used.
180
ProduactW Pounds Pe r Y-#.mar S
lbSYea r
Phemn 1 34.0 7.41

0-(rpsol 9.8 0. Oj 4.9J
N, P-C resol .2.7. )11..1
Xylenols 12.4 u. 5u. b.24
Lignite Tar 53.4 .10___
35-
25--
20-
10
5-.
0 1 4
10181
TABLE 88
CAPITAL INVrS7=Cr FOR GRADY/EL PASO

N"N-HI)RDTa~aIN CASE

Offsites @ 35% 11,446,000

Subtotal 44,150,000

52,980,000
Contingency @ 35% 18,5439000

lv
-R194 646 PRODUCTION OF JET FUELS FROM COAL DERIYED LIQUIDS 3/'3
VOLUME I MARKET ASSESSM.. (U) NORTH DAKOTA UNIV GRAND
FORKS ENERGY RESEARCH CENTER J E SINOR AUG 07
LAShhhhhh
UNCLASSIFIED AFUL-TR-67-2042-YOL-1 MI PR-FY1455-96-NSES? F/G 21/4 N
m700h00h010I
1111n W2. 2.5
L3
11111i~ W-0'It
1.25 Jill LA1-

3
1111121.6
MICROCOPY RESOLUTION TEST CHART
NATIONAL
BUREAUOF STANDARDS-1963-A
W % % ,.%~~* . .* *~ a - * * t*** $
.t ?o . tJ*rtr a.a' %-.
TABLE 89
OFPMtTIN1 WM IM FOR 3 /DY3PASO

N-EIUNGTIM CASS
Fixed Operating Costs 39811,000
Total Fined + Overhead Cost 6,884,000

Utilities 1,769,000
Liquids Transfer Cost @ $I/WBTU 3,330,000
Total Variable Cost 0 $/AMWMIi 6,037,000
Total Variable Cost 0 $2/IVMM 9,367,000
Total Variable Cost 4 $3/IMNTU 12,697,000
183
cI-
U. w
Z4 U. I
-j 0
0 1
0 1
z w
49 w Z .
-h z wi 0
0
ALN
lu N Z
z tU
z 0
IL
0
4c~
I.- z
0.O LU c 0
au 0
9-0 0
z ILUSW
A. z Z
z w
0 I0
P 0
z
z zI
0 0
P.
0L
ww
z z
I-z
x N N
z
l* 0
184
TABLE 90
xMriMC PA'M~r= IM DEMME WY Cne

(Processing Naphtha Only--Litol Hydrodealkylation)
Product
Benzene 0 7.34 lb/gal 32,000,000 lb/year
Revenue C $0.157/ib $5,024,000
Cqpital Cost
Depreciable Capital Cost $7,500,000
Working Capital 837,000
Pre-Production Startup Expense 300,000
$8,637,000
Operating Cost
Fixed Operating Cost $432,000
Hydrogen @ 1.70 380,000
Liquids transfer net cost @ $1/MMBTU 576,000
Total Operating Cost @ $1/MOM $1,388,000
Total Operating Cost C $2j1U $1,964,000
Total Operating Cost 0 $3/MITU $2,540,000
I 1l
Aida.,b
_V * *
TABLE 91
moawuc PA*.NKm2 POR in Chan

(Processing Naphtha Only--LT Unibon Plus Sulfolane)
Prodmet slate
Million Dollars Dol lars

Product Pounds/year Per Pound Per Year
Benzene 21.2 0.157 $3,328,000

Toluene 8.8 0.125 1,100,000
Xylene 1. 8 0.160 288,000
$4,716,000
Capital Cost
Working Capital 786,000
$7,066,000
Operating Cost
Hydrogen 190,000
Liquids Transfer Net Cost @ $1/F4STU -554,000
Total Operating cost 4 #ijivr $1,.147,000
Total Operating Cost a $2'flJ $1,701,000
Total Operating Cost 0 $3/jW1U $2,255,000

35.
I 30-
25-
S20-
15
10
04
UXWS TRANSFER PUCE ($1111111

ROMJE 73
LITOL BENZENE
(NAPHTHA OW"Y
I 30-
25-
S 20-
00
0 1 234
UOWDTAS M PRICE C$/AU
FIGURE 74
UNIBON & SULFOLANE
(NAPHTH4A ONLY)
187
Having denmnstrated the attractiveness of producing only benzene as a product
In a relatively simple processing ache=m, It was desirable to study the effect
of co-producing phenol frain the crude phenol stream. Because phenol is worth
even nwre than benzene, producing it as a co-product instead of converting it
to benzene seean logical.
RIl has a process still In the deveiopiment stage, Dynaphen, which co-produces
benzene and phenol fromn a mixed stream containing aikylated aromatics and
phenols.
Estimated economnic paramesters for a Dynaphen facility processing the Great

Plains naphtha and phenols streams are given in Table 92. With a simplistic
hssumption of 100 percent recovery, the product slate would be 35 million
pounds of benzene and 51, million pounds of phenol per year.
Results of the economic analysis are seen in Figure 75. The maximum rate of
return Is 35.9 percent at a liquids transfer value of $1.00 per million BTU.
At a transfer cost of $3.00 per million BTU, the rate of return drops to
22.5 percent. If the capital cost is then doubled, the return drops to
13.6 percent.
Although this appears to be the most attractive of the schemsa investigated,

it depends on a suitable definition of capital costs. Since the process has
not yet been operated cotinrercially, there will always be uncertainty In the
cost estimate.
In order to provide the rnaxinuim econonWr of scale for the Dynaphen process,
along with the greatest cash flow, a second Dynaphen case was constructed in
which the phenolics from the tar oil stream were processed in addition to the
naphtha and crude phenol streams. The product slate, capital cost and operat-
ing cost are given in Table 93.
Figure 76 presents the economic results. A discounted cash flow rate of

return of 33.8 percent is achieved at a liquids transfer price of $1.00 per
million BTU. This falls to 21.0 percent at a liquids transfer value of
$3.00 per million BTU. This processing scherne, like the previous Dynaphen
case, is rather sensitive to capital costs, with the maxinum return dropping
from 33.8 percent to 21.5 percent if the capital cost is doubled.
It should be noted that no allowance was made for the cost of a tar oil dis-
tillation colun to cut out the phenolic streamn. This omission was justified
on the grounds that this case would probably be implemented only if jet fuel
were rnude fromn the tar oil and that the still would be provided as part of the
jet fuel manufacturing facilities.
188
mA.'u 9
TABLE 92
8XDIOIC PhiM 08 FRMM CAll

(Processing Naphtha Plus Phenol Stream)
Product Slate
Million Dollars Dollars

Product Pounds/Year Per Pound Per Year
Benzene 35.0 0.157 $5,495,000

Phenol 50.9 0.213 13,387,000
$18,882,000
Capital Cost
$14,660,000
Operating Cost
Fixed Operating Cost $1,888,000
Hydrogen 1,340,000
Liquids Transfer Net Cost C $1/hMBTU 2t414,000
Total Operating Cost 0 */iiBTiU $5,640,000
Total Operating Cost 0 $2/lMT $8,056,000
Total Operating Cost @ $3/i U $10,470,000
189
TABLE 93
RON=Cci ?hARUTES FR DINAPMDI I CASE

(Processing Naphtha + Phenol + Light Oil Streams)
Product Slate
Million Dollars Dollars

Product Pounds/Year Per Pound Per Year
Benzene 35.0 0.157 $59495,000

Phenol 72.9 0.213 19,000,000
$24,495,000
Capiital Cost
$19,734,000
Operating Cost
Hydrogen 2,010,000
Liquids Transfer Net Cost 4 $1/MMBTU 2,993,000
Total Operating Cost 0 $IMOJ $7,544,000
Total Operating Cost 0 $2/B1EJ $10,537,000
Total Operating Cost 0 $3/01 J $13,530,000
190
,mz S.wz--p ~ *' ~ -..-.- Vl

35 -
30-
25-
S20-
0
5 15-
10-
0 1 2 34
OWDS TRANSFER PRICE (S/TU)
FIGURE 75
DYNAPHEN BENZENE PLUS PHENOL
NAPHIH a PINOLS"
35-
I 30-
25-
3t 20-
0
* 15-
10-
0 1 2 4
UQtXDS TRANSFER PRICE ($/MNBTU)
FIGURE 76
DYNAPHEN BENZENE PLUS PHENOL
(NAPHTHA, PHENOLS, TAN OIL)
191 *
OWOa
O ' WOGUuIC RU TS
Summary pages from the computer printout for each run may be found in the Ap-
pendix. The major financial parameters of interest are the discounted cash
flow rate of return on investment, the net present value, and the total net
cash flow over the project life. These parameters are listed for the base
case (best estimate for capital cost) runs in Table 94. The calculation of
net present value depends upon the assumed value for threshold rate of return.
Because both eight percent and ten percent are often used for threshold rate
of return, the net present value was calculated both ways.
BUISITIVIT 70 ThAWUPOKATIGN (DST8
Four different cases showing attractive rates of return are based on the flow
schemes which were referred to as the Benzene case (Benzene), the BTX case
(BTX), the Dynaphen II case (Benzene + Phenol) and the Grady/ANG case
(Maltiproduct ).
The Benzene and BTX cases treat only the naphtha stream and produce essen-
tially a single product; the Benzene + Phenol treat the naphtha and phenol
streams and part of the tar oil stream to produce two products; and the Multi-
product case treats all by-product streams to produce a wide range of
products. It was of interest to see how these four different cases might vary
in their sensitivity to the cost of transporting products to market.
Sensitivity cases were run for each of these four process concepts by changing
the product transportation costs by plus or minus 100 percent. The starting
point in each case was a liquids transfer cost of $2 per million BTU. The
results of the sensitivity analysis are listed in Table 95 and plotted in
Figure 77. As noted in Table 95 a doubling of the product transportation cost
over the base case estimate leads to a decrease of three to four percentage
points in the discounted cash flow rate of return on investment. This con-
firms initial assumptions that product transportation is a significant factor
in marketing by-products from the Great Plains plant.
Figure 77 indicates that all four product slates are affected to about the
same degree.
SINSITIVITY 70 PwUCxr PRICES
The same four cases as above were also tested for sensitivity to product
prices. A sensitivity range of -25 percent to +50 percent was tested, using
the same base cases as before. Results are shown in Table 96 and Figure 78.
The effects were very similar in all cases, with the rate of return on invest-
ment increasing alnst in proportion to product prices. The multiple product
cases benefit somewhat rmre than the single product cases.
192
TABLE 94
SUARY OF ECCN(NIC RZSULTS

(Results in 1987 Dollars)
NPV at NPV at Net

DCFROR 8% DCFROR, 10% DCFROR, Cash FLow,
Case Percent $Million $Million $Million
El Paso, (multiproducts schem)

$1 Transfer Cost 14.3 50.1 29.8 219.7
$2 Transfer Cost 9.3 9.1 -4.3 123.3
$3 Transfer Cost 2.6 -31.8 -38.4 26.9
El Paso, mDdified IhWdrogen source (multiproducts)

$1 Transfer Cost 20.2 99.9 72.2 329.5
$2 Transfer Cost 15.8 59.0 38.1 233.2
$3 Transfer Cost 10.7 18.1 4.0 136.8
El Paso, non-hydrotreating, BTX to plant fuel

$1 Transfer Cost 10.8 12.4 3.1 91.9
$2 Transfer Cost 3.3 -17.3 -21.7 22.1
$3 Transfer Cost -12.7 -47.0 -46.4 -47.8
Grady/AlM, (mua I t lproducts scheme)

$1 Transfer Cost 28.9 122.8 95.3 351.3
$2 Transfer Cost 24.2 88.9 67.0 271.5
$3 Transfer Cost 18.9 55.0 38.8 191.7
Grady, using El Paso cost for tar acid extraction

$1 Transfer Cost 15.4 66.6 42.3 269.7
$2 Transfer Cost 11.9 32.7 14.0 189.9
$3 Transfer Cost 7.8 -1.2 -14.2 110.1
Grady, non-hydrotreating case

$1 Transfer Cost 32.0 99.1 77.7 275.9
$2 Transfer Cost 28.6 81.0 62.7 233.3
$3 Transfer Cost 24.9 62.9 47.6 190.8
Grady, non-hydrotreating, using 91 Paso tar acid unit

$1 Transfer Cost 17.4 67.3 45.9 245.0
$2 Transfer Cost 15.2 49.2 30.8 202.5
$3 Transfer Cost 12.8 31.1 15.8 160.0
Cresylic Acids Only, treating phenols plus tar oil strewn

$1 Transfer Cost 12.8 29.2 14.9 148.3
$2 Transfer Cost 9.3 7.3 -3.3 96.8
$3 Transfer Cost 5.0 -14.6 -21.5 45.4
193
".p
TABLE 94 (Continued)
i OF Bu (MxIC RRI5LTS
(Results in 1987 Dollars)
NPV at NPV at Net

DCFROR 8% DCFROR, 10% DCFROR, Cash FLow,
Case Percent $Million $Million $Million
Cresylic Acids Only, treating phenols stream only

$1 Transfer Cost 11.0 10.5 3.1 72.4
$2 Transfer Cost 8.1 0.4 -5.3 48.6
$3 Transfer Cost 4.7 -9.7 -13.7 24.9
Bew eme, (Litol treating naphtha stream only)

$1 Transfer Cost 23.3 13.0 9.8 39.9
$2 Transfer Cost 20.2 9.9 7.2 32.6
$3 Transfer Cost 16.9 6.8 4.6 25.2
Irll, (Unibon + Sulfolane extraction, naphtha only)

$1 Transfer Cost 26.8 13.9 10.7 40.2
$2 Transfer Cost 23.6 10.9 8.2 33.2
$3 Transfer Cost 20.0 7.9 5.7 26.1
Dynphen, (benzene + phenol) fran (naphtha + phenols) strem

$1 Transfer Cost 35.9 46.2 36.6 125.1
$2 Transfer Cost 29.6 33.1 25.7 94.3
$3 Transfer Cost 22.5 19.9 14.8 63.4
Dynsphen, treating naphtha, phenols, tar oil stream

$1 Transfer Cost 33.8 56.2 44.3 154.3
$2 Transfer Cost 27.8 39.9 30.8 116.0
$3 Transfer Cost 21.0 23.7 17.2 77.8
194
TABLE. 95
31 SITlV1T 70 P3WU"r TiANSFCUrATION OSTS

(Liquids Transfer Value Set at $2/Million BTU)
Net Present
Value at Total net
DCFROR, 10% DCFROR Cash Flow,
Case Percent $ Million $ Million
nzene
Low Transport Cost 23.1 9.63 39.5
Base Case 20.2 7.18 32.6
High Transport Cost 17.1 4.72 25.6

Base Case 23.6 8.16 33.2
Benzene + Pbenol
Base Case 27.8 30.8 116.0
Mult iproduct
Base Case 24.2 67.0 271.5
195
00
0 T
10-
-100%"BE 100%
CASE
RELATIVE PRODUCT TRANSPORTATION COST
MiGMR 77
SENSITIVITY TO PRODUCT TRANSPORTATION COSTS
35-
25-
S 20-
0
40
a 5
-5BAE*25% #50%
CASE
RELATIVE PRODUCT PRICE (S/MMBTU)
FIGUR 78
SENSITIVITY TO PRODUCT PRICES
19 b
TABLE 96
SWITIVITr 70 PNWic PRICE8

(Liquids Transfer Value Set at $2/Million BTU)
Net Present
Value at Total Net
DCFROR, 1096 DCFIK)R, Cash Flow,
Case Percent $ Million $ Million
High Product Prices 32.1 18.6 64.9

Base Case 20.2 7.2 32.6
Low Product Prices 12.6 1.5 16.6
rm
Base Case 23.6 8.2 33.2
Deinmem + Phenol
Base Case 27.8 30.8 116.0
aM
it iprodmet
Base Case 24.2 67.0 271.5
19
U
197
RISK ANALYSIS
The effects of uncertainties in cost estimates and in price forecasts can be

evaluated by the technique of risk analysis. A Monte Carlo type economic risk
analysis was carried out for the same four process configurations as were used
in the economic sensitivity runs. Triangular probability distribution curves
were jlefined for capital costs and for product prices. In these distribution
curves, the most likely value, defining the peak of the distribution, is the
value which was used in making the base case runs. The upper and lower limits
on the distribution were specified as -20 percent to +100 percent for the
capital cost and -25 percent to +75 percent for the product prices.
Five hundred sinulations were conducted for each evaluation at a liquids

transfer value of $2 per million BTU. This resulted in probability distribu-
tion curves for both the DCFRDR and the NPV. The results are sumnrrized in
Table 97 and in Figure 79 (DCFROR) and Figure 80 (NPV).
The expected values for DCFROR and NPV were changed very little by the risk
analyses, although the Dynaphen (Benzene + Phenol) case showed a gain of a
little over one percentage point in DCFROR.
This makes it even more clearly the case with the best rate of return.
Table 97 suggests that it Is also the case with the most risk (uncertainty)
associated with the estimated rate of return. The standard deviation about
the mean DCFROR is 4.4 percentage points for the Dynaphen case, conpared to a
standard deviation no larger than 3.6 percentage points for the other cases.
However, this is partly due siply to having a larger expected DCFROR. On a
relative basis, the Benzene case, with a DCFROR of 20.2 +/- 3.35, exhibits a
wider range of uncertainty.
TABLE 97
RESULTS OF ECONOMIC RISK ANALYSIS

(Liquids Transfer Value set at $2/Million BTU)
DCFROR, Percent NPV, $ Million

Case Expected Std. Dev. Expected Std. Dev.
Benzene 20.2 3.35 13.8 2.9
BTX 23.2 3.43 14.8 2.2
Benzene + Phenol 29.1 4.44 60.9 9.1
Maltiproduct 25.4 3.63 136.7 15.0
198
....' .-. r,.," '.'." .. ' '. , . .- ' ,' ."-,a" " . .". ' ' '.'.' .,, -. ,,, ,' ¢a,%' ,.,,,.,' . , '. . ' .'. ,- , , ,
I aaa C ImOO
o So SS.o S moaoo ma,
I. . . , a. .wfwOg
i.o, .
ob op IgI~
C.!I 0!eS.. al
. .masial
M! M ma,
a I O- IOI mamIO ainal
IWOOmI a ILIeamel
I
m maImemoo I4 3 In as IIIi
I =,.iW, II "W;: |I, b,
I
a- ".40.
Be.ee - - 6 ..
a44
ft
-m-a
oo.q-
a m.C!
am a
a1! ml.
oma
- ~a *,aaamaam' goaa, ~
S,- .oo.oo
ainflmflaa, U aim lam*.tm
Na mma eOmama a I Na aaOaI nmmO mmmO0
mI nfOI amOIamamaee, -I Imamaammamaamaaam

a. ................. a Im
a a aaIamaI~~l inmaaa
~-o. m am-NM I
MONSM.... , -m
,m n.-.,-
0 ft 0 0 0 v . M a o.a an IN
I ,,
a aaa4a
a;
U*NNNNI4~~M
W7
ai
all
n
rO IM,
aO . M Cnn~
a
-I
Ia .
--
.a
- -Na-I-N
I
O IIaINa INI-I
lm.ma.
D c. Ma.. . . ... o. . . .A.. . . . . tw.--

. . .
a *aaam in1aama
a4-aa a ""r4
m
a *m Ia mam lIa O I
ae aajC!Ctm mm l Ia aaammaami ma
a4am Ia ai1 Iaamm 4aa mmaa
I
199 SI a al i n m S S S *
U0+E
o . 5* 0.oat
2 0e seene, e00e0,,
a ss OI ss m 0 -P1 1' 0 09 00 0 V) 0 I
I--
"i. !. t I . . . C.se.0 . C .
It-caswe vs IP , W%
40 s
02 4LC01, 0e e c
lCe
N., e ceec a, irn -sac acs.

C4 eccea A saca
WD a:::::::*as ea s.
4L as0s s c 0.0S cs s .
3-Wi e~e~ ese es .. fg a~c~ CCca
a~s
C e e~
ce Ut, e
ItCC
ce ea
e
, ces"Te"~e
: : " A "' C5 eC wz1
We n4 e w sa0. l5Cl5to5
se e 0 V; 0C;
to. s~eee caaa * t . se sses200e
aewc
9). Ir.
14~~~~~~~~~e
aa ~es , m ae a sCu
e as6aees
ec e c e P..,CC5
OUTLOOK 70 1995
The market outlook for individual products which might be produced from the
Great Plains liquid by-product streams was discussed in the "Market
Assessment" section for each product. Each product forms a part of, and is
keyed to, the state of the overall econon-'. The overall econom swings
through both short-term and long-term business cycles, which define generally
favorable and unfavorable periods for capital investment. In undertaking to
produce a basic coniodity, it is important to analyze how that connvdity will
be affected by general trends in the econorr. It is always desirable to try
to separate short-term effects from the longer-term outlook, over which any
capital investment would have to be repaid.
The period 1984 to 1986 saw a sharp world-wide decline in the prices of mst
non-fuel basic comnwdities. From the second quarter of 1984 to the third
quarter of 1986, comwdity prices, as measured by the International Monetary
Fund index of prices for 34 non-fuel conndities, fell by nearly one fourth in .3
terms of United States dollars and by ore than one third in terms of Special
Drawing Rights. In the third quarter of 1986 these indices were at their
lowest levels since the first halt of 1976, when prices were low in the wake
of the 1975 recession. It is difficult to make precise comparisons for longer .
term novements of comnodity prices in real terms, i.e. prices adjusted for in-
flation, because of the difficulties in obtaining consistent historical data.
However, it seems clear that real commodity prices in the third quarter of
1986 had fallen to a level not experienced since at least the 1930s. Some ex-
aryples include copper, down 37 percent since 1980, aluminum, down 35 percent
since 1980, iron ore down 20 percent, tin down 46 percent, zinc down
24 percent.
Because these price drops pre-date the fall in oil prices, they are obviously
not a consequence of the latter. They are a consequence of the fact that
major industries, world-wide, have over-invested in production capacity.
This, combined with a downturn in denand, has created the pricing problem.
Theoretically, the forces of supply and demand will produce prices which even-
tually cause the shutdown of "excess" capacity until the remaining producers
can operate at a profit. This has been slow to occur, however, because of
governmental and institutional incentives in many countries to keep producing
regardless of price levels.
As an example, it has been estimated that the world overcapacity in steel is

75 to 200 million metric tons per year, out of a total capacity of 570 million
tons in non-conmunist countries. Chase Econometrics states that only if the
entire United States steel industry shut down would demand equal supply in the
non-comunist world. For Great Plains, this creates a positive aspect, in
that supplies of by-product coal tar from the steel industry should continue
to diminish, through 1995 and beyond.
OW3#IIC C ICALS
The chemical sector is likely to be the first amjor comnodity inchstry to

break the trend of declining prices and profitability, and excessive over-
capacity. Chemical companies have shut down scores of plants in recent years,
201
and are finally seeing some growth In plant capacity utilization. According
to the Chemical Manufacturers Association chemical plant capacity factors
rose to 80.5 percent from 66 percent five years ago. The association expects
plant use to reach 82.3 percent in 1987.
The wave of plant closures in the early 1980's represented a dranmtic tur-
naround from the industry's building boom in the mid-1970's. After six years
of restructuring, cutbacks now appear to have succeeded in bringing capacity
back in line with demand. Total production in 1986 at 540 billion pounds was
essentially unchanged from 1985, and was still below the peak of 575 billion
pounds reached in 1979, but was decidedly inproved over the decade low point
of 464 billion pounds registered in 1982. Although overall demand was flat
between 1985 and 1986, the total was conposed of a 9.2 percent increase in
demnd for organic chemicals, coupled with a further 5.1 percent decline in
inorganic chemicals. Thus, the organic chemicals sector, which would encom-
pass all the Great Plains by-products, appears to be fully recovered.
Some organic chemicals are already in short supply. Polystyrene plants are
calculated to be running at 94 percent of capacity. Recent sharp price in-
creases for benzene and phenol were discussed earlier. Several plans for new,
smll, specialty chemicals plants are under consideration. Business cycle
analysis suggests that now would be an ideal time to invest in new organic
chemical plant capacity. Such an investment should only be made in a project
which is economically feasible now; but the investment can be ade with a high
degree of confidence that conditions in general will inprove rather than
deteriorate by 1995.
FEZ oLNN
Producers of organic chemicals derived from petroleum nmy worry less about oil
prices than about excess mnufacturing capacity. Their business is based on
conpetition over the mnufacturing mrkup on petroleum prices, regardless oL
the price of petroleum itself. Producers of coal-based chemicals, such as
Great Plains, however, must be vitally concerned about the absolute level of
petroleum prices. Because world oil prices are subject to mnipulation by the
Organization of Petroleum Exporting Countries, it is difficult to make price
forecasts based strictly on a supply/denmnd analysis.
The dramatic break in oil prices in 1986 created serious problems for the
petroleum industry. It was, of course, almost totally disastrous for economi-
cally mrginal sources of oil such as synthetic fuel processes based on coal.
For the long-term picture, little has changed. Petroleum is a finite resource
which will eventually be consumed.
Although the economics profession has been unable to successfully predict

world oil prices, there are some fundamental facts about world energy markets
which have been nmde nore clear by recent events.
Prices are ultimately set by supply/dennnd relationships, and it is now ob-

vious that energy demand is limited at any given price. In the 1970's,
projections of future petroleum demand were based on continued exponential
growth in total energy demand. Thoughtful persons realized that continuous
exponential growth is inpossible, but no one could predict why or when growth
in denund would cease. For the time being, the world's potential supply of
202
petroleum at the 1985 price of $28 per barrel clearly exceeds the demand at
that price. It is not yet clear what the equillibriam supply would be at typi-
cal 1986 prices of $14 per barrel. Supply has increased in many areas during
the last few years. In 1973 there were only 16 non-OPEC countries in the non-
conmnist world capable of producing at least 25 percent of the oil which they
used. Today that nuner is over 30. Six more non-OPEC countries have become
net oil exporters in the same period.
Faced with an inexorable decline in the demand for their oil, several oil ex-
porting countries abandoned, in late 1985, the strategy of curtailing produc-
tion to keep prices high and, instead began to increase production for sale at
market-determined prices. The resulting significant excess of supply over
demand drove prices down sharply in 1986. The innediate effect of this price
drop, in terms of decreased exploration activity, canceling new production
plans and shutting in marginal producing wells, indicates clearly that 1986
oil prices would be insufficient to maintain supplies at current levels. We
can therefore safely conclude that the economically optimum world price for
oil in the near-to mid-term future is somewhere between $14 and $28 per ba, -
rel. There is a mild consensus by industry analysts that for the near-term it
is probably in the neighborhood of $18 to $20 per barrel, although wide dif-
ferences of opinion, both higher and lower, exist. By 1995, prices should
again be approaching the top of the optimum range, or $28 per barrel.
With respect to the Great Plains situation, the difference between the price
of oil and the price of coal is of fundamental concern. It is this difference
which will justify or not justify burning coal instead of tar oil in the plant
and using the tar oil to make petroleum-related products.
The relative share of the energy market captured by coal and oil undergoes
slow but massive changes over time as a function of relative prices. Conpar-
ing oil and coal prices over a period of only a few years can be extremely
misleading because of the lagging effects of long-term purchase contracts, the
effects of inflation on capital costs for new mines, etc.
Any such general conparison also suffers from the non-fungible nature of coal,
intermixing of data for thermal coal and metallurgical coal, contract versus
spot market prices, world versus domestic markets, regional differences, etc.
Nevertheless, a long-term conparison can be instructive. Figure 81 shows
inflation-corrected prices for oil and coal over approximately the last
120 years. Although the possible shortcomings of such data must be remern-
bered, there is a rough overall correlation between oil and coal prices.
Because there is no application where coal and oil conpete in which coal would
be preferable to oil at the same price per BTU, coal will take market share
from oil only if it is available at a substantially lower price. Another way
of looking at the situation is that converting either the form of coal (so
that it can be used in oil-fueled equipment) or the form of equipment (so that
it can use coal instead of oil) requires a differential between coal and oil
prices. If oil is lower priced, no substitution will occur. Even when oil is
higher priced, the coal/oil differential (COD) must be large enough to make
some conversions feasible.
The constant-dollar COD is plotted in Figure 82. When the COD is positive, it
indicates the amount of cash flow per BTU of fuel consumed which can be
203
•
mi'," . ' , l ",- m ,° *• ' kq.'.*-' l '. .' ",J% ' '%j' -. ";%""
- "',. ''. %"% , "' .. " "•.% ,","' "
-COAL
4.0 -OL
1.6
0.8 _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _
1850 1 960 YEA 150 200
FIGURE 8 1
INFLATION -CORRECTED COAL & OIL PRICES
3
0
*w 2
IS
0)
-1
18460 1900 YEA 1940 19 80
FIGURE 82
THE COAL - OIL DIFFERENTIAL
204
WS*E~sw% NIWiNW ~naW
devoted either to manufacturing coal fuels with improved qualities or to

direct coal utilization. It is easy to see why the period from the 1930's to
1950 saw a massive substitution of oil for coal in applications such as rail-
road locomotives.
Only after 1970 did the COD show a great increase, sparking widespread inter-
est and activity in synthetic fuels from coal. Viewed from this perspective,
the 1986 drop in oil prices appears to have returned the COD to its historical
range. It appears prudent to conduct planning on the basis of a long-term COD
of much less than the $5 per million BTU evidenced in 1980. The COD ap-
plicable to a particular coal would have to be adjusted for the difference be-
tween the production cost for that coal and the average production cost repre-
sented in Figure 82.
The coal/oil differential may be viewed as roughly equivalent to the price

that a by-products operation could afford to pay as a transfer cost for the
raw liquid stream. If the transfer cost (exclusive of the value of the
lignite) is more than the coal/oil differential, then it will not be feasible
to produce liquid products which compete directly with petroleum products.
Figure 82 suggests that a coal/oil differential of less than $1 per million
BTU will prevail in the period to 1995. Unless a liquids transfer cost of ap-
proximately this order of magnitude can be achieved, the production of
petroleum-type products such as gasoline, diesel fuel, processing oils, etc.
is unlikely to be profitable. Only those products whose characteristics or 4
manufacturing costs derive some benefit from the specific chemical composition
of the Great Plains liquids should be pursued.
(IUTLGO( PM SIC IINUSRImKS USI5 GRIST PLAINS PIIX-S
Several of the major products likely to be produced by the Great Plains

Gasification Project can be related to the four-digit industry codes published
in the Standard Industrial Classification Manual, prepared by the Statistical
Policy Division and published by the Office of Management and Budget. (The new
Standard Industrial Classification Manual (SIC) will be published in mid-
1987.)
Using the SIC manual and the latest published national input-output tables,
it is possible to determine which industries are significant purchasers of
the products which could be produced at Great Plains. This section describes
the national input/output relationships between industries and the outlook
for industries which are purchasers of potential Great Plains products.
Relationships were found for the following:
Product Industry
Carbon black Rubber tires - SIC 3011
Rubber processing oils Synthetic rubber - SIC 2822 Il
Phenols Plastics and resins - SIC 2821
Cresylic acids Paints and varnishes - SIC 2851
Cresols Adhesives and sealants - SIC 2891
205
r, V *,,
Q.,;,:
V.,,, f,, ,, -:,. , . .:'.*.,* '? .." .,...,.;...-.:. - *. V"-...,
.-. . ..- .. .
Binder pitch Aluminum smelting - SIC 3334
Creosote Wood preservative - SIC 2491

(railroad ties, utility poles, etc.)
Coal tar pitch Roofing compounds - SIC 2952
Outlook For Rubber Tires And Inner Tubes - SIC 3011
Rubber products shipments are highly dependent on the motor vehicle market.
Approximately 85 percent of shipments go to passenger cars, trucks, and buses. "
Other markets include aircraft, agricultural, industrial, motorcycle, and
bicycle uses. Inner tubes represent less than two percent of shipments now
that tubeless tires dominate most markets.
As the rubber tires industry has shifted to longer-lasting radial tires, more
than 20 older tire plants have closed. Employment in the rubber tire industry
has dropped from 114,000 in 1977 to 66,500 in 1985. There have also been sig-
nificant gains in imports of foreign automobiles with tires mounted and in
imports of replacement tires. Import of replacement tires increased from 20
million annually during 1972-82 to 37 million in 1984. This trend is expected
to continue. Tire exports go mainly to Canada; imports are shifting away from
Canada and Europe to Asia. Asia accounted for 44 percent of United States
tire imports in 1985.
Slow growth is seen for the rubber industry, based on a number of conflicting
forces. Tires will be longer lasting. But demand will be enhanced by in-
creases in the driving-age population and in motor vehicle mileage driven.
Motor vehicles will remain a primary form of transportation and the number in
operation will grow. S.
Outlook For Synthetic Rubber - SIC 2822
Production of synthetic rubber is also suppressed by reduced demand from the

automotive industry. The only major exception is production of specialty
elastomers. The overall output of the synthetic rubber industry has been ad-
versely affected by the same factors affecting the rubber tire industry. P
However, demand for specialty rubbers has grown as markets have developed for
new automotive applications. Performance requirements in automotive and other
applications have forced the substitution of specialty rubbers for general-
purpose rubbers and rubbers have been increasingly alloyed with plastics to
form new materials.
During 1985, shipments of synthetic rubber declined six percent from 1984;
there also was a three percent decline in exports and a 17 percent increase in
imports. The United States has lost world market share over the last few
years but exports still amounted to $655 million in 1985.
Synthetic rubber prospects depend on the demand for rubber products in general
and on the industry's ability to develop new rubber-based alloyed materials.
Growth areas include roofing and oil additives, plus new products created from
alloys of rubber and plastics. %
S.
206
e*'. ".. , * -.,., °jN N ,". V , , -... .--

,. "N " , J ' . % ,%- %,
%n, .', , % * " %''"
%. % .% S .""'," "•" . '
Outlook For Plastics And Resins - SIC 2821
Plastics represent a wide variety of materials, with many built-in properties

making them adaptable for diverse markets. Packaging and
building/construction account for over 50 percent of the plastic end-use
market. In addition, plastics are used in furniture, autcm biles, toys,
electronics, and many other applications. Special properties and high-output
processing methods, such as injection nmlding, conbine to give plastics a com-
petitive edge over wood, metals, paper and glass.
The plastics materials industry is capital intensive, and its highly autoated
production facilities need relatively few production workers. Shipments of
plastics in 1985 increased approximately 4 percent in real terms over 1984.
U.S. exports of plastics accounted for about 11 percent of U.S. output in
1985. Saudi Arabia has recently entered the world rrmrket, a negative factor,
but the formerly strong dollar is declining and weak overseas economies seem
to be strengthening.
United States imports of plastics materials have been in an upward trend since
1977. Imports in 1985 were up nearly 16 percent following increases of 42
percent in 1984 and 75 percent in 1983. Major sources of imported plastics
materials are Japan, Canada and West Germany. Producers in oil-rich countries
have a favorable cost position in feedstocks and will continue to build up
their capacity.
One of the mre dynamic sectors of the plastics industry involves reinforced
plastics or composites. Major fibers currently being used ire glass, cel-
lulose, aranid, and carbon or graphite. Consunption of plastics composites
has been increasing rapidly.
Outlook For Paints And Varnishes - SIC 2851
The paints and coatings industry uses more than a thousand different raw
materials to produce binders, pigments, solvents and additives. Strengthened
economic conditions improved both the quantity and value of shipments in 1985.
Most of the sales effort of the United States industry is directed towards the
domestic market. Its ability to compete in foreign trade is limited by the
high cost of transportation and fragmentation of the industry. Major export
markets for the United States industry have been Canada and Mexico. Import
conpetition is small, but growing.
Shipments by the industry should continue to grow about three percent per
year. Growth areas include specialty coatings to pretreat polymer conposites
surfaces for bonding and new improved paints and specialty coatings for use by
electronics manufacturers. Advances in corrosion protection should increase
sales of special-purpose coatings.
The nunber of conpanies in the industry will probably decrease because of the
continued investment needed for production and research and development.
Outlook For Adhesives And Sealants - SIC 2891
Adhesives and sealants are formulated products, similar in conposition and

design to paints and coatings. Adhesives are highly specialized and complex
207
products tailored to meet the needs of many end-use markets. Packaging is the
major market for adhesives, in addition to the construction, transportation,
furniture, textile, shoe, bookbinding, and electronics industries.
Strong demand boosted industry shipments by five percent (constant dollars) in

1985. Synthetic resins and rubber-based adhesives accounted for 62 percent of
these shipments; sealants and caulking coffpounds, 20 percent; natural-based
adhesives, six percent; and other adhesives and sealants, 12 percent.
International trade in adhesives, both exports and inports, is small and

primarily with other industrialized countries. Inports rose 23 percent in
1985; exports slipped because of the high value of the dollar. Canada is the
major importer of United States adhesives and sealants, taking 30 percent of
United States exports in 1985.
Long term prospects for the industry are favorable. High performance ad-
hesives and sealants are gaining acceptance in a wide variety of applications,
and the development and use of new materials are creating demands for inproved
and new adhesives. Growth will be based on the ability of adhesives to reduce
the weight, increase the strength, and improve the quality of products, while
also reducing manufacturing costs. Research and development will continue to
be a controlling factor.
Outlook For Aluimmi Smlting - SIC 3334
The total volume of aluminum industry shipments reached a record level in

1985, but primary aluminum smelters continued to close or to reduce output be-
cause prices for primary metal were well below United States producers'
average cost of production. The domestic primary aluminum producing sector of
the industry is buffeted by high costs (especially for electricity), capacity
expansion in low-electricity-cost countries, oversupply, and low prices.
These factors have caused the industry to reduce primary output and close
capacity that is no longer competitive.
Thus, the domestic aluminum industry is facing a future that should see sig-
nificant changes in the structure of the industry. In the wake of new off-
shore capacity are shut-downs of domestic smelters, which could not conpete
against relatively low electricity costs, low labor costs, and/or increased
foreign government involvement.
The rate of growth in demand for primary aluminum has significantly declined
due to increased secondary recovery, saturation of certain end use markets,
and product improvements. These developments mean that the reopening of smel-
ters that are currently closed is unlikely.
Outlook For The Wood Preservatives Industry - SIC 2491
The value of shipments of treated wood increased by an estimated two percent

in 1985 over 1984. The similar increase in 1984 was 5.1 percent. Housing
starts are a principal factor affecting demand. Markets for treated lumner
and plywood have been expanding, while markets for treated telephone poles,
marine pilings, and railroad ties tend to be cyclical and grow slowly over the
longer term.
208
-- """
"
The railroad tie industry is complying with EPA directives that preservatives,
such as creosote, should be used only by certified treatnent plants or ap-
plicators. The EPA has also expressed concern over possible leakage of
creosote during railroad tie treatment and in storage yards. In addition,
there is concern over proper and safe disposal of used creosote-treated rail-
road ties by railroads.
Exports of treated wood products fell about 20 percent in 1985, due nostly to
the strength of the United States dollar. Most of these exports were treated
telephone poles that went to Bangladesh, Canada, the Caribbean, and Saudi
Arabia. Exports of these products probably will not increase nuch unless
foreign countries begin new rail, telephone, or port developnent projects.
Outlook For The Roofing Compomnds Industry - SIC 2952
The demand for coal tar pitch used in roofing compounds depends primErily on
the construction industry, principally the residential construction industry.
The inflation-adjusted value of new construction put in place rose by six per-
cent in 1985 and exceeded the previous peak year of 1973.
Overall construction spending will continue to reach record levels. Public

works construction has ended its long-term decline and will grow steadily over
the longer term.
p~* . 'a ~ ~'' 4 V 'V
209
DIWINITION OF OFTIEN PEU f SlIATE
Many factors enter into the definition of an optinum product slate. The
"optimim" slate could be chosen strictly on the basis of the highest indicated
rate of return on investment. It could be chosen on the basis of the highest
net cash flow above a threshold rate of return. Other factors influencing the
choice could be the amount of process (technical) risk, the long-term
prospects for market demand, the prospects for future increase in product
prices, etc.
DIOXJN 1 RATE OF R1TM ON INVW=IT
Considering only the discounted cash flow rate of return, the case involving
HRI's Dynaphen process to nuke benzene and phenol from the naphtha and crude
phenols streams would be assigned the highest ranking. A DCFROR of 22.5 to
35.9 percent was calculated. The only other process configurations achieving
a axinrun rate of return above 25 percent are the BTX case, based on UOP's
Unibon hydrotreating plus Sulfolane extraction technology for the naphtha
stream alone, and two nultiproduct cases based on cost estimates prepared by
R.B. Grady & Associates.
In all of these cases, there is considerable doubt about the accuracy of the
capital cost estinate. The Dynaphen process is not yet conmercialized, the
Unibon and Sulfolane units are usually built in nuch larger sizes and have not
been operated on similar feedstocks, and the Grady estimates were not based on
detailed engineering. If the capital cost estimate is doubled, only the
Dynaphen schemes show a rate of return above 20 percent.
NfT FPD EDS VALUE
The DCFROR criterion determines the greatest rate of return per dollar in-
vested. However, a highly profitable but very small project will not con-
tribute nuch to the total cash flow of a large corporation. It is also inpor-
tant to consider the net present value of all the cash flows generated over
the lifetime of the project. When the cash flows are discounted back to the
present, using a discount rate of 10 percent, the highest ranking goes to the
Grady nultiproducts case.
Of course the larger the project, the mere noney at risk if the project runs
into difficulty.
SEMITIVITY 70D CAPITAL ODST
Because of the large uncertainties regarding capital cost estinmtes, it is

desirable to have a project whose rate of return will still be robust if the
capital cost turns out to be nch larger than expected. The Dynaphen process
shows the highest DCFROR after a doubling of the capital cost estimates.
ABILITY 70 MWir PBCEXSS SPEMIFIC&TIONS
Processes which produce a single fungible chemical as product are likely to

have less difficulty in meeting customer specifications than a process which
produces conplex mixtures of chemicals. The coiplex mixtures present dif-
ficult quality control problems, and it ay be necessary to allow the customer
210
to make extensive sample runs before purchasing a product. In this respect,
the multiproduct schemes are at a disadvantage because there will be many dif-
ferent specifications that have to be met in order for the plant to be
prof itable.
NAKE! SIZE
The most desirable market to enter is one in which the amount of product which
will be produced is small in relation to the total market volume. A product
manufactured in large volumes cannot be introduced into a small volume market
without depressing prices.
Those products nost sensitive to market volume are the cresylic acid cases,
where the Great Plains output could be of the same order of magnitude as the
United States demand. Creosote could also be affected by market size. Poten-
tial benzene and phenol production from Great Plains is very small with
respect to the total size of the market.
PNOCESS RAMIi S
Each of the process configurations was ranked with respect to the others in
accordance with the preceding discussion. The results are seen in Table 98.
The two Benzene/Phenol cases (Dynaphen) clearly outranked all the other cases.
The Benzene and BTX cases were next, in spite of a very low ranking for net
present value. The lowest ranked cases were the Cresylic Acids Only cases.
Although this analysis is helpful in pointing out the more promising avenues
to explore, and it points out problem areas which may require attention in
some of the cases, it should not be accepted as a definitive judgment on the
relative merits of each case. Although the ranking criteria used in Table 98
are commonly encountered, other criteria could be defined which would yield
different results. Special situations can greatly change the ranking under
any criterion. For exanmple, a sales contract with a cresylic acid processor
could renove rost of the uncertainty concerning the ability to meet product
specifications. This would materially improve the ranking for this process.
There are many unknowns with respect to the high-ranked Dynaphen process. The
economic analysis was based on 100 percent conversion of all alkyl phenols to
phenol. This could never be achieved in practice and a much lower yield would
greatly reduce the estimated rate of return. A 75 percent yield, for exanple,
would reduce the maximum rate of return from 34 percent to about 25 percent,
and drop Dynaphen out of the top economic ranking.
OrrIiM PROM= SLATE

Accepting that the words "optimum product slate" are incapable of exact
definition, it nevertheless appears that a benzene plus phenol slate would
have a number of intrinsic advantages. Of particular interest is the fact
that It could be implemented either with or without the simultaneous produc-
tion of military jet fuel.
There are strong national security arguments for demonstrating the production
of military jet fuel at the Great Plains plant. However, the benzene and
phenol components of the raw liquid streams are not of value for jet fuel
211
ca0
4-4b 4; +
41 41 . .
. . .4(
-4'-
4.50
t4 4-b A
II " '-4-4
411
-*04-
'Cad)~' 0-4-4
14-4- .- 0,U)
MW
K 212
production. Benzene is too low boiling for jet fuel and the phenolic con-
stituents, if hydrotreated, would be converted to additional light aromatics.
These conponents are not only of the highest value in themselves, but are of
the least value for making jet fuel. Therefore an optimum strategy clearly
suggests that the light aromatics and phenols in the naphtha and crude phenols
streams be treated separately from the tar oil stream. Although it may even-
tually be possible to recover additional phenols from the tar oil stream, in-
cluding this source does not appear to increase profitability, and would add
additional complexity and technical risk.
V The optinum product slate is therefore defined as follows:
Great Plains
Plant Stream Disposition Final Product
Raw Naphtha Hydrodealkylat ion Benzene

Crude Phenols Hydrodealkylation Benzene + Phenol
Tar Oil Hydrotreating Military Jet Fuel
.4
-
.,...Although routes are known for hydrodealkylation of the naphtha stream, the
only route discovered for dealkylation without dehydroxylation of the crude
phenols stream is the Dynaphen process. Any heavy residue from the tar oil
stream could perhaps be developed as a coal tar product for roofing tar or
driveway sealers. These applications are purely speculative, however.
If the possibility of making military jet fuel does not materialize, the tar
oil stream can best be marketed as creosote, provided that a new specification
can be created with the American Wood Preservers Association.
A *213
. U
0ONCLUSIGS AND R c TIIs
1EASIBILITY OF BY-PKFrV PRODUTIGN
The major conclusion of this study is that there are profitable opportunities
for by-product development, even if the transfer price assigned to the crude
liquid stream from the gasification plant is as high as $3 per million BTU.
The profitability of any by-products production scheme is largely affected by
the liquids transfer price, and becomes questionable for almst all schemes if
the transfer price rises over $3.
Unless petroleum prices increase strongly in the near future (not forecast to
occur) none of the potential Great Plains liquid products which would conpete
directly with crude oil or heavy fuels are likely to be economically practi-
cal. This includes such products as refinery feedstock carbon black
feedstock, gasoline, fuel oils and rubber processing oil.
SUITABILITY FOR MILITARY JET FUKIS
It was not within the scope of this study to investigate the economics of
producing military jet fuels from the gasification plant streams. It was
determined that a suitable demnd exists within the state of North Dakota to
absorb the potential production without having to ship for excessive dis-
tances. It is highly unlikely that a jet fuel mnufacturing operation could
be successful on a conmercial basis. A firm purchase contract with the
Department of Defense and having a price floor would be required. Since some ..
of the jet fuel supplied to Minot Air Force Base is currently trucked in from
Canada, supplying this base from the Great Plains plant would reduce overall
petroleum inports.
Long-term national security considerations make the demonstration of a small-

scale jet-fuel-from-coal operation highly desirable. This demnstration could
utilize the tar oil stream from Great Plains.
ZVi= ON ROQIO" WZFINE) PROLUfTS BlLA=
The Great Plains mrket region, consisting of five states and two provinces,
is a net exporter of crude oil, and is approximately in balance with respect
to production and consunption of refined products. It is concluded that any
new production of refined products in the region, including jet fuel,
gasoline, fuel oils and residual oil, would probably result in a slight
decrease in regional refinery runs. The decrease In refinery purchases of
crude would probably result in mrre crude oil being exported to other states.
A mxinum of 5,000 barrels per day of liquid products could be produced at

Great Plains, and this would be less than one percent of regional refining
capacity. Therefore no discernible effect on average product prices within
the region would be expected, although some effect could be seen at the point
of introduction.
rauc SLATE
The economic evaluations carried out in this study suggest that if jet fuel Is
produced, the naphtha and crude phenols streams are the most suitable for by- w.
214
product development. Because these streams are not desirable feedstocks for
jet fuel production, the optinum product slate consists of jet fuel from tar
oil plus chemicals from the naphtha and crude phenols streams. Based on the
preliminary economic analyses carried out with the limited cost data avail-
able, benzene and phenol would appear to be the nmst promising candidates.
However, this appearance can be affected by a number of factors.
TAR AIDS MlU rIGN
A najor uncertainty in the estination of capital costs for a nultiproduct

operation is the cost of a process unit for extracting tar acids from the
gasification plant streams. Three major process contenders for this operation
are:
1. Caustic extraction and springing with carbon dioxide.
2. Solvent extraction by the Lurgi Phenoraffin process.
3. Solvent extraction by a nethanol/hexane solvent, as formarly prac-

ticed by Pitt-Consol Company and later by Merichem.
No good capital cost estinates for this process block are available. Using
the available information, switching from the first process listed above to
the second one made a difference of over 14 percentage points in the maximum
calculated discounted cash flow rate of return. It is not possible to make
sound judgments on process development routes to follow until the data in this
area are improved. It is therefore reconnended that some preliminary en-
gineering design effort be expended to arrive at credible up-to-date costs for
the tar acid extraction step. If jet fuel is not produced, extraction of
cresols from the tar acids would produce the largest total revenue stream.
PXSS DATA IR BMZM PMUJCTION
There are several processes available which might be suitable for making ben-
zene from the naphtha stream. Both thermal and catalytic hydrodealkylation
processes have been connercialized, or a solvent extraction process could be
used. However, there are large question marks concerning the ability of these
processes to handle the Great Plains liquids and concerning the costs for such
processes if constructed at the relatively small scale dictated by the Great
Plains flows. Further definition of the engineering parameters and costs is
required.
*O IN Pi MUxrIN OF BUIZUEMAND
) a
The combined production of benzene and phenol by the HRI Dynaphen process is
theoretically intriguing. It would make it possible to take advantage of the
natural characteristics of the gasification by-product liquids (i.e. high
phenol content) rather than simply using brute force to transform the con-
stituents Into other, saleable materials.
Deonstrating the feasibility of the Dynaphen process on the Great Plains

feedstock will require some basic laboratory research and development because
the process is not yet conmercialized. The potential advantages of the
215
q~V 's
I '*.*.* .. _. N. -.
process would appear to justify such an effort, and it is recomnended that
this be initiated.
PARALLEL NRKYINC SNORTS
It is strongly emphasized that the apparent economic advantage of the "optinum

product slate" defined by this study may be illusory. The cost data used for
the upgrading processes are simply not of sufficient quality to draw firm con-
clusions. Further refinement of the cost data could significantly change the
rankings of different processes. Therefore, it is not recommnded that any
potential by-products be eliminated from consideration at this time.
Creosote, cresylic acids and rubber processing oils could be developed as
profitable products if the liquids transfer price is low enough.
If the liquids transfer price is above $3.00 per million BTU, however, it will
be difficult to develop a profitable operation based on significant upgrading
unless the product selling prices are at least $5 per million BTU, or ap-
proximetely $30 per barrel ($0.10 per pound). Products clearly meeting this
criterion include benzene and other light arommtics, phenol and cresylic
acids. Almost all other products become marginal under these conditions.
MICK OUTLOOK
Prices for most chemical and petroleum commodities have suffered significant
declines over the past few years. Further significant declines appear un-
likely because present levels are low enough to have resulted in the shutting
down of most "excess" capacity. Demand for some chemicals (such as benzene)
in the petrochemical sector has recently surged, and there are solid prospects
for future growth In demand.
On the whole, the downside risk for future prices of most products should be
relatively small. Projects which can be shown to be profitable under current
conditions should be even more profitable in the future. It is recomiended
that planning for by-products production be given a high project priority.
4.
1, 2 1 . 1. 1
0U.S.Governmfnt Printinq Office: 1987

%.
7;48-061/61063 216 '". .- .
_
I W
- -%
'IP
%

Production of Jet Fuels From Coal Derived Liquids Forks Energy Research Center

Uploaded by

Copyright:

Available Formats

Production of Jet Fuels From Coal Derived Liquids Forks Energy Research Center

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Production of Jet Fuels From Coal Derived Liquids Forks Energy Research Center

Uploaded by

Copyright:

Available Formats

R194 9 PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS V/3

VOLUMIEI MRCET ASSESSN. (U) NORTH DAKOTA UNIV ORON

r UNCLSSIFIED AFWAK-TR-8-2042-VOL-i MJPR-FY1435-SE-NW.S? F/O 21/4 M

MICROCOPY RESOLUTION TEST CHART

AD-A 184 84-6

PRODUCTION OF JET FUELS FROM COAL DERIVED LIQUIDS

VOL I - Market Assessment for Liquid By-Products from the

INTERIM REPORT FOR THE PERIOD 20 January 1987 - 30 April 1987

APPROVED FOR PUBLIC RELEASE; DISTRIBUTION IS UNLIMITED

AERO PROPULSION LABORATORY

This report has been reviewed by the Office of Public Affairs

WILLIAM E. HARRISON CHARLES L. DELANEY, Actir Chief

FOR THE COMMANDER

Copies of this report should not be returned unless return is required

REPORT DOCUMENTATION PAGE TOMSF8 M. 070O-OM,

Is. REPORT SECURITY CLASSIFICATION lb. RESTRICTIVE MARKINGS

7960 Niwot Rd Suite 12 Wright-Patterson AFB Ohio

rl UNCLASSIFIEDUNLIMITED 0 SAME AS RPT. 0 DTIC USERS UNCLASSIFIED

In September 1986, the Fuels Branch of the Aero Propulsion Laboratory at

QUALY By ...... ............

A', I,:v C. des

Grady/ANG Case 171

Figure Title Page

1 Air Force Bases in Great Plains Region 20

Figure Title Page

50 Phenoraffin Extraction Process 145

Table Title Page

1 Liquid By-Products from the Great Plains Plant 7

Table Title Page

51 United States Consumption of Aluminum 110

The University of North Dakota Energy Research Center requested J. E. Sinor

Factors to be considered in the inplementation of a by-products nmnufacturing

FM A TUIICAL STAINPOINT, TWE VARIETY OF PRODUS MIICH COULD

TIE GREAT PLAINS REGICN IS BASICALLY SELF-SUFFICIEINT IN REFINED

ONLY ONE REFINERY IN TE REGION WILL HAVE THE CAPABILITY TO ACCEPT

THE ACCURACY OF THE INFORMATION AVAILABLE FOR EVALUATING THE

A MAJOR DEFICIENCY IN THE COST DATA AVAILABLE IS A RELIABLE ES-

RZDDVERY OF THE PHENOLS IN A PURE STATE IS THE KEY TD A PROFITABLE

$LL INKE VOLUS CONSTITUlE A RISK FR SOM FQINTIAL BY-

(IARISON OF 1985 AN) 1987 PHOOX1IC PRICES

Product 1985 1987

Gasoline 0.14 0.08

W1E PEIXCfIN OF REFINED FUEL PROUUXrF IS UNLIKELY 710 BE OWETI-

CCNTlTIONS OF M)ST Il)IVIDUAL CIPJI IN THE TAR OIL STREAN

33KJIWIM PU-EJNUFACTURIre NDTICES IRI)U THE TOXIC SUBSTANS ON-

TH OFTIEM PRODUCT SLATE WOULD BE MILITARY JET FUEL PFDUCED ROM

saNaMICs WILL BRE 7ONLY DEPDNDIT ON THE TRANSFER VALUE IICH

DCFROR, Percent, DCFROR, Percent,

Benzene 23.3 16.9

TIUMSIRATION OD'TS ARK A &JOR DETRRENT 70 PRDING LOf-VALUE

EMar OF TAiNSPor 00Ms ON RATE OF RZTURN

DCFROR, Percent, DCFROR, Percent,

Benzene 23.1 17.1

TH P ucr PRICE OUTLOOK IS GOIALLY POSITIVE. The worldwide

m OF PROUCT PRICES ON RATE OF RETUR

DCFROR, Percent, DCFROR, Percent,

Benzene 12.6 32.1

LIqUID BY-PRWOUl eM I GEAT PLAINS PLANT

Tar Oil 1.016 319.9 Million pounds per year