EPA-AA-SDSB-82-04
A Technical Report
Refining of Coal-Derived
Synthetic Crudes
by
John McGuckin
February 1982
NOTICE
Technical Reports do not necessarily represent final EPA decisions
or positions. They are intended to present technical analysis of
issues using data which are currently available. The purpose in
the release of such reports is to facilitate the exchange of tech-
nical information and to inform the public of technical develop-
ments which may form the basis for a final EPA decision, position
or regulatory action.
Standards Development and Support Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
-------
Table of Contents
Page
I. Introduction 1
II. Petroleum/Coal-Liquid Integrated Refineries 1
versus Grass Root Coal-Liquid Refineries
III. Analysis of Grass Root Coal-Liquid Refineries 2
A. Introduction 2
B. Overview of Coal-Liquid Processing Cost 3
C. Studies Eliminated From Consideration in Development
of Refining Cost 7
D. Refinery Configurations 8
1. Chevron Refinery Configuration 8
2. UOP Refinery Configuration . 10
3. ICF Refinery Configuration 12
E. Discussion of Parameters Effecting Product 12
Costs for Coal Liquid Refineries
1. Level of Engineering Design for 12
Investment Estimates
2. Refinery Feedstocks 14
3. Product Slate 18
4. Product Qualities 22
5. Coal-Liquid Refining Costs 22
F. Reconciliation of EDS, H-Coal and 25
SRC-II Refining Cost Differences
IV. Analysis of an Integrated 27
Coal-Liquid/Petroleum Refinery
A. Introduction 27
B. Feedstocks 28
C. Product Slate and Specifications 28
-------
Table of Contents (cont'd)
Page
D. Petroleum Refinery Configuration 28
E. H-Coal Liquid Petroleum Refinery Configuration ... 28
F. Material Balance 35
G. Economic Comparison 35
V. Economic Comparison of Grass-Roots vs 39
Integrated Coal-Liquid Refineries
A. H-Coal 39
B. SRC-II 41
C. EDS 41
VI. Conclusion 41
VII. References . . 43
-------
I. Introduction
This report will cover the economics of refining coal-liquids
(synthetic crude oils) obtained from the direct liquefaction of
coal. These coal-liquids must undergo considerable upgrading
before they can be considered environmentally and functionally
acceptable end products. These syncrudes are upgraded in opera-
tions analagous to those applied to petroleum crudes with the key
step being hydrotreating to reduce high nitrogen, oxygen, and
sulfur contents. The major products from the coal-liquid refin-
eries are gasoline, distillate and residual oil.
This discussion of the refining of coal-liquids consists of
five main sections. The first section discusses the refining of
coal syncrude/petroleum crude blends in existing petroleum refin-
eries versus the refining of coal syncrudes in grass-root coal-
liquid refineries. The next section is an analysis of grass-root
coal-liquid refineries for H-Coal, SRC-II, and EDS syncrudes. It
includes an overview of refining costs based on design studies al-
ready performed, a discussion of refinery configurations, and a
discussion of parameters affecting refining cost. The third
section presents an analysis of an integrated coal-liquid/ petro-
leum refinery. The fourth section compares the costs of proces-
sing synthetic crudes in existing refineries versus the costs of
processing them in new grass-root refineries. The last section
presents an economic summary of the most representative refining
costs.
II. Petroleum/Coal-Liquid Integrated Refineries versus Grass-Root
Coal-Liquid Refineries
There are two main options for the upgrading of coal
syncrudes. One option is to perform the upgrading in grass-roots
coal-liquid refineries, applying technology used in many petroleum
refineries, but with processing capabilities specifically designed
to refine coal-liquids. The other option is to upgrade coal-
liquids together with petroleum crudes in existing petroleum
refineries by making process alterations as required.
The coal-liquid/petroleum refinery integration option appears
to have a number of advantages over the grass-roots coal-liquid
refinery. Some of these advantages are:
1. The construction of coal-liquid refineries requires a
large capital investment relative to processing at existing refin-
eries, and projected capital shortages makes their construction
unlikely.[1]
2. Since the United States' refinery utilization has been
declining and refineries are currently operating at about 70
-------
-2-
percent of capacity, the excess capacity could be used to upgrade
syncrudes. Construction of new refineries would seem unwar-
ranted. [2] [3]
3. Since coal syncrudes may well be produced by owners of
existing refineries who have the necessary refining technology,
coprocessing these syncrudes together with petroleum in their
refineries would be a convenient syncrude processing method.[1]
4. Existing refineries have a comfortable degree of flexi-
bility to meet shifts in market demands, thus ensuring a market
for the syncrudes.
5. A syncrude/petroleum crude blend would minimize
specialized processing required for the synthetic components, and
it reduces performance, compatibility, and quality risks associat-
ed with using a new product.[4]
A disadvantage is that the feasibility of processing coal
syncrudes together with petroleum has not been demonstrated even
on a laboratory scale, although it appears to be feasible to
process up to 23 percent syncrude with petroleum crude based on
the results of a high level refinery computer model.[1] The
processing of 100 percent coal liquids in a grass-roots refinery
has been shown feasible on a laboratory scale.[5]
From this, it would appear that the most likely path for
growth for an oil from coal industry would be through the refining
of syncrude/petroleum crude blends in existing refineries.
However, there has been only one study[6] performed to date on
coal-liquid/petroleum refinery integration. This study will be
used to compare the processing of petroleum in a typical large
refinery with the processing of 10 percent H-Coal product in an
integrated coal-liquid/petroleum refinery.
However, since there has been so little work performed on
integrated refineries, this refining discussion will mainly
investigate the processing of whole EDS, H-Coal and SRC-II coal-
liquids in grass-root coal-liquid refineries, where there is much
more information. These refineries would have high capital costs
but low operating costs relative to processing in existing refin-
eries. The capital costs derived from the grass-root coal-liquid
refineries will then be compared to the capital costs of retro-
fitting an existing refinery to process coal-liquid/petroleum
crude blends. Also, the efficiency of the retrofitted refinery
will be compared to the efficiencies of a coal-liquid refinery and
a petroleum refinery.
-------
-3-
III. Analysis of Grass-Root Coal-Liquid Refineries
A. Introduction
A few studies have been conducted to estimate the cost of
refining H-Coal, SRC-II, and EDS coal liquids. These studies will
be discussed here in order to determine representative refining
costs for each syncrude and to obtain some idea of the processing
requirements for the grass-root refineries.
First, an overview of the processing cost estimates for each
of the studies will be presented; all costs have been placed on
the same economic basis as discussed in a previous report.[7a]
Secondly, a brief discussion of the refinery configurations will
be presented to obtain some idea of the processing schemes for
coal-liquid refineries. Also, the thermal efficiencies of coal-
liquid refineries will be discussed. Lastly, the refining studies
will be analyzed with respect to some important parameters which
affect the cost of refining; these parameters include feedstock
properties, product slate, product qualities and level of
engineering design conducted for the investment estimate. This
discussion will help to determine a representative refining cost
for each of the coal syncrudes.
B. Overview of Coal-Liquid Processing Costs
Investment and operating costs for coal-liquid refineries
have been reported in a few different studies. The cost estimates
for the SRC-II syncrude were made by Chevron and ICF.[5][8] The
estimates for the H-Coal syncrude were made by UOP, IGF, and
Exxon.[7][8][9] The only estimate available for the EDS syncrude
was made by ICF.[8] In addition to these specific studies, Exxon
has prepared a rough study which presents a range of costs for
upgrading a coal liquid in general.[10]
The economic basis for the refining costs is essentially
identical to the basis discussed in a previous report.[7a] In
addition the following criteria were used for estimating the
refinery processing costs:
1. The plant size for the refineries was adjusted to a
feedrate of 54,500 barrels per calendar day (BPCD) using the same
capital scaling factor of 0.75, which was discussed in a previous
report.[7a]
2. No credit was taken for the sulfur and ammonia byprod-
ucts. These costs are small and would have little effect on
refining costs.
3. A 1990 real cost (in first quarter 1981 dollars) of
$5.39/MMBtu for natural gas was used in those cases where natural
gas is consumed.[8]
-------
—4—
Tables 1 and 2 present economic summaries of the processing
costs for H-Coal, SRC-II, and EDS coal liquids, based on invest-
ment and operating costs from the studies mentioned above. Table
1 represents an optimistic case with a capital charge rate of 11.5
percent; Table 2 is based on a higher capital charge rate of 30
percent. The operating costs do not include the cost of the coal
syncrude. Therefore, to determine the total product cost, the
refining costs would be added to the cost of the syncrude from
direct liquefaction and the distribution cost of the products.
Table 3 presents the breakdown of refinery products based on a
syncrude feedrate of 54,500 BPCD.
With a 11.5 percent capital charge rate, refining costs vary
from an average of $1.50/mBtu of product for the H-Coal syncrudes
to about $2.00/mBtu for the SRC-II syncrude. With a 30 percent
capital charge rate, refining costs vary from an average of
$2.30/mBtu for the H-Coal syncrude to about $3.40/mBtu for the
SRC-II syncrude.
A comparison of Table 1 with Table 2 shows the effect of
increasing the capital charge rate on product cost. For capital
intensive processes, as in the Chevron and UOP cases, product
costs double with the increase in capital charge rate. For cases
with high operating costs, as in the.ICF and Exxon (H-Coal) cases,
the capital charge rate does not have as great an effect on prod-
uct cost.
The reason for the large differences in capital and operating
costs between the studies lies in the methods used to supply
refinery fuel and hydrogen for upgrading. The Chevron refinery
utilizes the undesirable residual oil from the syncrude for refin-
ery fuel and hydrogen production via partial oxidation. The UOP
refinery uses steam reforming of light napthas for hydrogen prod-
uction. The ICF refineries purchase refinery fuel and natural gas
for hydrogen production via steam reforming. The purchasing of
natural gas by the ICF refineries results in relatively high oper-
ating costs, while the capital costs for hydrogen production is
relatively higher for the Chevron refinery than the UOP and ICF
refineries (a partial oxidation unit has a higher capital cost
than a steam reformer). A major result of these differences in
refinery configuration is that the more capital intensive refine-
ries (Chevron and UOP) produce a higher quality product than the
less capital intensive ICF refineries since they consume the
residual oil in the process.
Offplot investment is another reason for the large capital
cost differences between the studies. Offplot allowances repre-
sent 30 and 48 percent of the total capital investment for the
Chevron and UOP refineries, respectively; while they represent
only about 12 percent for the ICF refineries. This may be the
result of differences in the actual designs'of the refineries or
could be due to the use of different cost estimation procedures.
-------
-5-
Table 1
Processing Cost Estimates Using a 11.5% Capital Charge Rate
(Millions of First Quarter 1981 Dollars)
SRC-II H-Coal EDS
Exxon (1974 Exxon (1974
Millions of Dollars Chevron[5] ICF[8] UOP[7] ICF[8] Study)[9] ICF[8] Study)[10]
Total Instantaneous 780.9 396.9 454.4 280.4 270.0 328.7
Investment
Total Adjusted 1033.9 525.5 601.6 371.2 357.5 435.2
Capital Investment*
Annual Capital 118.9 60.4 69.2 42.7 41.1 50.0
Charge
Annual Operating Cost 57.7 204.7 41.8 110.4 175.1 158.4
Total Annual Charge 176.6 265.1 111.0 153.1 216.2 208.4
Refining Cost
$/bbl of Product 9.76 13.42 5.75 8.19 10.94 11.03 6.90-14.70
$/mBtu of Product 1.84 2.19 1.06 1.42 2.00 1.87 1.31-2.46
Includes working capital and start-up costs.
-------
-6-
Table 2
Processing Cost Estimates Using a 30% Capital Charge Rate
(Millions of First Quarter 1981 Dollars)
SRC-II H-Coal EDS
Exxon (1974 - Exxon (1974
Millions of Dollars Chevron[5] ICF[8] UOP[7] ICF[8] Study)[9] ICF[8] Study)[10]
Total Instantaneous
Investment 780.9 396.9 454.4 280.4 270.0 328.7
Total Adjusted
Capital Investment* 1022.2 519.5 594.8 367.0 353.4 430.3
Annual Capital Charge 306.7 155.9 178.4 110.1 106.0 129.1
Annual Operating Cost 57.7 204.7 41.8 110.4 175.1 158.4
Total Annual Charge 364.4 360.6 220.2 220.5 281.1 287.5
Refining Cost
$/bbl of Product 20.13 18.25 11.41 11.80 14.22 15.22
$/mBtu of Product 3.80 2.98 2.10 2.05 2.60 2.58
Includes working capital and start-up costs.
-------
-7-
Products, BPCD
LPG
Unleaded Regular
Gasoline
Unleaded Premium
Gasoline
No. 2 Fuel Oil
Table 3
Product Slates Based on 54,450
BPCD of Syncrude Charged to Refineries
SRC-II
H-Coal
EDS
Exxon
Chevron[5] ICF[8] UOP[7] ICF[8] (1974 Study)[9] ICF[8]
2,828
49,590 9,708 23,585 17,072 36,021
10,122
16,871 22,018 18,147
21,054
Residual
TOTAL
44,434 — 12,109
49,590 54,142 53,406 51,199
54,168
30,710
51,764
Product Energy, mBtu/CD*
LPG
Unleaded Regular 262,827
Gasoline
Unleaded Premium —
Gasoline
No. 2 Fuel Oil
11,538
51,454 125,001 90,481
53,647
190,911 111,586
97,852 127,703 105,253
Residual
TOTAL
279,936 — 76,285
262,827 331,389 288,038 294,469
296,164
193,472
305,058
* The following energy contents (mBtu per barrel) have been assumed;
Gasoline - 5.3
Fuel Oil No. 2 - 5.8
Residual - 6.3
LPG - 4.08
-------
-8-
C. Studies Eliminated from Consideration in Development of
Refining Costs
As mentioned above, Exxon determined a rough estimate of the
cost of distilling, upgrading, and handling the products from
direct liquefaction by assuming that the total upgrading cost was
proportional to hydrogen consumption.[10] Since this Exxon/1980
study was prepared to give a rough indication of coal liquid
refining costs, and was not considered by Exxon to be specific for
the H-Coal, SRC-II or EDS syncrudes, it will not be considered
further in the discussion of refinery configurations, and refining
costs.
Another study which will not be considered further is the
Exxon/H-Coal study published in 1974. This study is now outdated,
since the feedstock property data were based on a 1967 study by
Hydrocarbon Research Institute.[11] The property data indicate
low nitrogen and oxygen contents compared to more recent data
supplied by Mobil Research and Development Corporation.[7][12]
High nitrogen and oxygen contents account for much of the high
refining costs of coal liquids. Another factor considered in not
using this study is its high operating costs. These costs are the
result of purchasing approximately 11,280 fuel oil equivalent bar-
rels per calendar day (FOEB/CD) of refinery fuel, steam reformer
feed and fuel, and utilities while yielding about 41,600 FOEB/CD
of product. The thermal efficiency of this refinery is only 74
percent. Today's refineries are becoming much more energy self-
sufficient and energy efficient and would generate much of this
energy from within the refinery.
D. Refinery Configurations
The key step in the refining of syncrudes is hydrotreating
which removes the heteroatom impurities and increases the hydrogen
content of these materials. Based on feedstock properties and
product slates, each study utilized a different refinery configu-
ration. In this section each refinery configuration will be
briefly described.
1. Chevron Refinery Configuration[5]
Chevron analyzed six different cases in which different prod-
uct slates and types of processing were imposed upon the SRC-II
syncrude refinery. The two most practical cases are discussed
here. The first case in which 100 percent gasoline is produced is
shown in Figure l.[5] Initially, the whole SRC-II syncrude is
hydrotreated which provides essentially complete nitrogen
removal. The hydrotreated SRC-II oil is then sent to a fraction-
ator where light naptha, heavy naptha, and gas oil fractions are
obtained. The light naptha fraction from distillation is sent
directly to motor gasoline blending, and the heavy naptha fraction
is hydrotreated and then catalytically reformed; the reformate is
-------
SCHEMATIC FLOW DIAGRAM
REFINING OF SRC-II OIL BY
HYDROTREATING AND HYDROCRACK ING - CASE 4
DOE CONTRACT EF-76-C-01-2315
Gas to Refinery Fuel
1 SRC- MOil
. .. __»_„ nyuiuyeii iu
Refinery Gas It r n 1 Hydrogen Hydrolreaters
ir Waler from H2b
fining Units Waste Water
Treating
1 H
Recycled Waier
to Refining Units c
c
•«-
n
-#-
e SRC- II Oil Intermediate 1
• '» jeveniy • *
Hyd rot real ing
Hydrogen
• iii J HGIAJVCI y - ni««i —
HeavyJJaphlha ^ Planl
1'
H ^
. _.., .....,,..» Sulfur Plinl ...,,...., —
Ammonia
Light Naphtha
To Hydrogen
i Cas Plant Hydrogen Gas
1 t I t t
; ||pfluy 1 | , .1 ... .1 , .
' Naplilha Naphtha ! r Catalytic , Motor Gasoline^
Hydrolreating Reforming
Hydrogen
~\
Hydrocracking
Hvdronpn -
1
1 Refinery Fuel
i _ — . — _ — •
I
VO
I
-------
-10-
sent to motor gasoline blending. The gas oil fraction is hydro-
cracked and then recycled to the fractionator. Total hydrogen
consumption is about 2,633 scf per barrel of SRC-II feed.
Refinery furnace fuel and hydrogen plant feed requirements
are met with untreated SRC-II oil. However, steam generating
boiler plants were assumed to be coal-fired in compliance with DOE
regulations. The thermal efficiency of this refinery has been
estimated at 83 percent. The thermal efficiency was calculated by
considering the higher heating value of all input streams to the
refinery, and all product output streams. By-products were not
considered as product energy output.
In addition to the above case, Chevron analyzed a case where
the requirement to produce 100 percent gasoline (50,000 BPCD) was
relaxed to 36,400 BPCD of gasoline and 13,100 BPCD of fuel oil No.
2. This is shown in Figure 2. For this case the refinery utiliz-
ed a fluid catalytic cracker (FCC) in place of the hydrocracker.
However, the FCC processing route appeared economically unattrac-
tive. For coal liquid refineries it seems preferable to rely on a
flexible hydrocracking system rather than on fluid catalytic
cracking, since coal distillates fail to meet the basic require-
ment for a FCC feed, high hydrogen content.[13] To add sufficient
hydrogen to the FCC feeds, severe hydrotreating conditions are
required. Therefore, utilizing FCC as a route to gasoline prod-
uction requires simultaneous installation of hydrocracking facili-
ties resulting in an economic preference for a flexible hydro-
cracking system rather than a joint FCC/hydrocracking system.[13]
2. UOP Refinery Configuration[7]
The refinery configuration for the UOP/H-Coal case is
presented in Figure 3.[7] The key refining step is hydrocracking
the total H-Coal distillate to produce gasoline and distillate
fuel.
The H-Coal syncrude is first charged to an atmospheric
distillation unit where a light naptha overhead cut (C^C^)
and bottoms cut (0^-880°F) are obtained. The light naptha is
hydrotreated to remove sulfur and nitrogen and then used as a
feedstock for hydrogen production, blended into gasoline, or split
into a C4 and C5/C5 fraction for LPG and gasoline blending.
The fractionator bottoms (C6-880°F) are charged to a
naptha/distillate splitter where a variable naptha cut is taken
overhead and a variable distillate cut is recovered as bottoms.
Variable cuts were incorporated to provide flexilibity in achiev-
ing a range of gasoline/distillate product ratios.
The variable naptha cut is processed with a two-stage hydro-
treater. The H-Coal naptha requires severe hydrotreating relative
to petroleum naptha because of its higher nitrogen and oxygen
-------
-11-
content. The hydrotreated H-coal naptha is sent to a catalytic
reformer. Coal-derived naptha is easily reformable to high octane
levels because of its large quantity of high octane aromatics and
cycloparaffins.
The variable distillate cut from the naptha/distillate split-
ter is charged to a hydrocracker. Hydrocracking severity is set
by the gasoline/distillate ratio desired. The hydrogen consump-
tion required to hydrocrack the distillate is very high and can
range from 2.5-5.0 weight-percent of the charge. Napthas from the
hydrocracker are sent to motor gasoline blending while the distil-
late is sent to the No. 2 fuel oil pool.
Utilities such as fuel and steam are produced internally in
the refinery. Only power and water are purchased. The thermal
efficiency of the UOP/H-Coal refinery was estimated at 95
percent. This efficiency seems high considering the high severity
processing required for the H-Coal oil.
3. ICF Refinery Configuration^]
ICF assumed that the EDS, H-Coal, and SRC-II syncrudes would
be charged to a distillation unit located within the liquefaction
battery limits, and that the straight run products from the
distillation unit would be the charge feedstocks to a refinery
complex. This refinery complex is separate from the liquefaction
complex. In the refinery the naptha, distillate, and residual
fractions are all charged to hydrotreaters. These hydrotreated
streams were then assumed to be the refinery products. Thus, the
refining facility was assumed to simply consist of a natural
gas-charged steam reformer to produce hydrogen, hydrotreaters,
emission control and effluent control equipment, and general
offsite allowances. Natural gas, power, water, refinery fuel and
catalysts would all be purchased. An analysis of the EDS, H-Coal,
and SRC-II feedstock and product properties indicated thermal
efficiencies of 79, 84, and 79 percent, respectively.
E. Discussion of Parameters Affecting Product Costs for
Coal Liquid Refineries
A number of parameters affect the estimated cost of refin-
ing. Feedstock properties, desired product slate, product qual-
ities, and the level of engineering design conducted for the
investment estimate are among the most important parameters. To
determine representative refining costs for each syncrude, the
refining studies will be analyzed with respect to the above para-
meters.
1. Level of Engineering Design for Investment Estimates
When comparing the level of engineering design work used for
the investment estimates, it was found that none of the studies
-------
FIGURE Z
SIMPLIFIED FLOW DIAGRAM
REFINING OF SRC-II OIL BY
MODERATE SEVERITY HYDROTREATING - CASE 5
DOE CONTRACT EF-76-C-01-2315
R"°
1
Sour Water
from
Refining Units
Refinery Gas
He
2 b
Gas
H2S
Recovery
IH2S
1
u
Hydrogen
Plant
Hydrogen Ic
Hydrotreatin
Gas to Refinery Fu=
Waste Water
Treating
Plant
Ammon
I
(N
t
Whole SRC- II Oil
i • *•
I
1
1
j Hydi
1
Recycled Water
o Refining Units
Modi
Sew
Hydro!
rogen
irate
jrily
real ing
u<
jhl Napn
T
Gas
Heavy ... 1
Naphtha Napt
* Hydro!
Hydrogen
itha
reating
tha
oHydrog
Plant
t
! .
en
Gas Hydroge
1 I
Catalytic
Reforming
n
Motor Gasolii
L _J
Gas Oil
and Refinery Fu:
-------
FIGURE 3
CQAL OIL REFINERY
I lYDROCRACKING CASES WITH LOW TEMPERATURE FRACTIONATION
1
H2
PRODUCTION
C^CB I 11
• MnvDRoiREATERl-Lr>
OIL
IOC
r
R
A
C
1
I
O
N
A
T
O
H
cBi;o c
S
NAPHTHA
178204'C
»•» .S .• — — — -«
F
R
A
176 470 C
DISIIILATE
C
I
O
N
A
t
O
II
f
(04 470 C
-fc.
HvbnOTREATER
H
HVDROTriEATER
.. ..... 1 Cc/C« LIGHT NAPIIIHA
CRACKER 1 | NAPHTHA _
1
| DISTILLATE
>
NO 2
FUEL OIL
GASOIINE
POOL
UNI FADED
PIIEMIUM
" REUUi AR
-------
-14-
represented detailed engineering designs.[5][7][8] However, it
was found that the Chevron and UOP refining studies were based on
more engineering work than the ICF study. The ICF study was
intended simply to provide a rough indication of the effect of
refining on the cost of liquefaction products.[8]
Chevron based their study on laboratory data, along with
general petroleum processing and cost correlations based on
refineries constructed by Chevron. The UOP cost estimate was
based on their company's experience with refinery construction.
ICF's investment costs are based on an estimate of the amount of
hydrotreating required to bring the coal liquid products up to the
hydrogen levels of existing petroleum products.
Therefore, even though the ICF estimates are based on reason-
able methods, these methods are not considered as accurate as
those used by Chevron and UOP. However, before it can be conclud-
ed that the latter estimates represent the best refining costs,
feedstock properties and product slates need to be discussed since
these also have significant effects on refining costs.
2. Refinery Feedstocks
A key parameter needed to determine an accurate estimate of
refining cost is knowledge of the feedstock proper/ties. In this
section the coal liquid properties used by the various studies
will be discussed.
Table 4 lists inspections of the feedstocks as reported by
the Chevron and UOP refining studies. The hypothetical SRC-II
syncrude used in the Chevron refining analysis was derived from
Pittsburgh Seam coal. It is a blend of three fractions of SRC-II
direct liquefaction product. The fractions were blended to
constitute a syncrude typical of SRC-II operation. The H-Coal
syncrude used for the UOP refining analysis was derived from
Illinois No. 6 coal. It is a C4~878°F crude obtained from the
atmospheric column overhead and bottoms product of the Hydrocarbon
Research Institute H-Coal process and inspected by the Mobil
Research and Development Corporation.[12] The ICF study did not
report feedstock data properly, but did report a breakdown of the
syncrude into its basic petroleum product fractions. EDS coal
liquid property data as reported by Exxon are presented in Table 5
to obtain some idea of what this refinery feedstock will be
like. [14] The EDS coal liquid was derived from Illinois No. 6
coal.
A significant difference between the various studies lies in
the fraction of the syncrude represented by the naptha, distil-
late, and residual cuts. Table 6 presents the volume percent of
the syncrudes distilled as a function of boiling point range as
used by the various studies. Also included in this table are the
product cuts as reported by the main contractors of the direct
-------
-15-
Table 4
Refinery Feedstock Property Data
Specific Gravity
Gravity, °API
Total Nitrogen, Wt-%
Oxygen, Wt-%
Sulfur, Wt-%
Carbon, Wt-%
Hydrogen, Wt-%
Ramsbottom Carbon, Wt-%
Conradson Carbon Residue, Wt-%
Benzene Insolubles, Wt-%
Cj Insolubles, Wt-%
Ash, ppm
Bromine Number
Pour Point, °F
Viscosity, CS at 100°F
ASTM D 86/D 1160 Distillation,
at Vol-% Distilled:
Start/5
10/30
50
70/90
95/End Pt.
Distillation, °F vs. Vol-%
Distilled
13.87 Vol-%
30.84
10.4
40.89
3.99
H-Coal [7]*
0.8733
30.5
0.37
1.72
0.15
86.7
11.0
0.10
0.10
67
41.7
SRC-II [5]**
0.9427
18.6
0.85
3.79
0.29
84.61
10.46
0.70
0.03
40
70
-80
2.196
154/217°F
281/382
438
484/597
699/850
C6/350°F
350/399
399/650
650/880
*
**
Derived from Burining Star Mine, Illinois No. 6 coal.
Derived from Blacksville No. 2 Mine, Pittsburgh Seam coal.
-------
-16-
Table 5
EDS Coal Liquid Property Data [14]*
Wt% of Whole Crude
Specific Gravity
Gravity, "API
Carbon, Wt-%
Hydrogen, Wt-%
Sulfur, Wt-%
Nitrogen, Wt-%
Oxygen, Wt-%
Higher Heating Value,
Btu/lb
Million Btu/B
15/5 Distillation,
Wt-% off, °F
5
10
30
50
70
90
95
Naptha Fuel Oil Total C5+ Product
39.0
0.77
52.3
85.2
13.16
0.43
0.06
1.15
20,076
5.41
61.0
1.03
5.7
87.72
8.89
0.51
0.75
2.13
17,837
6.43
100.0
0.928
21.0
-
10.56
0.479
0.48
1.75
18,710
5.89
92°F
117
184
234
277
322
340
271°F
359
442
623
885
1030
1081
Derived from Monterrey Illinois No. 6 coal.
-------
-17-
Table 6
Coal Liquid Products as a. Volume
Percentage of the Whole Coal Syncrude
Naptha
Initial/
(350-380°F)
H-Coal
UOP [1]
ICF [7]
Fluor Corp.
SRC-II
Chevron [6]
ICF [7]
PMC [14]
[15]
EDS
ICF [7]
Exxon [13]
53
35
35
30
19
23
42
43
Distillate
(350-380°)/
650°F
43
42
42
60
71
22
Residual
(650° F+)
4
23
23
10
81
6
58
35
-------
-18-
liquefaction processes (Exxon/EDS,[14] Pittsburg and Midway Coal
Mining Company (PMC)/SRC-II,[15] and Fluor Corporation/
H-Coal,[16]).
The greatest variability lies in the fractions of residual
and distillate reported. The correct percentage of residual in
the coal liquid syncrude is debatable, and will be determined by
future economic conditions. The direct liquefaction processes
have the capability of recycling the heavy liquids to the lique-
faction reactor or utilizing this heavy material for hydrogen
production, thus decreasing the fraction of residue in the
syncrude product and yielding a lighter syncrude.
Since the direct liquefaction costs for this study are based
on the main contractors' estimates of the naptha, distillate, and
residual portions of the syncrudes, for consistency their esti-
mates must be used as the basis for this refining discussion.
Therefore, since the H-Coal product fraction used by IGF is iden-
tical to that reported by Fluor, it would be considered most
accurate. The H-Coal syncrude as reported by UOP should contain
less of the naptha fraction and more of the residual fraction.
Their refinery would need to have the capability of handling this
extra residual material. This would result in higher operating
and capital costs for the UOP refinery; therefore, their current
refining cost would appear to be an underestimation.
The SRC-II syncrude product fractions used by Chevron are
similar to those reported by PMC, while those used by ICF are
significantly different. The net effect of the feedstock differ-
ences on refining cost for the Chevron case would not be very
significant. However, for the ICF case there would be much less
residual processing and more distillate processing. The overall
effect on the ICF product cost would be to decrease it, since
residual processing is more severe than distillate processing.
Also, there would be more distillate product from the ICF case and
less of the undesirable residual oil.
The EDS syncrude fractions used by ICF differ from Exxon's
mainly in the percentage of distillate and residual reported. The
effect of this on the ICF case would be to transfer a portion of
the residual processing capability to distillate processing.
Again the ICF product cost would decrease and there would be more
distillate product and less residual oil.
3. Product Slate
The projected future product demand from petroleum refineries
is listed in Table 7 with the 1980 product demand listed for
comparison (see Chapter VIII).[17] As shown in Table 7, the
future demand for gasoline is expected to decline. This decline
is largely compensated for by the rise in diesel fuel consump-
tion. The demand for jet fuel is expected to increase slightly
-------
-19-
Gasoline
Jet Fuel
Diesel Fuel
Kerosene
Distillate
Residual
Liquefied Gases
Other*
Table 7
Petroleum Product Demand [17]
1980
Million
Barrels /Day
6.8
1.1
1.2
0.2
2.0
2.4
0.8
3.2
Percent
38.6
6
6.8
1.1
11.4
13.6
4.5
18
2000
Million
Barrels/Day
5.1
1.6
3.4
0.2
1.2
1.4
0.7
4.9
Percent
27.6
8.6
18.4
1.1
6.5
7.6
3.8
26.4
* Other includes still gases, petroleum coke, asphalt and road
oil, lubes and waxes, special napthas, and miscellaneous products.
-------
-20-
while the demand for other distillate (fuel oil No. 2) and resid-
ual oil is expected to decrease. The percentage of total refinery
product represented by transportation fuels increases from 53
percent in 1980 to about 56 percent in 2000. The greatest change
lies in the G/D* ratio which changes from 1.5 in 1980 to about 0.8
in 2000.
The key products from refineries will continue to be trans-
portation fuels (gasoline, jet fuel, and diesel fuel) which repre-
sents 56 percent of the total refinery output in 2000. Therefore,
it is reasonable to require that coal liquid refineries produce at
least 56 percent transportation fuels. On the other hand, since
properties of coal syncrudes are significantly different from
petroleum crudes, syncrude refineries should not neces- sarily
have to meet exactly the same product slate as a petroleum
refinery. Therefore, a reasonable product slate for coal liquid
refineries will be developed below.
Coal liquids are not expected to be good feedstocks for jet
fuel production or diesel fuel production. For diesel fuel prod-
uction, coal liquid distillates need to be severely hydrotreated
to reduce the aromatic content to an appropriate amount (less than
25 percent aromatics) so that a product with a cetane number of
36-39 can be obtained.[18][19] These cetane numbers are still
lower than the minimum ASTM specification of 40; [20] therefore,
additives to boost the cetane number would be required. Similar-
ly, for jet fuel production the aromatic content must be less than
20 percent aromatics; thus, even more severe refining may be
required for jet fuel than for diesel fuel.[20] However, since
coal liquids are highly aromatic (i.e., high octane), they make
good feedstocks for gasoline production. Thus, if the coal liquid
refinery would not produce jet fuel, diesel fuel, or kerosene, it
would be reasonable to require it to produce at least 56 percent
gasoline, thus meeting the overall transportation fuel requirement.
The coal syncrudes contain from 35-60 percent of 350°-650° F
distillate. This could conveniently be used as fuel oil No. 2
after hydrotreatment. However, this is far greater than the year
2000 demand of only 6.5 percent fuel oil No. 2. Some of the coal
distillate along with some residual oil will have to be processed
into gasoline to meet the gasoline requirement. However, there
may still be more than 6.5 percent distillate product. Since the
coal liquid refineries may not produce much residual oil, because
some of it may be hydrocracked to gasoline and other portions may
be used for refinery fuel and feedstock for hydrogen production,
the year 2000 demand for resid of 7.6 percent could reasonably be
added to the 6.5 percent distillate demand to allow a total fuel
oil demand of 14 percent.
* G/D is the ratio of gasoline to distillate. Distillate
includes fuel oil No. 2, diesel fuel, jet fuel, and kerosene.
-------
-21-
Of the remaining 30 percent of the year 2000 product slate, 4
percent is represented by liquefied gases and 26 percent by the
"other" category (special napthas, asphalt, still gases, and
miscellaneous products).
As a result of the above discussion, the following year 2000
product slate for coal liquid refineries can be specified:
Liquefied Gases - Up to 4 percent
Gasoline - At least 56 percent
Distillate and Residual - Up to 14 percent
Other - 26 percent
The above product slate will be used as the basis for discus-
sing the product yields reported in the refining studies. The low
distillate demand (6.5 percent, or 14 percent if no resid is prod-
uced) may be the most difficult specification for coal liquid
refineries to meet. The product slates from the coal liquid
refineries have already been presented in Table 3.
The Chevron refinery produces 100 percent gasoline. Although
this is more than what is required, this case does provide an
upper limit for processing costs and is certainly reasonable with
respect to the production of transportation fuels.
The UOP refinery output is designed to meet a G/D ratio of
2.0, and the only products are gasoline, No. 2 fuel oil and LPG.
Sixty-three percent of the total product is gasoline. Therefore,
this is more gasoline than required. The only other major product
is fuel oil No. 2 (about 30 percent). This is about twice as much
as the year 2000 demand for distillate or residual. However,
because of the high yield of gasoline from this refinery, its
overall product slate can be considered reasonable.
For the ICF cases, the product slates are as follows:
H-Coal EDS SRC-II
Gasoline (%) 31 37 16
Fuel Oil No. 2(%) 43
Residual (%) 26 63 84
The high quantity of residual produced from these refineries is
considered unacceptable since future demand is projected to be
only 7.6 percent and because residual is not a premium product.
To produce an acceptable product slate, the ICF refineries would
need to become more integrated, thus the capital cost for their
refineries would increase significantly.
-------
-22-
The conclusions drawn from the above discussion are summariz-
ed below:
1. The Chevron/SRC-II refinery represents an upper limit
for refinery processing cost with respect to product slate since
100 percent gasoline is the product.
2. The UOP/H-Coal product slate is considered reasonable.
3. The product slate from the ICF refineries is unaccept-
able because of the large amount of residual product. These
refineries would need to become more integrated in order to meet
an acceptable product slate, thus the capital cost for their
refineries would increase significantly.
4. Product Qualities
All of the products from the UOP and Chevron refineries meet
ASTM product specifications with the exception of the fuel oil No.
2 for the UOP case. The API gravity of this fuel oil is slightly
lower than the specification.
The gasoline from the ICF cases is simply hydrotreated naptha
which probably does not meet automotive octane requirements.
Laboratory tests have indicated that research octane numbers for
EDS, H-Coal, and SRC-II hydrotreated napthas lie between
65-70. [12]. This is far below the present 87 (R + M)/2 octane
rating for unleaded gasoline. Catalytic reforming of the hydro-
treated naptha would be necessary to meet an 87 (R + M)/2 octane
rating. The addition of a catalytic reformer would increase the
capital costs of the ICF refineries. This would result in an
increased product cost.
To conclude this brief discussion, it is noted that upgrading
coal liquids to acceptable end products is technically feasible.
However, although coal or refinery products meet ASTM specifica-
tions, product properties can still be significantly different
from petroleum refinery products.[7]
5. Coal Liquid Refining Costs
Based on the above discussions of grass root coal liquid
refineries, representative refining costs can now be determined.
Since the Chevron/SRC-II and UOP/H-Coal studies were based on a
higher level of engineering design, their cost estimates will be
used as the initial basis to determine refining costs for the
SRC-II and H-Coal syncrudes. The ICF/EDS case will be used for
the initial EDS syncrude refining cost.
Table 8 lists a breakdown of the investment and operating
costs in first quarter 1981 dollars for the coal liquid refin-
eries. The operating costs do not include the cost of the
syncrudes. Table 9 lists the refining costs based on the costs
-------
-23-
Table 8
Investment and Operating Costs for Refineries
(Millions of First Quarter 1981 Dollars)
Inve s tment Costs,
Millions of Dollars
Onplot Investment
Offplot Investment
Prepaid Royalties
Contingency
Initial Catalyst and
Chemicals
Total Instantaneous
Investment
Working Capital
Total Capital Investment
SRC-II
Chevron[5]
455.6
227.8
-
75.9
21.6
780.9
108.2
889.1
H-Coal
UOP[7]
189.2
218.9
-
40.8
5.5
454.4
76.5
530.9
EDS
ICF[8]
261.7
22.7
1.6
42.7
—
328.7
26.7
355.4
Operating Costs, Millions
of Dollars per year
Interest on Working
Capital 6.5 4.6 1.6
General and Administrative - 3.7 3.3
Taxes and Insurance 17.4 11.4 9.9
Maintenance 16.3 9.1 12.2
Catalyst and Chemicals 5.5 13.0* 6.4
Labor 4.4 - 1.8
Utilities 7.6 - 9.47
Refinery Fuel - - -
Hydrogen Plant - - 107.6
Overhead - - 6.1
Other - - -
Total Annual Operating
Costs 57.7 41.8 158.4
Includes catalyst and chemicals, labor, and utilities.
-------
-24-
Table 9
Refining Cost of Coal Liquids
Chevron/SRC-II
UOP/H-Coal
ICF/EDS
Revised EDS Cost
11.5 Percent CCR
$/FOEB
9.76
5.75
11.03
7.87
$/mBtu
1.84
1.06
1.87
1.47
30 Percent CCR
$/FOEB
20.13
11.41
15.22
16.02
$/mBtu
3.80
2.10
2.58
3.00
Hydrogen
Consumption
scf/bbl
2633
1150
1728
-------
-25-
reported in Table 8 and on capital charge rates of 11.5 and 30
percent. The refining costs are added to the direct liquefaction
product costs and the distribution costs of the synthetic products
to obtain the total product cost. The revised "EDS refining cost"
listed in Table 9 will be discussed in the next section.
It must be understood that the Chevron/SRC-II refinery repre-
sents an upper limit to the refining cost since it produces 100
percent gasoline. The UOP/H-Coal refinery would need to be more
integrated to enable it to process the greater quantity of resi-
dual oil in the H-Coal feedstock as reported by Fluor Corpora-
tion. Therefore, the UOP/H-Coal refining cost is likely to be
low. It is difficult to determine whether the ICF/EDS refining
cost estimate is high or low. Based on Exxon's analysis of the
EDS syncrude, the ICF/EDS refinery would not have to handle the
quantity of residual oil they assumed; this would result in lower
refining costs. However, to meet an acceptable product slate
their refinery would need to become more integrated, thus increas-
ing the refining cost.
F. Reconciliation of EDS, H-Coal, and SRC-II Refining Cost
Differences
Table 8 indicates that there are significant differences
between the capital investment and operating costs for the SRC-II,
H-Coal, and EDS refineries. The total instantaneous capital
investments (including working capital) is $889 million for the
SRC-II refinery, $531 million for the H-Coal refinery and $355
million for the EDS refinery. Much of the capital and operating
cost differences can be attributed to product slate and quality,
level of engineering design and feedstock properties. In the
following paragraphs these differences will be reconciled.
With respect to the level of engineering design, there is a
greater probability that the actual capital cost of a project will
be more than the estimated cost rather than less as the level of
engineering design decreases.[21] For example, a design study
prepared to estimate the capital cost to within 30 percent of the
actual cost is likely to have a wider positive error for the
estimate than negative, eg., +40 and -20 percent. The Chevron
study represents the highest level of engineering design between
the three studies followed by the UOP study, and then the ICF
study. On this basis it is believed that the ICF estimate has the
greatest probability of being lower than the actual capital cost
followed by the H-Coal and SRC-II studies.
It is believed that ICF's estimate of the offplot investment
for the EDS refinery is too low. Offplot investment amounts to 30
percent of the total instantaneous investment for the Chevron
study and 48 percent for the UOP study, but only 7 percent for the
ICF study. This finding is also confirmed by typical offplot
refinery costs. [6] Thus, the resulting ICF/EDS refining costs
will also be low.
-------
-26-
The high operating cost of the EDS refinery relative to the
H-Coal and SRC-II refineries as reported in Table 8 is primarily
the result of the EDS refinery purchasing natural gas for hydrogen
production. The H-Coal and SRC-II refineries produce hydrogen
directly from the coal syncrude or from the light refinery prod-
ucts. The latter process is much more likely to occur in real
life since it utilizes low quality products rather than high qual-
ity products to produce hydrogen.
With respect to product slate, the Chevron refinery produces
100 percent gasoline while the H-Coal refinery produces 2/3 gaso-
line and 1/3 fuel oil No. 2, and the EDS refinery produces about
40 percent hydrotreated naptha and 60 percent residual fuel. The
SRC-II refinery which produces 100 percent gasoline will consume
the largest amount of hydrogen. The total hydrogen consumption
for the SRC-II refinery is 2633 scf per barrel of syncrude,
whereas the H-Coal/UOP refinery consumes about 1150 scf per
barrel, and the EDS refinery about 1728 scf/barrel. Since capital
and operating costs are generally proportional to hydrogen
consumption, one would expect the Chevron product cost to be the
most expensive, followed by the EDS cost, and then the H-Coal
cost. Table 9 shows that this is the general trend in the refin-
ing costs. Although much of the differences in hydrogen consump-
tion is due to product slate, it must be noted that feedstock
properties also have a significant effect on hydrogen require-
ments.
If the refinery product slate and other parameters were held
constant, the general effect of feedstock properties on refining
costs can be determined. The refinery feedstock property data are
reported in Tables 4 and 5. The most important feedstock property
data to compare are the nitrogen, oxygen, and hydrogen contents.
Tables 4 and 5 show that the H-Coal oil has the most desirable
properties of the three syncrudes, followed by the EDS crude, and
then the SRC-II crude. High nitrogen and oxygen contents, and a
low hydrogen content account for the high severity hydrotreating
and hydrocracking required for the coal syncrudes, and this high
severity refining correlates directly with refining cost. The
theoretical hydrogen requirement necessary to bring the hydrogen,
nitrogen, oxygen, and sulfur levels of the EDS and SRC-II
syncrudes up to the quality of the H-Coal oil is 248 scf per
barrel for the EDS crude and 469 scf/bbl for the SRC-II crude.
This indicates that the SRC-II refinery would require the highest
refining cost followed by the EDS and H-Coal refinery. This is
roughly confirmed by the cost estimates shown in Table 9.
There is much uncertainty in the refining cost based on the
ICF/EDS refinery. As discussed earlier, reasons for this uncer-
tainty include: 1) a low level of engineering effort resulting in
uncertain capital and operating costs, 2) the poor quality of the
EDS feedstock assumed by ICF relative to that most recently
reported by Exxon, and 3) the highly unreasonable product slate
-------
-27-
from this refinery. For these reasons a cost estimate based on
the SRC-II and H-Coal refineries and on the overall quality of the
EDS feedstock relative to the H-Coal and SRC-II feedstocks will be
determined and used in preference to the ICF-based estimate. The
measure of feedstock quality will be based on the theoretical
hydrogen requirement necessary to bring the hydrogen, nitrogen,
oxygen and sulfur levels of the EDS and SRC-II syncrudes up to the
quality of the H-Coal oil. As discussed above this requirement is
248 scf/bbl for the EDS crude and 469 scf/bbl for the SRC-II
crude. Therefore, on a linear scale the EDS crude lies 52.9
percent (248/469) of the scale between the other crudes. On this
basis the refining cost for the EDS process can be approximated by
linearly interpolating between the refining cost shown in Table 9
for the H-Coal and SRC-II crudes. The resulting cost is presented
in Table 9 as the "revised EDS cost." This cost ($1.47 -
$3.00/mBtu) will be used in preference to the ICF-based estimate
($1.87 - $2.58/mBtu).
In this section studies which have presented refining costs
for coal syncrudes have been critiqued. One must realize that
these refining costs are the best available at the present time.
There are shortcomings with respect to level of engineering
effort, in addition to uncertainties in feedstock qualities and
product slates. These problems make it difficult to compare cost
between studies. It would be desirable if all costs were on a
similar basis and derived from a high level of engineering design.
In conclusion the refining costs of whole SRC-II, EDS and
H-Coal syncrudes in grass-root refineries has been determined to
range from $1.84-3.80, $1.47-3.00, and $1.06-2.10 for each process
respectively. The Chevron/SRC-II refining cost represents an
upper limit to the refining cost since it produces 100 percent
gasoline. The UOP/H-Coal refining cost may be low since it would
need to be more highly integrated to enable it to process a great-
er quantity of residual in the H-Coal feedstock than they assum-
ed. Lastly, the EDS cost is based on the SRC-II and H-Coal costs
using the overall quality of the crude feedstock as an indicator
of refining cost.
IV. Analysis of an Integrated Coal Liquid/Petroleum Refinery[6]
A. Introduction
Gilder and Burton (UOP Inc.) evaluated the economic feasibil-
ity of co-processing H-coal liquid and a petroleum crude in a
typical large petroleum refinery located on the Gulf Coast with a
refinery capacity of 285,000 barrels per calendar day (BPCD).[6]
They analyzed the co-processing of 3, 5, and 10 percent H-Coal
liquid charges to the refinery. Only the 10 percent H-Coal liquid
charge will be discussed in this report, as even at this H-Coal
feedrate (28,500 BPCD) two such refineries would be needed to
handle the output of a typical 50,000 BPCD liquefaction plant.
-------
-28-
Linear programming techniques were used to provide material
balances, capital costs, utilities, and operating cost informa-
tion. The estimated investment costs reflect prices that were
derived by scaling detailed estimates prepared for similar units
to UOP standards and specifications.
Their first step was to design a grass-roots petroleum refin-
ery. Then, to co-process the H-Coal liquid in the petroleum
refinery the following questions were asked:
1. Can the operating severity of the petroleum refinery
process units be increased in order to produce an acceptable prod-
uct?
2. What modifications to existing units are required to
facilitate the processing of H-Coal liquids?
This discussion summarizes their work.
B. Feedstocks
The coal liquid modeled was a 64 to 880°F distillate
material obtained from the atmospheric column overhead and bottoms
products of the 3 ton per day H-Coal Process Development Unit
located in Trenton, New Jersey. An analysis of the H-Coal liquid
has already been presented in Table 4.
The petroleum crudes were selected to represent typical Gulf
Coast processing. Louisiana Delta and West Texas Sour are the
domestic crudes while a 65/35 Light/Heavy Arabian crude is the
foreign crude. Analyses of these crudes are presented in Tables
lOa, lOb, and lOc. When added, the H-Coal syncrude took the place
of a portion of the foreign crude.
C. Product Slate and Specifications
All refinery products met ASTM specifications (except for LPG
which was only required to meet a RVP restriction.) The two
refineries were not required to meet identical product slates, but
they did meet similar slates, as shown in Table 11.
D. Petroleum Refinery Configuration
The model of the petroleum refinery configuration (in which
H-Coal liquid is not processed) is a typical Gulf Coast refinery
representative of the complexity and processing flexibility
exhibited in the larger refineries of that area. Figure 4
presents a flow diagram of the major processing units utilized in
this all-petroleum refinery. A discussion of this flow scheme
will not be included here. Interested readers are referred to the
referenced study.[20]
-------
-29-
E. H-Coal Liquid/Petroleum Refinery Configuration
A diagram depicting the integrated H-Coal/petroleum refinery
design is shown in Figure 5. In this refinery the H-Coal liquid
(10 percent of total petroleum crude to the refinery) is co-mingl-
ed with the Arabian Crude and then fractionated. The resulting
fractions are processed in the same general manner as in the
petroleum refinery. However, because of the characteristics of
the H-Coal liquid, the petroleum refinery had to be modified to
include an additional high-pressure naptha hydrotreater (see
Figure 5). Operating conditions, product properties, and product
yields for each process unit were adjusted to reflect the feed-
stock quality differences in the various cases.
The H-coal liquid contains much less residual material than
the Arabian crude; however, processing to obtain acceptable inter-
mediate products (gasoline, diesel, etc.) from the H-Coal/Arabian
crude mixture is more severe than for the Arabian crude alone.
The H-Coal liquid properties which have the greatest impact on its
ability to be processed are its high nitrogen and oxygen contents,
and low hydrogen content.
F. Material Balance
The overall feed and product summary is presented in Table
11. Because the H-Coal material contains high concentrations of
naptha and distillate, and a low concentration of resids (unlike
the Arabian crude), there is a shift from atmospheric bottoms
processing (fluid catalytic cracking, vacuum distillation, etc.)
to distillate processing (distillate hydrotreating). Table 12
shows the processing shifts required to achieve the stated
yields. Utility usage is listed in Table 13. The refinery is
self-sufficient with respect to utilities except for power.
G. Economic Comparison
In this subsection the cost of refining 10 percent H-Coal
liquid in the integrated refinery will be compared to processing
100 percent petroleum. The economic basis used here is the same
as that used for the grass-root refineries except:
1. No by-product credits were taken
2. A power cost of 4.5^/kw-hr was used
3. Interest on working capital was not charged as an
operating cost.
4. Depreciation was included as an operating cost.
5. The refineries were not scaled to 54,500 FOEB/CD of
product.
-------
Table lOa
Volume % of Crude
Weight % of Crude
Specific Gravity
Sulfur, Wt. %
Research Octane
Number
Motor Octane
Number
Smoke Point
Pour Point °C
Viscosity Index
at 50°C
Conradson Carbon
Wt-%
Paraffins, Vol-%
Waphthenes, Vol-%
Aromatics, Vol-%
Nitrogen, Wt-%
Arabian Blend Crude
65% Light/35% Heavy[6]
C3
0.7
0.4
0.508
C4 C5-80
1.7 6-8
1.17 5.2
0.57 0.6617
0.01
63.
62.
85.5
10.7
3.8
80-190
17.8
15.5
0.7504
0.04
31-
29.
64.9
20.3
14.8
190-250
9.0
8.4
0.8025
0.23
17.
16.
23.
-48.3
6.7
57.7
21.1
19.7
250-340
15.0
14.6
0.8453
1.19
8.
-17.8
15.5
20.4
9.3
2.5
340-565 565+ Total
31.6 17.9
33.3 21,5 100
0.9123 1.035 0.8645
2.46 5.0 2.09
26.7 48.9
28.9 48.2
0.614 22.85 5.1
i
OJ
o
0.01
0.07
0.31 0.091
-------
-31-
Table lOb
Louisiana Mixed Delta Crude[6]
^3 £4 C5-80 80-193 193-250 250-340 340-565 565+
Volume % of Crude 0.3 0.6 2.6 16.5 9.8 27.4 35.1 7.7
Weight % of Crude 0.2 0.4 2.0 14.4 9.3 27.3 37.3 8.9
Specific Gravity 0.508 0.57 0.671 0.754 0.820 0.864 0.912 1.0
Sulfur, Wt. % 0.18 0.04 0.11 0.38 0.80
Research Octane
Number-Clear 89. 66. 73.
Motor Octane
Number-Clear 92. 65. 75.
Smoke Point 19.7
Pour Point Index 0.35 6.0 100 1000
Viscosity Index
at 50°C 7.52 17.12 32.04 42.42
Conradson Carbon
Wt-% 0.47
Paraffins, Vol-% 100 59.9 38.6
Naphthenes, Vol-% 35.3 50.2
Aromatics, Vol-% 4.8 11.2
Nitrogen, Wt-%
-------
-32-
Volume % of Crude
Weight % of Crude
Specific Gravity
Sulfur, Wt. %
Research Octane
Number
Motor Octane
Number
Smoke Point
Pour Point Index
Viscosity Index
at 50°C
Conradson Carbon
Wt-%
Paraffins, Vol-%
Naphthenes, Vol-%
Aromatics, Vol-%
Nitrogen, Wt-%
Table lOc
West Texas Sour Crude[6]
C3
0.5
0.5
0.508
C4 C5-80
0.2 7.5
0.2 5.8
0.57 0.670
0.12
70.2
66.5
80-190
22.8
20.6
0.775
0.31
57.8
190-250
10.3
9.9
0.825
0.74
30
0.35
37.8
37.6
24.6
8.32
250-340 340-565 565+
16.3
16.5
0.868
1.37
6.0
16.7
0.07
29,
31,
0.928
2.19
12.8
15.0
1.0
3.95
100 1000.
32.04 42.24
0.26 19.35
-------
-33-
Table 11
Feed and Product Summary[6]
Yields, BPCD 100% Petroleum 10% H-Coal
Feedstocks
Louisiana 85,500 85,500
West Texas Sour 85,500 85,500
Arabian Blend 114,000 85,500
H-Coal Liquid 0 28,500
Butanes 12,255 10,260
Total 297,255 295,260
Products
Unleaded Regular 93,195 91,485
Unleaded Premium 39,900 39,330
Jet Fuel 19,950 19,950
No. 2 Fuel Oil 93,765 98,325
No. 6 Fuel Oil 7,125 6,840
LPG 17,385 15,390
Total 271,320 271,320
-------
FIGURE V
GULF COAST REFINERY
BASE CASE
4
vsss _J AMiNi
ENDS —MIBEATINO UNII
k
SUlflJH
PI AMI
FUEL OAS
OIIISIANA
CHUOE •
Oil
I
SI If HAS
CR
on
AN BLENO .
IDt Oil
L>. 1O
AINVIAIION
MEROSINE
. MtflOX
H
HiSID IO f Uil
' Oil BUNDING
NO. i
NO. a
1
1 _
V
• »»
HI nr.in INIIS I
NFROMI
J coHcn ' F^
-^•BFSIDTOFUEl
Oil BIENOINO
soirun
t
f
u
-*
1
J
— 1
1
^L^
•' IIGM? „ LIGHT
"J ENDS Ml ENDi
* 4 * 4
^NAPHIU IJ MAT I
NO. 1 1 r| NO. 1 |
1
M, tlGllf u LIGHT 4
1 ENDS * ENDS 1
It 44
.1 NAPM NT IJ rtAf 1 , , ! ft
\ 1
1 1
I
C,S-fr
C^'l*^
1
IPO
OASOIINES
70% UNLEADED
HEGUIAB
M% UNIEAOEO
PREMIUM
JET FUIl
*>
I
1
n
•
•
l
-
• •>
I i
.
J
FCC NO. 1
OAS CON
1 L
1
1
1
^AIKUAIION t
UNIT | '
k|FCCOASOHNEl ^
•'I MEHOX f ' ^
t
* * 1
rut i oua
NO. I
NO. •
-------
FIGURE 5-
GULF COAST REFINERY
RAW H COAL LIQUID CASE
IN"! _^ *M"« U 80
ENDS » IREATINOUNlim «
LfUEl OAS
4
IPO
RECOVERY
.
. NERO
OUISIANA MEt
Oil
^-~-'/ |^^^^
T ri
I LI
y
^*Oll BIENO
1
.-A.
Oil 1
Tl *l
11
kl
Y
f .
1
N UISNO — >
•II LUAL ^ ^ j ^
T PI
1 U -r-
|
LFUR 1
1
• « •» loiro
••ipfr IO H'G
AlKVlAllON
SINE 1 .
OX I""*
OH 1
NO |
HI IIGIII ENDS
* 4
i
1
H] •'<•"!
1 rTs
i i
u IIGIII
"1 INOS '
* *
M IIOHI
H] (NOt
4 t
i
UCHT 1
tNOS ^ 1
-* coKin
1 .. IICI" .. IIOMT
' ^v "1 ENDS Ml END!
J NAPHHI IJ flAf 1
1 F| NO 1 1 *| NO. 1 f~
1
T
1 1
XH, UCHT M,
I ENDS M>
it *
>>. k| NAPHHI 1 .J f
i*l NO. i \y\ *
^-^ 1 l
1 1
** ,
I»
1
.
1 1
1
1 _ 1 „, UOHT
— • K 1 HI ENO<
- . I .... k A A
r ft
• • A NAPtlUT 1
F| NO. 3 |
UCIIIErjDal t\ *l"»l*"ON 1
1 KCNO 1 fc|»CI
'1 GASCON *|
I 1
1
Oil BUNUINO
UOHT
ENDS
f
iiAT
„,
•
: GASOU
MEMO*
WEI |
f
1
\
'
.
1
i
1 I
1 1
[ 1
'
'
c
«s-fr
^
— — J^
suLfun
IPO
OASOLINES
JET rUEL
run oil s
10% UNICAOED
REGULAH
Vm, UNIEAOEO
PREMIUM
JET Al
NO. 1
Ul
I
HOP 485 I
-------
-36-
Table 14 lists the investment costs and operating costs for
both the 285,000 BPCD petroleum refinery and the 285,000 BPCD
integrated refinery. The total instantaneous investment (exclud-
ing working capital) in first quarter 1981 dollars is $1,383
million for the petroleum refinery and $1,423 million for the
integrated refinery. This $40 million investment difference
reflects the cost of retrofitting an existing refinery and thus
enabling it to co-process 10 percent H-Coal liquid. Part of the
capital cost increase for the H-Coal case is due to necessary
increases in the capacities of catalytic reforming, kerosene
hydrotreating and diesel hydrotreating. The annual operating
costs of the integrated refinery are $18 million less than those
of the petroleum refinery because of the lower feedstock butane
requirement. Part of the lower butane requirement is due to the
reduced HF (hydrofluoric acid) alkylation requirement for the
integrated refinery; also less butane is needed for gasoline
blending on account of the lower gasoline yield for the integrated
case. These investment and operating cost differences essentially
cancel each other with respect to product cost, as can be seen in
Table 15. The product cost/FOEB is nearly the same for both the
petroleum and integrated refineries.
V. Economic Comparison of Grass Root vs. Integrated Coal Liquid
Refineries
A. H-Coal
The instantaneous capital cost of retrofitting the integrated
H-Coal/petroleum refinery is $40 million. The operating cost
savings is $18 million. These costs are based on feeding the
refinery with 28,500 BPCD of H-Coal liquid which represents 10
percent of the total refinery feedstock. Since liquefaction
plants are being designed to produce on the order of 50,000
FOEB/CD of syncrude, approximately two typical large refineries
(285,000 BPCD) must be retrofitted in order to co-process H-Coal
liquid. Therefore the capital cost increase of retrofitting two
existing refineries, each processing 28,500 BPCD of H-Coal liquid,
would be $80 million. Likewise the operating cost savings result-
ing from retrofitting two large petroleum refineries would be $36
million. The overall effect of retrofitting on product cost is
negligible since the increased capital cost is balanced by a
decreased operating cost.
A capital cost comparison between retrofitting existing
refineries to co-process 57,000 FOEB/CD H-Coal versus building new
refineries to process 100 percent H-Coal (57,000 FOEB/CD) shows
that there is a $390 million capital cost savings for the retro-
fitting alternative. However, the UOP study upon which these
figures are based, analyzed a new grass-roots petroleum refinery
with modern technology which was then retrofitted to co-process 10
percent H-Coal liquid. The great majority of the actual refine-
ries being retrofitted will not be new since the United States has
-------
-38-
Table 13
Utility Summary[6]
Electric Power, kw
Steam (42 k/cm2) ton/hr
Steam (10.5 k/cm2) ton/hr
Steam (3.5 k/cm2) ton/hr
Cooling Water, M^/hr
Fuel mBtu/hr
100% Petroleum
31,100
288
ir 177
r 111
11,800
5,754
10% H-Coal
29,900
261
161
100
11,000
5,674
-------
-39-
Table 14
Investment and Operating Costs
(Millions of First Quarter 1981 dollars)
100% Petroleum 10% H-Coal
Investment Cost
(Millions of Dollars)
Onplot Investment 552 568
Offplot Investment 691 711
Contingency 124 128
Initial Catlyst and Chemicals 16 16
Total Instantaneous Investment 1383 1423
Working Capital 436 436
Total Instantaneous Capital 1819 1859
Investment
Operating Cost (Millions
of Dollars per Year)
Butanes ($32.70/bbl) 146 123
Labor, Catalyst, Chemicals, 74 75
and Utilities
Maintenance, Taxes, Insurance, 165 170
GA and Depreciation
Total Operating Cost 385 367
-------
-40-
Table 15
Refining Costs
(Millions of First Quarter 1981 dollars)
Capital Charge Rate, %
(Millions of Dollars)
Refining Cost
$/FOEB
$/mBtu
100% Petroleum
11.5
Total Instantaneous Investment 1383
Total Adjusted Capital Investment 1568
Annual Capital Charge 180
Annual Operating Cost 386
Total Annual Charge 560
6.14
1.04
30
9.20
1.56
10% H-Coal
11.5
5.97
1.01
30
1383
1543
463
386
849
1423
1614
186
367
552
1423
1588
476
367
843
9.11
1.55
-------
-41-
excess refining capacity; the retrofitted refineries will be older
refineries, and the cost of retrofitting the older refineries will
probably be more than the cost of retrofitting the new grass-root
petroleum refinery modeled by UOP. Therefore the $80 million
retrofitting capital cost for co-processing 57,000 FOEB/CD of
H-Coal represents a minimum cost. Likewise the operating cost
resulting from the retrofitted new refineries also represent a
minimum cost. Thus the overall refining cost for processing coal
liquids in retrofitted refineries will be more for older petroleum
refineries than for new ones.
B. SRC-II
There has not been any work examining an integrated SRC-II
petroleum refinery. However, if the results obtained from the
H-Coal processing are applied to the SRC-II syncrude by using the
ratio of the H-Coal retrofitting cost to grass-root refinery cost,
a cost estimate may be obtained for retrofitting two existing
refineries enabling them to co-process 10 percent SRC-II syncrude
(57,000 FOEB/CD). The instantaneous retrofitting cost calculated
by following this procedure is $138 million (1 Q 1981) which is
$670 million less than the total instantaneous capital cost of
building a new grass-roots, 100 percent SRC-II refinery. It must
be noted that the capital cost of the SRC-II grass root refinery
is based on a 100 percent gasoline product .slate and therefore
represents an upper limit to the capital cost.
C. EDS
Because of the uncertainty of the capital cost for the
refining of EDS coal liquids developed by ICF, it is difficult to
approximate a retrofitting capital cost. However, since the qual-
ity of the EDS syncrude is between the quality of the H-Coal and
SRC-II crudes, it is expected that the retrofitting cost would
also be between those for H-Coal and SRC-II.
VI. Conclusion
The Chevron/SRC-II and UOP/H-Coal studies, together with the
revised EDS cost have been found to best represent refining costs
for the grass-root coal-liquid refineries. The refining costs
vary from about $1.00 per mBtu for the H-Coal syncrude to about
$2.00 per mBtu for the SRC-II syncrude when using a capital charge
rate of 11.5 percent; they vary from about $2.00 per mBtu for the
H-Coal syncrude to $4.00 per mBtu for the SRC-II syncrude when
using a capital charge rate of 30 percent. The wide variance in
the refining costs of the syncrudes is due to differences in feed-
stock properties, product slates, and level of engineering
design. Based on feedstock properties only, the refining cost of
the SRC-II syncrude is expected to be the most expensive, followed
by that of the EDS crude and then the H-Coal crude.
-------
-42-
The cost of co-processing a 10 percent H-Coal/90 percent
petroleum crude blend in a typical large refinery was discussed
and compared to the cost of refining 100-percent petroleum. The
refining costs are nearly identical and vary from about $1.00 to
1.50 per mBtu depending on the capital charge rate.
The cost of retrofitting existing refineries to process
syncrudes was estimated to be $80 million for H-Coal and $138
million for SRC-II. Capital cost savings resulting from process-
ing the synthetic crudes in existing petroleum refineries rather
than processing them in new grass root coal liquid refineries
range from $390 million for the H-Coal syncrude to $670 million
for the SRC-II syncrudes.
When comparing the refining cost of syncrudes from a new
grass-roots refinery to those costs from a retrofitted petroleum
refinery, it appears that their refining costs are nearly the
same. For example, the H-Coal refining cost varies from
$1.00-2.00/MBtu for a grass-roots refinery to $1.00-1.50/mBtu for
a retrofitted refinery. However, of most importance is the
savings in capital investment when refining syncrudes in retro-
fitted petroleum refineries.
-------
-43-
References
!• "Identificaton of Probable Automotive Fuels
Composition: 1985-2000," Prepared by SwRI for DOE,
HCP/W3684-01/1.
2. Oil and Gas Journal, pp. 22, 45, 79, June 1, 1981.
3. Chemical Engineering, pp. 10, 75, 88, May 18, 1981.
4. "Aviation Turbine Fuels from Shale and Coal Oils,"
Shaw, Kalfadelis, Johnig, ER&E, Energy and Environment, 3rd Nat.
Conf., Sept. 29-Oct. 1, 1975, Hueston Woods State Park, Ohio,
AICHE/ENE-E-P3.
5. "Cost Comparisons of Alternative Routes for Converting
SRC-II Oil to Distillate Fuels," Sullivan, Frumkin, Chevron
Research Co., Presented at the EPRI Conference, San Francisco CA,
October, 13-16, 1980.
6. Coal Liquid/Petroleum Refinery Integration, Gilder,
Burton, UOP Inc., American Petroleum Institute, Preprint No.
38-80, May, 1980.
7. "Crude Oil vs. Coal Oil Processing Comparison Study,"
Conser, Garret, Weiszmann, UOP, Presented at AICHE, Boston, MA,
August, 19-22, 1979.
7a. "The H-Coal and SRC-II Processes," Reiser, D.,
EPA-AA-SDSB-81-14, February 1982.
8. Methanol From Coal: Prospects and Performance as a
Fuel and as a Feedstock, December 1980, IGF Incorporated.
9. Feasibility Study of Alternative Fuels for Automotive
Transporatation, Exxon Research and Engineering Company, June
1974, prepared for EPA, EPA-460/3-74-009-a,b,c.
10. Alternate Energy Sources for Non-Highway
Transportation, Exxon Research and Engineering Company,
DOE/CS/05438-T1, June 1980.
11. "Project H-Coal," Hydrocarbon Research, Inc., 1967;
Office of Coal Research Contract No. 14-01-0001-477.
12. Analytical Studies for the H-Coal Process, Mobil
Research and Development Corp., Prepared for DOE, FE-2676-1,3.
13. "Upgrading of Coal Liquids," Annual Report, UOP Inc.,
Prepared for DOE, FE-2566-42.
14. EDS Coal Liquefaction Process Development, Phase V, EDS
Commercial Plant Study Design Update, FE-2893-61, March 1981.
-------
-44-
15. "The SRC-II Process," Presented at a discussion meeting
on "New Coal Chemistry," Schmid, Jackson, Organized by The Royal
Society, London, England, May 21-22, 1980, The Pittsburgh Midway
Coal Mining Company.
16. "Engineering Evaluation of a Conceptual Coal Conversion
Plant Using the H-Coal Liquefaction Process," Prepared by Fluor
Engineers and Constructors, Inc. for the Electric Power Research
Institute, EPRI-AF-1297, December, 1979.
17. Energy Review, Data Resources, Inc., Vol. 4, No. 1,
Winter, 198CL
18. Refining and Upgrading of Synfuels from Coal and Oil
Shales by Advanced Catalytic Processes, Third Interim Report,
Processing of SRC-II Syncrude, Chevron Research Co., Prepared for
DOE, FE-2315-47, March 1980.
19. "Exxon Donor Solvent Coal Liquefaction Process Develop-
ment Program Status III," Epperly, Plumlee, Wade, Exxon Research
and Engineering Co., Presented at EPRI Conference on Synfuels, San
Francisco, CA, October, 13-16, 1980.
20. "American Society for Testing and Materials, Standard
Classification of Diesel Fuels Oils (D975)," 1978 Annual Book of_
ASTM Standards, Part 23, Philadelphia, Pa., Nov. 1977, pp.449
21. Plant Design and Economics for Chemical Engineers,
Peters, Timmerhaus, 2nd Edition, 1968, McGraw-Hill.
------- |