A comparison of direct and indirect
liquefaction technologies for making fluid
fuels from coal
Robert H. Williams and Eric D. Larson
Princeton Environmental Institute, Princeton University
Guyot Hall, Washington Road, Princeton, NJ 08544-1003, USA
E-mail (Williams): rwilliam@princeton.edu
Direct and indirect liquefaction technologies for making synthetic liquid fuels from coal are com-
pared. It is shown that although direct liquefaction conversion processes might be more energy-
efficient, overall system efficiencies for direct and indirect liquefaction are typically comparable
if end-use as well as production efficiencies are taken into account. It is shown that some synfuels
derived via indirect liquefaction can outperform fuels derived from crude oil with regard to both
air-pollutant and greenhouse-gas emissions, but direct liquefaction-derived synfuels cannot. De-
ployment now of some indirect liquefaction technologies could put coal on a track consistent with
later addressing severe climate and other environmental constraints without having to abandon
coal for energy, but deploying direct liquefaction technologies cannot. And finally, there are much
stronger supporting technological infrastructures for indirect than for direct liquefaction tech-
nologies. Prospective costs in China for some indirect liquefaction-derived fuels are developed but
not costs for direct liquefaction-based synfuels, because experience with the latter is inadequate
for making meaningful cost projections. Especially promising is the outlook for the indirect lique-
faction product dimethyl ether, a versatile and clean fuel that could probably be produced in China
at costs competitive with crude oil-derived liquid fuels. An important finding is the potential for
realizing, in the case of dimethyl ether, significant reductions in greenhouse gas emissions relative
to crude oil-derived hydrocarbon fuels, even in the absence of an explicit climate change mitigation
policy, when this fuel is co-produced with electricity. But this finding depends on the viability of
underground storage of H2S and CO2 as an acid gas management strategy for synfuel production.
Many ‘‘megascale’’ demonstration projects for underground CO2 storage and H2S/CO2 co-storage,
along with appropriate monitoring, modeling, and scientific experiments, in alternative geological
contexts, are needed to verify this prospect. It is very likely that China has some of the least-costly
CO2 sources in the world for possible use in such demonstrations. It would be worthwhile to
explore whether there are interesting prospective demonstration sites near one or more of these
sources and to see if other countries might work with China in exploiting demonstration oppor-
tunities at such sites.
1. Introduction
China, with its rapidly growing demand for transportation
fuels, scant domestic oil and natural gas resources but
abundant coal, is likely to turn to coal as a basis for pro-
viding synthetic fluid fuels for transportation, cooking,
and other applications that are not easily served by elec-
tricity.
Two very different approaches to providing fluid fuels
from coal are described and compared in this paper: direct
coal liquefaction (DCL) and indirect coal liquefaction
(ICL). For both approaches a major challenge is to in-
crease the hydrogen-carbon ratio. For finished hydrocar-
bon fuels such as gasoline and diesel, H/C ~ 2 (molar
basis). For petroleum crude oil, the ratio ranges from 1.3
to 1.9. For typical bituminous coals, H/C ~ 0.8.
Making a comparison of DCL and ICL technologies is
not an easy task because of the very different stages of
development for these two classes of technologies. ICL
technologies (Fischer-Tropsch (F-T) liquids, methanol
(CH3OH or MeOH) and dimethyl ether (CH3OCH3 or
DME)) are either commercially proven or made up of
proven modules, and there is an extensive literature on
these technologies and modules. In contrast, DCL tech-
nologies are not yet commercially proven, and informa-
tion available in the public domain is limited -- with quite
different findings coming from the few assessments that
have been made. Despite this difficulty, enough is known
about DCL technologies to offer policy-makers guidance
in understanding the fundamental distinguishing aspects
of these two classes of coal conversion technologies.
DCL technology involves making a partially refined
synthetic crude oil from coal, which is then further refined
Energy for Sustainable Development l Volume VII No. 4 l December 2003
Articles
103
kozinsky
Reproduced with permission from Energy for Sustainable Development
into synthetic gasoline and diesel as well as LPG -- hy-
drocarbon fuel products similar to hydrocarbon fuels de-
rived from petroleum crude oil.
ICL technology involves first gasifying coal to make
synthesis gas (‘‘syngas’’, mainly carbon monoxide (CO)
and hydrogen (H2)) and then making synthetic fuels from
this syngas; the label ‘‘indirect’’ refers to the intermediate
step of first making syngas. ICL technology can also pro-
vide hydrocarbon fuels that resemble crude oil-derived
products. One possibility is synthetic middle distillates de-
rived via the F-T process that can either be used directly
as diesel or in blends with petroleum-derived diesel. An-
other possibility is gasoline via the route of first making
MeOH from syngas and then converting MeOH into gaso-
line via the Mobil process. But MeOH can also be used
directly as a fuel, and other oxygenates (fuels containing
some oxygen) such as DME can also be provided via ICL
process technology and used directly as fuels.
Making conventional hydrocarbon fuels from coal via
either DCL or ICL processes has the advantage that the
fuel infrastructures already in place for petroleum crude
oil products can be used unchanged when a shift is made
to coal-derived fuels. However, prospective air-pollutant
regulatory constraints worldwide give high value to clean
synthetic fuels with emission characteristics superior to
those for petroleum crude oil-derived fuels. Moreover, the
oft-cited advantage of fuel infrastructure compatibility of-
fered by synthetic hydrocarbon fuels is not so great in
China at present, where a liquid hydrocarbon fuel infra-
structure for transportation fuels is at an embryonic state
of development. For these reasons and because some oxy-
genates offer performance and emission characteristics su-
perior to those for hydrocarbon fuels, the focus of ICL
analysis in this paper is on the oxygenates MeOH and
DME.
1.1. Direct coal liquefaction
With DCL technology the H/C ratio is increased by adding
gaseous H2 to a slurry of pulverized coal and recycled
coal-derived liquids in the presence of suitable catalysts
to produce synthetic crude oil. A slate of partially refined
gasoline-like and diesel-like products, as well as propane
and butane, are recovered from the synthetic crude oil
mainly by distillation. Each of the products is made up
of not one but many different large molecules that are
recovered via distillation in different temperature ‘‘cuts’’.
Hydrogen is needed in the DCL process both to make
synthetic crude oil (which might be represented in a sim-
plified manner as CH1.6) and to reduce the oxygen, sulfur,
and nitrogen in the coal feedstock. These elements are
removed from the liquid fuel products in the forms of
H2O, H2S, and NH3. The oxygen is removed so that hy-
drocarbon fuels can be obtained. The nitrogen and sulfur
compounds are removed because they would otherwise
poison the cracking catalysts in the refining operations
downstream of the DCL plant.
The amount of H2 needed is crudely estimated as fol-
lows for Yanzhou bituminous coal[1], which can be repre-
sented as CH0.81O0.08S0.02N0.01:
CH0.81 + 0.395 H2 ® CH1.6 (1a)
0.04 O2 + 0.08 H2 ® 0.08 H2O (1b)
0.02 S + 0.02 H2 ® 0.02 H2S (1c)
0.005 N2 + 0.015 H2 ® 0.01 NH3 (1d)
Thus 0.5 kmol (1.0 kg) of H2 plus 1 kmol (14.9 kg) of
coal are required to produce 1 kmol (13.6 kg) of synthetic
crude oil. The H2 might be made from natural gas via
steam reforming or from coal via gasification; the latter
is a suitable option for China, where natural gas is scarce.
The DCL products are only partially refined. They must
be further refined into finished liquid fuel products at con-
ventional refineries, where additional H2 is added (to
bring the H/C up to ~ 2 for the final products), and energy
is consumed to provide the refinery’s heat and power
needs.
DCL technology was invented by Friedrich Bergius in
1913 and commercialized in Germany and England in
time to provide liquid fuels for World War II. The activity
was abandoned when low-cost Middle East oil became
available in the early 1950s. R&D was revived in the
United States, Germany, and Japan after the Arab oil em-
bargo of 1973. Interest in DCL declined again in the mid-
1980s with the decline of the world oil price. None of the
industrialized countries are now pursuing DCL technology
to meet their own liquid fuel needs. Most global interest
in alternatives to crude oil is focused on gas-to-liquids
(GTL) technology, which aims to exploit low-cost
‘‘stranded’’ natural gas resources in various parts of the
world. However, the Clean Coal Technology Program of
the US Department of Energy is pursuing several projects
that involve liquid fuel production via indirect coal lique-
faction (ICL).
Modern DCL technology is not proven at commercial
scale. The largest scale at which there has been experience
with DCL in the United States is a Process Development
Unit at the Hydrocarbon Technology, Inc. (HTI) R&D fa-
cility that consumes 3 tonnes (t) of coal per day.
In 2002, China announced a $ 2 billion investment for
a DCL plant in Inner Mongolia based on HTI technology.
The plant was expected ultimately to produce 50,000 bar-
rels per day (b/d) (1 barrel = 0.1364 t) of partially refined
gasoline and diesel and was to be made up of three reactor
trains, each processing 4,300 t of coal daily. The first re-
actor train, which represents a scale-up by a factor of 1400
from the previous largest plant, was to start up in 2005.
However, as this article was going to press in early De-
cember 2003, it was announced that construction of the
plant is being suspended. The future of the project is now
uncertain, but at the very least the project will be scaled
down and its timing stretched out.
1.2. Indirect coal liquefaction
The first step in indirect liquefaction is to gasify coal in
oxygen (partial oxidation) to produce syngas. The CO and
H2 molecules in the syngas are then combined catalyti-
cally to produce compounds that can be used as fuels --
either hydrocarbon fuels such as synthetic gasoline or syn-
thetic diesel, or oxygenated fuels. The challenge of in-
creasing the H/C ratio is addressed by using the
water-gas-shift (WGS) reaction (CO + H2O ® H2 + CO2)
and removing the CO2 thereby produced from the system.
Energy for Sustainable Development l Volume VII No. 4 l December 2003
Articles
104
At present the most important options are hydrocarbon
fuels synthesized via the F-T process, MeOH, and DME.
1.2.1. Fischer-Tropsch liquids
The F-T process for making synthetic hydrocarbons can
be summarized, in a simplified manner, by the following
two catalytic reactions that build up large hydrocarbon
molecules from the small CO and H2 molecules produced
by gasification, with the oxygen in the CO feed being
rejected in steam:
n CO + 2n H2 ® n H2O + CnH2n (olefins) (2a)
n CO + (2n + 1) H2 ® n H2O + CnH2n+2 (paraffins) (2b)
The slate of products generated depends on the catalysts
used and reactor operating conditions. Olefin-rich prod-
ucts with n in the range 5 to 10 (naphtha) can be used
for making synthetic gasoline and chemicals in high-tem-
perature F-T processes. Paraffin-rich products with n in
the range 12 to 19 (distillates) are well suited for making
synthetic diesel and/or waxes in low-temperature F-T
processes. Development has emphasized making synthetic
diesel because the raw distillate product is an excellent
diesel fuel, whereas the raw naphtha product requires sub-
stantial subsequent refining to make an acceptable gaso-
line.
F-T technology is well established commercially and is
the focus of global GTL efforts to exploit low-cost
‘‘stranded’’ natural gas to make synthetic liquid transpor-
tation fuels. Sasol in South Africa has extensive construc-
tion and operating experience with F-T technology based
on coal gasificaton and converts annually about 42 million
t (Mt) of coal into 6 billion liters (Gl) of synthetic fuels
and 2 Gl of chemicals [Geertsema, 1996]. And there is
growing interest in coal-based F-T technology in the
United States. Sasol F-T synthesis technology along with
a Shell gasifier will be used in a $ 0.6 billion US Depart-
ment of Energy-sponsored demonstration project in Gil-
berton, Pennsylvania, that will make from coal waste
materials 5,000 b/d of F-T liquids plus 41 MWe of elec-
tricity. Recently, a detailed assessment was carried out for
the US Department of Energy of the co-production of F-T
liquids and electricity from coal via gasification at large
scales [Bechtel et al., 2003a; 2003b]. This study, based
on the E-Gas gasifier (now owned by PhillipsConoco) and
slurry-phase reactors for F-T liquids synthesis, estimated
that for an optimized plant[2] built in the US Midwest, the
internal rate of return would be 10 % if the electricity
were sold for $ 0.04/kWh and the F-T liquids for $ 30
per barrel; the equivalent crude oil price would be up to
$ 10 per barrel less than this $ 30 per barrel cost (de-
pending on refinery configuration and relative oil product
demands), because the F-T liquids would already be partly
refined [Marano et al., 1994].
Sulfur and aromatic-free F-T middle distillates are al-
ready being used as blend stock with conventional crude
oil-derived diesel in California to provide fuel that meets
that state’s stringent specifications for diesel.
1.2.2. Methanol
MeOH is a well-established chemical commodity used
throughout the world. It can potentially also be used in-
directly or directly (see Box 1) as a fuel.
The primary reactions involved in making MeOH from
syngas are:
CO + H2O ® CO2 + H2 (water gas shift) (3a)
CO + 2 H2 ® CH3OH (methanol synthesis) (3b)
The MeOH produced can be further processed to make
gasoline by the Mobil process (a commercial technology
that can provide gasoline at attractive costs from low-cost
stranded natural gas [Tabak, 2003]) or DME by MeOH
dehydration (see below), or the MeOH can be used
Box 1. MeOH as a synthetic fuel for transportation
Because of its high octane rating[28], MeOH is well-
suited for use in SIE vehicles (see discussion in main
text)[29]. It can be used in such vehicles with rela-
tively modest modifications of the basic vehicle.
Used in SIE vehicles, MeOH offers air-quality bene-
fits that are thought to be comparable to those of-
fered by reformulated gasoline [Calvert et al., 1993].
The ozone formation potential from formaldehyde
emissions of MeOH is thought to be less than the
ozone formation potential of unburned hydrocarbon
emissions; NOx emissions from MeOH engines op-
erated at the same compression ratio as for gasoline
would be less than for gasoline, because of the lower
flame temperature, but when the compression ratio
is increased to take advantage of MeOH’s higher oc-
tane rating, thereby improving engine efficiency, this
advantage may be lost [Wyman et al., 1993]. And
just as some of the unburned hydrocarbon emissions
for gasoline are carcinogenic, the US Environmental
Protection Agency has classified formaldehyde as a
probable human carcinogen, on the basis of evidence
in humans and in rats, mice, hamsters, and monkeys
[EPA, 1987].
The major drawbacks of MeOH as a transport fuel
are its low volumetric energy density (half that of
gasoline -- see Table 4), its affinity for water, its cor-
rosiveness, and its toxicity -- a fatal dose is 2-7 %
MeOH in 1 litre (l) of water, which would defy de-
tection by taste.
Drawing upon HEI [1987] and Malcom Pirnie
[1999], the following provides a perspective on the
MeOH toxicity issue: MeOH is classified as a poison
(it is rated as slightly more toxic than gasoline), and
it is infinitely miscible with water (forms mixtures
in all concentrations), allowing ready transport in the
environment. Chronic low-dose MeOH vapor expo-
sure from normal vehicle operations is not likely to
cause health problems. However, exposure through
MeOH-contaminated drinking water is a concern. In
the event of a spill, MeOH would probably be less
likely to reach drinking water supplies than gasoline,
because natural processes would degrade it more
quickly, but if MeOH-contaminated drinking water
had to be treated (which it might if an underground
MeOH tank leaked into groundwater), remediation
would be more difficult than with gasoline.
Energy for Sustainable Development l Volume VII No. 4 l December 2003
Articles
105
IBMUSER
附注
烯烃
IBMUSER
附注
脂肪烃
IBMUSER
附注
致癌的
IBMUSER
附注
甲醛
directly as fuel. This last option is the focus of the present
study (see also companion paper in this issue by Larson
and Ren [2003]).
In most parts of the world MeOH is made by steam
reforming of natural gas, but in gas-poor regions such as
China it is made mainly from coal-derived syngas via
gasification.
Under the US Department of Energy’s Clean Coal Tech-
nology Program, Air Products and Chemicals, Inc., has
brought to commercial readiness slurry-phase reactor
technology for MeOH production [Heydorn et al., 2003].
Following successful proof-of-concept in 7,400 hours of
test operation at a scale of 12,000 l/day at the DOE-owned
process development unit at LaPorte, Texas, the technol-
ogy has been demonstrated successfully at near-commer-
cial scale (300,000 l/day rated capacity) at the Eastman
Chemical Company’s coal gasification facility in King-
sport, Tennessee; during the 69-month demonstration pro-
gram since start-up in April 1997 the plant availability
averaged 97.5 %.
1.2.3. Dimethyl ether
DME is a non-carcinogenic and virtually non-toxic chemi-
cal produced at a rate of 143,000 t/year for chemical proc-
ess uses and one significant final consumer market: as an
aerosol propellant that replaced fluorinated hydrocarbons
phased out because of concerns about ozone-layer dam-
age[3]. It is also usable as a fuel (see Box 2).
Currently DME is made by MeOH dehydration:
2 CH3OH ® CH3OCH3 + H2O. But DME can also be
made (prospectively at lower cost) in a single step by
combining mainly three reactions in a single reactor [Lar-
son and Ren, 2003]:
CO + H2O ® CO2 + H2 (water gas shift) (4a)
CO + 2 H2 ® CH3OH (methanol synthesis) (4b)
2 CH3OH ® CH3OCH3 + H2O (methanol dehydration).(4c)
Haldor Topsoe in Denmark [Bøgild-Hansen et al., 1995;
1997] is developing a single-step process for making
DME from natural gas. NKK Corporation in Japan [Ohno,
1999; Adachi et al., 2000] and Air Products and Chemi-
cals, Inc., in the United States [Peng et al., 1997; APCI,
2002; Heydorn et al., 2003] are developing single-step
processes for large-scale DME manufacture from coal-de-
rived syngas using slurry-phase reactors.
In China, the Institute of Coal Chemistry (ICC) of the
Chinese Academy of Sciences together with the Shanxi
New Style Fuel and Stove Company constructed a 500
t/year DME plant in Xi’an based on MeOH dehydration
for use as a domestic cooking fuel as an alternative to
LPG (see Box 2); also, since 1995, the ICC has been car-
rying out R&D on one-step DME synthesis based on
slurry-phase reactor technology [Niu, 2000].
The Ningxia Petrochemical Industry Lingzhou Group,
Ltd., is pursuing plans to build a 830,000 t/year DME
production plant in Lingwu City, Ningxia Province, based
on use of a Chevron-Texaco coal gasifier and the slurry-
phase reactor technology of Air Products and Chemicals,
Inc. [Lucas and Associates, 2002]. The proposed plant
would be built in two phases: during the