Two Approaches to Converting Coal to a Liquid Fuel Indirect

Gasify coal and rebuild small molecules to desired product

Direct
Break coal down to

maximize correct size of molecules for liquid products

Direct Liquefaction
Reacts coal with H2 sually in presence of a liquid sol!ent "##ressi!e reaction conditions $emperatures % &''() *ressures % +'' atm

 "ppropriate catalyst *roduces a syncrude )an be refined to produce #asoline or diesel fuel ,ore con!ersion than indirect process

Indirect Liquefaction
-n!ol!es #asification of coal to produce a syn#as ,ixture of ). and H2

 /yn#as con!erted into a liquid fuel !ia processes such as 0ischer1$ropsch 20$3 process ,obil ,ethanol1to1Gasoline 2,$G3 process
$wo !ery different approaches to pro!idin# fluid fuels from coal are described and compared in this paper4 direct coal liquefaction 25)63 and indirect coal liquefaction 2-)637 0or both approaches a ma8or challen#e is to increase the hydro#en1carbon ratio7 0or finished hydrocarbon fuels such as #asoline and diesel9 H:) 2 2molar basis37 0or petroleum crude oil9 the ratio ran#es from +7; to +7<7 0or typical bituminous coals9 H:) '7=7 1.1. Direct coal liquefaction >ith 5)6 technolo#y the H:) ratio is increased by addin# #aseous H2 to a slurry of pul!erized coal and recycled coal1deri!ed liquids in the presence of suitable catalysts to produce synthetic crude oil7 " slate of partially refined #asoline1like and diesel1like products9 as well as propane and butane9 are reco!ered from the synthetic crude oil mainly by distillation7 ?ach of the products is made up of not one but many different lar#e molecules that are reco!ered !ia distillation in different temperature @@cutsAA7 Hydro#en is needed in the 5)6 process both to make synthetic crude oil 2which mi#ht be represented in a simplified manner as )H+7B3 and to reduce the oxy#en9 sulfur9 and nitro#en in the coal feedstock7 $hese elements are remo!ed from the liquid fuel products in the forms of H2.9 H2/9 and CH;7 $he oxy#en is remo!ed so that hydrocarbon fuels can be obtained7 $he nitro#en and sulfur compounds are remo!ed because they would otherwise poison the crackin# catalysts in the refinin# operations downstream of the 5)6 plant7 $he amount of H2 needed is crudely estimated as follows for Danzhou bituminous coalE+F9 which can be represented as )H'7=+.'7'=/'7'2C'7'+4
)H'7=+ G '7;<H H2 )H+7B 2+a3 '7'& .2 G '7'= H2 '7'= H2. 2+b3 '7'2 / G '7'2 H2 '7'2 H2/ 2+c3 '7''H C2 G '7'+H H2 '7'+ CH; 2+d3

$hus '7H kmol 2+7' k#3 of H2 plus + kmol 2+&7< k#3 of coal are required to produce + kmol 2+;7B k#3 of synthetic crude oil7 $he H2 mi#ht be made from natural #as !ia

steam reformin# or from coal !ia #asificationI the latter is a suitable option for )hina9 where natural #as is scarce7 $he 5)6 products are only partially refined7 $hey must be further refined into finished liquid fuel products at con!entional refineries9 where additional H2 is added 2to brin# the H:) up to 2 for the final products39 and ener#y is consumed to pro!ide the refineryAs heat and power needs7
Coal is a solid with a high carbon content but with a hydrogen content of ~5%. Coal may be used to produce liquid fuels suitable for transportation applications by removal of carbon or addition of hydrogen, either directly or indirectly. The first approach is usually known as carbonisation or pyrolysis and the second as liquefaction. in outh !frica. "ere, with large reserves of coal but no oil or gas, trade embargoes over three decades to the mid#$%&'s drove large#scale application( up to )'% of transportation fuel requirements have been met from coal.

Indirect coal liquefaction first gasifies coal and then converts the coalderived syngas into fuels and petrochemicals using Fischer-Tropsch technology. There are several technology and process alternatives for this type of approach to CTL. Direct liquefaction, by contrast, brea s do!n the comple" coal structure into smaller component molecules !hich then can be further refined into clean liquid fuel products by reducing the contents of sulfur and nitrogen.
#osted by$ %ill &illard ' (ov )*, )++, -$+.$+/ 0&

I !ould li e to see the environmental outputs from the resultant energy products. Coal is largely carbon. 0dd hydrogen, ma e hydrocarbons, burn them, and you have C1), masses of it. 0nd !hat about the energy required in the manufacturing process - does it come from burning coal2 I suspect this is no !ay an environmentally friendly !ay to ma e liquid hydrocarbon fuel3
#osted by$ 4ugh Collins ' (ov )5, )++, 6$,)$+- 0&

4ugh, 7ou need to read the fine print I guess. Diesel fuel from coal

gasification 8or 9indirect liquefaction9: is almost ;ero sulfur. It is also lo! in aromatics. 4ence there are some air pollution benefits. 1f course, you could also capture the C1) produced in the liquefaction plant and sequester it 8ho!ever you may do that at an inland coal mine:. (o!, I !on<t be betting money on the Chinese going out of their !ay to capture C1)... The Chinese are moving for!ard !ith this pro=ect mainly to reduce the effect of volatile oil prices on their economy, I suspect. In that they seem to be some!hat ahead of >? politicians.
*any different +direct, processes have been developed, but most are closely related in terms of the underlying reaction chemistry. Common features are the dissolution of a high proportion of coal in a solvent at elevated temperature and pressure, followed by hydrocracking of the dissolved coal with hydrogen gas -"./ and catalyst. 0irect liquefaction is the most efficient route currently available. 1iquid yields in e2cess of 3'% by weight of the dry, mineral matter#free coal feed have been demonstrated in favourable circumstances. 4verall thermal efficiencies -% calorific value of the input fuel converted to finished products/ for modern processes are generally in the range )'#3'% if allowance is made for generating losses and other non#coal energy imports. These processes generally have been developed to process development unit -506/ or pilot plant scale and the main technical issues have been resolved. "owever, no demonstration or commercial#scale plant has yet been built. The commercial viability of coal liquefaction rests with the overall economics of the process. This depends on the availability of significant quantities of poor quality, low cost coal, and the unavailability or otherwise relatively high cost of oil -and gas

BENEFITSOFTHE TECHNOLOGY
7 Coal liquefaction offers the following benefits8 7 1argely proven technology for the manufacture of useful liquid products. 7 !bility to manufacture transportation fuels from abundant coal. 7 9nsurance against depleting oil stocks and oil supply problems.

INTRODUCTION
The hydrogen content of Transportation fuels varies from ~$..5% in some gasolines to $:.5% in aviation turbine fuels. ;or coal to replace them, it must be converted to liquids with similar hydrogen content. This can be achieved either by removing carbon or by adding hydrogen, either directly or indirectly. The first approach is known as carbonisation or pyrolysis and the second as liquefaction. 1iquid fuels have long been produced from coal. !s the cost of converting coal into useful liquid fuels is higher than the cost of refining crude oil, it is the relative price of the raw feedstocks that has provided the main incentive to pursue the technology. The ma<or e2ceptions to this principle resulted directly from the isolation of a country from reliable, secure sources of crude oil. =ermany produced substantial amounts of coal#derived fuels during the econd >orld >ar, as did embargoed outh !frica between the mid#$%5's and $%&'s.

*ost development derives from the early $%''s, when two distinct approaches were pursued. The earliest process route involved hightemperature and high#pressure dissolution of coal in a solvent to produce high boiling point -bp/ liquids. ?o hydrogen or catalyst was used at this time. This approach, known as direct liquefaction, was patented by @ergius in $%$A and commercialised in the early $%.'s( it also became known as either the 5ott#@roche or 9 = ;arben process. ;rom the mid#$%)'s, at a time of mounting environmental concern over emissions from power generation, there was continued interest in developments of the original uncatalysed 5ott#@roche process. *ost work was carried out in the 6 !, where the more developed e2amples were the BC#9 - olvent Befined Coal/ and BC#99 processes, although other processes were also developed to a smaller scale in Capan and the 6D. The BC#9 process was very similar in concept to the 5ott#@roche process and aimed to upgrade coal to produce a clean boiler fuel with a much lower ash and sulphur content than the original coal. The BC#99 process, however, was to produce distillate products( the distinguishing characteristic of the process was the recycle of vacuum bottoms. Two other direct coal liquefaction processes were under development in the 6 ! at the same time8 the E22on 0onor olvent -E0 / process and the "#Coal process. The distinguishing feature of the E0 process was a separate solvent hydrogenation step to carefully control the hydrogen donor characteristics of the solvent, and the most important feature of the "#Coal process was its use of an ebullated bed reactor. 9n the early $%3's, political changes in the ma<or oil#producing regions resulted in dramatic oil price rises and interest switched once more to the production of transportation fuels from coal. 4il prices continued to increase throughout the $%3's and remained high until late $%&5. ignificant BF0 programmes were started, mostly based on variants of direct liquefaction. The main participants were the 6 !, Capan and the 6D, which had been developing coal#upgrading processes, and =ermany. Coal liquefaction, via indirect processes, attracted less interest, although there were parallel developments in synthesis reaction catalysis and technology for other purposes. !lmost all of the current liquefaction processes originate from this period.

CARBONISATIONAND PYROLYSIS
Carbonisation & Pyrolysis
High temperature carbonisation is the oldest process for producing liquids from coal. Coal is heated to around 950C in a closed container decomposition takes place, and the volatile matter is driven off. his is t!pical of the cokemaking process, and the h!drocarbon liquid "coal tar# is predominantl! a b!$product. he process produces ver! lo% !ields and upgrading costs are relativel! high so the coal tar is not traditionall! used in the transportation fuel sector. Coal tar is used %orld%ide for the manufacture of roofing, %aterproofing, and insulating compounds and as ra% materials for man! d!es, drugs, and paints. &ild p!rol!sis is a lo%er temperature carbonisation, or decomposition, process. Coal is heated to bet%een '50C and (50C, driving off volatile matter and forming other compounds through thermal decomposition. )iquid !ields are higher than for hightemperature carbonisation, but reach a ma*imum of +0,. he main product is a char.

his process has been used to upgrade lo%$rank sub$bituminous coals in the -./ it increases calorific value and reduces sulphur content.
"igh#temperature carbonisation is the oldest route for the production of liquids from coal, wherein hydrocarbon liquid is predominantly a byproduct of coke#making. The low yields -G~5%/ of liquid product and relatively high upgrading costs mean that traditional high#temperature carbonisation is not an option for the production of liquid fuels on a commercial basis. *ild pyrolysis is also a carbonisation technology but with less severe operating conditions. *ild pyrolysis consists of heating the coal to a temperature in the range ~:5'#)5'HC -compared with ~%5'HC in hightemperature carbonisation/, driving off volatile matter from the original coal and generating other volatile organic compounds formed by thermal decomposition during the treatment. 1iquid yields are higher than for high#temperature carbonisation, but are still no more than $5#.'% at most. The main product is a char with a reduced hydrogen and heteroatom content. The 6 ! has led the development of this process, primarily as a means to upgrade low#rank sub#bituminous coals and lignites, to increase calorific value -CI/ by re<ection of the coal,s o2ygen as carbon dio2ide -C4./ and reduce the sulphur content. !t least one process has been developed to a semi#commercial scale. =iven the large potential market for these technologies as fuel#upgrading processes, there is significant potential for these processes to contribute to the global supply of liquid fuels, despite their low liquid yields. ! higher yield of liquids can be obtained by rapid pyrolysis. These processes operate at temperatures up to $.''HC, but the residence time of the coal is significantly reduced, to a few seconds at most. Bapid pyrolysis is aimed at producing chemical feedstocks rather than liquid fuels and process economics are likely to be highly unfavourable for production of liquid fuels. There also appear to be unresolved engineering difficulties. The disadvantage of all pyrolysis and carbonisation processes is that, if the hydrogen content of the feedstock coal is to be increased to the point at which it becomes a distillable liquid, the hydrogen content of the remaining solid must be reduced. 9n any case, the liquids produced are still of low quality and require, at least, additional treatment to remove solid contaminants and water. The resulting liquid products can then be blended to produce heating fuels and fuels for stationary turbine use. ;or the raw products to be used unblended, or to be used in transport fuels, they require still further treatment. @lending and processing with conventional refinery feedstocks is likely to be needed for this to be economically viable. This has not been successfully demonstrated to date. *ost of the recent interest in these processes has been in the 6 ! as a method for upgrading coals. !t least three technologies have been developed to pilot scale. 5rocesses differ mainly in the design of the pyrolysing reactor. 4ne process, the 1iquids from Coal -1;C/ process, has been in commercial#scale operation since $%%.. The 1;C process -;igure ./ is a mild pyrolysis method for upgrading coal and was developed by =9 9nternational. There are two saleable products8 a low#sulphur, high heating#value solid known as +process#derived fuel, -50;/ and a hydrocarbon liquid known as +coal#derived liquid, -C01/. 50; yields are considerably higher than C01 yields.

The Cd i "al fed into a

perfotatcd mary gtate dryer where it is heated by a hot gas stream. .%~ tuidena time of the coal and temperatureo f the inlet gas haveb eens electedto tufttee tho moisturec ontento f the coal without initiating chemicacl hanges.T he solid bulk temperatureis controlleds o that no

significanta mounto f methanec, arbonm ono2ideo r >bon diide is relea2edfr om the coal. The solidsf rom the dryer are then transfertedt o the pyroly2erw bem the temperatureo f the dried coat is raisedt o about l44JTb y a hot reoycledg ass tream. The rate of hating of the solids i.e., the inlet tempcmtum and flow rate of the hot recycled &as atream, is catd6ly contmlledb ecausiet determinetsh e pmpettieso f the solii andi iquid pmduc~. 9n the pyroly2er. a chemicalr eactionc ccursw hich tests9tsin the mleaseo f volatile &asemmm aterialsf rom the coal. olidse 2itingt he pyrolyse areq uickly quenchedto ,stopt he pyrolysisr eaction. Theya te thenc ooleda ndt ransfFredt o the 50; storages ilo. incet he solidsh aven o surfacet t2risture and, theseforea, re likely to be dusty. a dust suppressnneta liedh :B is addeda s they leavet he 50; product silo as foal product. The gas produced in the pytoly2er is aent through a cyclone for removal of pyticulates. 9t is thens entt o a quencht ower to stop any secondaryte aetiw end to condenseth e desired1 iquids. 4nly C01 is condensedin this step( the condensationo f water is avoided. The gas stream leaving the quench tower may contain some C01 in the form of a fine mist. 9n order to recever the llquld mist,t hreee leotrostatipcr ecipitators-B 5,s/o peratingin pamBewt ereb tsteCied.T he finished C01 product is pumped from the battom of the quench tower to storage. Tlte residual- earf mm the electrostaticp recipitatorsis dividedi nto threes treams.* ost of it is recycledd irectly to the pymlyKer, omei s buntedi n the pyrolymr combuatnar nd then mi2ed with the recycledg ast o provideh eat for the mild gpsffiuttionr ation. The remaindeor f the gasi s burnedi n the dryer combustorw, hich convertss ulfur compoundsto sulfur o2ides- 4,/ and hydrocarbontso carbono 2ides- Cod. ?itrogeno 2ide -?4C and carbonm ono2ide- C4/ emissionsa n controlledb y appropriated esigno f the combustor. The hot flue gas from the

dtyer combustoirs blendedw ith the recycledg asf rom the dryer to provide heatf or drying. Tho off#gas from the dryer tlrst passesth mu&ha cyelonet o removet he entrainedp artieulate2. . 9rte misa tihnedne dri siv itdreeadt iendto w i9twho s sotdreiuammc sa.r *boonsat toefs oitl uitsio rna icny cal otdwd oir#esctatl&y etso c trhueb bderyrs eyrs wtehmile. t@hye sprayingt he off#gas with sodiumc arbonates olution, the first stages crubberc apturesth e fine patticulatesth at escapeth e dryer cyclone,a nd the seconds tages orubber emovesm osto f the L sulfur o2ides from the gas streamb y converting it into sadiurns ultlte. l,lte
sultiteisthen

o2idi2edi nto sodiums ulfate. The treatedg asi s ventedt o the atmospherteh mugha stackw hile the spents olutioni s senti nto a nondiirchatgingp ond for evaporation. %.$ !9B 54169T!?T

0uring normal opemtkm,m y purgeg as that is dossed to the atmospherem ust passt brougk the destdfubtion unit, and any purge gas that is prmh2ed in the pyrolyKer loop mm be sent to the Fsulfuhation unit only after it has been incinuatal in the dryer combustor. Therefore, air polhnantse mittedh orn the 50; buildings tackc onsisto nly of minor amountso f ?4., 4,, C4, hydrocarbonasn dp articulates.E missionsfo r eachp ollutanta reb elow the $@4to n per year9 evel for named- fuel conversion/s ources. Bigorousp armittinga nd monltoting rsquimmentsw ere

thus avoided. This was achieved through careful sekction of appropriate JbFon eoatro9 technologieas nd processc ontrol equipment. ?4,, C4, and hydrocarbonse missionsa re minimiKedb y timrmal decompositionin the dryer combustor. 4i and particulatese missionsfr om the processa re controlledb y the purge gas treatment described in ection ).$A. 9n addition, particulates emissions from the crushingMscreeninbgu ilding, raw coai storages ilo, and 50; storages ilo am controlledb y wet scrubbers.T he effrciencicso f theses crubbersa re over %%%. ).$A 56B=E =! TBE!T*Em

+9n order to meete nvimnmcntasl tandardst,h ep urgeg asb eingd ischargedto the atmospherbey the 1;C processm ustb e treated. The off#gasf rom the processc, ontainingm ostlyw aterv a~or,

nitrogen, carbon dio2ide and small amounts of sulfur o2ides, -Table ).$/ is vcnlcd to a derulfuriKation unit -;igure ).A./ which consists of a wet gas scrubber and a horiKontal to sodiutir sulfF. 9n the wet gas scrubber, dual atomimtion noKKles disperse the sodium carbonates olutioni nto tine particlesw ith compresseadi r and sprayi t into the purgeg as stream to reducet he entrainedp articulatesa nds ulfur o2ides, The horiKontals crubberf urther reduces sulfur o2ides in the purge gas as it flows hori2ontaliy through spray curtains of sodium carbonate solution. The gas then passesth rough mist eliminators< ust prior to leaving the horiaoma< scrubberto capturea ny entrainedt reatings olutioni n the form of fine mist. -;igure ).AA The spent treating solution is discharged to an evaporation pond - 5igum) .A:/ where the sodium sulfite will o2idiKe into sodium sulfate. !t the end of the 5ro<ect life, the sodium sulfate, dried by evaporationi,s enolpsulatedin a clay linti pod, coveredw ith top soil and reclaimed.
! demonstration plant owned by Encoal Corporation is located near =illette, >yoming, built with support from the 6nited tates 0epartment of Energy,s -6 04E,s/ Clean Coal Technology 0emonstration 5rogram. This was commissioned in $%%. and has a ma2imum capacity of $''' tonnesMday of sub#bituminous 5owder Biver @asin coal with a high moisture and low sulphur content. The coal is crushed and screened and then heated by a hot gas stream on a rotary grate dryer. Control of the solid bulk temperature is critical at this point, as the intention is to reduce the moisture content without initiating structural changes in the coal, thereby minimising premature gas production. The dried coal is then fed into the main rotary grate pyrolyser where it is heated to ~5:'HC by a hot recycled gas stream. !gain, accurate control of the solid heating rate and the residence time is required, as these parameters affect the product properties. 4n discharge from the pyrolyser the solids are cooled and pass to a deactivation step, consisting of treatment in a vibrating fluidised bed with a gas stream of controlled o2ygen content. The 50; is thereby stabilised by slight o2idation, reducing the risk of subsequent pyrophoric behaviour. The gas stream leaving the pyrolyser is cooled in a quench tower, condensing C01 but leaving water in the gas phase. *ost of the residual gas is recycled to the pyrolyser, some being burned in the pyrolyser combustor to provide the necessary heat. The remaining gas is burned in the dryer combustor and passes into the dryer gas recycle loop. The purge from this loop is wet#scrubbed to remove particulates and sulphur o2ides. 5urge liquor from the scrubbers is discharged to ponds for evaporation. !fter being stabilised by mild o2idation, the 50;, a low#sulphur reactive fuel suitable for pulverised coal#fired boilers, is shipped by rail to power plant. C01, roughly equivalent to a ?o.) fuel oil, is shipped by rail to a fuel oil distributor. 9n $%%) =9 signed contracts for studies on commercial 1;C demonstration plant in 9ndonesia, Capan and China. 5lans for a ) million tonsMyear plant, to be built in the 6 !, were announced in $%%3.

The 1;C technology uses a mild pyrolysis or mild gasification process which involves heating the coal under carefully controlled conditions. The process causes chemical changes in the feed coal in contrast to conventional drying, which leads only to physical changes. >et subbituminous coal contains considerable water, and conventional drying processes physically remove some of this moisture, causing the heating value to increase. The deeper the coal is physically dried, the higher the heating value and the more the pore structure permanently collapses, preventing resorption of moisture. "owever, deeply dried 5owder Biver @asin coals e2hibit significant stability problems when dried by conventional thermal processes. The 1;C process overcomes these stability problems by thermally altering the solid to create 50; and C01. The 50; is a stable low sulfur, high @T6 fuel similar in composition and handling properties to bituminous coal. C01 is a heavy, low sulfur liquid fuel similar in properties to a ?umber ) fuel oil. @riefly, in the 1;C technology, coal is first deeply dried to remove water physically. The temperature is further raised in a second stage which results in decomposition reactions that form the new products. This chemical decomposition -mild gasification/ creates gases by cracking reactions, leaving residual solids. The soli.ds are cooled and further processed to make 50;. The gases are cooled, condensing liquids as C01, and the residual gases are burned in the process for heat.

DIRECTLIQUEFACTION
Direct Liquefaction
H!drogen is added to the organic structure of coal, breaking it do%n to the point %here distillable liquids are produced. here are a number of different methods, but the basic process involves dissolving coal in a solvent at high temperature and pressure follo%ed b! the 0h!drocracking1 "i.e. adding h!drogen over a catal!st#.

)iquid !ields can be in e*cess of 20, of the dr! %eight coal feed, %ith thermal efficiencies of around (0$20,. he liquids produced from direct liquefaction are of much higher qualit! than those from p!rol!sis and can be used unblended in po%er generation or other stationar! applications. Ho%ever, further upgrading is required for use as a transport fuel. here are t%o main groups of direct liquefaction processes3 44 .ingle$stage3 provides the distilled liquids "distillates# through one primar! reactor or reactor chain. &ost of these have been superseded b! t%o$stage processes to increase production of lighter oils. 44 %o$stage3 provides distillates through t%o reactors or reactor chains. he first reaction dissolves the coal either %ithout a catal!st or %ith a lo%$activit! disposable catal!st, producing heav! coal liquids. hese are further treated in the second reactor, %ith h!drogen and a high$activit! catal!st to produce additional distillate.
0irect liquefaction processes aim to add hydrogen to the organic structure of the coal, breaking it down only as far as is necessary to produce distillable liquids. *any different processes have been developed, but most are closely related in terms of underlying reaction chemistry. Common features are the dissolution of a high proportion of coal in a solvent at elevated temperature and pressure, followed by the hydrocracking of the dissolved coal with ". and a catalyst. 0irect liquefaction is the most efficient route currently available. 1iquid yields in e2cess of 3'% by weight of the dry, mineral matter#free coal feed have been demonstrated for some processes in favourable circumstances. 4verall thermal efficiencies for modern processes are generally in the range )'#3'% if allowance is made for generating losses and other non#coal energy imports. The liquid products from direct liquefaction processes are of much higher quality than those from pyrolysis processes and can be used unblended for most stationary fuel applications. They do, however, require further upgrading before they can be used directly as transportation fuels. This upgrading utilises standard petroleum industry techniques, allowing the products from a liquefaction plant to be blended into the feedstock streams of a petroleum refinery. 0irect liquefaction processes can conveniently be divided into two main groups, depending on whether the initial dissolution of the coal is separated from the conversion of the dissolved coal into distillable products8 7 ! single#stage direct liquefaction process gives distillates via one primary reactor or a train of reactors in series. uch processes may include an integrated on#line hydrotreating reactor, which is intended to upgrade the primary distillates without directly increasing the overall conversion. 7 ! two#stage direct liquefaction process is designed to give distillate products via two reactors or reactor trains in series. The primary function of the first stage is coal dissolution and is operated either without a catalyst or with only a low#activity disposable catalyst. The heavy coal liquids produced in this way are hydrotreated in the second stage in the presence of a high#activity catalyst to produce additional distillate. ome processes were designed specifically to co#process coal with petroleum#derived oils and these may fall into either group. !lso, coal liquefaction processes from both groups have been adapted for coprocessing.

Single-stagePro

esses

9n the mid# to late $%)'s, as interest was growing, all of the available

processes were single#stage. *ost development therefore continued to adopt a single#stage approach. ome developers added a second stage during the $%3's, following the oil crisis, to increase the production of light oils. The single#stage processes developed furthest are8 7 Dohleoel -Buhrkohle, =ermany/ 7 ?E041 -?E04, Capan/ 7 "#Coal -"B9, 6 !/ 7 E22on 0onor olvent -E0 / -E22on, 6 !/ 7 BC#9 and 99 -=ulf 4il, 6 !/ 7 9mhausen high#pressure -=ermany/ 7 Conoco Kinc chloride -Conoco, 6 !/. *ost of these have since been superseded and abandoned. Two e2ceptions are the Dohleoel and ?E041 processes, both of which are considered ready for commercialisation by their developers. everal other, less important, processes were developed to a modest scale in the 6 !. 4ther countries, notably Bussia and 5oland, also carried out BF0 on single#stage liquefaction at a significant scale( the approaches adopted are believed to be similar in most respects to the Dohleoel process.

T!e "o!leoel Pro ess

The Dohleoel process -;igure A/ with 9ntegrated =ross 4il Befining -9=4BN/ is a relatively recent development, by Buhrkohle != and IE@! 4E1 !=, of the process used on a commercial scale in =ermany until $%:5. 0evelopment proceeded via a '.5 tonnesMday and '.. tonnesMday continuous unit at @ergbau#;orschung -now 0*T/ and a .'' tonnesMday plant at @ottrop. The @ottrop plant operated from $%&$ to $%&3, producing over &5,'''t of distillate products from $3','''t of coal over appro2imately ..,''' operating hours. The technology can therefore be considered to be fully demonstrated. 9n $%%3 the China Coal Besearch 9nstitute -CCB9/ signed a two#year agreement with =ermany to carry out a feasibility study for a 5''' tonnesMday demonstration plant. The suitability of sites for a liquefaction plant in Ounnan 5rovince was investigated, including the potential market for products. Coal is slurried with a process#derived recycle solvent and a +red mud, disposable iron catalyst, pressurised and preheated. ". is added and the
0ryer Coal Cyclone Cyclone Caustic crubber tack =ases 50; -Char/ C01 -1iquids/ 0ryer Combustor 5yrolyser Combustor Botary Cooler Condensation E 5 Char 0eactivation 5yrolyser

Figure 2. Encoals LFC process

mi2ture passed to an up#flow tubular reactor, operating typically at A''bar and :3'HC. The specific coal feed rate to this reactor is in the range '.5# '.)5 tonnesMmAMhour. 5roducts from the top of the reactor pass to a hot separator. The overheads from this separator remain in the gas phase and are hydrotreated at a temperature of A5'#:.'HC in a fi2ed#bed reactor at the same pressure as the main reactor. The hydrotreated products are depressurised and cooled in two stages. The liquid product from the first of these stages is recycled to the slurrying step as part of the solvent. The liquid product from the second stage is routed to an atmospheric distillation column, yielding a light oil -C5 # .''HC bp/ and a medium oil -.''#A.5PC bp/ product. The bottoms from the original hot separator pass to a vacuum distillation column to recover distillable liquids. These are added to the hydrotreating reactor feed, and are subsequently largely recycled as solvent. The vacuum column bottoms consist of pitch, mineral matter, unreacted coal and catalyst, and in commercial operation would be used as a gasifier feedstock for ". production. =reater than %'% conversion can be obtained when processing bituminous coals, with liquid yields in the range 5'#)'% on dry ash#free coal. 5rocess yields and quality, when using 5rosper, a =erman bituminous coal, are summarised in Table $.
Table 1. Kohleoel process yields and pr oduct quality with Pr osper coal

T!e NEDOL Pro ess

;rom $%3& to $%&A, three direct coal liquefaction processes were developed by Capanese companies ?ippon Dokan, umitomo *etals 9ndustries and *itsubishi "eavy 9ndustries under a Capanese =overnment initiative. The initiative was managed by the ?ew Energy and 9ndustrial Technology 0evelopment 4rganisation -?E04/. @y $%&A these processes had been tested at scales ranging from '.$ tonnesMday to ..: tonnesMday. Bather than support each individually, ?E04 amalgamated features of all three processes to produce the ?E041 process -;igure :/, targeted principally at sub#bituminous and low#rank bituminous coals. ! consortium of .' companies was then established, under the name ?ippon Coal 4il Company 1td, to design, build and operate a .5' tonnesMday pilot plant. "owever, the pro<ect was terminated in $%&3 due to budgetary constraints. ! $ tonnesMday process support unit -5 6/ was completed to schedule in $%&&, at a cost of ~6 QA' million, and has operated on an intermittent campaign basis since that time. ! redesign of the pilot plant, at a reduced scale of $5' tonnesMday, started in $%&&. Construction of this plant commenced in 4ctober $%%$ at Dashima, and was completed early in $%%). The pilot plant was commissioned, followed by five operation runs, from *arch $%%3 to eptember $%%&. 0uring these runs, three types of coal of different rank -Tanito "arum and !daro coals from 9ndonesia and 9keshima coal from Capan/ were liquefied without serious problems. *any results were obtained, including the achievement of &' days, continuous coalcharging operation, 5&% by weight -dry ash#free coal basis/ of oil yield, the successful use of slurry with a concentration of 5'% and ).'' hours of cumulative operation. CCB9 signed an agreement with ?E04 and the Centre for Coal 6tilisation, Capan, to carry out a feasibility study for a 5''' tonnesMday demonstration plant. everal tests have been completed at laboratory# and 506#scale with Oilan coal. !t these scales, oil product yields of ).% were predicted. ! pyrite from "eilong<iang province was used as the catalyst. 5rocess evaluation, environmental impact assessment and economic analysis will be carried out in the ne2t stage of the pro<ect, which is due for completion in .'''. Coal is ground with .#:% by weight of a synthetic iron#based catalyst and slurried with a recycled solvent. The slurry is mi2ed with "., preheated and fed to the primary reaction stage. The reactors are simple tubular upflow reactors operating at temperatures in the range :A'#:)5PC and pressures in the range $5'#.''bar. The nominal slurry residence time is ~$ hour, with actual liquid#phase residence times in the range %'#$5' minutes. The products from this primary reaction step are cooled, depressurised and distilled in an atmospheric column to take off a light distillate product. The atmospheric column bottoms pass to a vacuum distillation column, where a middle distillate and a heavy distillate product are removed. *ost of the middle distillate and all of the heavy distillate are recycled as solvent via a hydrogenation step. The vacuum column bottoms, which contain unreacted coal, mineral matter and catalyst, are discharged and in commercial operation would be used as a gasifier feedstock for ". production. The ma2imum solids loading which can be achieved in this stream is ~5'%, but in practice it is understood that a loading of A5% is more typical. ince the pitch discharged with the solids represents a substantial loss of potential product, the process is limited to coal of relatively low ash content. !fter mi2ing, the middle and heavy distillates from the vacuum column are fed to solvent re#hydrogenation reactors. These are down#flow packed catalyst beds operating at A.'#:''HC and $''#$5'bar. The catalysts used are variants of those used in the conventional petroleum industry for the hydrodesulphurisation of crude oil fractions. The nominal feed residence time is ~$ hour. The reactor products are depressurised at temperature into a flash distillation vessel, where a hydrotreated naphtha product is taken off. The liquid product from the flash distillation is recycled to the slurrying step as the solvent.
"ydrogen Catalyst Coal lurry *i2ing "ydrogen Becycle 9ntegrated "ydrogenation

eparation =ases ;lash eparator 5rimary Beaction Iacuum 0istillation Becycle olvent !tmospheric 0istillation 15= 1ight 4il *iddle 4il Iacuum @ottoms

Figure . The Kohleoel process

"ydrogen "ydrogen Becycle Catalyst Coal lurry *i2ing 5rimary Beaction Iacuum 0istillation Becycle olvent !tmospheric 0istillation =ases 1ight 0istillate *iddle 0istillate "ydrotreated ?aphtha Iacuum @ottoms "ydrogen olvent "ydrogenation

Figure !. The "E#$L process

5rocess yields Oield "ydrocarbon gases -C$#C:/ $%.' 1ight oil -C5#.''HC/ .5.A *edium oil -.''#A.5HC/ A..) 6nreacted coal and pitch ...$ 5roduct quality 1ight 4il *edium 4il "ydrogen -%/ $A.) $$.% ?itrogen -ppm/ A% $3: 42ygen -ppm/ $5A &: ulphur -ppm/ $. G5 0ensity -kg m#A/ 33. %$. 5roduct yields vary with the type of coal being processed, although the primary reaction operating conditions are ad<usted to minimise this. 0istillate product yields of 5'#55% -dry ash#free basis/ are obtained with all these coals. The liquid products are of relatively low quality and require more substantial upgrading than those from other processes, although this is not likely to present new technical difficulties.

T!e H-Coal Pro ess

The "#Coal process -;igure 5/ was developed by "B9 -now "ydrocarbon Technologies 9ncorporated, "T9/ from the commercialised "#4il process used to upgrade heavy oils. The process was developed to a .'' tonsMday pilot plant, which was built at Catlettsburg, Dentucky, in $%&' and operated until $%&A. 0esigns were subsequently prepared for a commercial#scale plant, to be built at @reckinridge, Dentucky. The "#Coal process remained the basis of most subsequent process development sponsored by the 6 04E and was effectively incorporated into the Catalytic Two#stage 1iquefaction -CT 1/ process. Coal is slurried with a recycle solvent that consists of a mi2ture of a solidscontaining hydrocracker product with heavy and middle distillates obtained by product fractionation. ". is added and the mi2ture is preheated and fed to an ebullated bed hydrocracker, which is the distinguishing feature of the process. This reactor operates at temperatures of :.5#:55PC and a pressure of .''bar. 9t contains a conventional supported hydrotreating catalyst, either nickel#molybdenum or cobalt#molybdenum on alumina. The catalyst is fluidised by ". and a pumped internal recycle stream, for which the intake is positioned above the upper limit of the e2panded bed of catalyst but still within the reactor liquid Kone. This recycle stream contains unreacted coal solids. ince the hydrocracking reactions required to produce distillable liquids are highly e2othermic, accurate temperature control is crucial for engineering scale#up. The ebullated#bed reactor system offers substantial advantages

over fi2ed#bed reactors in this, as the reactor contents are well mi2ed and temperature monitoring and control is more easily effected. !lso, ebullated#bed reactors allow catalyst to be replaced while the reactor remains in operation, enabling a constant catalyst activity to be maintained. This is particularly important with supported catalysts as, although these have a high initial activity, they deactivate relatively rapidly at coal liquefaction conditions. The reactor products pass to a flash separator. 1iquids in the overheads are condensed and routed to an atmospheric distillation column, producing naphtha and middle distillate. The flash bottoms are fed to a bank of hydrocyclones. The overheads stream, which contains $#.% solids, is recycled to the slurrying stage. The underflow is routed to a vacuum distillation column. olids are removed with the vacuum column bottoms, while the vacuum distillate forms part of the product for e2port. !s with other processes, yields are dependent on the coal. R%5% overall conversion can be obtained with suitable coals, with liquid yields up to 5'% -dry basis/.

T!e E##on Donor Sol$ent Pro ess

E22on Corporation started E0 process development in the $%3's and progressed to the construction of a .5' tonsMday pilot plant at @aytown, Te2as, in $%&'. !t this point E22on considered that the process -;igure )/ was ready for commercialisation, although development was discontinued. 1iquid yields were lower than in more recent processes. !s a result the process showed relatively high specific capital costs and apparently uncompetitive economics. The pilot plant was operated until $%&., with further research continued until at least $%&5. Coal is slurried with a distillable recycled solvent that has been rehydrogenated to restore its hydrogen donation capacity. This improves the effectiveness of the solvent, and this is the key distinguishing feature of the process.

The slurry is mi2ed with "., preheated and fed to a simple up#flow tubular reactor that operates at :.5#:5'HC and $35bar. ?o catalyst is added. The reactor effluent passes to gas#liquid separators, from which the liquid product is fed to a vacuum distillation column. ?aphtha and middle distillate products are recovered, although most of the middle distillate is recombined with the heavy distillate to form the basis for the recycle solvent. The vacuum column bottoms, containing the solid residues, are discharged and fed to a proprietary E22on +;le2icoker, unit. This combines pyrolysis and gasification steps to produce additional distillate product and a fuel gas, which would be used for ". production. The pyrolysis step is carried out at temperatures in the range :&5#)5'HC. ;le2icoking is now in commercial use. Behydrogenation of the recycle solvent is carried out in a fi2ed#bed catalytic reactor, using either nickel#molybdenum or cobalt#molybdenum on an alumina support. The reactor is operated at conditions in the region of A3'HCM$$'bar, although conditions are varied to control the degree of hydrogenation of the solvent and thus maintain its quality. 5rocess yields are closely related to the characteristics of the coal being

processed, but can be varied to some e2tent by altering conditions. Typically, overall liquid yields -including the liquids produced by ;le2icoking/ are ~A)% for lignites, ~A&% for sub#bituminous coals and ~A%#:)% for bituminous coals -all dry ash#free basis/. 1iquid yields can be increased further by recycling part of the vacuum bottoms stream to the slurrying step, although this was not tested on the .5' tonsMday plant. 6sing this technique, yields of up to :3% for lignites, 5'% for sub#bituminous coals and )'% for bituminous coals could be achieved. The boiling range distribution of the liquid products can also be varied within a wide range, depending on market requirements.

T!e SRC-I an% SRC-II Pro esses

The BC processes were originally developed to produce cleaner boiler fuels from coal. ! '.5 tonsMday plant was built in $%)5 and scaled up in $%3: to two separate pilot plant. These were located at >ilsonville - BC#9, ) tonsMday see ;igure 3/ and ;ort 1ewis, >ashington - BC#9, 5' tonsMday/. The ;ort 1ewis plant was later converted to a BC#99 unit, although because of the more severe conditions required for BC#99, the capacity was downgraded to about .5 tonsMday. The ob<ective of the BC#99 process is to produce distillate products. 0etailed designs for large#scale plant were subsequently prepared, although these plant were not built. The >ilsonville plant continued to be funded by the 6 04E until $%%. as a pilot#scale test facility for the whole 6 direct liquefaction development programme. The BC processes have now been abandoned in their original form, but elements have been incorporated in more recent 6 processes.
"ydrogen Becycle Coal lurry *i2ing Iacuum 0istillation !tmospheric 0istillation =ases Iacuum @ottoms ;lash eparator Ebullated @ed "ydrocracker 5reheating 1ean lurry Becycle "ydrocyclone ;resh "ydrogen ?aphtha 5roduct 0istillate 5roduct *iddle 0istillate olvent Becycle

Figure %. &che'atic diagra' o( the )*Coal process

"ydrogen Becycle Coal lurry *i2ing Iacuum 0istillation =ases 5reheating ;resh "ydrogen ?aphtha 5roduct 0istillate eparation ;uel 4il Tubular Beactor "ydrogen ;le2icoker Coke to =asification ;i2ed @ed "ydrotreater Becycle olvent

Figure +. &che'atic diagra' o( the E#& process

T!e I&!a'sen Hig!-(ress're Pro ess

The development of the 9mhausen high#pressure process commenced in $%&. and a $''kgMday 506 was commissioned in $%&:. The process operating conditions appear very severe -:3'#5'5PC and )''#$'''bar/. !s a consequence, it seems unlikely that the process could be successfully commercialised unless yields were e2ceptionally high.

T!e Cono o )in C!lori%e Pro ess

9n the late $%3's and early $%&'s Conoco worked on the development of a process which uses molten Kinc chloride to hydrocrack coal directly to give good yields of gasoline in a single step. This process is one of the very few direct liquefaction processes that is not a direct derivative of prewar technology. The process was taken to the $ tonMday pilot plant scale,

although this was operated for only a short period and with limited success. *a<or metallurgical difficulties were e2perienced as a result of the highly corrosive nature of Kinc chloride and other chloride salts that formed in the system. These difficulties remain, but if they could be resolved this is one area in which further development might have the potential to make a significant improvement in liquefaction economics.

T*o-stage(ro

esses

*ost two#stage direct liquefaction processes were developed in response to the oil price rises of the early $%3's, often as a development of earlier single#stage processes. >ork was carried out in many different countries, but relatively few processes were developed beyond the laboratory scale and many were generically very similar. 5rocesses include8 7 Catalytic Two# tage 1iquefaction -CT 1/ -6 04E and "B9, now "T9, 6 !/ 7 1iquid olvent E2traction -1 E/ -@ritish Coal Corporation, 6D/ 7 @rown Coal 1iquefaction -@C1/ -?E04, Capan/ 7 Consol ynthetic ;uel -C ;/ -Consolidation Coal Co, 6 !/ 7 1ummus 9T 1 -1ummus Crest, 6 !/ 7 Chevron Coal 1iquefaction -CC15/ -Chevron, 6 !/ 7 Derr#*c=ee 9T 1 -Derr#*c=ee, 6 !/ 7 *itsubishi olvolysis -*itsubishi "eavy 9ndustries, Capan/ 7 5yrosol - aarbergwerke, =ermany/ 7 !moco CC#T 1 -!moco, 6 !/ 7 upercritical =as E2traction - =E/ -@ritish Coal Corporation, 6D/. 4nly the CT 1, 1 E and @C1 processes continued in development beyond the late $%&'s.

T!e Catal+ti T*o-stage Li,'e-a tion Pro ess

The CT 1 process is a development of the "#Coal single#stage process. 9n the pilot plant at the >ilsonville liquefaction test facility, the process was further developed over a period of almost $5 years. This facility was closed in $%%., but development work funded by the 6 04E continues, including continuous bench#scale and 506#scale testing at "T9,s 1awrenceville facility. The process has now evolved into a generic composite of much of the liquefaction development work funded by the 6 04E in the $%&'s and $%%'s. *ost recently, a close#coupled configuration has been adopted in which both stages use an active supported catalyst. Earlier configurations included a thermal first dissolution stage, sometimes assisted by a low#activity disposable catalyst similar to that used in many single#stage processes. @oth short contact and longer residence times were tested. 9n recent work "T9 has included an in#line hydrotreater after the second#stage reactor to improve product quality. 9n $%%3, "T9 signed a two#year agreement with CCB9 to carry out a feasibility study for a direct coal liquefaction plant using enhua coal. "T9 has support for this from the 6 04E. 1aboratory#scale tests and prefeasibility studies were undertaken and potential sites and product markets investigated in the henfu 0ongsheng area. ! schematic diagram for the most recent version of the process is given in ;igure &. Coal is slurried with a process#derived recycle solvent, preheated, mi2ed with ". and fed to the bottom of an ebullating bed reactor. This reactor contains a supported catalyst, generally nickel#molybdenum on alumina, which is fluidised by an internal recycle in the reactor. The reactor, therefore, has the characteristics of a uniform#temperature continuous#stirred tank reactor. The solvent acts as a hydrogen donor and solubilises the coal by breaking down its structure to a substantial e2tent in the first reactor. This first reactor substantially rehydrogenates the solvent. Typically, operating conditions of $3'bar and a temperature of :''#:$'HC are used when processing bituminous coals. >ith sub#bituminous coals, higher temperatures may be required to effect this structural breakdown and the hydrogen donor effect becomes of less importance. 9n one process variant, a dispersed molybdenum or iron catalyst may be used in the first stage in these circumstances. The reactor products pass directly into the base of a second, ebullating#bed reactor stage, operating at the same pressure as the first stage but at a higher temperature -~:A'#::'HC/. This reactor also contains a supported catalyst, generally but not necessarily the same as that in the first reactor. !fter separation and depressurisation steps, the products from the second reactor enter an atmospheric distillation column, where distillate products

boiling up to :''HC are removed. The bottom stream from this column contains solvent, unreacted coal and mineral matter. These solids are removed by one of several possible techniques and the solvent is recycled to the slurrying step. 9n some process variants, only part of the atmospheric column bottom stream is routed to the solids removal step, resulting in the recycle solvent containing mineral matter and any dispersed catalyst that may have been used. The process is designed to operate without the need to remove a separate pitch stream as a product. *ost operating e2perience at the >ilsonville plant was gained with the Derr#*c=ee critical solvent de#ashing -C 0/ process, also known as the residual oil solvent e2traction -B4 E/ process. ince the process has now been taken back to the developmental stage, recently reported yields vary considerably depending on the coal processed, the process configuration adopted and the particular processing
Figure ,. &-C*1 + tons.day direct lique(action plant at /ilson0ille
"ydrogen Becycle Coal lurry *i2ing =ases 5reheating ;resh "ydrogen 1ight 0istillate econd# tage Ebullating Beactor Becycle olvent ;irst# tage Ebullating Beactor Besidues olids eparation *iddle 0istillate "eavy 0istillate

Figure 1. The CT&L process

conditions. 0istillate product yields of )5% or higher on dry ash#free coal can be obtained, although the product is relatively high#boiling. The residue of unconverted coal and the heavy preasphaltenic material re<ected with the C 0 ash concentrate is appro2imately .'%, in equal proportions. The final operations at >ilsonville provided a well#defined base case for process conditions, yields and product quality.

T!e Li,'i% Sol$ent E#tra tion Pro ess

The 1 E process was developed by @ritish Coal Corporation between $%3A and $%%5. ! ..5 tonnesMday pilot plant was built and operated for four years at 5oint of !yr, ?orth >ales -;igure %/, but has since been decommissioned. !n outline design for a )5 tonnesMday demonstration plant has been produced in sufficient detail to allow a contractor to proceed directly to the detailed design stage. "owever, there are no current plans to construct such a plant. ;igure $' shows a schematic diagram of the process. Coal is slurried with a process#derived recycle solvent, preheated and passed to a non#catalytic digestion step, which consists of two or more continuous#stirred tank reactors in series. These reactors operate at a temperature of :$'#::'HC and a pressure of $'#.'bar, required solely to reduce solvent vaporisation. ?o ". is used in this step, but the solvent acts as a hydrogen donor, transferring up to .% by weight of hydrogen to the coal. The digester product is partially cooled and filtered in a vertical#leaf pressure filter to remove unreacted coal and ash. The filter cake is washed with a light recycle oil fraction to recover product and dried under vacuum. The dried filter cake contains only a small proportion of residual nondistillable liquids( the process is therefore relatively insensitive to the ash content of the feed coal or the e2tent to which it can be dissolved. 9n commercial operation the filter cake would be gasified to produce ".. The filtered coal e2tract passes to a distillation column to recover the light oil wash solvent and is then preheated, mi2ed with ". and routed to one or more ebullating#bed reactors in series. There is no inter#stage separation and the reactors operate at nominally the same conditions8 ~.''bar, :''#::'HC and a space velocity in the range '.5#$.'h#$ -kg feed per kg catalyst per hour/. The reactor products are cooled, depressurised and passed to an atmospheric distillation column to recover a distillate product. The cut#point of this column is ad<usted to maintain solvent balance, resulting in a product that typically boils below A''HC. The column bottoms are routed partially to a vacuum distillation column, used to control the level of pitch in the recycle solvent. The overheads from this

column are recombined with the main atmospheric column bottoms stream and recycled as the solvent to the slurrying step. ! significant proportion of saturated species may build up in the recycle solvent under certain conditions, reducing the effectiveness of the e2traction step. Thermal cracking is used to control this aspect of solvent quality. The pilot plant operating programme concluded with a steady#state demonstration run. The results from this run indicate that it would be possible to operate the process in an all#distillate product mode, removing the requirement for a vacuum column. olvent quality was maintained without the need for a separate, thermal cracking step. 5rimary productquality data from this demonstration period are summarised in Table .. The total distillate product yield is in the range )'#)5% -dry ash#free coal/, most of which boils below A''HC. The total filter cake yield includes ~3% of undistillable pitch.
Table 2. L&E process conditions2 yields and product quality using Point o( 3yr coal

daf8 dry ash#free

T!e Bro*n Coal Li,'e-a tion Pro ess

The @C1 process -;igure $$/ was developed by ?E04 of Capan to a 5' tonnesMday pilot#plant scale, constructed at *orwell in Iictoria, !ustralia. 9t was operated over the period $%&5#$%%', processing a total of ~)','''t of coal. 4perations ceased in 4ctober $%%'. The plant was decommissioned in $%%$ and demolished in $%%..
Coal lurry *i2ing Becycle olvent 0esaturated olvent Thermal Cracking ;ilter Becycled "ydrogen olvent Becovery ;resh "ydrogen Ebullated @ed "ydrocracker !tmospheric 0istillation Iacuum 0istillation =ases 0istillate 5roduct ;ilter Cake 5itch C TB Beactors

Figure 14. The L&E process Figure 5. L&E pilot plant (acility at Point o( 3yr

4perating conditions olventMcoal ratio ... 0igestion pressure -bar/ $5 0igestion temperature -HC/ :A$ ?ominal residence time -min/ 5' "ydrocracking pressure -bar/ .'' "ydrocracking temperature -HC/ :A: pace velocity -kg feedMkg catMhour/ '.3) 5roduct yield C$#C: hydrocarbon gases -% daf coal/ $5.: C5#A''HC distillate product -% daf coal/ :%.% A''#:5'HC solvent surplus -% daf coal/ $..: 5itch -R:5'HC/ -% daf coal/ '.& ;ilter cake organics -% daf coal/ .A.% 5roduct analysis "ydrogen -wt%/ $..$: ?itrogen -wt%/ '.$: ulphur -wt%/ '.':
"ydrogen "ydrogen Catalyst Coal Iacuum 0istillation olvent Becycle olvent Becycle !tmospheric 0istillation olvent "ydrogenation >et *illing lurry 0ewatering 5rimary

Beaction >ater =ases ?aphtha "ydrotreated ?aphtha Besidue olvent 0eashing

Figure 11. The 6CL process

The process is designed specifically to handle very low#rank coals such as those found in the 1atrobe Ialley of Iictoria, which may contain R)'% moisture. !s a result, a crucial aspect is the efficient drying of the coal. The 5' tonnesMday rated throughput of the pilot plant required ~$3' tonnesMday of raw coal to be processed. ;ollowing e2tensive pilot plant operation, BF0 using a '.$ tonnesMday bench#scale continuous liquefaction test facility and related equipment was carried out until $%%3 to improve the reliability, economics and environmental compatibility of the coal liquefaction process. @ased on the BF0 results an improved @C1 process was proposed. This comprises slurry de#watering, liquefaction, in#line hydrotreating, and de#ashing, with the following features8 7 use of a high#active and ine2pensive catalyst such as limonite ore pulverised in oil 7 use of a heavy fraction solvent -bp A''#:.'HC/ 7 adoption of coal liquid bottom -C1@/ bpR:.'HC recycling. Compared with the results of the pilot plant, the increase of oil yield, improvement of product oil quality and suppression of scale formation in reactors were proved using the bench#scale unit with G$% -dry ash#free coal/ catalyst addition. 9t was estimated that the improved process could decrease the crude oil equivalent nominal price by .:% compared with the @C1 process at the !ustralian pilot plant.

Co-(ro

essing

+Co#processing, is generally a variant on other direct liquefaction processes. 9t involves simultaneous upgrading of coal and of a non coal#derived liquid hydrocarbon. The liquid hydrocarbon also serves as the slurrying and transport medium for the coal. This is usually a low#value high#boilingpoint material, such as bitumen, an ultra#heavy crude oil or a distillation residue or tar from conventional crude oil processing. There is no solvent recycle loop and the underlying process may be either single# or two#stage. 9n general, co#processing technologies are based on adaptations of pree2isting direct liquefaction processes to a once#through non#recycling basis. 9n these processes most of the liquid product derives from the oil rather than from the coal. The overall aim of co#processing is to upgrade the petroleum#derived solvent at the same time as the coal is liquefied, thereby reducing capital and operating costs per unit of product. "owever, the non coal#derived solvents are both poor physical solvents for coal and poor hydrogen donors. This results in a relatively low conversion of the coal to liquid products. The economics of co#processing, therefore, depend predominantly on the differential between the heavy liquid feedstock cost and the price of conventional crude oil. The addition of a low#price coal to the feed improves the process economics by reducing the average feedstock cost. Compared with other liquefaction routes, capital costs are generally significantly lower per unit of product, since most of the product is derived from the oil feedstock. 9n practice, the true competitors for coprocessing are likely to be heavy oil upgrading processes. !lthough some co#processing technologies have been developed to several tonnesMday 506, or pilot plant scale, they have not been developed to the same degree as other liquefaction processes. ?one has been demonstrated at significant -~$'' tonnesMday/ scale. 5rocesses include8 7 *9T9 *ark 9 -Capan/ 7 *9T9 *ark 99 -Capan/ 7 the Cherry 5 5rocess -4saka =as Co., Capan/ 7 olvolysis -*itsubishi "eavy 9ndustries, Capan/ 7 *obil -6 !/ 7 5yrosol - aarbergwerke, =ermany/ 7 Chevron -6 !/ 7 1ummus Crest -6 !/ 7 !lberta Besearch Council -!BC, Canada/ 7 C!?*ET !4 TB! -Canada/ 7 Bheinbraun -=ermany/

7 T6C -Technical 6niversity of Clausthal, =ermany/ 7 645 lurry#catalysed -645, 6 !/ 7 "T9 -6 !/. The most important of these are !BC, C!?*ET, "T9 and 1ummus Crest.

L'&&'s Crest Co-(ro essing

1ummus Crest Co#processing was developed to a '..5 tonsMday 506 in the early and mid#$%&'s. 9t is a derivative of the 1ummus 9T 1 process. ! key feature of the process is that the petroleum heavy oil is hydrogenated prior to its use as slurrying solvent for the coal. This generates some hydrogen donor capability, increasing the e2tent to which the coal is dissolved and reducing repolymerisation and coking reactions. Coal is slurried with the hydrogenated petroleum residue and reacted in an uncatalysed, short contact time reactor at a temperature of :A'#:5'HC and a hydrogen pressure of $:'bar. The reactor products pass directly to the second stage 1C#;iner ebullated#bed reactor, which operates at the same pressure and a temperature of :''#:A5PC with a supported hydrotreating catalyst. !s with many other co#processing options, the intention appears to be to integrate the process within an e2isting oil refinery. ~%'% of the coal is dissolved in the first stage on a dry ash#free basis, with overall conversions approaching %5%. The overall conversion of heavy material in the petroleum residue is 3'#&'%. The total net yield of distillable products is in the range 5'#55% on fresh feed.

Al.erta Resear ! Co'n il Co-(ro essing

!BC, in collaboration with Canadian Energy 0evelopments, originally developed the two#stage Counter#flow Beactor -C;B/ process for upgrading tar#sand bitumen. ubsequently, the process was adapted to co#process subbituminous coals with bitumen. 9ncorporation of coal is said to increase distillable oil yields in comparison with those obtainable from bitumen alone. The process is unusual as it uses a C;B and, in place of "., uses C4 and water in the first stage. The first stage has been tested at a scale of '..5 tonsMday for co#processing and 5 tonsMday for bitumen alone. Coal is first cleaned by an oil agglomeration technique and then slurried with bitumen, water and a disposable alkali metal catalyst. The mi2ture is fed to the top of the counter#flow reactor, which operates at A&'#:''HC and &3bar. C4 is fed to the base of this reactor and travels upwards, the shift reaction generating ".. The high o2ygen content of sub#bituminous coals is reduced by use of C4 and steam and the process is claimed to be more effective and lower cost than the direct use of ".. The second stage, in principle, consists of a second C;B reactor system operated at ~:.'#:&'HCM$35bar. Either ". or C4Msteam could be used in this stage. There is no recycle of product from the second stage. The conversion of the coal depends primarily on the coal characteristics, but conversions of up to %&% on dry, ash#free coal can be obtained in some cases. The overall product yield from the two stages is appro2imately 3'% on the combined weight, dry ash#free coal and bitumen#fed.

CAN/ET Co-(ro essing

The C!?*ET hydrocracking process was intended to hydrocrack heavy oils and was developed to a 5''' barrels -bbl/Mday commercial scale at the 5etro Canada *ontreal refinery by $%&5. ! variation of the process was adapted for co#processing and was taken to the '.5 tonsMday pilot plant scale in a three#year BF0 consortium programme sponsored by Bheinbraun !=, !moco Corporation and the !lberta 4il ands Technology and Besearch !uthority -!4 TB!/. C!?*ET discontinued BF0 in $%%A. Coal and a disposable coal#based catalyst are slurried with a petroleum vacuum residue or bitumen, mi2ed with ". and fed to a single#stage upflow reactor. Typical operating conditions are reactor temperatures from ::'#:)'HC, pressures from $'#$5*5a with feedstock coal concentrations of A'#:'wt% -mineral matter#free basis/. The reactor product is separated and fractionated to recover distillate products and an undistillable residue. The e2tent to which coal is converted is highly dependent on coal characteristics, but may be as high as %&% on a dry ash#free basis. The conversion of high#boiling material in the bitumen or vacuum residue may be up to 3'%, depending on reaction severity. 4verall net distillable oil yields of up to &'% on dry ash#free slurry feed are reported.

HTI Co-(ro essing

"T9 -previously "B9/ has carried out test work on co#processing since $%&5, treating it as a simplified version of the mainstream two#stage direct CT 1

coal liquefaction process. 9t differs only in that there is no recycle solvent loop. *ost work has been carried out with lignites and other low#rank coals. Becent work includes the use of "T9s =elCatS, an iron#based dispersed catalyst . Coal conversions of up to %$% -dry ash#free basis/ have been reported. The conversion of heavy material in the petroleum residue varied from &'% to %'%, with overall distillable product yields in the range 33#&)% by weight on the total feed.

INDIRECTLIQUEFACTION
The only +core, unit specific to indirect liquefaction is the synthesis reaction step. "ere, a consensus has developed that slurry#phase fluidised#bed reactors are preferable. The ma<ority of recent work in the field has therefore concentrated on the development of improved catalysts, which need not be specific to a particular process. ynthesis reaction technologies have applications outside coal liquefaction. 9n particular, there has been substantial BF0 on processes intended to convert natural gas to liquids. ince these processes all involve the preliminary partial o2idation or steam reforming of natural gas to produce synthesis gas, there seems no reason in principle why they could not be applied to a synthesis gas derived from coal. Two of these, developed by *obil and hell, have been put into commercial#scale operation. 4thers have been tested only to a pilot plant scale, including processes from @5Dvaerner, E22on and yntroleum Corporation, all similar to the hell process -or the asol process/ e2cept for their use of different and proprietary catalysts. asol also offers a natural gas synthesis process, the asol lurry 5hase 0istillate - 50/ process, based on its asol process synthesis technology. 9ndirect liquefaction involves, as a first step, the complete breakdown of the coal structure by gasification with steam. The composition of the gasification products is then ad<usted to give the required mi2ture of ". and C4, and to remove sulphur#containing catalyst poisons. The resulting +synthesis gas, is reacted over a catalyst at relatively low pressure and temperature. The products may be paraffins, olefinic hydrocarbons or alcohols -particularly methanol/, depending on the catalyst selected and the reaction conditions used. The only commercial#scale coal liquefaction process currently in operation is the indirect asol process of outh !frica. asol produces gasoline, diesel fuel and a wide range of chemical feedstocks and wa2es from three plant.

T!e Sasol Pro ess

The asol process is based on the ;ischer#Tropsch -;T/ liquefaction process operated on a large scale in =ermany and elsewhere in the $%A's and $%:'s, although over :' years, continued development by asol has resulted in substantial improvements in the synthesis step. The production of synthesis gas by the 1urgi gasification of coal has remained relatively unchanged e2cept for an increase in gasifier siKe. The first plant, asol $, was built at asolburg in outh !frica in the mid#$%5's with a capacity of ~)'''bblMday of gasoline. The much larger asol . and A plant were completed at ecunda in $%&' and $%&. respectively. These plant were each designed to produce 5','''bblMday of gasoline, together with substantial quantities of other products for use as chemical feedstocks, from the processing of A',''' tonsMday of coal. ! schematic diagram of the asol process, based on the asol . and A plant, is shown in ;igure $.. The developments by asol have resulted in several changes to the processes now used. asol uses low#temperature ;ischer Tropsch -1T;T/ and high#temperature ;ischer Tropsch -"T;T/ for synthesis. 1T;T is used e2clusively at asolburg and comprises the older !rge fi2ed#bed technology as well as the newer#generation slurry#phase ;T process. !t ecunda the "T;T process is used e2clusively and comprises older circulating fluidisedbed technology as well as new#generation asol !dvanced ynthol - ! / technology. ! technology, first used in ecunda in $%%5, gives the asol . and asol A plant a capacity of ~$5','''bblMday of crude equivalent of white products plus chemicals. The incoming coal is wet#screened and the G5mm fraction directed to the steam boiler whilst the R5mm fraction goes to the 1urgi gasifiers. Baw#gas clean#up on both sites is carried out using conventional 1urgi Bectisol with cold methanol as the wash medium. The 1T;T process operates at .''# .5'HC and .'#A'bar, and produces paraffins and wa2es using an ironbased

catalyst. The "T;T process operates at A''#A5'HC and .'#A'bar, again with an ironbased catalyst, and produces a lighter, more olefinic product slate including gasoline, petrochemicals and o2ygenated chemicals. The gasoline produced by upgrading the primary products is of particularly good quality. The 50 process was developed for monetising natural gas and comprises natural gas reforming, slurry#phase ;T and mild hydroprocessing to produce naphtha and an e2cellent diesel. The 50 process uses a cobaltbased catalyst specially developed for the slurry#phase system. The naphtha, because of its paraffinic nature, has a low octane number and as such is poor quality for gasoline, but it is a very good cracker feedstock. tudies carried out by asol have shown that "aldor TopsTe autothermal reforming, used to reform the natural gas with o2ygen, is the most appropriate for the ;T process. "ydroprocessing of the products is very mild and the slurry#phase ;T has been demonstrated commercially in a .5''bblMday unit at asolburg. asol believes that this relatively simple three#step process is superior in all aspects and can be built and operated economically. ! schematic diagram of the 50 process is shown in ;igure $A.
*ethane Beforming Cryogenic eparation eparation and ;ractionation lurry 5hase Beactor Bectisol ?aphtha 9somerisation 0ewa2ing =asoline 0iesel =ases !sh Coal Becycled ynthesis =as

Figure 12. The &asol process

!ir eparation "ydrogen 5roduction !ir weet ?atural =as weet ?atural =as E2ternal Becycle ;uel =as Tail =as ?aphtha 0iesel Treated >ater 42ygenates 5lant Effluents @;> team yngas Beaction >ater >a2 N Cond 5lant Condensates team F @;> to 4ther 6sers 4. 5roduct >ork#6p ;#T ynthesis ynthesis =as 5ower =eneration 6tilities Effluent Treatment

Figure 1 . &&P# process

T!e /o.il /TG Pro ess

The *obil methanol#to#gasoline -*T=/ process produces gasoline from coal or natural gas in two distinct steps. The process has been taken to a commercial scale in a $.,5''bblMday plant built in ?ew Uealand to process gas from the *aui field. !lthough this plant is still operating, it has recently been used solely for methanol production, for which the economics are currently more favourable. 9n the first step, synthesis gas produced by steam reforming of natural gas or by coal gasification is reacted over a copper#based catalyst to produce methanol in near to $''% yield. The reaction is carried out at .)'#A5'HC and 5'#3'bar. The second step involves partial dehydration of methanol to dimethyl ether at A''HC over an activated alumina catalyst, followed by reaction over a fi2ed#bed Keolite U *#5 catalyst. These reactions are strongly e2othermic, with the feed entering the reactor at A)'HC and leaving at :$5PC. The

reactor pressure is ..bar. ! series of reactions converts methanol and dimethyl ether to olefins and then to saturated hydrocarbons. Oields of material in the gasoline boiling range represent ~&'% of the total hydrocarbon product. >ith alkylation of by#product propane and butane, total gasoline yields of %'% at %A.3 B4? -octane number/ were achieved at the ?ew Uealand plant. The use of a fluidised#bed reactor offers advantages for temperature control and maintenance of constant catalytic activity over a fi2ed#bed system. The fluidised#bed reactor operates at an almost isothermal temperature of :$'HC but at a pressure of only Abar. 5rimary, gasoline#range liquid yields are lower, but there is little difference in final gasoline yields after alkylation.

T!e S!ell S/DS Pro ess

The hell *iddle 0istillate ynthesis - *0 / process produces a highquality diesel fuel from natural gas. 9t is a process being considered in many =as#to#1iquids -=T1/ processes in oil production. !s the process produces liquids from synthesis gas, although the technology is primarily aimed at a natural gas feedstock, synthesis gas generated from coal gasification would, presumably, be equally suitable. ! schematic diagram of the process is shown in ;igure $:. ?atural gas is first partially o2idised in an o2ygen#blown hell gasifier to produce synthesis gas. This gasification approach is preferred over steam reforming, despite the considerably higher capital cost and lower thermal efficiency, because it produces a synthesis gas with the correct C48". ratio of $8.. team reforming produces e2cess "., which in a stand#alone operation can only be used as fuel. The cleaned synthesis gas is then reacted over a proprietary hell catalyst in a fi2ed#bed tube#bundle reactor that is cooled in boiling water. The product is almost e2clusively paraffinic. The catalyst formulation and operating conditions in this step are deliberately chosen to give a much higher#boiling product than usual, since this minimises the production of hydrocarbon gases. 9n the final step, the wa2y heavy paraffin is catalytically hydrogenated, isomerised and hydrocracked in a single trickle#bed reactor over a proprietary catalyst to give products that are mainly middle distillates. The reactor operates at A''#A5'HC and A'#5'bar. ! high degree of product recycle is used to minimise the production of light products and to ensure that higher bp products are recycled to e2tinction. @y varying the hydrocracking severity and the e2tent of recycle, the product distribution can be ad<usted to give up to )'% diesel, with .5% kerosene and $5% naphtha. !lternatively, up to 5'% kerosene can be produced, with .5% each of naphtha and diesel.

THER/ALEFFICIENCIESAND C O 0E / I S S I O N S
The thermal efficiency of a plant is defined as the percentage of the CI of the input fuel that is in the finished products. Typically therefore -as shown in ;igure $5/, in an 1 E plant, coal is the only energy input providing feedstock to reactors and utilities and the products are finished transport fuels. ;or the 1 E process the thermal efficiency is )3.&%. This can be improved to <ust over 3'% if integrated gasification combined#cycle power generation is used. 4ther single# and two#stage direction liquefaction schemes have values in the range )'#3'%, lower values being usually caused by low#grade coal feeds which take up to 5% of the coal,s energy to drive off its moisture. "owever, the only processes currently under consideration with multiple reactors are e2pected to give )5#3'% to finished fuels( )3% is thus a fair average figure for direct liquefaction. 9ndirect liquefaction of coal is much less efficient, asol $ being only A3%. "owever, asol . and A are said to be much better, perhaps in the mid# 5's%. 5yrolysis is even less efficient because the solid product is only equivalent to coal, thus the thermal efficiency of finished transport fuels from the net coal input is likely to be :5% at most. "owever, in pyrolysis, the liquids are merely by#products of a solid fuel upgrading process. ;or the purposes of deriving the C4. impact of liquefaction, +standard, bituminous coal and crude oil can be selected -Table A/, each of which may be used to generate transport fuels.
Table . 7&tandar d bitu'inous coal and crude oil

The crude oil in Table A is similar to that marketed as heavy fuel oil.

Befineries with the necessary thermal or catalytic crackers have claimed thermal efficiencies of %A#%:% when allowance is made for electrical power generation. Thus an efficiency of %'% may be assumed for a hypothetical refinery producing only transport fuel. The standard coal may be converted to the standard crude oil, by direct and indirect liquefaction, at thermal efficiencies of 35% and )$% respectively, with all the carbon in the coal, not ending up in the +crude oil,, converted to C4. -either in power stations or furnaces/. ;or direct or indirect liquefaction this amounts to $.&: tonnes or A.''tonnes C4.Mtonne crude. ?ote that, even at $''% thermal efficiency, there is some C4. produced because of the release of '.') tonnes of hydrogen from coal.
;uel =as Becycled "ydrogen ;ractionation "ydrocracking "eavy 5roduct Becycle ?aphtha Derosene 0iesel ;resh "ydrogen Tube#@undle Beactor Cleaned ynthesis =as

Figure 1!. &che'atic diagra' o( &8#& process

$''% V 5:)$*> ..) $3.5 "ydrogen 5roduction Coal E2traction and ;iltration "ydrocracking 1oss 1oss
1oss

1oss
1oss

@utane Befining 5ower tation :.)

$..3

$&.) .$.: '.3 5rincipal )'.3 5roducts @y#products inc. 15= -:.$/ 5ower E2port $''.' ).) '.$ '.$ :..
..) '.)

$.5 $... ..5 ..A :.$ 5.5

Figure 1%. Ener gy (low diagra' (or a conceptual L&E plant

Stan%ar% -'els @ituminous coal Crude oil Carbon -%/ &) &).$ "ydrogen -%/ 5.5 $$.& 42ygen -%/ ).' # ?itrogen and sulphur -%/ ..5 ..$ ?et CI -*>hMtonne/ %.)) $$.:% 9n refining crude oil to finished fuels there would be a further '.A. tonnes C4.Mtonne of crude oil processed, giving $'.A$*>h of CI in the finished products. 9n all, ~3#$' times as much C4. is emitted in converting coal to transport fuels, when compared with crude oil.

PROSPECTSFOR I/PLE/ENTATION
9n $%&5 oil prices fell suddenly and, with brief e2ceptions, have remained low since. 9nterest in coal liquefaction for the production of transportation fuels has declined accordingly. !t present, only Capan is active in largescale process development, with a $5' tonnesMday plant in operation at Dashima near Tokyo, and China is participating in several collaborative feasibility and process development studies. 9t remains to be seen whether a future increase in oil prices will result in renewed interest and possibly in full#scale commercialisation. 9n the meantime, the emphasis has returned to coal upgrading.

Coal liquefaction, by whichever route, is capital#intensive and therefore benefits substantially from economies of scale. *ost studies on process economics have assumed that a full#scale commercial plant would produce 5','''#$'','''bblMday of liquid products. uch a plant would process $5,'''#A5,''' tonnesMday of bituminous coal or up to double that amount of sub#bituminous coal or lignite. !t the lowest end of this range, the annual consumption would be 5 million tonnesMyear of bituminous coal. The output from this plant would still be small relative to that from a typical modern crude oil refinery, where a throughput of R.'','''bblMday is common. Countries that might implement coal liquefaction must, therefore, have ready access to large quantities of coal. The economics of liquefaction depend strongly on coal costs and this coal must be delivered to the plant at a low price. ince coal is more difficult to transport than oil, it would, as a general principle, be better for liquefaction to be carried out in the country of origin and preferably at the point of origin. There will, however, be e2ceptions to this generalisation, particularly where construction costs near the point of origin are likely to be high due to remoteness or where there are already good rail links. 9f liquefaction is to be carried out in the country of origin of the coal, there must be sufficient reserves to last for a pro<ect lifetime of ~.5#A' years. The coal feedstock required for a commercial#scale liquefaction plant equates to A#: million tonnes oil equivalent -mtoe/Myear. This means a minimum reserve of about $''mtoe or $5' million tonnes of coal or A'' million tonnes of lignite for the required reserves and minimum reserveMproduction ratio -BM5/ of .5. ome .: countries meet this criterion -Table :/. The single most important factor behind the decision to implement coal liquefaction is the price of crude oil. The discussion and the analysis summarised in Table : provides a comparison of the likelihood of individual countries pursuing coal liquefaction given an economically#favourable crude oil price.

FUTUREDE1ELOP/ENT
The scale of oil#refining industry operations is such that it would be likely to be impractical to produce the same amount of liquids via coal liquefaction. 1iquefaction can only really offer a means of providing a supplementary source of hydrocarbon liquids. This is likely to be most attractive as a source of transportation fuels once the supply of crude oil can no longer meet demand for these products. 6nder these circumstances, it is likely that liquefaction#derived fuels would be distributed through the e2isting infrastructure which, inevitably, is controlled and operated by the oil industry. This would greatly simplify the distribution and marketing of liquefaction products. 9t is, therefore, likely that coal liquefaction will initially be implemented by the oil industry rather than by the coal industry. This trend is demonstrated by the fact that most of the private#sector process developments have been undertaken by oil companies. =iven this scenario, it is likely to be beneficial for a liquefaction plant to e2port its products directly to an e2isting refinery, whether as raw materials for further processing or as finished fuels for simple blending into the refinery product pool. !s many of the ancillary requirements, such as Country 1arge 1ow cost 1arge oil BM5 Bestriction on oil availability 5otential for implementation coal coal demand BatioWW reserve production Oes ?o 6 ! W %.& ecurity of supply and possible political and environmental considerations Canada %.. 1ow reserves *e2ico AA.) @raKil $5.5 ?et oil importer with low reserves Colombia $$.3 9ncreasing demand with low reserves IeneKuela 5%.5 @ulgaria ' ?o oil or gas CKech Bepublic ' ?o oil or gas =ermany ' ?o oil reserves =reece ' ?o oil reserves "ungary ' ?o oil reserves 5oland ' ?o oil reserves

Bomania A..5 pain ' ?o oil reserves Turkey ' ?o oil reserves 6D 5.. 1ow oil reserves ;ormer oviet 6nion .5 5olitical and economic considerations outh !frica ' ?o oil reserves Uimbabwe X ' ?o oil reserves but no demand( economic considerations !ustralia & 1ow oil reserves China .'.5 4il reserves remote from demand 9ndia $5.) 4il reserves remote from demand 9ndonesia % 1ow oil reserves with increasing domestic demand Capan ' ?o oil reserves W 4il demand in 6 ! is ten times that of most other countries WW @5 tatistical Beview of >orld Energy $%%&
Table !. 3nalysis o( potential i'ple'enting countries

power and chemical feedstock supplies, would be common to both the refining and liquefaction processes, there would be opportunities for sharing some of the necessary facilities. Equally, many of the unit operations are likely to be similar in both the refinery and the coal liquefaction plant. !gain, this favours implementation by the oil industry. 9ntegration may involve simply a shared product#blending and e2port facility. Even this minimal integration of facilities would offer significant capital cost reductions. tudies have shown that a $.','''bblMday refinery could accept the whole output from a 5','''bblMday liquefaction plant with minimal change in overall product quality, so the typical refinery of .'','''bblMday would cope easily. ! more comple2 integrated approach could involve the liquefaction plant and the refinery transferring products for further processing. 5rimary distillates from the liquefaction plant could be blended and processed with the equivalent feedstocks in the refinery. 9n addition to capital savings, the greatest benefits would come from e2porting low#value products from the refinery to the liquefaction plant, for gasification to produce ".. These materials may include vistar from a visbreaking unit or high#sulphur coke from a delayed coker, for which the refiner may otherwise have no outlet. This also increases fle2ibility in determining the refinery product mi2, possibly by making additional use of coking to increase distillate production. The liquefaction plant would no longer require additional coal as a supplemental utility fuel. The <oint benefit has been estimated to be equivalent to reducing the cost of liquefaction by 6 Q.#AMbbl. This level of integration does not risk compromising the operability of an oil refinery. *uch more comple2 schemes are possible, although refiners may be reluctant to implement these until they have gained confidence in the liquefaction plant operation. ome e2amples are the use of common ". production and power generation facilities, steam, cooling and other utility systems. ! total reduction in overall liquefaction costs of ~6 Q5Mbbl has been suggested as achievable. There are a relatively small number of unit operations specific to a particular coal liquefaction process. These are generally serviced by a large number of ancillary units. The ancillary units and, in many cases, the core units are now well proven on a large scale in other applications. ince )'# 35% of the total estimated capital costs is accounted for by these units, overall process costs have become comparatively well defined. This is one area in which liquefaction technologies have advanced significantly since the mid#$%&'s. >hilst this is a considerable advantage in some respects, the potential for dramatic reductions in overall costs are limited as the core processes have evolved slowly through incremental improvements. 9t seems likely that this trend will continue meaning that further cost reductions will probably result from general engineering advances as much as from genuine process improvements. The cost of catalysts is so high that what appear quite e2pensive means for decreasing it could be worthwhile, eg increasing reactor siKe substantially in order to reduce the catalyst operating temperature.

CONCLUSIONS
De$elo(&ent Stat's
7 1iquid transportation fuels are characterised as having a hydrogen content of ~$.#$5%, whilst typically coal has a 5% hydrogen content

and higher carbon content. Coal#to#liquid conversion processes require hydrogen to be added or, alternatively, carbon to be removed. ?umerous +direct, and +indirect, processes have been developed for this purpose. 7 Considerable BF0 into liquefaction processes was carried out during the $%3's and early $%&'s, mainly in the 6 !, Capan, the 6D and =ermany spurred on by oil price shocks. ince then, development has largely been put on hold. 7 outh !frica is the only country presently operating liquefaction plant. 9t has large reserves of coal but no oil and gas. Trade embargoes over three decades to the mid#$%&'s drove large#scale application and up to )'% of transportation fuel requirements have been met from coal. 7 Coal liquefaction is now a largely proven technology for the manufacture of useful liquid products.

Bene-its o- t!e Te !nolog+

7 Coal is abundant but, in comparison with liquid fuels, inconvenient to handle and unsuited to some applications particularly transport. Coal liquefaction provides the ability to manufacture transportation fuels from coal. 7 The technology provides insurance against depleting oil stocks and oil supply problems.

Disa%$antages o- t!e Te !nolog+

7 1iquefaction processes typically achieve an energy conversion -% CI of the input fuel converted to finished products/ of )5#3'% -direct liquefaction/ and 55% -indirect liquefaction/. 7 Converting coal to transportation fuels results in ~3#$' times as much C4. being emitted, compared with converting crude oil. This increase in C4. emissions at the processing stage has the effect of raising overall C4. emissions from transport by ~5'%, compared with transport based on conventional, refined petroleum products.

/ar2et Pros(e ts
7 5resently, only Capan is active in large#scale process development, whilst China -increasingly a net oil importer and containing areas remote from sources of oil/ presents the strongest adoption prospects. 7 The commercial viability of coal liquefaction rests with the overall economics of the process. This depends on the availability of significant quantities of poor#quality, low#cost coal, and the unavailability or otherwise relatively high cost of oil -and gas/. !ppropriate market conditions are likely to emerge around .'$'( many countries are then likely to be involved.

AC"NO3LEDGE/ENTS
This review was funded by the 0T9,s Cleaner Coal Technology 5rogramme.

BIBLIOGRAPHY
7 ynthetic ;uels from Coal8 tatus of the Technology. Bomey, 5aul and 9marisio -Eds/. 5ublished by =raham and Trotman for the European Communities -$%&3/. 9 @? $#&5AAA#$'A#$. 7 Coal 1iquefaction8 ! Besearch ?eeds !ssessment. 6 04E Beport 04EMEB#':'' -$%&%/. 7 6 04E Coal 1iquefaction and =as Conversion Contractors, Beview Conference. 5roceedings. 7 Energy for Tomorrow,s >orld. >orld Energy Council. 5ublished by Dogan 5age -$%%A/. 9 @? '#A$.#$')5%#%. 7 The Coming 4il Crisis. C C Campbell. 5ublished by *ulti# cience 5ublishing Co -$%%3/. 9 @? ' %')5..#$$#'. 7 @5 tatistical Beview of >orld Energy -$%%&/. 7 Beport of 5hases 9#9Ib of @ritish Coal 1iquefaction 5ro<ect. 0T9 Cleaner Coal Technology 5rogramme Beport ?o. C4!1 B'3%. 7 !ndus, Daarstad and Dowal, tuttgart $%%). 7 !utomotive ;uels urvey, 9E! !;9 Iols.$#A, $%%) Iol.: $%%%. 7 5etroleum Befining, =ary F "andwerk $%35. 7 Technical Besponses to Climate Change. 9E! =reenhouse =as BF0 5rogramme $%%3.
5epartment of $rade and -ndustry

(or the 0T9M5ub 6B? %%M$$.'

Direct and Indirect Liquefaction

There are two methods of deriving liquid fuel from coal: direct coal liquefaction (DCL) and indirect coal liquefaction (ICL). The major difference between the two is that DCL is the rocess of creating a liquid fuel from coal while ICL involves gasif!ing coal first and then using this s!ngas to roduce liquid fuel. "lthough DCL methods are more efficient# the rocess involves large amounts of aromatic com ounds. "s a result# an! liquid fuels obtained through DCL methods would have worse e$haust emissions from the combustion rocess than etroleum fuels. "lso# DCL does not give an! o tions for ca turing C%& emissions.

Li,'e-a tion

9ncreasing "MC ratio is a must( Two options

5 Be<ect carbon
*ost pyrolysis processes

5 !dd hydrogen
0ry pyrolysis with hydrogen not effective Coal liquefaction( use a solvent to effect hydrogenation

5roducts often solid at room temperature 1iquefaction defined by solubility

Coal Con$ersion Pro esses

CarboniKation and 5yrolysis
Y 1ow severity -*ild =asification/ Y "igh temperature

0irect 1iquefaction
Y 4ne stage reactor technology Y Two#stage reactor technology Y Co#processing Y "ybrid

9ndirect 1iquefaction
Y =as reactors Y lurry reactors

Coal Li,'e-a tion A((roa !es

5yrolysis or mild gasification 0irect coal liquefaction 9ndirect coal liquefaction Co#processing

@ioliquefaction

&ubstantial o0erlap in the che'istry o( 'ild gasi(ication2 direct coal lique(action2 and co* processing