Overview of Fischer-Tropsch Synthesis in Slurry Reactors

Chinwse J. of Chem. Eng, 8 (3) 255- 266 (2000)REVIEWSOverview of Fischer-Tropsch Synthesis in Slurry ReactorsDing Baiquan(丁百全)* and Li Tao(李涛)Department of Chemical Enginering, East China University of Science and Technology, Shanghai 200237, ChinaA.A.C.M. Beenackers and G.P. van der LaanDepartment of Chemical Eagineering, Univernity of Groningen, Nijenborgh 4, 9747 AG Groningen, The NetherlandsAbstract A brief review of Fiecher-Tropech syntheais specially in slurry reactors is presented, covering reactionkinetics, activity and selectivity of catalysts, product distribution, effects of process parameters, mass transferand solubility of ga8. Some important apects of further research are propoed for improving both tbeories andproduction.KeywordaFischer-Tropech aynthesis, slurry reactor, kinetice, catalyst, mass transfer1 INTRODUCTION260C, the pressure from 0.1 to 3 MPa, preferably 1.0Fischer-Tropsch Synthesis (FTS) reaction has beento 2.0MPa. About half of the processes use octa-known for about 90 years. Earlier in the twenty cen-cosane as liquid phase and some adopts paraffin. Thetury research work on FTS was focused on Co catalystratio of H2 to CO is from 0.5 to 3.5. It is depended onand mainly by German scientists. The first commer-the methods or producing syngas: low ratio for syn-cial plant for producing hydrocarbons by FTS reactiongas by coal and high ratio when produced by naturalusing Co catalyst was started in 1936 in Germany.gas or others. The catalysts used in FTS reaction areThen, from the middle of this century, scientists ofFe, Co and some bimetal such as RuFe, promoted bymany countries such as Germany, France, Britain,Cu, K, Al, Si, Na and Zn. Fe catalysts are the mostUnited States, China, Japan etc. were developed topopular and the cheapest. While Co catalyst is usedFTS research. The catalyst study bad been shiftedin a few cases.to Fe-based ones and two types of reactor, ARGE andSince a slurry reactor with 5 m diameter has beenCFB reactors used by SASOL FTS plant, both with Feused for FTS in Sasol, South Africa, from 1985, it iscatalysts, designed by Ruhrchemie in Germany, wereimportant to summarise the research of FTS reactionconstructed in South Africa.in slurry reactors in order to do further study theoret-Many scholars have done the research work onical basis and improve the process fficiently. Table 1FTS. At the beginning, the reactor used in FTS stud-is the extracts from twenty papers on FTS in slurryies and industry was only fixed-bed reactor. Lately,reactors.fixed-bed with recycle reactor, fuidized reactor, ertrained reactor, tube wall reactor, trickle-bed reactor,2 REACTION MECHANISM AND KINET-slury reactor were gradually used in R&D work. TheICSreactors used in experiments were mainly fixed-bed reHuf and Satterfield4) developed a rate expressionactor (FBR), continuous stirred tank reactor (CSTR),from the experimental data measured in a CSTR,bubble column slurry reactor (BCSR) and the reactorswhich indicated the inhibiting effect of CO and H2Oused in industry were mainly FBR, BCSR, entrainedquantitativelybed reactor (EBR), and tube wall reactor (TWR).- Rga+co = ab'pco嘱。/(PH2o + b'pcoPHz) (1)Among these, the slurry reactors are considered to bemore efficient because of it's supper advantages suchIn the model, b' is a function of temperature only,as high activity and eficiency, fAexibility to variablebutc the model reduces to the form proposed byratio of H2/CO, low pressure drop, easy withdrawalAnderson[21] if b' in his expression is inversely pro-of reaction heat and control of temperature, catalyst中国煤化工renewal in-line, simple structure and low investment.Generally, the temperature in slurry reactors forTYHCN MH G + bpa2o)FTS ranges from 200 to 280°C, most from 220 toSatterield and his co workersl22] studied the efectReceived 1999-09- 25, accepted 2000-04-13.* To whom correapondence should be addreased.Table 1 _ Brief information of Flecher-Tropsch bynthosis in slurry reactorsRef. T, 0P. MPaSolventCat. componentsHa/COProductsReactor type Xvextowaxprecipitated≤0.8waxCSTR, 1L,0.70- -0.92SH 105100Fe, 1.3KzO, CaO, Al2O3600r.min -12] 220- -280 2.0parafn oil Co2(00)g. 0.035 Zn(OPr), .0.5- -2.0C1- Os,Ce+, hydrocarbonCSTR,0.34- -0.710.066Si02 or Al2O33] 250.5octacosaneruhrchemie LP33/810.66paraffine, olefinsCSTR100Fe/4.3Cu/4.1K/25SiO24) 232- -263 0.445- -0.48 octacosaneFused-magnetite cat. C-73, 2.0- -3.0Al2O3, 0.55- 0.81 liquid hydrocarbons (C>4) and wster0.33--0.980.5- 0.8K2O, 0.7- -1.2 CaO, <0.4 SiO25] 2630.79f.-m. cat., 0.58K/100Fe/0.74Ca/0.76A10.69- -0.90 Cs:olefn, parafio, alcohol, beavy wax6] 2681.2parafnFe-based0.67BCSR7] 200- -300fused-ironcat. C730.71- -1.38} 275.2parafinprecip. Fe cat.0.7gsoline, diceel, alcohol, heavywax9| 235--280 0.68- -1.31 bard parafin promoted fused iron cat, C-730.6- -2.3C1- 4:olein, parafins,→C3ocSTR0.16- -0.93[10] 250--270 1.5precip. 100Fe/0.3Cu/0.5K0.72C13.2, C2 411.4, Cs-120.4, C12- 1865.00.71[1] 266- -268 12unsupported precip. iron cat., Fe/Cu/K2O 0.67C13.2, C2-431.3, Cs-1153.6, C12- 1810.0,0.89C10+1.9[12]0.8- -1.5ruhrchemie cat, (precip. Fa)C15.7, Cz- 427.0, Co-1240.0, C1s27.30.336”(13]1.5precip. Fe cat,, Fe/Cu/K2OC17.8, C2-424.5, C-1141.5,5 C12 -1815.4,0.868C19+10.8, wax[14]1.48n-octacosane precip. iron cat.0.660.70 C1-20.54- -0.82, CH40.25- -0.37, C2-40.10 - CSTR0.74- -0.81100Fe/5Cu/6K/24SiO20.15, C-110.13- -0.2, C12+0.62- -0.74(15]王1.0 -2.0vestowaxprecip. Fe cat, 99Fe/1K0.5- -1.9msinly wax0.2-0.7sh 105([16]团1.0- -1.5n-octacosane precip. Fe cat., 100Fe/22Cu/1KgCO31.0olefin/parafin C20.33- -2.11, C3.11-0.2- -0.65.52, C42.83- -5.26n-octacosane Co cat., 21.4Co/3.9Mg1.5- -3.5u to C3oOSTR0.17- -0.73diatomactour earth[18] 206.8tetralinultrafine Fe2OsCH46.4, C22.0, C32.5, CO217.5,0.608Alkene/Alkane C20.26, C30.92,[19] 230- -265 0.77- -2.98 n-octacosane precip.Ruhrchemie at, 47.3F/2.36CuC15.3, Cn 423.1, Cs-1234.1, C13- 120.6, CSTR0.1- -0.65/2.69SiO2/1.94K/0.02A1/34.360C20+ 16.9(20] 235-265 1.5-3.0noctacoane RuFe,100F/4.3Cu/4.1K/0.67- -1.0 CH48.5, C2-429.8, C6-125.3,0.28- -0.57259i02C12+36.6Overview of Fischer -Tropsch Synthesis in Slurry Reactors257of liquid composition on the reaction rate of slurryrich feeds were usedFTS. They used the kinetic expression of Hf'g(4 andthought that at conversion of about 60% formation of- RHa+Co = kcH,/(1 + Kc1H2o/cocH2)water was low and the equation reduced to first orderYates and Satterfieldlzn] used a CSTR to measureexpressionthe intrinsic kinetics of the FTS over cobalt cata-RH2+co = koPHzlyst at 220- -2409C, 0.5- -1.5 MPa, H2/CO feed ratiosof 1.5-3.5, hydrogen conversion 6%- -68% and car-Deckwer et al!1] used potassium promoted precip-bon monoxide conversion 11%- -73%. The Langmuir-itated Fe catalyst to study the rate inhibition by CO2Hinshelwood type equation was found to be the bestand H2O in a stirred autoclave and slurry phase. Theone to represent the experiment and well fts the datarate data were measured under variations of tempera-from previous studies on this catalystture and H2/CO inlet gas ratios. They confirmed thatthe syngas consumption rate of FTS could be inhib- RHa+co = apcop12/(1 + bpco)2ited by competitive adsorption of CO and CO2. AtZimmerman et al!3) obtained the experimentallow H2/CO (≤0.8), water was converted by the waterdata in a laboratory stirred tamk reactor with the com-gas shift (WGS) reaction and inhibition by water ad-mercial iron-based catalyst Ruhrchemie LP33/81 SuS-sorption on the surface site was negligible, while atpended in the molten octacosane. They suggested ahigher H2/CO ratios or if water vapor was introducedstirred slurry tank model and developed a reactionwith the inlet gas, the rate inhibition was also by ad-rate equationsorption of part of the water. They thought that athigher H2/CO ratios the CO insertion mecbanism(23)R=2 Unjrheand the carbide theory|24] seemed to be more appro-priate.Kinetic models of FTS reaction in slurry reactorsSanders et al15] had checked the data repro-from literature are listed in Table 2, which indicatesducibility of FTS on K-promoted Fe catalyst underthat the reactions and mechanism involved in the FTSconditions of arbitrarily alternating operating param-process are very complex, but the overall reaction rateeters in three laboratory BCSRs. They evaluated theequations they used are very simple and the WGS re-rate constants as first order (k = kH/Ccat) based onaction plays an important role in the process. Mostthe mixed slurry phase model25] and the back mixedof them used Langmuir type equation as the kineticslurry phase model26). The results showed surpris-model for FTS reaction. Thus , generally, we can useingly that it was not possible to discriminate betweena common equationthe two models and deviation to the plug fow wereexceeded 3.2%. For the same batch of catalyst the一RH2+co =apcoPH2/(K'coPco + KHpPH2+reproducibility of data on K content experiments wasKcorPCoz + KHzOPHzo)"fairly good under all experimental conditions and themean deviation was about 15%. If runs with differentIf the adsorption of some components is weak corn-batches of catalyst, the rate constants may difer by aparing with other components, they can be eliminatedfactor of 2. .from the equation. Thus, the equation usually can beZimmerman et al.10calculated the catalyst effec-simplified to many forms that are suitable to specialtiveness factors by assuming frst order reaction kinet-conditions. The simplest form is first order equation.ics (RH2+co = kopH;) and single reaction stoichiome-Normally, water inhibits the FTS reaction, especiallytry and compared to those obtained from experiments.for the catalysts that have high activity of WGS reThe calculated and experimental results were in goodaction. So, if the high ratio of H2/CO to be used oragreement, but the calculated values were consistentlyif water vapor is introduced into the system, it musthigher than the experimental values. They thought be careful that the inhibition effect by water might bethe over-prediction might because by basing the Thielestrong.modulus on the difusivity of H2 or by neglecting rateinhibition by water.3 ACTI中国煤化工y or CAT-Withers et al2] verifed a kinetic expression thatALYST出CNMHG，took water inhibition into account as below andThe activty aru seicllviuy ul calalyous are the firstshowed that the cobalt catalyst had low activity forcriteria when commercial production by FTS withthe WGS reaction and performed poorly when COhigh eficiency is in mind.ChineweJ. Ch. E.8 (8) 265 (2000)258Chinene J. Ch. E. (Vol. 8, No. 3)Table 2 Kinetic models of FTS reaction rate in slurry reactorsef.P, MPaT,c。Hz/COCatalystModel, RHg+co=21]reduced, fused FeapcoPH2/(Pco + bpH2o)41]0.445232- -2630.55- -1.81fused, magoetite Fearoo喉/(mH2o + VpcoPH2)-1.482]1.480.69magnetite Fe28]1.220- 2600.48- -0.61precipitated FebrcH2/(1 + bzccoz/oco)1]0.220- -260≤0.8procipitatod Fek= kn/cas10]1.5250- -270fused Pekopco唱」/(bpH2o + PH2Pco)2.07220- 2800.5-- -2.0 .cobalthcu2/(1 + Kcr2/coc;2)17]0.5- -1.5220- -240.5- -3.5apcopr/(1 + bpco)23}]2500.66Ruhrchemie LP33/81 Fe29}240- -2650.5- 0.8FkOPCOpH2/(pco + 0PHzo)30)250- -280precipitateda1 exp(a2/ RT)CH/(1 + 3GH20/cco)Kolbel et al11] and Kolbel and Ralek(31] had done and V(STP) = 0.3- -0.6Lmin-1. The FTS catalystthe research work of the Rheinpreussen demonstra- maintained steady -state activity for the entire run oftion plant performance that is regarded as the most about 400h on stream at a level comparable to thesuccessful bubble column slurry reactor (BCSR) op-cobalt catalyst operating alone. The rate of the FTSeration by the literaturel14]. All subsequent attempts, reaction in the mixture fllowed a rate expression forMitra and Royl32), Farley and Rayl33), Kuo[84] andthe cobalt catalyst alone. The wGS catalyst exhib-Kuol13], to reproduce or exceed the Rheinpreussen ited a very slow loss of activity. The mixed catalystdemonstration plant performance have not been suc- system exhibited stable long-term FTS selectivity, butcessful. A high single pass conversion 89% of syngasthe presence of the WGS catalyst increased the ex-was achieved in Rheinpreussen's test and the bulk cat- tent of the secondary reactions, such as 1-alkene hy-alyst activity (space-time-yield) was also high.drogenation and isomerization, thereby lowering theWithers et al[2] described the preparation, charac-overall selectivity to fuel range products.terisation and performance of cobalt carbonyl clusterHuang et. al[18] reported their work of the inAu-based catalysts for use in the slurry phase FTS tech-ence of different methods for catalysts activation onnology. They thought that using metal carbonyls asthe activity and selectivity of catalysts, using ultra-active metal precursors allows for the control of metal fine iron oxide upon catalyst and tetralin as solventparticle size on the support surface, thus offering the in a CSTR. The results showed that the activation ofpotential for better control of activity and selectiv-Fe2O3 with high surface area in CO resulted in acity of Cσ- C20 hydrocarbons in the FtS reaction. tivity that was three times that when activated in H2They concluded that silica as the support provided the or directly in the syngas. But methane and CO2 sehighest catalyst activity and 3.5% cobalt, 6.6% zirco- lectivity obtained on these catalysts were similar andnium in the catalyst Coz(CO)g/Zr(OPr)4/SiO2 was seemed independent of the catalyst activation. Theythe most active system when used in the FTS slurry tbought that the active catalytic phase might be thereactor and gave the best diesel liquid fuel selectivity. same for the three pretreatments and the activity dif-Selectivity patterns correlated to the Schulz Flory pre-ference of the differently activated catalysts could bediction. Diesel fuel product produced by this catalystattributed to difference in the concentration of activewas high quality.sites on the catalyst surface rather than the surfaceChanenchuk et a,l35] established the feasibility of area since the particle size varied within 30% only.using a mechanical mixture of a Co/MgO/SiO2 FTs中国煤化工l tbe performance (ac-catalyst and a Cu-ZnO/Al2O3 water gas shift catalystof the precipitatedfor hydrocarbon synthesis in a CSTR. The mass路HHC NM H Gra stedy state comtios of two catalysts were 0.27 and 0.51. The operating ditions with a long run (up to 560h of continuousconditions were 240C, 0.79 MPa, H2/CO is 1.0 -2.0 test) in a CSTR. High syngas conversions (>80%)万男数辖ber, 2000Overview of Fischer Tropech Synthesis in Slurry Reactors259were achieved and the catalyst deactivation rates were expected to give high yield of Cs- 18 products only.moderate (1.3%- -2.1% per day). Selectivities wereGuan et al.(37]studied the activity change of F-fairly stable with time. Methane selectivity (2.1%- T synthesis over modified Fe/MnO catalysts. They3.7% of total hydrocarbons) and gaseous hydrocar. found that the activity was increased when 1%Cu wasbon (Cz- -C4) selectivity (9%- ~15%) were very low， added to the catalyst and the conversion of CO andwbereas the selectivity of C; bydrocarbons was high. the selectivity of CF- C were raised.They found that the catalyst with the components ofAnother important criterion for evaluating the se-100Fe/5Cu/6K/24SiO2 was higher than that of any lectivity of catalyst is the ratio of olefins to paraffins,other known iron FTS catalyst developed for slurry and the high ratio is preferred. But only a few au-phase operation.thors listed it. Satterfield and Stenger/16) gave theComparison of the activities for several FTS cat- ratios of olefin to parafin from the experimental re-alysts in slurry reactors is shown in Fig.1. The com- sults on precipitated iron catalyst. The values weremercial Ruhrchemie catalyst 100Fe/3Cu/4K/8SiOzC2 0.33- -2.11, C3 3.11- -5.52, and Cs 2.83- -5.26.tested by Bukur maintained higher activity up to 600 hwhile the activity level of unalkalized precipitated Fe 4 PRODUCT DISTRIBUTIONcatalyst measured by DonellylI9l was lower, which The distribution of products is another important fac-can be seen from Fig. 1tor to the FTS process. lron catalysts form mostlystraight -chain hydrocarbons. Such products (in the8range of about C10 to C2o) are of particular value asdiesel and jet fuel. Heavier waxy components (C20)are of low value as fnal products, but can be hydro-罗40cracked to lower molecular weight fuel. Low methaneyield is desired. 1-Alkenes can be used as chemical400800200feedstock (especially C2 to C4) or can be reformed togasoline (Cs to C12). Oxygenated organic are formedFigure 1 Comparison of the activity of FTS catalystsT-O-S,hin modest amounts and generally must be removedfrom the final product.in slurry reactors■ultrafne iron oxide catalyat, 260 C, 0.8 MPa,The first attempts to analyze the Fischer -Tropsch3.2L-h-1.g-', Hz/CO is 1.0(98product distributions mathematically as a polymeriza●commercial Ruhrchemie catalyst, 100Fe/3Cu/4K/8SiO2,tion process that could be treated in terms of relative260%C, 1.48 MPa, H2/CO is 0.58, SV=2.5Lb-1.g-1[14]probabilities appear to be Herington/38). He postu-100Fe/4.3Cu/4.1K/258iO2, 250 °C, 1.5 MPa,lated that n-parafins and a- and B-olefins were formed21 mmolh 1.g~1, H2/CO is 0.87120]by stepwise addition of a methylene radical onto a◆unalkalized precipitated Fe catalyst,growing chain on the catalyst surface.260°C, 1.48 MPa, 1.2Lh-1g-1, H2/CO=1.0l19)Weller and Friedel(39]proposed a method basedon probability considerations tbat methyl-substitutedTable 3 gives tbe comparison of selectivity of FTS isomer distributions for saturated bydrocarbonsreaction for several catalysts in slurry reactors.within any one C number in the range Cs to Cg formedMost of the catalysts had good selectivities since on a cobalt catalyst could be correlated. Anderson etthe products of C2-11 and C12- 18 were high, which all40] extended this 8ssumption to include the carboncan be seen from Table 3. However, the best catalysts oumber distribution.Table 3 Comparison of selectivity of FTS reaction in slurry reactorsRef. Cat. T, 0p, MPaH2/CO SV,Lh-1g-1 Ci C2- 4C3-11C12-18C19+(11] Fe2671.20.67.231.353.610.01.9[13]2602..824.541.515.4[12] Fe2630.8 -1.50.75.7中国煤化工[10]Fe 250- -2701.50.721.3.2[19]230- -265 0.77- -2.98 0.6- -2.4DHC NMH G.8 C20+:16.9 .[20] Fe 235- -265 1.5- -3.0 0.67- -1.01.0 -4.0412+: 00.6[14Fe261.480.66- -0.72.2- -3.42.3- -3.7 10-15 13- -20C12+:62-74ChinomJ, Ch. B.8 (3) 288 (2000)260ChineseJ. Ch. E. (Vol. 8, No.3)Satterfield and Hfl42] studied carbon number dis- catalysts could be accurately described by a modifedtribution of Fischer Tropsch products formed on anSchulz-Flory distribution with two chain growth prob-lron catalyst in a well-stirred slurry reactor at 234 abilities.to 269C and 790kPa. A precise linear relationshipZimmerman et al.3I developed the slurry stirredbetween the log mole fraction mn of products of car-tank model (to be mentioned above) to account forbon number n, as predicted by the Flory molecular the formation of parafines and olefins and also for theweight distribution was obtained. It held over morewater-gas shift reaction and secondary hydrogenationthan four orders of magnitude of mn, values of n of of olefins. Non-Schulx Flory distribution was obtainedfrom 1 to 20, and over a wide range of gas composition. due to olefn re adsorption fllowed by chain growth.The chain growth probability factor a was increasedThe Schulz-Flory diagram of product distributionslightly from 0.67 to 0.71 in the temperature rangeof from 7 reports is displayed in Fig.2. It can be seen35 C as tbe ratios of H2/CO varying from 0.85 to 16.from Fig.2 that whatever catalysts were used, theAndersonl42] pointed out that olefins from Fe werecurves in Schulz-Flory diagram for product distribu-often found in amounts exceeding 50% in each carbon tions have no significant diference under the normalnumber and more tban 60% of these were a olefins.conditions of FTS reaction.The olefn fractions were more or-less independent ofcarbon number. For products from Co, both the frac-tion of total olefins and the fraction of a-olefins weresmaller, and both decreased with increasing carbonnumber. A suggestion for the reason of carbon num-ber distribution is that it is related to the pore sizeof the catalyst or to the size of the metal crystallites,particularly, that the length of the growing chain is120 30 40limited by the dimensions of the pores or the metalcarbon numbercrystallites.Figure 2 Schulz-Flory diagram of productsSatterfield and Stenger[16] compared the perfor-distribution for FTS reactionmance of precipitated iron- copper-potassium catalyst◆unalkalizxed precipitated iron catalyst, 260 C,in a CSTR with that of the same catalyst from1.48 MPa, 1.2L.h-1g-i, H2/CO is 1, T-O-S=665 b[18];■Ruhrchemie catalyst(100Fe/4.3Cu/4.1K/25SiO2),fixed-bed reactor studies and that of a reduced fused250 C, 1.5 MPa, H2/CO is 0.67，magnetite catalyst in the same CSTR. They con-21 mmolh-lg-1, T-O-s=619hl20];cluded that the ratios of olefn/parafin for Cz- C4▲precipitated and reduced fused iron catalyst,250C, 0.987MPa, 1.1Lbh-1g-', H2/CO is 1.42[161;of precipitated iron catalyst and fused magpetite cat-x sulfur poisoned reduced fused mognetite catalystalyst in slurry reactor under similar conditions ofsulfur loading (S/Fe) is 4.0mg;g -1, 263 C, 1.48 MPa,Huf43] at 225- 232 C were essentially the same. At0.95L.h-1.g-1, H2/CO is 0.69[36);248- 250C, the O/P ratios were slightly higher on. cobalt catalyst, 230°, 0.79MPa, 1.2L-h-1g-1,H2/CO i 1.55[17];the precipitated catalyst. In general, the presence) reduced fused magnetite catalyst, 232 C,f potassium, or an increase in potassium content1.48MPa, 0.475Lh-1.g-1, H2/CO is 0.69, T-O_S- :300 h[5;in an iron catalyst increases the O/P ratio. A higher+ precipitated iron catalyst(100Fe/3Cu/4K/16SiO2).H2/CO ratio tends to reduce the O/P ratio, depend-263 C, 1.48 MPa, 1.39L.h-i.g -1, H2/CO is 0.70[14]ing in part upon the degree of secondary reactions.Donnelly and Satterfieldl19] studied the productBased on classical Anderson-Schulz-Flory (ASF)distribution of the FTS on alkalised and unalkalizeddistribution, the maximum yield of transportation fu-precipitated iron catalysts in a CSTR. A commer-els Cs- -C18 hydrocarbons is about 66% (correspond-cial Ruhrchemie catalyst with 1.9%(w) potassium anding to chain growth probability factor of 0.816). This12.0%(w) silica showed stable activity for 1300 hours-yield was nearly achieved in the Rheinpreussen BCSRon- stream and the average molecular weight of prod-done by Kolbel et al11] and Kolbel and Ralek[31] inucts from this catalyst decreased with time on streamtheir research work on the demonstration plant perfor-but the oxygenated production increased substantiallyman中国煤化工rocarbons (gasolineat the same time. In contrast, a precipitated iron cat-naswas accormpaniedalyst with neither potassium nor silica showed stableby:MHCN M H G(3.2% oly. Thisselectivity but decreasing activity for 2000 hours-on-product distribution is very unusual and could not bestream. The carbon number distribution of bhydro-reproduced in any subsequent studies. In all othercarbon products from both alkalized and unalkalizedstudies low methane yields were always accompanied矛野数鹅"，2000Overview of Fischer-Tropsch Synthesis in Sturry Reactors261by high yields of C12+ products. An alternative wayplay armong many variables.t increase total transportation fuel yield is throughThe effects of temperature, H2/CO, space velocity,maximization of the reactor wax yield, while simulta~ concentration of H2O, catalysts reduction conditions,neously minimizing the light gas yield. Reactor wax liquid phase, poisons and potassium on the process ofthen can be upgraded via conventional processes intoFTS in slurry reactors are to be reviewed separatelyhigh quality diesel fuel.as follows.5.1 Temperature5 EFFECTS OF PROCESS PARAMETERSNormally, increasing reaction temperature accel-AND MATERIALSerates reaction rate according to the Arrhenius equa-There are many factors that can affect the result oftion, but the conversion of reactants depends on bothFTS process. Bukur and Brown间studied the ef- the infuence of temperature on reaction rate and onfects of process variables temperature, pressure, spaceequilibrium conversion.velocity and H2/CO feed ratio on the FTS reactionFig. 3 shows that the conversion of H2 and CO isusing a promoted fused iron catalyst in a CSTR over increased with temperature for both Fe and Co cata-the wide range of operating conditions. The tempera- lysts, but the slope for Fe catalyst is much higher thanture was up to 280 C, the pressure up to 2.86MPathat of Co catalyst.and the H2/CO ratio from 0.6 to 2.3. The resultshown that, at high ratio of H2/CO, very good agree-ment was obtained between their data and the dataof Huf's[43)] under the similar conditions. But withCO rich feed gas (H2/CO is 0.64- 0.72) the rate ofFTS reaction in the study was lower than Stenger andSatterfieldl4]. The water-gas shift reaction was nearequilibrium at higher conversions (Xco+H, > 50%),but significant departures from equilibrium were ob-210 230250 270290served when Xco+H2 less than 30%. WGS reactionalways proceeded at a slower rate than the FTS reac-Figure 3 The infuence of temperature on conversiontion in a CSTR.of FTS in slurry reactorsBukur et al.20] tested the efects of process con-●cobalt catalyat (Co2(CO)s/Zr(OPr)4/SiO2),3.5%cobalt, 6.6%zirconium,ditions similar to those used icommercial reactorsp=2.07 MPa, Hz/CO i 1.0, SV=1.0Lh-'g-14;on the catalyst activity, products distribution and se6 I precipitated Fe catalyst[promoted with 1%(w)K],lectivity. They wsed a commercial Ruhrchemie cat-p=2.0MPa, Fz/C0 is 0.58, u=0.0105 ms-alyst for FTS and a CSTR. The apparent reaction▲iron based catalyst, p=2.0 MPa, H2/CO is 0.5,rate constant was calculated by assurning the reac-tion rate has a first order dependence on hydrogen5.2 H2/COpressure. The catalyst was deactivated slowly withSatterfield et al7l thought that the syngas pro-time-on-stream while the selectivity shifted towardsduced from certain second generation coal gasiferslower molecular weight products. Secondary reactionsthat use minimum amounts of oxygen and steam with(1-alkene hydrogenation and isomerization) increasedthe lowest cost would have the H2/CO ratio possiblygradully with time-on-stream.as low as 0.5. But the conventional FTS or methanolYates and Satterfield17] used a cobalt FTS catasynthesis processes cannot accept such low H2/COlyst to study the bydrocarbon selectivity in a CSTR atratios. The H2/CO ratios used in industry such asthe conditions of 220- 240C, 0.5-1.5 MPa, H2/COSASOL are conventionally 1.8 to 6.0 for fixed- bed oris 1.5- 3.5. They found that increasing space veloc-entrained bed reactors. Kolbel and Ralek(31} reportedity (decreasing conversion) or decreasing H2/CO ratiothe satisfactory operation of a large slurry reactor u1s-resulted in the decreased yield of C1 products (unde-ing an iron catalyst with H2/CO ratio of 0.67. So,sired) and increased yield of C10+ products (desired),hey had studied the infuence of low ratio H2/CObut reactor temperature and pressure had lttle ef-(about中国煤化工:tor. They usedfect on the carbon number distribution of the prod-octaco: CSTR and theucts. The extent of incorporation of re-adsorbed ofcatalysYHcNMHGAowrateofgas1-alkenes, into gtowing chains on the catalyst surfacein liquid was 50-- 430m3.m~ -3.h-1. They concludedagainst hydrogenation to the corresponding alkane orthat tbe ability of a slurry type FTS reactor to pro-isomerization to 2-alkene reflected an intricate inter-cess satisfactorily feed syngas of low H2/CO ratio wasChineweJ. Ch. E.8 (3) 255 (2000)262ChineseJ. Ch. E. (Vol.8, No. 3)intimately associated with a high degree of mixing incatalyst in slurry reactor based on analytical calcula-the reactor. When the feed ratio exceeded the con-tion. He thought that water gas-shift reaction playssumed ratio of H2/CO, termed the usage ratio in suchan important role in the FTS reaction and derived aa reactor, the highest degree of conversion resulted inrate model of slurry reactor which accounted for thethe highest ratio of dissolved H2 to CO in the liquid.kinetics of both FTs and wGS reactionsThis should minimize the extent of carbon formationand catalyst disintegration caused by disproportion--. Rn+co=k(2 + 2/){(mH)nation of CO. Thus, high conversion was desired notonly for economic performance but also for helping to[(1 + m/2m)rPrs - rwcsldt}maintain catalytic activity.It is interesting in seeing from Fig. 4 that differentThe equation suggested that if the rwGs is large, thenresults were reported even though all using Fe cata-the rate - RH2/co becomes large, as shown in Fig. 6.lysts. Xu's curve has a minimum percentage conver-Increasing the concentration of water in the feed gassion at about H2/CO being 1.0136). Sanders's resultincreased CO conversion but decreased H2 conversion.shows that the global conversion of syngas (XH2/co )As a global efect, the conversion of syngas was in-was slightly increased from 0.31 to 0.34 with the ratiocreased with the concentration of water at low reac-of H2/COlI5), but Deckwer found that the tendencytion temperature and low H2/CO ratio (≤0.7).of change of conversion with the ratio of H2/CO wasalways decreasingl. Logically thinking, the ratio ofH2/CO should infuence product distributions. Per-haps more research on this important aspect should0Hbe done experimentally to verify this infuence and toprovide data for maximizing.10SV,L"h'g-'Figure5 The infuence of space velocity onconversion of syngas●promoted precipitated Fe catalyst, CSTR,260 C, 1.0MPa, H2/CO doesn't exced 0.8[1]■K promoted Fe catalyst, BCSR, 2609C,1▲promoted fused Fe catalyst, CSTR, 265 C,H:/CO1.28- -1.55 MPa, H2/CO is 0.64--0.72|@Figure 4 The infuence of H2/CO on conversion ofFTS in sury reactors●ultraine iron oxide catalyet, 2600, 0.7MPa,■,▲precipitated Fe catalyet, 260C,1.0MPa, u=0.0125 m-1[15]◆precipitated Fe catalyst, 260C, 1.0 MPa[1l争70-5.3 Space velocityThe conversion of H2/CO in syngas is decreased0一十一古一+with increasing space velocity normally accordingmole frnction of water, %to engineering knowledge. The results of three researchers confirmed the same tendency (shown inFigure 6 The infuence of concentration of H2O onyogu conversion in alurry reactorFig. 5) though the slopes of curves are diferent.(260C, p 2.026MPa, H2/COi 0.5, FR =0.152 m-12)5.4 Concentration of waterIt seems that the presence of water in the syngas中国煤化工always inhibits the reaction rate of FTS or the con-d a model to de-version of H2/CO, which is reflected in some kineticscribMYHCNM H Goolumn sury化models in Table 2.actors used for FTS or accounted for the dispersionPrakash's result was diferent[29), who studied theof gas and liquid and the axial distribution of cata-effects of water-gas-shift reaction rate in FTS over iron lyst pellets. Distribution of products, reactants and劈嘮数搪2000Overview of Fischer-Tropsch Synthesis in Slurry Reactors263H2/CO ratio was calculated by the model. It was in- catalyst became encapsulated and was preferentiallydicated that the calculation was in good agreement wetted by hydrocarbons in the presence of a perfuor-with the result of Prakash!2and the water-gas-shift polyether (Fomblin) in which hydrocarbon waxes werereaction played a significant role providing reasonable immiscible, the reaction occurred in wax capsules, inexplanation why slurry reactor can use lean-hydrogen which wetted catalyst particles were suspended wassyngas in FTS. Their conclusion was that the opti- highly limited by mass transfer resistance. Thus, inmum concentration of water in the feed gas increased accordance with the theory, secondary reactions werewith decreasing H2/CO ratio at a given reactor tem-greatly enbanced by mass transfer resistance.perature and decreased with increasing reactor tem- 5.7 Poisonsperature. Increasing the WGS reaction rate improvedStenger and Satterfield46] examined the infuencethe syngas conversion at low reaction temperature.of two poisons: H2S and dibenzothiophene (DBT) on，5.5 Reduction conditionsthe Fe catalyst used in FTS. When H2S was added intoSatterfield and Stengerl16] compared the perfor- synthesis gas with catalyst suspended in octacosanemance of a precipitated iron-copper-potassium cata- the catalyst activity was increased by 60% at a sulfurlyst and a reduced fused magnetite catalyst in the loading of 1.3mg:g-1 (S/Fe) beyond which catalystsame CSTR, whereas the precipitated iron-copper- activity decreased. But DBT presented in pbenan-potassium catalyst was also tested in a fixed-bed re threne during catalyst activation did not inhibit thisactor. The activities of the three cases were approx- effect. The two poisons reduced activity of catalystimately the same, but, if H2 was used to reduce the comparably at loading of about 10 mg:g-1 (S/Fe). Un-catalyst instead of the 1:1 H2/CO mixture, the activ- like H2S, DBT caused a marked drop in selectivityity of former was ten times of that of the latter both to form methane and increase in olefin/parafin ratiofor precipitated Fe catalyst and fused magnetite cat- of products, indicating that it decreased the hydro-alyst. .genation capability of the catalyst. They noticed thatAlso Withers[2] observed that when H2 was used olefin isomerization is enhanced by both H2S and DBTas activation gas the activity was substantially larger but the efect of DBT is much more marked. DBT, un-than when H2/CO mixture was applied as activation like H2S, markedly decreased methane formation andgas for Co/Zr/Al2O3 catalyst in the slurry reactor.increased the olefin/ parffin ratio of the products.5.6 Liquid phase5.8 PotassiumVan der Laan et al.45] considered that liquid phaseDonnelly and Satterfeldl19] considered that potas-might infuence heterogeneous reaction. The presence sium was not the cause of the double-a (chain growthof the slurry liquid appeared to affect the product se probabilitie a1 and a2) Schulz-Flory distribution inlectivity of the gas -solid system. The slurry phase their studies of products distribution of the FTS on al-systerm yielded a higher olefin fraction at comparablekalized and unalkalized precipitated iron catalysts in areaction conditions. The corresponding model param- CSTR. This conclusion was in contradiction with theeters, the readsorption constant and the termination suggestion of Schliebs and Caube(47] that potassium-constant to olefins are lower under the similar condi- promoted and potassium-free sites accounted for thetions.two chain growth probabilities observed on catalysts.Stenger and Satterfieldl5l used a reduced fusedPotassium inhibited both hydrogenation and isomer-magnetite catalyst to study the effect of liquid phase ization on the catalysts and signifcantly increased theon the products selectivity of slurry FTS reaction. yield of heavy products by increasing both a1 and a2.They focused mainly on secondary reactions such asVariations in operating parameters were shown to af-olefn hydrogenation, olefin isomerization, incorpora fect both carbon number distributions and chemicaltion of ethylene and/or ethyl alcohol. These reactions selectivities, specially to 1-alkenes. Low temperatureoccupied slightly greater extent in octacosane than inoperation increased a1 and tbus the yield of heavyphenanthrene at 2329C but esentially the same at products. On the other hand, high temperature oper-263 °C. They thought that this was attributed to com- ation increased the 1-alkene to n-alkene ratio of prod-petitive adsorption effects with phenanthrene, which ucts from the alkalized catalysts.adsorbed significantly onto the catalyst at 232 C, butAm中国煤化工of temperaturenot at 2639C. They found that all secondary reac- and spa; clear compar-tions were enhanced by the conditions minimizing CO atively.YHCNMHGH2/COonFTSadsorption, oxygenation was greatly diminished atreaction rate and product distributions, the mecha-high conversions and C2 incorporation into productnism of H2O in acting the FTS reaction, the optimumstopped chain growth by a scavenging efect. When conditions for catalyst reduction, the efects of the liq-Chinese J. Ch. E.8 (&) 285 (2000)264Chinese J. Ch. E. (Vol. 8, No.3)uid medium on the FTS process and the function of 60/100 mesh) showed strong mass transfer limitationpotassium are still not clear and deserve further study.and the activation energy was roughly half of the in-trinsic one.6 HEAT AND MASS TRANSFER AND SOL-In the slurry processes of converting syngas intoUBILITY OF GASthe mixture of alcohol and hydrocarbons, the selec-The problem of heat removal should be paid much tion of the inert solvent with good properties suchattention to choose a reactor for FTS process since as high boiling point, thermal stability, good mobil-the FTS reaction is very high exothermic, AH =ity and particularly the suitable solubilities of gases- 165kJ.mol-', OTsdisbatice ≈1600C. But heat re is critical. But research work on this subject is verymoval or temperature control is a problem for three scarce in the open literature, especially under synthephase reactor, epecially for slurry reactor, since the sis conditions of high temperature and pressure.turbulence in the liquid phase results in effective heatBeenackers and Swaij48] presented a review ar-transfer.ticle about the mas transfer characteristics of gas-Three phase reactors can be categorized into three liquid slury reactors. They discussed mainly the in-types: multi- tubular trickle bed reactor, ebulliating fuence of the presence of solid particles in slurry reac~bed (three phase fuidized bed) reactor and slurry (no-tors on the mass transfer parameters, gas-liquid masstably the slurry bubble column) reactor. Because of transfer cofficients (or,a, kca), liquid-side mass trans-the small size of the catalyst particles (typically less fer cofficients (kL, ks) and specific ga-slurry contactthan 0.2 mm diarmeter) used in slurry reactor intra-area (a) and at the mas transfer enhancement causedparticle difusion is not limiting. And also because by physical adsorption on small particles, fast homo-very fine catalyst is suspended in a liquid medium, it geneous reactions in the slurry due to inert particles,allows catalyst to be replaced easily during operation.homogeneous reaction in the liquid with dissolvingThe multi-tubular trickle bed reactor has another sig particles, reactive particles and catalytic particles innifcant advantage that its maximum scale-up factorheterogeneous reactive systems.is above 10000 and the maximum factor of safe up-Breman et al49 studied the solubilities of syngasscaling for slurry reactor is less than 500. The ebul- in the solvents systematically. In their measurements,lating bed reactor occupies an intermediate positionthey used fourteen solutes which are all reactants orbetween trickle bed reactor and slurry reactor.products relevant to synthesis gas conversion via gas-Mass transfer has certain efect on the process of slurry process and five solvents which are being used orFTS reaction. Deckwer et al[6] declared that theyseemed to be used in hydrocarbon synthesis. Experi-analyzed 8 diferent conditions of catalysts or bubble mental conditions were varied from 293 to 553K andcolumns in the FTS studies using new and reliable from 0.06 to 5.5 MPa, covering typical process con-results on hydrodynamic properties. They used theditions used in synthesis gas conversion. They gotreactor model in the literature and experimental data1533 data of gas-liquid solubilities divided into overto calculate mass transfer resistance. Their conclusion60 binary systerms and compared these with few lit-was that tbe FTS slurry process in bubble columnserature data, as gas-liquid solubilities in this systemwas governed mainly by reaction rate, ie. catalyst ac were hardly satisfactory reported previously The etivity, catalyst concentration and temperature. Masssults of experiments and comparison were good withtransfer rsistance was found to be small compared to an average deviation of 7.6% and a maximal deviationreaction resistance and were negligible for the knownof 15.0%. They got the conclusion from their workcatalysts under normal operating conditions. The ab-that for both water and alcohol, the strongly polarsence of mass transfer limitation was mainly aresult of tetraethylene alycol is the best solvent, and the apo-the high gas holdups atainable with molten paraffn lar solvent hexadecane and octacosane are the poorestas liquid phase.solvents. In contrast, tetraethylene glycol is the poor-Zimmerman et aL[10] had studied the efect of par-est solvent for hydrogen and the poorest but one forticle size on the activity of a fused iron catalyst usedcarbon monoxide, whereas hexane and octacosane arefor FTS. The main result of their work was that therelatively good for these solutes.significant resistance was found to be intra-particlestudied the massmass transfer, which was caused by the low difusiv-中国煤化Iyphase FT sytheity of reactants in wax flled catalyst pores. When sis. |C N M H Gtance of mass trans-the particle size of catalyst was 0.078 mm (170/230fer trom H2 and CU were attected by fAuid properties,mesh), the catalyst activity approached its intrinsicreactor types and operation conditions. FTS reactionvalue. Particles of 0.48 mm and 0.21 mm (30/60 andoccurs in the kinetic region when the solid content was伊数辖r，2000Overview of Fischer-Tropsch Synthesis in Slurry Reactors265from 5% to 12%.p8ystem pressure, MPapartial pressure of component i, MPa7 RESEARCH SUGGESTIONS- RH2+Co reaction rate of FTS, molh-'gStudy of Fischer-Tropsch synthesis has been under-Rgtotal reaction rate, molh-'ggone for a long time and remains an attractive projectreaction rate of FTS, mol.hrate of reaction kfor the importance to industrial production. SlurryrWGsrate of water gas shit reaction, mol:h-'g-'reactors used for FTS have been investigated in some3Vspace velocity, Lrh-'g1concerned aspects.temperature, CKinetics models for FTS in slurry reactors areT-0-Stime on-stream, hmainly Langmuir type. Thus, it seems that a com-flow rate, m:-1mon equation including all adsorbed components canV(STP)flow rate in standard temperature and pressure,be used conveniently and be simplified in many situ-L.min~ations. The inhibition by water due to WGS reactionXH2+coconversion of Hz+CO, %should be dealt carefully.cofficient of the rate equation of FTSActivity and slectivity are very important crite-thk,sstoichiometric coeficient of specie j in reactionria in FTS production and affected by many factors.Normally, conversion of H2 and CO about 40%- -80%Subscriptscan be achieved. Most catalysts have high selectivityinlet of the reactorfor the desired products，but the ratios of olefin toparaffin are not always high. There are some furtherREFERENCESresearches, such as adding best promoters in Fe cata1 Deckwer, W. D. Kokuun, R., Sanders, E, Ledakowicz, s.,lyst, finding optimum process parameters and condi-Ind. Eng. Chem. Process Des. Dev., 25, 643- 649 (1986).tions, need to be done get higher selectivity of olefin2 Withers, H. P, Jr, Eliexer, K. F., Mitchell, J. w., Ind.and hydrocarbon with middle carbon numbers.Eng. Chem. Res., 29, 1807- 1814 (1990)3 Zimmerman, W. H, Bukur, D. B., Ledakowicz, S., Chem.The characteristics and functions of liquid or sol-Eng. Sci, 47 (9- -11), 2707- 2712 (1992)vents in the three-phase system are not fully under-Huf, G. A., Jr, Satterfield, C. N.. Ind.ng. Chen. Pro-stood. These properties are related to diffusion andreaction rate. Further studies of solubility of reac-5 Stenger, H. G., Jr, Sterfield, C. N, Ind. Eng. Chem.tants and products, the mass transfer coficients be-6 Deckwer, W. D., Serpemen, Y, Ralek, M., Schmidt, B.,tween gas-liquid and liquid-solid, the adsorption ofChem. Eng. Sci, 36, 765- 771 (1981).gas and liquid on the solid surface, the beat transfer7 Satterfeld, C. N. Huf, G. A., Jr., Can J. Chem. Eng., 60159- 162 (1982).in gas-liquid-solid system, the separation of diferent8 Rickena, M. L., Vickers, A. G,, Haun, E. C.. Koltz, R. C.phases and products and the special effect of super-AIChEJ, 86- .90 (1982).critical fAuid on the process(51] are not only of theoret-9 Bukur, D. B, Brown, R. F. Can. J. Cbem. Eng., 65,ical value, but also provide the technical measures to10 Zimmerman, W. H, Rossin,J. A, Bukur, D. B., Ind. Eng.604- 612 (1987).improve the production of FTS in slurry reactors.Chem. Res, 28, 406- -413 (1989).11 Kolbel, H, Ackerman, P,lhardt, F. Proc. FourthWorld Petroleum Congress, Section IV/C, Carlo ColomboNOMENCLATUREPublishers, Rome (1955).rate constant of FTS reaction, molg-1.h-112 Satterfeld, C. N, Proc. 7th World F-T Meeting, Pitte-0,b' coficients of the rate equation of FTSburgh (1987).concentration of component, molL- 113 Kuo, J. C. W, Final report prepared for DOE ContractNo. DE-AC22-83PC600019, Mobil Research and Develop-ment Corp,, Pauisboro, NJ (1985).△H heat of the reaction, kJ-mol-14 Bukur, D. B., Nowicki, L, Lang, X. S., Report at the IS-absorption constantCRE 13 Meeting in Baltimore, Maryland (1994).k,ko reaction rate constants15 Sanders, E, Ledakowicz, S, Deckwer, w. D., Can. J.Chem. Eng, 64, 133- -140 (1986)gas phase mass transfer coeficient, Ls- 116 Satterield, C. N., Steger, H.G.Ind. Eng. Chem. Procesreaction rate constantkIliquid-side mas transfer coeficient, m:s~117 Ya中国煤化工Prels,e, 308- -314ksolid- aide mass transfer coeficient, m-s- 1(199maverage number of hydrogen atoms per hydroca-8 HueCNMH G, PrelSci. Tech.INTL, 11 (5&6), 639- 664 (1993).rbon molecule19 Donelly, T. J.. Satterfield, C. N., Appl. Catal, 52, 93-average carbon chain length of the hydrocarbon114 (1989).product0 Bukur, D. B., Patel, s. A., Lang, x. s, Appl. Catal, 61,ChineseJ. Ch. E 8 (3) 285 (2000)266Chinese J. Ch. E. (Vol. 8, No. 3)329- -349 (1990).B. H, Puel Sci. Tech. INT'I, 12 (10), 1323- 1353 (1994).21 Anderson, R. B., Catalysis, Vol.4, Emmett, P. H., ed, Rein-37 Guan, N. J., Zheng, H. S. Li, M., J. Fuel Chem. Tech.. 26hold, New York (1958).(4), 377-380 (1998). (in Chinese)l2 Satterfield, C. N., Stenger, H. G, Jr., Ind. Eng. Chem.8 Herington, E. F. G., Chem. Ind, 65 346 (1946)、 (inProcess Des. Dev, 24, 407- -411 (1985).Chineae)23 Pichler, H., Schulz, H., Chem. Ind. Tech., 42, 1162 (1970).39 Weller, S, Frield, R. A, J. Chem. Phys., 17, 801 (1949). .24 Rofer-Depoorter, C. K., Chem. Rev., 81, 447 (1981)40 Anderson, R. B., Frield, R., Storch, H, J. Chem. Phys,5 Deckwer, W. D,D, Serpemen, Y., Ralek, M., Schmidt, B., Ind.19, 313 (1951).Eng. Chem. Process Des. Dev, 21, 222- 231 (1982).41 Satterfield, C. N, Huf, G. A, Jr., J. Catal, 73, 187- -19726 Bukur, D. B. Chem. Eng. Sei, 38, 441- 446 (1983).27 Yates, I. C, Satterfield, C. N,' Energy Fuels, 5, 168- 17342 Anderson, R. B, The Fischer Tropech Synthesis, Academic(1991).Press, INC., Orlando, Florida (1984).28 Ledakowicz, S., Nettelhoff, H., Kokuun, R., Deckwer, W.43 Huf, G. A., Jr, Ph. D. Thesis, MIT, Massachusetts (1982).D., Ind. Eng. Chem. Process Des. Dev., 24, 1043--104944 Wang, Y. N, Li, Y. W., Zhao, Y. L, J. Fuel Chem. &(1985).Tech, 27 (3), 193- 292 (1999). (in Chinese)29 Prakash, A., Chem. Eng. Comm, 128, 143- 158 (1994).45 Van der Laan, G. P, Beenackers, A. A. C. M., Ding, B. Q,30 Hou, W. J.. Zhao, Y. L., Zhang, B. J, J. Fue! Chem. 8Strikwerda, J, C., Cataly. Today, 48, 93 -100 (1999).ech., 25 (1), 42- 48 (1997). (in Chinese)46 Stenger, H. G., Jr., Satterfield, C. N., Ind. Eng. Chem.31 Kolbel, H, Ralck, M, Catal. Rev. Sei. Eng, 21 (2),Process Des Dev, 24, 415- 420 (1985).225- -274 (1980).47 Schliebs, B., Caube, J., Bunsenges, B., Phys. Chem., 89,32 Mitra, A. K, Roy, A. N, Indian Chem. Eng, (7), 127--142668(985).18 Beenackers, A. A. C. M., Swaaji, w. P. M. V, Chem. Eng.33 Farley,R, Ray, D. J. Inst. Petr, 50, 27- -46 (1964).Scit, 48 (18), 3109 -3139 (1993).34 Kuo, J. C. w., Final report prepared for DOE Contract49 Breman,B. B., Beenackers, A.A. 0. M., Rietjens, E. W.J,No. DE AC22-80PC30022, Mobil Research and Develop-Stege, R.J. H,J, Chem. Eng. Data, 39, 647- -666 (1994).ment Corp, Paulsboro, NJ (1983).50 Lu, Y. J, Zhou, J. L, Tianranqi Huagong, 23 (4), 28- -3235 Chanenchuk, C. A, Yatea, I. C., Satterfield, C. N, Energy(1998). (in Chinse)Puels, 5, 847- -855 (1991).51 Yan, s. R, Fan, L., Zhang, z. X., J. Fue1 Chem. & Tech.,36 Xu, L. G, Bao, S. Q, Brien, R. J. O., Houpt, D. J, Davis,26 (6), 510- 515 (1998). (in Chinese)中国煤化工MHCNMHG劈唠数据2000