Influence of the Feed Gas Composition on the Fischer-Tropsch Synthesis in Commercial Operations

Availableonlineatwww.sciencedirect.comScienceDirect圖Journal of Natural Gas Chemistry 16(2007)329-341SCIENCE PRESSReviewInfluence of the Feed Gas Composition on the Fischer-TropschSynthesis in Commercial OperationsYijun lu’, Theo LeeR&D Center, Headwaters Technologies Innovation Group, 1501 New York Avenue, Trenton, New Jersey, USAI Manuscript received July 9, 2007; revised September 13, 2007 1Abstract: Key technical challenges relating to theYijun Lu was born in 1972. HeFischer-Tropsch(F-T)synthesis applied in the commer-received his chemical engineering/gas-to-liquids(CTL/GTL)teademic of Sciences(CAS) ingies have been reviewed. Based on the experiences ac-1999. He was assistant and ascumulated from pilot plant, semi-work test and lab re-sociate professor of the Institutesearches, the influences of the H2/Co ratio and the cOof Coal Chemistry, CAS in 2000in the feed gas on the F-T process as well as on CTL/GTLand from 2001 moved to the hydrorbon Technologies Incorporation,complex in terms of product yields, energy efficiency andUSA, to become a research engi-carbon utilization efficiency have been studied. Beingneer and project designer of the In-contrary to the current design schemes for F-T processusing the coal derived syngas and the iron-based cata-esearch activity was focused on thelyst, it is suggested to feed the F-T synthesis unit with a Fischer- Tropsch catalysis, the core process engineering ofsyngas having a H2/CO ratio of 0.5 and then adjusting gas-liquid-solid reaction, petroleum product characterizato 1. 4 via the recycling process. As a result, the carbontion and upgrading processes. He has invented the novelslurry F-T synthesis catalyst, entitle as skeletal iron basedefficiency of the whole plant could be reached to as highcatalyst. His work was witnessed by 8 US and Chinese50%. For the issue of CO2 addition to the feed gaspatents and 20 more publicationsit is proved that only a diluting role is played under theDr. Theo L.k. Lee is vice presicurrent commercial slurry phase F-T processdent and Chief Technology oficerKey words: Fischer-Tropsch synthesis; gas-to-liquidfeed gas composition; carbon efficienbility Company. He is responsiblefor the development and promoting of Headwaters'coal-to-diquids1. Introduction(CTL) technology. He leads a CTLproject development team and a de-sign/engineering group in evaluatOil supply security and price concerns have leding CTL project opportunity in theto a renewed interest in coal and natural gas as alUnited States and asian countriesHe has his bachelor of chemicalternative feedstocks for the production of transportfuels and chemicals. Synfuels derived from naturalUniversity (Taiwan) and his doctoral degree also in chemical engineering degree from Uni-gas, coal and biomass can also improve air quallty. versity of New Brunswick. He has conducted research andto-liquids(GTL), the world has stepped into a huge bon zi opment in heavy oil and adsorption of light hydrocar-By using natural gas conversion technologies, gas-中国煤化工mre8ddm水and promising synfuel production commercializationAnother well-known energy conversion route, coal-to- facticCNMHGeer direct kcna liquequids, describes both coal gasification and Fischer- He led the designing work of a 1000 t/y capacity of gtlTropsch(F-T) synthesis to produce liquid fuels to direct/indirect coal liquefaction techiCorrespondingauthorTel:001-609-807-9308;Fax:001-609-807-9320or9334;E-mail:yijunlv@yahoo.comYijun Lu et al. Journal of Natural Gas Chemistry Vol. 16 No. 4 2007(indirect CTL). And there are also the less developed, owned 2000 more patentsdirect coal liquefaction technologies. As has been pub-In the race of alternative energy commercializa-lished, the world s vast coal resources could become an tion, apparently the countries that have owned hugeimportant alternative to crude oil via the CTL routes. coal or natural gas resources are in the technologyThe F-T synthesis, first developed in Germany dur- frontiers. According to the published data, it can being the early decades of the 20th century, has been listed briefly here for the current status of the F-Tfurther developed and improved in South Africa by technology development as well as for the ICTL/GTLSASOL. Since the 1950,'s, 40 million tons of coal per commercial projects world widerear have been converted into 160,000 barrels per day . Chinaor crudeoil equivalent [1]Starting from the biggest investor, China(theFor any conversion routes, including the indirect ShenHua Group) is constructing a 60,000 bpd(di-coal-to-liquid(ICTL)and GTL technologies, the first rect)CTl plant and, with plans for further projects,step is to generate cleaned synthesis gas and followed aims to produce one million barrels per day by 2020by F-T synthesis, in which the two main characteris- China also has signed MOU with SASOL for a simi-tics are the unavoidable production of a wide range lar scaled ICTL plant in northwest China. SASOL isof hydrocarbon products and the liberation of a large considering the building of two CTL plants in Chinaamount of heat from the highly exothermic synthesis at a total cost of S14 billion. As a domestic repre-reactions. Product selectivities are influenced by tem- sentative technology vendor, the Shanxi Institute ofperature, feed gas composition(H2/Co), pressure, Coal Chemistry, Chinese Academy of Sciences(ICCand catalyst type 2CAS) dominates most of the F-T process researchersOver the years, four types of F-T reactors have and pilot plant experiences, covering from the fixedbeen designed and used commercially. The fixed bed bed reactor technology to the slurry ones [101tubular reactor, known as the ARGe reactor, oper-. USAates at 220-260C and 2-3 MPa. The high tempera-In the US, new incentives have been introducedture circulating fluidized bed reactors, known as Syn- for coal-based transport fuels, and coal companiesthol reactors, have been developed for gasoline andlight olefin production. Decades ago, the SASOL Ad-re now assessing the commercial viability of newvanced Synthol reactor has been developed. It is a projects. Medicine Bow Fuel Power LLC has en-tered into a long-term contract to sell 100% of itsfixed fluidized bed reactor with similar operating con- ultra-low-sulfur diesel fuel from its planned ICTL fa-ditions as the Synthol reactor, but at half the capital cility in Wyoming to the Sinclair Oil Corporationcost and size for the same capacity. The fourth reac-tor design is the low temperature slurry reactor whichThe planned project will use a few F-T technologyis a three-phase reactoring of a solid catalystvendor options. One is the Rentech, Inc. who hasentered into a joint development agreement for theliquid. Syngas is bubbled through the liquid phase for development of a F-T ICTL fuels plant to be locatedachieving excellent contact with the catalyst, whilein the Mingo County [7-9keeping the catalyst particles dispersed. Slurry reac-The coal producer Peabody Energy has pledgetors are widely accepted in the rising world business $10 million in cash to the F-T technology developerof F-T technology commercialization [ 3, 4Rentech toward a future ICTL plant in Illinois in2007. When the construction is completed in 20102. Progresses in F-T technology commercial. Rentech expects to produce 1, 250 b/d of high-qualityizationdiesel and other middle distillates. ConocoPhilliplans to build a $75 million test plant at Ponca City,Today, with the reappearing of high oil prices, the Okla. ExxonMobil ran a pilot GTL plant in Batonwhole world steps into an era of alternative energy in- Rouge, La, in the early 1990s, and has proposed thdustry. As a core technology of the GTL and ICTL buildinant to make as much as 100,000 b/d ofprocesses, the F-T synthesis technology has never syncr中国煤化ny, howeverbeen so intensively researched by scientists and pro- droppCNMHGcarry out its GTLmoted to commercialization by engineers 5-9. As proJect Ina reference, the core technology vendors of the F-T Middle Eastprocess are grouped into less than two dozens but haveSASOL and Chevron Texaco signed a global GTLJournal of Natural Gas Chemistry Vol. 16 No. 4 2007joint venture in 1999. As a result, Qatar Oryx gave pilot plant in Chicago. Besides, HTI continues to con-birth to the worlds largest commercial GTL plant duct credible R&d to fulfill the commercial technolwith a capacity of 34,000 bpd. A failure in the ogy requirements as well as to establish a greater un-refinery's steam superheater has been the most sig- derstanding of the technological issues (91nificant problem in the start-up phase(June, 2006to date. The plant uses SASOL's proprietary Slurry 3. Key technical issues of the F-T process dur-Phase Distillate process[12ing commercializationShell operated a gTl plant in Bintulu, Malaysia,from 1993 to 1997, when a Christmas Day explosionclosed it down. After its re-opening in 2000, the carAs the world is stepping into a high speed in thepacity was reduced from 50,000 bbl to 35,000 bbl. GTL/ICTL business, it is more practical for engineersThe second major milestone was the launching ofand scientists to focus on the very key technical is-the world-scale integrated Pearl GTL project along sues that are affecting the success and economy ofwith Qatar Petroleum in July 2006. The Pearl GTL a GTL/ICTL plant. In general, the commercializa-project includes the development and processing of tion of the F-T technology suffers from four majoroffshore natural gas resources to clean liquid hydro- Imitations:(1)limited selectivity for the main prod-carbon fuels through the construction of the worlds ucts;(2)catalyst deactivation;(3)high capital costlargest integrated GTL complex[13]. Pearl GTL will(4)less-than-optimum thermal and carbon efficiencyThe forbreak many records, such as, it will have the world'snly attributed to catalystlargest capacity to produce premium quality base oils, development in the R&D work, while the latter twothe largest hydrocracking capacity in one location, the are obviously related to the design and operationallargest ASU in terms of high purity oxygen and the strategies of the scaling-up of the F-T technologylargest system for full recovery of industrial processThe technical problems at Oryx are a real eye-water, thus achieving zero-liquid dischargeopener for other GTL developers, demonstrating thateven the best-laid plans are subject to change in thisemerging industry. Meanwhile, SASOL has investedSome GTL plans, however, have not panned out. millions of dollars in the construction of an innova-Syntroleum and Sustec have entered into an tive F-T reactor design at its R&D facilities in Sasol-greement to develop jointly a 20,000 bpd F-T plant burg to support the engineering design of the nextand is called the Spreetel project [14]generation F-T reactors for SASOLs GTL and ICTLBritish Petroleum(BP) has taken the wraps off technologiesdramatic cost-saving technologies that will be testedIt has been reported that Statoil's proprietaryat its 300 bpd demonstration GTL plant on the Kenai GTl technology was verified in a semi-commercialPeninsula [15demonstration plant at Mossel Bay, South Africa (18ENI(Italy), having a partnership with The reactor was erected in 2003. The reactor diamIFP/Axens, patented a proprietary GTL F-Ttechnol- eter was 2.7 meters and had an overall height of 40ogy based on a highly efficient cobalt catalyst within meters. The demonstration of Statoil,s F-T technol-a slurry reactor. Currently they possess a pilot plant ogy was combined with commercial production of upin Sannazzaro, Italy. The plant capacity is 20 bpd of to 1000 bpd productsparaffins [16On the other hand although the fluid dynam-The Japan National Oil Corporation (JNOC)an- ics of slurry bubble columns has been the subjectnounced in 2002 that it ventures with five private of numerous academic studies, the behavior of com-Japanese companies to produce successfully the coun- mercial reactors with diameters around 5 to 10 me-try's first manufactured GTL products at their pilot ters and using F-T processing conditions is not wellplant(7 bpd)in Yufutsu, Hokkaido, Japan [17]known. The main designing philosophy of a properHydrocarbon Technologies Incorporated(HTD) GTL demonstration plant was to use the same tech-as developed a gTL process that can be integrated中国煤化工 al commercial designinto the existing fertilizer plants to co-produce ultra andCNMHGclean transportation fuels and ammonia. Under the thelnat the essentialfunding of doe and accompanying with WMPI and ponents were tested and ready for commercial appliSyntroleum, HTI joined the primary design of 20 bpd cation. The reactor requires a feed gas with an optiYijun Lu et al. Journal of Natural Gas Chemistry Vol. 16 No, 4 2007mal H2/Co ratio for its best performance, and much cooling tubes to remove most of the reaction heateffort is being put into the design of an efficient cat- in the reactor, and was installed as a concen-alyst production plant, mainly to achieve continuous tric reducer. Typical operation conditions wereproduction rather than a batch operation for individ- 2.0-3.0 MPa, 220-270C, catalyst loading 1 kg, andual processing stepssuperficial gas velocity 10-20 cm/s. The cooling sys-For all engineering efforts involved in the com- tem was also employed as the heater at the start-upmercialization of GTL/ICTL plants, the concerns stage and kept at 190-210C to remove exothermicare mostly focused on reactor hydrodynamics and heat. The tail gas after the separation vessels wasscaling-up strategies as well as a successful package analyzed by on-line GCsthat includes reliable catalysts and operation proce-dures 19. Typical design and operating conditions 3. 1. Influence of H2/Co ratio in the feed gasof the F-T slurry bubble column for an optimally de-signed reactor can be obtained from the informationThe issue being firstly addressed here relates togiven by Krishna②20the syngas composition just after gas generation, pu-In the present paper, the core technology in the rification and prior to the principal F-T synthesis re-ntegrated plant-the F-T synthesis unit-is singled actor. The conclusions in terms of catalytic perforout and diagnosed in terms of factors affecting the en- mance as a result of H2/ CO ratio changing are brieflyergy efficiency and carbon efficiency of a GTL/ICTL summarized here. In general, for the iron-based cat-alyst, one can obtain the following trends with theTo support the study in the present discussion, increase of the H2/CO ratio in the syngas[21-28F-T synthesis unit in the HTI R&D center. The unit whereas syngas ratio is decreased arently enhancedsets of experiments were carried out in a slurry phaseThe CO conversion iswas a bubble column reactor with complex systemsb)Heat transferring/removal due to the exotherof feed gas supply, cooling and products separatiomic reaction is relatively improvedThe catalyst was a hTi proprietary iron-based for-c)Catalyst carburization is minimized, hence itsmula composed of iron and promoters. All test con- stability is extendedditions were designed simulating the commercial unitd) The ratio of paraffin to olefin in the hydrocar-plus recycling operationbon products is increased slightlyThe slurry F-T synthesis reactor had an ID ofe)Reaction volume contraction is decreased and2.5 cm and a height of 800 cm (Figure 1). It is helpful to the slurry reactor operationwas simply made of a SS316 tube with an extendedf)Methane selectivity is increased whereas the de-reducing tube on the top To prevent plug flowsired product, C:the reactor during the whole test, a sinter plate withFor the cobalt-based catalyst similar trends couldmicron pores was used as the gas distributor andbe observed. Typical suggestion from the Shell In-ternational is to choose the circumstances in such acontacted with a synthgas having a substantially lower H2/CO ratio thanhe feed synthesis gas. As a result, the selectivity tolonger hydrocarbon chains is improved 28Obviously, these experiences relate only to theF-T reaction performance regardless of the effect onthe gross energy efficiency and carbon efficiency ofthe integrated process. As a reference, typical valuesof the principal characteristics and the H2/co ratiofor different processes are listed in Table 1 (29 . It中国煤化工io,from1 to nearlyesis processes, 2.e.Figure 1. Process flow diagram of HTI F-T synthe-CNMHGsis set-upl-Gas supply, 2-Preheater, 3--Slurry bubble column reactor generation and the synthesis process, so as to ensure(SBCR), cOoling system, 5--Product separationthe optimum overall conversion rateJournal of Natural Gas Chemistry Vol. 16 No. 4 2007Table 1. Synthesis gas specificationsreaction. Thus, this will re-balance the H to COSynthesisH2/coratio to another optimum macro stoichiometric valueDowUnion CarbideCO+H2O←→H2+CO2ARGE13-3.0Gulf-BadgerEven back to the last century, it has been widely1.5-2.0accepted that the iron based catalyst can cause theF-T reaction to combine with the wGs reaction into2.6an overall reaction as follows26, 32, 33Owing to the high cost of syngas, it is important2CO+H2=>(-CH2-)+COthat the maximum amount is converted in the F-treactors. The H2/CO ratio in the exit of the F-T syn-This point has been widely accepted all over thethesis can be lower or higher than the initial H2/C(world by most of the researchers in universities andratio at the inlet of the reactor, depending on whether companies. Thus, almost all reported F-T synthe-sis test conditions preferred a H2/co ratio of 0.7 inthe initial H2/CO ratio is lower or higher than the us- the syngas, which was just the one derived from coalage ratio. This requires the composition, particularlythe H2/co ratio, of the syngas matches the overallsource as the test material 34. As has been reported,usage ratio of the reactions.the direct processing of this gas in the F-T synthesisBasically, the syngas generated from natural gaseliminates the need for an additional WGS step be-fore the F-t unit for increasing the H2 /co ratio. Thefollowed by multi-stage purification, has the optimum inherent WGS activity possessing by the F-Tiron cat-composition of hydrogen and carbon monoxide, wita ratio of 2.0 or so. TheF-T reaction thusalysts allows the direct processing of low H2/CO ratiodescribed as following the macro equation accordingsynthesis gas in a slurry reactor without an excessiveto the stoichiometric chemistrycoking of the catalyst.It is noticeable toSASOL experiences ob-2H2+OO=(-CH2-)+H2O(1) tained from the ICTLWith the hydrodynam-ics and catalyst selectivities changing, there are twoshia he syngas from coal gasification, followed by a typical H2/Co ratios for two respective commercialreaction and cleaning, has a broad H2/Co ra- cases. For the LTFT process the H2/Co usage ratio range from 0.5 to 2.0, depending on the catalyst tio is typically about 1.7. In the high temperaturesystem of the F-T synthesisF-T (HTFT) the WGS is rapid and goes to equilib-For the cobalt based catalyst, CO2 does not rium, and it is believed that CO2 is also convertedparticipate in the chemical reactions of the low- into F-T products via the reverse WGS. Thus, if thetemperature F-t(LTFT)processes. The influence syngas has a ratio of H2/2C0+3CO 2) equal to aboutof the syngas composition on the initial behavior of 1.05, all of the H2, CO and CO2 can in principle bea Co catalyst in the F-T reaction has been studied converted to F-T products. In practical operationsin a continuous and perfectly mixed slurry reactor for of tubular fixed bed reactors, a value of 1. 8 is cho-an inlet H2/CO ratio between 1.6 and 3.35, keeping sen as the syngas H2/CO ratio. In the case of theother conditions constant. It was observed that de- HTFT synthesis producing gasoline and light olefins,activation increases with the increase of the H2/ co the combined gas has a H2/(2C0+3CO2)ratio closeratio[30. The optimum synthesis gas for the LTFt to the desired value of 1.05 and is suitable as the feedprocess would have H2/CO ratios slightly higher than for the HTFT reactors 12 and a low content of CHa and cO2. These syntheGenerally, for the F-T synthesis over iron-basedsis gas characteristics would allow the achievement of catalysts, the relationships between the consumptionthe highest overall carbon efficiency, combining high ratio of H2/Co with a fraction of the Co converted toconversion per pass in the F-T reactor and with good.- F-Trodd cn can he demonstrated linearlyselectivity towards the desired liquid products 31中国煤化工 nthesis reactor theFor the iron-based catalyst system, it is well emplCNMH Ge recognized as theknown that besides the main F-T reaction(1)and macro H2/C consumption ratio at commercial op-a number of side reactions producing oxygenates and eration conditions. The two main reaction fractionscarbides, there also exists the water-gas-shift (WGs) according to the Co conversion are listed as follows33Yijun Lu et al. Journal of Natural Gas Chemistry Vol. 16 No, 4 2007CO conversion fraction by combining two main reactions2H2+CO=(-CH2-)+H2O07(4)1.4H2+CO=→08(-CH2-)+0.6H2O+0.2CO2CO+H2O←→H2+CO20.3(5)To make the calculation more realistic by takingIt should be noticed that the ratio of H to CO inthe H2/CO stoichiometric ratio of F-T reaction to be2. 15, the overall H2/co consumption ratio will be inthe main F-t reaction is more reasonable to be 2.15creased to 1.52(top line in Figure 2), according to the stoichiometriThis result is in accordance with the observationeactionsthat the H2/Co ratio in the recyclereactor depends on both the H2/CO ratio in the freshF-T reaction, H/CO-2.0feed and the usage ratio of the catalyst. For negligible2.0F-T reaction, H/Co-2.change in the F-T selectivity during external recyclethe H2/co ratio in the fresh feed should match theusage ratio of the catalyst 37]Espinoza has illustrated the change in the exiting81.0H2/CO ratio of a F-T reactor with a cobalt-based catalyst for different conversions when the initial H2/Coratio is below, at, or above the usage ratio of 2.1538]. Boelee has reported similar trends for an iron-based catalyst, albeit the extent of the WGS reactionis higher for iron-based catalysts 39. However, suchFigure 2. The H2/CO consumption as a function ofplots for varying the recycle ratios have not yet beenreported in the open literature. Proofs supportingthe calculated value can be seen in the Table 2, whichTo address the above equations, for every unit were-obtained from internal exchange documents ofof CO converted on a typical iron based catalyst and F-T technology vendors 40)under commercial operation conditions, it can be sepApparently, the issue of deciding an optimumarated into two parts in the main F-T reaction, with H2/Co ratio for the commercial slurry F-T processa fraction of it around 0.7, and another WGS frac- can be clarified to some degree basing on the abovetion around 0.3. However, for special cases and at description. For different catalyst types such as fusedhigh CO conversion levels(about 90%)per pass, CO2 iron and Cu promoted or Mn promoted iron based cat-production accounts for about 42% of the total Co alysts, there will be unavoidably different H2/Co us-conversion [35]. The overall equation representing the ages and hence will require different feed gas H2/COF-T synthesis can be adjusted to as followsratios. This certainly will result in significant effects11H2+CO=0.7(-CH2-)+0.4H2O+0.3C02on the gross energy efficiency, carbon efficiency and(6) operation cost, etc. In other words, by feeding thesame H2/Co ratio in the feed gas to the reactor con-It indicates that the practical consumption ratio taining different iron catalysts, the consumption ratioof H, to CO is about 1.1, while the WGS fractionof the Co converted to CO2 is. 3. Raje and Davis of H2/CO could vary from 0.6 to 1.5. In conclusion,the varying of optimum H2/CO ratio is dependent onhave illustrated the benefits of low single-pass synthe- two factors. one is the catalyst properties which dom-sis gas conversion with recycle. The recycle reactorcan process more than double the volume of syntheinate the competitive activity of the F-T reaction ver-sis gas per weight of iron and produce twice as much sus the WGS reaction. The other one is the solubilityhydrocarbons as the single pass reactor 36of the feed gas components and the mass transferringproperties. The same H2/CO ratio in the feed gas hasIn practice, for a typical iron F-T catalyst pro- diffemoted by Cu, K or Si, the CO conversion can be conH中国煤化工: erring rates in thetrolled to around 50%60% at lower temperatures andtalyst poreshigh space velocities, which can then generate a lowerCNMHGof gas solubility offraction of CO converted to CO2, say 0.2, thus the syngas and product components in n-hexatriacontaneand the measurements were extended to gas solubilityconsumption of H2/Co equal to 1. 4 can be obtainein different waxes(see Table 3& 4)[41]Journal of Natural Gas Chemistry Vol. 16 No. 4 2007Table 2. Variation of H2/Co ratio in different cases of commercial projectsLaPorte II RecycleHTI RecycleSASOL RecycleICC CAS Recycle*SOff gasOff gasyngasTail Gas0.361.070.400.23399540.9548.590.1223.75570629.2548.1138.8018.1037,4212.5920.850.152.0020.22H2/co3.48*After recycling, the entrance feed gas composition has a H2/co ratio of 2.0Table 3. Solubility of F-T reactants and light products in n-Cs6 H74Componentsr, mole fraction0.04760.06140.1180.1150.2040.234Table 4. Solubility of H2 in various types of solventSolvent waxn-C20n-C28Mobil wax, n-C6r(H2), mole fraction0.03620.04020.04760.0661All the above solubilities were measured at 200C, moval are sent to the combined cycle. Hydrogen re-which is somehow lower than the practical operation moved from the F-T reactor tail gas is recycled to thetemperature(higher than 230C). In mole fractions F-T reactor to increase the H2/Co ratio in the syngasthe solubility of a gas increases slightly with the mole- produced from the carbon-rich coal. For the claimedcular weight of the wax under the same temperature process, the syngas H2/ Co ratio is maintained atand pressure. And the molecular weight of the wax about 1.0[43]has a relationship with the chain propagation proba-In conclusion, to match the consumption of thebility, a, which is again related to the catalyst prop- H2/Co ratio of the reactants with the feed gas com-ertiesposition, the best reference is the SASOL run dataSimilar work has been done by Song. It was found listed in Table 2. For customerized (iron) catalyst,that the solubility of syngas components increases one should also concern about the phenomena thatwith increasing pressure, but decreases with increas- the H2 mass transfer efficiency is 10%0-30% highering temperature, except for H2 which increases with than that of the Co. The representative catalyst ofthe temperature[42ICC CAS under commercial conditions would need aCombining these observations, it is reasonable to lower H2/CO ratio, reasonably about 1.6 in the feedthink that the H2 solubility in practical F-T wax un- gas after recycle. To guarantee this value, the H2/coder commercial conditions(250C)is close to CO, and ratio in the fresh syngas is suggested to be about 1.1could be higher than CO at elevated temperatures. Meanwhile, for the Laporte example, a high WGS acThe moderate idea for commercialization design tivity catalyst was applied, the H2/co in the syngasis to consider the optimum product selectivity as well was inevitably selected as 0.7as the gross energy efficiency and carbon efficiencyFor a cobalt-based catalyst at H2/C0=2, the COi.e. adjusting the H2/Co ratio to achieve higher de- is removed prior to the F-T process and the ideal carsired product yields and avoid poor H2/Co balance bon efficiency(based on the assumption that eitherin the tail gas(huge amounts of H2 in the tail gas is a H2 or CO is less than 100% conversion) is 50% basedbig drawback of the ICTL plant in view of the energy on 100 Cornal oasification(Equation 8 ineffiTabl中国煤化工To maximize the yield of the diesel fuel frac-CNMHGSS, the impact of alltion, Rentech has given a flowsheet starting from coal unit operations on the entire F'I synthesis process wasgasification. The permanent gases and light hydro- investigated by turning all their material, energy, andcarbon gases from the F-T unit after hydrogen re- work requirements into one variable, namely carbonYijun Lu et al. Journal of Natural Gas Chemistry Vol 16 No. 4 2007efficiency. Experimental data on a cobalt catalyst(Equation 10), the gross carbon loss basing on thewas used, the effect of the H2/ Co ratios as well as scheme without the WGS step after gasification willthe space velocities in the F-T synthesis loop on the lower the gross carbon loss and increase the productverall carbon efficiency were investigated 44yield according to the calculation(Table 5). However,For the iron-based catalyst, the case study startsthe commercial case that shifts the syngas H2/ co ra-from a condition that the syngas introduced to the tio to 1. 4 prior to the F-T unit and taking a WGSWGS step, while the unavoidable WGS reaction in carbon efficiency of 3 a t (Equation 12)results in aF-T synthesis has a H2/co ratio of 0.5 without the fraction of 0.2 in the unthe F-T synthesis unit will re-adjust the H2/Co ratioiously the suggested scheme without the wGSThus, each unit of Co from the gasification results in step prior to the F-T unit has a higher carbonthe same yield as the cobalt based catalyst case by efficiency and product yield than the one which istaking the WGS fraction(CO converted CO 2)as 0. 3 followed by a shift reaction. It also indicates that the(Equation 9). To extend the calculation range, tak- gross carbon efficiency is lowered with the increas-ng a wGs reaction fraction as 0.2 and the feed gas ing WGS activity( CO2 fraction)of the catalyst(ironprior to the F-T reactor having a H2/Co ratio of 1.4 based)Table 5. Comparison of productivity and carbon efficiency via differentf ICTL processGasification +WGS => Syngas Feedgas→ Products Carbon2CO+H2+H20 CO+2H2+COCO-+2H2(-CH2-)+H2OcO+H22CO+H21.43C0+157H2(CH2-)+0.57H2O+043CO2C0+H22CO+H21.25CO+1.75H2(CH2-)+0.75H2O(10)+0.25CO22C0+H21.25CO+175H20.875c0+1.75H20.61(-CH2)+0.4H2O30%+0.75H2O+075CO+0.3CO2Co+H1.25CO+175H20.875c0+1.75H20.7(-CH2-)+0.6H2O12+0.75H2O+0.75CO+0.2CO2CO2 removal3.2. Influence of CO2 in the feed gasof liquid hydrocarbons and electricity from coal usincombined cycle power generation facilities has beerTo many F-T synthesis researchers and proposed as a clean and efficient approach [45]GTL/ICTL plant engineers involving in the worldThe addition of CO2 to the feed gas is not an op-frontier technology, the highest fraction of the syngas tion for research any more, as more and more ICTLgeneration cost in the whole plant pushes them to plants or pilot projects prefer to believe the benefitsmake efforts in dealing with the non-reactant com- of CO2 addition. Meanwhile, there are numerous neg-ponents in the feed gas. Foally viable ative claims in the public information. To addressoperation of an iron-based F-T technology, there are this unclear point in the application of the F-T techtwo apparent options available.nology, catalyst selection has to be singled out first(1)Using a diluted feed such as nitrogen-rich syn- Iron based catalysts are extensively utilized for the F.thesis gas, thereby saving the synthesis gas costT synthesis through coal derived synthesis gas due to(2)Recycling the unconverted synthesis gas that its significant WGS reaction which allows the usingaves the reactor after condensation of the liquid of a feed gas with lower H2/Co ratio without an external shift step 3. For the sake of emphasizing CO2The most common F-T plant schemes involve re- additan between iron andcycling of the tail gas back to the synthesis gas prepa- cobal中国煤化了res is not describedration section with or without CO, removal. One fact hereCNMHGthat cannot be avoided is that every ton of carbon inIt is helpful to understand the overall influence ofthe mined coal will sooner or later end up as 3.67 tons CO2 addition by reviewing the R&D conclusions inof carbon dioxide in the atmosphere. Co-production terms of the apparent F-T synthesis performanceJournal of Natural Gas Chemistry Vol. 16 No. 4 2007337One recent study was undertaken to understandHereby it looks like that the CO2 has very negligithe importance of CO2 addition to the feed gas for ble effect on the wGS reaction as wellF-T synthesis over an iron-based catalyst [37 It was reaction. Nevertheless, whether the CO2 in the sys-revealed that the presence of CO2 in the recycled gas nthesis gas contributes to the hydrogenated productdid not change the overall CO2 selectivity; hence, CO2 or not is still unclearis regarded as an inert under these conditions. SASOLa special work, a microemulsion technology wasclaimed a process for an option by using a Co-based used in order to study the CO and CO2 hydrogenationF-T catalyst with the reaction temperature being in reactions on an improved iron catalyst. The catalyststhe range of 200"C to 280C. The CO2 was claimed with several promoters were tested in a fixed-bed mi-to behave as an inert gas in the reactor 46croreactor(Table 6)[48Dry has attempted recently to explain the effectof CO2 on the product selectivities via a mecha-Table 6. CO2 effect on wGS reaction on thenism analysis and simulation 1. It was thought thatcatalyst Zn-K-Cu/Fe at 235'C and 2.17 MPathe effect of the presence of CO 2 was oversimplifiedCO2/Co CO conv. CO2 selectivitySince the chemisorption of CO is much stronger than pressure, MPa pressure ratio (%that of H2, the presence of CO2 could have a greaternegative effect on H2 than on CO chemisorption22.6Li et al. have conducted the Co2 addition study0.152n a fixed bed reactor on iron catalysts. The work wasfocused on conducting the synthesis with a range ofCO2 pressures in the synthesis gas fed to the reactoraiming to define whether the presence of high partial0.8111.21pressures of CO2 will decrease the contribution of thewater gas shift to the synthesis. The results suggestthat the partial pressure of CO2 required to stop theAlso, the hydrogenation of Co, CO2 and theirwater gas shift reactions may be too high to be of mixtures has been comparatively studied with a Copractical use [47]. Thus CO2 is likely to be recycled and Fe catalyst at the University of Karlsruhe. F-Tto the synthesis gas generation rather than just to the CO2 hydrogenation with iron catalysts has been ad-F-T reactordressed in a number of investigations(Table 7)[49Table 7. Comparison of carbon oxides hydrogenation activity and selectivityCatalystCO hydrogenation, H2/CO=2CO2 hydrogenation, H2/CO2=3CO conv (% CO2 sel (%) CH4 sel (%CO2 conv (% CO sel (% CH4 sel (%)*95Fe21.38.15495Fe-K51814218928.395Fe-Na10.127.195Fe-Cu54.995Fe-CuK20.24 CO2 free selectivity, * Co free selectivityOther reaction conditions, 573 K, 1.01 MPa, GHSV=0.0023 L/(gs)It was claimed that at least the main route of larger reactor volumes and or higher required gas rehydrocarbon formation from CO2 includes a first(re- cycle rates. The consecutive reaction mechanism im-verse CO shift)conversion of CO2 to CO which then plies the difficulty that optimum conditions for eachreacts to form hydrocarbons [50]. For the F-T CO2 of the reactions are different. Reactor configurationshydrogenation the operating temperature must be there中国煤化工 the two reactionsrather high because of the equilibrium constraints for areCNMHG volumes, each un-the reverse CO shift reaction. This limits the apply- der optnnditionsng of temperatures in the lower-range for the FTDavis has carried out studies in a different wayconversion. Lower overall reaction rates will lead to 14CO2 was added to the syngas, and was observed toYijun Lu et al. Journal of Natural Gas Chemistry Vol, 16 No. 4 2007initiate chain growth to produce both oxygenates andhydrocarbons. This observaticF-T synthesis with iron catalysts containing promot sa so f layexamination of the production of alcohols during theers 35. Xu et al. had made similar efforts to assessthe contribution of the WGS to the F-T reaction 51As expected, if the WGS reaction was rapid comparedto the F-T synthesis, 14C that added to the CO2 wasrapidly redistributed and was present in the Co. Sur-prisingly, in the performed experiment it was foundthat the radioisotope distribution was such as to in-dicate that chain initiation could occur from an in-termediate derived from CO2, but that chain growthoccurred only by using carbon derived from CO. Furthermore,of the CO2 that was converted, half, or 2 40/more than half, was converted to hydrocarbonsThe present data, together with their earlier dataare consistent with the fact that high partial pressureof CO is inhibiting the adsorption of CO2 and thiscombined with the low H2/CO ratio, will allow CO2to only initiate chain growth. However, the lower COpartial pressure in the present study, together withthe higher H2/CO ratio, caused significant WGs to Figure 3. Effect of CO 2 contained syngas on theoccur. Thus, the synthesis was conducted with theF-T synthesis performanceadded i4C distributed about equally between the coand the CO2. and this caused a linear increase in theIt was found that the presence of CO2 in the wholeradioactivity/mol for the C1-Ca hydrocarbons 48syntheis process did not influence the net CO2 selec-The conversion of(CO+CO2)/H2 mixtures us- tivity(WGS reaction), but slightly lowered the rateing cobalt based catalysts under typical F-T synthesis of the F-T synthesis reactionconditions has also been conducted [ 52. The resultsFor most of the researchers, it is basically exshow that in the presence of Co, CO2 hydrogenation pected that CO2 addition to the feed gas will inhibitis slow. In the case of only CO or CO2 hydrogenation, CO conversion to CO2 via the WGS reaction. It hassimilar catalytic activities were obtained, but the se- been cited that the SASOL CFB process has about 6%lectivities were very different. In contrast to the Co COz in the feed gas and operates at 340C, 2.0 MPa,hydrogenation results, the CO] hydrogenation prod- with a fused iron catalyst and performs at high con-ucts contained about 70% or more of methaneversionIn the liquid phase synthesis, lower temperaturesHowever, according to the above conclusions, inare utilized with either a cobalt or iron based catalyst spite of the fact that CO2 can lead to the oxidationand the WGS may be far from the equilibrium value. of Fe carbides, the presence of the added COz in theTo make a pracitcal confirmation, the study on feed gas would not influence the CO conversion ratesthe influence of CO, addition was conducted in a hTIIn the designing work for ICTL/GTL commer-bench-scale slurry bubble column reactor, and the re- cialization on the basis of pilot experiments and R&Dsults are demonstrated in Figure 3. An experiment studies, there is no fixed technology relating to CO2with a feed gas of CO/H2=50 /50 is shown in the up- removal and/ or to their percentage in the recycledper figure. The CO conversion at the steady state feed gas. Among the ICTl technology vendors, un-was about 60%. In another set of experiments, the like theial one SASOl that operates at afeed gas was composed of the same Co /H2 ratio and cor中国煤化工ut1.6 without cO2was changed after a period (8 h)to a feedgas contain- rerCNMHGch as Caer in USAing 50%CO2, while the pressure in the whole process and IdC UAS in China strongly suggested that CO2was elevated to keep the H2 and CO initial partial is preferred to be removed, at least partially. Thepressure constantdifference in operation strategies can be attributedJournal of Natural Gas Chemistry Vol. 16 No. 4 2007to the different catalyst activities for the WGS reac-For other components in the syngas, there are spe-tion,and without detailed explanation to the gross cial reviews and research results. Small amount ofrbon efficiency basing on the coal resource. As CHg is thought to have less effect on the F-T processmentioned above, at the condition of a catalyst with Most concerns are focused on the impurities in thegh F-T activity but a lower WGS activity, it could syngas and can be abstracted briefly here. By far theprovide feasible strategy to keep CO2 in the recycle most abundant, the most important, and the mostlyline. However, it still requires an optimum balance in studied F-T synthesis catalyst poison is sulfur. Ide-gross energy efficiency and carbon efficiency starting ally, there should be no sulfur in the syngas. Therefrom the gasification or the reforming unit. More im- is, however, always a small amount that will contactportant, it needs to prevent the inert gas build-up in with the catalyst. The level of gas cleaning requiredthe reactor and to reduce the circulation cost. The based on economic considerations, namely, how longcarbon efficiency takes the chemical feedstocks and the catalyst remains active versus the investment forenergy streams into account [53 It is a good mea. gas cleaning. Fischer has recommended 4 ppm as thesure of plant operating economics, because the higher maximum sulfur content in the F-T synthesis feed gasthe carbon efficiency, the higher the usage of the feed 58. Dry recommended a maximum sulfur contentmaterial and energy for the production of the desired of 0.2 ppm based on commercial experience in theSASOL plantsFor cobalt-based catalyst systemsis not nec.ssary to remove CO2 from the synthesis gas before 4. Conclusionsusing it as feedstock for the hydrocarbon synthesisstage. Besides to SASOL, BP has patented a processThe H2/co ratio of the syngas is an importantcomprising continuously introducing a synthesis gasdesigning variable for maximizing the production ofed stream having 0. 1 to 50%v/v of carbon dioxide high quality diesel. For the GTL process on cobalt-into a continuous stirred reactor system(cobalt cata-based catalyst, it is well recognized that the H2/ colyst)54ratio in the feed gas should be close to 2.0, i.e, thestoichiometric reaction ratio. For the ICTl process3.3. Effects of N2 in the feed gasto date the iron based catalyst is the onlyable option. After gasification the syngas has to beThe use of nitrogen-rich syngas could be an alter- adjusted by the water gas shift unit to provide"suit-native to classical processes with nitrogen-free syn- able"H2/Co ratio for the F-T synthesis. The sig-gas, because the investment costs are probably lower, nificant loss of carbon and energy efficiency occurredi. e, syngas is produced by partial oxidation with air, in this shift step, and then doubled in the F-T syn-which eliminates the need for an air separation plant, thesis due to the unavoidable WGS reaction regardand a process with nitrogen-rich syngas does not need less of a high or low H2/CO ratio. As a result, theto utilize a recycle loop and a recycle compressorH2/co ratio in the feed gas for an iron based catalystFor the sake of studying the effect of nitrogen gas could be in the range of 1.1 to 1.7, depending on thein the syngas, kinetics of the F-T reaction was stud- WGS activity and recirculation ratio. Furthermoreed on an iron based catalyst, which indicates that it could be a bold idea to suggest in feeding the F-Titrogen only dilutes the syngas, and therefore, has reactor with a syngas just after the gasification andno influence on the kinetics if the partial pressures of purification, but without any shift process, namely, acarbon monoxide and hydrogen are kept constant 55. H2/Co ratio about 0. 7 will be applied for undergo-Similar conclusion was drawn by Jess and Kuntze ing the F-T reaction and accompanied with the WGSet al, their experimental results of F-T synthesis with reaction. Thus the recycling feed gas to the F-T reac-nitrogen-rich syngas indicate that nitrogen does not tor is increased to about 1. 1 at a recirculation of 2.0affect the reaction kinetics, provided that the partial The overall carbon loss as well as the facility investpressure of carbon monoxide and hydrogen are kept ment andinn cost (shift unit removal) will beconstant [56, 57中国煤化工 detailed design,In conclusion, nitrogen gas only dilutes the syn- figtCNMHGICTL process overgas, and therefore, has no influence on the kinetics if particular catalyst will be provided in another reportthe partial pressures of carbon monoxide and hydro-For the issue of CO2 addition to the feed gas priorgen are kept constant.to the F-T synthesis, it has been studied from two340Yijun Lu et al. Journal of Natural Gas Chemistry Vol. 16 No 4 2007perspectives. One is the impact of CO2 in the react- [13] Shell-Qatar internet information, Qatar Petroleumng gases on the competitive activities of F-T versusand Royal Dutch Shell plc today announced theWGS, and the other is how far the CO2 hydrogena-launch of the world-scale integrated Pearl Gas totion to the product could go. It has been proved thatLiquidsprojectinQatar.2007,07,www.shellobserved hydrocarbon formation from CO2 involvescom/home/content/qatar/news-and library /press.a first reverse wGs conversion of cO, to CO, whichreleases /2006/integrated-pearl- gtl-project. htmlthen reacts to hydrocarbons. For Co2 hydrogenation, [14 Thackeray F. F-T GTL Approach Threshold, Worldit was observed that the operating temperature mustbe rather high because of the equilibrium constraints15 Statistical Review of World Energy, LondeReview Autumn 2003, P32, Turning Natural Gas intofor the reverse WGs reaction. This limits the occurLiquidrence in the LTFT process, which has been scaled-up [16] Holmen A. GTL introduction, Norwegian Universityworldwide. This indicates the sole role of CO2 in theof Science &z Technology, 2003, Trondheim, Norwayfeed gas as a diluting gas. Furthermore, the extra CO2 [17] Yoshifumi S, Kazuto K, Ken I Japan Oil, Gas Andin the feed gas apparently leads to larger reactor vol-Metals National Corp, Annual Report of TRC's Ac-umes and /or higher gas recycle rates, and compressortivities,2005,2004:13[18 Buller A T, Schanke D, Rytter E, Hansen R. Gas-to-with cobalt-based catalysts, CO2 is not producedLiquids Technology, Memori No. 8, Statoil Researchduring the F-T synthesis, and is believed that it onlyCentre, Statoil ASA, December 2006dilutes the reacting gases under most conditions.19 Steynberg A P. wO Patent 2006097905, 2006(20 Krishna R, van Baten J M. Topics in Catalysis, 2003,References26:121 Davis B H. DOE Technical report, DE-FT40308, 1999, Technology development forA-FC26-l Dry M E. Catalysis Today, 2002, 71: 2272 Anderson R B. The Fischer-Tropsch Synthesis. Lonand cobalt F-T catalystsdon: Academic Press. 198422 Davis B H. DOE Technical report, DE-AC223 Steynberg Andre, Dry M. Studies in Surface Science94PC94055, 1999, Technology development for ironand Catalysis, 2004, 152: 4064 Shen J, Schmetz E, Stiegel G J, Winslow J C,Ko-VAZQUEZ V H C. MS Thesis, Texas A&M Univer-rnosky R M, Madden D R, Jain S C. Studies in Sur-sity, Spray drying and attrition behavior of iron catface Science and Catalysis, 2007, 163: 315alysts for slurry phase F-T synthesis5] Rahmim I. Recent Advances, Economics, Recent Ad. [24] Datye A K. DOE Technical report, DE-FG22-vances, Economics, Prospects 26th IAEE Annual In-95PC95210, 1997, Slurry phase iron catalyst for in-ternational Conference, Prague-June 20036) Borodko Y, Somorjai G A Applied Catalysis A: Gen- (25] Xu J, Iglesia E. Technical progress report,DE-FC26-eral,199186:35503NT41964, Design, Synthesis, and Mechanistic Eval-7 Workshop Report, IEA Coal Industry Advisory Boarduation of Iron-Based Catalysis for Synthesis Gas Conworkshop, IEA Headquarters in Paris, 2 Nov 2006version to Fuels and Chemicals[8] Miller C. Lowell, Coal conversion-Pathway to alter- (26 Bhatt B L. DOE Technical Report,DE-FC22nate Fuels, Congressional Briefing Washington, D C,95PC93052, 1999, Liquid phase Fischer- Tropschanuary 19, 2007demonstration in the la porte alternative fuels devel9 Geertsema A. Barriers and opportunities relating topment unitn of coal to liquids, Presentation at Pur27 Li S, Krishna S, Li A. Journal of Catalysis, 2002, 200due University Energy Security Summit, August 2920200628Shena J, Schmetza E, Kawalkin G J, StiegeG J. Top-[10 ShenHua Group Has 8 CTL Projects in Pipeline, newson16June2006,NewsofGreenCarCongresshttp[29]derBurgtV,SieST.lIquidHydrocarbonsfromNat-//www.greencarcongress.com/coaltoliquids-ctl/ural Gas, Petro Pacific 1984[11] Jacobs F E. Workshop on Trends in Oil supply and 30 PCay j, Allais schwe-Demand and Potential for Peaking of conventional Oil中国煤化工6:137Production, 2005, Oct. 20-21, Washington DCCNMHG106:3412 Sasol-Q2information, website: 32 Davis B H. DOE Technical report, DE-AC2om/sasol_ internet /downloads91PC90056, 1993, Preparation of precipitated iron F.HWC20010404_1089876828459.pdfT catalysisJournal of Natural Gas Chemistry Vol. 16 No. 4 200747 The first quarterly meeting between Dr. B H Davisand Dr. E Iglesia, University of California, BerkeleyTechnology Development for Iron and Cobalt Fischer-34 Raje A, Inga J R, Davis B H. Fuel, 1997, 76(3): 273Tropsch Catalysts, DE-FC26-98FT40308, December35] Xu L, Bao S, Houpt D J, Lambert S H, Davis B H18,1998Catalysis Today, 1997, 36: 34748 Herranz T, Rojas S, Perez-Alonso F. Applied Cataly-i Raje A, Davis B H. Catalysis Today, 1997, 36sis A: General, 2006, 311: 6637 Govender N S, van Vuuren M Janse, Claeys M49 Lee M D, Chern W S, Lee M D, Dong T Y. AppliedSteen E Ind Eng Chem Res, 2006, 45: 8629Catalysis A, 1991, 72: 26738] Landis S R, Espinoza R L, Mohedas S R, Belt B A, 50 Riedel T, Claeys M, Schulz H, Schaub G, Nam S,JunMelquist V H, Goodwin III R T. US Patent 6872753K, Choi M, Kishan G, Lee K. Applied Catalysis A:2005Genera,199,186(1-2):20139 Boelee J H, Custers J M G, Van Der Wiele K Applied 51 Xu L, Bao S, Tan L M, Chawla B, Dabbagh H Pro-Catalysis A, 1989, 53: 1ceedings of the 11th Annual International Pittsburgh40 SASOL internal exchange documents, Technical studCoal Conference. 1994.88case for ICTL plant, 20062 Zhang Y, Jacobs G, Sparks D E, Dry M E, Davis B[41 Chao K C, Lin H M, DOE-AC22-84PC-70024-T9H. Catalysis Today, 2002, 71: 4111988, Synthesis gas solubility in Fischer-Tropsch [53 Mukoma P, Hildebrandt D, Glasser D Ind Eng Chemslurry: Final reportRes,2006,45(17):592842] Song H, Zhang H, Ying W. Journal of East-China [54] Albert Beorge. US Patent 7138434, 2007University of Science and Technology, 2001, 30: 555 Jess A, Popp R, Hedden K. Applied Catalysis, 199943 Benham C B, Bohn M S. Maximization of diesel fuel186:321production from an iron based F-T catalyst, AIChE [56 Hedden J A. Prepr Am Chem Soc Diu Pet Chemspring national meeting, Houston, TX, March 14-18,2000,45(2):20257] Kuntze T, Hedden K, Jess A. Erdol Kohle erdgas44 Mukoma P, Hildebrandt D, Glasser D Ind Eng Chem1995,111(2):67Res,2007,46:15658 Fischer-Tropsch Synthesis, DOE report, 3rd Quarter45 Steynberg A P, Herman G. FueL, 2004, 83: 765Harvard Energy Journal Club meeting, Harward Uni-Steynberg A P. Europe Patent 20010982639, 2001versity中国煤化工CNMHG