甲烷空气部分氧化制合成气喷动床反应器的研究

hinese J, Chern. Eng, 11(6)643-648(2003)Catalytic Partial Oxidation of Methane with Air to Syngasin a Pilot-Plant-Scale Spouted Bed Reactor*WEI Weisheng(魏伟胜), XU Jian(徐建), FANG Dawei(方大伟)andBAO Xiaojun(鲍晚军)The Key Laboratory of Catalysis, China National Petroleum Co., University of Petroleum, Beijing 102249, ChinaAbstract On the basis of hydrodynamic and scaling-up studies, a pilot-plant-scale thermal spouted bed reactor0 mm in ID and 1500 mm in height)was designed and fabricated by scaling-down cold simulators. It was tested foraking syngas via catalytic partial oxidation(CPO)of methane by air. The effects of various operating conditionssuch as operating pressure and temperature, feed composition, and gas flowrate etc. on the CPO process wereinvestigated. CH4 conversion of 92. 2% and selectivity of 92.3% and 83.3% to CO and H2, respectively, were achievedat the pressure of 2.1 MPa. It was found that when the spouted bed reactoflow regime, the temperature profiles along the bed axis werethan those operated within thefixed-bed regime. The CH4 conversion and syngas selectivity wse to thermodynamic equilibriummits. The results of the present investigation showed that spouted bedconsidered as a potential type ofemical reactor for the CPO process of methaneKeywords spouted bed, methane, catalytic partial oxidation, syngas1 INTRODUCTIONfeaturcs of CPO process lie in its slightly exothermicSyngas, a mixture of hydrogen and carbon monox. with very fast reaction rate, and H, to CO molaide with varying compositions, is an important chem- ratio approximate to 2: 1 in produced syngas, whichical feedstock in many chemical processes, such as in is appropriate for methanol production and Fischer-the production of methanol, ammonia and hydrocar- Tropsch synthesis. With these features, CPO processbons. Lowering operating cost and investment capi- encounters the problem of hot spot formation in catal has been challenges to both chemists and chemical alyst beds resulting in catalyst melting and deactiva-engineers. Among various routes proposed for making tion. Thus the innovation of suitable types of chemicalsyngas from methane up to date, the catalytic partial reactors that satisfies all requirements of the reactionoxidation(CPO)route seems to be more promising system is one of the technological keys for the com-onell. The main reactions occurring in the CPO pro- mercialization of CPO process. As a potential reactorofchane are listed beloyfor the CPo process in recent years, special attention(1) Main reactionhas been paid to spouted bed reactors for their structural and How characteristics, which are desirable forCH4+0.502→CO+2H2,those chemical reaction systems demanding strict con-△H298k=-3559kJ·mol-1(1) trol of contact time between gas and solids and fastremoval of a large amount of reaction heat[3-5.In2)Combustion reactionthe past, the possibility to employ spouted beds asCH4+202→CO+2H2O,chemical reactors had been explored for different pro-cesses, such as the combustion of fuel-rich mixture△H298K=-802kJ.mol-1(2) catalytic polymerization of olefins, fue gas desulfurization,and so on[6-8. The spouted bed reactor(3)Reforming reactionshas been tested for CPO process by reacting methaneCH4+CO2→2CO+2H2with oxygen in spouted bed with conical bottom(25-△H2K=247 kJ. mol-1(3) particle sizes of 0.6-1.2 mm. Depending upon thefeed flow rate and the operating pressure, methaneCH4+H2O→CO+3H2conversions of 65%0-95% were achieved with Cha toH206kJ(4) O2 molar ratios varying in the range of 1.39--1.77leading to a syngh a h to o ratioCompared with the steam reforming (SMR) by close to 23). The CPO process had also been inves-which syngas is mainly produced nowadays, the salient tigated in a conical spouted bed reactor(50mm ID)Supported by the Doctorate Discipline Foundation of the Ministry中国煤化工63)& China NationalPetroleum CoTo whom correspondence should be addressed. E-mail: baoxjebHCNMHGChinese J, Ch. E(Vol, ll, No 6)by using Na/La/Al2O3 catalysts with particle sizes ofto GC exhaust0. 15 to 0.6mm at pressures of 0. 1-1 MPa and tem-peratures up to 1323 K, CH4 to O2 ratios of 1.6-2.5and 40% N2 dilution(4. Both of the above investiga-tions reported lower methane conversions and syngasselectivities than the thermodynamic equilibrium.Although numerous case studies have beerported, the reaction engineering aspects of spoutedbed reactors remain almost untouched. To further un- distilled waterderstand spouted beds and their advantages for somekinds of chemical reaction systems, e.g., the partialgoxidation of methane, a thorough investigation in-cluding hydrodynamic studies under ambient pressureand temperature in cold simulators, and scaling-up atFigure 1 Schematic diagram of the experimentalelevated temperature and pressure in spouted beds1--compressor; 2-gas cleaner; 3-buffer tankof different sizes have been carried out in previouworks(]. The purpose of the present investigation-mass How controller; 7--air preheater;is to construct a pilot-plant-scale spouted bed reactor8--stean preheater; 9-natural gas preheater10--spouted bed reactor; 11-safety valve:on the basis of cold model studies and thus to obtain12-backthe systematical information on the applicability ofthe spouted bed as the chemical reactor for the CPO2 EXPERIMENTALa thermal pilot-plant-scale spouted bed reactor formaking syngas via the CPo of methane by air wasdesigned and fabricated by scaling-down cold simula-tors. The overall schematic diagram of the experimen-tal equipment is shown in Fig. 1. Natural gas from thecity pipeline was desulfurized first in a fixed bed filledwith zinc oxide, then compressed to about 10 MPa bya natural gas compressor, and finally stored in a seriescool ing waterof buffer gas cylinders that are mutually connectedThe air as oxidant was compressed to 3 MPa by anair compressor. Before entering the pilot-plant-scalespouted bed reactor, the pressure of natural gas andthe purified air was reduced to the operating pressureby two separate reducers and then preheated to 673KTheir flowrates were controlled and measured by twoseparate mass flow controllers(MFCs). Natural gaswas then mixed with air in the feeding nozzle locatedat the bottom of the spouted bed and then enteredthe bed reactor. In the case dealing with operationswith rich oxygen, purified oxygen from oxygen stor-age cylinders was fed through an MFC into the airline after the MFC for the air flow. If steam addion was needed in experiments, distilled water wasfed into the steam feeding line by a metering pumpthen preheated to 673 K, and finally mixed with theFigure 2 The detailed structure of theilot -plant-scale spouted-bed reactorair before entering the spouted bed reactor1--safety valve: 2-pressurized steel shell; 3--spouted bedThe structure of the pilot-plant-scale spouted-bedtemperature and pressure measurernent ports;reactor is shown in Fig. 2. In the bottom section of5-heater; 6-gas inletthe bed there is a detachable chassis with a collect-ing pipe of 8 mm ID used as a nozzle to which all the mixing in the collecting pipe, all feed gases enteredfeeding tubes(8 mm ID)were connected. After fully中国煤化工December 2003YHCNMHGCatalytic Partial Oxidation of Methane with Air to Syngas in a Pilot-Plant-Scale Spouted Bed Reactor 645with an included angle of 60 and a ceramic cylin. developed by the research group of the present inve-der of 50 mm ID and 0.5 m in height. The outer side tigation. The detailed procedure can be referred toof the ceramic cylinder was surrounded by an electric Ref. 12. The physical properties of the catalyst par-heater to provide heat during the reaction startup ticles used are given in Table 1The heater was controlled by a thermocouple T6 lo-cated on the outer surface of the cylindrical sectionTable 1 Physical properties of catalyticas shown in Fig 3. Just above the ceramic cylindermaterials uswas a cooling cylinder of 50 mm in ID and 1000 mm inlength with a shell-tube structure, in which the reacCatalykgion product, syngas, was quickly cooled down by thewater as coolant flowing in the shell side to avoid the1.7830called methanization reaction, a reverse reaction ofthe CPO reaction, thatTo monitor the operation mode of the spouted bedconversion of methane. The two cylinders were con- reactor, the time series of differential pressure fuctuanected by a flange and installed inside a carbon steel tion signals of the bed were sampled and analyzed bypipe that was insulated with thermal barriers. Seven the method of power spectrum analysis as describedthermocouple ports spaced at 50 mm apart starting in detail in Ref [ 13] to identify the flow regimesfrom the intersection between the conical bottom andmpositions of CH4, O2, N2, CO and CO, inthe bottom cylindrical section, two pressure tapping reactants and or products were analyzed by an on-lineports located at the bottom of the conical section and gas chromatograph, while those for H2 and H2O werethe upper cylindrical section, respectively, and a sam- calculated from the balances of C, H andO atomspling port between the two cylindrical sections were All data given in the present paper were average re-arranged, as shown in detail in Fig 3. The thermo- sults of at least two samples. It was estimated thatcouples T'l, T2 and T3 were inserted into the spouted the accuracy of component composition analyses wasbed 10mm in depth and thus the temperatures mea- within +3%, as it was discussed in Ref. 12. The con-sured indicate the reaction temperature profile in the version of methane and the selectivity to CO and Hon the wall of the spouted bed. The temperature at are defined as followsthe entrance of the nozzle to the bed was measured byHa ithe thermocouple TOSco=Foo, et Fco, ut[air[preheater]3 RESULTS AND DISCUSSION回p[rea3.1 Effects of bed temperature and operatingpressureFigure 3 Arrangement of the temperature andTable 2 lists the CH4 conversion and selectivitiesfor CO and H2 obtained at different bed temperaturesThe whole spouted bed was encased in a heavy and operating pressures with O2/CHa molar ratio atsteel shell that surrounds the spouted bed and its con- 0.58 in the feed. As shown in Table 2, CHA conver-riections. The space between the spouted bed and the sion and the selectivities to Co and H2 increase withshell was filled with the syngas produced. The pres- the increasing bed temperature at a given operatingsure in the space was controlled equal to the pressure pressure. With O2/CHA molar ratio in the feed ofin the spouted bed so that the net pressure exerted on 0.58 and the bed temperature at 1073 K, the effectthe bed wall is zero. In the lower part of the shell there of the operating pressure is shown in Fig. 4. It canwas installed with a cooling coil to avoid the overheat. be seen that increasing pressure results in considering of the shell due to heat transfer The shell protects able decrease of CH, conversion and syngas selectiv-against potential hazards due to reactor failure when ity. The results presented by Table 2 and Fig 4 arehomogeneous explosive reactions take place inside the in good agreement with those predicted by thermodyreactor. A handheld explosive gas alarm detector was namic analyses 2l which stated that lower pressure andused to monitor the atmosphere around the reactor. highfavorahle to methaneThe CPO experiments were conducted with a ver中国煤化工 The higher syngasNi/La/Al]O3 catalyst (containing 7%(by mass)Ni] pressCNMHGeoperating costsChinese J, Ch. E. 11(8)843(2003)646Chinese J. Ch. E.(Vol 11, No 6)because most of the downstream processes are oper- bed experienced the transition from fixed bed regimeated at elevated pressure. The decrease of methane to stable spouting regime, mass transfer behavior wasconversion and syngas selectivity at higher operating gradually improved, and consequently the CPO reac-pressure should be compensated by increasing the re- tion was gradually enhanced, resulting in higher bedaction temperatureperature which is also favorite to the CPo reac-tion and yields better reaction performance. However,Table 2 Effect of operating temperCHwhen the stable spouting regime was reached, furtherconversion and syngas selectivityincrease of space velocity gave rise to the decrease ofthe residence time of the reactants in the spouted bedMPa K9 and finally led to the decrease of CH4 conversion and1138 1157 1109 893 87.4 83.5 91.9 syngas selectivity. In all experiments, the measured9L788.792.4H4 conversion and syngas selectivity were a little bit1.0 1150 1166 1095 1012 95.6 91.3 93.7 lower than those predicted by thermodynamic analy751102108497492894.088.9sis, suggesting that long contact time should be em-390.3loyed to obtain better results. Although lower than89.6 88.5 91.0 the results predicted by thermody120212091139101293.189993.9ethane conversion and syngas selectivity obtained in1185 1198 1119 724 83.8 88.7 86.4 the present investigation were much better than the83.686.587.4by Marnasidou etbelieved that11881207109584283.987287211971197110792185.2the improvement in methane conversion and syngas38.6 87.6 selectivity should be attributed to the use of coarserparticulate catalysts in the present investigation,com-pared with finer particulate catalysts of sizes rangingfrom 0.15 mm to 0.6mm in diameter, which may beeasily entrained into the freeboard region of the bedand thus responsible for the failure to reach thermo-dynamic equilibrium conversion under the assumptionof adiabatic operation. On the other hand, the poorspouting property of finer particles may also affect thereaction behavior in the bedTable 3 Effect of space velocity on CH4 conversionFigure 4 Effect of pressure on CHa conveand syngas selectivityH CHA conversion;-+CO selectivity;Flow regiFH2 selectivity27000113689.190.7851 fixed bed32000114491,490.4869 fixed bed3.2 Effect of space velocity1190 93.6 92.0 87.6 stable spoutingpace velocity is defined as the ratio of the gas vol-460001185 93.5 91.2 87.3 stable spoutingumetric flowrate at standard state through the nozzle1187 91.6 91.5 87.0 stable spoutingto the catalyst volume in the spouted bed and withthe unit h-.With O2/CHA molar ratio in the feed of 3.3 Effect of O2/CHa ratios in the feed0.58, catalyst loading 220 ml, pressure 1.0 MPa, andowShe effects of O2/CH4 molar ratIothe bed temperature(T6)about 1073 K, the effect of in the feed on CHa conversion and syngas selectivityspace velocity on CH4 conversion and syngas selec- with temperature from 673 K(To)to 1023 K(T6)andtivity is shown in Tablc 3. It was found that when pressure at 2.1 MPa. With the increase of O2/CHathe how regime in the reactor changed from fixed bed molar ratio in the feed, CH4 conversion increased sigto stable spouting regime with increasing the space nificantly, and syngas selectivity decreased slightly. Tovelocity both CH, conversion and syngas selectivity avoid the deposition of soot on catalyst, the O2/CH,increased, and after the stable spouting regime was molar ratio in the feed should be above 0.5, as sug-eached, the further increase of space velocity resulted gested by the thermodynamic analyses by the authorin slightly decrease in CH4 conversion and syngas ge- of the present article 21. However, it should be keptlectivity Similar results were reported by Marnasidou in mind that the higher O2/ CH4 molar ratio in theet al., but no explanation was given. The most prob feed than the stoichiometric ratio of the CPO reac-able reason for the lower CH4 conversion and syngas tion means poorer syngas selectivity, and thus a com-electivity might be attributed to the relatively highere made between CH, conversionbed temperature when operated with higher space ve-中国煤化工 maximum syngaslocity. With the increasing velocity, the flow inside theindicated that theCNMHGdecember. 2003Catalytic Partial Oxidation of Methane with Air to Syngas in a Pilot-Plant-Scale Spouted Bed Reactor 647suitable O2/CH molar ratio should be in the range the experimental results were given in Table 6.Itfrom 0.58 to 0.62showed that as the O2/N2 molar ratio in the feed wasincreased, the reaction temperature, CH4 conversionTable 4 Effect of O2/CHa ratio on CH4 conversion and CO selectivity increased obviously, but H2 selec-tivity decreased slightly. This tendency could be in-terpreted as a consequence of the increase in adiabaticreaction temperature due to the decrease of redundantnO2/ncH4 CCH4 Sconitrogen into the reaction system. On the other hand8,586,683.192,689.9however, the use of oxygen-enriched air will obviously86.192.689.7increase the contact probability between the methane6.192,186.60.292789and oxygen and thus lower the selectivity to H2. Be3892688.7cause of the relative larger size of the reactor, using92292.383.397692.487.2xygen cylinders as oxygen stoinsufficient to maintain a long-term steady operationDetail thermodynamic analyses showed that the to check the infuence of O /N Mma k was conducted3. 4 Effect of steam additionddition of steam could considerably improve the per-formance of the CPO reaction and it is also beneficialto the heat balance due to the enhancement of the enTable 6 Effect of O2/N2 ratio on CHA conversionand syngas selectivitydothermic steam reforming reaction. When O2/CH4molar ratio in the feed, pressure and total flowratof air and methane were fixed at 0.62, 2.1 MPa, and220L min", respectively, the effect of steam addi1/4(air)119295198.7l/2127397.1tion was investigated and quite positive results wereyielded. As reflected in Table 5, significant improve-ent in both CHa conversion and CO selectivity was 3.6 Temperature distributions within the bedachieved. It should be noted that the decrease of theThe present investigation was initially motivatedselectivity to H2 as listed in Table 5 may be attributed by the consideration that because of the structuralto the inappropriateness of Eq (7)for the calculation and hydrodynamic characteristics of spouted bedof H2 selectivity, in which the item FH,O, out in the enu- reactors, they should be beneficial to the CPo processmerator on the right-hand side includes steam added development in avoiding the formation of hot spotsto the reactor and reduces the values of the H2 selec- as it was frequently encountered in fixed bed reactorstivity. It was also observed that H2/co molar ratios as reported by many researchers. To assess this adbecame larger with the increasing of H2O/CH, ratio vantage of spouted bed reactors over other parallelin the feedtechnologies in solving the problem of hot spot forma-3.5 Effect of O2/N2 ratioion, comparative experiments were conducted for theWhile the CPo process with air as oxidant has spouted bed operated at fixed bed and stable spout-the advantage to omit the expensive oxygen gener- ing regimes, respectively, and the typical results wereation over other competing technologies for syngas represented by Figs. 5 and 6. It should be pointedmaking, it suffers the problem that the introduction out that the temperatures measured at the locationsof unnecessary large amount of nitrogen may intro- of 270 mm and 350 mm from the nozzle were aboveduce a number of problems to the downstream pro- the upper bed surface, so the temperature differencescess. To overcome this shortcoming, it is reasonable through the bed could be roughly represented by theto consider using of oxygen-enriched air as oxidant. temperature difference between locations of 50 mmTo this end, experiments were conducted to check the and 130 mm from the nozzle. By the analyses of theeffects of O2/N2 molar ratio in the feed on CH4 con- differential pressure fuctuation signals, it was conion and syngas selectivity under the conditions of firmed that at p= 1.0 MPa and Q= 120Lmin-5MPa, no,/ncH.=0.58, T0=673K, T6=873 K, the bed was operated at fixed bed regime, and whencatalyst loading 220 ml, and Q= 90 L. min-, and the total volumetric flowrate of the reactants enteringTable 5 Effect of steam addition on the CH4 conversion and syngas selectivityHyo/nCHO,%TH,/nco90.5127686.3127992284.11279125992.68791278YH中国煤化工CNMHGJ.Ch.E.11(6)643(2003)into the bed reached at Q= 150 L min -, the stable beds as a potential chemical reactor for the CPOspouting regime emerged. As shown in Fig. 5, when cess. The experimental results also revealed that thethe bed was operated within the fixed bed regime, gas-solid contact mode in the spouted bed presented athe highest temperature could reach to about 1230 k better solution for the large amount of heat producedand the bed temperature difference was 95 K, whereas during the catalytic partial oxidation reaction, whithe highest temperature and the bed temperature dif- indicates an advantageous route for the industrialference were 1176 K and 80 K when operated within actor development of the CPO processthe stable spouting regime, respectively, indicating thecyclic motion of particles inside the spouted bed couldbe helpful in avoiding hot spot formation and made NOMENCLATUREthe temperature profiles much more even. Similar be-havior was observed when the bed was operated at F.particle diameter.molar flow rate of species iP=2.0MPa, H=120 mm, and no, /ncH, =0.62,static bed height, mmshown in Fig. 6mole number of species i,operating pressure, Pa1300gas flow rate through the central orifice understandard state, Lmin-1ity to species i,pbarticle density, kg m-3REFERENCES1 Marschall, K.J., Mleczko, L,"Short-contact-time reactortion in thfor catalytic oxidation of methane, Ind. Eng. Chem.Res38(5),18131821(1999Ox/ncH=0.622 Xu, J., Wei, w.s., Bao, X.J. ,"TheQ,Lmin-:120;-·150Chinese J. Chem. Eng, 10(1),56-62(2002)3 Brophy, J H, Telford, C D, Process for producing synthe-sis gas by partial combustion of hydrocarbons, wO Pat4 Marnasidou, K.G., Voutetakis, S.S., Tjatjopoulos, G. .salos, I.A.,"Catalytic partial oxidation of methane tosynthesis gas in a pilot plant scale spouted-bed reactor"hem.Eng.Si,54,369l-3699(1998)J, Vasalos, G,"Catalartial oxidation of methane in a spouted-bed reactor, De-sign of a pilot plant unit and optimization of operating con6 Weinberg, F.J., Bartleet, T G, Carleton, F B, Rimbotti,distance from nozzle, mmP,"Partial oxidation of fuel-rich mixture in a spouted bedFigure 6 Axial temtre distribution in theombustor", Combust, Flame, 72(3), 235-239(1988)spouted-bed at p=2.0MPa, H= 120 mm, and7 Lazar, M, Arandes, J M, Zabala, G, Aguayo, A.T., Bil-no2/ncH4=0.62D,J,"Design and operation of a catalytic polymerizationQ, Lmin-1:-Hspouted bed regim637-1643(1997)8 Ma,X. Kaneko, T,, Xu, G, Kato, K, "Infuence of gas4 CONCLUSIONSonents on removal of So2 from flue gas in the semi-dA pilot-plant-scale spouted bed reactor was de-FGD process with a powder-particle spouted bed", Fuel, 80signed and fabricated based on the previous hydrody(5),673-680(2001)namic and scaling-up studies and was used for makingarticles at elevated pressure", J. Chem. Ind. Eng, 53 (3)syngas via catalytic partial oxidation of methane by281284(2002),( in Chinese)air in the present investigation. The effects of oper10 Wang, S, Xu, J., Wei, w.S., Shi, G, Bao, X.J. " Gas spout-g hydrodynamics of fine particles", Can. J. Chem. Eng.ating pressure and temperature O 2/CH4 molar ratio78(1),156-160(2000)steam addition, use of oxygen-enriched air, and flow 11 Wang, M, Xu, J, Wei, WS, Bao, X.J. Bi, x.T."Preregime on methane conversion and syngas selectivitywere deternined experimentally. The results showed 12 Zhang, Z.B. "Ni-base catalysts for syngas making via cat.that with the bed operated within the stable spout-lytic partial oxidation of methane",Ph. D. Thesis, Uni-ing regime, methane conversion and syngas selectivityclose to the thermodynamic equilibrium limit could beniversity of Petroleiobtained. It demonstrates the capability of spouted中国煤化工December, 2003CNMHG