喜歡就充值下載吧。。資源目錄里展示的文件全都有，，請(qǐng)放心下載，，有疑問(wèn)咨詢QQ:1064457796或者1304139763 ==================== 喜歡就充值下載吧。。資源目錄里展示的文件全都有，，請(qǐng)放心下載，，有疑問(wèn)咨詢QQ:1064457796或者1304139763 ==================== Chemical Engineering and Processing 46(2007)918923Ammoxidation of propylene to acrylonitrile in abench-scale circulating fluidized bed reactor Yongqi Hu a,b,Fengyun Zhao b,Fei Wei a,Yong Jin aa Department of Chemical Engineering,Tsinghua University,Beijing 100084,Chinab Institute of Chemical and Pharmaceutical Engineering,Hebei University of Science and Technolgy,Shijiazhuang 050018,China Received 16 March 2007;received in revised form 21 May 2007;accepted 22 May 2007 Available online 29 May 2007AbstractThe ammoxidation of propylene to acrylonitrile over Mo-Bi/?-Al2O3catalyst was investigated in a bench-scale hot model riser reactor with7mm i.d.and 30m in length.Propylene conversion and product yields were investigated under various operation conditions and the optimumconditions have been found for the new type reactor.The results show that the efficiency of catalyst is increased by four times and the yield ofacrylonitrileisincreasedby3%fortypeAcatalystandby6.5%fortypeBcatalystincomparisonwithacommercialturbulentfluidizedbedreactor.The yield of acrylonitrile can be further increased through staged air feeding strategy.2007 Elsevier B.V.All rights reserved.Keywords:Propylene;Ammoxidation;Acrylonitrile;Circulating fluidized bed;Riser reactor1.IntroductionTheheterogeneousselectiveammoxidationofpropyleneintoacrylonitrile(AN)is one of the most commercially significantreactions.CH3CH CH2+NH3+32O2713723K,CatalystCH2CHCN+3H2O(1)The features of this reaction include that:(1)it is a highlyexothermalreaction,?H=512.5kJ/mol;(2)thedesiredprod-uctacrylonitrileisaintermediatewhichmayfurtherbeoxidizedintoCO2orCO,gasbackmixinginreactorwillcausetheoverox-idation of AN and thus,the decrease of AN yield;(3)it followsredox mechanism,i.e.,oxygen is supplied by catalyst in theform of lattice oxygen and subsequently the reduced catalyst isreoxidized(regenerated)by molecular oxygen 1.Corresponding author at:Institute of Chemical and Pharmaceutical Engi-neering,Hebei University of Science and Technolgy,Shijiazhuang 050018,China.Tel.:+86 311 88632175;fax:+86 311 88632175.E-mail addresses:,yongqi (Y.Hu).Turbulent fluidized bed(TFB)reactor has been employedfor propylene ammoxidation to synthesize AN for decades.Effective heat and mass transfer in TFB makes it advantageousover packed bed reactor on the control of reaction temperature.However,TFBstillsuffersfromsevereaxialgasandsolidsback-mixing,insufficientgasandsolidscontactandsmallthroughput.Moreover,it is difficult to build a catalyst regeneration region tomeet the requirement of redox reaction mechanism in TFB dueto highly backmixing of gas and solids.To solve these problems,obstacles such as shaped metal-lic articles,screens,grids,perforated plates,horizontal plates,pipes or the likes were laid in a catalyst bed to prevent thecoalescence or growth of bubbles,or to prevent the back mix-ing of gas,thereby improving the contact between the feedgas and the catalyst particles 26.However these methodsare not practical because the construction for laying the obsta-cles is complicated,and the mixing of the catalyst particlesis prevented by the obstacles and the distribution of the cat-alyst in the reactor becomes uneven in terms of space andtime,so that it is difficult to stably and continuously con-duct the operation 7.A loop fluidized bed reactor with bafflefor propylene ammoxidation was proposed and experimentallyexaminedbyChenetal.8.Atwostagefluidizedbedwasdevel-oped for improving gassolid contact,which can be applied in0255-2701/$see front matter 2007 Elsevier B.V.All rights reserved.doi:10.1016/j.cep.2007.05.009Y.Hu et al./Chemical Engineering and Processing 46(2007)918923919redox catalytic reaction such as the ammoxidation of propylene9.ThedisadvantagesofTFBcanbeovercomeinahigh-densitycirculating fluidized bed(CFB)riser reactor 7,1012.Circu-lating fluidized bed allows the spatial separation of propyleneammoxidation in the riser reactor and catalyst regeneration inthe downcomer to maintain catalyst in oxygen-rich state for fur-ther ammoxidation reaction.CFB riser operates under severaltimes higher gas velocity than TFB,which decreases gas back-mixing significantly.Staged addition of air can be permittedin CFB to control oxygen concentration along riser for opti-mal performance 12.In conventional fluid catalytic cracking(FCC),in which CFB is employed,the density of catalyst bedis relatively low,however,high density in riser reactor,aver-age catalyst fraction higher than 10%1315,is needed for thereaction of ammoxidation of propylene to acrylonitrile 7,10for which longer gassolid contact time is required than that forFCC.Moreover,high-density operation allows higher mass andheat transfer to guarantee the conversion under higher operatinggas velocity,and smaller in reactor size resulting in the decreasein construction cost 7.This paper reports hot-model experiments on the selectiveammoxidation of propylene to acrylonitrile over Mo-Bi/?-Al2O3in a CFB riser reactor under high-density condition.Conversions and product yields obtained in the reactor are com-pared with those measured in a commercial turbulent fluidizedbed.2.ExperimentalMo-Bi/?-Al2O3catalyst was used in the hot-model exper-iments.In order to obtain representative experimental results,the catalyst was taken from a commercial TFB reactor.Afteremployed for several months,the catalyst had been reaching itsstable state of activity.The properties of the catalyst are listedin Table 1.The bench-scale circulating fluidized bed reactor is shown inFig.1.It consisted of a riser reactor,a separator,a regenerator inwhich catalyst was reoxidized by air,and an electrical heatingfluidized bed bath.The riser,0.007m i.d.and 30m in length,was spiraled round the regenerator.The long riser can simulta-neouslyguaranteeahighergasvelocity(3m/s)andenoughgasresidencetime.Theriserinspiralingtypeinstalledinafluidizedheating bath,in order to maintain the isothermal conditions ofthe 30m riser reactor.The inclination of the spiral tube againstthe horizontal was about 5.The temperature fluctuation of thefluidized heating bath was controlled within 1K.There was aTable 1The physical properties of the Bi/Mo catalystParticle size distribution(%)45?m and 90?m8.5Density of particle(kg/m3)1800Specific surface area(m2/g)0.68Fig.1.Theschematicofthelaboratory-scalehigh-densitycirculatingfluidized-bed reactor.side air inlet on the riser at 10m from the entrance of feed toexamine the staged air feed.Theflowratesofpropylene,airandammoniawerecontrolledby mass flow controllers.The pipe of reactant air was placed inthe fluidized bed bath for preheating the reactant air,and thenthe air entered the bottom of the regenerator to fluidize the cat-alyst particles which flowed down to an injector.In the injector,reactant air and catalyst particles are mixed with propylene andammonia up to the riser.The amount of carried catalyst wascontrolled by the flow rate of secondary air at the bottom of theinjector.In the riser the ammoxidation of propylene occurred,then the catalyst particles were separated from the gas in theseparator and stored in a catalyst-collector.The catalyst parti-cles were returned to the regenerator in batch to be reoxidizedbytheregeneratingair.Thegaseousproductsfromtheseparatorwent to a combustor for venting or to an absorbing system foranalysis.The residence time distribution of gas in the riser reactor wasmeasuredbyapulseresponsemethodusingathermalconductiv-ity detector.The measured residence time distribution indicatedthat axial Peclet number(Pe)increased with gas velocity andwas larger than 1000 when gas velocity was higher than 2m/s,indicating that gas flow in the riser approached plug flow.Theaverage catalyst volume fraction in the reactor was determinedby weighing the catalyst in the riser after suddenly closing thegas feed.When gas velocity was 2.53.0m/s and pressure dropwas set to 0.03MPa,the typical average catalyst fraction was0.100.12,and the density of catalyst bed was 180216kg/m3,which was 510 times higher than that of FCC riser.Nakamuraet al.7 gave the density of catalyst bed was 100kg/m3ormore,andpreferable200kg/m3ormorefortheammoxidationofpropyleneinacirculatingfluidizedbedreactor.Althoughthereisa world of difference on the structure between the above experi-920Y.Hu et al./Chemical Engineering and Processing 46(2007)918923mentalriserreactorandcommercialscaleriser,thesimilarityontheflowregimes,gasvelocityandaveragecatalystfractiongivessupport to the demonstration on the effectiveness to improvethe yield of AN in fast fluidizing flow regime compared to inturbulent fluidizing flow regime.The gaseous products were collected in three 400ml 0.1NHNO3scrubbers at 273K.A temperature programmed FID gaschromatograph was used to analyze AN,ACL,ACN and ACA.The gaseous products were analyzed by a TCD gas chromato-graph.Yield of HCN was determined by the addition of NaOH,followed by titration with 0.01M AgNO3.Ammonia break-through was measured by titration of the HNO3scrubber with0.1N NaOH.3.Results and discussionsIn order to demonstrate the features of CFB riser reactor forthe ammoxidation of propylene to AN,hot model experimentswere made under different operation conditions:contact time,temperature,and feed ratio.The experiments with side feed ofair were also carried out.The results are discussed as follows.3.1.Effects of contact time on the product distributionProduct yield distributions are shown in Fig.2 as a functionof the contact time,W/F.Steep increases in both propylene con-version and AN yield are observed in the beginning stage ofreaction.A complete conversion is nearly reached at the contacttime longer than 125gcat.h/mol C3H6.WWH,the weight(kg)of reacted propylene per kilogram catalyst per hour,is usuallyused to represent the efficiency of catalyst in a certain reac-tor.For the catalyst used in the experiments,WWH is 0.065 incommercial TFB reactors.The contact time of 125gcat.h/molC3H6is equivalent to 0.33 in WWH,indicating that the capac-Fig.2.Changesinyieldofproductswithcontacttime:P=0.10MPa,T=718K,air/C3=10.5,NH3/C3=1.15.Fig.3.Changes in yield of AN with temperature:P=0.10MPa,air/C3=10.5,NH3/C3=1.15,W/F=125gh/mol.ity of the catalyst in CFB is about four times higher than that incommercial TFBs.As can be seen in Fig.2,increasing contact time,AN yieldincreasessignificantlyatbeginningreachesamaximumandthendecreasesgradually.TheyieldsofbothCOxandHCN,however,increase continuously with contact time,indicating that AN canfurther be oxidized to HCN and COxunder long contact time.ANisintermediateproduct,severegasbackmixingdecreaseANselectivity and yield.3.2.Effect of reaction temperatureFig.3 plots the effect of temperature on AN yield.The high-est yield can be achieved within the range of 708718K.Underlower temperature,AN yield is low due to the slow reactionrate and thus,the low conversion level;under higher temper-ature,overoxidation causes the decrease of AN yield.On theother hand,COxproduction steadily increases as the reactiontemperature increases.HCN yield changes insignificantly withthe increase of temperature.3.3.Effect of feed ratioAir/C3is an important controlling factor in industrial pro-cesses.Theoretically,air/C3ratio of 7.5 is enough for the mainreaction to produce AN,in which 1.5mol O2reacts with 1molpropylene.Due to the existence of side reaction,air/C3ratiois usually maintained between 10 and 10.5 in commercial TFBreactors.Fig.4 presents the effect of air/C3on AN yield.Anoptimumratioof9.510isfound.Itisslightlylowerthanthatintheindustrialprocess.Thisispartlybecausetheregeneratedcat-alystcarriedsomeamountofoxygeninlatticetypeintotheriser.Less overoxidation of AN resulted also in a low consumption ofoxygen.Y.Hu et al./Chemical Engineering and Processing 46(2007)918923921Fig.4.Changes in yield of AN with air/C3:P=0.10MPa,T=718K,NH3/C3=1.15,W/F=125gh/mol.NH3/C3ratio is also important in the synthesis of AN.MoreNH3in the reacting gas promotes N-containing products,espe-cially AN,and impedes O-containing products like COxandACL.The monotonous increase of AN yield and decrease ofCOxyield can be seen in Fig.5.NH3/C3ratio of 1.11.2 ispreferred.High NH3/C3ratio will increase the operation costin the product separation.3.4.Comparison with commercial TFB reactorTwo type of Mo-Bi/Al2O3catalysts was used in theexperiments.The experimental results from the experimentalhigh-density riser reactor and the data from commercial TFBreactors are shown in Table 2.The results from a small bubbleFig.5.Change in yield of AN with NH3/C3:P=0.10MPa,T=718K,air/C3=10.5,W/F=125gh/mol.fluidized bed(BFB)reactor for the evaluation of type A cat-alyst activity,provided by catalyst manufacturer,is also givenin this table.The experimental results indicate that comparedto the commercial TFB reactors,the high-density riser reactorhas the following features(1)the operating gas velocity reaches2.23m/s,the throughput of reactants is increased by more thanfourtimes;(2)theefficiencyofthecatalysts,WWH,isincreasedbyfourtimes;(3)ANyieldisincreasedby3%forthetypeAcat-alystandby6.5%fortheusedtypeBcatalyst,andtheproductionof COxare obviously decreased.The increases in reactor capac-ity and in WWH are mainly due to better gassolid contact inriserunderhigh-densitycondition,fullyregenerationofcatalystin regenerator and the higher concentration of reactant gases intheinletoftheriserreactor.TheincreaseinANyieldistheresultTable 2Comparison with commercial TFB reactorCatalyst(type A)Catalyst(used type B)HDCFB(reactor)TFB(reactor)BFB(reactor)HDCFB(reactor)TFB(reactor)T(K)716718718724722P(MPa)0.0810.050.050.050.05Air/C39.5210.59.710.08NH3/C31.191.151.121.03U(m/s)2.290.500.012.990.51Yield(%)AN83.3580.3376.875.268.68ACL0.310.10.00.79ACN2.793.081.592.48HCN5.665.9110.984.21COx6.9810.2111.5120.79X99.0999.6399.2896.95Balance in C1.041.08Balance in O0.990.98WWH0.3490.0650.4510.065922Y.Hu et al./Chemical Engineering and Processing 46(2007)918923Fig.6.Variations of AN yield with the fraction of oxygen side feed.ofnearlyplugflowintheriserunderhighergasvelocitythanthatin TFB.3.5.Experiments on staged air feedingEssentialplugflowandthustheexistenceofgradientofreac-tant concentration along riser make it possible to optimize thefeed policy and to get high selectivity of the desired product.Low concentration of oxygen along riser will reduce oxida-tion and favor ammoxidation,and thus will increase AN yield.However,considering that too low O2/C3ratio will cause over-reduction of catalyst to lost its activity,only 1030%of totalair was introduced through one side inlet at 10m of the riser inthe experiments.In order to maintain similar state of flow in theexperimentsunderdifferentsidefeedratio,theairfromsideinletwas replaced by 79%N2+21%O2and only O2was introducedfromthesideinletwhileN2enteredtheriseratthebottomoftheriser.Fig.6 shows the results of side oxygen feed experiments.Whentheexperimentsonsidefeedwerecarriedout,thecatalysthad been run for several months and had experienced too manytimes of the rise and drop in temperature due to the start andstop of experiment.The catalyst activity was not so good as atthe beginning,and AN yield under no side feed condition wasonly about 75%,as shown in Fig.6.Nevertheless,the increaseofnearly5%inANyieldwasobtainedwhen30%oxygenasthesidefeed,confirmingtheeffectivenessofstagedoxygenfeedingstrategy.4.ConclusionsHotmodelexperimentsweredoneonalaboratory-scalehigh-density CFB reactor with two kinds of Mo-Bi/?-Al2O3catalystundervariousoperationconditions.Theoptimumoperationcon-ditionsforhigh-densityriserreactoraretemperature708718K,air/propyleneratio9.5,NH3/propyleneratio1.11.2andcontacttime 125gh/mol C3H6.Compared with the commercial turbu-lent fluidized bed reactor,a high-density riser reactor has theadvantages including that(1)gas velocity reaches 2.23m/s,and the throughput of CFB reactor is increased more than fourtimes;(2)the efficiency of catalyst,WWH,is increased by fourtimes;(3)TheyieldsofANisincreasedby3%fortypeAcatalystand by 6.5%for type B catalyst,and the yield of COxis obvi-ouslydecreased.Stagedoxygenfeedingcanfurtherpromotetheincrease of AN yield.AcknowledgementsFinancial support from Petrochemical Company of China isgratefully acknowledged.The authors are also grateful to Pro-fessor Zhanwen Wang and Professor Zhiqing YU for their veryuseful discussion and to Mr.Hongwei Dian,Xiaotao Wan andYanhui Yang for their help on the experiments.Appendix A.NomenclatureACAacrylic acidACLacroleinACNacetonitrileair/C3mole ratio of air and C3H6(dimensionless)ANacrylonitrileCOxCO2+CONH3/C3mole ratio of NH3and C3H6(dimensionless)Ppressure at the exit of riser(MPa)Treaction temperature(K)Ugas velocity in riser(m/s)W/Fcontact time on the basis of C3H6,gcat.h/mol C3H6WWHthe weight(kg)of reacted propylene per kilogram cat-alyst per hourXconversion ratio of propyleneReferences1 S.P.Lankhuyzen,P.M.Florack,H.S.van der Bean,The catalytic ammoxi-dation of propylene over Bismuth Molybdate catalyst,J.Catal.42(1976)20.2 J.R.Krebs,Fluidized solids contacting apparatus,US Patent 2,893,849(1959).3 C.C.Georgian,L.M.Tex.,Powered catalyst contacting unit,US Patent2,893,851(1959).4 J.L.Callahan,E.C.Milberger,Proceeforpreparingolefinicallyunsaturatednitriles,US Patent 3,230,246(1966).5 J.L.Callahan,E.C.Milberger,Proceeforpreparingolefinicallyunsaturatedaldehydes and nitriles,US Patent 3,427,343(1969).6 H.R.Sheely,Reactor staging,US Patent 3,783,528(1974).7 T.Nakamura,H.Arai,H.Inaba,H.Yamamoto,Process for producingacrylonitrile,European Patent 0,842,922 A1(1998).8 B.H.Chen,Q.L.Dai,D.W.Lu,Development and modeling of a loop flu-idized bed reactor with baffle for propylene ammoxidation,Chem.Eng.Sci.51(11)(1996)1983.9 V.R.Choudhary,T.V.Choudhary,Method for gassolid contacting in abubble fluidized bed reactor,US Patent 7,022,307 B2(2006).Y.Hu et al./Chemical Engineering and Processing 46(2007)91892392310 F.Wei,F.B.Lu,Y.Jin,Z.Q.Yu,Mass flux profiles in a high-densitycirculating fluidized bed,Powder Technol.91(1997)189.11 F.Wei,H.F.LIN,Y.Cheng,Z.W.Wang,Y.Jin,Profiles of particle velocityandsolidsfractioninahigh-densityriser,PowderTechnol.100(23)(1998)183.12 L.Zhou,W.P.Dennler,A.R.Oroskar,B.V.Vora,H.Abrevaya,L.O.Stine,Reaction process in hybrid reactor for propylene ammoxidation,US Patent6,143,915(2000).13 H.T.Bi,J.X.Zhu,Static instability analysis of circulating flu-idized beds and concept of high-density risers,AIChE J.39(1993)1272.14 J.X.Zhu,H.T.Bi,Distinction between low density and high density circu-lating fluidized beds,Can.J.Chem.Eng.73(1995)644.15 J.R.Grace,A.S.Issangya,D.Bai,Situating the high-density circulatingfluidized bed,AIChE J.45(10)(1999)2108.