research article evaluation of the parameters and...

Research ArticleEvaluation of the Parameters and Conditions ofProcess in the Ethylbenzene Dehydrogenation with Applicationof Permselective Membranes to Enhance Styrene Yield

Paulo Jardel P Arauacutejo1 Manuela Souza Leite2 and Teresa M Kakuta Ravagnani3

1School of Petroleum Engineering Tiradentes University 300 Murilo Dantas Avenida 49032-490 Aracaju SE Brazil2Institute for Research and Technology (ITP) Tiradentes University (UNIT) 300 Murilo Dantas Avenida49032-490 Aracaju SE Brazil3College of Chemical Engineering State University of Campinas 500 Albert Einstein Avenida 13083-852 Campinas SP Brazil

Correspondence should be addressed to Manuela Souza Leite slmanuelagmailcom

Received 9 November 2015 Revised 1 February 2016 Accepted 9 February 2016

Academic Editor Antal Tungler

Copyright copy 2016 Paulo Jardel P Araujo et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Styrene is an important monomer in the manufacture of thermoplastic Most of it is produced by the catalytic dehydrogenationof ethylbenzene In this process that depends on reversible reactions the yield is usually limited by the establishment ofthermodynamic equilibrium in the reactor The styrene yield can be increased by using a hybrid process with reaction andseparation simultaneously It is proposed using permselective composite membrane to remove hydrogen and thus suppress thereverse and secondary reactionsThis paper describes the simulation of a dehydrogenation process carried out in a tubular fixed-bedreactor wrapped in a permselective composite membrane A mathematical model was developed incorporating the various masstransport mechanisms found in each of themembrane layers and in the catalytic fixed bedThe effects of the reactor feed conditions(temperature steam-to-oil ratio and the weight hourly space velocity) the fixed-bed geometry (length diameter and volume) andthe membrane geometry (thickness of the layers) on the styrene yield were analyzedThese variables were used to determine exper-imental conditions that favour the production of styrene The simulation showed that an increase of 4098 in the styrene yieldcompared to a conventional fixed-bed process could be obtained by wrapping the reactor in a permselective composite membrane

1 Introduction

Styrene is an important intermediate product in the petro-chemical industryThe commercial demand for styrene growsat 6 per year This popularity is mainly due to its recyclablecharacter which is not shared by other thermoplastics About90 of the total production of styrene is based on the cat-alytic dehydrogenation of ethylbenzene (EB) This reactionis reversible and endothermic and the maximum conversionof EB under real process conditions is limited by thermody-namic equilibrium to about 45 The industrial process isconducted at a high temperature usually between 550 and650∘CThis temperaturemust be carefully controlled to avoidthermal stress on the catalyst which can cause irreversibledamage and most importantly the loss of selectivity by the

formation of coke [1ndash4] The operating pressure is usuallyatmospheric or lower and a catalyst of Fe

2O3 associated with

other metallic oxides is used [5ndash7]To supply the high energy load consumed by the highly

endothermic reaction of dehydrogenation a great quantity ofsteam is used in the reactor feed It was shown that increasein styrene production over that of the industrial conventionalfixed-bed unit is achievable if design operating parametersand catalysts are properly chosen [8ndash10]

Recently with improvements in the catalyst compositionthe ratio of steam to oil (EB) in the feed has fallen consider-ably from 20 to 6 The kinetic model of the dehydrogenationof ethylbenzene used here is based on the presence of tenspecies involving six stoichiometrically independent linearreactions The values of the corresponding reaction rate

Hindawi Publishing Corporatione Scientific World JournalVolume 2016 Article ID 4949183 7 pageshttpdxdoiorg10115520164949183

2 The Scientific World Journal

constants were given by Abdalla and Elnashaie [11] whoobtained them by the adjustment of experimental data Thereaction generating styrene is reversible When the concen-tration of hydrogen in the reactor is decreased the equilib-rium will be pulled to the right increasing the conversion ofEB and the selectivity to styrene since the rate of conversionto toluene will be slower Therefore for separating hydrogenmembrane reactor can help the thermodynamic equilibriumto shift favourably in dehydrogenation reactions [12ndash14]The potential for enhancement of styrene yield by catalystimprovement appears to be limited due to the fact that themain bottleneck is related to the thermodynamic equilib-rium Therefore great efforts are needed to improve the per-formance through process design modification and mem-brane separation technology has been intensively investigatedin both modeling and experimental studies [9 12 15ndash21]

This paper proposed the removal of hydrogen from thecatalytic medium through a permselective composite mem-braneThe driving force for removal of hydrogen is a gradientof the chemical potential of hydrogen between retentate andpermeate sides To create this gradient a pressure drop wasmaintained across the membrane wall It presents an analysisof the parameters that control the performance of the fixed-bed reactor surrounded by a permselectivemembrane in thedehydrogenation of EB In a process simulation the operationconditions fixed-bed geometry andmembrane geometry areassessed Such variables were used to optimize the yield ofstyrene which is related to the conversion and the selectivityto the primary reaction The demand for higher conversionof ethylbenzene high yield and selectivity of the desiredreaction products especially styrene had led to new ingeniousconfiguration and design of reactors for the dehydrogenationprocess

2 Mathematical Modeling

For the simulation a computer program was developed inFortran The kinetic model mentioned above was used inwhich the rate constants were multiplied by an empiricalfactor observed in a commercial styrene catalyst providedby Stid-Chemie AG Munchen Germany These constantsrepresent the effective rather than the intrinsic kinetics inthat the influence of diffusion is built into the rate constantsaccording to a pseudo-homogeneous model of plug flow

The configuration of the reactor consists in catalyticfixed bed filled with the commercial 3mm styrene catalystwrapped by a membrane which is compound of a stainlesssteel macroporous support the microporous layer and apalladium thin film

It is assumed that there is no resistance to heat transferacross the membrane around the reactor

21 Axial Transport On the reaction side the axialmolar flowrate of species 119894 is controlled by its reaction rate and its rate oftransport through the membrane

119889119877

119894

119889119911= 119860119877

120588119887

119899

sum

119895=1

V119894119895119903119895minus 2120587119877

119877119869119894

10038161003816100381610038161003816100381610038161003816100381610038161003816119877119877

(1)

In adiabatic conditions 119889119879119889119911 can be calculated by theenthalpy changes of the system The pressure gradient(119889119875119889119911) through the catalytic bed is calculated by the Ergunequation The axial profiles of concentration temperatureand pressure through the fixed bed (reaction side) areobtained by integrating this equation by the fourth-orderRunge-Kutta method

22 Radial Transport The radial mass transport involvesdifferent transport mechanisms for the various membraneregions

221 Stagnant Gas Film It is supposed that a stagnant filmof gas is formed next to the membrane wall exhibitingmulticomponent diffusion Stefan-Maxwellrsquos equation is usedto describe the concentration profile of species across thisstagnant film Temperature and pressure are assumed to beconstant within the stagnant film

222 Macroporous Support The dusty gas model was usedto describe the profile in the macroporous support Thismodel assumes multicomponent diffusion of the molecularand Knudsen types and also viscous flow contributionsAssuming that the mass transport through the macroporoussupport is not due to the pressure gradient but to theconcentration gradient the dusty gas model for an idealgaseous mixture is given by

119889119910119894

119889119903=R119879119877

119875(

119899

sum

119895=1

119869119895119910119894minus 119910119895119869119894

(120576120591)119863119877

119894119895

minus119869119894

119863119890

119894119870

) (2)

where 119863 Knudsenrsquos effective diffusion coefficient of Fullerrsquoscorrelation was used to determine the binary diffusivity 119863119877

119894119895

for the components of the process

223 Microporous Layer It is assumed that Knudsenrsquos diffu-sion [22] controls the flux of the various species through themicroporous layer

224 Palladium Film As the palladium film is permeableonly to hydrogen the boundary condition [23] is applied tothe other components in the membrane wall To describe thetransport of hydrogen through the palladium film Sievertrsquoslaw was used (3)

It is assumed that the membrane is inert and operates ina stationary state so that 119869

119894is constant through the various

layers After discretization by finite differences and imposingthe limit conditions for the concentration of each species onthe reaction side the system of equations was solved by thegeneralized Newton-Raphson method

119869H2 =119863H21198620

11987512

0

ln (119903ex119903in)1

119903[(119875)12

1198981

minus (119875)12

1198982

] (3)

3 Simulations and Results

The styrene yield (percent ethylbenzene converted to styrene)was used as ameasure of the performance of the process Such

The Scientific World Journal 3

process was conducted by removal of the hydrogen producedin the primary reaction This removal was established bypressure drop of 199 kPa applied across the membrane fromretentate to permeate side of the reactor This way it waspossible to determine the operating conditions and optimalconfiguration of the reactor and the membrane throughsimulations to assess their influence on the processThe otheroperating ranges adopted are in Table 1 as well as the rangesof variation adopted for the configuration of the systemproposed

Each range was chosen on the basis of data from indus-trial operation and the restrictions of the process Dataof a commercial catalyst were used as a reference for thecatalytic properties The membrane properties were basedon probable values for the types of material included in themembrane (stainless steel with palladium) These propertiesare described in Table 2

For comparison a simulation of the proposed model wasrun for a conventional fixed bed operating under standardconditions of industrial process [2] resulting in a styreneyield of 5276

31 Analysis of Operating Conditions

311 Energy Available to the System Among the processconditions analyzed the temperature and amount of steamare the parameters that reflect the energy available to bringabout the reaction To analyze these the systemwas simulatedwhile they were varied within the ranges in Table 1 tryingto meet the energy requirements for the process withoutdegrading the equipment The molar ratio (steamEB) variesindustrially between 6 and 12

The steamEB ratio does not have to be the highestvalue tested in the analysis (Figure 1) but it (and the inlettemperature of the reactant) should be high enough to supplythe energy needs of the reaction and the ranges analyzedmust not result in any damage to the equipment or to theproducts

The best styrene yield was obtained at the highest valuesof temperature and steam-to-oil ratio tested which were95315 K and 12 As a rise in temperature leads to an increasedrate of hydrogen permeation through the permselective Pdmembrane the increased production of styrene at highertemperatures may be attributed to the removal of hydrogenfrom the reaction medium

312 Weight Hourly Space Velocity (WHSV) To concludethe simulation of the influence of operating conditions onthe system a factor associated with the reaction velocitywas analyzed namely the WHSV over the range given inTable 1 The maximum yield was observed for the minimumspace velocity (10 hminus1) This behaviour was expected sincethis parameter is the ratio of the feed rate of reactants (kgh)to the mass of catalyst (kg) in the reactor This proportionincreases when either the reactants flow more quickly orthe catalyst mass decreases In either case there is reducedcontact between the reactant and the catalyst resulting in alower styrene yield Once theWHSVwas fixed the fixed-bed

Table 1 Ranges of variation of the operating conditions and systemconfiguration

Operating conditionsInlet temperature 87315ndash95315 KSpace velocity 1ndash16 hminus1

Steam-to-oil ratio 6ndash12Reactor configuration

Length 039ndash12mInner diameter 3410158401015840ndash210158401015840

Equivalent diameter 5mmMembrane configuration

Thickness (macro and micro) 1ndash3 and 10ndash30mmThickness (Pd) 1ndash20 120583m

Table 2 Properties of the catalyst and membrane

Properties of the catalystPorosity 05Density 2150 kgsdotmminus3

Diameter 3mmProperties of the membrane

Porosity (macro and micro) 05Tortuosity (macro and micro) 3Pore diameter (macro) 02 120583m

68

1012

880900

920940

9604550556065707580

Steam-to-oil ratio (SO)Temperature (K)

Styr

ene y

ield

()

Figure 1 Influence of inlet temperature and steam-to-oil ratio onstyrene yield

geometry became directly responsible for the flow of thereactor feed

32 Analysis of the Fixed-Bed Geometry To optimize theperformance of the process it is necessary to analyze thegeometry of the fixed bed and the membrane that wraps it Inthis section the influence of the reactor length and the fixed-bed inner diameter will be examined


321 Inner Diameter times Reactor Length As the radial dis-persion in the fixed bed was not considered in this study itwas necessary to restrict this analysis to cases in which thereactor inner diameter is at least 8 (eight) times bigger thanthe diameter of the catalyst particle The reactor used has acatalytic fixed bed filled with pellets of 3mm (catalyst parti-cle) in diameter of a catalyst commercial dehydrogenation ofethylbenzene The fixed bed was filled with a permselectivemembrane and a thin layer of palladium

In this simulation the reactor inner diameters testedwere those available on the market for the selected material(stainless steel) while the shortest possible reactor length waschosen because thematerial cost is decisive for the viability ofa project As can be seen in Figure 2 a smaller inner diameterresults in a higher styrene yield but since radial dispersion isneglected one must choose the smallest diameter satisfyingthe above restriction (00254m)

It was also observed that in the range of reactor lengthsanalyzed the styrene yield varies little from 7237 at 039mto 7271 at 12m This range of values is not sufficient todefine an optimal reactor length Therefore it was decidedto analyze the fixed-bed geometry at a constant reactant feedflow rate instead of a predetermined space velocity

322 Analysis of the Reactor Volume at Constant Feed RateThis analysis was carried out with a constant reactor innerdiameter of 00254m so that the reactor volume was variedby changing its length Adopting a fixed mass flow at theentrance to the reactor of 908 times 10minus5 kgsdotsminus1 determinedfrom the data in Tables 1 and 2 simulations were run withreactor lengths from 04 to 12m corresponding to volumesfrom 19762 to 60805 cm3 and the styrene yield gain (relativeto the shortest reactor) was calculated for each volumeAccording to Figure 3 the ideal reactor volume that satisfiesthe material restrictions on the process and minimizes thecost of implementation is 30402 cm3 That represents areactor length of 06m for an inner diameter of 110158401015840 (00254m)

33 Analysis of Membrane Geometry

331 Thickness of the Macroporous Layer A thinner layeris expected to improve styrene yield However bearing inmind the membrane manufacturing techniques availableon the market today the tested range of thickness of themacroporous layer was selected from values available in theliterature [24]

In the simulation presented in Figure 4 as expected thebest value of styrene yield was obtained (7270) with thethinnest macroporous layer Its thickness (10minus3m) can beachieved by the technique of bimetallic multilayer deposition[25] without making the macroporous layer lose its capacityto support the membrane mechanically

332 Thickness of the Microporous Layer Studies of perms-elective composite membranes usually fail to distinguish thethickness of the macroporous from that of the microporouslayer but as the latter affects directly the flux of speciesthrough the membrane it was analyzed separately here

002003

004005

006 0406

081

1260

65

70

75

80

Reactor length (m)

Inner diameter of fixed bed (m)

Styr

ene y

ield

()

Figure 2 Analyses of reactor geometry

0005101520253035404550

15000 25000 35000 45000 55000 65000

Incr

ease

in th

e sty

rene

yie

ld (

)

Reactor volume (cm3)

Figure 3 Influence of reactor volume on styrene yield at a constantmass flow

The effect of the thickness of the microporous layeron yield was not significant As it rose from 10 to 30 120583m(threefold thicker) styrene yield decreased almost linearlyfrom 7270 to 7263 as shown in Figure 4 As expectedthe best styrene yield (727) was achieved with the thinnestmicroporous layer which allows a slightly greater flow ofhydrogen through themembrane since there is a shorter routeto permeate Therefore a 10 120583m thick microporous layer maybe recommended for this system

333 Thickness of the Dense Metal Layer In the transportof different species through the membrane the stage withthe greatest influence on hydrogen removal from the reac-tion medium is the dense metal layer In dehydrogenationreactions palladium is used in this layer in view of its goodpermselectivity to hydrogen The thickness of the palladiumlayer chosen for the simulation ranged from the thinnest(10 120583m) available with modern membrane manufacturingtechniques [25] up to thicknesses already studied in theliterature

The simulation results indicate that the effect of reductionof the dense metal layer thickness in the range from 10 to


Macroporous layerMicroporous layer

Pd-film6900700071007200730074007500

Styr

ene p

rodu

ctiv

ity (

)

Growth thickness of layerMacroporous layerMicroporous layer

Pd-film

Growth thickness f l

Figure 4 Influence of thickness of each membrane layer on styreneyield

200120583m (twenty times bigger) resulted in a styrene yield of3 approximately Hence the thinner the palladium layerthe better the styrene yield This is ideal for the applicationof palladium composite membranes in industrial processesgiven the very high cost of this metalThese results reflect thefact that this layer imposes greater resistance to flow throughthe membrane

Thus in light of the simulation results and the restrictionson the manufacture of the membranes mentioned abovethe dense metal layer thickness recommended for the bestperformance is 10 120583m

34 Analysis of the Thickness of the Stagnant Gas LayerBesides analyzing the thickness of the membrane layers it isimportant to consider the stagnant gas layer next to themem-braneThe thickness of this stagnant gas layer depends on thefixed-bed geometry and the feed conditions (compositiontemperature and pressure) The proposed model assumesthat within the reactor bed resistance to mass transfer in theradial direction can only be found in the stagnant gas layerThus the internal radius of the fixed bed must be higher thanthe calculated thickness of the stagnant gas layer As can beseen in Figure 5 nowhere does the thickness of the stagnantgas layer approach the internal radius of the fixed bed sothe application of the film theory is valid showing a centralregion that allows gas to flow through the fixed bed (withinthe bulk gas phase) and a surrounding region where the flowmeets resistance (stagnant gas)

35 Profile of Hydrogen through the Membrane The simu-lated radial profile of partial pressure of hydrogen through thelayers around the reactor is plotted for various positions alongthe reactor in Figure 6 Note that the different membranelayers are not shown to scale (the Pd layer is 1000 timesthinner than the macroporous one)

The largest drop in the partial pressure of hydrogen isacross the stagnant gas and macroporous layers Such layersprovide high rates of flow of the hydrogen permeating themembrane

Analyzing the difficulty of transport in the microporouslayer one can assume this is partly related to the pore diame-ter adopted in the simulation but mainly to the accumulationof components due to the phenomenon of polarization ofthe concentration next to this layer As far as the transport

0

0005

001

0015

002

0 01 02 03 04 05 06

Stag

nant

gas

thic

knes

s (m

)

Reactor length (m)

Internal radius (fixed bed) 00127m

Figure 5 Profile of the stagnant gas layer in the fixed bed

050

100

150

200

250

300

350

400

450

500

0 1 2 3 4 5 6 7 8 9Membrane discretization

Stagnant gas

Mac

ropo

rous

Mic

ropo

rous

Pd-fi

lm

Part

ial p

ress

ure o

fH2

(Pa)

Zf = 001mZf = 005m

Zf = 030mAf = 060m

Figure 6 Profile of partial pressure of hydrogen in radial flowthrough the different layers

through the palladium is concerned although it is totallypermeable only to hydrogen this permeability is quite smallTo allow an increase in this flux it would be necessary toreduce the resistance to flow imposed by the outer layerswhich means altering their properties However within theexisting restrictions on the forms of manufacture of thechosen material (stainless steel) and on the flux imposed bythe actual phenomenon of hydrogen transport through thePd the configuration adopted gives the best yield for thissimulation

4 Conclusion

Using the styrene yield as a criterion to assess systemperformance the simulations carried out enabled both theoperating conditions and the reactor and membrane dimen-sions that provided the best styrene yield to be determined

By simulating the process with the recommended valuesfor each parameter the styrene yield was raised to 4098higher than the yield achieved with a conventional fixed


bed (5276) This improvement demonstrates the impor-tance of determining the favourable conditions for processdevelopment Analyzing the hydrogen profile across thedifferent layers of the membrane it can be concluded that themicroporous and palladium film layers prevent the furtherincrease of the styrene yield due mainly to the polarizationof the hydrogen concentration next to the microporouslayer However this inconvenience cannot be overcome withthe configuration adopted for the membrane as currentmanufacturing techniques restrict the characteristics of theselayers

Competing Interests

The authors declare that they have no competing interests

References

[1] S S E H Elnashaie T Moustafa T Alsoudani and S SElshishini ldquoModeling and basic characteristics of novel inte-grated dehydrogenation-hydrogenation membrane catalyticreactorsrdquo Computers and Chemical Engineering vol 24 no 2ndash7pp 1293ndash1300 2000

[2] C Hermann P Quicker and R Dittmeyer ldquoMathematicalsimulation of catalytic dehydrogenation of ethylbenzene tostyrene in a composite palladium membrane reactorrdquo Journalof Membrane Science vol 136 no 1-2 pp 161ndash172 1997

[3] Y H Ma B C Akis M E Ayturk F Guazzone E E Eng-wall and I P Mardilovich ldquoCharacterization of intermetallicdiffusion barrier and alloy formation for PdCu and PdAgporous stainless steel composite membranesrdquo Industrial andEngineering Chemistry Research vol 43 no 12 pp 2936ndash29452004

[4] G Sznejer and M Sheintuch ldquoApplication of a carbon mem-brane reactor for dehydrogenation reactionsrdquo Chemical Engi-neering Science vol 59 no 10 pp 2013ndash2021 2004

[5] A Makaruk M Miltner and M Harasek ldquoMembrane systemsfor the recovery of hydrogen from multicomponent gas mis-tures obtained in biomass gasificationrdquo Chemical EngineeringTransactions vol 25 pp 833ndash838 2011

[6] R Dittmeyer V Hollein and K Daub ldquoMembrane reactors forhydrogenation and dehydrogenation processes based on sup-ported palladiumrdquo Journal of Molecular Catalysis A Chemicalvol 173 no 1-2 pp 135ndash184 2001

[7] Y L Becker AGDixonWRMoser andYHMa ldquoModellingof ethylbenzene dehydrogenation in a catalytic membranereactorrdquo Journal of Membrane Science vol 77 no 2-3 pp 233ndash244 1993

[8] E Gobina K Hou and R Hughes ldquoMathematical analysisof ethylbenzene dehydrogenation comparison of microporousand dense membrane systemsrdquo Journal of Membrane Sciencevol 105 no 3 pp 163ndash176 1995

[9] M E E Abashar ldquoCoupling of ethylbenzene dehydrogenationand benzene hydrogenation reactions in fixed bed catalyticreactorsrdquo Chemical Engineering and Processing vol 43 no 10pp 1195ndash1202 2004

[10] M M Hossain L Atanda N Al-Yassir and S Al-KhattafldquoKineticsmodeling of ethylbenzene dehydrogenation to styrene

over a mesoporous alumina supported iron catalystrdquo ChemicalEngineering Journal vol 207-208 pp 308ndash321 2012

[11] B K Abdalla and S S E H Elnashaie ldquoCatalytic dehydrogena-tion of ethylbenzene to styrene in membrane reactorsrdquo AIChEJournal vol 40 no 12 pp 2055ndash2059 1994

[12] C Yu and H Xu ldquoAn efficient palladium membrane reactor toincrease the yield of styrene in ethylbenzene dehydrogenationrdquoSeparation and Purification Technology vol 78 no 2 pp 249ndash252 2011

[13] B N Lukyanov D V Andreev and V N Parmon ldquoCatalyticreactors with hydrogen membrane separationrdquo Chemical Engi-neering Journal vol 154 no 1ndash3 pp 258ndash266 2009

[14] M H Khademi M R Rahimpour and A Jahanmiri ldquoDif-ferential evolution (DE) strategy for optimization of hydro-gen production cyclohexane dehydrogenation and methanolsynthesis in a hydrogen-permselective membrane thermallycoupled reactorrdquo International Journal of Hydrogen Energy vol35 no 5 pp 1936ndash1950 2010

[15] M E E Abashar ldquoCoupling of ethylbenzene dehydrogenationand benzene hydrogenation reactions in fixed bed catalyticreactorsrdquo Chemical Engineering and Processing Process Intensi-fication vol 43 no 10 pp 1195ndash1202 2004

[16] C Kong J Lu J Yang and JWang ldquoCatalytic dehydrogenationof ethylbenzene to styrene in a zeolite silicalite-1 membranereactorrdquo Journal of Membrane Science vol 306 no 1-2 pp 29ndash35 2007

[17] N S Abo-Ghander J R Grace S S E H Elnashaie and C JLim ldquoModeling of a novel membrane reactor to integrate dehy-drogenation of ethylbenzene to styrene with hydrogenation ofnitrobenzene to anilinerdquo Chemical Engineering Science vol 63no 7 pp 1817ndash1826 2008

[18] B Xiang C Yu H Xu and W Li ldquoDehydrogenation ofethylbenzene in the presence of CO

2

over v catalysts supportedon nano-sized aluminardquo Catalysis Letters vol 125 no 1-2 pp90ndash96 2008

[19] E Pourazadi D IranshahiM R Rahimpour and A JahanmirildquoIncorporating multi-membrane tubes for simultaneous man-agement of H

2

HC and hydrogenation of nitrobenzene to ani-line in naphtha heat exchanger reactorrdquo Chemical EngineeringJournal vol 184 pp 286ndash297 2012

[20] CNederlof V Zarubina I VMelian-Cabrera EH J Heeres FKapteijn and M Makkee ldquoApplication of staged O

2

feeding inthe oxidative dehydrogenation of ethylbenzene to styrene overAl2

O3

and P2

O5

SiO2

catalystsrdquo Applied Catalysis A Generalvol 476 pp 204ndash214 2014

[21] M Farsi A Jahanmiri and M R Rahimpour ldquoOptimalconditions of isobutane dehydrogenation in radial flowmovingbed hydrogen-permselective membrane reactors to enhanceisobutene and hydrogen productionrdquoChemical Engineering andProcessing Process Intensification vol 75 pp 126ndash133 2014

[22] A Makaruk M Miltner and M Harasek ldquoMembrane gaspermeation in the upgrading of renewable hydrogen frombiomass steam gasification gasesrdquoAppliedThermal Engineeringvol 43 pp 134ndash140 2012

[23] J G Palencia and E M Verruschi ldquoComparison of a modelfixed-bed bioreactor plating a model of N-reactors in seriesperfectly blended and a tubular reactor model and simulationmodelsrdquo Chemical Engineering Transactions vol 27 pp 379ndash384 2012


[24] K Shashi S Shankar P R Shah and K Surendra ldquoA com-prehensivemodel for catalyticmembrane reactorrdquo InternationalJournal of Chemical Reactor Engineering vol 4 no 1 pp 1ndash272006

[25] Y Ma F Liu M Zhu and Z Zhang ldquoAligned microcrystallinesilicon nanorods prepared by glancing angle hotwire chemicalvapor deposition for photovoltaic applicationsrdquo Physica StatusSolidi (C) vol 7 no 3-4 pp 537ndash540 2010

Submit your manuscripts athttpwwwhindawicom

Computer Games Technology

International Journal of

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Distributed Sensor Networks


Advances in

FuzzySystems

Hindawi Publishing Corporationhttpwwwhindawicom

Volume 2014


ReconfigurableComputing

Hindawi Publishing Corporation httpwwwhindawicom Volume 2014


Applied Computational Intelligence and Soft Computing

thinspAdvancesthinspinthinsp

Artificial Intelligence

HindawithinspPublishingthinspCorporationhttpwwwhindawicom Volumethinsp2014

Advances inSoftware EngineeringHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Electrical and Computer Engineering

Journal of

Journal of

Computer Networks and Communications


Hindawi Publishing Corporation

httpwwwhindawicom Volume 2014

Advances in

Multimedia


Biomedical Imaging


ArtificialNeural Systems

Advances in


RoboticsJournal of



Computational Intelligence and Neuroscience

Industrial EngineeringJournal of


Modelling amp Simulation in EngineeringHindawi Publishing Corporation httpwwwhindawicom Volume 2014

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


Human-ComputerInteraction

Advances in

Computer EngineeringAdvances in



constants were given by Abdalla and Elnashaie [11] whoobtained them by the adjustment of experimental data Thereaction generating styrene is reversible When the concen-tration of hydrogen in the reactor is decreased the equilib-rium will be pulled to the right increasing the conversion ofEB and the selectivity to styrene since the rate of conversionto toluene will be slower Therefore for separating hydrogenmembrane reactor can help the thermodynamic equilibriumto shift favourably in dehydrogenation reactions [12ndash14]The potential for enhancement of styrene yield by catalystimprovement appears to be limited due to the fact that themain bottleneck is related to the thermodynamic equilib-rium Therefore great efforts are needed to improve the per-formance through process design modification and mem-brane separation technology has been intensively investigatedin both modeling and experimental studies [9 12 15ndash21]

This paper proposed the removal of hydrogen from thecatalytic medium through a permselective composite mem-braneThe driving force for removal of hydrogen is a gradientof the chemical potential of hydrogen between retentate andpermeate sides To create this gradient a pressure drop wasmaintained across the membrane wall It presents an analysisof the parameters that control the performance of the fixed-bed reactor surrounded by a permselectivemembrane in thedehydrogenation of EB In a process simulation the operationconditions fixed-bed geometry andmembrane geometry areassessed Such variables were used to optimize the yield ofstyrene which is related to the conversion and the selectivityto the primary reaction The demand for higher conversionof ethylbenzene high yield and selectivity of the desiredreaction products especially styrene had led to new ingeniousconfiguration and design of reactors for the dehydrogenationprocess

2 Mathematical Modeling

For the simulation a computer program was developed inFortran The kinetic model mentioned above was used inwhich the rate constants were multiplied by an empiricalfactor observed in a commercial styrene catalyst providedby Stid-Chemie AG Munchen Germany These constantsrepresent the effective rather than the intrinsic kinetics inthat the influence of diffusion is built into the rate constantsaccording to a pseudo-homogeneous model of plug flow

The configuration of the reactor consists in catalyticfixed bed filled with the commercial 3mm styrene catalystwrapped by a membrane which is compound of a stainlesssteel macroporous support the microporous layer and apalladium thin film

It is assumed that there is no resistance to heat transferacross the membrane around the reactor

21 Axial Transport On the reaction side the axialmolar flowrate of species 119894 is controlled by its reaction rate and its rate oftransport through the membrane

119889119877

119894

119889119911= 119860119877

120588119887

119899

sum

119895=1

V119894119895119903119895minus 2120587119877

119877119869119894

10038161003816100381610038161003816100381610038161003816100381610038161003816119877119877

(1)

In adiabatic conditions 119889119879119889119911 can be calculated by theenthalpy changes of the system The pressure gradient(119889119875119889119911) through the catalytic bed is calculated by the Ergunequation The axial profiles of concentration temperatureand pressure through the fixed bed (reaction side) areobtained by integrating this equation by the fourth-orderRunge-Kutta method

22 Radial Transport The radial mass transport involvesdifferent transport mechanisms for the various membraneregions

221 Stagnant Gas Film It is supposed that a stagnant filmof gas is formed next to the membrane wall exhibitingmulticomponent diffusion Stefan-Maxwellrsquos equation is usedto describe the concentration profile of species across thisstagnant film Temperature and pressure are assumed to beconstant within the stagnant film

222 Macroporous Support The dusty gas model was usedto describe the profile in the macroporous support Thismodel assumes multicomponent diffusion of the molecularand Knudsen types and also viscous flow contributionsAssuming that the mass transport through the macroporoussupport is not due to the pressure gradient but to theconcentration gradient the dusty gas model for an idealgaseous mixture is given by

119889119910119894

119889119903=R119879119877

119875(

119899

sum

119895=1

119869119895119910119894minus 119910119895119869119894

(120576120591)119863119877

119894119895

minus119869119894

119863119890

119894119870

) (2)

where 119863 Knudsenrsquos effective diffusion coefficient of Fullerrsquoscorrelation was used to determine the binary diffusivity 119863119877

119894119895

for the components of the process

223 Microporous Layer It is assumed that Knudsenrsquos diffu-sion [22] controls the flux of the various species through themicroporous layer

224 Palladium Film As the palladium film is permeableonly to hydrogen the boundary condition [23] is applied tothe other components in the membrane wall To describe thetransport of hydrogen through the palladium film Sievertrsquoslaw was used (3)

It is assumed that the membrane is inert and operates ina stationary state so that 119869

119894is constant through the various

layers After discretization by finite differences and imposingthe limit conditions for the concentration of each species onthe reaction side the system of equations was solved by thegeneralized Newton-Raphson method

119869H2 =119863H21198620

11987512

0

ln (119903ex119903in)1

119903[(119875)12

1198981

minus (119875)12

1198982

] (3)

3 Simulations and Results

The styrene yield (percent ethylbenzene converted to styrene)was used as ameasure of the performance of the process Such




















68

1012

880900

920940

9604550556065707580


Styr

ene y

ield

()













002003

004005

006 0406

081

1260

65

70

75

80

Reactor length (m)


Styr

ene y

ield

()


0005101520253035404550

15000 25000 35000 45000 55000 65000

Incr

ease

in th

e sty

rene

yie

ld (

)








Pd-film6900700071007200730074007500

Styr

ene p

rodu

ctiv

ity (

)


Pd-film









0

0005

001

0015

002

0 01 02 03 04 05 06

Stag

nant

gas

thic

knes

s (m

)

Reactor length (m)



050

100

150

200

250

300

350

400

450

500


Stagnant gas

Mac

ropo

rous

Mic

ropo

rous

Pd-fi

lm

Part

ial p

ress

ure o

fH2

(Pa)

Zf = 001mZf = 005m

Zf = 030mAf = 060m



4 Conclusion





Competing Interests


References




















2



2



2


O3

and P2

O5

SiO2















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in






















68

1012

880900

920940

9604550556065707580


Styr

ene y

ield

()













002003

004005

006 0406

081

1260

65

70

75

80

Reactor length (m)


Styr

ene y

ield

()


0005101520253035404550

15000 25000 35000 45000 55000 65000

Incr

ease

in th

e sty

rene

yie

ld (

)








Pd-film6900700071007200730074007500

Styr

ene p

rodu

ctiv

ity (

)


Pd-film









0

0005

001

0015

002

0 01 02 03 04 05 06

Stag

nant

gas

thic

knes

s (m

)

Reactor length (m)



050

100

150

200

250

300

350

400

450

500


Stagnant gas

Mac

ropo

rous

Mic

ropo

rous

Pd-fi

lm

Part

ial p

ress

ure o

fH2

(Pa)

Zf = 001mZf = 005m

Zf = 030mAf = 060m



4 Conclusion





Competing Interests


References




















2



2



2


O3

and P2

O5

SiO2















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in












002003

004005

006 0406

081

1260

65

70

75

80

Reactor length (m)


Styr

ene y

ield

()


0005101520253035404550

15000 25000 35000 45000 55000 65000

Incr

ease

in th

e sty

rene

yie

ld (

)








Pd-film6900700071007200730074007500

Styr

ene p

rodu

ctiv

ity (

)


Pd-film









0

0005

001

0015

002

0 01 02 03 04 05 06

Stag

nant

gas

thic

knes

s (m

)

Reactor length (m)



050

100

150

200

250

300

350

400

450

500


Stagnant gas

Mac

ropo

rous

Mic

ropo

rous

Pd-fi

lm

Part

ial p

ress

ure o

fH2

(Pa)

Zf = 001mZf = 005m

Zf = 030mAf = 060m



4 Conclusion





Competing Interests


References




















2



2



2


O3

and P2

O5

SiO2















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Pd-film6900700071007200730074007500

Styr

ene p

rodu

ctiv

ity (

)


Pd-film









0

0005

001

0015

002

0 01 02 03 04 05 06

Stag

nant

gas

thic

knes

s (m

)

Reactor length (m)



050

100

150

200

250

300

350

400

450

500


Stagnant gas

Mac

ropo

rous

Mic

ropo

rous

Pd-fi

lm

Part

ial p

ress

ure o

fH2

(Pa)

Zf = 001mZf = 005m

Zf = 030mAf = 060m



4 Conclusion





Competing Interests


References




















2



2



2


O3

and P2

O5

SiO2















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in





Competing Interests


References




















2



2



2


O3

and P2

O5

SiO2















Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in













Advances in

FuzzySystems


Volume 2014












Journal of

Journal of





Advances in

Multimedia


Biomedical Imaging



Advances in


RoboticsJournal of










Advances in



research article evaluation of the parameters and...

Documents