2007) 774780

Reactive distillation: The front-runner of indA full review of commercial appl

scale-up, design and opG. Jan Harmsen a,b

a Shell Global Solutions, Shell Research and Technology Center Amsterdam, P.O. Box 38000, 1030 BN Amsterdam, The Netherlandsb

Abstract

Most indureported incollaboratio

The businreduced byto the envirosociety impa

These indlarge rangesheterogeneo

The systemass flow, a

The scaleTechnolo

removed byand developidentify theexamples.

Academicpacking typdynamic sim

The rapidcan also be s 2007 Else

Keywords: ReCatalytic; Sus

E-mail ad

0255-2701/$doi:10.1016/jRijksUniversiteit Groningen, Nijenborgh 4, 9747 AG Groningen, The NetherlandsReceived 19 June 2007; accepted 20 June 2007

Available online 23 June 2007

strial scale reactive distillations (presently more than 150), operated worldwide today at capacities of 1003000 ktonnes/y, and arethis paper. Most of these plants started up less than 15 years ago. The drivers, processes, systems, scale-up methods and partnerns for this rapid invasion of a new process intensified technique are explained in this paper.ess drivers are (a) economical (prosperity): variable cost, capital expenditure and energy requirement reduction. In all cases these are20% or more, when compared to the classic set-up of a reactor followed by distillation. (b) Environmental (planet): lower emissionsnment. In all cases carbon dioxide and diffusive emissions are reduced and (c) social (people): improvements on safely, health andct are obtained by lower reactive content, lower run away sensitivity and lower space occupation.ustrial reactive distillation systems comprise homogeneous and heterogeneous catalysed, irreversible and reversible reactions, coveringof reactions, notably hydrogenations, hydrodesulfurisation, esterifications and etherification. Various commercial methods for packingus catalyst in columns are now available.ms comprise amongst others: multiple catalyst systems, gas and liquid internal recycle traffic over these catalyst systems, separation,nd enthalpy exchange. These are integrated optimally in a single vessel, a characteristic feature of process intensification.-up methods applied from pilot plants to commercial scale are brute force and modelling.gy providers CDTECH and Sulzer Chemtech have used these scale-up methods successfully. Barriers perceived and real have also beenthese companies. Chemical manufacturing companies have also developed their own specific reactive distillations by their own researchment. These companies, both on their own and in consortia, also developed heuristic process synthesis rules and expert software toattractiveness and technical feasibility of reactive distillation. Heuristic rules and expert software will be presented and supported by

research also produced design methods to identify the feasibility of reactive distillation, to determine the feed locations, to selectes, to sequence columns optimally and also produced methods to design, optimise and control the columns with steady state andulation models.commercial scale implementation of reactive distillation by co-operation of partners in research, scale-up, design and reliable operationeen as a model for rapid implementation of other process intensification techniques in the chemical industry.vier B.V. All rights reserved.

active distillation; Industrial; Process intensification; Multi-functional; Applications; Research; Scale-up; Design; Operation; Innovation; Stakeholders;tainable development; Triple P; Environment; Society; Economic

dress: jan.harmsen@shell.com.

1. Introduction

Reactive distillation, also called catalytic distillation, can beconsidered as reaction and distillation combined into one newunit operation. Distillation itself is here considered in the widesense, i.e. the separation by use of vapourliquid composition

see front matter 2007 Elsevier B.V. All rights reserved..cep.2007.06.005Chemical Engineering and Processing 46 (ustrial process intensificationications, research,eration

G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780 775

Table 1Catalytic distillation in commercial operation in 2006, licensed from CDTECH[1]Process Number

Ethers: MTBE, TAME, ETBE 69Hydrogenation of aromatics and light sulfur 50Hydrodesulfurisation 21Isobutylene production from C4 stream 3Ethyl benzene production 3

Total 146

difference. So it includes distillation columns, flashers, strippersand condensers. The reactions in reactive distillation consideredinclude heterogeneous catalysis reactions, homogeneous catal-ysis reactiocases reacting place inphase of th

In this ptillations odesign metto implemeapplied profor implemare derived

1.1. Reactoperation

CDTECprovider, licesses. Ofof 2006 [1Mid 2005of 79 [3],Sulzer repocations: synof methyl afrom formaregional spdisclose nu

Eastmanstarted upof about 4

Table 2commercial reChemtech [4]Process

AcetateSynthesis oHydrolysis

AcetalisationSynthesis oRemoval oFatty acid e

is made of five different materials of construction (ranging tozirconium). It has an annual capacity significantly in excess of200,000 metric tonnes per year methyl acetate [5]. The secondplant, built 7 years later with the benefit of operating experience,is virtually identical (geometry, staging, dimensions, capacity)except for some sieve tray detail in the upper non-reactive zones[5]. It contains five functions in one column resulting in capi-tal expenditure and energy reductions by a factor 5 relative toconventional unit operation design [6,7].

Except for the Eastman process the author knows of sev-eral other large-scale reactive distillations in the petrochemicalindustry, which are not disclosed to the public [8]. One wasstarted up as early as 1953. Combining the reported commer-cial applications with these the total number of commercialapplications exceeds 150.

otential process applications

dmacher has a whole chapter containing over a 100 dif-industrial applications [9]. However, most of these stemsatents and little information is given on their industrial sta-

H reactive distillations

o Selective hydrogenation using catalytic distillation: MAPD reduction in mixed C3s C4 diolefins reduction in mixed C4s

(hydroisomerisation option) C4 acetylenes reduction in mixed C4s C5 diolefins reduction in mixed C5s

(hydroisomerisation option) C6 diolefins reduction in mixed C6s Benzene reduction in LSR and reformate streams Mercaptan reduction in C4/C5/C6 olefinic streams Hydrogenation of benzene to produce cyclohexane

andDS+

Desulfurisation using catalytic distillation:

FCC gasoline Jet fuel/kerosene

cttSM

SM

e

rs

rol

ns, and thermal (non-catalyst) reactions. In nearly allions take place in the liquid phase, but reactions tak-

the gas phase and locate the catalyst in the vapoure column is conceivable.aper, we review the commercial scale reactive dis-n their application areas, advantages, scale-up andhods and roles of various partners in the developmentntation. As reactive distillation is the most widelycess intensification technique today, lessons learnedentation of other process intensification techniques.

ive distillation processes in commercial

H, the major commercial process technologycensed up to now over 200 commercial scale pro-these 146 are in commercial operation at the end]. The process applications are shown in Table 1.this number was 121 [2] and in 2002 the numberso the rate of implementation is still increasing.rts the following industrial commercial scale appli-thesis of ethyl, butyl and methyl acetates, hydrolysiscetate, synthesis of methylal, removal of methanolldehyde, formation of fatty acid esters [4] with theirread shown in as shown in Table 2. Sulzer does notmbers.s first methyl acetate reactive distillation towerin 1980. It is more than 80 m tall with a diameterm. The process (including condenser and reboiler)

active distillation applications with Katapak licensed from Sulzer

1.2. P

Sunferentfrom p

Table 3CDTEC

CDHydr

CDHDSCDH

HDSeleCDSelecCDAlkyCDMtbeCDEtbeCDTamCDTaeeCDEthe

CDEtheIndustrial column location

f ethyl, butyl and methyl acetate Europeof methyl acetate Europe and Asia

f methylal Europe and Asiaf methanol from formaldehyde Europesters Asia

ISOMPLUSDimerSM8CDCumeneCDIBCDTECH

EBBASF SELOP

CDAcrylamidSulzer

Butene-1Selective desulfurisation of mid catalytic naphthaLow cost desulfurisation of low sulfur catalytic naphthaSulfuric acid alkylation of olefinsMTBE production from mixed C4s and methanolETBE production from mixed C4s and ethanolTAME production from mixed C5s and methanolTAEE production from mixed C5s and ethanolCo-production of ethers from mixed C4C7s andmethanolSelective hydrogenation of diolefins within etherprocesses aboveIsomerisation of n-olefins to iso-olefinsDimerisation of iso-olefins in C4 streamsAlkylation of benzene with propyleneDecomposition of MTBE to high-purity isobutyleneAlkylation of benzene with ethylene

Selective hydrogenation of C4s and C5s for diolefinreduction

eSM Hydration of acrylonitrile to acrylamideSuperfractionation of mixed C4s to produce high-puritybutene-1

776 G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780

tus. Table 3 shows the process applications CDTECH presentlyoffers for licensing.

1.3. Opera

No majcesses. Givthis is remapaying extrtrollable opprocess theetry. If thiswill contamco-productfrom these

1.4. Triple

1.4.1. EcoPresent

ing large sset-ups betwpresentersto less rotatin general,

For hyappeared [3

(1) Lower(2) Lower(3) Lower

(515 b(4) Lower

to vapo(5) Longe

reducenone-r

1.4.2. EnvBecause

ment and thhave less diit will also

1.4.3. SociIn react

evaporationrate, but thbehavioursevere thandistillationSafety of cacceptancehelp in ach

A sustaiprocess inte[12].

Fig. 1. Packing types in a reactive distillation column.

vailable commercial catalytic packings andeneous internals

alyst particles as such cannot be used directly as distil-packing because they form too compact a mass for thed flow of the vapour and downward flow of liquid. There-he catalyst particles are placed in various ways in then as shown in Fig. 1. A short description will follow foracking type.ale packing the catalyst particles are placed in pockets inbelt and these are supported in a distillation column reac-h the help of open mesh knitted stainless steel wire so thatuisite flow of the vapour and liquid can be ensured. This

f a structure prevents the loss of catalyst, it allows for theg of the catalyst particles and it also prevents the break-the catalyst particles due to mechanical attrition [13].

acking structure was the first commercially used packingctive distillation applications for MTBE production. Theontainers that hold the catalyst particle should be inert toctants and products and to the conditions in the reaction. Cotton, polyester, nylon and the like can be used for theontainers, but fiberglass has been reported to be the mostused [13,9, p. 179].bales are placed on trays, so that they can be easily

ed. The technology is licensed by CDTECH. The hydro-tion experiences

or operational problems are reported on these pro-en the complex phenomena occurring in the columnsrkable. The designers, knowing the complexity, werea attention to all critical details, to ensure robust con-eration. For instance in the Eastman methyl acetatereactants flows are exactly controlled on stoichiom-would not be precisely controlled either methanolinate the product or acetic acid will contaminate thewater. The process has no back-up ability to recoverkinds of errors [5].

P (prot, planet, people) business drivers

nomic advantagesreactive distillation applications in general are show-avings in capital cost and energy over conventional

een 15 and 80% [3,6,7,1012]. Moreover, industrialoften mention the higher reliability of the system, dueing equipment in particular and due to less equipmentrequiring less maintenance.drogenations the following specific advantages]:

capital expenditure due to no recycle compressor H2.capital expenditure due to integrated heat removal.capital expenditure due to lower H2 pressurear i.s.o. 2540 bar).feedstock cost due to higher selectivity by productur by lower temperature across reaction zone.

r catalyst life due to (a) washing with cleans reflux, (b)d oligomer formation, because dimer goes to bottomeacting zone.

ironmental advantagesreactive distillation reduces the number of equip-e number of connections between equipment it willffusive gas emissions. Because it requires less energyreduce carbon dioxide emissions.

al acceptanceive distillation the heat of reaction is removed by. A higher reaction rate means a higher evaporation

e reaction temperature changes very little. Runawayof a reactive distillation is therefore in general less

a conventional reactor. This means that reactives in general require less safety precaution measures.hemical processes is the key critical factor for social; so reactive distillation implementation is likely toieving social acceptance.nable development assessment for industrial cases ofnsification, including reactive distillation is available

1.5. Ahomog

Catlationupwarfore, tcolumeach p

In ba clothtor witthe reqtype oswellinage ofThis pfor reacloth cthe reasystemcloth cwidely

Thereplac

G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780 777

dynamics and mass transfer characteristics are published. Forreferences see Refs. [2226] in Ref. [9, p. 188]. These balepackings htillation proof reduceddistillationachieve theof other pamately asscontainersreplaced wcant loss incycle life o

Catalystwire gauze(Katamax)transfer is a

In theseby sandwicpairs of coof wire gauals, such asview and thIt is imporpackings, kby CDTECbut whosetioned prevpacking de

SulzerKATAPAKnates catalthis modulfraction caspecific pro

1.6. Homo

For hommulti-trayhold-up cafor reactionfor instancIn cases of

1.7. Multi-

Reactivetional integThe classicman ChemAnother exand olefin cumn contaienters beloas vapour tto di-alkyl-

reaction section where it is hydrogenated to H2S and hydrocar-bons. The H2S and clean light naphta leave the column at the

e heavy naphta leaves the column at the bottom. The heattioncyclour

or co

ull dided]. Ation

ISERs Ds syol ofessibhermcapases seousSYeric

urac

wledns t

edge

sforictin

ejectstigarteniinitia

matudiecessive oracticessy assBASe distativownrapi

easibinestion

e reator isere ilighte coatioave been used for a long time in the reactive dis-cesses and they provide good performance in termspressure drop and reduced channelling within thecolumn. But it has been reported that they fail todistillation performance achieved with many types

ckings [14]. Also as the catalyst containers are inti-ociated with the support structure; both the catalystand the support structure must be dismantled andhen the catalyst is spent. This can lead to a signifi-

the operating time even when the catalysts have af several months.particles sandwiched between corrugated sheets of

are licensed from Sulzer (KataPak) and Koch-Glitsch. Information on fluid dynamics, mixing and massvailable (Refs. [3034] in Ref. [9, p. 188,].

catalytic structured packing the catalyst bed is formedhing a layer of solid, particulate catalyst between

rrugated plates. Preferably the plates are comprisedze or metal screen material but other types of materi-plastic gauze and ceramics are also utilised. The tope side view of such a type of packing are shown here.tant to note here that a class of catalytic distillationnown as the M-series packings, have been developedH which belong to the class of structured packingconfiguration is markedly different from that men-iously. The M-series packing has a higher catalystnsity than their conventional counterparts.Chemtech has extended their product line with-SP (SP for separation performance), which alter-yst bags with conventional distillation layers. Withar concept separation efficiency and catalyst volumen be varied to perfectly fit the requirements of eachcess [15].

geneous reaction column internals

ogeneous reactions taking place in the liquid phasecolumns are commercially used, because the liquid-n be adjusted to achieve the required residence time[9, p. 169]. In the methyl acetate column of Eastman

e trays with bubble caps of 25 cm high are used [5].fast reactions packings could be used [9, p. 169].

function integrated design

distillation can also be considered as a result of func-ration design rather then just a new unit operation.

example is the methyl acetate process by East-ical, which combines five functions into one column.ample is the hydrodesulfurisation of crude mercaptanontaining gasoline process from CDTECH. The col-ns two different catalytic reaction sections. The feedw these two sections. Mercaptan and diolefins go upo the top section, where they react in the liquid phasesulfides. This component goes down to the second

top; thof reacThe reby vappump

A fis provsen [11integraTHESProcesprocesThe toily accbasic tsor) isproceserogen

Thetic numthe acctic knodecisioknowl

Tran Pred A r

inve Sho

ble

Sumative sof Proattractnot-att

Proquicklcases.

reactiva qualiand sh

Fornical fguideldistilla

- If threac

- If this a

- If thformis directly used inside the column for evaporation.e mass flow of sulfur containing species is obtainedand liquid countercurrent flow without requiring ampressor [2].escription of a general method of functional design

by Siirola [6]. A summary is provided by Harm-n embracing systematic investigation of functional is facilitated by a software package called SYN- [16]. The software is developed by the company

esign Center (PDC) and is used to support theirnthesis services in the area of reactive distillation.fers several benefits: commencing with limited, eas-le information concerning the reaction system andodynamic data the SYNTHESISER (and its succes-ble of developing catalytic and reactive distillationystematically. Besides non-catalytic reactions het-or homogeneous catalysis can also be investigated.NTHESISER applies a hybrid, so called heuris-approach. This sophisticated approach combines

y of numerical results with the versatility of heuris-ge. The advantage of this approach is obvious: the

aken are based on numeric calculation. Heuristicbases support the generation of these results by:

ming qualitative information into quantitative data.g lacking parameter by conclusion of analogy.

ion of unsuited processes in an early phase of thetion.ng numerical calculations by the generation of feasi-l values.

rizing the authors experience with PDC, four cooper-s were carried out at Shell, revealing that the serviceDesign Center is useful in identifying feasible andpportunities and also in rejecting non-feasible or

ve options.ing companies have developed their own methods foressing the feasibility and attractiveness for particularF for instance provides a rapid way of identifyingtillation as a feasible and attractive option is by usinge graphical method provided by Schoenmakers [17]in Fig. 2.dly assessing the economic attractiveness and tech-ility of cases the author has developed a set of 15for Shell. Here are a few of those rules. Reactiveis attractive:

ction is so exothermic that cooling in the conventionalrequired.

s un undesired consecutive reaction and the productboiler.nventional separation is expensive due to azeotropen and by reaction the azeotrope is prevented.

778 G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780

Fig. 2. Princi

For desiSundmachea book by Dpurposes. Fa reactive a

Severalable for dyndistillationcomprehen

1.8. Colum

Columntransfer, herequired peligent Coluprinciple imparticular rin a speci[21]. The kof Computtion with rresulted in[22], a softCFX-INTI[23]. All m[21].

Reactiohold-up. Sphold-up inapplied inscarce, sincA large amtion reactioas Amberlyparticles wFrom variomated thatis in the ran

Table 4CDTECH pilot plant scale reactive distillation columns [30,2]Diameter (in.

cess

TECchno

plaey hs rapis ponne

4 sumzer Cengthe Eof reses prial pfar aesigned frk-S

l reapedercianam20-c

condencety) e[5].

cale-ples of equipment choice reactor and distillation from Ref. [17].

gn and optimisation the book ReactiveDistillation byr [9] is available. A chapter on reactive distillation inoherty [18] is very useful for insight and educationalrey provides a method for testing the occurrence ofzeotrope [19,20].commercially available computer packages are avail-amic and steady state flow sheet modeling of reactive

. Due to the limited time for preparing this paper asive overview was not possible.

n internals selection and design

design requires relations for reaction kinetics, massat transfer and impulse transfer in relation to therformance. A large European project INTENT (Intel-mn Internals for Reactive Separations) aimed atprovement of column internals and their fitting to

eactive separation processes. The results are foundal issue of Chemical Engineering and Processingey theoretical innovation has been the applicationational fluid Dynamics programming in combina-igorous, rate based process simulation. The projecta software tool for internal pre-selection ADVISER

1334

2. Pro

CDtion teMTBEthen th

Thitationtheir cTable

Sultractorpart oftestingprocesindust

Sotheir dobtainKatapasevera

develocomm

nal dy(up toThe seexpericapacizones

2.1. Sware tool for the fluid flow and reaction simulationNT [21] and a rate based flow simulator, PROFILERodels were verified for five reactions in pilot plants

n kinetics in the end leads to the required catalystecial attention should be paid to the feasible catalystthe column. Reported data about catalyst amountscommercial reactive distillation (RD) processes aree these data are regarded as proprietary information.ount of open literature on RD deals with etherifica-ns for which mostly an ion exchange resin catalystst is used. This catalyst is typically used as sphericalith particle size ranging between 0.5 and 0.84 mm.us publications on MTBE RD-modeling it is esti-the amount of catalyst in a commercial RD-columnge of 710 tonnes [24].

For reliamethods ar

- Brute for- Rate mod

In the Bplant are kediameter anthe catalytiavoid wallrelatively c

In the Rand reactioplant scaleings are ap) Height feet Number Comment30 >1 Exp. dev. units50 5 Com dev. units CDU50 2 Top sections CDU50 2 Bottom sections CDU

development in pilot plants

H is a very successful developer of reactive distilla-logies. Their first commercial implementation was ant at Charter Oil, Houston, TX USA in 1981. Sinceave implemented 123 commercial units [2].id development and reliable commercial implemen-

robably due to their large pilot plant facilities andction with the engineering contractor ABB/Lummus.marises their main experimental facilities.hemtech has pilot plant facilities for testing and con-

ineering capability for commercial scale design. Asuropean consortium project INTINT comprehensiveactive packings was carried out and for the study twoilot plants were build [25]. Currently more than 20rocesses are in pilot testing phase [4].ll companies seem to use pilot plants to validates. BASF has a reactive distillation pilot plant facilityom Sulzer Chemtech in Ludwigshaven with Sulzerpacking [25]. Shell Global Solutions has currentlyctive distillation pilot plants in operation. Eastmantheir first methyl acetate plant without the benefit ofl reactive distillation codes at that time (1980). Inter-ic simulation software and piloting at several scalesm diameter) were then used to prove the concept.plant, built 7 years later with the benefit of operating

, is virtually identical (geometry, staging, dimensions,xcept for some sieve detail in the upper non-reactive

up methods

ble scale-up from pilot plant facilities two scale-upe typically taken into consideration:

ce;el.

rute force method all critical parameters in the pilotpt equal to the commercial scale design, so only thed the feed flow rates are scale down. Care is taken that

c distillation packing is fitted correctly to the wall, toflow. The Brute force method is very reliable, butostly.ate model method an elaborate model of mass transferns is made for the commercial scale and for the pilot. In the commercial scale and in the pilot plant pack-plied for which the mass transfer performances are

G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780 779

Table 5Scale-up methods

Commercial s

Feed qualityDiameterHeightPackingVelocitiesFeed locationGood distribu

gas and liq

known. Thin which alis fitted cosolution, issummarise

2.2. Barrie

The bar

- Complex- Complex- Expensiv- Validated- Difficult

Let usresearch) to

Reactivetor or a difacilitatingcated. Thethe design pno methoderature forwould be v

Reactivebut for allappear to bCDTECHcomplex. Inlation to deif the columgral way, aachievable

Extensivthe EastmaHowever, tin one colusimulationmodest sizesure drop ageneral theand othersflows. So p

Validated scale-up knowledge is needed. However, the scale-up methods, as indicated by the author, are available and also

omm

idatificulperi

reactkelyin fale de

rom

toandctivever,

acadtionign ans, rimpic mrtuahe saeral.longistillns, h

an

ver,

ns waceent

ale-re dgralargergycom

ces,iculaign ktiond cao thied.cale Pilot plant Brute force Pilot plant + models

Same SameSmaller SmallerSame SmallerSame SmallerSame Smaller

s Same Same model locationtionuid

Avoid wall flow Avoid wall flow

e elaborate model is then validated in the pilot plantso care is taken that the catalytic distillation packingrrectly to the wall, to avoid wall flow. This scale-upsuggested by Schoenmakers in Ref. [9, p. 38]. Table 5s the key features of both scale-up methods.

rs to commercial implementation

riers (real and perceived) are:

design.control design with less degrees of freedom.e pilot plant development is needed.scale-up knowledge is needed.

to start-up and to operate.

briefly discus these barriers and ways (includingreduce them.distillations are more complex to design than a reac-

stillation column. However, methods and softwarethe design are becoming increasingly more sophisti-large number of commercial applications shows thatroblems have been solved. For irreversible reactions

s and little information are available in the open lit-conceptual design. Academic research in this areaery useful.

distillations have indeed less degrees of freedom,cases reported so far sufficient variables for controle available. The design of the control is, according toLummus/ABB in their commercial applications, not

other cases this may be the case and dynamic simu-termine a feasible control design is needed. However,n design and the control design are done in an inte-

dditional cost savings and increased robustness are[26].

many cfor val

Difown ex

phaseIt is liphasetrollab

2.3. F

Duepricesing reaMoreoand bydistilla

Desreactiofer andDynamitate vi[21]. Tin gen

Thetive) dreactiolibriumMoreoreactioto replimplemable sc(pressu

Inteshow l

Exedesigning forin part

Desof reac[27] an[29]. Sextende pilot plant development appeared to be needed forn process developed in the nineteen seventies [5].his is about the most complex process (five functionsmn) imaginable and at that time no comprehensivepackages were available. Presently pilot plants ofs and scale-up knowledge on mass transfer and pres-re sufficient to reduce the risk to normal levels. Inpilot plants applied by CDTECH, Sulzer Chemtech

have a small diameter of typically 5 cm, so small feedilot plant cost are low.

Reactivethe field ofimplementup in mind

3. Conclu

Reactiverapidly impof the petroercial scale applications, whose results can be usedon of models.t start-up and operations are not reported. From myence at a manufacturing site operation of the three-ive distillation column was not particularly difficult.that the increased attention and focus in the designct reduces the operational difficulty, by a robust con-sign.

reactive distillation to process intensication

increased worldwide competition, increased fuelmore stringent emission limits, the benefits of apply-

distillation and process intensification will increase.due to increased knowledge by industrial experienceemic research, the barriers for implementing reactiveand process intensification will be further reduced.nd simulation tools need to be developed that includeesidence time distribution, mass transfer, heat trans-ulse transfer. In the long-term Computational Fluidodelling packages integrated with reaction will facil-

l prototyping of reactive distillation column internalsme approach can be applied to process intensification

er term forecast for the application of catalytic (reac-ation in the petrochemical industry, is that manyomogeneous and heterogeneously catalysed, equi-

d irreversible, will be implemented commercially.in some separations, selective reversible chemicalill be applied in reactive extractive distillation set-upsexpensive distillations [27]. Barriers to commercialations are lowered, due to pilot plant testing and avail-up knowledge on mass transfer and impulse transferrop).design, optimisation, control by dynamic simulationcost savings [26].loss minimisation should be included in conceptualputer packages to include a more even spread of driv-causing lower exergy losses for reactive distillationr [28] and process intensification in general.nowledge is being developed for novel combinationsand separation, like reactive extractive distillation

talytic reaction in dividing wall distillation columnss shows that the field of reactive distillation is being

distillation can also be seen as the forerunner inprocess intensification. The general lessons for rapidation are operation of pilot plants designed with scale-. Either by brute force or via model validation.

sions

distillation for many different processes has beenlemented in more than 150 commercial operationschemical and chemical industry.

780 G.J. Harmsen / Chemical Engineering and Processing 46 (2007) 774780

These implementations show large capital and energy costreductions, a lower environmental impact and safe and reliableoperation.

Reliable scale-up from to pilot plant facilities without inter-mediate demonstration scale seems to have been obtained in allcases. In most cases the scale-up is carried out by the majortechnology providers CDTECH and Sulzer Chemtech.

Academic research has lead to methods and tools to identifydesign and optimise new applications, however all for reversiblereactions o

The pretive distillatechnology

Reactivethe field of

4. Recomm

Processare typicalltions in rea

Scale-upvalidation mtechnologie

Acknowled

JeffreyfunctionalLeslie Mucon CDTECCDTECHGerla and oother inforSYNTHESfor reactiveShell Globcomments.ing catalytSolutions fHesselink o

Reference

[1] L. Much[2] M.E. Lo[3] K.L. Ro

distillati2022, 2

[4] J. Gerla,10 Augu

[5] J.J. Siirola, Eastman methylacetate information, Email to Harmsen, 21 July2005.

[6] J.J. Siirola, Industrial applications of chemical process synthesis, in: J.L.Anderson (Ed.), Advances in Chemical Engineering, Process Synthesis,vol. 23, Academic Press, 1996.

[7] J.J. Siirola, Synthesis of equipment with integrated functionality, in: Syl-labus First Dutch Process Intensification: Profits for the Chemical IndustrySymposium, Novem, 7 May, 1998.

[8] G.J. Harmsen, L. Chewter, Industrial applications of multi-functional,multi-phase reactors, in: International Symposium on Multi-functionalReactors I, Amsterdam, 2628 April, Chem. Eng. Sci. 54 (1999)

11545.Sundns, W. Hatribus.), R3.. Harm

g. Pro. Harent,

cessi4, pp. Sm. Gel

Gotze208mannh ed.,. Sch

l persF. Dol, BosFrey,

e dist.rey, e

g. TecGorakAvramKloe

Taylo352oritz5 (2

. Almes in r.D.-pr.M. Kreact99.

Sauarlicatihnica

SandeGroteecem

searchnly.sent status seems to warrant the conclusion that reac-tion is now an established unit operation in process.

distillation can also be seen as the front-runner inprocess intensification.

endations

synthesis heuristics on feasibility and attractivenessy recommended to be developed for irreversible reac-ctive distillation.based on pilot plants with the brute force or by modelethod are recommended for process intensification

s in general.

gements

Siirola of Eastman, for detailed information on hisdesign method and Eastman process design details.ha, ABB-LUMUS/CDTECH for latest informationH applications. Mitchell Loescher and Will Groten offor information on CDTECH applications. Johannesthers, Sulzer Chemtech on industrial applications andmation. Stephen Tlatlik of Process Design Center onISER tool and other process synthesis informationdistillation conceptual design. Harry Kooijman of

al Solutions for providing information and criticalAshees Sastry of Shell Global Solutions for provid-

ic package information. Tim Nisbet of Shell Globalor reactive distillation pilot plant information. Wimf Shell Global Solutions for reviewing the paper.

s

a, ABB-LUMUS/CDTECH, Email to author, 20 March 2007.escher, Email to author, 25 July 2005.ck, CDTECH, Selective hydrogenation of MAPD via catalyticon, in: ERTC Petrochemical Conference, Amsterdam, February002.S. Chemtech, Reactive distillation, Emails to author, 26 July andst 2005.

154[9] K.

tio[10] G.J

con

(Ed200

[11] G.JEn

[12] G.JopmPro200

[13] L.A[14] A.P[15] L.

201[16] Ull

ent[17] H.G

tria[18] M.

Hil[19] T.

tiv618

[20] T. FEn

[21] A.[22] Y.[23] M.[24] R.

518[25] P. M

CIT[26] C.P

turPh

[27] N.Jby88

[28] E.appTec

[29] S.[30] W.

9 DRemacher, A. Kienle, Reactive Distillation Status and Future Direc-iley-VCH, Weinheim, 2003.rmsen, G. Korevaar, S.M. Lemkowitz, Process intensificationtions to sustainable development, in: A. Stankiewicz, et al.e-engineering the Chemical Processing Plant, Dekker, New York,

sen, Industrial best practices of conceptual process design, Chem.c. 43 (5) (2004) 671675.msen, Process intensification contributions to sustainable devel-in: A. Stankiewicz, et al. (Eds.), Re-engineering the Chemical

ng Plant: Process Intensification, Marcel Dekker Inc., New York,. 495522.ith, US Patent 4,215,011 (1980).bein, et al., EU Patent Application EP 0428 265 (1991)., et al., Reactive distillation with Katapak, Catal. Today 69 (2001)., Ullmanns Encyclopedia of Industrial Chemistry Release, sev-2005.oenmakers, et al., Reactive and catalytic distillation from an indus-pective, Chem. Eng. Proc. 42 (2003) (2003) 145155.herty, et al., Conceptual Design of Distillation Systems, McGraw-ton, 2001 (Chapter 10, Reactive distillation).

et al., Reactive azeotropes in kinetically controlled reac-illation, Trans. Inst. Chem. Eng. Part A 77 (7) (1999) 613

t al., Thermodynamic fundamentals of reactive distillation, Chem.hnol. 22 (1) (1999) 1118., et al., Chem. Eng. Process. 44 (2005) (Editorial and whole issue).enko, et al., Chem. Eng. Process. 44 (2005) 609616.

ker, et al., Chem. Eng. Process. 44 (2005) 617630.r, et al., Modelling reactive distillation, Chem. Eng. Sci. 55 (2000)29., A. Gorak, Aus zwei mach eins, Reaktive Trennung senkt Kosten,002) 3637.eida-Rivera, Simultaneous optimisation of process control struc-

eactive distillation, in: H.A. Akse (Ed.), OSPT Process Technologyojects in Miniposters Form, 2003 (http://ospt.ct.utwente.nl).uipers, et al., Functionalized solvents for olefin isomer purification

ive extractive distillation, Chem. Eng. Res. Des. 85 (A1) (2007)

, Energy efficient process design by equipartion of forces: withons to distillation and chemical reaction, Ph.D. Thesis, Norwegianl-Natural Science University of Trondheim, 1998.r, et al., Chem. Eng. Res. Des. 85 (2007) 149154.n, CDTECH, Catalytic distillation Technologies, Slides presentedber 1998 at mini-symposium network reactor/process design, Shell& Technology Center, 1998.

Reactive distillation: The front-runner of industrial process intensificationIntroductionReactive distillation processes in commercial operationPotential process applicationsOperation experiencesTriple P (profit, planet, people) business driversEconomic advantagesEnvironmental advantagesSocial acceptance

Available commercial catalytic packings and homogeneous internalsHomogeneous reaction column internalsMulti-function integrated designColumn internals selection and design

Process development in pilot plantsScale-up methodsBarriers to commercial implementationFrom reactive distillation to process intensification

ConclusionsRecommendationsAcknowledgementsReferences