.eactor Design
`Chemical
`Technology
`and
`
`Edited by
`Hugo I~ de Lasa
`
`NATO ASI Series
`
`Series E: Applied Sciences - No. 110
`
`Akermin, Inc.
`Exhibit 1018
`Page 1
`
`

`
`he NATO Science
`scientific and technological
`ientific communities.
`
`~hers in conjunction with the
`
`"ation
`
`party
`
`Chemica~ Reactor Design and
`Technology
`Overview of the New Developments of
`Energy and Petrochemical Reactor
`Technologies. Projections for the 90’s
`
`edited by
`
`Hugo I. de Lasa
`
`Faculty of Engineering Science
`The University of Western Ontario
`London, Ontario
`Canada N6A 5B9
`
`1986 Martinus Nijhoff Publishers
`Dordrecht / Boston / Lancaster
`Published in cooperation with NATO Scientific Affairs Division
`
`Akermin, Inc.
`Exhibit 1018
`Page 2
`
`

`
`Proceedings of the NATO Advanced Study Institute on "Chemical Reactor Design
`and Technology", London, Ontario, Canada, June 2-12, 1985
`
`Library of Congress Cataloging in Publication Data
`
`NATO Advanced Study Institute on "Chemical Reactor
`Design and Technology (1985 : London, Ont.)
`
`(NATO ASI series. Series E, Applied sciences ; 110)
`"Proceedings of the NATO Advanced Study Institute on
`"Chemical Reactor Design and Technology", London,
`Ontario, Canada, June’2-12, 1985"--T.p. verso.
`"Published in cooperation with NATO Scientific
`Affairs Division."
`i. Chemical reactors--Design and construction--
`Congresses. I. De Lasa, Hugo I. II. North Atlantic
`Treaty Organization. Scientific Affairs Division.
`III. Series: NATO ASI series. Series E, Applied
`sciences ; no. II0.
`TPI57.N296 1985 660.2’81 86-8345
`ISBN 90-247-3315-4
`
`ISBN 90-247-3315-4 (this volume)
`ISBN 90-247-2689-1 (series)
`
`Distributors for the United States and Canada: Kluwer Academic Publishers,
`190 Old Derby Street, Hingham, MA 02043, USA
`
`Distributors for the UK and Ireland: Kluwer Academic Publishers, MTP Press Ltd,
`Falcon House, Queen Square, Lancaster LA1 1RN, UK
`
`Distributors for all other countries: Kluwer Academic Publishers Group, Distribution
`Center, P.O. Box 322, 3300 AH Dordrecht, The Netherlands
`
`All rights reserved. No part of this publication may be reproduced, stored in a
`retrieval system, or transmitted, in any form or by any means, mechanical,
`photocopying, recording, or otherwise, without the prior written permission of the
`publishers,
`Martinus Nijhoff Publishers, P.O. Box 163, 3300 AD Dordrecht, The Netherlands
`
`Copyright © 1986 by Martinus Nijhoff Publishers, Dordrecht
`
`Printed in The Netherlands
`
`NATO A~
`
`CHEMICAL PC!
`
`Overview of t~
`Petrochemical Reacto~
`
`DIRECTOR
`
`Prof0 Hugo de Lasa,~
`
`CO-DIRECTOR
`
`Prof0 Alirio Rodrigu~
`
`ORGANIZING ADVISORY
`
`Prof0 MoAo Bergougno~
`Prof0 JoR° Grace~ Th~
`Dro R0I~0 Koros, Exxo~
`Dr, M0 Ternan, Energ~
`
`Akermin, Inc.
`Exhibit 1018
`Page 3
`
`

`
`TABLE OF CONTENTS
`PREFACE
`
`Ao RODRIGUES, C. COSTA and R. FERREIRA
`Transport Processes in Catalyst Pellets
`
`x!
`
`VII
`
`LoJo CHRISTIANSEN and JoE. JARVAN
`Transport Restrictions in Catalyst Particles with Several
`Chemical Reactions
`
`Mo TERNAN and R.H, PACKWOOD
`Catalyst Technology for Reactors used to Hydrocrack
`Petroleum Resldua!
`
`H. HOFMANN
`Kinetic Data Analysis and Parameter Estimation
`
`M.Po DUDUKOVIC
`Tracer Methods in Chemical Reactors° Techniques and
`Applications
`
`Jo VILLERMAUX
`Macro and Micromixing Phenomena in Chemical Reactors
`
`JoRo GRACE
`Modelling and Simulation of Two=Phase Fluidized Bed
`Reactors
`
`R.O. FOX and LoTo FAN
`A Stochastic Model of the Bubble Population in a Fluidized
`Bed
`
`M. Ao BERGOUGNOU, CoLo BRIENS and Do KUNII
`Design Aspects of Industrial Fluidized Bed Reactors
`
`H0 DE LASA and SoL.P. LEE
`Three-Phase Fluidized Bed Reactors
`
`UoCo TOSYALI and BoZo UYSAL
`Liquid Phase Mass Transfer Coefficients and Interfacial
`Area in Three-Phase Fluidization
`
`Wo-Do DECKWER
`Design and Simulation of Bubble Column Reactors
`
`AoAoCoMo BEENACKERS and W°PoMo VAN SWAAIJ
`Slurry Reactors, Fundamentals and Applications
`
`Mo CRINE
`Hydrodynamics of Trickle-Bedso The Percolation Theory
`
`35
`
`53
`
`69
`
`107
`
`191
`
`245
`
`291
`
`305
`
`349
`
`393
`
`411
`
`463
`
`539
`
`Akermin, Inc.
`Exhibit 1018
`Page 4
`
`

`
`XII
`
`R,Mo KOROS
`Engineering Aspects of Trickle Bed Reactors
`
`Ao GIANETTO and Fo BERRUTI
`Modelling of Tcickle Bed Reactocs
`
`DoLo CRESSWELL
`Heat Transfer in Packed Bed Reactors
`
`JoJo LEROU and G,~, FROMENT
`The Measurement of Void Fraction Profiles in Packed Beds
`
`A0 RAVELLA~ Ho DE LASA and E. ROST
`Converting Methanol into Gasoline in a Novel Pseudo-
`adiabaticCatalytic Fixed Bed Reactor
`
`C.S. YOO and AoGo DIXON
`Maldistribution in the Radial-Flow Fixed Bed Reactor
`
`Co KIPARISSIDES and Ho MAVRIDIS
`Mathematical Modelling and Sensitivity Analysis of High
`Pressure Polyethylene Reactors
`
`Fo KIRKBIR and B0 KISAKUREK
`Dynamic Analysis of an Ethane Cracking Reactor
`
`I,Bo DYBKJAER
`Design of Ammonia and Methanol Synthesis Reactors
`
`LECTURERS, PARTICIPANTS~ LOCAL ORGANIZING COMMITTEE
`
`SUBJECT INDEX
`
`579
`
`631
`
`687
`
`729
`
`737
`
`749
`
`759
`
`779
`
`795
`
`821
`
`825
`
`TRANSPORT PROCESSES IN
`
`Allrio RODRI
`
`Department
`University
`
`INTRODUCTION
`
`The subject of rea
`a well established bra
`such as by Aris [I] an
`long been recognized s
`
`In this paper we w
`them to the design of
`regime.Then we will co
`traparticle convection
`this question will be.
`thermal and nonisother~
`sivities and the impli~
`the design and operati,
`
`From a qualitative
`and diffusion in an is~
`parameter governing th~
`ratio between time con~
`If Td << Tr the reacti~
`the concentration prof
`equal to the external ~
`is around unity.Howevm
`llet will be lower tha~
`effectiveness factor w~
`order is n> O. In the J
`kinetic controlled reg~
`ting in the diffusion ~
`the Thiele modulus i i~
`irreversible nth order
`
`Akermin, Inc.
`Exhibit 1018
`Page 5
`
`

`
`463
`
`SLURRY REACTORS, FUNDAMENTALS AND APPLICATIONS
`
`AoAoCoMo Beenackers*) and WoPoMo van Swaaij**)
`
`*) Chemical Engineering Laboratories, University of Groningen,
`Nijenborgh 16, Groningen, The Netherlands
`**) Laboratory of Chemical Engineering, Twente University of Tech-
`nology, PoOo Box 217~ Enschede, The Netherlands
`
`io INTRODUCTION
`Slurry reactors are applied nowadays in a wide range of chemical
`processes~ both on a laboratory scale and in industrial practice°
`Applications range from catalytic hydrogenation of edible oils,
`coal hydrogenation, Fischer- Tropsch reactions to biological reac=
`tors (eogo single cell proteins) and polymerization reactors° Do=
`raiswamy and Sharma [I] identified over 50 different processes in
`which slurry reactors have been used and more applications are to
`be expected~ A lot of experimental work has been devoted to slurry
`systems° Overviews have been given in textbooks on three phase
`reactors (see Doraiswamy and Sharma [i] and Shah[[~])o Recent
`view articles are: Chaudhari and Ramachandran Deckwer and
`Alper [4]~ Hofmann [5]~ Shah et alo [6]o Even more information is
`available on three phase fluidizationo This regime is related to
`slurry reactors and sometimes no distinction is made with slurry
`reactors° In the present lecture series there is a separate contri-
`bution on three phase fluidization and therefore it will not be
`discussed in this paper° After a discussion of the different pro-
`perties of slurry reactors as compared to other three phase systems
`and the different modifications of slurry reactors~ hydrodynamic
`aspects are considered: minimum suspension criteria, hold=up frac=
`tions and axial mixingo Then regimes of mass transfer in relation
`to chemical reactors will be treated in more detai$o Subsequently,
`developments in reactor modelling will be discussed and progress in
`application of such models to real life systems will be summarized°
`The paper concludes with prospective new developments and a com=
`prehensive list of conclusions°
`
`Akermin, Inc.
`Exhibit 1018
`Page 6
`
`

`
`464
`
`2o PROPERTIES OF SLURRY REACTORS
`In. slurry reactors an attempt is made to realize intensive and
`intimate contact between a gas phase component, usually to be dis-
`solved in the liquid phase, a liquid phase component and a finely
`dispersed solid. With respect to this purpose slurry reactors are
`related to packed bed reactors with the different gas/ liquid flow
`regimes that can be realised (such as trickle flow, pulsed flow,
`dispersed bubble flow etco) Also, there is a lot of similarity with
`three phase fluid bed systems. These latter systems share many
`properties with .slurry reactors but the main difference is the fact
`that in fluid beds with upward fluid flows the drag force acting on
`the solids by the gas and liquid flow is on the average balanced by
`the net weight of the particles, while in slurry reactors the over-
`all liquid-solid slip velocities are practically zero and particles
`remain suspended by the action of the turbulence in the liquid
`phase. This usually implies that somewhat larger particles are
`applied in three phase~ fluid beds allowing for a more or less re-
`stricted height of the expanded particle bed under the action of
`the gas and liquid-flow only, thus creating a liquid/gas freeboard
`(in slurry reactors usually dr~ < 100-200 ~m [7] and no clear free-
`board exists). Typical properties of slurry reactors, and for com-
`parison packed bed co-current downflow trickle flow reactors, are
`summarized in Table Io Most properties indicated for slurry reac-
`tors also hold for the three phase fluidized beds° An important
`difference is related to the somewhat larger particle sizes normal-
`ly used in three phase fluidization allowing for a simple separa=
`tion between fluid and particles. The properties indicated in Table
`I can be advantageous or disadvantageous depending on the applica=
`tiono We will shortly discuss the different properties in the light
`of examples of industrial applications°
`
`Table I. Typica! properties of three phase reactors
`
`Property
`
`Slurry reactors
`
`macro-mixing liquid
`
`macro-mixing gas phase
`
`solids mixing
`
`solids replacement
`
`intermediate to ideally
`mi×ed
`usually intermediate to
`ideally mixed
`usually’intermediate to
`ideally mixed
`continuous
`
`particle size
`
`small to very small
`
`pressure drop
`wetting
`radial heat transport
`heat removal/addition
`
`high (hydrostatic pressure)
`complete
`very fast
`easy
`
`danger of plugging
`gas/liquid mass transfer
`rate constant
`liquid solids mass transfer
`rate constant
`liquid solids separation
`
`small
`
`intermediate
`
`very high
`difficult, costly
`
`Co-cb~rent downflow
`trickle flow reactors
`
`close to plugflow
`at high liquid rates
`close to plugflow
`at high liquid rates
`no mixing
`
`usually intermitted
`(difficult)
`restricted by pressure
`drop
`low
`intermediate to complete
`slow
`difficult (interstag
`cooling or cold shot)
`can be high
`
`high
`
`high
`easy
`
`2.1.
`liquid phase reactant
`a decrease in reaction
`cial. Satterfield and
`thesis in a slurry reac!
`er effective H2/CO rati
`the reactor than in th~
`the feedstock over the
`in minimizing carbon f~
`disproportionation° If
`is desired cascades of
`other hand the degree
`increased by applying i!
`2.2. Macro-mixiqg_~_%i
`high gas phase conversi~
`rent operation is aimed
`(C02, NOx, S02) in limi
`not specially suited f<~
`conversion is aimed for
`be a solution if slurry
`2,3. ~!~ds mixing_~n~
`where the solids have a
`solid phase is to be
`the catalyst is rapidly
`demetallation of oil
`tots have a definite ad
`ly if the residence
`putting slurry vessels
`tion is possible. Rece~
`as applied by Shell in
`continuous removal and
`
`even down to the (sub)
`effectiveness factors
`therefore can have hi~
`Other advantages can be
`the large external sur
`and also in the case of
`catalytic deposition o~
`lysts for desulfurizat~
`2.5. Pres~K~_~K~ in
`dependent of the gas f]
`course, there is also
`tributoro In trickle
`by the gas flow rate a
`size the pressure drop
`In three phase reactor
`and ~ere pressure dro
`lime scrubbing of flue
`special contactors all
`
`Akermin, Inc.
`Exhibit 1018
`Page 7
`
`

`
`realize intensive and
`ent, usually to be dis-
`component and a finely
`)se slurry reactors are
`ferent gas/ liquid fl0w
`~kle flow, pulsed flow,
`~ lot of similarity with
`ter systems share many
`~ difference is the fact
`~he drag force acting on
`the average balanced by
`~urry reactors the over-
`ally zero and particles
`cbulence in the liquid
`: larger particles are
`for a more or less re-
`ed under the action of
`a liquid/gas freeboard
`~ [7] and no clear free-
`reactors, and for CON--
`~kle flow reactors, are
`[cated for slurry reac-
`zed beds° An important
`particle sizes normal-
`tg for a simple separa-
`
`pending on the applica-
`properties in the light
`
`Co-current downflow
`trickle flow reactors
`
`close to plugflow
`at high liquid rates
`close to plugflow
`at high liquid rates
`no mixing
`
`usually intermitted
`(difficult)
`restricted by pressure
`drop
`low
`intermediate to complete
`sl ow
`difficult (interstag
`cooling or cold shot)
`can be high
`
`high
`
`high
`easy
`
`465
`
`2.1. ~a~£sz~ixi~g_~f_li~uido Normally this reduces the average
`liquid phase reactant concentration which results in most cases in
`a decrease in reaction rate. Sometimes this effect can be benefi-
`cial. Satterfield and Huff [8] showed that in Fischer Tropsch syn-
`thesis in a slurry reactor, due to the intense mixing, a much high-
`er effective H2/CO ratio can be present within the liquid phase of
`the reactor than in the feedstock at a given excess H2/CO ratio in
`the feedstock over the usage ratio. This has important advantages
`in minimizing carbon formation and catalyst deactivation due to CO
`disproportionatlon. If overall low axial mixing in the liquid phase
`is desired cascades of slurry reactors should be applied. On the
`other hand the degree of mixing in a trickle flow reactor can be
`increased by applying liquid recycle.
`2.2° Macro-mixing. of the g~_~h~ is an important problem only if
`high gas phase conversions are desired and specially if countercur-
`rent operation is aimed for. Examples are absorption of acid gases
`(CO2, NOx, S02) in lime slurries. A single stage slurry reactor is
`not specially suited for these conditions and if a high gas phase
`conversion is aimed for, multiple stage or slurry tray columns may
`be a solution if slurry operation is to be selected.
`2.3. ~!~_~i~!~g ~nd sq!!~_K~lacement are important in cases
`where the solids have a short lifetime. This can be the case if the
`solid phase is to be converted such as in coal hydrogenation or if
`the catalyst is rapidly deactivating as in hydrodesulfurization and
`demetallation of oil residues° For these applications slurry reac=
`tots have a definite advantage over trickle flow reactors, special=
`ly if the residence time distribution can be reduced, eogo, by
`putting slurry vessels in series. Then also countercurrent opera=
`tion is possible° Recent modifications of the trickle flow reactor
`as applied by Shell involve bunkerflow of the solids allowing for
`continuous removal and addition of catalyst (see [9]~ [i0])o
`2.4° ~a£%iG!~_~%~ in slurry reactors can be small to very small~
`even down to the (sub)micron range° This allows for high particle
`effectiveness factors even at high reaction rates. Slurry reactors
`therefore can have high conversion rates per unit slurry volume°
`Other advantages can be related to the small particle size, such as
`the large external surface as, e.g., in edible oil hydrogenation,
`and also in the case of poremouth plugging, as~ eog., in the (auto)
`catalytic deposition of metals (Ni, V) in the pore mouths of cata-
`lysts for desulfurization of residues (see [Ii])o
`2.5. ~K~£~_~ in slurry reactors is usually more or less in-
`dependent of the gas flow and close to the hydrostatic pressure° Of
`course, there is also the pressure drop required for the gas dis-
`tributoro In trickle flow the pressure drop is strongly influenced
`by the gas flow rate and because of the generally larger particle
`size the pressure drop will often be lower than in slurry reactors.
`In three phase reactors where large gas flows are to be processed
`and where pressure drops are an important cost factor, such as in
`lime scrubbing of flue gases simple slurry reactors are avoided and
`special contactors allowing for a low pressure drop are preferred
`
`Akermin, Inc.
`Exhibit 1018
`Page 8
`
`

`
`466
`
`(see e.g. Coea and Diaz [12]).
`2.6. Particle we~g in slurry reactors is always complete while
`in trickle flow, specially at low specific liquid flow rates, wet-
`ting can be uncertain and a problem. This is nearly always a dis-
`advantage, as partial wetting implies dead zones, hot spots and
`channeling° Only in cases where part of the fluid to be processed
`is a gas, and part a liquid under reactor conditions, partial wet-
`ting may allow for the gas phase to be in better contact with the
`solids. This happens, e.g., in hydroprocessing of light oil frac-
`tions (such as,hydrodesulfurization).
`2.7° Radial heat traq~N£K~ may be a problem in trickle flow but is
`never a problem in slurry reactors. Also the heat
`in slurry reactors is much more easy because high heat transfer
`rates to cooling surfaces are possible, In trickle flow reactors
`interstage cooling, cold shot techniques and cooling with an eva-
`porating solvent are possible; and the latter technique can also be
`applied with slurry reactors.
`2°8. In case ~!~gg~g_~L~_fouling is a problem, trickle flow
`clearly has a disadvantage over slurry reactors. A well known phe-
`nomenon is the pressure drop built-up during operation of a trickle
`flow hydrodesulfurization reactor limiting the operating period.
`Similar problems do not occur in slurry reactors.
`2°9° ~!iN~id mass transfer depends on many factors, as will be
`described below. Related to the lower liquid hold-up the interfa-
`cial area per unit volume liquid in trickle flow columns can be
`very high which is an important advantage in mass transfer.
`2o10o ~iR~%~~_~K~i~n is not a problem in trickle flow but
`can be a difficult and costly operation in slurry reactors, especi=
`ally if very fine particles have to be removed from (viscous) li-
`quids, eogo, by filtration° As already stated, it is in this re-
`spect that three phase fluid bed reactors, which have properties
`rather similar to slurry reactors, are markedly different°
`
`3. DIFFERENT TYPES OF SLURRY REACTORS
`Slurry reactors can be classified according to the phases where the
`reactants are present~ Table II gives an overview° The most impor-
`tant distinction is whether the solid phase is a reactant or a
`catalyst. In principle, the solids could also be inert and only
`present to increase mass transfer between phases as is often the
`case, eogo, in trickle flow reactors. In slurry reactors the intro-
`duction of solids for this purpose only is not worthwhile, with the
`exception of solids like zeolites and activated carbon for enhance-
`ment of mass transfer or improvement of selectivity [21, 22] but in
`such a system the solid is not really inert° Another e~ample is the
`turbulent contactor in which large but light balls are moved by a
`gas flow and irrigated by a liquid phase, However, this regime
`falls outside the scope of the present presentation° If the solid
`is a reactant as well as the gas phase and liquid phase~ the situa=
`tion becomes rather complex; nevertheless~ it corresponds to many
`practical situations (see eogo Shah [2])° A rather exceptional
`
`Table ll, Slurry reactors class
`
`gas liquid and solids
`all reactants
`
`gas and solids are reactants
`
`gas phase is reactant
`solid is a catalyst
`liquid is inert
`
`gas phase and liquid phase
`are reactants
`solid is a catalyst
`
`situation is that the
`phase are reactants. A
`absorption/desorption
`(etasinski et alo [161
`introduce a liquid p~
`resistances~ In the c~
`al!ows for rapid heat
`ticles from being ent
`circulate the solids
`ero
`For similar reasons
`solid is a catalyst~ h
`reaction° Apart from
`dispersion of catalyst
`bution over the avai!a
`In a large class of s
`phase are reactants ar
`development work has
`this class of reactors
`A different classif~¢~
`contacting pattern ant
`the contacting patter~
`view° We will not cc
`phase sparged slurry
`reactor. The liquid a
`continuously° For con
`liquid/solid emulsion
`macromixing close to ~
`ditionso This can be
`
`Akermin, Inc.
`Exhibit 1018
`Page 9
`
`

`
`always complete while
`.iquid flow rates, wet-
`nearly always a dis-
`zones, hot spots and
`fluid to be processed
`)nditions, partial wet-
`,etter contact with the
`ing of light oil frac-
`
`in trickle flow but is
`~_~m~X~!L~dition
`Lse high heat transfer
`trickle flow reactors
`cooling with an eva-
`technique can also be
`
`~ problem, trickle flow
`ors° A well known phe-
`operation of a trickle
`the operating period.
`:ors.
`tny factors, as will be
`J hold=up the interfa=
`e flow columns can be
`mass transfer.
`.em in trickle flow but
`_urry reactors~ especi=
`ved from (viscous) li=
`ed~ it is in this re=
`which have properties
`fly different°
`
`~o the phases where the
`rviewo The most impor=
`.e is a reactant or a
`[so be inert and only
`hases as is often the
`’ry reactors the intro~
`t worthwhile~ with the
`ed carbon for enhance=
`tivity [21~ 22] but in
`Another example is the
`~ balls are moved by a
`However, this regime
`~ntationo If the solid
`quid phase~ the situa-
`[t corresponds to many
`A rather exceptional
`
`Table ll. Slurry reactors classified according to the chemical system
`
`Typica! examples
`
`ref. remark
`
`467
`
`gas liquid and solids
`all reactants
`
`- thermal coal liquifaction
`- CO2 absorption in a lime
`suspension
`- single cell protein
`
`gas and solids are reactants - hydride formation and de-
`composition in a slurry
`
`[12]
`[]13
`
`[14]
`
`[15]
`
`liquid to improve heat
`transport properties
`avoid dust entrainment,
`allowing to operase an
`absorber-desorber con-
`tinuously
`
`gas phase is reactant
`solid is a catalyst
`liquid is inert
`
`- Fischer-Tropsch process
`- hydrogenation of ethylene
`using a suspended Raney
`nickel catalyst
`
`[8]
`[16 ]
`
`liquid phase used
`suspend catalyst and
`improve heat ~ranspor5
`
`gas phase and liquid phase
`are reactants
`solid is a catalyst
`
`many industrial reactions:
`-hydrogenation of edible oil
`- hydro-desulphurization
`- oxygen consuming reactions
`in activated carbon slurries
`
`active carbon may also
`enhance oxygen transfer
`(see D2)
`
`situation is that the liquid phase is inert and gas phase and solid
`phase are reactants° A recently published example is the continuous
`absorption/desorption of hydrogen in hydridable metal/oil slurries
`(Ptasinski et alo [16]). Of course~ there must be a good reason to
`introduce a liquid phase here as it adds to the total transport
`resistances° In the case of the hydride slurries the liquid phase
`allows for rapid heat removal~ prevents metal or hydride dust par~
`ticles from being entrained with the gas and makes it possible to
`circulate the solids continuously between an absorber and desorb-
`ero
`For similar reasons an inert liquid phase is often used if the
`solid is a catalyst, heterogeneously catalyzing a desired gas phase
`reaction. Apart from enhancing the heat transport rates~ the fine
`dispersion of catalyst and the control of the reactants (re)distri=
`bution over the available catalyst can be important factors°
`In a large class of slurry reactors both the gas phase and liquid
`phase are reactants and the solid is a catalyst° Much research and
`development work has been carried out and is still going on for
`this class of reactors°
`A different classification of slurry\reactors can be based on the
`contacting pattern and the mechanical devices applied to influence
`the contacting patterns and mass transfer. Fig° 1 gives an over-
`view. We will not consider three phase fluidizationo The three
`phase sparged slurry reactor is in fact the most simple slurry
`reactor° The liquid and solid phase een be operated batchwise or
`continuously° For continuous operation the intense mixing of the
`liquid/solid emulsion may be a disadvantage because it renders the
`macromixing close to a single ideal mixer operating at outlet con=
`ditions. This can be counteracted by putting more slurry spargers
`
`Akermin, Inc.
`Exhibit 1018
`Page 10
`
`

`
`468
`
`o ~ 0
`
`Fig. I Slurry reactors as classified by the contacting pattern and mechanical
`devices (a) three-phase sparger (g) tray column
`(h) rotating disc contactor
`(b) countercurrent column
`(c) co-current upflow
`(i) three phase spray column
`(d) co-current downflow
`~ ~
`liguid flow gas flow
`(e) stirred vessel
`(f) draft tube reactor
`
`in series. If the particles are removed between the stages, the
`contacting pattern of liquid and solid can also be manipulated
`(eogo countercurrent contact) which may be advantageous in relation
`to catalyst aging (eogo in hydrodesulfurization of residues)° If
`the contact time of liquid/solid suspension with the gas phase is
`relatively short and also in case of reactors of high L/dt values,
`it may be useful to distinguish a co- and countercurrent contacting
`pattern and, specially in relation to the gas hold-up, upflow and
`downflowo To improve the solids suspension and/or to improve mass
`or heat transfer in many cases a stirrer is added to the system. In
`some cases a stirrer can be avoided and yet the solid can be kept
`in suspension by a relatively small amount of gas by means of a
`
`draft tube placed in
`external slurry recyc]
`staging and counterc~
`column and a rotating
`[23]) to shape a slur~
`In flue gas purificat
`llme or CaC03 slurrie
`phase (see [12]). So~
`phase but also simple
`
`4. HYDRODYNAMICS OF
`Hydrodynamics of slur~
`velocity or power inp~
`¯ homogeneously suspend
`up fractions of gas,
`blem is the large nun,
`I) and the fact that
`tially of an empirica2
`to sparged slurry coL~
`problem is the diffe~
`too much overlap we
`cial liquid velocitie~
`sion of the particles
`
`4.1. Minimum suspensi~
`For the design of
`flowing gas or assist
`which the particles
`overall contacting
`increases rapidly if
`the condition of min
`laying on the bottom
`input for suspension <
`ciencyo Another aim
`suspension° This cond
`applications and an
`Therefore~ generally
`deredo
`In ~g~_~!~ w
`sion gas velocity ca~
`just suspended. Thre~
`used to obtain an op~
`Velocities:
`io Particles do not
`than say one or
`observation througl
`2. The lowest gas w
`liquid phase ind,<
`(see Fig° 2).
`
`Akermin, Inc.
`Exhibit 1018
`Page 11
`
`

`
`469 ’
`
`draft tube placed in the slurry reactor. Other possibilities are
`external slurry recycling which can be combined with jet mixing° If
`staging and countercurrent operation are essential a slurry tray
`column and a rotating disk column have been suggested (Ghim et alo
`[23]) to shape a slurry reactor towards a countercurrent contactor.
`In flue gas purification where a low pressure drop is essential,
`lime or CaC03 slurries are used in spray contacting with the gas
`phase (see [12]). Sometimes it is meant to evaporate the water
`phase but also simple slurry scrubbing is envisaged.
`
`4. HYDRODYNAMICS OF SLURRY REACTORS
`Hydrodynamics of slurry reactors includes the study of minimum gas
`velocity or power input to just suspend the particles (or to fully
`homogeneously suspend the particles), bubble dynamics and the hold-
`up fractions of gas, solids and liquid phases. A complicating pro-
`blem is the large number of slurry reactor types in use (see fig.
`I) and the fact that most correlations available are at least par-
`tially of an empirical nature. We will therefore restrict ourselves
`to sparged slurry columns and slurries in stirred vessels° A second
`problem is the difference with three phase fluidizationo To avoid
`too much overlap we will only consider those cases where superfi-
`cial liquid velocities are so low that its contribution to suspen=
`sion of the particles is relatively unimportant°
`
`4olo Minimum suspension criteria
`For the ~design of slurry reactors, whether agitated only by the
`flowing gas or assisted by one or more stirrers, the conditions at
`which the particles are just suspended are very important. If the
`overall contacting efficiency with the solids is considered, it
`increases rapidly if more and more solids are suspended but once
`the condition of minimum suspension is passed and no solids are
`laying on the bottom of the vessel, further increase of the energy
`input for suspension only moderately increases the contacting effi=
`ciencyo Another aim could be to reach for complete homogeneous
`suspension. This condition is generally difficult to reach for many
`applications and an arbitrary criterium should then be introduced°
`Therefore, generally only a minimum suspension criterium is consi-
`dered.
`In ~a~g~=£~!~ with a stagnant liquid medium a minimum suspen-
`sion gas velocity can be defined at which all solid particles are
`just suspended° Three different experimental techniques have been
`used to obtain an operational definition of minimum suspension gas
`Velocities:
`io Particles do not remain on the bottom of the vessel for more
`than say one or two seconds. This has to be checked by visual
`observation through a transparent bottom°
`2o The lowest gas velocity at which the pressure drop over the
`liquid phase indicates that all particles ’are just suspended
`(see Fig. 2)°
`
`o ~ 6
`
`0 0
`
`ing pattern and mechanical
`
`disc contactor
`ase spray column
`
`gas flow
`
`)etween the stages, the
`.an also be manipulated
`advantageous in relation
`cation of residues)° If
`a with the gas phase is
`3rs of high L/dt values,
`3untercurrent contacting
`gas hold-up, upflow and
`and/or to improve mass
`added to the system. In
`~t the solid can be kept
`t of gas by means of a
`
`Akermin, Inc.
`Exhibit 1018
`Page 12
`
`

`
`470
`
`AP
`
`onset of slugging
`
`ws
`
`Fig. 2 Experimental technique to
`find the minimum gas velocity for
`particle suspension in gas-liguid-
`solids sparged columns..
`
`3. Observing the solids concentration at a given position above the
`bottom of the vessel where it passes through a maximum value or
`a discontinuity [24].
`Narayanan et alo [25] used a visual observation technique and have
`given relations for the minimum gas velocity to suspend the parti-
`cleso To obtain a theoretical basis, they compared a pick-up velo=
`city previously derived [25] on the basis of a force balance:
`
`vi = (2g(pS=PL)[~L + wsLsL ])~
`Ps+WsPL~
`
`which was taken equal to the liquid velocity in upward direction
`for which it was derived [25]:
`
`vu = uG + 1/3 (2gLSL~G(~))0°5
`
`(2)
`
`An empirical relation was introduced to relate the bubble hold=up
`with the superficial gas velocity uGo
`£G = 6°2 uG for uG < 0°067 m/s
`~G 0"765u~°38 0o067<UG<0O22 m/s
`
`(3)
`
`Combining equations (i), (2) and (3) results in a minimum gas velo-
`city necessary to suspend the particles: uG (min, Theoro)o However,
`it was clear from comparison with the experimental values that a
`correction factor should be introduced for particle concentration
`and vessel diameter, dt:
`
`d
`-lOWSuG
`uG(min, actual) = 4o3(U~0~)ne
`
`(min, Theoro)(WS < 0oi) (4)
`
`uG (min, actual) = 1
`
`in which n = 0.2 if
`
`These correlations ~
`of 0°006 Ns/m2 with
`Roy et alo [26], ap
`large variety of gas
`systems and particl~
`coal, catalyst pow-
`ups which correspon(!
`just suspended at a
`that in many cases t~
`rendering the comp]
`slurry columns° A
`fuku et alo [28]o P;I
`draft tube applied
`rate of gas was requ
`In stirred vessels
`required to keep th
`criteria for a thre
`those obtained in t!
`by Chapman et alo [
`the gas phase which
`absence of a gas ph~
`studied by Zwieteri~
`rials and experimen
`relatively old (195
`sion, its results a~
`to check new theoret
`
`~min= C1 v0°I dp0°2
`
`The exponents were f
`size~ impeller clea
`dimensionless consta
`metry (e.g. on dt/ds
`leading to minimum ~
`stirrer geometries°
`one given by Nienow
`few other investiga
`phenomena has been ~
`close to those of Z~.
`successful theoretic
`that the suspension
`tain critica! scale°
`of the particle size
`
`Akermin, Inc.
`Exhibit 1018
`Page 13
`
`

`
`ig. 2 Experimental technique to
`ind the minimum gas velocity for
`article suspension in gas-liquid-
`olids sparged columns.~
`
`iven position above the
`~ugh a maximum value or
`
`tion technique and have
`to suspend the parti-
`3mpared a pick=up velo=
`force balance:
`
`(1)
`
`ty in upward direction
`
`(2)
`
`.ate the bubble hold-up
`
`(3)
`
`in a minimum gas velo-
`(min, Theor.). However,
`rimental values that a
`particle concentration
`
`Theoro)(WS < 0oi) (4)
`
`=
`uG (min, actual) 1.25(DqD~O]T)n
`
`uG (min, Theoro) (wS > 0oi)
`
`in which n = 0.2 if dp<100 ~m and n = 0.5 if dp>200 ~m.
`
`These correlati