1-s2.0-s0926860x00008516-main

8/22/2019 1-s2.0-S0926860X00008516-main

1/23

Applied Catalysis A: General 212 (2001) 129151

Consequences of catalyst deactivation forprocess design and operation

S.T. SieLaboratory for Chemical Technology, Delft University of Technology, Julianalaan 136, 2628 BL, Delft, The Netherlands

Abstract

Catalyst deactivation has important consequences for the design of a process and the way it is operated. The nature of

the deactivation, and in particular the question whether it can be reversed under conditions which are compatible with the

normal operation or whether a separate regeneration treatment of the catalyst is required to restore its activity, as well as

the time-scale of the deactivation determine the type of technology that is feasible and process options like reactor type andprocess configuration. This relationship between deactivation behaviour and process lay-out forms the subject matter of the

present paper.

Thegeneral principles that guide thechoices of process type andparameters areillustratedin more detailwith examples from

the fields of catalytic reforming of petroleum naphtha and hydroprocessing of petroleum residues. In these fields, different

catalyst deactivation mechanisms are operative and catalyst deactivation rates can vary widely depending upon feedstock

and process parameters. Consequently, different reactor technologies and process configurational choices are possible. The

relation between catalyst deactivation behaviour and process design and operation can be viewed from two sides: on the one

hand, the deactivation behaviour may dictate the choice between viable process options and may provide an incentive for the

development of novel technology that can cope optimally with the demands set by the deactivation of the catalyst. On the other

hand, the introduction of novel technological options may widen the scope of a process, e.g. by opening the possibilities to

apply novel catalysts or to operate under unconventional conditions that lead to a more economic or otherwise better process,

possibilities that were previously barred by catalyst deactivation. 2001 Elsevier Science B.V. All rights reserved.

Keywords: Catalyst deactivation; Regeneration; Process design; Process configuration; Reactors; Catalytic reforming; Residue hydroprocessing

1. Introduction

Deactivation of the catalyst during the course of a

process is often unavoidable and has to be reckoned

with in the design of a process. Catalyst deactivation

represents a technical as well as an economic factor in

the process and its effect on the performance of a given

type of reactor or the economics of a certain process

has been the subject of a number of quantitative analy-

ses, e.g. [1,2]. However, catalyst deactivation may af-

fect the process design in an even more principal way:the selection of process options such as the process

configuration, reactor type, and mode of operation of

an industrial process can be influenced or even dictated

by the deactivation of the catalyst [3,4]. The nature of

catalyst deactivation, the possibility of regaining lost

catalyst activity either during the operation of the pro-

cess or in a separate regeneration step, and the speed

of deactivation are factors that determine these process

options. The present article reviews these factors and

seeks to provide a link between deactivation behaviour

and the choice of the appropriate process technology.

The general principles set out will be illustrated

with some examples, viz. from the fields of catalyticreforming of petroleum naphtha and the hydropro-

cessing of residual oils. In both processes, different

0926-860X/01/$ see front matter 2001 Elsevier Science B.V. All rights reserved.

PII: S 0 9 2 6 - 8 6 0 X ( 0 0 ) 0 0 8 5 1 - 6

8/22/2019 1-s2.0-S0926860X00008516-main

2/23

130 S.T. Sie / Applied Catalysis A: General 212 (2001) 129151

mechanisms of catalyst deactivation are operative and

these are coped with in different ways. Moreover,

depending upon reaction conditions chosen or char-

acteristics of the feedstock to be processed, the speed

of catalyst deactivation can vary widely and conse-

quently lead to different process configurational and

operational choices.

2. Nature of catalyst deactivation and remedial

actions

Catalysts can lose their activity during a process

for a variety of reasons. Most common reasons are

poisoning or reaction inhibition by impurities in the

feed or by reaction byproducts, deposition of poly-

meric material including coke on the catalyst as

a result of side or consecutive reactions, and loss of

catalyst dispersion by sintering of small particles of

the active material. In addition, catalysts may becomedeactivated by loss of active components by leaching

or vaporisation, or by changes in their porous texture.

Fig. 1. Flow scheme of BPs paraffin isomerisation process, showing the dosing of organic chloride to compensate for catalyst deactivation

by stripping of hydrogen chloride [5].

Changes in porous texture that can affect the perfor-

mance of a catalyst are, for instance, loss of specific

surface area through sintering of the carrier or loss of

permeability through plugging of pores.

An important distinction between the deactivation

processes is whether they are reversible or irreversible

under the conditions of the process. A reversible de-

activation caused by leaching of active material from

the catalyst in a continuous process can be coped

with by adding the leached product to the feed. At the

correct dosing rate, the supply and leaching rates are

in balance so that the net loss of active material from

the catalyst is zero. An example of such a process is

the isomerisation of paraffins with a catalyst based on

chlorided alumina. Loss of chlorine from the catalyst

as a result from some hydrolysis by moisture is coun-

teracted by feeding hydrogen chloride or a hydrogen

chloride generating compound to the feed to make up

for the hydrogen chloride lost, see Fig. 1.

An irreversibly deactivated catalyst has either to bediscarded or to be subjected to a reactivation treat-

ment to restore its performance. The latter treatment

8/22/2019 1-s2.0-S0926860X00008516-main

3/23

S.T. Sie / Applied Catalysis A: General 212 (2001) 129151 131

Fig. 2. Periodic catalyst reactivation during catalytic hydrodewaxing of Arabian heavy way distillate of 5565

F pour point [6].

Fig. 3. Viable process technologies as determined by rapidity of catalyst deactivation.

8/22/2019 1-s2.0-S0926860X00008516-main

4/23


is termed regeneration or rejuvenation. Rejuvenation

is a term commonly used for a treatment which can

be carried out under conditions that are rather similar

or not principally different from the operating condi-

tions. For instance, polymeric material that had been

deposited on a catalyst under relatively mild condi-

tions of hydroprocessing may still be sufficiently reac-

tive to be hydrocracked under somewhat more severe

conditions. Thus, the deposited polymeric material

may be removed in the form of cracked fragments

at somewhat increased temperature and/or increased

hydrogen rate (hydrogen stripping). An example of

such a catalyst rejuvenation is shown in Fig. 2.

When the deactivation process is irreversible and

reactivation requires conditions that are incompatible

with those of the main process, the catalyst has either

to be regenerated in a separate step or it must be

disposed off. Common examples of such irreversible

deactivation processes are deposition of coke on

the catalyst and sintering of metals. Coke depositedon the catalysts at high temperatures in the form of

a carbon-rich polyaromatic material is generally too

unreactive to be hydrocracked in a hydrogen stripping

step and has to be removed by oxidation to carbon

oxides. Sintering of active materials can be undone

by redispersing, e.g. sintered platinum on alumina

may be redispersed by treatment with chlorine. These

decoking as well as redispersion steps are carried

out in an oxidative atmosphere, which is incompati-

ble with the reductive atmosphere of a hydrocarbon

conversion process.

If the regeneration option is not chosen, spent cat-

alyst has to be disposed off. In catalyst disposal, thechoice between dumping and working up of spent

catalysts for recovery of valuable materials is dictated

by the economics of materials recovery and does not

affect the main process other than via the catalyst

replacement costs. Regeneration of a catalyst can be

carried out either ex-situ or in-situ. In the former

case, the regeneration may be carried out in a separate

facility at a different location from the main process

and may even be performed as a service by an outside

company. In the case of in-situ regeneration, the re-

generation facilities are much more an integral part of

the process installation. The regeneration procedure

may even be carried out in the same reactor as used inthe main process, but with the latter temporarily out of

regular service during the regeneration campaign (this

mode of operation is often termed semi-regenerative

operation). Another possibility is to install more than

one reactor, with each reactor being alternately op-

erated in a process mode or in a regeneration mode

(swing operation). Yet another possibility is not to

keep a batch of catalyst in the same vessel all of

the time, but to circulate catalysts between reaction

and regeneration vessels (continuous regeneration).

This implies that the catalysts must be able to move

in and out of a reactor, which is made possible by

moving-bed or fluid-bed technology.

3. Rate of catalyst deactivation

In batch processes, the rate of catalyst deactivation

determines catalyst consumption and is in general

only a factor of economic importance. Aside from the

Fig. 4. Compensating for catalyst deactivation in catalytic reform-

ing by increasing the reactor temperature during the run [7].

8/22/2019 1-s2.0-S0926860X00008516-main

5/23


Fig. 5. (A) Progress of deactivation front through a fixed-bed of Mobils methanol-to-gasoline (MTG) process, as indicated by the movement

of the exothermic reaction zone [8]. Note the time-scale of the deactivation. (B) Flow scheme of the fixed-bed MTG process, showing thearrangement to switch individual reactions from the operating to the regeneration mode [8]. (C) Cumulative gasoline yield of individual

reactors as a function of time in the fixed-bed MTG process [9].

8/22/2019 1-s2.0-S0926860X00008516-main

6/23


Fig. 5 (Continued).

fact that it may have an impact on the sizing of cat-

alyst storage and dosing vessels and the capacity of

separating equipment to remove spent catalyst from

the products, the rate of catalyst deactivation hardly

affects the process configuration and reactor choice

in batch processes. The situation is quite different,

however, for continuous processes where catalyst de-

activation rates may determine the choice of process

options in a more principal way. This is especially true

for processes in which the conditions of operation and

regeneration are incompatible, so that the regenerationhas to be carried out as a separate step. Fig. 3 shows

the relation between possible reactor technologies

and rapidity of catalyst deactivation for these cases.

If the deactivation rate in a fixed-bed reactor is

sufficiently low, no special on-site facilities for re-

generation are required, and after an operating period

spent catalyst can be discarded or regenerated else-

where for reuse. This situation pertains if the life of

the catalyst is about a year or longer. Assuming an

average of 10 days downtime in 1 year for a catalyst

change in a process operating with oxidisable mate-

rial under pressure (which change involves cooling

down, depressurising, and opening of the reactor, re-moval of spent catalyst, reloading with fresh catalyst,

closing the reactor, repressurising and reheating), the

availability of the unit is reduced by about 3%, which

is generally considered acceptable. The loss of avail-

ability can be further reduced if the time of catalyst

change can be made to coincide with the generally

required periodic safety inspection of the unit.

When catalyst life becomes shorter, e.g. about half

a year, dedicated facilities for on-site regeneration be-

come of interest, particularly when it concerns an ex-

pensive catalyst. In the so-called semi-regenerative

mode of operation, the catalyst remains in the reactor,

which is taken out of service during the regenerationcampaign. The required facilities for carrying out the

regeneration, such as compressors for inert gas circu-

lation and dosing of air may be a permanent part of

the unit or be installed on a temporary basis.

With the relatively slow rate of catalyst deactiva-

tion in the above non- and semi-regenerative process

modes, the effect of deactivation during a catalyst life

cycle is generally counteracted by raising the reactor

temperature during operation so as to maintain the

desired level of conversion or quality of the product.

An example of such a temperature programme is

shown in Fig. 4.

With still shorter catalyst lives, i.e. less than a fewweeks, the required frequency of regeneration dictates

the installation of dedicated permanent facilities for

8/22/2019 1-s2.0-S0926860X00008516-main

7/23


Fig. 6. Fluid-bed variant of the methanol-to-gasoline process [10].

regeneration which are an integral part of the processinstallation. The catalyst can either remain within

the same reactor, which is switched between normal

operation and regeneration modes (swing operation)

or be transported as a more or less continuous flow

of solids to a separate regeneration vessel and back

to the reactor after having been regenerated (contin-

uous regeneration). In the swing-type of operation

two or more reactors are generally installed to ensure

an almost uninterrupted product flow. With multiple

reactors installed, it is customary to operate them in

a staggered fashion so that at any moment in time a

mix of products from differently aged catalyst-beds

is obtained. This merry-go-round way of operationensures an almost constant product quality, notwith-

standing the aging of the catalyst. An example of

such an operation is shown in Fig. 5AC, pertaining

to Mobils methanol-to-gasoline (MTG) process. In

this process, the catalyst (ZSM-5 zeolite) deactivates

as a result of deposition of carbonaceous material in

the course of a few hundred hours and is regenerated

by an oxidative treatment [8,9].

The movability of catalyst in continuous regener-

ation systems sets a limit to the achievable catalyst

circulation rate in practice and, therefore, determines

a minimum catalyst lifecycle. Fluidised catalyst can

be moved much faster than catalyst particles in amoving-bed. Another important limiting factor is the

operating pressure: when the process operates in a

Fig. 7. Scheme of a fluid catalytic cracking unit with cracking in a

dilute fluidised phase riser and catalyst stripping and regeneration

in a dense fluidised-bed.

8/22/2019 1-s2.0-S0926860X00008516-main

8/23


reducing atmosphere at elevated pressure, the required

positive sealing against the oxidative atmosphere in

the regenerator requires the use of sluice vessels or

lock-hoppers. This is in contrast to processes operat-

ing at (near) atmospheric pressure, where pressure dif-

ferences over standpipes or seal legs are sufficient to

ensure effective sealing between spaces with different

atmospheres. Hence, much higher catalyst circulation

rates can be achieved in low pressure processes, and

much shorter catalyst lives are allowed. Fig. 6 depicts

a process scheme for a process at elevated pressure,

viz. the fluidised-bed version of the previously dis-

cussed MTG process [10]. A widely applied process

operating in the fluidised mode at near atmospheric

pressure with a very short-lived catalyst circulating

at a high rate between a riser reactor and a bubbling

fluidised-bed regenerator is the modern fluid catalytic

cracking process (FCC process, see Fig. 7).

4. Catalytic reforming of naphtha

4.1. Basic reactions and deactivation processes

Catalytic reforming of naphtha, a petroleum frac-

tion boiling between about 80 and 180C is a very

Fig. 8. Relation between catalyst chloride content and the water/

hydrogen chloride ratio in the gas for a typical reforming catalyst.

Fig. 9. Effect of operating pressure on the stability of a reforming

catalyst [11].

Fig. 10. Effect of operating pressure on yields in reforming of

naphtha with the Rheniforming process [7].

8/22/2019 1-s2.0-S0926860X00008516-main

9/23


important refinery process in the production of

high-octane gasoline. The process produces aromatic

hydrocarbons by dehydrogenation of cyclohexanes,

by dehydro-isomerisation of cyclopentanes, and by

dehydrocyclisation of alkanes. These reactions are

catalysed by a bifunctional (metal + acid) catalyst.

The metal function is generally represented by finely

dispersed platinum or an alloy of platinum with a

second metal (mostly Sn, Re, Ge, Ir), while the acid

function is provided by chloriding the alumina car-

rier. In addition to the formation of aromatics, the

production of hydrogen is important since for many

refiners the catalytic reformers are the main and often

only source of the hydrogen needed for hydrotreating

processes. A generally undesired side reaction is the

formation of gaseous hydrocarbons by hydrocrack-

ing, which lowers the reformate yield and adversely

affects the yield and purity of the hydrogen produced.

The catalytic reforming process is carried out at

elevated temperatures and moderately high pressures

Fig. 11. Flow scheme of a typical semi-regenerative reformer [12].

in the presence of circulating hydrogen. Catalyst deac-

tivation is generally caused by stripping of hydrogen

chloride from the catalyst under operating conditions,

and by deposition of carbonaceous material on the

catalyst. In addition, the dispersion of the active metal

may be negatively affected, e.g. by high temperatures

especially during carbon burn-off.

The loss of acidity by stripping of hydrogen chlo-

ride is a reversible deactivation and is generally coun-

teracted by dosing of a chlorine-containing compound

to the feed, e.g. hydrogen chloride or an organic com-

pound that is easily converted to hydrogen chloride in

the reactor, such as dichloro-ethane. The dosing rate

depends on the (actual + potential) water content of

the feed and the desired steady-state chloride level on

the catalyst, see Fig. 8.

The rate of carbon deposition and, therefore, the

speed of the irreversible deactivation for a given

catalyst and feedstock depends on the operating con-

ditions, in particular on the operating pressure. At

8/22/2019 1-s2.0-S0926860X00008516-main

10/23


relatively high pressures (e.g. 2540 bar) less car-

bon is deposited and consequently long catalyst lives

are obtained. However, the high pressures are un-

favourable for the dehydrogenation reactions that

produce the desired aromatic hydrocarbons and hy-

drogen. Moreover, due to an increased degree of

hydrocracking, liquid yields are lowered and the

hydrogen purity is decreased.

At low pressures (e.g. 515 bar) the formation of

aromatics is enhanced, and liquid yield, hydrogen

yield and hydrogen purity are all improved as com-

pared with the operation at higher pressures. How-

ever, these advantages of low pressure operation are

obtained at the cost of a decreased catalyst stability.

Fig. 9 compares the stabilities of the reforming cat-

alyst at different pressures, while Fig. 10 shows the

effect of pressure on yields.

4.2. Technological options

The relatively low catalyst deactivation rates at

high pressures makes a semi-regenerative type of

operation a logical choice. In the semi-regenerative

Fig. 12. Flow scheme of a fully-regenerative reformer [12].

catalytic reformer, the catalyst is placed in fixed-bed

reactors which are operated adiabatically. Because

of the strong endothermicity of the process, three to

four reactors are generally applied in series, and the

process stream is reheated between the reactors, as is

shown in Figs. 11 and 12. In the course of the run,

temperatures are gradually increased so as to main-

tain the octane quality of the liquid product on target

(see Fig. 4). The run ends at a point in time where

the required temperature reaches the design limit of

the unit, or when the yields have become uneconomi-

cally low. The unit is taken out of normal service and

with the required safety precautions (such as using

blind flanges for isolating certain parts) the catalyst

in the respective beds is subjected to a carbon-burn

regeneration. This step may, if required, be followed

by an oxidative chlorination treatment to redisperse

agglomerated platinum or to re-establish the required

interaction between platinum and the other alloying

metals. Thereafter, the unit can be brought back to thenormal operating mode and after catalyst reduction,

optional presulfiding and adjustment of the chloride

level on the catalyst a new operating cycle is started.

8/22/2019 1-s2.0-S0926860X00008516-main

11/23


Fig. 13. Flow scheme of U.O.P.s continuous regenerative reforming process [12].

The semi-regenerative catalytic reformer has the

advantage of simplicity of plant hardware, but

the relative long interruption of the production and the

laborious character of the regeneration are clear dis-

advantages. For operation at the yieldwise attractivelow pressures, the required frequency of regeneration

renders this type of operation no longer viable, and

either the fixed-bed swing reactor type of operation

or the moving-bed type of operation with a sepa-

rate, dedicated reactor for regeneration becomes a

logical choice. In both types of operation the regen-

eration facilities are an integral part of the process

installation.

Fig. 12 is a simplified flow scheme of a swing-type

fully regenerative reformer. The unit features a num-

ber of reactors, each of which can be operated in

a regeneration mode during a certain period. The

other reactors operate in the main processing modeduring that same period. The reactors are switched

between the two modes by means of values that are

actuated automatically according to predetermined

sequence.

At a relatively low operating pressure with larger

gas volumes to be circulated, the pressure drop over

the catalyst-beds becomes a more constraining factorand the reactors are, therefore, designed for low pres-

sure drop accross the beds. Designs that feature a low

pressure drop are shallow, large diameter fixed-beds

that are accomodated in spherically shaped pressure

vessels, or radial-flow reactors in which the catalyst

occupies the annular space between concentric cylin-

drical screens through which the process stream flows

in a radial direction.

Fig. 13 shows a simplified flow scheme of the

continuous regenerative catalytic reformer (CCR) of

Universal Oil Products Co. (UOP) [13]. The reactors

are radial-flow reactors through which the catalyst

moves downward by gravity. In the reaction section,successive reactors are stacked above each other. Cat-

alyst flows by gravity from the top of the stack to

8/22/2019 1-s2.0-S0926860X00008516-main

12/23


Fig. 14. Comparison of catalyst activity and stability in processing of a distillate and residual feed from the same Middle East crude oil

over a conventional hydrodesulfurisation catalyst. Note the much more severe operating conditions for processing the residual feed.

the bottom. Discrete portions of the spent catalyst are

removed out from the bottom of the last reactor via a

lock-hopper system and are transferred pneumatically

to the top of the regenerator reactor which is like-

wise a radial-flow reactor. The regenerated catalyst

is taken out from the bottom of the regenerator via a

lock-hopper and transported pneumatically to the top

of the first reactor where it is reduced before being

reused in the reforming process.

The continuous regenerative catalytic reforming

process can be applied under conditions that lead to

Table 1

Characteristics of some heavy distillates and residues

Feedstock Sulfur

content (wt.%)

Asphaltenes

C5, (wt.%)

Nickel

(ppm, weight)

Vanadium

(ppm, weight)

Kuwait vacuum gasoil 2.9 Nil

8/22/2019 1-s2.0-S0926860X00008516-main

13/23


Fig. 15. (A) Carbon on catalyst as a function of catalyst age. The data were obtained from fixed-bed experiments with a large liquid

recycle, terminated after processing different quantities of feed over a batch of catalyst. The gradientless reactor in these experiments isequivalent to a continuous stirred tank reactor. (B) Metals on catalyst as a function of catalyst age, from the same experiments as in

Fig. 15A. Residue B contains about four times as much metals as residue A.

8/22/2019 1-s2.0-S0926860X00008516-main

14/23


Fig. 16. Steady-state level of carbon on a catalyst as a function

of hydrogen pressure.

5. Hydroconversion of residual oils

5.1. Deactivation mechanisms

Catalyst deactivation is a major problem in the cat-

alytic treatment with hydrogen of the residuals from

atmospheric or vacuum distillation of petroleum, for

the purpose of sulfur removal (hydrodesulfurisation)

or cracking to distillates. Whereas the established Co/

Mo/alumina or Ni/Mo/alumina catalysts can remain

active for several years in hydrotreatment of distillates,

their activity decays in a matter of weeks or months

Fig. 17. Radial concentration profiles of vanadium as determined from electron microprobe scans along the diameter of catalysts used in

hydroprocessing of residues. The profiles A, B and C are typical for the catalyst types distinguished by the same letters in Table 2.

Fig. 18. Correlation between maximum uptake capacity for vana-

dium and utilised pore volume of catalysts. PV: specific pore

volume; F is an effectiveness factor as determined by electron

microprobe analysis.

when used in the processing of long (atmospheric) or

short (vacuum) residues, see Fig. 14.

Residual oils differ from distillate oils in that they

contain much more material with condensed polyaro-

matic structures and organically bound metals such as

nickel and vanadium, see Table 1. These two factors

8/22/2019 1-s2.0-S0926860X00008516-main

15/23


are responsible for the much lower activity and stabil-

ity of the catalyst in residue procesing. The condensed

polyaromatics as present in so-called asphaltenes

(material insoluble in apolar solvents like n-pentane

or n-heptane) can act as precursors for coke, and cat-

alyst coking in residue processing is, therefore, much

more severe than in processing of distillates. However,

the deactivation by coking is reversible. This can be

seen from Fig. 15A, which shows that after an intitial

build-up period a steady-state coke level is estab-

lished on the catalyst. For a given feed and catalyst,

this steady-state coke level is inversely proportional

to the hydrogen pressure, as is shown in Fig. 16.

The important consequence of the reversibility of

deactivation by coke is that one can cope with it by

operating at a sufficiently high hydrogen pressure.

Thus, the amount of coke on the catalyst is controlled

at a level where the residual activity of the catalyst

is still sufficient for the intended duty. However, this

remedy is not effective for the deactivation caused by

Fig. 19. Effect of pore size on useful life and hydrodesulfurisation activity of a set of catalysts with narrow monomodal pore size

distributions, tested under standard conditions.

the presence of metals. Even though the absolute con-

centratios of metals in the residues are much lower

than that of carbonaceous matter, their role in catalyst

deactivation is much more serious. The metal-organic

bonds by which these metals are attached to organic

structures such as porphyrin groups that are also

part of the asphaltenes are broken in the catalytic

hydrotreatment so that in the presence of hydrogen

sulfide the metals deposit on the catalyst surface as

metal sulfides such as Ni3S2 or V2S3. Catalyst de-

activation by the deposition of nickel and vanadium

sulfides is irreversible in contrast to coking.

An important feature of the deactivation by metals

is that the process of deposition of metal sulfides is

not stopped when the original active sites of the cat-

alyst are covered with deposited metal sulfides. The

reason for this is that the process is autocatalytic in

that it is also catalysed by the deposited metal sulfides

themselves [14]. Hence, metals from the feed continue

accumulating on the catalyst as is shown in Fig. 15B.

8/22/2019 1-s2.0-S0926860X00008516-main

16/23


Table 2

Types of catalysts for hydroprocessing of residuesa

Catalyst type A B C

Pore size Wide Intermediate Narrow

Metals penetration Deep Intermediate Shallow

Metal storage capacity High Medium Low

Stability Very good Fair Poor

HC/HDS activity Low Fair HighApplication HDM catalyst HC/HDS catalyst HC/HDS catalyst

Location in reactor train Front end Middle Tail end

a HDM: hydrodemetallisation; HC: hydroconversion; HDS: hydrodesulfurization.

Another feature of the feed demetallisation reaction

is that it is subject to pore diffusion limitation, as is

understandable considering the bulkiness of the as-

phaltene structures. As a consequence, the metals are

generally deposited in an outer zone of the catalyst

particle, as is demonstrated by electron microprobe

analyses of used catalysts, see Fig. 17. Depending

upon the degree of pore diffusion limitation, the met-als deposition zone within the catalyst particle may

be shallower or broader.

Fig. 20. Quick catalyst replacement (QCR) system according to Shell design [15]. Left: loading of reactor with fresh catalyst. Right:

unloading spent catalyst.

The end of the catalyst life is reached when the

catalyst pores in the deposition zone are completely

blocked by deposited metal sulfides. Hence, the

ultimate life of the catalyst will be determined by the

pore volume available for storage of sulfidic metal de-

posits. Fig. 18 shows that for a given feed processed

under the same conditions, the lives of different cat-

alysts correlate well with the pore volume that isavailable for accomodating the deposited metals from

the feed.

8/22/2019 1-s2.0-S0926860X00008516-main

17/23


5.2. Technological consequences: catalyst aspects

The mechanism of catalyst deactivation by pore

plugging as described above has important conse-

quences for catalyst selection and design. Conven-

tional distillate hydrotreating catalysts generally have

relatively narrow pores so as to maximise their spe-

cific surface area and thereby their activity. With

such narrow-pore catalyst, the demetallisation reac-

tion is strongly limited by pore-diffusion, and the

shallowness of the metal sulfide deposition zone in

the catalyst particles causes the catalyst to be very

shortlived. Catalyst life can be extended by allow-

ing deeper penetration of metals containing species

by increasing the effective pore size of the catalyst.

However, this goes at the expense of catalyst activity,

since the non-fouled inner part of the catalyst particle

becomes relatively smaller. The opposite effects of

pore size variation on catalyst stability and activity

are illustrated by Fig. 19. It follows that a compromisehas to be reached between catalyst life and activity

for the hydroconversion reaction such that neither

of them are maximal, but both sufficient for the

intended duty.

In a fixed-bed reactor as generally applied for dis-

tillate hydrotreating the removal of metals from the

feed causes a decrease of the metals concentration in

the process stream. This descending axial metals con-

centration profile implies that no single catalyst can

be optimal in the whole reactor, but that in theory the

metal sulfides accomodating capacity and the residual

activity have to be tailored to the local conditions

within the reactor. This leads to the concept of usinga combination of different catalysts rather than a sin-

gle one in the reactor or reactor train. A highly active

catalyst (with necessarily limited metals tolerance)

may be applied at the downstream part of the reactor,

where the metals content is low due to the demetalli-

sation action of the preceding catalysts. This tail-end

catalyst may be preceded by a catalyst which has

a greater tolerance towards metal sulfide deposition

obtained at some cost of activity. In the front end, a

highly metal tolerant catalyst capable of coping with

the full metal content of the feed, but necessarily with

a limited activity for the main hydroprocessing reac-

tion can be used. Since its main function would be theremoval of metals from the feed so as to protect the

downstream catalyst-beds, this type of catalyst can

be termed a demetallisation catalyst. These different

types of catalysts are distinguished in Table 2.

5.3. Technological implications: reactor aspects

Fixed-bed reactors operated in the trickle-flow

regime as used in hydrotreating or hydrocracking of

heavy distillates are attractive on account of theirrelative simplicity. The use of combinations of suit-

ably tailored catalysts as described above allows

sufficiently long run lengths (e.g. 0.51 year) with

fixed-bed reactors except for very demanding cases

of residual oil hydroprocessing.

The demands on minimum run length in practice

can be lessened by reducing the downtime needed for

catalyst change. A quick catalyst replacement (QCR)

design by Shell [15] that avoids the opening and clos-

ing of the reactors and that shortens the cooling and

heating-up periods as well as the catalyst loading and

Fig. 21. Ebullated-bed reactor [16] (reprinted from Fuel Process-

ing Technology, 35 (1993), p. 22, with permission from Elsevier

Science).

8/22/2019 1-s2.0-S0926860X00008516-main

18/23


Fig. 22. Bunkerflow reactor. Adapted from [17].

unloading time is shown in Fig. 20. Catalyst transport

into the reactors occurs as a slurry in a stream of gasoil.

A reactor that allows portions of catalyst to be addedto or taken out of the reactor during normal operation

is the so-called ebullated or ebulliating-bed reactor

[16], depicted in Fig. 21. The catalyst inventory is kept

in a fluidised state by strong upward liquid and gas

streams. With this reactor, catalyst life is no longer an

issue as far as achievable run length is concerned but

it has become an economic factor relating to catalyst

consumption rate.

With the ebullated-bed reactor, conventional cata-

lysts as used in distillate hydrotreating may be used

which can have the advantage of high activity (cf

Table 2, catalyst type C). A disadvantage is the back-

mixed character of the reactor. The large liquid recycle

causes the reactor to behave as a continuous stirred

tank reactor, which is a less ideal reactor for reactions

of positive order. The fluidisation of the solid also

results in solids backmixing, which implies that the

catalyst removed also contains fresh catalyst particles.

The latter disadvantage is absent in the so-called

bunkerflow reactor depicted in Fig. 22, which is a

moving-bed reactor operated in the trickle-flow mode

with the possibility to remove or add portions ofcatalyst during operation by means of valves and

lock-hoppers. A special design of the reactor internals

ensures that the catalyst can move down in plug flow.

The narrow residence time distribution achievable

with this type of reactor is shown in Fig. 23.

The bunkerflow reactor has been specifically de-

signed for application as front-end demetallisation

Fig. 23. Residence time distribution of solid particles in a bunker-

flow reactor. The experimental distribution is about the same as

the theoretical distribution in a series of about 200 mixers, which

indicates a good approach to ideal plug-flow.

8/22/2019 1-s2.0-S0926860X00008516-main

19/23


Fig. 24. Axial deactivation profiles at end-of-run conditions in fixed-beds and during operation in a continuous stirred tank or bunkerflow

reactor. F: feed, P: product, C(f): fresh catalyst, C(s): spent catalyst. The hatched area denotes the deactivated part.

Fig. 25. Relative catalyst life vs. metals content of the feed [18].

8/22/2019 1-s2.0-S0926860X00008516-main

20/23


reactor. Operating conditions and catalyst addition/

withdrawal rate can be chosen so as to ensure that

the catalyst taken out is completely spent, whilst still

retaining an acceptable average activity in the reactor.

Thus, better use is made of the metals sulfide acco-

modating capacity of the catalyst as compared with

other reactor systems, see Fig. 24.

Fig. 26. (A) General process scheme for Shells residue hydroprocessing technology [19]. (B) Configuration of the reactor section in

Shells residue conversion technology for feeds of different metals content. Adapted from [19].

5.4. Technological consequences: process

configuration

Since the metals content of the residual feedstock

largely determines the catalyst deactivation rate, dif-

ferent feedstocks with widely differing metals con-

tent may require different catalyst combinations and

8/22/2019 1-s2.0-S0926860X00008516-main

21/23


Fig. 27. Simplified scheme of Shells demetallisation catalyst regeneration process [19].

reactor technologies. This can be seen from Fig. 25,

which shows the relation between feed metals contentand catalyst life.

The wide range of metal contents in residual feeds

and consequently the large differences in deactivation

rates lead to the adoption of different process config-

urations in which different catalysts and reactors are

applied. This is borne out by Fig. 26A and B pertai-

Fig. 28. Reactor configuration in IFPs Hyvahl-S process with swing reactors in the hydrodemetallisation section [20].

ning to Shells residue hydroconversion technology.

Fig. 26B shows the different catalyst and reactor com-binations in relation to the feeds metal content, while

the general process scheme is depicted in Fig. 26A.

The scheme of Fig. 26B also includes the option

of regenerating the spent demetallisation catalyst by

removal of the deposited metal sulfides. With a spe-

cially developed carrier on the basis of silica, removal

8/22/2019 1-s2.0-S0926860X00008516-main

22/23


Fig. 29. Reactor configuration for hydrotreating of resids according to Chevron, featuring the onstream catalyst replacement (OCR) system

for removal of metals [18].

by acid leaching becomes a realistic option, and a

demetallisation catalyst regeneration (DCR) process

has been developed by Shell on this basis, see Fig. 27.

This regeneration option makes the processing of feed-

stock of very high metals content feasible and more

economic compared to the use of demetallisation cat-

alysts on a once-through basis.

An alternative solution for the processing of feed-

stocks with moderately high metals content is to

use a swing reactor system in the demetallisation

stage instead of the bunkerflow reactor. Fig. 28shows the swing reactor system as used in IFPs

Hyvahl-S process [20]. Yet another alternative for

Shells bunkerflow reactor where the gas/liquid feed

and the catalyst move in the same, i.e. downward,

direction is an operation with gas/liquid upflow and

catalyst downflow. The latter feed/catalyst counter-

current operation features in the Hyvahl-M process of

IFP/Asvahl [20,21] and in the online catalyst replace-

ment (OCR) system of Chevron [18], see Fig. 29.

6. Conclusion

Catalyst deactivation is often inherent in the reaction

mechanisms underlying a process and while it can be

minimised by proper choice of catalyst and process

variables, it can in many cases not be entirely avoided.

In such cases, the process technology has to be able

to cope with the deactivation at hand.

The nature of catalyst deactivation, in particular

the question whether the deactivation mechanism is

reversible or not under normal operating conditions,

and the rapidity of catalyst performance decay have a

large bearing on the choice of process options. Aside

from catalyst parameters, these include the type of re-

actor and its hydrodynamic regime (fixed-, moving-,fluidised-bed, co- or counter-current, plug flow or

mixed) as well as the configuration of the process.

The relation between catalyst deactivation and pro-

cess technology can be viewed in two ways: while on

the one hand, the deactivation behaviour of a given

type of catalyst in a certain conversion reaction may

dictate the process technology to be applied, on the

other hand, it is possible that the development of a

novel specific technology widens the scope of the

conversion process. The widening of the scope of a

process may, for instance, encompass the feasibility

to operate under conditions that are more desirable in

terms of yields but inherently lead to faster catalystdeactivation. Or it may allow using alternative cata-

lysts or the processing of more contaminated feeds

8/22/2019 1-s2.0-S0926860X00008516-main

23/23


posing more severe deactivation problems. The devel-

opments of the continuous regeneration technology

in catalytic reforming and the bunker-flow reactor in

residue hydroprocessing are cases in point.

References

[1] A. Sadana, L.K. Doraiswamy, Effect of catalyst fouling in

fixed-, moving-, and fluid-bed reactors, J. Catal. 23 (1971)

147157.

[2] J.M. Douglas, E.K. Reiff Jr., J.R. Kittrell, Process design with

catalyst deactivation, Chem. Eng. Sci. 35 (1980) 322329.

[3] W.H.J. Stork, Molecules, catalysts and reactors in hydro-

processing of oil fractions, in: G.F. Froment, B. Delmon,

P. Grange (Eds.), Hydrotreatment and Hydrocracking of Oil

Fractions, Elsevier, Amsterdam, 1997, pp. 4167.

[4] J.W. Gosselink, J.A.R. Van Veen, Coping with catalyst

deactivation in hydrocarbon processing, in: Proceedings of

the 8th International Symposium on Catalyst Deactivation,

Brugge, Belgium, 1013 October 1999.

[5] B.W. Burbidge, P. Watson, A.H. Richardson, BPs isomerisa-

tion process takes on a new look, in: Proceedings of the NPRA

Annual Meeting, San Antonio, Texas, 2325 March 1975.

[6] N.Y. Chen, R.L Gorring, H.R. Ireland, T.R. Stein, New

process cuts pour point of distillates, Oil Gas J. June, 6 (1977)

165170.

[7] R.C. Robinson, L.A. Fredrickson, R.L. Jacobson, Cat-reform-

ing pressure drops again, Oil Gas J. May, 15 (1972) 124136.

[8] W. Lee, S. Yurchak, N. Daviduk, J. Maziuk, A fixed-bed

process for the conversion of methanol-to-gasoline, in:

Proceedings of the NPRA Annual Meeting, New Orleans,

LA, 2325 March 1980.

[9] D.E. Krohn, M.G. Melconian, in: D.M. Bibby, C.D. Chang,

R.F. Howe, S. Yurchak (Eds.), Methane Conversion, Elsevier,

Amsterdam, 1988, p. 679.

[10] J.E. Penick, W. Lee, J. Maziuk, Development of the methanol-

to-gasoline process, in: Proceedings of the Inter national

Symposium on Chemical Reactor Engineering (ISCRE-7),

Boston, MA, 4 October 1982.

[11] J. Barbier, E. Churin, P. Marecot, J.C. Menezo, Appl. Catal.

36 (1988) 277.

[12] Royal Dutch/Shell Group of Companies, The Petroleum

Handbook, 6th Edition, Elsevier, Amsterdam, 1983.

[13] E.A. Sutton, A.R. Greenwood, F.H. Adams, A new promisingconcept: continuous platforming, Oil Gas J. May, 22 (1972)

5256.

[14] S.T. Sie, Catalyst deactivation by poisoning and pore

plugging in petroleum processing, in: B. Delmon, G.F.

Froment (Eds.), Catalyst Deactivation, Elsevier, Amsterdam,

1980, pp. 545569.

[15] W.C. Van Zijll Langhout, C. Ouwerkerk, K.M.A. Pronk,

New process hydrotreats metal-rich feedstocks, Oil Gas J.

December, 1 (1980) 120126.

[16] R.M. Eccles, Residue processing using ebullated-bed reactors,

Fuel Process. Technol. 35 (1993) 2138.

[17] A.J.J. Van Ginneken, M.M. Van Kessel, K.M.A. Pronk, G.

Renstrom, Shell process desulfurizes resids, Oil Gas J. April,

28 (1975) 5963.

[18] G.L. Scheuerman, D.R. Johnson, B.E. Reynolds, R.W.Bachtel, R.S. Threlkel, Advances in Chevron RDS technology

for heavy oil upgrading flexibility, Fuel Process. Technol. 35

(1993) 3954.

[19] P.B. Kwant, W.C. Van Zijll Langhout, The development of

Shells residue hydroconversion process, in: Proceedings of

the Conference on the Complete Upgrading of Crude Oil,

Siofok, 2527 September 1985.

[20] A. Hennico, G. Cohen, T. Des Courires, M. Espeillac, Hydro-

conversion process for high metal residues, Hydrocarbon Int.

(1993) 3337.

[21] J.P. Euzen, Moving-bed process for residue hydrotreating,

Rev. Inst. Franc. du Ptrole 46 (1991) 517527.

1-s2.0-s0926860x00008516-main

Documents