Documenti di Didattica
Documenti di Professioni
Documenti di Cultura
OCTAVE LEVENSPIEL
This paper concerns the experimental detrrmination of the ordrrs of reaction and
deactivation in the IIth order rate equation which represents the four broad classes
of deactivation of catalyst particlrs : parallrl, series, side-by-side. and independent.
Wr discuss the various rractor types which may he usctf, hoth hatch-solids and
flowing solids. For batch-solids systems WP show that when deactivation is concen-
tration independent then using a batch of fluid or any flow of fluid gives rc,sulta
which are simple to interpret. However, \vhrn dcactivation is concentration dr-
pcndent then only one very particular form of contacting is recommended in that
it alone allows decoupling of conrrntration and activity effects, and their study one
at a time.
In csscnce we show how the methods for ?Ith order homogrnrous reactions c’an
br rxt.ended in a simple way to catalytic systems with deactivating porous solids.
7’ = temperat,ure (“I<)
a = act)ivity of a catalyst’ pellet, defined V = volume of fluid (liters)
in Eq. 3 W = weight, of catalyst (g)
A = gas phase reactant, ,Y.& = fractional conversion of reactant A
C = concentrat,ion (mol/lit,er)
Greek letters
tl = order of deact,ivation, see Eqs. 4-X
F.40 = feed rat,e of reactant A (mol/sec) Q = fractional change in a volume of an
1~= rat’e constant for the reaction of A, element of react,ant, fluid on com-
see Eqs. 4-X (literll/molrl-‘.sec.g rllete conversion
catalyst) ,
7 = p,
kc,"
capacity factor of a reactor
1;’ = x;c’; FAO
/c" = liJ@/J' called the n-eight-time (g cat.sec/
/i,L = rate constant for the deactivation liter of fluid)
reaction, see Eqs. 4-S (literfi’/
sec.moP’) Xubscripts
li,l' = kJ2..yn’ 0 = initial time, or for fresh catalyst
r~ = reaction order for conversion of m = at’ infinite time
reactant A A desirable form of kinetic equation for
W’ = order of concentration dependency describing the behavior of deactivating
of the deactivation porous catalyst particles should have the
LVn = moles of A (mol)
following properties:
1’ = poison
-.-TA’ = rate of reaction of A (mol/sec.g 1. It should be general enough to repre-
catalyst) sent the various types of deactivation
R = product of reaction of A encountered in practice.
t = time (set) 2. Its parameters should be related in a
265
@ 1972 b.v Academic Press, Inc.
266 OCTAVE LEVENSPIEL
simple way with the mechanisms of of feeds which contain poisonous impurities
action of the deactivation.
3. Its form should be such that its A-R -r’~ = kCA”a )
parameters can be found directly by P+PJ + (6)
experiment.
4. Its form should be useful for those Independent deactivation, a result of
concerned with reactor design. Thus, structural modifications or sintering of the
the equation must be easy to use to catalyst surface which is exposed to ex-
predict performance of reactors, opti- treme conditions, and so called because the
mum operating schedules, etc. decay is unaffected by the concentration of
materials in the gas stream but is depend-
A recently proposed form of kinetic
ent on the length of time spent in the high
equation gives promise to satisfy these re-
temperature environment
quirements. Its general form for the rate
of disappearance of any reactant A in the -#A = kCA”a
environs of a representative pellet of cata- (7)
lyst is given by --= da
dt
k clud *
The activity starts at unity and usually However since CA + CR remains constant
drops smoothly with time to zero. for a specific feed, this type of deactivation
Let us illustrate this formulation by reduces to the simple-to-treat independent
using nth order kinetics with the four major deactivation of Eq. 7.
types of deactivation. Thus, if the material In the above expressions the constants
which deposits on the surface to cause it to n, n’, and d are the orders of reaction and
lose its activity is called the poison P, we of deactivation. The rate constants k and
have: Icd are normally taken to be Arrhenius tem-
Parallel deactivation in which reactant perature dependent
decomposes to produce the poison Ic = kOe-EIRr and ka = kdOe-EalRT,
A--+R -?+‘A = kc.&% ) where E and Ed are the activation energies
of the main reaction and of the
deactivation.
Series deactivation in which product de- Let us now turn to the status of this
composes to produce the poison formulation with respect to the above-men-
tioned four properties of a desirable rate
form. First Szepe (1) has shown that this
rate form generalizes many of the previ-
ously proposed kinetic equations for deac-
Side-by-side deactivation, representative tivating catalyst pellets: the linear, ex-
RATER FROM EXPERIMENT FOR DEACTIVATING SOLIDS 267
ponential, hyperbolic, Voorhies, Elovich, lyst pellet, and many more parameters to
and ot,hers (property 1). In addition pre- evaluate. Nevertheless, by proper choice of
liminary studies by Khang (2) indicate that reactor type and design of experiment it
in many cases this formulation is related becomes possible to directly evaluate and
to the mechanisms of deactivation, account- test rate equations of various kinds, just
ing for pore diffusion effects without par- as with homogeneous systems.
ticular difficulty (property 2). Finally, in
recent years numerous studies, referred to Hatch-Solids us Flowing Solids
in Levenspiel (3) have uncovered the fact Experimental Reactors
that this formulation is eminently suitable
for reactor design purposes, allowing simple First, the available experimental devices
analytical solution to optimal scheduling fall into two broad groups, those which use
problems for maximizing conversion and a fixed batch of catalyst, and those which
yield. Such problems have been impossible use a flow of particles through the reactor,
to solve with other equation forms see Fig. 1.
(property 4) . Because of the ease of experimentation
As yet no one has examined how well the setups which use a fixed batch of solids
this form of kinetic equation satisfies the are much preferred, however these can only
third property, a desirable rate form. This be used when deactivation is slow enough
is the aim of this paper. so that the changing fluid concentration
This paper shows how the parameters of can be followed for a sufficiently long time
this formulation, n, n’, d, k, Icd, E, Ed can before exhaustion of the catalyst. Deacti-
be found directly by experiment. Xote that vat,ion in the order of minutes or longer
for any condition of operations the values can use the fixed batch of solids.
of these parameters will account not only Whenever deactivation is very rapid, in
for the intrinsic surface kinetics, but also the order of seconds or less, then a flowing-
for pellet size and geometry, as well as port solids system must be used. Cracking cata-
diffusion effects for both reactants and lysts whose act’ivity half-life is in the order
poisons. of 0.1-l set fall into this category.
We illustrate the search strategy by com- This paper is primarily concerned with
parison with the simple and well known the batch-solids reactor. Here the choice of
homogeneous reaction. There the rate is de- convenient reactor type again splits into
pendent, on two factors, the temperature two classes depending on whether the de-
and composition, and these dependencies activat.ion expressions da/dt of Eqs. 4-7
are normally determined one at a time in are concentration dependent or not. When
separate experiments. Different rate forms concentration independent (or n’ = 0’1 then
are guessed at, starting with the very any type of batch-solids system may be
simplest, first order, second order, and so used; but when concentration dependent,
on, until a reasonable match between the then unless one particular type of reactor
rate form and experiment is obtained. The is used the analysis of the experimental re-
experiments themselves are usually con- sults becomes awkward and difficult.
ducted in a batch reactor and are analyzed We treat these two classes of rate equa-
either by differential analysis (direct test tion in turn.
with experiment of the different,ial rate ex-
pression) or by integral analysis (integra- K atch-Solids.. Finding the Rate Equ,ation
tion of the guessed differential rate expres- when the Deactivation is Independent
sion and a test of the resulting expression of Concentrations
with experiment).
Now -in catalytic systems with deacti- Let us illustrate how to interpret experi-
vating catalysts matters are somewhat more ments from the various batch-solids re-
complicated since we have three independ- actors of Fig. 1, how to manipulate the
ent factors to contend with: the temper- basic performance equations for these re-
ature, composition, and activity of the cata- actors by testing the fit for the simplest
268 OCTAVE LEVENSPIEL
All Fluidized A
bed
Solids: mixed flow
Fluid: questionable
flow pattern
-plug tlow
FIG. 1. Slow deactivation may use a batch of solids in experimentation; fast deactivation requires a flow
of solids.
FIG. 2. Test of the kinetic expressions of Eq. 9 FIG. 3. Test) of the kinetic expressions of Bq. !I
using a batch-solids, batch-fluid reactor. using a bat,ch of solids and mixed flow of fluid of
constant flow rate.
slower than the conversion, then CAm be-
comes very low and difficult’ to measure We should mention that this and the
accurately. following derivations for a batch of solids
Batch-solids: Completely mixed fluid of are based on the pseudo-steady-state as-
constant flow rate (Fig. lc). Inserting the sumption which views that conditions
rate equation of Eq. 9a into the performance change slowly enough with time so that
expression for a completely mixed flow re- the system can be considered to be at
actor gives steady state at any instant. Since a batch
w x* x*
of solids can only be used in experimenta-
-= I = hzA’ (16) tion if deactivation is not too rapid this
FAO --;r A
assumption is reasonable.
and on rearrangement Batch-solids: Plug constant flow of fluid
(Fig. lb ). Here the performance expression
C
Ao=l+ka combined with the rate of Eq. 9a gives
= 1 + kur’, (17)
C,A
experimental device which allows the de- For second order deactivation
activation order to be studied without in- = c.40 - CA + (CAU- CAP’d
terference of concentration effects. The key
then is to keep constant the concentration
term in the da/dt equation of Eqs. 4, 5, For
and 6 while searching for the deactivation
order. After this the concentration effect 7’2 =
can be studied. As examples of the concen-
trations to be kept constant:
For parallel deactivation (Eq. 4) keep
C,%constant; For dth order deactivation
For series deactivation (Eq. 5) keep CR (7’)d--1 = (32)
Cl + c2t.
constant;
Figure 6 then shows how to test for dth
For side-by-side deactivation (Eq. 6)
order deactivation. If the data fall on a
keep CP constant;
st,raight line the guessed mechanism is cor-
For independent deactivation (Eq. 7)
rect, and the slope and intercept, gives the
no concentration need be kept constant’.
constants I?’ and k’d.
The completely mixed flow reactor with After the order of deactivation is found
controlled and changing flow rate of feed then the concentration and temperature de-
can satisfy this requirement for all these pendencies can be determined in turn. The
kinetic forms. procedure for this is similar to that for
Deactivation order for parallel deacti- nondeactivating catalytic systems, is well
vation. We illustrate the method of analysis known, and needs no discussion here.
with general n-, n’-, and d-order kinetics for Deactivation order for other types of de-
parallel deactivation, or Eq. 4. At constant, activation. The above development shows
C, Eq. 9 becomes how to treat parallel deactivation. The
analysis is quite similar for series deacti-
(25)
vation if CA, hence CR is kept constant, and
- 2 = (kdCAn’)ad = Vda”. for side-by-side deactivation if C, and, if
(26)
possible, C, are kept constant. If CA cannot
So for the mixed flow reactor and the rate be kept constant in side-by-side deactiva-
of Eq. 25 we obtain tion, analysis is still not particularly
difficult.
14’
-=
XA XA CA, - CA
F AU Flowing-solids Experimental Reactors
or Devices where the through flow of gas
CA0 - CA and solid, with flow patterns known, arc
T’ = (27)
h/a A
Next, integrating Eq. 26 for various orders WYd-l
much more difficult to build and operate. 2. TO test the general rate form with con-
They are only used when the simpler batch- centration dependent deactivation the most
solids system cannot be used, this means useful reactor set up uses a batch of solids
for reactions with very fast deactivation. and mixed flow of fluid with changing flow
The fluidized bed reactor is probably the rate so as to keep the concentration of the
simplest of these devices, and under steady- pertinent reaction component unchanged
state flow and with flow pattern known with time within the reactor. Thus by con-
(this usually is the difficulty with this type stantly lowering the flow rate in a basket
of set up) integration and testing of some or recycle reactor the activity and concen-
of the simpler rate forms is possible. Leven- tration dependencies can be decoupled and
spiel (5) and Kunii and Levenspiel (4) studied separately.
show how to test rate equations with the 3. The rate equations treated here cer-
following reaction and deactivation orders tainly do not cover all possible cases; how-
ever, they should suffice to show the gen-
n = 1; n’= 01.
) ) d = 1,3 erality of the method of analysis. Thus we
either in a single reactor or in reactor-re- conclude that the integral and differential
generator systems in which catalyst is con- methods of testing rate equations for
tinually recirculated between units. homogeneous reactions can be extended
Extension to other reaction and deacti- without undue difficulty to systems of de-
vation orders can be done. We need not activating catalysts as long as the proper
consider this problem here. reactor type is selected for the experiment
and as long as the rate is expressed in the
DISCUSSION form of Eqs. 4-7.