CHEMICAL ENGINEERING SCIENCE

GENIE CHIMIQUE
VOL. I

JULY

1952

NO. 4

Application of the Temkin kinetic equation to ammonia synthesis in large-scale reactors

D. ANNARLE
Bese8mh

Depertment,

Imperi8l

Ohem~al

(Reuwed

Industnes

Lmuted,

Bdlmghsm

Division

1 March 1952)

Summrry-The
rate of re8otlon in 8n industrial, high pressure ammom
synthesis reactor ls expressed in
terms of the km&lo equation formulated by TEAKKINand PYZHEV. An estunate of cetalyst activity is
derived, effordmg a meane of comp8rmg
tiferent
cat8lysts, and of measurmg cetalyst deterioration.
Optmuun temper8tures gmmg maximum rates of reaction 8re e&mated for certam conditions of catalyst
ectlvlty and mtml ges composltlon.
A reasoneble 8pproach to optimum conditions ls calculated for 8 partrxrhu design of catctlyat bed, fitted
with 8 me8ns of removmg some of the he8t of re8ctlon. Incorporation of the design m a plent reactor showed
th8t the best operatmg conditions conformed closely to those calculated

RBsumB-La vltesse de r&&ion

d8ns un convertrsseur mdustnel pour 18 synth&ee de Iammonlac sous

haute pressron, est exprun6e sous I8 forme de 16quation conetique de TEMKIN et PYZHEV On obtlent
une estlmatlon de lctctlvlt6 catalytique,
fourmssant un moyen de comparer divers catalyseurs
et de
mesurer la b8mm dactunte dun catalyseur.
On obtient une veleur approxun8ttlve des temp6r8tures
optun8, dormant les plus gr8ndes vltesses
de r&x&on, pour dee conditions don&es dactivlt6 catalytlque et de composition mltmle des gaz.
Une 8pproxmWlon
nrlsonnable des comhttlons optun8 8 et+5c8lcuMe pour une disposition partlcuhere
du ht de cetQseu.r,
comportant un moyen devacuer une partie des cctlories fourmes psr 18 rt%ctlon.
Lmtroduotion
de ce duxxxitlf d8ns un convertlsseur mdustriel a montr6 aue les mellleures comhtions
par le calcul
op6r8toues coincident fort bien 8vec oelles p&us

and pressure, experimental reaotlon data can be used

to determine values of the reaction velocity constants.
k2 is the constant usually calculated, probably because
experimenters, such as WINTER [3], first concentrated
on the kinetics of rtmmomla decomposition. k2 should
obey the Arrhenius equation

INTRODUCTION

TEMEINand PYZEEV [l] formulated s, kinetic equation

for ammonia synthesis assuming that the rate determining step was the process of activated adsorption of
nitrogen. A formula for the latter was postulated on
a semi-empirical basis, and the equation for ammonia
synthesis m a static system given as:

k, = b, e- %c/RT
where

E dCC= apparent activation energy for ammonia de-

where

composition cals/mol N, reacting.

NNS= no. of mole. of N,, per unit vol. of catalyst

after time of contact t hours.
PNH,9 PN? PH. = partial pressures of NH,, N,,
H, after tnne t.
kl and k, are the reactron velocity constants for
ammoma synthesis and decomposition, respectively
At pressures above atmospheric, the perfect gas
laws are assumed to hold. kl and k, are related to the
equihbrium constant l&, thus

= Ki

As Kp is

known accumtely [2] over a wide rkge of temperature

* TEMKW ormtted the fector 2 between the rate of NC
adsorptron and NB, synthesis, and neglected the change. 8t
con&&
pressure, of the tot.81 volume of gas with reaction
[see eq (21 =d
12

Chem

R = gas constant.
27 = absolute temperature K.

b, = frequency factor.
TEMKIN and PYZHEV [l] and EMMETTand KUBXMER [4] have apphed the kmetm equation to experrmental data at pressures up to 100 atm for various
space velocities, H,/N, ratios and degrees of approach
to equilibrium. The data were obtamed at three
temperatures, 370,400 and 450 C. The equation was
found to interpret the date satmfaotorrly, except that
k2 decreased with inoreasmg pressure, and some van&tion. 40000 to 55000-was observed m Edeo**.

*+ See

(311.

Esz 9ci Vol 1

145

Tables 4 and 6

ANNABUG'

Apphce;tionof the Temkin km&c equation to ammoniasynthesis111large-ecdereactors ~ngi~&?$%enoe

The present paper describes the flttmg of the

equation to approximate plant data, to fmd a measure

As the catalyst beds were a&abatlc, various slmph-

of catalyst activity and a means of estimating the best

practical reaction conditions The data were obtained

fymg assumptions could be made. Radial conduction of

heat was taken to be neghgble, so that catalyst temper-

at 245 and 300 atm pressure, and covered a temperature

range of 370 to 550 C

atures over any cross-section were considered uniform

TREATMENTOF DATA

Gm and catalyst temperatures were aasumed equivalent

and- longitudinal conduction

EXPERIMENTAL

of heat was neglected

In the case of the smgle bed reactor, for the purposes of analysis, the catalyet bed, of total depth

The plant reactor, from which the data at 245 atm

13 ft, was consldered as a series of small horizontal

were obtained, was an upright cylmdncal forgmg,

contammg a smgle vertical catalyst bed The basket
holdmg the catalyst was lagged and measurements

sections, of depth 6 or 1 ft, such that there were

temperature measurements at the begmnmg and end

showed the heat loss from the converter to be only

of each se&on

about 2% of the total heat evolved

The catalyst
bed was therefore considered as adiabatic

Smce all the heat of reactlon appeared as Increased

heat content of the reaction gas, the amount of reactlon
in each of the above sectlons could be found from the
rise m temperature of the gas over the se&on, usmg
spectilc heat and heat of reaction data The calculation
was commenced at the top sectlon, as the rate and composltlon of the gas were known only at the inlet of the
bed The result for the top section provided mlet gas
data for the second section Repeatmg the calculation
for successive sections gave a senes of amounts of ammonia synthesised, the sum of which should agree urlth
the total ammonia made m the reactor, as given by the
gas rate and the inlet and exit ammonia analyses The
agreement actually obtamed was usually wlthm &6%
In the case of the multi-bed reactor, each bed was
considered as one catalyst section The amounts of
reaction and the gas concentrations were calculated
by the same method as above, makmg allowance for
the extra gas added before each bed The mltlal reaction temperatures m the beds subsequent to the first
were calculated from the exit temperatures of the

Gas at 245 atm pressure, and of constant (imtlal)

gas composltlon flowed contmuously mto the reactor
After bemg heated to reactlon temperature, the gas
passed down over the catalyst bed, where It was
partially converted to ammonia, mth a nse m temperature due to the heat of reactlon
Condltlons were
steady, and the pressure constant, so that at each
level m the catalyst bed, there was a constant temperature and also a constant ammonia concentration,
both quantities mcreasmg with the amount of catalyst
traversed
After leavmg the catalyst, the gas gave up some
of its heat to the mcommg gas, before flowmg out of
the reactor
The vanables measured were the pressure, the inlet
gas rate and composltlon, the concentration of ammonia m the exit gas, and temperatures taken at short
intervals through the catalyst bed by a thermocouple
which could be moved m a vertical sheath near the
axis of the reactor
Except for a more comprehensive exploration of
temperature m the catalyst bed, the measurements
were the normal plant readmgs, and no great accuracy
for them is clanned

precedmg beds and the amounts of coohng produced

by the added coId gaE The agreement between the
calculated amount of ammoma made m the reactor
and the total amount measured was usually between
0 and -7%
The kmetlc equation was adapted to the contmuous
flow system and became

The data at 300 atm were obtamed from a reactor

slmllar to the above, but contaming several adiabatic
beds m senes, between each of which was provlsion
for addmg cold unconverted gas The measurements
used were the pressure, the lmtlal gas composltlon,
the amounts and temperatures of the gas added before each bed, the total amount of ammonia made
m the reactor, and the temperatures at the exit of

(2)
Direction of gas flow
& -1 J
&
---_M3------~- A kg mols/hr

----_,----1-zJ_
_
_

each bed
146

NH, & - J -~~

_._ - -

Entry to
catalyst

+-dW

Vol I
No 4 - 1062

D.ANNAJ.WZ:
Applmatlon
of theTemku~ kuwtlcequation
to ammoxua syuthesls
ti large-scale
reaotore

where A = flow of ammoma m kg mols/hr after the

With regard to the poor s.ucuracy, It LBpornted out

gs,s has traversed w, Me of catalyst.

that the purpose of this work was to find out how much

The psrtial pressures of the gases were expressed

m terms of total pressure P atm, and z the mol fraction
of NH, in the gas after flowing over w, M* of catalyst,

use could be made of the plant data as measured,

without mtroducing more elaborate ma rumentation
f
to improve the accuracy end frequency1 of measurement

and the equatron was rearranged exphcitly m k,, vm.

k

P4a16V

z(l-bz)6

*
(l+z)[L*(l-bbz)"-za]
dw

___--

48

I
TableI. Expersmenti data fw tAe detewns iola
of k,

(3)

Pressure 245atm. Temperature

range4Ok5 66'C. Efffrcsency
range.
009 to 090.
ma? fmetron
Repreaentattve rnhu.? gas unnpo&ton NH
0021
1
0674
H,
0225
K
CH,\ 0048
A'
0032
N1000

where, after traversing w, Ma of catalyst, the gas rate

is V, MS/hr, measured at 1 stm and 20 C, the temperature IS T K and the composition m mol fraction is
z of NH,,

a(1 --bz)

of H,,

L=

(1

and %(l - bz) of N,

::Gq)z

where zes is the equilibrium mol fraction of NH8 at

pressure P atm, and temperature T K. a and b are
constants determmed from the initial gas composition.
As mtegration of the equation for an admbatic
system, where z is a function of T, would be extremely
complex, the followmg approximation was adopted.
The average reaction velocity over each catalyst section, given by A z/A w, was assumed equal to the actual
velocity d z/d w at the erithmetmal average temperature
of the section, for the concentrations of reactants and
products correspondmg to that temperature.
The errors thus involved were, for the c&se of the
smgle bed reactor, generally less than 1 or 2 C m
temperature, and 0901 m the mol fraction of ammonia, and were within the errors of measurement
In the calculation of k2, except for conditions near
eqmhbrium, an error of 2 C in temperature will cause
an error of about 9 % in i&,, whilst an error of 0 001 m

Temp.
2
OC
?wl fmc

the mol fraction of ammoma wdl affect the msgmtude

of ka by less than 10%. As the errors act m opposite
directions, the final maccuracy of k, should not be
more than 10%. Errors of measurement may easily
c&use an additional error of the same order in the
k, values.
The reaction velocity constant has not been calculated for conditions very near equlhbrium, for the
errors produced by errors of measurement are then
very large
Using the approxnnatlon_to apply the kmetm equation to the data of the multi-bed reactor may mvolve
greater errors, because of the more scanty mformation
on catalyst temperatures It le estimated that m the
malority of the examples, the error mcurred m the k,
value should not be greater than 16%
12~

Chem.Eng.Sai.
Vol 1

147

404
404
414
414
416
420
423
432
433
437
444
448
462
463
469
472
476
476
484
496
498
502
506
612
618
528
629
530
638
547
550
565

00321
00263
0.0286
0.0339
00247
0.0368
OG439
00354
00408
00476
00690
0.0622
00484
00630
00666
00761
00631
00687
00743
00922
00833
0.0876
00842
00919
0.1088
00993
01050
01054
01067
0.1164
01149
01234

Effu:

4%
009
0.07
008
010
007
011
0.13
012
0.13
016
021
0.19
0.18
020
028
0.32
027
030
036
0 47
043
046
046
0.52
064
062
067
068
073
0.86
0 86
096

Az/Aw

0096
0096
0.106
0093
0093
0066
0.106
0.133
0.203
0138
0162
0209
0.176
0230
0182
0137
0163
0281
0.164
0136
0214
0176
0194
0134
0146
0126
0174
0126
0126
0030
0076
0062

3
---T
k&f'/

344
34.2
238
289
31.1'
33.8
340~
3091
23 6
286
3361
33 61
30 4~
232~
28 2i
33.0'
30.11
2291
3221
32.6
i
22.6'
27.6
296
317
321
290
22.3
271
288
267
286
21.7

k,

170
13.1
152
19.7
16.6
20.4
43.7
62.8
762
998
161
209
166
191
368
417
381
662
614
983
966
1112
1410
1360
2460
2140
2626
2481
3630
3360
6440
11760

D ANNABLE Apphcatlon of the Ten-dun km&o equation to ammoma synthe+w m large-scale

reactors

Cl1cmirXXl
E~neering Science

RESULTS
JO-

Tables 1 and 2 gwe the valuee of k, determined by the

above methods
reciprocal

The values of log, 4 are plotted

temperature

in Figs

wd,h

1 and 2
80-

Table 2. Experzmental data for the determcnatzon of k,

300 atm. Temperature range

Preesure
ctency

range

37449Q

70-

C. Effc-

0 19 to 0 76.

mol fractcon
znetial gas wmpoestwn

Represent&we

NH,

0060

H*

0 678

N*

0 192

CH,
A

0045
0 126

I 60 *
w
H
50.

x
Temp
OC

Gas mate
kM8/hr

1mol

frac

--

VO-

4
so-

374

0 076

0 19

0 008

46 2

155

376

0 131

0.34

0 008

64.1

446

379

0 120

0 31

0011

61 0

6.68

386

O-081

0 22

0 015

46 0

6 07

387

0 078

0 21

0 016

42.2

4 82

391

0 145

040

0007

64 5

7 84

396

0 088

0.26

0 016

46 7

790

395

0 076

021

0 013

50 4

6 10

396

0 107

0 30

0 016

46 1

908
848

396

0 087

0 26

0 015

618

397

0 074

0.21

0 021

32.2

403

0 152

045

0 008

67 4

404

0 074

0 22

0017

33 4

406

o-157

0 47

0006

66 6

410

0 082

026

0.049

320

22.0

412

0 079

0.26

0049

33.3

23 8

Fig 1
lyst

Log, k, versus I/T for five batches of new cataDifferent symbols denote different batches of
catalyst.

6 62
16.1
6 97
13.1

413

0100

031

0 014

51 9

16 4

415

0,148

047

0 014

616

36 7

420

0,113

0 37

0 015

52 7

26 0

421

0 090

0 30

0064

31 7

38 2

422

0 143

0 47

0016

60 1

43 4

426

0 123

0 41

0 021

64.9

49 1

430

0 134

046

0 016

66 1

47 3

432

0 097

o-34

0054

31.6

69.0

438

0164

0 69

0 010

66 5

67.1

446

0 161

0.62

0 010

65 3

87 6

448

O-099

0 38

0 034

33 2

451

0 109

0 43

0 082

313

463

0 147

0 63

0 027

68 1

292

471

0 124

0 66

0 071

30 8

332

60 7
174

487

0 131

0 66

0 072

318

603

496

0 142

o-74

0 083

306

1120

499

0140

0 75

0 062

316

893

Fig

Log, k, versus I/T for one batch

operatmg

148

at various temperatures

of catalyst
and effmlencles.

Vol. I
Ho. 4 - 1962

D; &~ABLB:

Application of the Temkin kinetic equation to ammonia synthesis in large-so&

All the data applied to new catalyst, which was

multi-promoted

iron oxide.

Poisons (oxygen contain-

reactors.

stant and pressure, mentioned in the subsequent section, is correct, and the concentration of poisons in the

ing compounds and sulphur) were probabIy present in

very small quantities in every example. Measurements

reaction gas is similar for the two sets of measurements.

of them were not undertaken, but their concentration

is regarded as being fairly steady between the experi-

able, amounting to a coefficient of variation in k, of

29% for the fimt set of data, and 31% for the second.

mental runs. The actual small variations which occurred will increase the so-called experimental error.
Table 1 and Fig. 1 include data at 245 atm from

Neverthel&, the results are good enough to demonstrate that the Temkin equation in conjunction

five charges of catalyst.

The experimental scatter in the data is consider-

with the Arrhenius equation give a good interpretation

Initial gas composition was

of the ammonia synthesis reaction in the plant reac-

very similar from one run to another, and as there

tors, making possible the measurement of catalyst

activityin terms of the reaction velocity constant k, ,
and the apparent activation energy for ammonia de-

was not much variation in initial reaction temperature,

the adiabatic nature of the reaction has resulted in
some interdependence of temperature and efficiency,
where efficiency = z/z,,.
Fig. 2 and Table 2 present the data at 300 atm pressure from a single batch of catalyst spread between
the series of adiabatic beds. Although initial gas
composition was again fairly constant, it was possible
to obtain some variation of temperature which was
independent from the variation in efficiency, thus providing a more comprehensive test of the Temkin
equation.
In Table 3 are summarised values of k, and E,,
obtained from the two sets of data by calculating the
best straight line for the relation log, k, versus l/T.
Table 3
k, at
Data

Table 1
Table 2

420 C

) 470 C
I

Edec

. . . .
. . . .
DISCUSSIONOF RESULTS

Fig. 2 and Table 2 show that, over the experimental

range of ammonia concentrations, efficiencies and total
rates of flow, the Temkin equation gave values of the
reaction velocity constant, k,, which conformed well
with the Arrhenius equation.
Fig. 1 and Table 1 prove that reproducible results
could be obtained with different batches of catalyst,
-and show that the Arrhenius equation was obeyed
under different sets of reaction conditions and wider
ranges of efficiencies than in the case of Table 2.
obtained at 245 aCm and
The values of E,,
300 atm are similar. The small difference between
the k, values is approximately that expected, if the
empirical relation between the reaction velocity con-

composition : E,, .
The large experimental scatter will make impossible the detection of real small differences in the reaction velocity constant, and will render approximate any
calculations on reactor operation.
AGREEMENTOF k, AND EdeCwrrn 0TnEn Po~Lrsnxn
RESULTS
The values for Edec, the apparent activation energy
for ammonia decomposition, accord well with estimates found by other workers (see Table 4).
As the reaction velocity constant varies with pressure, it is difficult to compare our results at 245 and
300 atm with others at pressures of 106 atm and less,
(see Table 5). Moreover, as k, is very sensitive to the
concentration of poisons in the reaction gas, it is not
justifiable to compare different estimates without
a knowledge of the gas purity.
Nevertheless it is interesting to note that our results
can be correlated fairly well with those of EMMETTand
Gknwpa~iem~of v&m

Table 4.

iPresswe

of data

Tsnxm ond PYZEEV . .

T&xm and PyZEEVB cakul&ions

for 2, kc
atm

@-mvarious
Edcc

4oooo

111

PI end [31
Cll.~nd161
PI and PI
II41
II41
141
141

on WINTERS re1

46600
436oa

. . . .

31-6

46600

EMMETT . . . . . . . .

33.3

45ooo

EMMETT .
EMXE~ .
Em&r
a
I.C.I. . .

66.6

46 600

, . . . . . . . .

LARSON and TOUR.

.LARSONand TOUR

I.C.I.

References

10

sults

8ource.s

.
.
.
.

.
.
.
.

.
.
.
.

. . .

.
.
.
.

.
.
.
.

.
.
.
.

.
.
.
.

. . . . . . . . .

100

48900

100

53000

246

46 800.

300

47900

D.

Apphcatlon of the Temkm kmetm equatzon to ammonm synthe& m large-tie

ANNABLE:

Table 5. &nqwwon
Te?np

460

I.C.I.

LARSON and Tow

&?dEFT

127

135

100

196-212

400

LARSON and TOUR

I.C.I.
. . .

E-m.

370

I.C.I.

EMMETT

30.1

33.9

10.7

12 6

. . .

2 02

O(
20

316

237-292

10

781-915

714-91s

223-208

248

34-17

16.2-216

24 l-30.8

37.2-47 4

1 27-2.91

3 194.96

7.33-8 96

. . . .

LAIWON and TOTJR, by an empirical relationship

k, cc P-0 =, whmh is very similar to that found by
EMMETT,k, bcPo5 (see Fq. 3).

IO-

33 3

666

424-613

I.C

245

420

I.

300

Chemical
Englnedng Boienoe

from vmwwawwca
Preaeure (am)

sfnkrce of data

of k, vdua

reactors

tammg a heavier concentration of poisons than that

appertammg to the data for Fig. 1 Although similar
catalyst of similar age was used in both cases, the

1=/Cl
E- Emme ft
I!=hson 8 row

IIW

1300

1200

I400

loft-

Fig 4. Effect of po~som on actmty of new catalyst

Different symbols denote Uferent batches of catalyst

activity of the catalyst subject to the greater amount

of poisons 1s only about one third of that operatmg
Fig 3 Varmtlon of k, with pressure
under the more favourable condrtlons. This rapid
poisonmg effect, at least partly due to oxygen-containUSE OF k, AS A MEASUXE OF CATALYSTAUTJJ-ITY mg compounds, IS to be expected from the experimenTwo examples are mentioned of this use of the reaction
tal results of ALMQUISTand BL~OK [6], on the poisonvelocity constant Fig 4 presents data for five charges
ing of ammonia catalyst by carbon monoxide and
of new catalyst which operated in a g&s system conwater-vapour .
30

IO
log,
P-D

50

a-0

150

Vol. I
No 4-1962

D. ANXABLE: Apphcatlon

of

the Temkin kmetio equation to ammoma synthesu 111large-soak maators

The change m activation energy due to the iucreas-

where Earn = apparent

ed poisonmg is obscured by the fact that the slope of

activation

energy

for am-

monia synthesis in cal/mol N, reactmg

the line correlatmg log, k, w&h l/T IS mflueneed by

the greater poisoning of those layers of the c&al-&t bed

The above expression gives

P

k
Eaec
rka -i&-a

which operate at the lower temperatures

Fig. 5 demonstrates the tiference in activity which
which reduces to

was observed for two catalysts of different manufacture operatmg under smular condltlons

(4)
This equation gives L, for any z
L, being related to the equihbrmm constant, IS an
accuraQly known function of pressure, temperature

0I

MO/fracfton

020

of NH- -

Fig 6 Relation between optunum temperature

and ammoma conoentratxon

Fig. 6
t%LOJLATION

Comparison of two catalysts

OF

&TllWM CONDITIONS

The most useful apphcation of the Temkin equation 1s

to calculate optimum reactlon con&tlons, and then
to find how nearly these may be achieved m a plant
reactor
The determination of optimum reaction conditions
1s a wide problem, but m the example gven below, the
vanables of catalyst activity, total pressure and initial
gas composition were fixed
Maximum reactlon velocity for a gven gas composltlon, is attamed at a temperature where the dtierential coefficient of reaction velocity with respect to
temperature 18zero.
Dfiferentlatmg eq. (2), under optimum condltlons

o=&(g)

and gas composition. Hence the optimum temperature T, correspondmg to z, can be deduced. For the
example quoted, a pressure of 245 atm was chosen,
with an mltlal gas composltlon
mol fraction
0 016

Total.

0686
0 228
0 036
0 034

..I000

E,, w&8 taken as 47400, and the function k, such

that its value at 470 C was 325. A value of 20800
was taken for Esyn, &~1
determined from the relation
E wn - E,, = Q, the heat of reaction pr mol N,
reacting. With this information, the optimum temperature correspondmg to any mol fraction of NH, could
be calculated. Fig 6 gives the relation obtained between these two quantltles, and Fig 7 the correspondmg

151

D ANNABLE Apphcation

of the Temkm km&o

ammoniasynthesla m

equation to

maximum reaction velocities. It is of interest that

the reaction velocities, whilst they are high and decrease rapidly at low ammoma concentrations, are
low and relatively constant

Integration

catalyst volume A w 1s then found by substituting the

dw=

P6a15V
48

tion to be apphed g

of the synof

the ideal temperature and

mto account

ammonia concentratiPm gra-

The resultmg ideal gradients of ammoma concentration and temperature are given m Fig 8 The unit
of the abcissae la volume of catalyst traversed/imtlal

a catalyst

of N Ii3 -L

reaction velocltles
condltlons

at optimum

reaction gas rate, chosen so that results cantbe easily

computed for various gas rates
THEBESTOPERATINGCONDITIONSINAF'LANTREACTOR

Attainment of ideal condltlons m a plant reactor would

be a complex practmal problem mvolvmg removal of
the heat of reaction from the catalyst bed at a rate
decreasmg with catalyst age, and varymg through the
bed from a high figure at the beginnmg of the reaction
to a low figure at the end of the reaction. Moreover,
the first part of the bed would be operated at such a
high temperature that the activity of the catalyst
there would probably be quickly impaired
A reactor had been designed, however, to grve a partial approach to the ideal condrkons, by transferring
some of the heat of reaction to partially heated gas
which flowed m a counter-current direction to the
reaction gas, through vertical tubes mserted m the
catalyst bed The rate of removal of the heat of reaction could be vaned by altering
the rate of flow of the cooling gas,
or its initial temperature, but the
coolmg of particular sections of
catalyst could not be varied independently from one another.
- 650

..

fd,"l

femperofufe
--__

--__

I
5

500
-----a__

Vof cafa&f fmvemed/Imtlf/o/yas

Fig

Ideal

at the average conditions.

Subsequent increments A w are calculated m a similar

construction

- 706
"C

L,

= g

manner until the whole catalyst bed has been taken

Mot thcf~on

(I%*

sable the

bed The equation memtegrated m a stepwise manner

commencmg wrth the nntial
reaction conditions of gas
rate, composition and pressure. For a small increment
d z m the mol fraction
of ammonia, the optimum
temperature is calculated,
for the arithmetical average z m the interval, from
eq (4) The increment of

_--

(I+t);:(lIIE&_*t]

thesis equation makes pos-

dients through

Maximum

Engin%~~?eience

A z 1s origmally chosen small enough for the approxuna-

trations

rfxwtora

average values of V, z, and T 111eq. (3) ucz

for high ammonia ooncen-

Fig

h@-s0a~0

temperature

and NH,

I
IO
rafe -W

oonoentratlon

gradients

152

The kinetic data and methods

described above were apphed to
forecast the output and the best
operatmg conditions of this reactor, by calculatmg the optimum
nntial synthesis temperature and
the optimum degree of cooling
The same catalyst activity and
initial gas composition were assumed as 111the calculation of the
ideal gradients.

Vol I
No 4-lQ.62

D. ANNABLE

Apphcatlon of the Temkm kin&o equation to emmome synthwa m large-scale reaotore

The catalyst bed was considered in small seetfons,

the commencement

and the kmetic equation was used to give, for each

of the reaction, and the cooling

of one section of catalyst cannot be adJusted independently from another These limltatlons stall, however,

sectlon, the maximum possrble reaction which was

consistent wrth a heat balance between the heat

permrt of a fair correspondence between the

and ideal condltlons over the
of the catalyst bed (see Frg
divergence occurs where the

practical
last 80%
9). The
optimum

temperatures are higher than 550 C so

that operation at the lower temperatures has the compensation of preVofcata$d traversed/hfla/gas

mar

002

I
00.5

I
0 10
MO/fiacf/on of

raie -

serving the life of the catalyst

I
045

feature, together wrth Frg 9 demonstrate

that m future developments of the design of reactor, the arm should be to
improve still further the correspondence
between practical and ideal condrtrons
m the region of the peak temperature
and subsequently

I
m 021

tin,-)

This

AUKNOWLEtiUEMENT

The author wuthes to

Fig 9 Comperlson of practlcel and ideal condltlons.
Ammonza Works for
the experimental data, and the
of reaction, the change in heat content of the reDepartment ad Research Department
action gas, and the heat transferred to the coohng

thank the staff of

provrdmg most of
staff of Technical
for helpful advme

medium.
NOTATION

The best operating conditions, calculated by the

above method, gave an output of ammonra from the
reactor, equivalent to an exrt mol fraction of 0 21, or

a = mol fraction of H, m the reaction gas,

correspondmg to zero NH, content
A = kg mols/hr NH,
b = constant, such that a (1 -bz)
1s the mol

a conversron .
(exit mol fraction-inlet
19 4% (see Frg 9)

mol fractron) X 100, of

frctctlon of H, m the reactlon gas correepondmg to a mol fraction z of NH,

Thus compared with a maximum ammonra conversron of 22 0%) calculated for the Ideal operatmg
conditrons, and a figure of 19 0% which was the maximum conversion actually achieved m the running of
the plant reactor

b, = frequency factor m the Arrhenms equa-

tion for k,, the reectlon velocity constant

for ammoma decomposltlon
E deo = apparent activation energy for ammomla
decomposltlon
E 8Yn = apparent activation energy for ctmmoma

The agreement between the calculated and achreved practical condrtions 1s good, considering the large
coeffroient of variation (30%) in the original k, values,
from which the k, function was determmed The actual
catalyst temperature gradient was quite close to that
calculated, as was the amount of heat transfer from
the bed.
The best practmal condltrons also represent quite
a reasonable approach to the ideal The maxrmum
further improvement in the design and operation of
the catalyst bed would only mcrease the output by
less than 14%) although wrth the present arrangement,
the rate of heat removal is least instead of greatest at

synthesis
kl = reaction velocity

constant for ammoma

synthesis L
k, = reaction velocity constant for ammonw
decomposltlon
Kp = eqmhbnum constant for ammoma eynthesls

N = number of kilo mols of ammoms

P = total preseure atm
PNK,, PN,, PE, = p&d
pressuresm atm of NEE,, Np, and

153

H, respeotlvely
& = heat of reactron,c&/gm mol N, reactmg
R = gae constant cals/deg., gm mol

M M. HARUNIand J. A.

Hydroextractlon IV

khORROW:

Radml &stnbutlon of permeability m

chemical

cakes

Enslneealng Bdenae

REFERENOES

t=tlme, llr
T = temperature, K
V = rate of reaction gas, Mz/hr measured at 1 atm
and 20 c
w = volume of catalyst: MS

z = mol fraction of NH,

zsq = eqmhbrium mol fraction of NH,

[l] TE~~KIN,M.1 and PYZHEV,V, A&a Phys~cochnn 1940 1%

327-366. [2] GILLESPIE,L T. and BEA-,
J A , Physic
Rev 1930 86 743. [3] WINTER, E , Z physlk Chem. 1931
13 401 143 EXMETT,P H and KUXBXER,J T , Ind Eng
Chem. 1943 86 677. [6] Laaso~, A T and TOWE, R S ;
Chem and Met Eng. 1922 86 647 [a] AIJSQIJI~T, J A. and
BLACK, C A , J Amer Chem Sot 1926 48 2814

Hydroextraction IV: Radial distribution of permeability in cakes

M M ISAR~I and J AND~ON

STORROW

Chemical Engmeenng Laboratory, College of Tcohnology, Manchester, England

(Rmmed

21 March 1952)

Summary-Tests have been made on a 9 m &&meter hydroextractor runnmg at 1060 and 1400 r p m
usmg rn&ize starch, preclpltated chalk and kmselguhr
RBsnmXLBtude dun hydroextracteur tournant 8,105Ot/m ou B 1400 t/m a Bti falte en ut&sant de lanudon
de meis, du carbonate de chaux pr&lplte et du kmselguhr
It has been shown [l], [2], [3], [a], [5], [6] that
observed flow rates, q, of wash water through hydroextractor cakes follow closely the relationship

Q=

4d.na.KcX

6-r;

(u-g
I

loge z

(1)
I

and the small whnhng cell (Fig 2) showed hydroextra&on permertbilities withm f15%
of filtration
permeabllrtles, but there remained van&Ion mthm
this 15% which appeared systematic [4]. The variations of hydroextraction permeabihty wAh cake thickness, or cake weight, was not of the form expected

/nches
I
0
II I a 1

I
I

Fig 1 The centrifuge basket with the probe arrangement

for streambne flow with a hqmd layer of thickness

(rC- rL) w&hin a cake of thickness (rO- re), as shown
m Fig. 1 The results of tests on malze starch and
chalk in both the 9 in dm hydroextractor (Fig 1)

as a result of the varymg compressive stresses m the

centrifugal field There may be a significant resistence assocmted with the interface between packed
material and the cloth retaining the cake.

154