-
Chemical Engineering Science, 1965, Vol. 20, pp. 281-292.
Pergamon Press Ltd., Oxford. Printed in Great Britain.
Steady-state simulation of an ammonia synthesis converter
R. F. BADDOUR, P. L. T. BRIAN, B. A. L~~E.-wJ, and 1. P.
EYMERY
Department of Chemical Engineering, Massachusetts Institute of
Technology
(Received in revised form 28 September 1964)
Abstract-A simple one-dimensional model of a T.V.A. ammonia
synthesis converter approximates closely the ammonia production
rate and the catalyst bed temperature profile of an industrial
reactor. Using this simulation, the effects of space velocity, feed
gas ammonia and inert contents, reactor heat conductance, and
catalyst activity upon reactor stability, ammonia production rate,
and catalyst bed temperature profile have been determined.
THE AMMONIA synthesis as it is performed in a Haber-Bosch
reactor belongs to the category of autothermic processes. This term
has been intro- duced by VAN HEERDEN [l] to describe an exother-
mic chemical reaction for which the temperature is maintained by
the heat of reaction alone. In order to achieve this condition, gas
flow and heat exchange are arranged to reduce the increase in
temperature associated with the exothermic reac- tion and to
suppress the need for an external source of heat once the reaction
is started. The Tennessee Valley Authority reactor (T.V.A. reac-
tor), which is a particular design of the Haber- Bosch reactor, has
been described elsewhere [2,3,4,] and a simplified diagram of it is
shown in Fig. 1. In the catalyst section the preheated gas flows up
inside a large number of small tubes. There it absorbs part of the
heat generated by the chemical reaction on the catalyst. At the top
of the reactor (C) the synthesis gas, now brought to a sufficient
temperature, reverses its direction and flows down the catalyst bed
where the reaction occurs.
The operating characteristics of a T.V.A. reactor have been
described by SLACK, ALLGOOD and MAW [4]. The process variables they
reported are: temperature of the feed, pressure, space velocity
(equivalent to the feed rate of gas), am- monia and inert content
of the feed gas, and hydrogen to nitrogen ratio in the feed. They
reported the existence of an optimum feed temper- ature, which
maximizes the production of ammonia and the existence for each
condition of operation of a maximum in the temperature profile,
called
the hot spot or peak temperature. Both the optimum feed
temperature and the hot spot temperature were found to vary with
the process variables and the catalyst activity. SLACK, ALL- GOOD
and MAUNE [4] reported that as the space velocity increases the
stability of the reactor decreases. The reactor tends to blow off,
causing the temperature and the ammonia mole fraction to decrease
monotonically. As the catalyst ages
SHELL COOLING GASES F
CATALYST BED
SECTlON
HEAT
Es%FR
GASES
.c
CATALYST
-G
-D
-B
ENTRANCE n
BY PASS INLET
FIG. 1. Simplified diagram of a T.V.A. reactor.
281
-
R. F. BAJXXXJR, P. L. T. BRIAN, B. A. LOGEAIS and J. P.
EYMERY
an increase in the average feed temperature is necessary to
avoid instability and keep the ammo- nia production up.
To investigate the steady state behavior of such a reactor, VAN
HEERDEN [l], ANNABLE [5], BEUTLER and ROBERTS [6], and KJAER [7]
have derived mathematical models of Haber-Bosch type reactors. VAN
HEERDEN, ANNABLE, BEUTLER and ROBERTS derived one-dimensional
models allowing for temperature and composition variations in the
longitudinal direction only. Even though their results approximated
experimental results, none investigated the effect of operating and
design variables on the production, stability and temper- ature
profiles in the reactor. KJAERS model took into account the
variations in temperature in both the longitudinal and radial
directions. His mathe- matical model consisted of three partial
differential equations which were solved by hand computation using
a double step integration technique, The temperature and
composition profiles in the reactor were computed for only one set
of operating condi- tions. The agreement of the computed production
rate and average bed temperatures with plant data was very good.
However, the KJAER model could not explain the radial temperature
gradient report- ed by SLACK, ALLGOOD and MAUNE. KJAER gave a
qualitative explanation of this discrepancy based on the location
of the various thermocouple wells with respect to the cooling tubes
in the catalyst bed. The work already published on the T.V.A.
reactor still leaves the following important areas to be
investigated. The effect of operating and design variables on:
(1) the optimum feed temperature, (2) the stability of the
reactor, (3) the temperature profiles in the reactor.
This information is necessary to determine the conditions of
maximum production. The objective of this paper is to present the
results of calculations using a computer simulation of a T.V.A.
reactor. The operating variables investigated were : space
velocity, ammonia and inert content of the feed, and catalyst
activity. The design variable was the heat conductance per unit
volume of reactor between the reacting gas and the gas in the
cooling tube. The pressure was kept constant, and the hydrogen to
nitrogen ratio was equal to 3.0.
Table 1. Parameters and their range of variation
Lower Upper limit Standard limit
Space Velocity, VO Ammonia mole fraction in
the feed, y* Inert mole fraction in
the feed, ant Catalyst activity, f Total heat conductance,
US
9000 13,800 18,000
0.01 0.05 0.10
0.0 0.08 0.15 0.4 . f 30,000 :&IO &IOO
The range of parameters investigated appears in Table 1.
MATHEMATICAL MODEL
A one-dimensional model was used, neglecting the temperature and
concentration gradients in the radial direction. The temperature of
the gas flowing through the catalyst at each location was assumed
equal to the temperature of the catalyst particles. With these
assumptions the T.V.A. reactor can be lumped radially into two
sections as shown in Fig. 2. The empty tube section repre- sents
the gas inside the cooling tubes, and the catalyst section includes
the catalyst particles and the gas flowing through them. The
temperatures TT and I, vary longitudinally in both sections. The
reaction rate expression used is that of TEM- KIN and PYZHEV [S]
with constants obtained by
T TOP
CATALYST
SECTON T Tc TT Y l--L 1 I T INLET
EMPTY TUBE
sEcnm
FIG. 2. Lumped model of T.V.A. reactor
282
-
Steady-state simulation of an ammonia synthesis converter
fitting [3] this expression to SmOROV'S experimental results[9]
measured at the same pressure (300 atm) and over the same range of
temperature and space velocities considered in this investigation.
In this equation the ?eaction rate is a unique function of
temperature and gas composition, neglecting mass transfer and pore
diffusional resistances.
In this model, both heat and mass diffusion in the longitudinal
direction have been neglected. It was also assumed that the heat
capacity of the gas is independent of temperature and that the
effect of pressure on enthalpy is negligible. The pressure drop
along the reactor was neglected. The validity of the major
assumptions will be dis- cussed in a later section.
A material balance written around a differential slice of the
catalyst section and enthalpy balances in the empty tube section
and in the catalyst section completely characterize the steady
state behavior of the T.V.A. reactor with the assump- tions
presented before. These three equations are presented in
dimensionless form.
(a) Material balance
fA 1 -20,300 1 m = m exp , ~ . d~
X {[ (K~P)2'~(~ -- y)1.5(c~ _ y)
Y
:.l (1 + Y) t (a~ - y )15] 1 + y* J
(b)
(e)
(1)
Energy balance in the empty tube section
dTr / US \ /AC\ T.
Energy balance in the catalyst section
[1 (AC~(y - y*~] dT c + -- [Cpo] \ 1 + y / J
I US \ /AC\ T +
r ( -AHo) -- T~AC -L +
y*,,dy,
In equations (1), (2), and (3) the temperature is normalized
with respect to the temperature at the top of the reactor where the
gas reverses its direction to enter the catalyst section.
The boundary conditions associated with this mathematical model
are specified at the top of the reactor by
Atg=O, To=l , T r= l , y=y* (4)
The numerical values of thermodynamic constants in the
dimensionless groups are reported in Table 2.
Solution of the mathematical model
This system of ordinary differential equations was solved on an
I.B.M. 704 digital computer using the Runge Kutta formulas. The
details of the computation are in reference [3]. Increment sizes
used in the computation were:
A~=0.05 for 0
-
R. F. BADDOUII, P. L. T. BRIAN, B. A. LOGEAIS and J. P.
KYMERY
top REACTOR LENGTH, (ft ) bottom
FIG. 3. Typical temperature profiles.
coefficient which yielded a value of 57,300 Btu/(hr) (F) for US.
The calculated ammonia production rate of 142 tons/day was 19 %
higher than the plant value. This discrepancy is discussed
later.
Table 3. Operating Conditioh
Parameter Actual
Converter Model
HB Mole fraction in feed 0.65 O-6375 Na Mole fraction in feed
0.219 O-2125 NHs Mole fraction in feed O-052 0.050 Inert Mole
fraction in feed 0.079 O-08 Space velocity, (hr)-l 13,800 13,800
Pressure, atm 286 300 Catalyst volume, ft8 144 144
The temperature profiles are shown in Fig. 3. At the outlet of
the converter the model tempera- ture is 12C higher than the plant
outlet tempera- ture; the hot point for the model is 1 ft lower and
the maximum temperature is 4C. higher than in the plant reactor.
Figure 3 also indicates the computed temperature profile inside the
tubes where the synthesis gases are heated from 228C to 421C. No
experimental data are available for comparison with these
temperatures.
RESULTS AND DISCUSSION
Before discussing the agreement between the computed and the
experimental results, each of the major assumptions made about the
model will be considered. As mentioned previously, ~AER [71
explained qualitatively the difference between the temperatures
reported by the center thermocouple and by the outer thermocouple
as due to the rela- tive positions of the thermocouple wells and
the cooling tubes. As shown in Fig. 4, the center thermocouple well
replaces a cooling tube, while each of the off-center wells is
located in the middle
COOLING TUBES
0 THERMOCOUPLE WELL
OUTER THERMCCOlJP,_E
FIG. 4. Thermocouple arrangement.
284
-
Steady-state simulation of an ammonia synthesis converter
of the equilateral triangle formed by 3 cooling tubes. The same
hexagonal area (hatched area on Fig. 4) is cooled by two cooling
tubes in the case of the center thermocouple and by three cooling
tubes in the case of the off-center thermocouple. For this reason
the measured center temperature prolYe should correspond
approximately to a heat transfer area decreased by one third, i.e.,
to a value of US equal to 37,000.
In order to examine this explanation, the effect of heat
conductance on the temperature profile was computed, and the
results are shown in Fig. 5. It can be seen that a conductance of
55,000 gives a good fit with the profiles measured by the south and
north thermocouples when the experimentally determined top
temperature is used for each profile. On the other hand, a
conductance of 37,000 is the value which gives a good fit with the
center thermocouple experimental temperature profile. The results
in Fig. 5 show that the difference in area available for heat
transfer accounts quantita- tively for the higher readings of the
center thermo- couple as reported by SLACK, ALLGWD and MAUNE [4].
Thus the computations performed by KJAER appear to be reasonable,
and they support the choice of a one-dimensional model.
The difference in temperature between the catalyst particle and
the gas flowing past it has been estimated by KJAER [7] and EYMERY
[2] for the synthesis of ammonia using the same type of catalyst.
The maximum temperature difference
has been computed to be 2~3C at the top of the reactor where the
rate of reaction is a maximum. This difference decreases as the gas
proceeds down the reactor to a value of 0.6C. near the middle and
0.4C. at the outlet. These small temperature differences justify
the assumption that the particle temperature is equal to the
temperature of the gas with which it is in contact.
The effect of longitudinal diffusion of enthalpy in the T.V.A.
reactor has been estimated by EYMERY [2]. A mathematical model
taking into account the axial diffusion of enthalpy by Taylor
diffusion has been derived and solved on a digital computer. The
inclusion of the longitudinal dis- persion term altered the
steady-state temperature profile by less than 0.6C. This effect is
consid- ered to be negligibly small.
In addition to the assumptions discussed above, the catalyst
activity was assumed to be uniform and constant with time. When a
catalyst activity factor of O-7 was used, the plant production of
120 tons/day could be obtained but with a temper- ature profile
unlike any of those measured.
Finally, the converter was operated at a lower pressure than
that used for the model, and the hydrogen-nitrogen ratio in the
plant was not exactly 3. Because of both these factors, the model
should give a higher production rate. Also, the data used in the
rate expression were taken from Russian experiments made using a
catalyst different from that in the T.V.A. reactor.
- MODEL ----- PLANT DATA
0 5 IO I5 top REACTOR L~~oTH, (ft 1 bottom
Fm, 5. Comparison of mode.1 with plant operation Y. = 13,800, y*
= 0.05, y; = O-08, f = I.
285
-
225
2001 I I I I 300 400 500
TOP TEMPERATURE, (C)
FIG. 6. Relationship between inlet and top tempera- tures.
In view of these differences, agreement between actual and
calculated production rates within 20 % corresponds to a reasonable
fit. Furthermore, it is possible to explain the difference between
measured and calculated profiles in the upper part of the converter
in terms of a decrease in catalyst activity in that region. For the
first 3ft of reactor length the experimental curves in Fig. 3
correspond to computed curves for a catalyst activity reduced by
about 50%. For the next 3ft the temperature rise in the actual
converter is faster than in the model, but this might be explained
by the fact that the gas mixture entering this section is farther
away from equilibrium because of the lower catalyst activity in the
upper part of the bed.
Thus, while the proposed model is only an approximation to the
actual reactor, it appears to give a fairly good prediction of the
ammonia production rate and the catalyst bed temperature profile.
The results obtained from this model concerning the effects of
operating and design variables on production, stability and
temperature profiles in a T.V.A. reactor are considered to be
reliable.
R. F. BADDOUR, P. L. T. BRIAN, B. A. LOGEAIS and J. P.
BYMERY
Existence of Several Steady States
For the operating conditions in Table 3, Fig. 6 presents the
relationship between the top tempcra- ture (computed at location C
in Fig. 1) and the inlet temperature (location G). Figure 6 shows
that there is a minimum inlet temperature called blow-out feed
temperature below which stable operation of the reactor is not
possible. For each inlet temperature higher than the blow-out feed
temperature there are two top temperatures, an example of which is
shown by the upper dotted line in Fig. 6. Likewise there are two
different temperature and composition profiles which satisfy the
steady state equations and which correspond to the same value of
the inlet temperature. VAN HIZERDEN [I] explained that the lower of
the two top temperatures corresponds to an unstable point. Although
the steady state equations are satisfied at this point, a small
perturbation would result in either reactor blow-out or reactor
heat-up and change-over to the stable operating point for that
inlet temperature. Thus in Fig. 6 only the branch of the curve to
the right of the minimum corre- sponds to stable operation of the
reactor.
Efects of parameters on production
The effect of top temperature on ammonia pro- duction rate is
shown in Fig. 7. The middle curve corresponds to the standard
conditions of opera- tion with a space velocity of 13,800. Figure 7
reveals the existence of an optimum top temperature equal to 425C
for the standard conditions of operation. As space velocity is
varied, the optimum top temperature changes as shown in Fig. 7.
When the space velocity increases from 9,000 to 18,000, production
rate increases as expected, but the production rate near the
maximum becomes more and more sensitive to a change in the top
tempera- ture. A deviation from the optimum top tempera- ture
affects the production rate more strongly at higher space
velocity.
An analogous result is found when the ammonia content of the
feed gas decreases. The production rate of the T.V.A. reactor is
quite sensitive to changes in ammonia mole fraction in the feed gas
but much less sensitive to changes in the inert content of the feed
gas. A decrease in catalyst
286
-
Steady-state simulation of an ammonia synthesis converter
350 400 450
TOP TEhtPEFfATLRE, (=C)
FIG. 7. Effect of top temperature on production rate y* = O-05,
g = 0.08, f= 1.0.
activity decreases the production rate and requires that the
reactor be operated at higher temperatures. Furthermore, at low
catalyst activity, the produc- tion rate is more sensitive to
changes in top temper- ature. The heat conductance has been found
to have a small effect on production. Although there is a value
which maximizes the production (US = 60,000), the maximum is very
flat.
Figure 8 presents the effects of the parameters on the maximum
production rate. For each set of parameters the converter is
operating at the opti- mum top temperature. Around the reference
conditions the changes in operating variables reported in Table 4
result in a decrease in pro- duction rate by 5 tons/day.
Efects of parameters on stability
It was shown previously that for each set of parameters there is
a minimum inlet temperature below which the reactor cannot be
operated. On Fig. 9 the relationship between the top and the inlet
temperatures is shown for three values of the
I 11 11 11 1 1 l0,000 15,000
, I, f I 81 8 I I
0 0.05 0.10
I$,, ,,,II,llllllll 0 0.05 0.10 0.15
I I 1 , t 0.5 1.0 1
I I I, 1 I I I I I I,
3QOoO 50,000 l30,ooo
0
Y'
Y;
f
US
FIG. 8. Effect of parameters on maximum production rate.
287
-
400 I I I I Table 5. Changes in parameters resulting in a
v. = SWCE VELOCITY +lOC change in the blow-ofl inlet
temperature
50[ 300 350 450 500 550
TOP TEMPERATURE (C)
FIG. 9. Effect of space velocity on stability.
Table 4. Changes in operating variables resulting in a 5
tons/day decrease in production rate
Operating variable Reference Change value
R. F. BADDOUR, P. L. T. BRIAN, B. A. LOGEAIS and J. P.
EYMERY
Operating variable Reference Change value
Space velocity, VO Ammonia mole fraction, y* Inert mole
fraction, yc* Catalyst activity, f
13,800 +600 0.05 +oGN3 0.08 $0.024 1.0 -0.05
The stability of the reactor can also be expressed in terms of
changes of the top temperature corre- sponding to the blow-off
inlet temperature. Figure 11 shows the effects of these same
parameters on the blow-off top temperature. It can be seen that
while the effect of space velocity and ammonia mole fraction on the
blow-off inlet temperature is relatively large, the corresponding
effect on the blow-off top temperature is much smaller. On the
other hand the catalyst activity a&&s the blow- off top
temperature more strongly. The heat transfer coefficient per unit
volume affects the blow- off inlet and top temperatures in opposite
directions: an increase in heat transfer coefficient allows a lower
blow-off inlet temperature but increases the blow-off top
temperature.
Eflect of parameters on temperature profire
Space velocity, VO. Ammonia mole fraction, y* Inert mole
fraction, yr Catalyst activity, f
13,800 -700 0.05 +0.01x 0.08 +0.02x 1.0 -10%
space velocity V,. The blow-off inlet temperature corresponds to
the minimum of the curve, and this temperature can be seen to
increase as the space velocity increases. Figure 10 presents the
effects of the parameters upon the blow-off inlet tem- perature.
Increasing the ammonia or the inert mole fraction in the feed
increases the blow-off inlet temperature. Increasing the catalyst
activity or the heat transfer conductance decreases the blow-off
inlet temperature. The changes in operating variables reported in
Table 5 result in a 10C increase in the blow-off inlet temperature
around the reference condition.
Rather than study the effects of the various parameters for a
fixed value of the inlet tempera- ture, the catalyst bed
temperature profile corre- sponding to each set of operating
conditions has been computed for the inlet temperature which
maximizes the ammonia production rate. These proties are presented
in reference [3]; they will be described qualitatively here.
An increase in space velocity shifts the hot spot downward with
only a small increase in the hot spot temperature. A larger
increase in the outlet temperature results, and the average
temperature of the bed increases. The amount of ammonia in the feed
gas influences greatly the location and the magnitude of the hot
spot. As the ammonia mole fraction in the feed increases, the
profile becomes flatter, the hot spot shifting downstream and
becoming cooler. The average temperature of the bed remains
approximately constant. A varia- tion in inert content of the feed
gas has little effect
288
-
Steady-state simulation of an ammonia synthesis converter
I_ f 100 - Y 01 I 1 1 1 1
10,000 15,000
YK P I
Y; 0.05 0.10 III! IIII ,!!I III)
0 0.05 0.10 0.15
f I I I I I { 05 1.0
US I, I I I I I I I I I 30.000 54000 00,000
FIG. 10. Effect of parameters on blow-off inlet temperature.
I- 0 0.05 0.10 0.15
I I I I I I I II 11 30,000 50,000 00,000
"0
Y' Y;
f
us
0
Y-
Y;
f
us
FIG. 11. Effect of parameters on blow-off top temperature.
289
-
R. F. BADDOUR, P. L. T. BRIAN, B. A. LOOEAIS and J. P.
EYMERY
0 I I I I I II I II
l0,000 15,000 "0
YS 0
I I I I 1 I I I Y 0.05 0.10
Yi* CO
, I I I I I IS I I I I f I I1 11 0.05 0.10 0.15
y;
f. I f 0.5 I .o I I I I I I I I 11 1
us 30,000 50,000 80,000 us
FIG. 12. Effect of parameters on peak temperature.
bottom
I I, ! I I II,,, 10,000 15,000
IO I I I I 1 I
0.05 0.10
0 I, I I I I ,,,I I I I I I I I I \
0.05 0.10 0.15 I I I I I
0.5 1.0 I I I I I * I I I I I
30,000 50,000 80,OO
FIG. 13. Effect of parameters on hot spot location.
290
0
Y'
Yi*
f
Us
-
Steady-state simulation of an ammonia synthesis converter
on the temperature profile; the hot spot does not move and its
temperature varies little. A decrease in the activity of the
catalyst results in an increase in the temperature at every
location in the catalyst bed. The heat transfer coefficient per
unit volume affects both the location and the magnitude of the hot
spot, but the average bed temperature does not vary very much. High
values of the heat transfer coefficient give higher hot spot
temperatures located nearer the top of the reactor. The effects of
the parameters on the magnitude and the location of the hot spot
are presented graphically in Figs. 12 and 13.
CONCLUSIONS
A simple mathematical model of a T.V.A. ammonia synthesis
reactor has been developed which approximates within 15 to 20% the
temper- ature profiles and the ammonia production rates of an
industrial reactor. With this model the effects of design and
operating parameters upon reactor stability, ammonia production
rate, and catalyst bed temperature profile have been studied.
It has been shown that an increase in space velocity increases
ammonia production rate but decreases reactor stability and
requires that the converter be operated at a higher temperature
level. Any increase in ammonia or inert content of the feed gas was
found to decrease both pro- duction rate and stability but not to
affect the aver- average temperature of the bed. The use of a less
active catalyst was shown to decrease both pro- duction rate and
stability and to necessitate opera- tion at a higher temperature
level. The heat transfer coefficient per unit volume of catalyst
was found to have a small effect on the production rate and the
average bed temperature but a marked influence on stability, a high
coefficient increasing stability and lowering the inlet temperature
of the reactor. The optimum temperature profiles were found to be
relatively insensitive to operating parameter variations. However,
the use of a high coefficient
of heat transfer increased local overheating of the
catalyst.
The reactor was shown to be sensitive to changes in the
operating parameters. Since under most conditions the difference
between the optimum feed temperature and the blow off feed tempera-
ture is very small (5C or less), the stability prob- lems
associated with small perturbations in the feed condition should be
investigated from a dynamic point of view. The results of such an
investigation appear in an accompanying paper.
Acknowledgement-The machine computations were per- formed at the
Massachusetts Institute of Technology Computation Center. H. Y.
ALLGOOD supplied the operating data, for which the authors are
appreciative.
NOMENCLATURE
(359) (1.75 x 1016) (P-0.5) (1.5 - Yl*,2)1.5(03 - Y;,)/(i +
y*)2,5
(Y$, + 1*5y*)l(1*5 - Y * a) (YNa) + @5Y*)/(@5 - YN2) molal heat
capacity of the feed gas, Btu/(lb mole) (F) (1 + Y*)l(1*5 -
Y&P5 molal feed rate lb mole/hr catalyst activity factor
equilibrium constant (atm)-1 length of reactor ft total pressure
atm. total heat transfer area fP top temperature R catalyst
temperature R empty tube section temperature, R base temperature
for enthalpy datum = 537R normalized catalyst temperature TC
P&p normalized empty tube section temperature T$/T&,,
normalized base temperature heat transfer coefficient
T,*I%, Btu/(hr)(ftz)(F)
space velocity = (359)(F)/v reactor volume ammonia mole
fraction
W-1 fts
NHa mole fraction in the feed gas Ha mole fraction in the feed
gas N2 mole fraction in the feed gas inerts mole fraction in the
feed distance from the top of the reactor, ft normalized distance =
z/L decrease in heat capacity resulting from the formation of one
mole of ammonia Btu/(lb mole)(oF) heat of formation of ammonia
Btu/lb mole
291
-
R. F. BADDOUR, P. L. T. BRIAN, B. A. LOGEAIS and J. P.
EYMERY
&iFERENCES
[l] VAN HEERDEN C., Zndustr. Engng. Chem., 45, 1242, 1953. [2]
EYMERY J. P., Clrem. Engng. Sc.D. thesis, M.I.T., 1964. [3] LOOEAIS
B., Chem. Engw. M.S. thesis, M.I.T., 1959. [4] SLACK A. V., ALLGOOD
N. Y. and MAIJNE H. E., Chem. Engng. Progr., 49, 393, 1953. [5]
ANNABLE D., Chem. Engng. Sci., 1, 145, 1952. [6] BEU~SR, J. A. and
ROBWTS, J. B., Chem. Engng. Progr., 52,69, 1956. [7j KJAER J.,
Measurements and Calculations of Temperature and Conversion in
Fixed Bed Catalytic Reactors, Jul.
Gjellerups Forlag, Copenhaeon, 1958. [8] TEMKIN M. I. and PYZHEV
V., Phys. Chem. Acta USSR, 12, 327, 1940. [9] Smo~ov I. P. and
Lxvsnrrs V. D., J. Phys. Chem., USSR, 21, 1177, 1947.
[lo] ALLGOOD H. Y., Private communication, Div. of Chemical
Operation, Tennessee Valley Authority, Wilson Dam, Alabama.
RtiLEn se basant sur un modele uni-dimensionnel du r&acteur
T.V.A. pour la synth&se de lammoniaque on a calcule leffet de
la v&cite spatiale, de la composition de lalimentation, de la
con- ductibilit6 thermique du lit et de lactivite du catalyseur sur
la stabilitt du reacteur, la production et les profils de
temperature.
292