364 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No.

3, 1979

Van't Riet, Boom, J. M., Smith, J. M., Trans. Inst. Chem. Eng., 54. 124 Zlokarnik, M., Chem. Ing. Tech., 47, 281 (1975).
(1976). Zwietering, Th. N., Ingenieur, 75, 63 (1963).
Votruba, J., Sobotka, M., Biotechnol. Bioeng., 18, 1815 (1976).
Votruba, J., Sobotka, M., Prokop, A., Biotechnol. Bioeng., 19, 435 (1977). Received for review June 26, 1978
Westerterp, K. R., Van Dierendonck, L. L, De Kraa, J. A., Chem. Eng. Sci.,
18, 157 (1963).
Accepted December 20, 1978
Wise, W. S„ J. Gen. Microbiol., 5, 167 (1951).
Yoshida, F., Miura, Y., Ind. Eng. Chem. Process Des. Dev., 2, 263 (1963). Part of this article was presented as a paper at the Engineering
Zlokarnik, M„ Chem. Ing. Tech., 42, 1310 (1970). Foundation “Mixing” Conference, August 1977, Rindge, N. H.

ARTICLES

Simulation of Ammonia Synthesis Reactors

See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Chandra P. P. Singh and Deokl N. Saraf*

Downloaded via UNIV OF CALGARY on October 11, 2022 at 20:19:37 (UTC).

Department of Chemical Engineering, Indian Institute of Technology, Kanpur-208016, India

A workable method to calculate diffusion effects within pores of a catalyst pellet, for a complex reaction, has been
developed. Suitable rate equations have been selected to describe the rate of ammonia synthesis reaction over
catalysts of different make. This, in conjunction with the method developed to calculate the effectiveness factor,
has been used to obtain a mathematical model for ammonia synthesis reactors. Reactors having ‘adiabatic catalyst
beds with interstage cooling as well as autothermal reactors have been considered. Plant data and simulation results
are, generally, in very good agreement.

Introduction A mathematical model considering all physical and

Synthesis of ammonia from hydrogen and nitrogen is chemical processes in the reactor has been described in this
one of the simplest kinetic reactions. The synthesis is paper. A method to solve transport equations to evaluate
straightforward, there is no side reaction, and the product the effectiveness factor, up to the desired accuracy, is
is stable. The physical and thermodynamic properties of developed and used in the model calculations.
the reactants and products are well known (Gillespie and Rate Expressions
Beattie, 1930). However, the mechanism of this reaction The literature contains innumerable rate expressions.
over the synthesis catalyst (iron catalyst) is not well
Those reported until the early thirties have been sum-
understood. This has led to numerous rate equations, all marized by Frankenburg (1933) and Emmett (1932,1940).
of which are of complex order. Because of the complexity In 1940 Temkin and Pyzhev developed a rate equation
of these rate equations it is difficult to account for the which offered a satisfactory kinetic approach to the
diffusional resistances to the transport of reactants and
synthesis and decomposition of ammonia over doubly
product in catalyst pores. This, in addition to limited promoted iron catalysts. Since then this rate equation as
reliability of rate equation, makes it difficult to have a such or in modified forms has been most extensively used,
mathematical description of the processes taking place some doubts about the generality of the equation
inside an ammonia synthesis reactor. although
have been raised (Adams and Comings, 1953; Emmett and
Major changes have taken place in the design of am- Kummer, 1943; Hays et al., 1964). The modified form of
monia synthesis reactors since the first commercial pro- the Temkin equation (Dyson and Simon, 1968) used in this
duction started in 1925 (Johansen, 1970). Most of these follows
work, is as
changes have been based on historical plant data rather
than an insight into the physical and chemical processes
taking place in the reactors. However, the use of com-
puters in design, optimization, and control made it nec-
essary to have a mathematical description of the process.
Simulation models for ammonia synthesis converters of
·~-'[*'·"-( )'(^)
where rNH3 reaction rate, kg-mol of NH3/(h m3 of cat-
=
“

different types have been developed for design, optimi- alyst), K2 velocity constant of the reverse reaction,
=

zation (Murase et al., 1970; Singh, 1975), and control (Shah, kg-mol/(h m3), /N¡¡, /Ha, /nh3 = fugacities of nitrogen, hy-
1967; Shah and Weisenfelder, 1969) purposes. drogen, and ammonia, respectively, ka = equilibrium
To describe the reactor operating conditions as accu- constant of the reaction: 1.5H2 + 0.5N2 NH3, and a =
—

rately as possible, the simulation model should take into constant. According to some workers a = 0.5 for all iron
consideration all the physical and chemical processes catalysts (Morozov et al., 1965; Temkin and Pyzhev, 1940;
taking place in the reactor. In order to avoid the com- Temkin et al., 1963) whereas others obtained values
plexity resulting from such a consideration, earlier workers ranging from 0.4 to 0.8 (Anderson, 1960; Bokhoven et al.,
have attempted only approximate simulations. 1955; Bokhoven and Van Raayen, 1954; Brill, 1951; Buzzi

0019-7882/79/1118-0364S01.00/0 © 1979 American Chemical Society

 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979 365

Ferraris and Donati, 1970; Cappelli and Collina, 1972; occur if the concentration and temperature throughout the
Dyson and Simon, 1968; Emmett and Hummer, 1943; pellet were the same as those at the outer surface, i.e.
Guacci et al., 1977; Huber, 1962; Livshits and Sidorov,
[molar flux of i across surf.]
1952; Love and Emmett, 1941; Mills and Bennett, 1959; 77
*
Nielson, 1968; Ozaki et al., 1960; Sidorov and Livshits, [pellet vol]
1947; Sholten, 1959). ___:-
[pellet surf, area ]
Having found different values of a, many authors suggest (5)
[rate of formation of i at surf, comp., T,P]
that it depends upon catalyst characteristics. It appears
reasonable, considering that catalysts of different make Several authors (Bokhoven and van Raayen, 1954; Kubota
differ in their promoter contents and physical charac- and Shindo, 1956; Kubota et al., 1959; Nielsen, 1968; Dyson
teristics. However, difference of the same order in the and Simon, 1968) have calculated this factor for ammonia
value of a has been reported for the same catalyst oper- synthesis reaction. In all these calculations but those due
ating under two different conditions (before and after the to Dyson and Simon (1968) either pseudo-first-order ki-
thermoresistancy test). Even for two arbitrary sets of runs netics was used or the bulk flow terms in the equations
over the same catalyst, values of a show similar differences for transport within the catalyst pellet were ommitted.
(Guacci et al. 1977). Dyson and Simon (1968) have formulated the problem by
Generally to evaluate a, K2 is expressed as a function considering all the aspects, but unable to solve it within
of temperature in Arrhenius equation form, i.e. practicable computation time limits, they also suggest the
use of an empirical correlation to obtain effectiveness
K2 =
K20e~E^T (2) factor. The development of transport equations, to follow,
where K20 is a constant, E2 is the energy of activation, Rg are similar to those done by Dyson and Simon (1968).
is the gas constant, and T is the absolute temperature. The Catalyst particles are assumed to be spherical (Nielsen,
experimental results are then fitted with K20, E2, and a as 1968) and isothermal (Prater, 1958). Knudsen diffusion
parameters. Such a method of evaluation could lead to is neglected (Bokhoven and van Raayen, 1954) and dif-
variations in values of a with a corresponding change in fusion coefficients of each component are assumed to be
values of E2 and K20. It has been suggested that there is independent of position within a particle.
a linear relationship between the value of log K2 and a. The reaction considered for the development of the
This shows that to differentiate the effect of catalyst diffusion equation is
characteristics or promoter content from those associated
with measurement errors or method of calculation is 72n2 + Y2H2 ^ nh3 (6)
difficult. Based on this, it has been suggested that a A mole balance for component i gives
constant value of a, independent of catalyst make, could
be used (Guacci et al., 1977). However, for a fixed value -—(r2N)
d rK
=
a^
ll-e (7)
of a, other parameters K20 and E2 are not available for the ,•2

two catalysts over which the present model has been tested. where r is the radial coordinate of spherical particle, N¡
The values used for the Montecatini Edison catalyst are is the molar flux of the ith component in the r direction,
a = 0.55, E2 = 39057 kcal/(kmol), log K20 = 14.7102, and
rNH is the rate of formation of ammonia given by eq 1, e
the corresponding rate equation is is the void fraction of the packed bed, and a¡ represents
the stoichiometric coefficient of the ith component in the
rNH, =
exp
(2
303 X 14.7102 -

RVT )
I X reaction scheme given by eq 6. Following the ordinary
convention, cq is taken to be positive if the component is
a product, negative if the component is a reactant, and zero

h-ter-ten
For the Haider Topsoe catalyst, a = 0.692, E2 =
(3)

42893,
if the component is an inert substance. Nitrogen, hy-
drogen, ammonia, methane, and argon have been desig-
nated as components 1, 2, 3, 4, and 5 respectively.
Boundary conditions for eq 7 are
and log Kw = 15.2059; the rate equation is J2:
II 0 at r = 0

4OQQO
dx,·
=
exp| 2.303 X 15.2059 I X
-

rNh3 3—
=
0 at r =
0
^2.303 dr

7'.«( (6
All the above values of a,
(4)

E2, and log K20 have been ob-

tained from the published work of Guacci et al. (1977).
X¿ =X¿g at r

where R' is radius of spherical particle, x¡ is the mole

fraction of component i at any point of the catalyst
particle, and xig is that at the surface. It is taken to be
= R'

same as that in a gas.

The molar fluxes of any two components i and j at
Effectiveness Factor steady state are related as follows
The above expressions represent the intrinsic rate of
a¡Nj
=
ajNi (9)
reaction, i.e., the rate of reaction for small particles in
which there is no resistance to transfer of mass or heat to This implies that molar flux of any inert component is
the active surface. The large size industrial catalyst equal to zero. The molar flux, N¡ of any active component
particles (6-12 mm for axial flow converters) are, however, i can be expressed in terms of its concentration gradient
subject to diffusion restriction in their pore structure. and the molar fluxes of other active components as
This effect can be taken care of by effectiveness factor,
d x¡
3
77, which is defined as the rate at which the reaction occurs +
in a pellet divided by the rate at which the reaction would
N¡ =
-CDie—
dr XiLNj (10)
;=1
366 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979

where C is the total concentration of reacting gas mixture

in kmol/m3 and Ae is the effective diffusion coefficient
of component i. Substitution of eq 9 into eq 10 and
utilization of the relationship

i=1
» = -1 (ID

gives
dx,
-CDic.
dr
Nt =
(12)
xi
1 +

Substituting for N¡ and Nj from eq 12 in eq 9 we have

Figure 1. Illustrating the instability of the numerical solution of the
equations for diffusion and reaction.

calculated value of dx3/dZ from the true one, i.e., direction

of shooting of dx3/dZ value as integration proceeds, de-
pends on whether the assumed value of dx3/dZ (at Z =
1) is smaller or greater than the true value. As shown in
Figure 1, if the assumed value of dx3/dZ is smaller (A)
Integrating this equation subject to boundary conditions integration would lead to shooting of its value in one
8 we get direction as the center is approached whereas it would have
the opposite direction if the assumed value (B) is greater
Xj
=
-Otj + (<Xj
+ Xyg) (14) than the true one. Therefore, by observing the direction
of shooting, it is possible to obtain a range (AB) within
which the true value lies. This range could be reduced to
As a consequence of this relationship, the solution of only desired accuracy (A'B') by repeated integration. After
one mole balance (eq 7) need be considered.
achieving the desired accuracy (range of dx3/dZ value
Substituting for N¡ from eq 12 into eq 7 yields reduced to A'B') no further integration is called for al-
d2x¡ i / dx, 2 2 dx, / a¡ + x, X rNH3 though the boundary condition at the center has not been
+ _

^ satisfied. In this work range A'B' representing less than

dr2 (a¡ + x¿) \ dr / r dr \ CDie /l -

e
0.5% of the true value of dx3/dZ has been used. The mean
of the two extreme values (A' and B') is taken to represent
Normalizing eq 15 by defining Z =
r/R' gives the true value of dx3/dZ.
The surface flux may now be substituted along with
other terms in eq 5 to get the effectiveness factor

(16)

and the boundary conditions in terms of Z are

dx¿
—
=
0 at Z =
0; x¿
=
x¿g
at Z = 1 (17)
From eq 18, is directly proportional to dx3/dZ. Therefore
This is a two-point boundary value problem and may the percent error in the evaluation of is the same as that
be solved by either choosing the unknown condition at the for dx3/dZ. With the range A'B' less than 0.5% of the
center (i.e., x¿ at Z = 0) and integrating eq 16 toward the absolute value of dx3/dZ, the maximum percent error in
surface {Z = 1) or choosing the unknown condition at the the evaluation of dx3/dZ and hence will be 0.25 in the
surface (dx,/dZ at Z = 1) and integrating inward. To solve present calculations.
the problem, in both cases, the value of the chosen quantity The total concentration c is obtained from
is varied and integration repeated until the boundary
condition at the other point is satisfied. It is convenient i =
/,l
to write eq 16 for the product component (NH3), thereby (19)
ensuring that the denominator in the second term of eq RJ
16 cannot go to zero. If the integration is carried out from where /¿ is the fugacity of component i. The radius, R',
center toward the surface, it is necessary to rewrite eq 16 for spheres equivalent to industrial size particles may be
because of the third term. This type of problem has been calculated from
treated by Weisz and Hicks (1962). We use the inward
integration procedure. It is reported (Dyson and Simon, R' =
(20)
1968) that to satisfy the boundary conditions at the center 2
with tolerable accuracy it is necessary to determine the flux where and are equivalent diameter and shape factor,
at the surface (i.e., dx3/dZ, which represents ammonia
respectively. The effective diffusion coefficient is cal-
concentration gradient at any point Z in the catalyst pellet) culated from the relation given by Wheeler (1955)
to within 1 part in 105. However, it may not be necessary
to satisfy tbe boundary condition at the center, if it is
Ae \e A (2D
=

recognized that the direction of rapid deviation of the

 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979 367

dT - 3( , P, T)
(25)
_

dV CP G
where G mass flow rate, AH
=
heat of reaction in =

kcal/kg-mol, V catalyst volume in m3, and £" is the

=

extent of reaction defined by

gjo + (26)
=
g¡ atjm¿"
where g¡, gj0 mass fraction of component j at any point
=

and at the inlet, and =

molecular weight of component
j·
Nonadiabatic Catalyst Bed. Assumptions the same
as those in the case of the adiabatic bed are made. Mass
balance equation 24 holds for this case also. The heat
balance is described by the equations
dTg U.A.(T -

Tg)
'
(27)
dV
dT -AHWnh3 U.A.(T -

Tg) -AHvr nh3

dV ~57 G (757 CP G
Figure 2. Schematic diagrams of a three-bed reactor and an au-
dTg CPg
tothermic reactor for synthesis of ammonia. (28)
"57 dV7
where is the intraparticle porosity and D¡ is the bulk
where Tg is the temperature of the feed gas in the im-
diffusion coefficient of component i. This coefficient at bedded cooling tube, U is the heat transfer coefficient^ A
0 °C and 1 atm is calculated from the relation
is the exchange area per unit volume of catalyst, and CPg
1 -

Xu is the specific heat of the feed gas. On the basis of the

DP = i =
1, 2, 3 (22) cooling tube inside diameter value of U used in this work
is 400 Kcal/(h m2 °C). The equilibrium constant Ka is
obtained from the Gillispie and Beattie equation (1930).
The fugacity coefficients are calculated by means of the
where Z)y, is the diffusion coefficient of component j in Cooper expression (1967) for nitrogen and hydrogen and
component i. The diffusion coefficients calculated from by the Newton expression (1935) for ammonia.
The composition and temperature of the reacting system
eq 22 are then corrected for the temperature and pressure
at the inlet of any layer being known, the material balance
at the surface of the catalyst pellet by
equation (24) is solved to obtain the extent of reaction
which in turn gives the mole fraction of individual com-
ponents (eq 26). For an adiabatic bed the solution of
energy balance equation (25) gives the outlet temperature
where P is in atmospheres. whereas for a nonadiabatic bed (cooling tubes imbedded
Description of the Mathematical Model in catalyst bed) eq 27 and 28 are to be solved simulta-
Since the reaction is exothermic all reactor designs have neously for the same purpose. The Runge-Kutta method
arrangements for removing the heat generated in the
of fourth order (Carnahan et al., 1969) has been used for
catalyst bed by the progress of the reaction. The reactor the simultaneous solution of mass and energy balance
also serves as its own heat exchanger to heat the incoming equations. The size of the integration step has been varied
synthesis gas. Apart from flow pattern (axial or radial) according to the volume of catalyst used.
the designs differ on heat removal strategy. In autothermic Calculation Program. On the basis of the mathe-
designs, heat is removed throughout the bed volume
matical model described above, a calculation program in
whereas other designs have adiabatic beds with cooling in Fortran IV has been prepared for the IBM 7044 computer.
between them either by external exchangers or by quench The calculation time for checking the performance of a
gas mixing. Figure 2 shows schematic diagrams of a reactor triple-bed reactor with interstage cooling is 3 min and 3
s when the step size for each bed is 1/100th of the bed
having three adiabatic beds with external cooling of the
reaction mixture and an autothermal reactor. Both volume. This program has been used only for checking
adiabatic as well as nonadiabatic beds, in an axial flow the performance of existing ammonia synthesis reactors,
reactor, have been considered in the present study. but it can conveniently be used for design purposes as well.
Adiabatic Catalyst Bed. The temperature and Results and Discussion
composition throughout a general cross section of the bed The calculation results for six different cases having
is assumed uniform. Axial diffusion of mass and heat is adiabatic as well as nonadiabatic beds (autothermal re-
neglected. Pressure drop is very small compared to the actors) working under different conditions of pressure and
total pressure; therefore a uniform pressure equal to the composition of the reacting system are presented along
average pressure in the reactor has been assumed. On the with the plant data in Tables I to V and Figures 3 to 5.
basis of these assumptions the material and heat balance For the purpose of discussion we consider the reactors
equations which describe the evolution of the composition having adiabatic beds and those with nonadiabatic beds
and temperature of the system along an adiabatic catalytic separately. The adiabatic reactors have Montecatini
layer can be written as catalyst whereas in autothermal reactors (nonadiabatic
T) beds) Haldor Topsoe catalyst is in use.
d£" nh3(£"> P>
Reactors with Adiabatic Catalyst Beds. Each re-
=
(24)
dV G actor under consideration has three adiabatic catalyst beds.
368 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979

Table I. Comparison between Experimental Data and Calculated Results. Case I. Triple Adiabatic Bed Reactor.
(Total Feed Flow, Nm3/h, 242 160; Pressure, atm, 226)
bed I II III
vol. of catalyst, m3 4.75 7.2 7.8
compn of gas, % inlet outlet inlet outlet inlet outlet
caled exptl caled exptl caled exptl
N2 22.19 20.03 20.1 20.03 18.58 18.2 18.58 17.55 17.8
h2 67.03 60.58 61.0 60.58 56.25 57.1 56.25 53.16 53.9
nh3 2.76 10.74 10.5 10.74 16.11 15.9 16.11 19.93 19.1
ch4 5.46 5.89 5.7 5.89 6.17 6.1 6.17 6.37 6.3
Ar 2.56 2.76 2.7 2.76 2.89 2.7 2.89 2.99 2.9
temp, °C 385 504 507 433 505 502 415 463 455

Table II. Comparison between Experimental Data and Calculated Results. Case 2. Triple Adiabatic Bed Reactor.
(Total Feed Flow, Nm3/h, 180 000; Pressure, atm, 177)
bed I II Ill
vol. of catalyst, m3 5.05 7.12 8.10
compn of gas, % inlet outlet inlet outlet inlet outlet
caled exptl caled exptl caled exptl
n2 19.6 17.71 17.2 17.71 16.63 15.7 16.63 15.70 15.5
h2 65.1 59.83 60.3 59.83 56.79 57.2 56.79 54.20 54.5
nh3 3.2 9.60 9.7 9.60 13.30 13.8 13.30 16.45 16.4
ch4 7.4 7.86 7.8 7.86 8.12 8.2 8.12 8.35 8.3
Ar 4.7 5.00 5.0 5.00 5.16 5.1 5.16 5.30 5.3
temp, °C 395 490 496 442 492 502 404 445 440

Table III. Comparison between Experimental Data and Calculated Results. Case 3. Triple Adiabatic Bed Reactor.
(Total Feed Flow, NM3/h; 19 500; Pressure, atm, 207)
bed I II III
vol. of catalyst, m3 5.25 7.35 8.30
compn of gas, % inlet outlet inlet outlet inlet outlet
caled exptl caled exptl caled exptl
n2 20.6 18.23 18.4 18.23 17.00 16.1 17.00 16.12 16.1
h2 65.1 59.33 59.0 59.33 55.62 57.4 55.62 52.90 53.0
nh3 3.2 10.60 10.7 10.60 14.90 14.2 14.90 18.28 18.3
CH„ 7.3 7.72 7.8 7.72 8.08 7.9 8.08 8.28 8.1
Ar 3.8 4.12 4.1 4.12 4.40 4.4 4.40 4.42 4.5
temp, °C 390 504 512 443 500 502 400 446 439
____

Volume ot bed { m3 )-*

Figure 3. Temperature profile in a triple-bed ammonia synthesis

reactor.

The results for these are presented in Tables I to III.

Figure 4. Ammonia concentration profile in a triple-bed synthesis
Figures 3 and 4 show the temperature and ammonia reactor.
percent profile along the beds for case I (Table I).
For case I, maximum difference in measured and cal- 8 °C less than the calculated one. Deviations in con-
culated ammonia concentration exists at the exit from bed centration as well as temperature at this point are of the
III (Table I, Figure 4). The measured value is 4.2% less same sign. However, in general there is no such corre-
than the calculated one. This error is well within ex- spondence between the difference in calculated and
perimental error limits. At the same point the difference measured values of composition and temperature. On the
in temperature is also maximum with a measured value contrary, for bed I, the calculated conversion is more than
 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979 369

Table IV. Comparison between Experimental Data and Calculated Values (Autothermic Reactors)
case 4 5 6
total feed flow, Nm3/h 47200 26350 242 59
pressure, atm 279 268 357
vol. of catalyst, m3 4.07 2.3 2.3
total heat transfer surface, m2 50.0 25.2 25.Í2
heat transí, coeff., kcal/(m2 °C) 400 400 400
compn of gas, % inlet outlet inlet outlet inlet outlet
caled exptl caled exptl caled exptl
n2 21.9 17.65 17.7 21.2 16.73 17.1 18.82 13.74 14.2
h2 65.0 52.11 52.8 65.8 52.73 52.3 59.45 44.75 45.1
nh3 5.2 21.16 20.5 3.0 18.99 19.1 3.00 20.55 20.0
ch4 4.3 4.93 4.9 6.7 7.74 7.8 13.21 15.62 15.9
Ar 3.6 4.15 4.1 3.3 3.81 3.7 4.52 5.34 5.3
temp, °C 421 438 438 437 456 460 417 447 443

Table V. Comparison between Experimental Data and Calculated Values (Autothermic Reactors)
case 4 case 5 case 6
bed temp, °C bed temp, °C bed temp, °C
bed length
(fractional) caled exptl caled exptl caled exptl
0 421 437 417
‘/24 444.88 443 467.84 475 444.53 441
3
/ 2 4 489.15 486 516.22 520 489.47 480
3
/ 2 4 517.82 516 532.89 536 516.95 514
2
/ 2 4 526.03 526 535.30 532 524.30 526
9
/ 2 4 522.73 523 534.10 530 523.80 522
n/2 4 516.03 514 530.91 516 520.21 518
* 3
/ 2 4 507.15 508 523.10 518 514.13 515
1 3
/ 2 4 497.48 498 513.28 514 505.55 502
1 7
/ 2 4 486.49 483 502.96 509 494.62 495
1 9
/ 2 4 474.57 475 490.85 500 482.36 480
W24 461.10 460 477.65 480 468.86 462
2 3
/ 2 4 446.29 446 464.30 468 454.88 456
1 437.64 438 456.03 460 446.90 443
and case 3 is less than 3.5%. This difference for case
1
2 is around 6%; however, the measured and exit con-
versions agree within 0.4%.
All these observations suggest that there is no con-
sistency in the error distribution. Measurements of high
temperatures around 450-500 °C in a plant could easily
involve an error of 8 to 10 °C, which is about 2% of the
temperature measured. All the variations of calculated
temperatures from the measured ones are less than this
allowable error and hence are insignificant.
Autothermal Reactors. Results for these reactors are
tabulated in Tables IV and V. The maximum difference
in measured and calculated composition exists at the outlet
of case 4 (Table IV). This difference of 3.2% is well within
the allowable error limits. For the same case, the measured
and calculated temperatures throughout the bed (Table
V, Figure 5) show very good agreement. This, from
Figure 5. Temperature of synthesis gas in catalyst bed and cooling thermodynamic considerations, would imply a close
tubes along the length. agreement between measured and calculated compositions
throughout the reactor, indicating thereby that the dif-
the measured one whereas the reverse is the case for ference observed in composition may be due to mea-
temperature. For first beds in case 2 and case 3, the surement errors. Also shown in Figure 5 is the calculated
concentration difference is very small whereas relatively temperature profile of the feed gas in the tubes embedded
high temperature differences, of 6 °C and 8 °C respectively, in the catalyst bed. The temperatures at various points
are obtained. Moreover, in the calculation procedure, the along the length of the reactor (Table V) and composition
input concentration to the second and third beds are the at inlet and outlet (Table IV) show that the difference
calculated output values at the exit of the first and second between measured and calculated values is small and could
beds, respectively, but while comparing we use measured be considered insignificant. At some points the difference
values. One can see that usually a higher calculated between measured and calculated temperatures is high;
conversion does not correspond to a higher calculated e.g., for case 5 at the fractional length of n/24 the tem-
temperature rise (bed III, Table III) and vice versa (bed perature difference is 14.91 °C. However, closer agree-
III, Table II). ments at most other points suggests that this could be
In the overall consideration for each case it is observed because of measurement errors and should be considered
that the difference between the measured and calculated insignificant. Thus in general the model can be considered
sum of temperature rise {ATs) over three beds for case reliable.
370 Ind. Eng. Chem. Process Des. Dev., Vol. 18, No. 3, 1979

The results very clearly show that in general the cal- Rg

=
universal gas constant
culated values are in very good agreement with the plant R' =
radius of spherical particle
data. It is to be noted that the deviations, in addition to T =
temperature, K
being small, are randomly distributed with respect to T„ =
temperature of gas in cooling tubes, K
actual plant observations both for temperature and if =
heat transfer coefficient, kcal/h m2
V =
catalyst volume, m3
composition. This confirms the reliability of the rate x¡ =
mole fraction of component i
equations used as well as the validity of assumptions made, xig
=
mole fraction of component i in gas (bulk phase)
under all plant conditions. This also shows that the effect
of diffusion of reactants and products inside the catalyst Greek Symbols
pores has been correctly accounted for. a, =
stoichiometric coefficient of component i
effectiveness factor
=

Conclusions £" = extent of reaction

A suitable pseudohomogeneous rate expression for e = void fraction of packed bed
ammonia synthesis reaction has been selected. A method A =
equivalent diameter
of calculating the effectiveness factor for industrial cat- =
shape factor
alysts by partially solving the transport equations has been
=
intraparticle porosity
developed. Subsequently, these have been used in a Literature Cited
mathematical model developed for design and simulation Adams, R. M., Commlngs, E. W., Chem. Eng. Prog., 49(7), 359 (1953).
calculations of axial flow ammonia synthesis reactors. Anderson, R. B., Ind. Eng. Chem., 52, 89 (1960).
Bokhoven, C„ VanHeerden, C„ Westrik, R., Zletering, P., “Catalysis”, Vol. Ill,
Reactors having adiabatic as well as those with nonadi- Relnhold, New York, N.Y., 1955.
abatic beds have been considered. Although the models Bokhoven, C., van Raayen, W., J. Phys. Chem., 58, 471 (1954).
obtained are capable of satisfactorily representing the Brill, R„ J. Chem. Phys., 19, 1047 (1951).
Buzz! Ferrarls, G., Donati, G., Ing. Chim. Ital., 6, 1 (1970).
reactor, under varying conditions of operation, each model Cappelll, A., Collina, A., I. Chem. E. Symp. Ser. No. 35, 5, 10 (1972).
is limited in its use due to the dependence of rate equation Carnahan, B„ Luther, . A., Wilkes, J. O., “Applied Numerical Methods”, Wiley,
New York, N.Y., 1969.
on catalyst make. More extensive work is required using
Cooper, H. W., Hydrocarbon Process. Pet. Refiner, 46(2), 159 (1967).
different catalysts to enable the development of a truly Dyson, D. C., Simon, J. M., Ind. Eng. Chem. Fundam., 7, 605 (1968).
Emmett, P. H., In Curtis “Fixed Nitrogen", Chapter VIII, New York, Chemical
general rate equation which can fit the data from any plant Catalogue Co., 1932.
irrespective of the catalyst in use. Emmett, P. H„ 12th Report of Committee on Catalysis, National Research Council,
Chapter XIII, Wiley, New York, N.Y., 1940.
Nomenclature Emmett, P. H., Kummer, J. T., Ind. Eng. Chem., 35, 677 (1943).
Frankenburg, W„ Z. Electrochem., 39, 45-50, 97-103, 269-81, 818-20 (1933).
a = constant Gillespie, L. J., Beattie, J. A., Phys. Rev., 36, 734 (1930).
A =
heat exchange area, m2 Guaco!, U., Traína, F., Buzz! Ferrarls, G„ Barisone, R„ Ind. Eng. Chem. Process
c = total concentration of reacting gas mixture, kg-mol/m3 Des. Dev., 16, 166 (1977).
Hays, G. E., Poska, F. L., Stafford, J. D., Chem. Eng. Prog., 60, 61 (1964).
CP = specific heat of reacting gas mixture, kcal/kg K Huber, A., Chem. Ing. Techn., 3, 147 (1962).
Cpg
=
specific heat of feed gas, kcal/kg K Johansen, K„ Chem. Eng. World, 5(6), 107 (1970).
Di diffusion coefficient of component i
= Kubota, H., Shlndo, M., Chem. Eng. (Japan), 20, 11 (1956).
=
diffusion coefficient of component i at 0 °C and 1 atm Kubota, H„ Shlndo, M„ Akehata, T„ Lin, A., Chem. Eng. (Japan), 23, 284 (1959).
D¿° Love, K. S., Emmett, P. H„ J. Am. Chem. Soc., 63, 3297 (1941).
= effective diffusion coefficient of component i Livshits, V. D„ Sidorov, I. P„ Zh. Fiz. Khim., 26, 538 (1952).
E2 = activation energy for ammonia decomposition, kcal/ Mills, A. K„ Bennett, C. O., AIChE. J., 5, 539 (1959).
kg-mol Morozov, N. M., Luk’yanova, L. I., Temkin, . I., Kinet. Kata!., 6, 82 (1965).
Murase, A., Roberts, H. L., Converse, A. O., Ind. Eng. Chem. Process Des.
/,·
=
fugacity of component i Dev., 9, 503 (1970).
/n2> /hj> /nh3
=
fugacity of nitrogen, hydrogen, and ammonia, Newton, R. H„ Ind. Eng. Chem., 27, 302 (1935).
respectively Nielsen, A„ “An Investigation on Promoted Iron Catalysts for the Synthesis of
g¡
=
mass fraction of component j Ammonia", 3rd ed, Jul Gjellerups Forlang, Copenhagen, 1968.
Ozaki, A., Taylor, H. S., Boudart, M., Proc. R. Soc. London, Ser. A, 258, 47
g,;0
= mass fraction of component j at inlet
(1960).
G = mass flow rate, kg/h Prater, C. D., Chem. Eng. Sci., 8, 284 (1958).
AH = heat of reaction, kcal/kg-mol Shah, M. J., Ind. Eng. Chem., 59, 72 (1967).
Shah, M. J., Welsenfelder, A. J., Automática, 5, 319 (1969).
K20 = frequency factor in Arrhenius equation for K2
Shorten, J. J. F., Thesis, Delft, the Netherlands, 1959.
K2 = velocity constant of reverse reaction Sidorov, I. P„ Livshits, V. D„ Zh. Fiz. Khim., 21, 1177 (1947).
=
equilibrium constant of the reaction: 0.5N2 + 1.5H2 = Singh, V. B., M. Tech. Thesis, I.I.T. Kanpur, India, 1975.
NH3 Temkin, . I., Morozov, N. M„ Shapatina, E. N., Kinet. Katal. 4, 260, 565 (1963).
Temkin, . I., Pyzhev, V., Acta. Physicochem., 12, 327 (1940).
rrij = molecular weight of component j
=
Weisz, P. B„ Hicks, J. S„ Chem. Eng. Sci., 17, 265 (1962).
N¿ molar flux of component i Wheeler, A., “Catalysis", P. H. Emmett, Ed., Chapter 2, Relnhold, New York,
P = pressure, atm N.Y., 1955.
r = radial coordinate of spherical catalyst particle
3 = rate of ammonia formation, kg-mol of NH3/h m3 of Received for review February 27, 1978
catalyst Accepted November 13, 1978