Professional Documents
Culture Documents
IJCCE - Volume 40 - Issue 6 - Pages 1800-1813
IJCCE - Volume 40 - Issue 6 - Pages 1800-1813
6, 2021
INTRODUCTION
Most of the reactions taking place in a catalytic fixed runaway’, which causes some deleterious effects
bed reactor are complex. These can be in the form of in the reactor. These include the promotion of undesired
highly exothermic reactions or highly endothermic side reactions, catalyst deactivation by sintering and
reactions. Usually, when a highly exothermic reaction productivity loss, and reduction of product selectivity.
takes place in a catalytic fixed bed reactor, some problems Thus, there is a need to develop a model that predicts
related to the operation of the reactor always occur since the outer-tube wall temperature profile for a given set of
most of the chemical reaction terms are a highly non-linear inputs, and the effect of heat loss through the wall on the
function of temperature and concentration and the value of magnitude of hotspot temperature which will help plant
heat may vary in the large amount. When heat is produced operators to reduce the risk of the reactor failure. Also,
faster than the rate it leaves the reactor walls, a rapid there is continuous generation or consumption of heat due
increase in the reactor temperature results [1]. This to enthalpies of adsorption and chemical reaction. The heat
excessive heat in the reactor is referred to as ‘temperature transfer that takes place during the adsorption of reactants
*To whom correspondence should be addressed.
+E-mail: aolatunde@unilag.edu.ng
1021-9986/2021/6/1800-1813 14/$/6.04
on the catalysts, the surface contributes to the thermal directions. It considers in addition to convection flow,
behavior of the fixed bed reactor as it increases the heat, and mass dispersion fluxes in the bulk fluid.
temperature of the reactor [2-3]. It is important to develop Mass balance:
a model that incorporates the effect of heat generated
Ci 2C 1 Ci
2
Ci
by adsorption of reactants on the catalyst surface in the reactor f R Ci,T (1)
i
uf Dr D e ,a x
z r2 r r dz
2
with a view to integrating its effect on the hotspot and
runaway phenomena.
Several models have been proposed in the literature to Energy balance:
study the behavior of fixed bed reactors. The models
dT 2T 1 T
are grouped into two major categories namely: the pseudo- uf f C pf r (2)
2
r r
homogeneous and heterogeneous models. Many works dz r
have been done on the modeling and simulation of 2
d T
catalytic packed bed tubular reactors using a two- e,a x
2
f H r R Ci,T
dz
dimensional pseudo-homogeneous model [4-14]. They
assumed the catalyst surface to be totally exposed to the
Boundary conditions
bulk fluid conditions, that is, there is no fluid-to-particle
C r, 0 C o ; T r , 0 To (3)
heat and mass transfer resistance between the solid and the
fluid phases. This implies there is a negligible difference at Z0 and 0rR
conditions in these phases are significant mostly highly at r0 and 0zL
Tf
2
Tf Tf Cf
f Cspf e ,a x U f f Cspf (8) (ii) U f C f (t) U f C o D e ,a x 0 t 0 (18)
z0
t z
2
z z z0
1 b 3 Cf
R f
Tf Ts (iii) 0, t 0 (19)
b z z zT
Solid-phase for temperature: The mass transfer effects included in the model are
intra-particle diffusion, inter-phase film diffusion, and
s s 2 Ts n
Tf Cs
(vi) 0, t 0 (15) (ii) 0, 0 z zT , t 0 (22)
z z zT r r0
Cf Cf
2
Cf
The energy balance applied to the bulk fluid phase
D e,a x Uf (16) (macrosystem)gives the following equation:
t z
2
z
1 b 3 2
Tf Tf Tf
k Cf Cs rR
R g f Cspf e ,a x U f f Cspf (24)
b t z
2
z
1 b 3 aw
The initial and boundary conditions needed to complete
R f
Tf Ts U w Tf Tw
the definition of the macrosystem are: b b
(i) Tf To , t 0, zT z 0 (25) Ci T r
Ci , T , ,
C i o To R
Tf
(ii) U f f C s p f Tf U f f C s p f To e ,a x 0 (26) rC , T
z i z tuf
i
Ci ,T
i
z0 R ,Z ,
t 0 ri o C i o , T o zT zT
r s
H ad
(28) zT U f
r
2
r r t P em (34)
D e ,a x
n Ts
s H rR i C i , Ti s s C s p s
i 1 t
1 b 3 kgzT
1 (35)
R U
The initial and boundary conditions needed to b f
(i) Cf 1 t 0, 1 Z 0 (36)
Ts
(ii) s f Tf Ts , (30)
r rR
1 Cf
(ii) C f (0, ) 1 0 (37)
0 z zT , t 0 P em Z Z0
Ts Cf
(iii) 0, 0 z zT , t 0 (31) (iii) 0 , t 0 (38)
r Z
r0 Z 1
Eq. (28) is the equation that shows the effect of heat For spherical pellets, Eq. (20) for solid-phase in terms
generated by adsorption of reactants on the catalyst surface of dimensionless variables yields:
while the second term on the left-hand side of the equation
Cs 1 2C 2 Cs
is the representation of the term. R i C s , Ts
s 2
s (39)
2 2
The energy balance around the column wall is given
by:
where
Tw
w C p,w hwaw Tw Tf (32) *
sq o Q *
t 1 (40)
2 C o s C s
a a m b Tw Ta m b
U am b
The Thiele modulus, , is given by
Eq. (32) explicitly predicts the effect of the outer-tube
wall temperature profile for a given set of inputs, and the
s ro C o , T o
effect of heat loss through the wall.
R (41)
D sC o
Non-dimensionalization of the model equations A restatement of the initial and boundary conditions
Introducing the dimensionless variables, defined as needed to solve Eq. (39) in terms of the dimensionless
follows: variables yields:
1 b 3 f zT
(47)
3 R C Which are used in equations (40) to obtain simple
b f spf U f
expressions 2 .
awU w
zT For an ideal gas, p V nR T p C R T , that is
4
(48)
b f U f C sp f C p RT . Using this relationship in Eq.(59) and
Subject to the transformed initial and boundary introducing dimensionless variables, we obtain:
conditions: Q
*
aLCo
, (60)
(i) Tf 1 , t 0, 1 Z 0 (49) Cs qo
*
U am ba am bzT
Ts 1 2T 2 Ts n
Tw Ta m b
R i C i , Ti
s 2
s (52) w C p,w U f
5 2
i 1
where
Application of the model to partial oxidation of methanol
H q
*
Q
* to formaldehyde
ad o
5 s (53)
C p s To Ts In this study, the exothermic reaction chosen was from
the pilot plant experimental work of Windeset al. [19],
D s H rxn C o which involves the partial oxidation of methanol to
s (54) formaldehyde accompanied by a side reaction of formaldehyde
s To
Table 1: The geometric, kinetic and parameters used for the partial oxidation of methanol to formaldehyde.
Physical properties and bed characteristics value unit
Fluid density 1000 [kg/m3]
Catalyst density 2000 [kg/m3]
Wall density 7800 [ kg/m3]
Specific heat of the fluid 952 [J/kg-K]
Specific heat of the catalyst 1904 [J/kg-K]
Thermal conductivity of the fluid 2.0 [W/m-K]
Superficial velocity 2.47 [m/s]
Feed temperature 523 [K]
Methanol in feed 1.7 [mol/m3]
Oxygen in feed 34 [mol/m3]
Fluid Temperature 523 [K]
Wall temperature 523 [K]
Ambient temperature 523 [K]
Reactor tube radius 0.0133 [m]
Particle diameter 0.0046 [m]
Thermal conductivity of the catalyst 2.0 [W/(m.K)]
Thermal conductivity of the bulk fluid 0.08 [W/(m.K)]
Void fraction in the bed 0.5
Void fraction in particles 0.57
Length of the reactor 0.7 [m]
Particle effective diffusivity 4.910-6 [m2/s]
Effective axial diffusivity 0.2619 [m2/s]
Ratio of mean surface area to volume of the column
Wall 270.0 [m-1]
Ratio of internal surface area to volume of the column
Wall 300.0 [m-1]
Heat transfer coefficient from wall to ambient 7.092 [W/(m2.K)]
Heat transfer coefficient from the fluid to solid phase 400 [W/(m2.K)]
Heat transfer coefficient of the wall 528 [W/(m2.K)]
Overall heat transfer coefficient of the wall 220 [W/(m2.K)]
Inlet pressure 1.55 [atm.]
Outlet pressure 1.3 [atm.]
Kinetic parameters
Heat of reaction for reaction 1 -158700 [J/mole]
Heat of reaction for reaction 2 -158700 [J/mole]
Heat of adsorption - 8368 [ J/mole]
Dimensionless number
Peclet number (mass) 6.6
Peclet number (heat) 8.6
Sherwood number 11.0
Biot number 5.5
0 .5 1 C
k 1C M 79496
rM D E N 1 1 k 2 o e x p 2 C o e x p
2 2 s
Cs
0 .5 0 .5
;
7
k1 1 2 5 1 0 e x p (63)
;
1 2 1 C s
0 .5
1 K 2C M RT
8368
K 2
1 .1 2 ex p
RT Methods for the numerical solution
The numerical method adopted for the solution
This is accomplished with an undesirable side reaction.
of the mass and enthalpy balance equations is the method
1 of orthogonal or optimal collocation. The orthogonal
C H 2O O 2 C O H 2O (64)
2 collocation method is applied by first choosing a new set
0 .5 of dimensionless variables, which have values lying
k 2C C H
r C H 2O
O
2
; k2 (65) between 0 and 1.
0 .5
1 0 .2 C C H 2O When the method of orthogonal collocation is applied
5 8368 to the space variables Z and then substituting u=2,
54 10 exp
RT Eq. (33) becomes:
1 C
A j,1
A
A 1, k j, N 2
A
A 1, k j, k
exp
1 1 s
C 0 .5
1 1 C s
s
2C
2 CS P e
1
1
s 2 P em B P e m B (70)
s
1
m j,1 j, N 2
2 P em
s
2 D E N1
P e m
A j,1
P e m A j, N 2
Ts A 1, N 2
(67) P e m
P em A 1,1 A N 2 ,1
A 1, N 2 A 1, N 2 A
1 C N 2,N 2
1 1 s
exp Cs
0 .5
1 1 1 C s
A 1, k A
1 2T
s 2 Ts 2
N 2 ,k
s 1 s A 1, N 2 A
5 2
D E N1 N 2,N 2
A 1,
k
P em A 1,1 A N 2 ,1
A 1, N 2 A 1, N 2 A N 2,N 2
B
P em A 1, k A N 2 ,k 1
H j, k
j,1
A 1,k B j, N 2 A 1, k B j, k (76)
A 1,1 P e m A 1,1 P e m A N 2 ,1 P eh
P e m ; A 1, k
A N 2,N 2
A 1, N 2 A N 2,N 2
A 1, N 2
A j,1
A
A 1, k j, N 2
A 1, k A j, k
A N 2 ,1 A 1,1 P e m A N 2 ,1 A 1,1 P e m
6
1
; A N 1, k
4
A N 1, k 9
U j, k
4u
j B j, k
B j, N 1 A
*
N 1 (77)
4 Shs
1
A N 1, N 1 *
S hi 6 A j, k
A j, N 1
A N 1
Peh
7 4 u j B j, N 1 6 A j, N 1 (73) A 1,1 P e h
P e h
A N 2,N 2 A 1, N 2
Following the same development of the collocation A A 1,1 P e h
N 2 ,1
points for the concentration profiles in the bulk fluid phase
A 1, k A N 2,k
and solid phase, the resulting expressions for the
A 1,1 P e h A N 2 ,1
temperature profiles in these phases are presented thus: A 1, k
A N 2,N 2 A 1, N 2
d Tf j N 1 NR A N 2 ,1 A 1,1 P e h
*
H j, k
Tf k A T
N 1, k s k
(74)
d k2 k 1 1 * 4
8 ; A N 1, k A N 1, k 8
3 3 8 4 Tf j P e h 4 Tw 1
4
A N 1, N 1
Nus
Nus
Ts 1 NR
s U j, k
Ts k 7 8 Tf j (75) RESULTS AND DISCUSSION
3 k 1
The results of the reaction network investigated
1 C
exp
1 1 s
C sj
0 .5 is presented and discussed in this section.
1 1 1 C s
The effects of heat generated by adsorption of the
2
1 s
D E N1 reactants on the catalyst and heat loss by the wall on the
magnitude of hotspot temperature using partial oxidation of
methanol to formaldehyde on FeO/MoO3 catalyst
Where
were discussed.
Temperature(K)
the conventional heterogeneous model without the effect 600
FeO/MoO3 catalyst and simulated results. The proposed Dimensionless axial distance
model predicted the location of the hotspot at the same
dimensionless distance of 0.4413 (= 0.0309 m) as the Fig. 1: The axial temperature profiles of different models
experimental value but with a temperature of 619 K as developed in the reactor.
compared with the experimental value of 622 K having
2
an Absolute Average Deviation (AAD) 0.48%. The conversion
585
Response of the reactor to change in feed temperature, 490 500 510 520 530
of the hotspot with little or no change in the position Wall heat transfer coefficient (W/m2K)
of the hotspot for the two changes.
Variation of the temperature inside the catalyst pellet and Fig. 4: Effect of change in wall heat transfer coefficient on
hotspot temperature.
with time along the reactor
Fig. 8 depicts the variation of the temperature inside
the catalyst pellet. The reaction being exothermic leads
to decrease in temperature from the center of the pellet to 640
To = 523 K
640 660
To= 523 K
620 640
To= 503 K
Temperature (K)
600 620
Temperature (K)
600
580
580
560
560
540
540
Tw= 503 K
520
Tw= 523 K 520
500 500
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Dimensionless axial distance Dimensionless radial distance
Fig. 6: Response of the reactor temperature to change in wall Fig. 8: Temperature profiles inside the catalyst pellet at
temperature from 503 K to 523 K. different feed temperatures (503K and 523K).
750 640
Co=1.74 mol/m3
600
650
580
600 560
540
550
520
500 500
0 0.2 0.4 0.6 0.8 1 0 0.2 0.4 0.6 0.8 1
Dimensionless axial distance Dimensionless time
Fig. 7: Response of the reactor temperature to change in feed Fig. 9: Temperature variation with time in the reactor.
concentration (methanol).
and parabolic partial differential equations took into model developed herein is useful in the design and analysis
account intra-particle and inter-particle diffusion of a catalytic packed bed tubular reactor, where mass and
resistances; heat generated by adsorption of reactants heat transfer resistances are accounted for.
on the catalyst surface and the heat transfer from the fluid
phase to the surroundings was included in the energy Acknowledgments
balance as it has a significant effect on the reactor We acknowledge the contribution of the Reaction
performance, especially reactor “hotspot”. The model Engineering and Energy Research group of the Chemical
shows that heat generated by adsorption of reactants on the and Petroleum Engineering Department of the University
catalyst surface has an effect on the magnitude of the hotspot of Lagos, Nigeria.
and reducing the temperature of the column wall leads
to a decrease in the magnitude of the hotspot. An increase Nomenclature
in the value of the wall heat transfer coefficient leads aamb Ratio of mean surface area to volume
to an appreciable reduction in the magnitude of the of the column wall, m-1
hotspot. It also shows that the model was able to accurately aF Freundlich adsorption isotherm constant, m3/kg
predict the experimental result of the reaction system, aL Langmuir adsorption isotherm constant, m3/kg
suggesting that the developed model represents the aw Ratio of internal surface area to volume
essential features of the real system. The comprehensive of the column wall, m-1
i, j, k Integer values [10] Lassak P., Labovsky J., Jelemensky L., Influence of
s Solid phase Parameter Uncertainty in Modelling of Industrial
w Wall ammonia Reactor for Safety and Operability
Analysis, Journal of Loss Prevention in the Process
Received :May 1, 2020 ; Accepted : Aug. 3, 2020 Industries,35: 280-288 (2010).
[11] Vasco de Toledo E.C., Morais E.R., Melo D.N.C.,
REFERENCES Mariano A.P., Meyer Jao F.C. A., Maciel Filho R.,
[1] Kursawe A., “Partial Oxidation of Ethene in Suiting Dynamic Models of Fixed-Bed Catalytic
Microchannel Reactor. Catalytic Pellet Reactor”, Reactors for Computer Based Applications,
Ph.D. Thesis, Technische Universiat Chemnizt, Engineering, 3: 778-785 (2011).
Fakulat fur Naturwissenschat, (2009). [12] Mahinsa P., Sadeghi M.T., Ganji H., Shokri S.,
[2] LL’in H., Luss D., Wrong-Way Behaviour of Packed Modeling and Sensitivity Analyses of
Bed Reactors: Influence of Reactant Adsorption on Hydrodesulfurization Catalyst Pellet, Petroleum and
Support, AIChE. Journal, 38: 1609-1617 (1992). Coal,54(2): 104 –109(2012).
[3] Bolis V.,” Fundamentals in Adsorption at the Solid- [13] Sadooghi P., Rauch R., Pseudo-homogeneous
Gas Interface, Concepts and Thermodynamics”. CA. Modeling of Catalytic Methane Reforming Process
Aurox (Ed.) Springer-Verlag Berlin Heidelberg, in a Fixed Bed Reactor, Journal of Natural Gas
(2013). Science and Engineering,11: 46-51 (2013).
[4] Ahmed A., Fahien R.W., Tubular Reactor Design I, [14] Fischer K.L., Langer M.R., Freund H., Dynamic
Two Dimensional Model, Chemical Engineering Carbon Dioxide Methanation in a Wall-Cooled Fixed
Science, 35: 889-895 (1978). Bed reactor: Comparative Evaluation of Reactor Models,
[5] Linke D., Wolf D., Baerns M., Zeyb S., Dingerdissen U., Ind. Res. Chem. Res., 58(42): 19406-19420 (2019).
Mleczko, L., Catalytic Partial Oxidation of Ethane [15] Liu Q.S., Zhang Z.L., Zhou J.L., Steady State and
to Acetic Acid over Mo1V0.25Nb0.12Pd0.0005Ox ; Reactor Dynamic Behaviour of Fixed Bed Catalytic Reactor
Operation, Chemical Engineering Science,57: 39-51 for Fischer-Tropsch Synthesis I: Mathematical Model
(2002). and Numerical Method, Journal of Natural Gas
[6] Aboudheir A., Akande A., Idem R., Dalai A., Chemistry, 8(2): 137-180 (1999).
Experimental Studies and Comprehensive Reactor [16] Celia X., “Catalytic Pellet Reactor Design under
Modelling of Hydrogen by Catalytic Reforming of Uncertainty”, Laboratory for Product and Process
Crude Ethanol in a Packed Bed Tubular Reactor over Design (LPPD). Project Final Report, University of
a Ni/Al2O3 Catalysts, International Journal of Illinois at Chicago, (2005).
Hydrogen Energy, 31: 752-761 (2006). [17] Cornelio A.A., Dynamic Modeling of an Industrial
[7] Amin N.S, Istadi., Yee, N.P., Mathematical Modelling Ethylene Oxide Reactor, Indian Chemical
of Catalytic Fixed-Bed Reactor for Carbon Dioxide Engineering. Section A, 48(3): 117-135 (2006).
Reforming of Methane over Rh/Al2O3 Catalyst. [18] Vasco de Toledo E.C., Morais E.R., Stremel D.P.,
Bulletin of Chemical Reaction Engineering Meyer F.C. A., Maciel Filho R., Development of
Catalysis,3(1-3): 21-29 (2008). Rigorous and Reduced Heterogeneous Dynamic
[8] Jess A., Kern, C., Modelling of Multi-Tubular Reactor Models for Fixed Bed Catalytic Reactor and Three-
for Fischer Tropsch Synthesis, Chemical Engineering Phase Catalytic Slurry Reactor, Chemical Products
Technology, 35: 1164-1175 (2009). and Process Modelling,3: 1-45 (2008).
[9] Suh J., Lee M., Greif R., Grogoropoulos C.P., [19] Windes L. C., Schwedock M. J., Ray W.H., Steady
Transport Phenomena in a Steam-Methanol State and Dynamic Modeling of a Packed Bed
Reforming Microreactor with Internal Heating, Reactor for the Partial Oxidation of Methanol to
International Journal of Hydrogen Energy,34: 314 – Formaldehyde I. Model Development, Chemical
322 (2009). Engineering Communication, 78:1-43 (1989).
[20] Zhu J., Araya S. S., Cui X., Sahlin S.L., Kaer S.K.,
Modeling and Design of a Multi-Tubular Packed-Bed
Reactor for Methanol Steam Reforming over a
Cu/ZnO/Al2O3 Catalyst, Energies, 13: 610-635
(2020).
[21] Bendjaouahdou C., Bendjaouahdou M. H., Control
of the Hotspot Temperature in an Industrial SO2
Converter, Energy Procedia, 36: 428-443 (2013).