You are on page 1of 5

CATALYSIS, KINETICS AND REACTORS

Chinese Journal of Chemical Engineering, 17(4) 630ü634 (2009)

Experimental Study and Modeling of an Adiabatic Fixed-bed Reactor


for Methanol Dehydration to Dimethyl Ether

M. Fazlollahnejad1, M. Taghizadeh1,*, A. Eliassi2 and G. Bakeri3


1
Department of Chemical Engineering, Babol University of Technology, 4714871167 Babol, Iran
2
Chemical Industries Research Department, Iranian Research Organization for Science and Technology (IROST),
Tehran, Iran
3
Catalyst Research Group, Petrochemical Research and Technology Company, National Petrochemical Company,
Tehran, Iran

Abstract One-dimensional heterogeneous plug flow model was employed to model an adiabatic fixed-bed reactor
for the catalytic dehydration of methanol to dimethyl ether. Longitudinal temperature and conversion profiles pre-
dicted by this model were compared to those experimentally measured in a bench scale reactor. The reactor was
packed with 1.5 mm Ȗ-Al2O3 pellets as dehydration catalyst and operated in a temperature range of 543603 K at an
atmospheric pressure. Also, the effects of weight hourly space velocity (WHSV) and temperature on methanol con-
version were investigated. According to the results, the maximum conversion is obtained at 603.15 K with WHSV
of 72.87 hˉ1.
Keywords methanol, dimethyl ether, modeling, fixed-bed reactor

1 INTRODUCTION In our previous work, the effects of temperature


and feed composition on catalytic dehydration of
Dimethyl ether (DME), as a multi-source and methanol to dimethyl ether over gamma-alumina were
multi-purpose product, has received growing attention studied and the results showed that the conversion of
considering global environment pollution and energy methanol strongly depended on the operating tem-
supply problem today. DME can be produced from perature in the reactor. Also, conversion of pure
syngas, which is from natural gas, coal or biomass. methanol and mixture of methanol and water versus
DME has wide applications, such as liquefied petro- time were studied and the effect of water on deactiva-
leum gas (LPG) substitute, transportation fuel, pro- tion of the catalyst was investigated [12]. In this work,
pellant, chemical feedstock and fuel cell fuel [1]. It has acidic gamma-alumina has been used as the catalyst
been identified as a potential diesel and cooking fuel for the dehydration of methanol to DME. Also, we
due to its many excellent characteristics. It has an reached the optimum WHSV that gives us maximum
oxygen concentration of 34.78% and can be burned conversion of methanol in three inlet temperatures of
without soot emission, while for traditional diesel fu- methanol gas.
els the simultaneous control of NOx and soot emis-
sions cannot be expected. It has a boiling point of 2 EXPERIMENTAL
 25°C, which is 20°C higher than LPG and can be
liquified at 0.54 MPa (20°C). Therefore, based on the 2.1 Apparatus
matured technology of LPG application, there is no
any problem for the storage, transportation and usage The schematic diagram of the used set up is
of DME [2]. shown in Fig. 1. Pure methanol was pumped from
Two processes are used for DME production, in- methanol storage tank at different rates from 44 g·h 1
ˉ
ˉ1
direct [38] and direct processes [911]. In indirect up to 1021 g·h to an evaporator and then to a supper
process, methanol is converted to DME in a catalytic heater before entering the reactor. The superheated
dehydration reactor over a solid-acid catalyst by the methanol was sent to an adiabatic fixed-bed reactor.
following reaction: The axial reactor temperature at any point of the cata-
2CH3OH o CH3OCH3  H 2 O lyst bed was measurable via a thermo-well using a
thermocouple. The reactor outlet products were passed
In the second process (direct process), a synthesis gas through an air cooler and a double pipe heat exchanger
(a mixture of H2 and CO gases) is used as the feed of to cool down to the ambient temperature. The cooled
the process. In this process, the synthesis gas is pri- products were sent to a gas-liquid separator. A back
marily converted to methanol and then it is followed pressure regulator (BP-LF690, pressure Tech2000, Eng-
by methanol dehydration to DME. The net reaction is land) was placed on this separator to regulate the system
as follows: pressure. Reaction products were analyzed by a gas
3CO  3H 2 o CH3OCH3  CO2 chromatograph (Varian CP-3800) equipped with ther-
mal conductivity detector (TCD) and flame ionization

Received 2009-01-30, accepted 2009-05-25.


* To whom correspondence should be addressed. E-mail: taghizadehfr@yahoo.com
Chin. J. Chem. Eng., Vol. 17, No. 4, August 2009 631

d p / d tube 0.0015 / 0.0127  0.1 , and the second


Ltube / d tube 0.25 / 0.0127  50 [13]. The mass and
energy balance equations for the bulk gas phase can
be written as follows:
dC b
u i K0 vi rM T b , Cib U B (1)
dz
dT b
uUcp K0 rM T b , Cib 'H r U B (2)
dz
Where, Ci is concentration in fluid phase, u is superfi-
cial velocity, ǻHr is heat of reaction, vi is stoichiomet-
ric coefficient, ȡ is gas phase density, ȡB is catalyst
Figure 1 A schematic diagram of the experimental appa- bed density, cp is specific heat of fluid, and z is reactor
ratus for catalytic production of DME from methanol longitudinal coordinate. In these equations the reac-
tion rate equation rM T b , Cib and its kinetic parame-
1ünitrogen cylinder; 2ümethanol feed tank; 3üdosing pump;
4üflow meter; 5ümixer; 6üevaporator; 7üpreheater; 8ü
adiabatic fixed-bed reactor; 9üair cooler; 10ücondenser; ters that expressed by following equations were got
11üliquid-gas separator; 12üback pressure regulator from Bercic and Levec [3]:

rM T , Ci
2
ks K M CM2  CW CE / K (3)
1  2
4
detector (FID). Also, the remaining methanol in the exit K M CM  K W C W
products was measured and the methanol conversion
was estimated with comparison to the entrance methanol. ks 5.35 u 1013 exp 17280 / T (4)

2.2 Chemicals KM 5.39 u 104 exp 8487 / T (5)


KW 8.47 u 102 exp 5070 / T (6)
Acidic gamma-alumina as dehydration catalyst
was prepared from Engelhard (Netherland). Methanol Where, Ș0 (effectiveness factor) that is expressed by
was obtained from Iran Petrochemical Company (IPC). Eq. (7) is supposed to be 1, because catalyst particles
Characterization of the used catalyst and methanol and are so small that concentration and temperature varia-
the operational conditions are reported in Table 1. tions in particles can be neglected. Thus, this factor
does not have any important effect on temperature of
reactor and conversion of methanol.
3 DEVELOPMENT OF REACTOR MODEL
V 1 ³ rM T , Ci dV
3.1 Model assumptions K0 v
(7)
rM T b , Cib
The mathematical model for adiabatic fixed-bed Equations (1) and (2) are subject to the initial conditions
reactor is based on the following assumptions: (1) The that specify the feed composition and temperature:
feed current in reactor is plug flow and the gas phase
is assumed to behave ideally; (2) The reactor is oper- Cib Cib,0 at z 0
ated at steady state conditions; (3) Heat transfer from b
(8)
reactor to environment is negligible; (4) The reactor is T T0b at z 0
in isobaric operation; (5) Diffusion limitation in cata-
lyst pores is negligible; (6) Radial gradients of con- 3.3 Numerical method
centration and temperature are absent.
The numerical integration of the reactor mass and
3.2 Mathematical model heat balances [Eqs. (1) and (2)] was performed by
using a fourth-order Runge Kutta while the surface
A one-dimensional and heterogeneous ideal plug conditions (CS, TS) in the heterogeneous model were
model is chosen in this work to simulate the adiabatic determined by a Newton iteration method. In these
fixed-bed reactor. The dispersion model in this work calculations, ȡ, ǻHr, u, and cp were considered as a
cannot be used because of two reasons: firstly function of inlet temperature.

Table 1 Properties of catalyst and operating conditions


Catalyst particle Catalyst bed Catalyst bed Catalyst bed Reactor Reactor Inlet Absolute
Catalyst
diameter/mm height/cm porosity density/ kg·m 3 diameter/m height/cm temperature/K pressure/Pa
ˉ

engelhard Ȗ-Al2O3 1.5 25 0.6 693.96 0.01905 90 543.15603.15 101325


632 Chin. J. Chem. Eng., Vol. 17, No. 4, August 2009

4 RESULTS AND DISCUSSION


4.1 Model validation, prediction and experiments

Model results and the obtained experimental data


are shown in Figs. 27. In Figs. 24 variations of
temperature versus reactor length are shown. When
the inlet temperature of the feed is increased from 543
to 603 K, the reaction rate is increased; therefore, the
reaction is reached to its equilibrium state faster. Ac-
cording to Figs. 24 the experimental data are well Figure 5 Methanol conversion versus reactor length (Inlet
predicted by the model. In Figs. 57 conversion feed temperature is 543 K, p 101325 Pa and WHSV 45.3 hˉ1)
üü Eq. (1);Ƶexp.
versus reactor length are shown. It can be observed
that conversion is increased by increasing the inlet
feed temperature.

Figure 6 Methanol conversion versus reactor length (Inlet


feed temperature is 573 K, p 101325 Pa and WHSV 63.5 hˉ1)
üü Eq. (1);Ƶexp.

Figure 2 The simulated reactor temperature versus reac-


tor length (p 101325 Pa and WHSV 45.33 hˉ1)
üü Eq. (2);Ƶexp. data

Figure 7 Methanol conversion versus reactor length (Inlet


feed temperature is 603 K, p 101325 Pa and WHSV 79.8 hˉ1)
üü Eq. (1);Ƶexp.

According to the obtained results, the best inlet


Figure 3 The simulated reactor temperature versus reac- feed temperature is 573 K. The maximum conversion
tor length (p 101325 Pa and WHSV 63.47 hˉ1) at this condition was 95% with WHSV of 48.85 h 1.
ˉ

üü Eq. (2);Ƶexp. data Our experiments showed that at higher temperatures,


methanol conversion increases negligibly, but the se-
lectivity respect to DME decreases.

4.2 Effect of WHSV on methanol conversion

The influence of WHSV on the conversion of


methanol at three different inlet feed temperatures
(543603 K) is shown in Fig. 8. At a constant inlet
temperature, methanol conversion increases with de-
creasing WHSV, however, when the WHSV is less than
36.4 h 1, the methanol conversion does not increase
ˉ

Figure 4 The simulated reactor temperature versus reac-


with decreasing WHSV. For instance, XMeOH increases
tor length (p 101325 Pa and WHSV 79.8 hˉ1) from 48.9% to 93.00% at 543 K when WHSV de-
creases from 100.4 h 1 to 36.4 h 1. Increasing the
ˉ ˉ
üü Eq. (2);Ƶexp. data
Chin. J. Chem. Eng., Vol. 17, No. 4, August 2009 633

Table 2 Experimental data for conversion of methanol at different inlet temperatures and atmospheric pressure

Experiment mMeOH/ T 543.15 K T 573.15 K T 603.15 K


No. g·h 1
ˉ Mass of used WHSV/ MeOH Mass of used WHSV/ MeOH Mass of used WHSV/ MeOH
catalyst/g h 1 Conv./% catalyst/g h 1 Conv./% catalyst/g h 1 Conv./%
ˉ ˉ ˉ

1 44.042 1.16 27.35 64 0.805 54.71 85.8 0.805 54.71 70


2 75.6966 3.22 31.34 66 1.16 94.03 73.8 0.805 94.03 80
3 163.58 3.22 50.8 52 2.415 67.74 75 1.61 67.74 92.2
4 275.2616 12.88 34.2 85 9.66 42.74 89 4.83 56.99 83.4
5 469.298 12.88 36.44 93 9.66 48.58 95 6.44 72.87 95.8
6 485.2068 4.83 100.45 48.9 4.03 120.55 73 3.22 150.68 76
7 1021.832 25.76 43.6 82 16.1 63.47 83 12.88 79.83 84

with WHSV of 48.58 h 1, and 95.8% at 603.15 K with


ˉ

WHSV of 72.87 h 1. Therefore, according to the ob-


ˉ

tained results, the maximum conversion is obtained at


603.15 K with WHSV of 72.87 h 1.
ˉ

ACKNOWLEDGEMENTS

This work was partially supported by National


Petrochemical Research and Technology Company of
Figure 8 Experimental data of methanol conversion versus
WHSV at three different inlet feed temperatures (Dashed Iran, NPC-RT. The authors specially thank to the
lines are trend of changes) Chemical Industries Research Department of Iranian
Tinlet/K:ƽ543;Ƶ573;Ʒ603.15 Research Organization for Science and Technology
(IROST) for their generous assistance with this project.
WHSV will cause an increase in the gas velocity,
which promotes mass transfer but leads to a decrease NOMENCLATURE
in the contact time of reactant species [14]. As a result,
the conversion of methanol increases with decreasing Ci concentration in fluid phase, kmol·m 3
ˉ

WHSV until it reaches a maximum value. For the cp specific heat of fluid, kJ·kg 1·K 1
ˉ ˉ

lower WHSV, although the decrease in the flow rate ǻHr heat of reaction, kJ·kmol
ˉ1

causes to increase the contact time, but in this case the K thermodynamic equilibrium constant
mass of catalyst is too low to increase the conversion. Ki adsorption constant, m3·kmol 1
ˉ

Also, in higher WHSV the operative temperature reaction rate constant, kmol·kg 1·h 1
ˉ ˉ
ks
ˉ1
should be increased to have better conversion of Mi molecular weight, kg·kmol
p pressure, Pa
methanol (Table 2).
gas constant, Pa·m3·kmol 1·K 1
ˉ ˉ
R
As it is reported in Table 2, 7 runs are performed. rM rate of methanol disappearance, kmol·kg 1·h 1
ˉ ˉ

A constant mass flow rate of methanol as a feed is T temperature, K


used at three inlet temperatures for each run. The u superficial velocity, m·h 1
ˉ

maximum conversion of methanol is 93% at 543.15 K V particle volume, m 3

with WHSV of 36.44 h 1, 95% at 573.15 K with


ˉ
z reactor longitudinal coordinate, m
ˉ1
WHSV of 48.58 h , and 95.8% at 603.15 K with Ș effectiveness factor
WHSV of 72.87 h 1.
ˉ
Ȟi stoichiometric coefficient
gas-phase density, kg·m 3
ˉ
ȡ
catalyst bed density, kg·m 3
ˉ
ȡB
5 CONCLUSIONS Subscripts
b bulk conditions
A one-dimensional heterogeneous model is de- i ith component (methanol, DME, water)
M methanol
veloped to simulate the adiabatic fixed-bed reactor.
s surface conditions
The axial temperature and conversion profiles in an W water
adiabatic fixed-bed reactor for DME production are 0 inlet conditions
well predicted by the proposed model. Also, the effect
of WHSV on methanol conversion in different inlet
feed temperatures are considered. The results show REFERENCES
that the maximum conversion of methanol is 93% at
543.15 K with WHSV of 36.44 h 1, 95% at 573.15 K
ˉ
1 Ng, K.L., Chadwick, D., Toseland, B.A., “Kinetics and modeling of
634 Chin. J. Chem. Eng., Vol. 17, No. 4, August 2009

dimethyl ether synthesis from synthesis gas”, Chem. Eng. Sci., 54, methyl ether”, US Pat., 5908963 (1999).
35873592 (1999). 9 Takeguchi, T., Yanagisawa, K.I., Inui, T., Inoue, M., “Effect of the
2 Lu, W.Z., Teng, L., Xiao, W.D., “Simulation and experiment study of property of solid acid upon syngas-to-dimethyl ether conversion on
dimethyl ether synthesis from syngas in a fluidized-bed reactor”, the hybrid catalysts composed of Cu-Zn-Ga and solid acids”, Appl.
Chem. Eng. Sci., 59, 54555464 (2004). Catal. A, 192, 201209 (2000).
3 Bercic, G., Levec J., “Catalytic dehydration of methanol to dimethyl 10 Ramos, F.S., Duarte de Farias, A.M., Borges, L.E.P, Monteiro, J.L.,
ether-Kinetic investigation and reactor simulation”, Ind. Eng. Chem. Fraga, M.A., Sousa-Aguiar, E.F., “Role of dehydration catalyst acid
Res., 32, 24782484 (1993). properties on one-step DME synthesis over physical mixtures”,
4 Xu, M., Goodman, D.W., Bhattacharyya, A., “Catalytic dehydration Catal. Today, 101, 3944 (2005).
of methanol to dimethyl ether (DME) over Pd/CabOSil catalysts”, 11 Jun, K.W., Lee, H.S., Roh, H.S., Park, S.E., “Highly water-enhanced
Appl. Catal. A, 149, 303309 (1997). H-ZSM-5 catalysts for dehydration of methanol to dimethyl ether”,
5 Yaripour, F., Baghaei, F., Schmidt, I.B., Perregaard, J., “Catalytic Bull Korean Chem. Soc., 24, 106110 (2003).
dehydration of methanol to dimethyl ether (DME) over solid-acid 12 Raoof, F., Taghizadeh, M., Eliassi, A., Yaripour, F., “Effects of tem-
catalysts”, Catal. Commun., 6, 14752 (2005). perature and feed composition on catalytic dehydration of methanol
6 Xu, M., Lunsford, J.H., Goodman, W.D., Bhattacharyya, A., “Syn- to dimethyl ether over gamma-alumina”, Fuel, 87, 29672971
thesis of dimethyl ether (DME) from methanol over solid-acid cata- (2008).
lysts”, Appl. Catal. A, 149, 289301 (1997). 13 Riggs, J.M., An Introduction to Numerical Methods for Chemical
7 Eliassi, A., Savadkoohi, L., Kargari, A., “Effects of various process Engineers, Texas Tech University Press, USA (1994).
parameters on temperature profile of adiabatic fixed-bed reactor for 14 Moradi, G.R., Ghanei, R., Yaripour, F., “Determination of the opti-
production of dimethyl ether (DME) from methanol”, Chem. Eng. mum operating conditions for LPDME from Syngas”, Int. J. Chem.
Commun., 194, 14951502 (2007). Reactor Eng., 5, A14 (2007).
8 Voss, B., Joensen, F., Hansen, J.B., “Preparation of fuel grade di-

You might also like