AJChE 2005, Vol. 5, No.

2, 176-182

Propane DehVdrouenalion in a ModDied Porous lembrane Reactor lor Producing ProPVIele wIh Chemical and PolVmer Grades

Azis Trianto Yazld Blndar Noezran

Department of Chemical Engineering Institut Teknologi Bandung

JI. Ganesa 10 Bandung 40132 INOONESIA Email: [email protected]

Propane dehydrogenation is a promising route for producing propylene to replace traditional cracking methods. A membrane reactor offers a possibility to produce not only chemical grade but also polymer grade of propylene. The purpose of the present study is to evaluate the performance of a Modified Porous Membrane Reactor (MPMR) in producing these two propylene grades simultaneously. The study involves evaluations based on thermodynamics and process flow sheeting. The performance of this reactor is compared to that of conventional reactor. At first, the thermodynamics is conducted using minimum Gibb's Energy approach. Then the process flow sheeting evaluation is built using the HYSYS simulator. The effect of inert gas (steam) concentration in both sweep and feed sides is investigated. The thermodynamics study results optimum temperature and inert gas concentration to obtain these two grades of propylene Simultaneously. The propylene with polymer grade above 99% is produced from the sweep side outlet. The propylene with chemical grade is produced from the feed side outlet. The simultaneous production of these two grades of propylene has benefit in vanishing propane-propylene splitter.

Keywords: Membrane reactor, porous membrane, propane dehydrogenation, propylene production, and process simulation.

INTRODUCTION

The increasing value of polypropylene as a versatile polymer is one of the major influences to the change of propylene price, since the polypropylene productions use over half of the global propylene consumptions (WeIrauch 1999). International Chemical Information Conference (ICIC) forecasts the rise on propylene demand and production capacity for regional and international markets until 2003 as given in the Table 1 (2001).

Global and national projections for propylene production give the increased in production quantity. There are several routes of propylene production. One of these routes is catalytic dehydrogenation of propane. This process has been commercialized by UOP as UOP Oleflex Process. This process produces propylene with polymer grade or chemical grade. This catalytic dehydrogenation of propane process has several problems with the major problem encountered in separation duty of propylene rectifying column,

Propane Dehydrogenation in a Modified Porous Membrane Reactor 177

Table 1. Propylene Demand and Production Capacity (ICIC 2001)

Percentage of Increase
Demand Production Capacity
World 4.6 3.3
Asia 5.9 4.5
North America 4.1 2.5
ASEAN 12.0 8.9
Western Europe 3.1 2.7
China 9.5 6.4
India 29.3 27.1
South Korea 4.9 1.8 or better known as propane-propylene splitter column. Propylene with high purity production needs Jarge separation duty. Furthermore, since the boiling point of propane and propylene is very close, the column must have a lot of stages.

Various methods have been proposed to solve the problem propane-propylene splitter (Cheng 2001, Cheng 2001, Cottrell 2001). However, they could not reduce the propane-propylene separation duty significantly. This gives opportunity for other methods with various approaches.

Propane dehydrogenation is an endothermic and equilibrium reaction. High conversion for this type of reaction can be achieved with high operating temperature and low hydrogen partial pressure. If hydrogen as reaction product is separated as fast as possible inside the reactor continually, the partial pressure of hydrogen will decrease and larger conversion will be achieved.

A membrane can be used to separate hydrogen because of its chemical properties or its smalt pore size. Reactor that uses the membrane wall inside is called as membrane reactor. In a membrane reactor, hydrogen diffuses through the wall from the catalyst zone. Wood introduces this method in 1968 to improve the yield of cyclohexane dehydrogenation (Wood 1968). Thenceiortn, many efforts have been done to further develop the membrane reactor technology. The motivation of those efforts is to achieve high conversion with modification of various parameters.

Trianto and Kokugan in 1999 gave a new modification of the membrane reactor with high performance for dehydrogenation reaction. They called the reactor as Modified Porous Membrane

Reactor (MPMR) (Trianto 1999, Trianto 2001). MPMR is sa new configuration of membrane reactor where the reaction took places in both sides of membrane reactor, i.e. the sweep side and the feed side. The difference between the Conventional Porous Membrane Reactor (CPMR) and MPMR is schematically described in Fig. 1. This MPMR can give higher conversion than CPMR due to the reactant permeation or reactant loss problem that occurs in CPMR (Trianto 2001, Kokugan 2001).

This present assessment suggests that MPMR is able produce propylene from propane by dehydrogenation reaction with polymer and chemical grades Simultaneously. The schematic description of the suggested process system is in Figure 2. This assessment also evaluates the performance 0/ a model MPMR process system and compares it to a model of UOP Ole/lex process system.

a. CPMR

b. MPMR

In!iH!:~1 catalysl

o non.catalyst

Figure 1. CMPR and MPMR Diagram

178 A. Trlanto, Y. Bindar, and Noezron

CHEMICAL GRADE PROPYLENE

FEED

SWEEP

POLYMER GRADE PROPYLENE

MPMR

Figure 2. MPMR Process System

METHOD OF EVALUATION

The present assessment performs evaluation in two phases. The first phase is thermodynamics evaluation of propane dehydrogenation. The results of this evaluation is an optimum operating condition such as operating temperature at atmospheric condition and inert concentration which give the highest conversion. The second step is evaluation of several scenarios of process systems. These process systems are built based on thermodynamics evaluation results. The evaluation of these scenarios will result an optimum scenario for the production of propylene with high purity and minimum consumption of feed energy.

Thermodynamic evaluation of propane dehydrogenation is carried out using minimizing method of Gibb's energy change due to temperature change. The pressure is constant and is held at atmospheric condition. The relation can be written as:

d( t..G" / RT) -Mf"

=

dT RT2

Reaction equilibrium constant K correlates with temperature and Gibb's energy as described in:

scr --=-lnK

RT

The change of reaction equilibrium constant affects the change of reaction conversion.

Therefore, the evaluation investigates the relation between the change of reaction conversion and the change of temperature. Thermodynamic evaluation is conducted at temperature range of ODC-1000DC. Thermodynamic evaluation is also made to investigate the effect of inert concentration, i.e. steam in this study, to the reaction conversion and propylene production. The evaluation uses ratios of steam/propane at 0, 1,5, 10, and 50.

Two scenarios of process systems were built based on the thermodynamic evaluation results. First scenario is a process system that produces propylene using the MPMR. Performance of this process system scenario is compared with the second scenario, i.e. a process system that produces propylene using a model of UOP Oleflex Process. The evaluation of these scenarios is tested at product basis of 100 kmole/hr polymer grade propylene and 1000 kmole/ hr chemical grade propylene. Both scenarios use pure propane as reactant and steam as inert for reactor feed. The operating temperature and steam concentration for both scenarios is defined from thermodynamics evaluation results. The effect of hydrogen separation inside MPMR is neglected.

The comparison between both scenarios is conducted for various operating variables such as: consumption of propane, consumption of steam, consumption of energy, and hydrogen production. For all evaluations carried out in this study, the products of propane dehydrogenation are assumed to be propylene, propadiene, methyl acetylene, and hydrogen.

RESULTS AND DISCUSSION

(1)

The thermodynamic evaluation at isothermal condition shows that level of propane conversion increases as the temperature increases. However, there is an optimum temperature at which propylene production reaches a maximum. Figure 3 shows propane dehydrogenation conversion, propylene production, methyl acetylene production, and propadiene production as a function of temperature. As shown in Figure 3, the optimum temperature for propylene production is obtained at temperature around 800 DC. Due to this reason, temperature of SOO DC is selected as the operating condition for the second phase evaluation.

(2)

1000

Propane Dehydrogenation In a Modified Porous Membrane Reactor 179

100
90
80
70
I:
.2 60
~
III 50
;:::.
I:
0 40
CJ
r:fl. 30
20
10 600 700 800

Temperature [deg C]

Figure 3. Effect of Temperature on Propane Conversion. Propylene, Methyl Acetylene,

and Propadiene Productions

500

900

100
90
80
70
I: 60
.2
~ 60
III
;:::. 40
I:
0
CJ 30
r:fl.
20
10
0
400 SiP':"ueamiptOpilPf mole ratio

500

900

1000

600 700 800

Temperature [deg C]

Figure 4. Effect of Steam/Propane Mole Ration on Propane Conversion

100
90
80
70
I:
0 60
.~
III
;:::. 50
I:
0
CJ 40
r:fl.
30
20
10
0
400 SIP=steamipropanc IOOlc ratio

600 700

Temperature [deg C)

800

900

500

1000

Figure 5. Effect of Steam/Propane Mole Ratio on Propylene Production

180 A. Trlanto, Y. Blndar, and Noezran

Table 2. Value o/Comparlson Variables/or Production Polymer Grade Propylene

Variables MPMRMSweep UOP
(Polymer Grade) (polymer Grade)
Feed Consumption
Propane (kglhr) 6875 4900
H20 (kglhr) 126390 -
Feed Preheating Energy Consumption (kJlhr) 1.3 x 107 8.0 x io'
Reactor Energy Consumption (kJlhr) 2.0 x 107 1.3x107
H20 Separation Energy Consumption (kJ/hr) - M
H2 Production (kg/hr) 425 215
H2 Separation Energy Consumption (kJ/hr) 3.9 x 106 3.5 X 106
MAPD Separation Energy Consumption (kJ/hr) 7.0 x lOS -
MAPD Column Stages 57 M
Propane-Propylene Splitter Energy Consumption (kJ/hr) M 1.5 x 107
Propane-Propylene Splitter Stages M 183 Table 3. Value oj Comparison Varlablesjor Production Chemical Grade Propylene

Variables MPMR-Sweep UOP
(Polymer Grade) (Polymer Grade)
Feed Consumption
Propane (kglhr) 48526 49266
H20 (kglhr) 26168 -
Feed Preheating Energy Consumption (kJ/hr) 1.3 x 107 1.0 X 107
Reactor Energy Consumption (kJlhr) I.3 x 108 1.3 X 108
HzO Separation Energy Consumption (kJ/hr) M -
Hz Production (kg/hr) 2194 2124
Hz Separation Energy Consumption (kJ/hr) 3.7 x 107 3.6 X 107
MAPD Separation Energy Consumption (kJ/hr) 1.8 x 107 -
MAPD Column Stages 50 -
Propane-Propylene Splitter Energy Consumption (kJ/hr) M 7.4 x 107
Propane-Propylene Splitter Stages - 156 Figure 4 shows the effect of steam to propane ratio towards the propane conversion. The result shows that high level of propane conversion can be achieved at lower temperature at high steam/propane mole ratio. This can be understood since propane dehydrogenation is a decomposition reaction, so the presence of inert will shift the reaction to the product side.

Figure 5 shows effect of steam/propane mole ratio towards the propylene production. It can also be seen in this figure that high level of propylene production can be achieved at lower temperature using large steam/propane ratio.

However, propylene production decreases at a certain point of temperature at each steam/ propane ratio as the side reactions that produce methyl acetylene and propadiene begin to occur.

Temperature of 8000C and atmospheric pressure will be used as the operating conditions at certain ratio of steam/propane for second phase evaluation. The reason for selecting operating temperature has been described previously. Meanwhile the sweep/feed mole ratio in MPMR (see Figure 2) is fixed to be 6.73. In this case sweep is steam, while feed is propane. The selection of such value is based

walt represents propane feed to the sweep side of MPMR. The fraction of propane permeation is assumed to be 0.12 of the total propane feed, while its value for steam permeation is assumed to be 0.17. These values were based on experimental results using Vycor glass membrane (Silalahi 2003).

The evaluation of those process system scenarios were conducted using HYSYS veT. 2.4.1 process simulator. Tables 2 and 3 show the comparison of variables.

Propane Dehydrogenation in a Modified Porous Membrane Reactor 181

on several considerations such as optimum steam consumption, high level of propane conversion, high level of propylene production, and minimum productions of methyl acetylene and propadiene (MAPD).

Both scenarios, MPMR scenario and UOP scenario, are configured to produce both of propylene with polymer and chemical grade simultaneously. The configuration of the MPMR is modeled as split plug flow reactor; the first reactor represents the sweep side of MPMR for producing polymer grade propylene and the second one represents the feed side of MPMR for producing chemical grade propylene. The scenario of UOP process system is modeled the same as that of MPMR. The schematic description of both scenarios is shown in Figures 6 and 7. The difference between both scenarios is feed component. Feed components for MPMR scenario (feed) are propane and steam, while feed component for UOP scenario is propane only.

The reactant permeation or propane permeation from feed side through membrane

CONCLUSIONS

A study on propane dehydrogenation for producing propylene with polymer and chemical grades simultaneously in Modified Membrane Porous Reactor has been carried out. It is observed that the reaction can be conducted thermodynamically at atmospheric pressure, at BOOoe with a certain steam/propane ratio. Propylene with polymer grade is produced at the

'---- ,",,,po

Figure 6. MPMR Process System Scenario

C3HB

PROPANE\--- ... PROPYLENe SPLITTER

H2

Figure 7. UOP Process System Scenario

182 A. Trianto, Y. Bindar, and Noezran

sweep side outlet and propylene with chemical grade is produced at the feed side outlet of the MPMR. As shown above, for producing both propylene grades, the MPMR process system feed consumption is larger than UOP process system. MPMR process system produces hydrogen greater than UOP Process system. For producing polymer grade propylene, the energy consumption MPMR process system is larger than UOp, but in producing chemical grade propylene, MPMR uses less energy than UOP. Therefore, the overall energy consumption used by MPMR for producing polymer and chemical grades of propylene Simultaneously is less than the UOP process system. MPMR process system does not use propane-propylene splitter, because both of reactor products have small quantity of unconverted propane, which meets both of the commercial product specifications, 0.05% wt. for polymer grade propylene and 0.5% wt for chemical grade propylene. MPMR only requires MAPD column for methyl acetylene and propadiene separation that need less stages than propane-propylene splitter.

NOMENCLATURE

DGo standard Gibb's energy change, J

moie'

DHo standard enthalpy change, J mole 1

K reaction equilibrium constant

R ideal gas constant, 8.314 J moi' KJ

T temperature, K

ACKNOWLEDGEMENT

The funding from Indonesian Research Council through the RUT project for this work is gratefully appreciated.

REFERENCES

Cheng, L. 5., and Wilson, S. T. (2001). "Process for separating propylene from propane," U.S. Pat. 6,293,999.

Cheng, L. 5., and Wilson, S. T. (2001). "Vacuum swing adsorption process for separating

propylene from propane," U.S. Pat. 6,296,688.

Cottrell. (2001). "Method for improving the operation of a propane-propylene splitter," U.S. Pat. 6,218,589.

Indonesian Chemical Industries Club. (2001).

ICIC Report, rCI, Jakarta.

Kokugan, T., and Trianto, A. (2001). "Catalytic reactor having partition wall comprising porous membrane, method of manufacturing chemical substances using the same, and reaction apparatus using reactor," Jap. Pat. 2001157834.

Silalahi, S. A., and Rahman, A. Z. (2003). Study on propylene production by propane dehydrogenation in a membrane reactor. Bachelor's thesis, Institut Teknologi Bandung.

Trianto, A (2001). The development and analysis of a new porous membrane reactor system for dehydrogenation reactions. Ph.D. Dissertation, Tokyo University of Agriculture and Technology, Japan.

Trianto, A, and Kokugan, T. (2001). "Method for improving the performance of porous membrane reactor," Journal Chern. Eng. Jpn., 34, 199-206.

Trianto, A., Kokugan, 1., and Fukushima, 1. (1999). "A novel method for improving the performance of porous membrane reactor fed with pure reactant," Proceedings of 8h APPChE Congress, 3, 1529-1533.

WeIrauch, W. (1999). "Chern. system petrochemical outlook," Hydrocarbon Processing, March, 29-35.

Wood, B. J. (1968). "Dehydrogenation of cyclohexane on a hydrogen-porous membrane," J. Catal., 11, 30-34.