Computers and Chemical Engineering 26 (2002) 563– 573

www.elsevier.com/locate/compchemeng

Optimisation of azeotropic distillation columns combined with

pervaporation membranes
Ana M. Eliceche a,*, M. Carolina Daviou a, Patricia M. Hoch a,
Inmaculada Ortiz Uribe b
a
Chemical Engineering Dept., PLAPIQUI, Uni6ersidad Nacional del Sur, CONICET, Camino La Carrindanga km 7,
8000 Bahı́a Blanca, Argentina
b
Dpto. de Quı́mica, Uni6ersidad de Cantabria, A6da los Castros s/n, 39005 Santander, Spain

Abstract

The main objective of this work is the analysis and optimisation of azeotropic distillation columns, when a liquid side stream
with the distributing non-key component is treated in a pervaporation membrane and the retentate is recycled to the column. The
objective is to separate the pure distributing non-key component using a pervaporation membrane, thus helping to improve the
purity in the top and/or bottom products. The operating conditions of the column such as reflux ratio, product and side draw
flowrates and pressure are selected optimally to minimise the operating cost of the hybrid system. The simulation of the
pervaporation membrane units is implemented solving the differential– algebraic equation system of mass transport and energy
balances. The optimisation of the operating conditions of the azeotropic distillation column in the hybrid distillation/pervapora-
tion system for Methyl tert-butyl ether production is presented. Numerical results with a significant reduction in the operating cost
are reported. © 2002 Elsevier Science Ltd. All rights reserved.

Keywords: Azeotropic distillation; Pervaporation; Optimisation; MTBE

Nomenclature

B bottom flow rate (kg mol/h)

Cc condenser operating cost ($/h)
Cm,r cost of membrane replacement ($/membrane unit)
Co,i intermediate heating operating cost ($/h)
Cp pervaporation membrane operating cost ($/h)
Cr reboiler operating cost ($/h)
Co condenser operating cost coefficient ($/kJ)
Co,ref operating cost coefficient for condensing the permeate ($/kJ)
Co,i operating cost coefficient for intermediate heating between modules ($/kJ)
Co,r reboiler operating cost coefficient ($/kJ)
D distillate flow rate (kg mol/h)
E sidestream (kg mol/h)
F1 fresh feed flow rate (kg mol/h)
F2 retentate flow rate (kg mol/h)
NE location of the side draw
NF1 location of fresh feed

* Corresponding author.
E-mail address: meliceche@plapiqui.edu.ar (A.M. Eliceche).

0098-1354/02/$ - see front matter © 2002 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 9 8 - 1 3 5 4 ( 0 1 ) 0 0 7 7 5 - X
564 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
NF2 location of the recycle stream (retentate)
Qc heat withdrew from the condenser of the column (kJ/h)
Qi heat added between pervaporation modules
Qr heat added to the reboiler of the column (kJ/h)
Qref refrigeration needed for the pervaporation modules (kJ/h)
R reflux ratio
recMTBE,B MTBE recovery in the bottom product
xMTBE,B liquid composition of MTBE in the bottom product
1. Introduction
Debottlenecking and azeotrope breaking are fruitful
fields for hybrid membrane systems. Some studies pro-
pose a combination between distillation columns and
pervaporation membranes as an interesting alternative
for solving these problems in industrial processes. The
original Hüls process to obtain pure Methyl tert-butyl
ether (MTBE) (Fig. 1) involves several azeotropic distil-
lation columns to obtain a high recovery and purity of
MTBE as end product and methanol to be recycled to
the reactor. Hybrid distillation-pervaporation processes
are generally less energy-consuming than distillation,
and the separation taking place through the membrane
is not influenced by the equilibrium between compo-
nents. Recent patents propose hybrid distillation/perva-
poration technologies for azeotrope breaking processes
involving the separation of alcohols and ethers (Chen,
Eng, Glazer, & Wensley, 1988; Chen, Markiewickz, &
Venugopal, 1989) applied to the MTBE process, replac-
ing the Hüls process. The pervaporation membrane
used shows high flux and high selectivity to the perme-
ation of Methanol, effectively breaking the azeotrope
Fig. 1. Schematic Hüls MTBE production process.
Methanol–MTBE. The process called ‘Total Recovery
Improvement for MTBE’ (TRIM™), is a combination
of an organophilic pervaporation membrane and distil-
lation, using two different layouts. One with the perva-
poration membrane between the reactor and the
distillation column, and the second layout with the
membrane treating a sidestream of the column (Fig. 2).
When the membrane is used to treat the sidestream, the
required area decreases and investment costs could be
reduced by 20%. The integration of the TRIM™ pro-
cess to an existing one would be attractive if the
production could be increased by 5%.
In a recent review of Lipnizki, Hausmanns, Ten,
Field, and Laufenberg (1999a) several processes of
organophilic pervaporation are described. Even when
the medical applications of membrane processes like
dialysis are more profitable than any other industrial
membrane process because of higher annual sales, in a
near future, due to its high potential to separate or-
ganic–aqueous and organic–organic mixtures, it is
forecast that pervaporation market share will grow at
rates higher than any other membrane application. The
authors also show a measure of the interest in pervapo-
 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
Fig. 2. Schematic TRIM™ process.
ration with a graph showing how the number of papers
on pervaporation increased over the last 20 years.
Lipnizki, Field, and Ten (1999b) present an extense
review of pervaporation-based hybrid processes, focus-
ing on industrial applications and pointing out the need
of optimisation of some of the analysed processes.
One of the fields where there is need of optimisation
is the distillation/pervaporation hybrid process. Höm-
merich and Rautenbach (1998) studied the integration
of pervaporation and vapour permeation into the Hüls
process. They analysed the influence of the operating
conditions in a hybrid distillation–pervaporation– va-
pour permeation system for the MTBE production and
implemented the simulation of the process in ASPEN.
González and Ortiz Uribe (2001a,b) carried out experi-
mental work to find a semi-empirical model for the
pervaporation membrane to separate methanol and
MTBE, and simulated the hybrid distillation/pervapo-
ration process using gPROMS (PSEnterprises, 2000)
performing a cost analysis.
However, the formal optimisation of hybrid distilla-
tion/pervaporation process has not been reported previ-
ously. In particular, the optimisation of the debutanizer
column with pervaporation membrane units to treat the
side stream is reported in this work. Previous work on
the optimisation of non-conventional distillation
columns (Hoch & Eliceche, 1991) neither contemplate
the azeotropic distillation case nor include simulta-
neously the rigorous simulation of pervaporation mem-
branes. Thus the main objective of this work is the
optimisation of the azeotropic distillation column in a
hybrid distillation/pervaporation system using a rigor-
ous simulation of the azeotropic distillation column and
the pervaporation membrane units. Recent work (Höm-
merich & Rautenbach, 1998; González & Ortiz, 2001b)
explored the influence of the optimisation variables
565
performing the simulation of the process at different
operating conditions. As previously stated, the optimi-
sation of the hybrid distillation–pervaporation process
has not been attempted in former work. The formal
optimisation of the hybrid system that includes an
algebraic and differential equation system is described
in this work. Numerical results of the optimisation are
shown for the separation of a mixture of butanes,
methanol and MTBE. This work was motivated by the
evaluation of the feasibility of revamping the MTBE
separation sector of a refinery.
2. Distillation –pervaporation process for MTBE
production
MTBE has been the chemical with the highest in-
crease in its rate of production, a notorious 20%/year
during the years 1985–1995 (Trevisan, 1995). MTBE is
used as a high octane fuel additive. It provides the fuel
with additional oxygen, decreasing the amount of CO
produced as it benefits combustion completeness.
The production process of MTBE consists of a reac-
tion sector, where i-C4H10 is combined with methanol
to form the ether, and a separation sector where all the
MTBE must be separated from unreacted methanol
and C4’s. Unreacted methanol forms azeotropic mix-
tures with MTBE and butanes. In the conventional
Hüls process, a sequence of azeotropic distillation
columns was used to break the azeotropes, thus recov-
ering MTBE and methanol to be recycled to the reactor
(Fig. 1). A hybrid distillation–pervaporation process
seems as an attractive alternative, because it combines
the advantages of both methods. The use of hybrid
systems can improve the cost of the traditional Hüls
separation sequence as reported recently (Hömmerich
& Rautenbach, 1998).
566 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
Different configurations for the hybrid distillation/
pervaporation process can be used, locating the perva-
poration membrane to treat the feed, products or side
stream. In this work, the pervaporation membrane is
located to treat the side stream and remove the dis-
tributing component as a permeate, helping to improve
the top and bottom purity.
The separation of pure MTBE as the bottom product
of an azeotropic distillation column from a mixture of
C4’s, methanol and MTBE is performed by means of a
combined distillation column and pervaporation mem-
brane process. The process being studied is shown in
Fig. 3, where the pervaporation process is synthesized
as a single unit.
The azeotrope formation of methanol with both
MTBE and C4 limits the purity of the products. A high
purity of MTBE is required in the bottom product (B)
of the column that separates a multicomponent mixture
(F1) including MTBE, C4’s and methanol. A side
stream (E) of the column, rich in methanol, is pro-
cessed through the membrane units. The membrane
selectivity allows the methanol in the column side-
stream to be permeated and then condensed and recy-
cled to the reactor (Permeate liquid stream), thus
helping to improve the MTBE bottom product purity.
The retentate is recycled to the column (F2). The
butanes do not permeate through the membrane, and
this is confirmed by some patents of the hybrid process
(Chen et al., 1988) and experimental results carried out
by González (2000).
3. Pervaporation membrane model
A scheme of a pervaporation module is shown in Fig.
4.
Fig. 3. Schematic flowsheet for the distillation/pervaporation process.
For a binary system, where the permeation of one of
the components Eq. (1) is favoured over the other
component Eq. (2), the equation system representing
the mass transport and energy balances can be synthe-
sized as (González, 2000):
Total flux: J(c, T)= Jtotal(c, T)= J1(c, T)+ J2(c, T),
(1)
where 1 refers to methanol and 2 refers to MTBE.
Recently, a mathematical model based on the general-
ized Fick%s law and the assumption that transport
through the membrane is the rate-limiting step was
developed in order to describe the pervaporation flux of
both components (González & Ortiz Uribe, 2001a). The
prediction of the flux of methanol needed a concentra-
tion-dependent diffusion coefficient whereas a simple
model with concentration-independent diffusivity was
sufficient for the description of the MTBE flux. Pre-
dicted results with the mathematical model and
parameters agreed with experimental data obtained in a
wide range of variables working with Sulzer’s Per-
vap2256 membranes. Further details of the pervapora-
tion modelling can be found elsewhere (González &
Ortiz Uribe, 2001a). Thus, the resulting equations for
the prediction of pervaporative fluxes are,
J1(c, T)
= [− ƒ1Do exp(~ƒ1)k2,1z1/(ln(k2,1ƒ1 + 1)(k2,1ƒ1 +1))]
× dƒ1/dzp (2)
J2(c, T)= − Dz2dƒ2/dzp. (3)
Boundary conditions: At zp = 0
ƒ1 = [exp(k1x1,1/k1,1)− 1]/k2,oh (4)
ƒ2 = k2x2,1/k1,2 (5)
 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573 567

Fig. 4. Pervaporation membrane.

At zp =t c%= zF (13)
ƒ1 =[exp(P2x1,2/(P v1(T)k1,1)) − 1]/k2,1 (6) T= TF (14)
ƒ2 =(P2x2,2/(P (T)k1,2)),
v
2 (7) Values for fluxes and enriching factor are obtained
from experimental data. The parameters were found for
where ki is the activity coefficient for component i concentrations of methanol between 5 and 40% in a
obtained with a gE model, and ƒi the composition of binary mixture methanol–MTBE. For different values
component i. of methanol composition in a binary mixture of
dm; methanol– MTBE, the values for fluxes were found at
Global mass balance: − =WJ(c, T) (8) different temperatures ranging between 30 and 50 °C.
dzA
The experimental results can be found in González and
Component mass balance: Ortiz Uribe (2001a). This allows to find the necessary
dci J(c, T) parameters for modelling the mass transport in the
− =[ii (c, T) −1]ci W, (9) membrane. For a certain composition, the flux of com-
dzA m;
ponent i varies with the temperature according to the
where W is the characteristic membrane width, ci is the expression:
mole fraction of component i in the permeate, T the
temperature along zA and ii is the enriching factor, Ji,T 1(c, T)= Ji,T 2(c, T)exp
 
Ea,i 1
−
1  , (15)
defined as: R T2 T1

c%i where Ea,i is the apparent energy of activation, found

ii (c, T)= (10) using the experimental results at different temperatures.
ci
The value of Ea,i is independent of the temperature,
dT WJ(c, T)u m V (c%, T) thus fluxes can be slightly extrapolated to other temper-
Energy balance: − = , (11)
dzA m; C m
pL(c, T) atures different than the experimental, while the whole
model should be used for concentrations on the reten-
being uV the heat of vaporization and CpL (J/kmol K)
tate side lower than the maximum experimental concen-
the mass heat capacity of the liquid mixture.
tration used to find the correlated values of the
For Eqs. (8), (9) and (11), at zA =0,
parameters used to describe the flux through the
m; = F (12) membrane.
568 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
As seen from the equations, the pervaporation mem-
brane is modelled in terms of a differential/algebraic
system involving lumped and distributed parameters.
The differential equations through the membrane must
be discretized in the direction of zp, and an integration
package for Initial Value Problems (IVPs) is used to
solve the equations along zA. As a simplification, it is
assumed that only two components permeate through
the membrane, while the rest of the components remain
at the retentate side and do not permeate. This simplifi-
cation was experimentally tested, and holds with the
pervaporation membrane Pervap2256 used to find the
experimental data.
The membrane favoured the permeation of methanol
over MTBE so that Eqs. (1)– (15) could be used. The
remaining components do not permeate through the
membrane, so the assumption of binary fluxes is hold.
Experimental data were available for the system being
studied, so there was information regarding the values
of the flux and retentate concentration (González,
2000), which allowed the implementation of this model.
4. Simulation of the hybrid distillation/pervaporation
system
The hybrid process of an azeotropic distillation
column combined with several membrane modules to
treat the side draw was simulated with HYSYS (1999).
Each membrane module was rigorously modelled using
Eqs. (1)–(15). The software gPROMS was used to
implement the differential– algebraic system, then the
whole model was translated into Fortran language, to
allow the implementation of a user extension unit in
HYSYS. The differential– algebraic equation solver
DASOLV (Jarvis & Pantelides, 1992) was used to inte-
grate the IVP along zA, while the differential equations
in the direction of zp were approximated using first
order backwards finite differences for the derivatives in
40 finite elements, thus obtaining a differential–alge-
braic non-linear equation system. The choice of using
an IVP integrator was made because this allows to
predict the required membrane area in order to achieve
a desired separation, by means of fixing the desired
composition of the permeant in the retentate stream or
by letting it free as optimisation variable. This is the
main difference between the implementation of the
proposed model and the model shown in González and
Ortiz Uribe (2001a). For the simulation, thermody-
namic properties for the distillation column and the
pervaporation membrane were predicted using Wilson
equation with user provided binary interaction parame-
ters, taken from Espinosa, Aguirre, and Pérez (1995).
This method allowed to accurately predict the thermo-
dynamic behaviour of the mixture being analysed.
The pervaporation takes place when a vacuum of
15–20 torr is maintained at the permeate side. For the
start-up a vacuum pump is needed, but when steady
state is reached, vacuum is maintained by the permeate
condensation. The distillation column has 24 theoretical
stages plus a condenser (stage 0) and a reboiler (stage
25). The membrane can be purchased in modules of 30
m2, so several modules are needed to perform the
separation task. The number of modules required is
automatically selected by fixing the required composi-
tion of the permeant in the retentate, thus allowing the
introduction of as many intermediate heaters as re-
quired. For the integration, it was assumed a character-
istic width of the module (W) of 50 m, and the
integration proceeds up to zA = 0.6 m for each module.
Intermediate heating is provided, and the retentate
stream is heated until the temperature reaches 83 °C.
The purpose of the intermediate heating is to avoid the
decrease of the flux across the membrane.
5. Formulation of the optimisation problem
The optimum operating conditions of the distillation
column, such as reflux ratio, side draw and product
flow rates and pressure are obtained solving an optimi-
sation problem to minimise the operating cost of the
hybrid distillation/pervaporation system. The operating
cost is calculated as the summation of column and
membrane units costs:
Co = Ccolumn + Cmemb = (Cc + Cr + Cp) [$/h] (16)
Cc = Co,cQcon,
Cr = Co,rQreb,
n
Cp = Co,pQperv + % Cc,i Qint,i + Cm, (17)
i=1
where Qcon is the heat (kJ/h) withdrew from the con-
denser of the column, Qreb (kJ/h) is the heat added to
the reboiler of the column, Qperv (kJ/h) is the heat
withdrew in the pervaporation unit to condense the
permeate, Qint,i is the heat added after the stage i, n is
the number of pervaporation modules used and Cm is
the cost of replacing the membrane. The operating cost
of replacing the membrane modules is included in the
objective function considering that they must be re-
placed every 2 years. For the operation of the con-
denser in the pervaporation membranes, a refrigerant
fluid is needed, because the methanol condenses at very
low temperatures (− 5 °C at the permeate pressure,
usually 0.02 bar). The refrigeration cost to condense the
permeate is a function of this temperature. Details on
the operating cost are presented in Appendix A. The
optimisation problem for the hybrid distillation/perva-
poration system is formulated as:
 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573 569

Table 1 Table 3
Feed to the column Cost results for the example shown

Component Flow rate (kg mol/h) Cost ($/h) Optimum Percentage

i-butene 325.82 Condenser 2.10 1.45

MTBE 155.00 Reboiler 127.94 88.42
Methanol 32.96 Refrigerant 4.53 3.13
F1 (kg mol/h) 512.78 Heating between PV modules 4.77 3.30
Temperature (K) 334.30 Membrane replacement 5.36 3.70
Pressure (kPa) 685.00
Total 144.70 100.00

min Co(x)
x configurations and combinations of units can be imple-
s.t. g(x)50
b L 5x5 b U,
where the optimisation variables x are the reflux ratio
(r), a product flow rate (B), the side draw flow rate (E)
and the column pressure (P), b L and b U are lower and
upper bounds on the variables x.
The separation objectives like minimum composition
or recovery in the products and operating constraints
related with the maximum sidedraw flowrate are formu-
lated as inequality constraints g(x).
The simulation of the hybrid system and the solution
of problem Eq. (18) is implemented in HYSYS. The
membrane modules are modelled in a user extension
unit in HYSYS. For each set of x variables provided by
the HYSYS optimisation code, a rigorous simulation of
the hybrid system is carried out including the solution
of the differential equations to model the pervaporation
membranes. HYSYS is a sequential modular simulator,
therefore the equality constraints modelling the pro-
cesses are not explicitly formulated as equality con-
straints in problem Eq. (18). This includes the
differential equations used to model the membrane
units.
The software developed allows the incorporation of
different membrane models as user extension units in
HYSYS. Thus the optimisation of similar technologies
like the production of ethyl tert-butyl ether (ETBE) or
other hybrid separation processes combining distillation
columns with pervaporation membranes in different
mented adapting the software developed. Once the rig-
orous simulation of the pervaporation membrane
(18)
module is generated and interfaced automatically in
HYSYS, the optimisation of different systems is feasi-
ble by modifying the membrane models and implement-
ing a solution strategy for the system of differential
equations.
6. Numerical results
A mixture of iso-butene, MTBE and methanol has to
be separated to obtain MTBE as a bottom product of
the column. A hybrid distillation/pervaporation process
is analysed to perform this task. The liquid side draw is
located above the fresh feed and the retentate from the
membrane is recycled to the column in the stripping
section. The distillation column has 26 stages, the liquid
side draw is located in stage 4, the fresh feed in stage 9
and the permeate from the membrane modules is recy-
cled in stage 20. The composition, temperature and
pressure of the fresh feed (F1) are shown in Table 1.
For the sake of clarity, the mixture of butanes of the
original process is replaced by one component, iso-
butene, which forms azeotrope with methanol, so a
ternary mixture is considered instead of the original
multicomponent mixture. MTBE also forms an
azeotropic mixture with methanol at the working
pressure.

Table 2
Optimisation results for the example shown

Initial point Optimum point Lower bound Upper bound

Side stream (kg/h) 4900 5542 3000 7500

Reflux ratio 1.22 0.9515 0.9 1.25
Pcondenser (kPa) 600 572.5 300 1200
Bottom flow rate (kgmol/h) 162.98 161.96 160 170
Bottom specification (wt.%) 98.02 98.02 98.02 –
MTBE recovery (%) 99.62 99 99 –
E/Liq ratio (%) 27.31 40.93 – 41
Number of PV modules 26 22
Cost ($/h) 160.26 144.7
570 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
In this work, the objective is to obtain grade gasoline
MTBE 98 wt.% with a recovery of 99% in the column.
The inequality constraints g(x) of problem Eq. (18) are:
Minimum MTBE recovery in the recMTBE,B
bottom product
] 0.99
Minimum MTBE composition in the xMTBE,B
bottom product
] 0.98
Maximum value of sidedraw E50.41 *
Lj−1
being j the stage from where the side draw is extracted
and Lj − 1 the liquid flow rate of stage ( j −1).
Optimisation results with the initial and final solution
points, upper and lower bounds used in problem Eq.
(18) are shown in Table 2. An important reduction in
the reflux ratio of 22% and a significant reduction of
9.7% in the operating cost have been achieved by
selecting the optimum operating conditions.
The operating cost of the distillation column is nearly
90% of the total cost and the membrane cost is nearly
10% as reported in Table 3. Thus the optimisation of
the operating conditions of the distillation column has
a great impact in reducing the energy consumption and
operating cost.
The minimum composition and recovery are active
constraints. The constraints related to the minimum
composition and recovery of MTBE in the bottom
product of the column are active because the cost of the
column increases with higher composition and recov-
ery, requiring a higher reflux ratio.
A fixed composition of methanol in the retentate of
6% wt C4-free basis (equivalent to 15% mole-basis) is
Fig. 5. Temperature profiles in the pervaporation modules.
assumed to perform the pervaporation process, while
González and Ortiz Uribe (2001b), Hömmerich and
Rautenbach (1998) have used 5 wt.% The number of
pervaporation modules required to perform this separa-
tion is calculated.
The HYSYS selected optimisation option to solve
problem Eq. (18) is the Mixed method. It starts the
optimisation with the Box method using a very loose
convergence. After convergence, an SQP method is then
used to locate the final solution using the desired toler-
ance. The optimisation variables selected were the side
draw flow rate, the reflux ratio and the pressure of the
column.
It is assumed that cooling water is used to condense
the distillate of the column and low pressure steam is
used to heat the bottom of the column and to provide
interstage heating between the membrane modules.
For this example, 22 modules are required to obtain
the specified composition of methanol in the retentate.
The temperature profile for the pervaporation modules
is shown in Fig. 5.
The retentate methanol molar composition decreases
from 40 to 15% (C4-free basis), while the permeate
composition remains above 99% throughout the mem-
branes. Composition profiles in the membrane are
shown in Fig. 6a and b. It is considered that only
MTBE and methanol permeate through the membrane,
while the iso-butene remains at the retentate side (Chen
et al., 1988).
Liquid composition profiles for the distillation
column, when a side stream is withdrawn and the
retentate is recycled to the column, are shown in Fig. 7.
Fig. 8 shows in detail the methanol composition
profile considering the ternary system, with a clear
 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573 571

Fig. 6. (a) Composition profiles of the permeate. (b) Composition profiles of the retentate.

reference to the stages where the fresh feed, recycle and
side stream are located. There is no formal exploration
of other alternatives for the location of these streams,
as the purpose of this work is to optimise the operating
conditions only, with no structural changes involved.
The extraction stream is optimally located, in the max-
imum value of the methanol composition profile in the
rectification section.
The location of fresh feed, recycle and side draw
have not been modified during the optimisation, as the
objective of this work is to improve the operating cost
without changes in these variables, although the place-
ment of the fresh feed, side stream extraction and
retentate recycle locations can be modified to improve
the objective function, while maintaining the total num-
ber of stages fixed. The procedure developed by Hoch
and Eliceche (1991) treat the number of stages as
continuous optimisation variables, allowing the simul-
taneous selection of feed and side stream locations with
the flowrates in a non-linear programming problem
572 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573

formulation. The implementation of this option is cur-
rently being studied, as it is not possible to do it within
the HYSYS environment in an algorithmic fashion.
Instead, a trial-and-error procedure can be performed
and was initially used to locate the side draw, fresh feed
and recycle from the membranes reported in Table 2.
The cost of the process has been improved from 156
to 148 $/h, representing a reduction of approximately
10% on the total cost of the process. It was not the
purpose of this work to compare the hybrid system cost
with the cost of the original Hüls process, but from the
data in Table 3 it can be seen that the operating cost of
the reboiler of the column represents almost 90% of the
total cost, thus replacing a distillation column with the
pervaporation membrane seems to be economically
attractive.

Fig. 7. Column liquid composition profile.

Fig. 8. Methanol liquid composition profile.

 A.M. Eliceche et al. / Computers and Chemical Engineering 26 (2002) 563–573
7. Conclusions
The optimisation of the operating conditions of an
azeotropic distillation column in a hybrid distillation/
pervaporation system has been presented. A rigourous
modelling of the pervaporation membrane units has
been implemented. Numerical results are reported for
MTBE production, and the main improvements that
can be expected by optimising the operating conditions
of the debutanizer column are shown. A reduction of
9.7% in the operating cost of the distillation/pervapora-
tion process has been achieved. Although in this work
the option of treating the side stream with pervapora-
tion membranes has been analysed, the same approach
can be used to study different configurations, for exam-
ple pre-treatment of the feed or end products purifica-
tion with pervaporation membranes. The methodology
can be extended to similar technologies like the produc-
tion of ETBE or other hybrid separation process
combining distillation and pervaporation, where similar
reductions in the operating cost can be expected.
Acknowledgements
This work was carried out under research grant PICT
14-04065 from ANPCyT, Argentina.
Appendix A. Operating cost evaluation
The operating cost of the process is calculated as the
sum of operating costs for the condenser and reboiler
of the column plus the operating costs of the pervapo-
ration membrane,
Co =Cc + Cr +Cp +Cm, (A.1)
where
Qcond
Cc =Co,c , (A.2)
DTw
is the operating cost of the condenser of the column,
Cr = Co,rQreb, (A.3)
is the operating cost of the reboiler of the column,
Cp =Cref + Cheat, (A.4)
is the operating cost of the pervaporation module,
Cref = Co,refQref, (A.5)
is the cost of condensing the permeate in order to
573
maintain vacuum,
Cheat = Co,i Qi, (A.6)
is the interstage heating operating cost, and
Cm = nCm,ra, (A.7)
is the cost of membrane replacement, with n, number of
pervaporation modules, Cm,r the replacing cost of each
module. The annuity a is calculated considering an
inflation rate of 10% a year. Cost coefficients can be
found in Seider, Seader, and Lewin (1999).
References
Chen, M. S., Eng, R. M., Glazer, J. L., & Wensley, C. G. (1988).
Pervaporation process for separating alcohols from ethers, U.S.
Patent 4774365.
Chen, M. S., Markiewickz, G. S., & Venugopal, K. G. (1989).
Development of membrane pervaporation TRIM™ process for
methanol recovery from CH3OH/MTBE/C4 mixtures. American
Institute of Chemical Engineering Symposium Series, 85, 82–88.
Espinosa, J., Aguirre, P. A., & Pérez, G. A. (1995). Product composi-
tion regions of single feed reactive distillation columns: mixtures
containing inerts. Industrial and Engineering Chemistry Research,
34, 853 – 861.
González González, B. PhD. Thesis, Universidad de Cantabria
(2000).
González, González B., & Ortiz Uribe, I. (2001a). Mathematical
modelling of the pervaporative separation of Methanol –Methyl
tert-Butyl Ether mixtures. Industrial and Engineering Chemistry
Research, 40, 1720 – 1731.
González, González B., & Ortiz Uribe, I. (2001b). Modelling and
simulation of a hybrid process (pervaporation – distillation) for the
separation of azeotropic mixtures of alcohol – ether. Journal of
Chemical Technology and Biotechnolgy, 76, 1 – 14.
Hoch, P. M., Eliceche, A. M. (1991). Optimal design of non-con6en-
tional distillation columns, Process Technology Proceedings vol. 10,
Computer Oriented Process Engineering, pp. 369 – 374.
Hömmerich, U., & Rautenbach, R. (1998). Design and optimization
of combined pervaporation/distillation processes for the produc-
tion of MTBE. The Journal of Membrane Science, 146, 53–64.
Hyprotech, HYSYS User Manual (1999).
Jarvis, R. B., Pantelides, C. C. (1992). DASOLV: a differential –alge-
braic equation solver, CPSE report series.
Lipnizki, F., Hausmanns, S., Ten, P. K., Field, R. W., & Laufenberg,
G. (1999a). Organophilic pervaporation: prospects and perfor-
mance. Chemical Engineering of Journal, 73, 113 – 129.
Lipnizki, F., Field, R. W., & Ten, P. K. (1999b). Pervaporation-based
hybrid process: a review of process design, applications and
economics. Journal of Membrane Science, 153, 183 – 210.
PSEnterprises (2000). gPROMS v. 1.8.4 User Guide.
Seider, W.D., Seader, J.D., Lewin D.R. (1999). Chapter 10:
Profitability analysis. Process design principles, John Wiley and
Sons (ed.).
Trevisan, M. (1995). Lead-free thanks to MTBE, Sulzer Technical
review, 1, pp. 35 – 37.