SlideShare a Scribd company logo
PRODUCTION OF n-PROPYL ACETATE BY
REACTIVE DISTILLATION
Experimental and Theoretical Study
M. Brehelin1
, F. Forner2
, D. Rouzineau1
, J.-U. Repke2
, X. Meyer1
,
M. Meyer1Ã
and G. Wozny2
1
Laboratoire de Ge´nie Chimique, INP ENSIACET, Toulouse, France.
2
Institute of Process and Plant Technology, Technische Universita¨t Berlin, Germany.
Abstract: First steps of the development of a catalytic reactive distillation process for the pro-
duction of n-propyl acetate based on experiments and simulations are proposed. The kinetics
for homogeneously (sulphuric acid) and heterogeneously (Amberlyst 15) catalysed reaction
were investigated and the constants for a pseudo-homogeneous model are presented. Pilot
plant experiments were performed using a homogeneous strong acid catalyst in a packed
column. A top-column decanter is used to withdraw the aqueous phase and to reflux the organic
phase. Simulation results are in good agreement with experimental data. Thermodynamics non-
idealities are taken into account using VLE and LLE NRTL interaction parameters. Alcohol conver-
sion and n-propyl acetate purity may be dramatically increased just by adding to the pilot plant a
stripping section in an additional column: six different configurations are identified to achieve such
a production. The startup is studied in order to determine the best strategy to achieve steady-state
conditions. The strong influence of the composition of the initial charging in the decanter can be
seen and an initial charging of the two-phase top product leads to the fastest startup.
Keywords: reactive distillation; kinetics; startup; experimental data; n-propyl acetate.
INTRODUCTION
The combination of reaction and distillation in
one reactive distillation (RD) unit can lead to
significant reduction of investment and oper-
ational cost. However, interactions between
reactions and separation increase complexity
and require a better understanding of the pro-
cess. Whereas RD is intensively studied for
years, only few works have been published
on vapour–liquid–liquid (VLL) systems, thus
our investigations focus on VLL reactive
systems.
Experimental investigations have already
been published for the production of butyl
acetate (Steinigeweg and Gmehling, 2002)
or amyl acetate (Chiang et al., 2002) but, to
the best of our knowledge, none for n-propyl
acetate reactive distillation; thus, information
about n-propyl acetate synthesis can hardly
be found in literature. Investigations on reac-
tive membrane separation have led Huang
et al. (2005) to publish data on chemical equi-
librium. Bart et al. (1996) have published a
kinetic modelling of the n-propyl acetate syn-
thesis catalysed by an ion exchange resin
(Dowex) as well as Steinigeweg (2003) with
the Amberlyst-15. Unfortunately, uncertainty
remains on the chemical equilibrium constant
value. Finally, no data about reactive distilla-
tion pilot-plant experiments is published.
However, reactive distillation column
design requires reliable kinetic data in
addition to other crucial information such as
phase equilibria. Methods carried out to
assess the reactive distillation process feasi-
bility are often based on residue curve maps
(RCM) for non-reactive distillation, and on
reactive residue curve maps (rRCM) for reac-
tive distillation. In addition to the four pure
components, three azeotropes are the singu-
lar points of the n-propyl acetate synthesis
reaction rRCM (Table 1). Two of these
azeotropes (PrOH/PrOAc and PrOH/H2O)
are remaining from the non-reactive RCM
although two others involving reactive species
(PrOAc/H2O and PrOH/PrOAc/H2O which
are both heterogeneous) disappear; the last
singular point of this rRCM is a quaternary
reactive azeotrope. The chemical equilibrium
constant value has a great influence on the
liquid phase stability of this point (but not on
its topological stability since it is always an
unstable node): when a value of 20 (Huang
et al., 2005) is used, the reactive azeotrope
is homogeneous, although a value of 48.8
(Okasinski and Doherty, 2000) leads to a
heterogeneous reactive azeotrope.
109 Vol 85 (A1) 109–117
Ã
Correspondence to:
Dr M. Meyer, Laboratoire de
Ge´nie Chimique, INP
ENSIACET, Toulouse,
France.
E-mail: michel.meyer@
ensiacet.fr
DOI: 10.1205/cherd06112
0263–8762/07/
$30.00 þ 0.00
Chemical Engineering
Research and Design
Trans IChemE,
Part A, January 2007
# 2007 Institution
of Chemical Engineers
These effects are clearly illustrated in Figure 1 for which
rRCM have been computed using a previous work on the
rRCM calculation taking into account rigorous VLLE equilibria
(Brehelin et al., 2006); the transformed variable coordinates
(Ung and Doherty, 1995) are used with n-propyl acetate as
the reference component. For the sake of better represen-
tation of the chemical and phase equilibria complexity, rRCM
in molar composition are represented too in Figure 1 (bottom).
Table 1. Singular points of n-propyl acetate synthesis reaction rRCM.
System T (8C)
Stability
Molar compositionTopological Physical
HOAc 117.9 SN
PrOAc 101.5 Sa
H2O 100.0 Sa
PrOH 97.2 SN
PrOH/PrOAc 94.7 Sa Homogeneous 0.625/0.375
PrOH/H2O 87.8 Sa Homogeneous 0.432/0.568
HOAc/PrOH/PrOAc/H2O 85.4 UN Homogeneous 0.069/0.309/0.146/0.476
84.0 Heterogeneous 0.059/0.245/0.262/0.435
Figure 1. n-Propyl acetate synthesis reaction rRCM in transformed (top) and molar (bottom) compositions showing the miscibility gap for two
values of the chemical equilibrium.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
110 BREHELIN et al.
As a consequence, great attention has to be paid to the
chemical and phase equilibria determination in order to
perform the design of such a process. Therefore, this article
presents the first steps of an approach to the development
of a RD process for the production of n-propyl acetate: the
catalysed reaction kinetics are firstly investigated in order to
employ a simulation environment based on an equilibrium
stage model; in parallel pilot plant experiments are carried
out in order to validate this simulation tool. On the basis of
these results, an optimal startup strategy is proposed and
several process configurations are studied in order to propose
the best setup for high purity n-propyl acetate production.
EXPERIMENTAL SECTION
Materials and Analysis
The chemicals used for the kinetic study were of analytical
grade (rectapur); Amberlyst 15 and concentrated sulphuric
acid were selected as catalysts. For the reactive distillation
experiments, the chemicals were of reaction grade (99.5%)
and sulphuric acid was the catalyst. All samples were analy-
sed by gas chromatography (Varian 3800) using cross-linked
polyethylene glycol CP-WAX 52 CB 30 m  0.32 mm column
with FID detector. For the homogeneously catalyzed kinetic
runs, the acetic acid concentration has been determined
using titration; additional titrations were performed with a
METTLER DL 35 Karl–Fischer-Titrator to measure the quan-
tity of water formed during the reaction.
A data reconciliation process was led in order to appre-
ciably increase reliability and consistency of results: the
elimination of aberrant measurements is performed on the
base of the atomic conservation (Fillon et al., 1995). An accu-
racy of molar fractions of 3% can thus be expected.
Kinetic Experiments
Kinetic experiments were conducted in a thermostated
three-neck glass reactor with a volume of 500 cm3
. One
neck was connected to a thermometer and another was
devoted to the GC sampling; a cold trap, in addition to a
reflux condenser, was installed in the third neck to avoid
any loss of volatile components. The temperature of the heat-
ing jacket was kept constant within +0.1 K. Before each
kinetic experiment was started, both reactants were brought
to reaction temperature separately; when the desired temp-
erature was reached, reactants and catalyst were added to
the reactor and time measurement was started. Liquid
samples of about 1 cm3
volume, without any catalyst in
case of Amberlyst 15, were taken using a syringe and
cooled to 253 K to avoid any further reaction. The contents
were stirred by a magnetic stirrer. The stirrer speed is fixed
in such a manner that it has no influence on the reaction
rate (Steinigeweg and Gmehling, 2002); thus, the external
mass transfer resistance seems not to be significant
(Pitochelli, 1980) and the internal mass transfer resistance
was neglected on the basis of a Weisz and Hicks criterion
calculation (Weisz and Hicks, 1962).
Reactive Distillation Experiments
Setup
Experiments were performed in a 4-m of glass column with
an inner diameter of 80 mm (Figure 2): the reactive section is
packed with 8 mm  8 mm glass Raschig rings and the recti-
fying section contained Sulzer-CY packing; the liquid phase
total hold-up estimated by the Engel’s method (Engel et al.,
2001) is of 80 ml. Each reactant is fed separately in the
column. Reaction was catalysed by homogeneous sulphuric
acid fed at the top of the reactive section with a weight
amount of 1.3% in the acetic acid flow. The reboiler, with a
liquid hold-up of approximately 10 dm3
was heated by an
oil-boiler and insulated by mineral wool whereas the column
was equipped with two couples of heating mantles; the maxi-
mal heating power measured was of 3 kW. Distillate was
collected in a decanter operated at ambient temperature
and all organic phase was refluxed at the top of the column.
Feed and product flows were measured by balances.
Temperatures were measured by Pt-100 thermometers
installed at the bottom of each section, as well as in the
feed stream, the reboiler and the condenser.
Startup
The startup strategy was to start with empty reboiler and
decanter; the preheated feeds and a gradual heat duty at
the reboiler led the mixture to boiling point. When a sufficient
amount of distillate was collected to the decanter the reflux
pump was turned-on and the set-point was adjusted in
order to reflux all the organic phase.
Results
Although this pilot plant has not been built especially for the
n-propyl acetate production by reactive distillation—its design
is far from optimal since an acetic acid conversion of 79%
with a molar purity of n-propyl acetate of 64% has been
achieved—dynamic and steady-state simulation can be per-
formed on the base of these experimental results.
Thus the aim of these runs was to obtain experimental data
useful to validate our simulations. The material balances of a
Figure 2. Conceptual schema of the pilot-plant used to perform
reactive distillation experiments.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 111
typical reactive distillation run summarized in Table 2 show
the reliability of a typical steady-state.
RESULTS AND ANALYSES
Kinetic Parameter Identification
A pseudo-homogeneous model [equation (1)] is used to
describe the reaction catalysed both with Amberlyst 15 or sul-
phuric acid. This choice is theoretically supported by the fact
that the heterogeneous catalyst swells in contact with polar
solvents making the sulfonic groups of the catalyst readily
accessible to reactants; in such a model, the ion-exchange
resin is thought to act as solvated protons; consequently,
rate constants can be expressed as a function of proton
amount; its temperature dependency is expressed by
Arrhenius’ law [equation (2)], so that four adjustable para-
meters (Kf,o, Ka,o, Ef, Ea) have to be fitted (Table 3).
r ¼
1
ni
dni
dt
¼ nHþ Kf aHOAcaPrOH À
aPrAcaH20
Ka
 
(1)
Ki ¼ Ki,0: exp À
Ei
R:T
 
(2)
The molar fractions of acetic acid, computed by the numeri-
cal integration of the kinetic equation, are compared with the
experimental ones. The activity coefficients were calculated
using the NRTL equation of which interaction parameters
are given in Table 4(a). Parameters are fitted by minimization
of the criterion defined by equation (3) using a least square
method. The mean value of this criterion, i.e., the mean rela-
tive deviation—assumed to be representing the accuracy of
the model—is inferior to 5%.
crit ¼ min
X
all samples
(xcalc
HOAc À xexp
HOAc)2
(3)
The theoretical dependence on temperature of the equili-
brium constant is shown in Figure 3: the obvious agreement
between our chemical equilibrium data and Huang’s ones
[almost 20 on the range temperature (80–1108C)] reinforce
the validity of our identification. Moreover experimental data
are related to the standard enthalpy of reaction by the van’t
Hoff equation: experimental and theoretical values of the
standard energy of reaction are concordant (21.02 and
21.69 kJ mol21
).
Simulation
Steady state
Steady-state simulations based on an equilibrium model
were carried out using the commercial software ProSim
Plus; the activity coefficients are derived from NRTL model
[Table 4(a)] and used for phase equilibriums as well as for
the kinetic model. It will be shown later that distillate stream
is free of acetic acid so specific liquid–liquid NRTL
parameters are used to compute phase equilibria into the
decanter [Table 4(b)]. Vapour phase association due to the
presence of acetic acid in the system has been accounted
for with the chemical theory (Marek and Standard, 1954).
Table 2. Material balances of a typical reactive distillation run.
HOAc
Feed
PrOH
feed
Aqueous
phase Bottoms I–O (%)
46.63 41.00 38.57 53.50 4.9
F (mol h21
) Molar composition Dn (mol)
HOAc 1 0 0.018 0.173 236.7
PrOH 0 1 0.008 0.131 233.7
PrOAc 0 0 0.004 0.638 þ34.3
H2O 0 0 0.970 0.058 þ40.6
Table 3. Parameters of the kinetic model for sulphuric acid and
Amberlyst 15 as catalysts.
Ko (mol s21.molHþ
21
) Ea (J mol21
)
Esterification (H2SO4) 7.699 Â 104
37,084
Esterification (A15) 2.576 Â 106
51,540
Chemical equilibrium 17.52 2366
Table 4. (a) NRTL parameters for the calculation of activity coeffi-
cients and phase equilibria (Huang et al., 2005; Okasinski and
Doherty, 2000).
I J aij (cal mol21
) bij(cal mol21
) a
HOAc PrOH 2 327.52 256.90 0.3044
HOAc PrOAc 2 410.39 1050.56 0.2970
HOAc H2O 2 342.20 1175.72 0.2952
PrOH PrOAc 2 340.02 111.74 0.3005
PrOH H2O 2 152.51 1866.34 0.3747
PrOAc H2O 2 667.45 3280.60 0.2564
Table 4. (b) Liquid–liquid NRTL parameters for the ternary PrOH/
PrOAc/H2O (Sørensen and Arlt, 1982).
I J aij (cal mol21
) bij(cal mol21
) a
PrOH PrOAc 286.17 467.23 0.2
PrOH H2O 21102.18 2949.20 0.2
PrOAc H2O 2602.82 3023.92 0.2
Figure 3. Natural logarithm of the equilibrium constant against reci-
procal temperature. This study (B) Huang et al. (A), Bart et al. (4),
Steinigeweg (...) and theoretical estimation based on enthalpy of for-
mation (– – –).
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
112 BREHELIN et al.
Figure 4 illustrates the good agreement observed between
experimental data and simulation (mean deviation of 3.5%
on molar composition).
Startup analysis
The startup of a reactive distillation column is a very com-
plex, time and energy consuming process. The dynamic pro-
cess behaviour becomes even more complex in case of a
miscibility gap at the column top which is influenced by the
reaction. Special attention has to be paid to the behaviour
of the decanter during startup in order to rapidly reach con-
ditions leading to the phase split. Reepmeyer et al. (2004)
have analysed the startup of homogeneously catalysed reac-
tive distillation in tray columns and have shown that initial
charging of product to the column trays can lead to significant
reduction of startup time. Tran (2005) have studied non-reac-
tive three phase distillation in a tray column with a decanter.
They have shown both by experiment and simulation that the
startup time depends strongly on the composition of the initial
hold-up in the decanter. Surprisingly, for the system ethanol/
water/cyclohexane, initial charging of ethanol (bottom pro-
duct) to the decanter led to a reduction of about 60% of
startup time. On the other hand charging with a two phase
mixture close to steady state composition extended the
startup time (Tran 2005).
The startup process which is considered in this contribution
begins when feed is entering the initially cold and empty
column. The dry packing is wetted and the column bottom is
filled up until the reboiler is switched on. Vapour starts rising
in the column and partly condenses on the cold internals.
When it reaches the top, reflux from the condenser enters
the column and finally all variables reach the steady-state
values. During this process the hydraulic and thermodynamic
variables underlie large changes. The hold-up in the column
and the flow rates change from zero to their steady state
values. The temperatures rise from ambient temperature to
the boiling temperature. Due to these transitions the startup
cannot be described by a standard dynamic equilibrium
stage model which is only valid for conditions close to the
operating point. Different sets of equations are therefore
needed for the different distinguishable phases of the startup,
requiring a switching between these model equations at cer-
tain points for every single stage j separately. This procedure
eventually leads to an equilibrium stage model for the operat-
ing range consisting of the MESH equations, reaction kinetics
and hydraulic correlations [pressure drop and liquid hold-up
correlations from Engel et al. (2001) for the stages, Toricelli
type correlations for the liquid hold-up in reboiler and decan-
ter]. During the startup simulation, switching concerning
liquid and vapour flow rates assures that the hydraulic corre-
lations are only considered if a certain amount of liquid has
accumulated on a stage (liquid hold-up correlation) or a cer-
tain pressure drop has built up (pressure drop correlation).
The corresponding flow rates are otherwise set to zero. At
low temperatures liquid and vapour are not in equilibrium,
and the pressure is therefore first set to a constant initial
value, the temperature is calculated independently from the
VLE and the vapour phase is not considered. When during
the heating-up the temperature calculated by the startup
model reaches the additionally calculated bubble point temp-
erature at this initial pressure and the current liquid compo-
sition, then the vapour phase is included in the model and
temperature and pressure correspond to the VLE [switching
according to equation (4)]. In the decanter model the possible
appearance of a second liquid phase is considered during the
whole startup by continuously checking the liquid composition.
If Tj ! TLV
j then
Tj ¼ TLV
j
Vsection ¼ HUL
j vL
þ HUV
j vV
Else
Tj = TLV
j
pj ¼ const:
HUV
j ¼ 0
(4)
The developed simulation model has been validated with
startup data from the presented experiment. Figure 5 shows
the comparison between the concentrations in the reboiler
and the organic phase in the decanter during startup. The
startup begins with the start of the feeding and the reboiler
duty ramp (t ¼ 0). Since reboiler and decanter are initially
empty, the first liquid samples are taken after 1–1.5 hs. The
simulation describes the dynamic behaviour very well, includ-
ing the instant formation of a miscibility gap when the first sub-
cooled condensate enters the decanter.
In the following the startup process with different initial
charging of the decanter is analysed using the validated
model. During all the simulation runs the reboiler is initially
empty and filled by the feeds entering the column. When
the level is high enough the reboiler duty is increased step-
wise to the steady-state value. In case of an initially charged
Figure 4. Liquid composition profile along theoretical stages. HOAc
(5), PrOH (A), PrOAc (4), H2O (W).
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 113
decanter, the reflux is turned on as soon as the first vapour
condenses, otherwise (case 0) the decanter is first filled
with the condensate. As in the experiments the organic
phase is completely returned to the column as the reflux,
the aqueous phase is withdrawn from the system. Specifica-
tions are taken from the experimental setup.
Figure 6 shows the comparison of the trends of the MX
function (calculated for the organic phase in the decanter)
which gives the deviation of the current from the steady
state composition for the different startup strategies. The
strong influence of the composition of the initial charging
can be seen. As expected, initial charging of the two-phase
top product leads to the fastest startup (case 3) and reduces
startup time compared to an initially empty decanter (case 0).
The composition of the organic phase deviates only very little
and only for a short period of time from the steady-state
values. All other initial charging results in a slower startup.
The prolonging of startup time due to charging of the two
phase product to the decanter as described by Tran (2005)
could therefore not be observed for this process.
Conceptual Design
The conceptual design of a high-purity n-propyl acetate
production process is initiated on the basis of the method
developed by Tang et al. (2005) assuming that any acetic
acid esterification reactive distillation process includes both
rectifying and stripping sections: for n-propyl acetate
system, when most of acetic acid is consumed in the reactive
section, composition at the top of the column is close to the
minimum boiling ternary azeotrope located inside the two
liquid phases region (Figure 7); thus a decanter is placed to
remove the aqueous phase of the distillate whereas organic
phase is further processed in a stripping column to obtain
high-purity n-propyl acetate: Figure 8 regroups all studied
flowsheets for this esterification.
The process design has been developed using a sequen-
tial approach based on steady-state simulations. The feasible
values of different parameters leading to a PrOAc molar
purity of 98% for each configuration [except (e) for which
water content in product stream is remaining slightly
higher than 2%)] are related in Table 5. The qualitative
analysis of all possible flowsheets gives the following
conclusions:
. the optimal feed tray for the acetic acid (high boiling reac-
tant) is the column base unlike the usual double-feed con-
figuration (the low-boiling reactant is fed in the reboiler and
Figure 7. RCM of the PrOH/PrOAc/H2O non-reactive mixture show-
ing the miscibility gap, the distillation boundaries, the unstable (W)
and stable nodes (†) and the saddle points (A).
Figure 5. Dynamic composition trend in the reboiler (top) and of the
organic phase in the decanter (bottom) from experiment and simu-
lation. HOAc (5,–dash-dot line–), PrOH (A, dashed line), PrOAc
(D, solid line), H2O (o, dotted line).
Figure 6. Comparison of the MXtop-function in the organic phase in
the decanter during startup for different simulation runs. (0): initially
empty decanter, (1): initial charging of n-propanol, (2): initial charging
of n-propyl acetate, (3): initial charging with two phase mixture
(steady state composition), (4): initial charging of acetic acid.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
114 BREHELIN et al.
Figure 8. Six possible flowsheets for the n-propyl acetate production by catalytic reactive distillation (single/multiple feeds). (a)  (b) Partial
withdrawal of the decanter organic phase towards an additional stripping column; (c)  (d) total withdrawal of the decanter organic phase
towards an additional stripping column; (e)  (f) total reflux of the decanter organic phase in a single reactive distillation column with an inter-
mediate PrOAc withdrawal.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 115
high-boiling above the reactive section); this can be
explained since the reboiler holdup is larger than the
column holdup;
. the bottom products may not always be neglected since
depending on PrOAc stream specifications (purity and
recovery rate), unreacted HOAc has to be withdrawn;
. a configuration with total withdrawal of the decanter
organic phase [(c) and (d)] is more suitable to produce
high-purity PrOAc; the last column in Table 5(c’) shows
that a quasi-pure PrOAc stream can be produced when
the stripper NTS and the reboiler duties are increased;
. the single-column configuration with an intermediate
PrOAc withdrawal (e) is powerful in terms of conversion
and recovery, but seems not to be adapted to pure
PrOAc production: water is still remaining in the product
stream (2%) which may not be suitable. Moreover, an
addition of 10 stages in the rectifying section results in
only a slight improvement of PrOAc purity (,0.1%).
Note that the number of theoretical stages in the rectifying
section, the reflux ratio of organic phase of the decanter and
the feed ratio are the main sensitive parameters. The rectify-
ing section is required to decrease the acetic acid amount in
the distillate (see liquid composition profiles Figure 9), but its
concentration in the bottom products depends on the reaction
conversion only: that’s why a small alcohol excess is set in
the feed. Some attempts of fractionating the alcohol feed
were carried out in order to increase reaction rate on the
upper stages of the reactive section, but no significant
effect was found.
CONCLUSIONS
This study is focused on the conceptual design of high-
purity n-propyl acetate production by reactive distillation.
The first part of this article is devoted to experimental results
obtained on two kinds of device: stirred reactor and RD pilot
plant. A kinetic model for both homogeneously and heteroge-
neously catalysed reaction has been developed and chemi-
cal equilibrium data have been successfully compared with
previous publications. The second part deals with the
Table 5. Feasible setups of the three single-feed configurations.
(a) (c) (e) (c’)
Design variables
NTS reactive section 15 15 16 15
NTS rectifying section 15 15 10 þ 10 15
NTS stripping section 15 10 — 20
Operating variables
FHOAc (mol h21
) 83.26 83.26 83.26 83.26
FPrOH (mol h21
) 84.03 84.03 84.03 84.03
Decanter temperature (8C) 20 20 20 20
Decanter reflux ratiob
0.923 0 1 0
RD reboiler duty (kW) 4.468 5.025 6.220 6.205
Stripper reboiler duty (kW) 1.819 1.115 — 2.000
Stripper reflux ratio 4.985 1.000 — 1.000
Products
PrOAc flow rate (mol h21
) 82.82 83.22 84.61 82.83
xHOAc 0.0018 0.0003 0.0002 0.0005
xPrOH 0.0182 0.0197 0.0023 0.0005
xPrOAc 0.9800 0.9800 0.9770 0.9990
xH2O 0.0000 0.0000 0.0202 0.0000
H2O flow rate (mol h21
) 82.24 82.11 82.33 84.22
xHOAc 0.0004 0.0001 0.0000 0.0000
xPrOH 0.0055 0.0023 0.0099 0.0109
xPrOAc 0.0029 0.0015 0.0031 0.0032
xH2O 0.9912 0.9961 0.9870 0.9859
Bottoms flow rate (mol h21
) 2.22 1.97 0.35 0.24
xHOAc 0.6551 0.6937 0.7402 0.7397
xPrOH 0.1918 0.1702 0.1463 0.1466
xPrOAc 0.1084 0.0972 0.0816 0.0817
xH2O 0.0447 0.0389 0.0319 0.0320
Molar conversion (%) 98.0 98.3 99.7 99.7
Recovery rate (%) 99.4 99.6 99.7 99.7
Performance indexb
(%) 97.4 98.0 99.3 99.4
a
Ratio of the organic stream refluxed in the RD column on the withdrawn stream sent to the stripper.
b
Ratio of molar PrOAc production on molar HOAc feed.
Figure 9. Liquid phase composition profiles and temperature along
reactive distillation for type (c) process (PrOAc purity of 99.9%).
HOAc (5), PrOH (A), PrOAc (4), H2O (W) and temperature (solid
line).
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
116 BREHELIN et al.
simulation of the reactive distillation pilot plant. Model vali-
dation has been carried out with steady-state and dynamic
data and an optimal startup strategy has been recommended
using dynamic simulations. On this basis, six different setups
have been identified as being able to product high-purity n-
propyl acetate; a single-feed configuration with a decanter
at the top of the column and the further purification of all
the decanter organic withdrawal has been noticed to be the
most suitable design to achieve such a production.
In the near future, an optimal design of a catalytic distilla-
tion column will be achieved based on our model and using
MINLP approach; finally, the performances of this new
design will be experimentally checked in an heterogeneous
catalysed column located in TU Berlin.
NOMENCLATURE
a liquid phase activity
Dn molar transformation, mol
E energy of activation, J mol21
F flow rate, mol h21
K equilibrium constant
Ko preexponential factor
HU hold-up, mol
n molar amount, mol
p pressure, bar
Q reboiler duty, kW
r rate of reaction, mol s21
T temperature, K
v molar volume, m3
/ mol
V volume, m3
x liquid phase composition
Subscripts and superscripts
a activity
f forward (esterification)
i component index
j stage index
L liquid
LV boiling point
o reference or initial state
V vapour
REFERENCES
Bart, H.J., Kaltenbrunner, W. and Landschutzer, H., 1996, Kinetics of
esterification of acetic acid with propyl alcohol by heterogeneous
catalysis, Int J Chem Kinetics, 28: 649–656.
Brehelin, M., Rouzineau, D., Thery, R., Meyer, X. and Meyer, M.,
2006, Calcul des courbes de re´sidu re´actif he´te´roge`ne de la
reaction de synthe`se de l’ace´tate de n-propyle, Proceedings of
SIMO’06.
Chiang, S.F., Kuo, C.L., Yu, C.C. and Wong, D.S.H., 2002, Design
alternatives for the amyl acetate process—coupled reactor column
and reactive distillation, Ind Eng Chem Res, 41: 3233–3246.
Engel, V., Stichlmair, J. and Geipel, W., 2001, Fluid dynamics of pack-
ings for gas-liquid contactors, Chem Eng Technol, 24(5): 459–462.
Fillon, M., Meyer, M., Pingaud, H. and Joulia X., 1995, Data reconci-
liation based on elemental balances applied to batch experiments,
Comput Chem Eng, 19S: 293–299.
Huang, Y.S., Sundmacher, K., Tulashie, S. and Schlu¨nder, E.U., 2005,
Theoretical and experimental study on residue curve maps of
propyl acetate synthesis reaction, Chem Eng Sci, 60: 3363–3371.
Marek, J.and Standart, G., 1954, Vapor–liquid equilibria in mixtures
containing an associating substance. II. Binary mixtures of acetic
acid at athmospheric pressure, Collect Czech Chem Comm, 19:
1074–1084.
Okasinski, M.J. and Doherty, M.F., 2000, Prediction of hetero-
genenous reactive azeotropes in esterification systems, Chem
Eng Sci, 55: 5263–5271.
Pitochelli, A.R., 1980, Ion exchange catalysis and matrix effects
(Rhom and Haas Co., Philadelphia, PA).
Reepmeyer, F., Repke, J.-U. and Wozny, G., 2004, Time optimal
start-up strategies for reactive distillation columns, Chem Eng
Sci, 59(20): 4339–4347.
Sørensen, J.M. and Arlt, W., 1982, Liquid-Liquid Equilibrium Data
Collection. Chemistry Data Series, (Dechema).
Steinigeweg, S., 2003, Zur Entwicklung von Reaktivrektifikationspro-
zessen am Beispiel gleichgewichtslimitierter Reaktionen (Shaker
Verlag Aachen Germany).
Steinigeweg, S. and Gmehling, J., 2002, n-Butyl acetate synthesis
via reactive distillation thermodynamic aspects, reaction kinetics,
pilot-plant experiments, and simulation studies, Ind Eng Chem
Res 41:, 5483–5490.
Tang, Y.T., Chen, Y.W., Huang, H.P., Yu, C.C., Hung, S.B. and Lee,
M.J., 2005, Design of reactive distillation for acetic acid esterifica-
tion, AIChE J, 51(6): 1683–1699.
Tran, T.K., 2005, Analyse des Anfahrens von Dreiphasenkolonnen
(Shaker Verlag, Aachen, Germany).
Ung, S. and Doherty, M.F., 1995, Vapor–liquid phase equilibrium in
systems with multiple chemical reactions, Chem Eng Sci, 50:
23–48.
Weisz, P.B. and Hicks, J.S., 1962, The behaviour of porous catalyst
particles in view of internal mass and heat diffusion effects, Chem
Eng Sci, 17: 265–275.
The Manuscript was received 18 July 2006 and accepted for
publication after revision 24 October 2006.
Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117
PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 117

More Related Content

What's hot

Factors affecting rate of chemical reactions
Factors affecting rate of chemical reactionsFactors affecting rate of chemical reactions
Factors affecting rate of chemical reactions
Abigail Sapico
 
Intermediate Organic Special Project Poster
Intermediate Organic Special Project PosterIntermediate Organic Special Project Poster
Intermediate Organic Special Project Poster
Viren Lad
 
heat capacity of sitric acid0c96051e8eb63eea58000000
heat capacity of sitric acid0c96051e8eb63eea58000000heat capacity of sitric acid0c96051e8eb63eea58000000
heat capacity of sitric acid0c96051e8eb63eea58000000
Tika Ningsih
 
CONVERSION OF LIGNIN TO LIQUID COMPOUNDS
CONVERSION OF LIGNIN TO LIQUID COMPOUNDSCONVERSION OF LIGNIN TO LIQUID COMPOUNDS
CONVERSION OF LIGNIN TO LIQUID COMPOUNDS
Magdalena Kalabova
 
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejjeConstantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
AndreaCspedes8
 
Chemical reaction engineering
Chemical reaction engineeringChemical reaction engineering
Chemical reaction engineering
1langshen
 

What's hot (19)

Factors affecting rate of chemical reactions
Factors affecting rate of chemical reactionsFactors affecting rate of chemical reactions
Factors affecting rate of chemical reactions
 
Rate of reaction for limestone and citric acid.
Rate of reaction for limestone and citric acid.Rate of reaction for limestone and citric acid.
Rate of reaction for limestone and citric acid.
 
1.4 rate of reaction(1.2d)...biology
1.4 rate of reaction(1.2d)...biology1.4 rate of reaction(1.2d)...biology
1.4 rate of reaction(1.2d)...biology
 
Chapter 1
Chapter 1Chapter 1
Chapter 1
 
Che 412 cre 1 notes
Che 412 cre 1 notesChe 412 cre 1 notes
Che 412 cre 1 notes
 
Solvent Effect on the Enthalpy and Entropy of Activation for the Hydrolysis o...
Solvent Effect on the Enthalpy and Entropy of Activation for the Hydrolysis o...Solvent Effect on the Enthalpy and Entropy of Activation for the Hydrolysis o...
Solvent Effect on the Enthalpy and Entropy of Activation for the Hydrolysis o...
 
Rate and order of reaction
Rate and order of reaction Rate and order of reaction
Rate and order of reaction
 
Effect of Operating Conditions on CSTR performance: an Experimental Study
Effect of Operating Conditions on CSTR performance: an Experimental StudyEffect of Operating Conditions on CSTR performance: an Experimental Study
Effect of Operating Conditions on CSTR performance: an Experimental Study
 
Intermediate Organic Special Project Poster
Intermediate Organic Special Project PosterIntermediate Organic Special Project Poster
Intermediate Organic Special Project Poster
 
247447699 reactors
247447699 reactors247447699 reactors
247447699 reactors
 
slow reaction
slow reactionslow reaction
slow reaction
 
Bachelor's thesis article
Bachelor's thesis articleBachelor's thesis article
Bachelor's thesis article
 
heat capacity of sitric acid0c96051e8eb63eea58000000
heat capacity of sitric acid0c96051e8eb63eea58000000heat capacity of sitric acid0c96051e8eb63eea58000000
heat capacity of sitric acid0c96051e8eb63eea58000000
 
Rate of reaction for limestone citric acid reaction.
Rate of reaction for limestone  citric acid reaction.Rate of reaction for limestone  citric acid reaction.
Rate of reaction for limestone citric acid reaction.
 
CONVERSION OF LIGNIN TO LIQUID COMPOUNDS
CONVERSION OF LIGNIN TO LIQUID COMPOUNDSCONVERSION OF LIGNIN TO LIQUID COMPOUNDS
CONVERSION OF LIGNIN TO LIQUID COMPOUNDS
 
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejjeConstantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
Constantedeequilibrioqumicoensistemashomogneos 170110024707 jejejeeeejejejje
 
Chemical reaction engineering
Chemical reaction engineeringChemical reaction engineering
Chemical reaction engineering
 
Catalysis- Physical Chemistry
Catalysis-  Physical ChemistryCatalysis-  Physical Chemistry
Catalysis- Physical Chemistry
 
Quimica 4.c.2
Quimica 4.c.2Quimica 4.c.2
Quimica 4.c.2
 

Similar to Production of-n-propyl-acetate-by-reactive-distillation-experimental-and-theoretical-study 2007-chemical-engineering-research-and-design

Optimal Production of Methyl Acetate using a Micro-Reactor
Optimal Production of Methyl Acetate using a Micro-ReactorOptimal Production of Methyl Acetate using a Micro-Reactor
Optimal Production of Methyl Acetate using a Micro-Reactor
Sasha Kozmonaut
 
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docxRunning Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
healdkathaleen
 
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 KLiquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
Reddysuresh Kolavali
 
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
Josemar Pereira da Silva
 
Two experiments-for-the-introductory-chemical-reaction-engineering-course
Two experiments-for-the-introductory-chemical-reaction-engineering-courseTwo experiments-for-the-introductory-chemical-reaction-engineering-course
Two experiments-for-the-introductory-chemical-reaction-engineering-course
Sachin Patil
 
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINALAn Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
Stroma Service Consulting
 
Activity coefficients at infinite dilution for organic solutes dissolved in t...
Activity coefficients at infinite dilution for organic solutes dissolved in t...Activity coefficients at infinite dilution for organic solutes dissolved in t...
Activity coefficients at infinite dilution for organic solutes dissolved in t...
Bihan Jiang
 
2014_Belkheiri et al._Cellulose Chemistry and Technology
2014_Belkheiri et al._Cellulose Chemistry and Technology2014_Belkheiri et al._Cellulose Chemistry and Technology
2014_Belkheiri et al._Cellulose Chemistry and Technology
Huyen Lyckeskog
 

Similar to Production of-n-propyl-acetate-by-reactive-distillation-experimental-and-theoretical-study 2007-chemical-engineering-research-and-design (20)

1_s2.0_S0255270118302393_main.pdf
1_s2.0_S0255270118302393_main.pdf1_s2.0_S0255270118302393_main.pdf
1_s2.0_S0255270118302393_main.pdf
 
B040608 806
B040608 806B040608 806
B040608 806
 
Optimal Production of Methyl Acetate using a Micro-Reactor
Optimal Production of Methyl Acetate using a Micro-ReactorOptimal Production of Methyl Acetate using a Micro-Reactor
Optimal Production of Methyl Acetate using a Micro-Reactor
 
G04103948
G04103948G04103948
G04103948
 
CONVERSION OF DIMETHYL-NITROBENZENE TO DIMETHY L ANILINE, EFFECT OF SOME PROC...
CONVERSION OF DIMETHYL-NITROBENZENE TO DIMETHY L ANILINE, EFFECT OF SOME PROC...CONVERSION OF DIMETHYL-NITROBENZENE TO DIMETHY L ANILINE, EFFECT OF SOME PROC...
CONVERSION OF DIMETHYL-NITROBENZENE TO DIMETHY L ANILINE, EFFECT OF SOME PROC...
 
Dynamic
DynamicDynamic
Dynamic
 
Dynamic
DynamicDynamic
Dynamic
 
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docxRunning Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
Running Head AZEOTROPIC EQUILIBRIUM AZEOTROPIC EQUILIBRIUM.docx
 
Development of Dynamic Models for a Reactive Packed Distillation Column
Development of Dynamic Models for a Reactive Packed Distillation ColumnDevelopment of Dynamic Models for a Reactive Packed Distillation Column
Development of Dynamic Models for a Reactive Packed Distillation Column
 
Simulation and Optimization of Cyclohexanone Ammoximation Process Over TS-1 C...
Simulation and Optimization of Cyclohexanone Ammoximation Process Over TS-1 C...Simulation and Optimization of Cyclohexanone Ammoximation Process Over TS-1 C...
Simulation and Optimization of Cyclohexanone Ammoximation Process Over TS-1 C...
 
Liquid liquid equilibrium data for n hexane ethylacetate acetonitrile ternay ...
Liquid liquid equilibrium data for n hexane ethylacetate acetonitrile ternay ...Liquid liquid equilibrium data for n hexane ethylacetate acetonitrile ternay ...
Liquid liquid equilibrium data for n hexane ethylacetate acetonitrile ternay ...
 
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 KLiquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
Liquid-Liquid Equilibria of Nitrobenzene-Inorganic Acid Systems at 298.15 K
 
An experimental method for micro particle ammonium nitrate (mp-an)
An experimental method for micro   particle ammonium nitrate (mp-an)An experimental method for micro   particle ammonium nitrate (mp-an)
An experimental method for micro particle ammonium nitrate (mp-an)
 
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
Isobaric vapor liquid equilibrium for binary mixtures of 3 methyl 1 butanol 3...
 
Two experiments-for-the-introductory-chemical-reaction-engineering-course
Two experiments-for-the-introductory-chemical-reaction-engineering-courseTwo experiments-for-the-introductory-chemical-reaction-engineering-course
Two experiments-for-the-introductory-chemical-reaction-engineering-course
 
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINALAn Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
An Experimental Method for Micro-Particle Ammonium Nitrate (MP-AN)_FINAL
 
Activity coefficients at infinite dilution for organic solutes dissolved in t...
Activity coefficients at infinite dilution for organic solutes dissolved in t...Activity coefficients at infinite dilution for organic solutes dissolved in t...
Activity coefficients at infinite dilution for organic solutes dissolved in t...
 
Enhancement of heat transfer rate using MgO nanofluid in heat exchanger
Enhancement of heat transfer rate using MgO nanofluid in heat exchangerEnhancement of heat transfer rate using MgO nanofluid in heat exchanger
Enhancement of heat transfer rate using MgO nanofluid in heat exchanger
 
2014_Belkheiri et al._Cellulose Chemistry and Technology
2014_Belkheiri et al._Cellulose Chemistry and Technology2014_Belkheiri et al._Cellulose Chemistry and Technology
2014_Belkheiri et al._Cellulose Chemistry and Technology
 
Martin1981
Martin1981Martin1981
Martin1981
 

More from Josemar Pereira da Silva

Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
Investigation on thermodynamics in separation for ethylene glycol neopentyl g...Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
Josemar Pereira da Silva
 
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
Josemar Pereira da Silva
 

More from Josemar Pereira da Silva (20)

Fmincon
FminconFmincon
Fmincon
 
Critical data
Critical dataCritical data
Critical data
 
Influence of temperature on the liquid liquid equilibria of methanol benzene ...
Influence of temperature on the liquid liquid equilibria of methanol benzene ...Influence of temperature on the liquid liquid equilibria of methanol benzene ...
Influence of temperature on the liquid liquid equilibria of methanol benzene ...
 
(Liquid liquid) equilibrium of systems involved in the stepwise ethanolysis o...
(Liquid liquid) equilibrium of systems involved in the stepwise ethanolysis o...(Liquid liquid) equilibrium of systems involved in the stepwise ethanolysis o...
(Liquid liquid) equilibrium of systems involved in the stepwise ethanolysis o...
 
Liquid liquid equilibria data for ethylbenzene or p xylene with alkane and 1 ...
Liquid liquid equilibria data for ethylbenzene or p xylene with alkane and 1 ...Liquid liquid equilibria data for ethylbenzene or p xylene with alkane and 1 ...
Liquid liquid equilibria data for ethylbenzene or p xylene with alkane and 1 ...
 
Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
Investigation on thermodynamics in separation for ethylene glycol neopentyl g...Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
Investigation on thermodynamics in separation for ethylene glycol neopentyl g...
 
A correlation for the prediction of thermal conductivity of liquids
A correlation for the prediction of thermal conductivity of liquidsA correlation for the prediction of thermal conductivity of liquids
A correlation for the prediction of thermal conductivity of liquids
 
A new model of thermal conductivity for liquids
A new model of thermal conductivity for liquidsA new model of thermal conductivity for liquids
A new model of thermal conductivity for liquids
 
Perda de carga escoamento bifásico 1
Perda de carga escoamento bifásico 1Perda de carga escoamento bifásico 1
Perda de carga escoamento bifásico 1
 
Tibirica c bhebulição
Tibirica c bhebuliçãoTibirica c bhebulição
Tibirica c bhebulição
 
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
On criteria for occurence of azeotropes in isothermal and isobraric binary sy...
 
Heteroazeotropic batch distillatioin feasibility and operation
Heteroazeotropic batch distillatioin feasibility and operationHeteroazeotropic batch distillatioin feasibility and operation
Heteroazeotropic batch distillatioin feasibility and operation
 
A new model of thermal conductivity for liquids
A new model of thermal conductivity for liquidsA new model of thermal conductivity for liquids
A new model of thermal conductivity for liquids
 
Nicolau maquiavel o príncipe
Nicolau maquiavel   o príncipeNicolau maquiavel   o príncipe
Nicolau maquiavel o príncipe
 
On the dynamics of distillation processes
On the dynamics of distillation processesOn the dynamics of distillation processes
On the dynamics of distillation processes
 
1 s2.0-0378381289800755fazeremcasa-main
1 s2.0-0378381289800755fazeremcasa-main1 s2.0-0378381289800755fazeremcasa-main
1 s2.0-0378381289800755fazeremcasa-main
 
Fluxograma de construção de mapas de curvas residuais
Fluxograma de construção de mapas de curvas residuaisFluxograma de construção de mapas de curvas residuais
Fluxograma de construção de mapas de curvas residuais
 
Mspf
MspfMspf
Mspf
 
68183043 apostila-simulacao-e-otimizacao-de-processos-quimicos
68183043 apostila-simulacao-e-otimizacao-de-processos-quimicos68183043 apostila-simulacao-e-otimizacao-de-processos-quimicos
68183043 apostila-simulacao-e-otimizacao-de-processos-quimicos
 
Sistemas equacoes lineares
Sistemas equacoes linearesSistemas equacoes lineares
Sistemas equacoes lineares
 

Recently uploaded

DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdfDR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
DrGurudutt
 
Digital Signal Processing Lecture notes n.pdf
Digital Signal Processing Lecture notes n.pdfDigital Signal Processing Lecture notes n.pdf
Digital Signal Processing Lecture notes n.pdf
AbrahamGadissa
 

Recently uploaded (20)

DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdfDR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
DR PROF ING GURUDUTT SAHNI WIKIPEDIA.pdf
 
Furniture showroom management system project.pdf
Furniture showroom management system project.pdfFurniture showroom management system project.pdf
Furniture showroom management system project.pdf
 
Software Engineering - Modelling Concepts + Class Modelling + Building the An...
Software Engineering - Modelling Concepts + Class Modelling + Building the An...Software Engineering - Modelling Concepts + Class Modelling + Building the An...
Software Engineering - Modelling Concepts + Class Modelling + Building the An...
 
Introduction to Machine Learning Unit-5 Notes for II-II Mechanical Engineering
Introduction to Machine Learning Unit-5 Notes for II-II Mechanical EngineeringIntroduction to Machine Learning Unit-5 Notes for II-II Mechanical Engineering
Introduction to Machine Learning Unit-5 Notes for II-II Mechanical Engineering
 
A CASE STUDY ON ONLINE TICKET BOOKING SYSTEM PROJECT.pdf
A CASE STUDY ON ONLINE TICKET BOOKING SYSTEM PROJECT.pdfA CASE STUDY ON ONLINE TICKET BOOKING SYSTEM PROJECT.pdf
A CASE STUDY ON ONLINE TICKET BOOKING SYSTEM PROJECT.pdf
 
The battle for RAG, explore the pros and cons of using KnowledgeGraphs and Ve...
The battle for RAG, explore the pros and cons of using KnowledgeGraphs and Ve...The battle for RAG, explore the pros and cons of using KnowledgeGraphs and Ve...
The battle for RAG, explore the pros and cons of using KnowledgeGraphs and Ve...
 
Quality defects in TMT Bars, Possible causes and Potential Solutions.
Quality defects in TMT Bars, Possible causes and Potential Solutions.Quality defects in TMT Bars, Possible causes and Potential Solutions.
Quality defects in TMT Bars, Possible causes and Potential Solutions.
 
Natalia Rutkowska - BIM School Course in Kraków
Natalia Rutkowska - BIM School Course in KrakówNatalia Rutkowska - BIM School Course in Kraków
Natalia Rutkowska - BIM School Course in Kraków
 
Introduction to Machine Learning Unit-4 Notes for II-II Mechanical Engineering
Introduction to Machine Learning Unit-4 Notes for II-II Mechanical EngineeringIntroduction to Machine Learning Unit-4 Notes for II-II Mechanical Engineering
Introduction to Machine Learning Unit-4 Notes for II-II Mechanical Engineering
 
1. Henrich Triangle Safety and Fire Presentation
1. Henrich Triangle Safety and Fire Presentation1. Henrich Triangle Safety and Fire Presentation
1. Henrich Triangle Safety and Fire Presentation
 
ONLINE CAR SERVICING SYSTEM PROJECT REPORT.pdf
ONLINE CAR SERVICING SYSTEM PROJECT REPORT.pdfONLINE CAR SERVICING SYSTEM PROJECT REPORT.pdf
ONLINE CAR SERVICING SYSTEM PROJECT REPORT.pdf
 
Lect 2 - Design of slender column-2.pptx
Lect 2 - Design of slender column-2.pptxLect 2 - Design of slender column-2.pptx
Lect 2 - Design of slender column-2.pptx
 
Digital Signal Processing Lecture notes n.pdf
Digital Signal Processing Lecture notes n.pdfDigital Signal Processing Lecture notes n.pdf
Digital Signal Processing Lecture notes n.pdf
 
Electrical shop management system project report.pdf
Electrical shop management system project report.pdfElectrical shop management system project report.pdf
Electrical shop management system project report.pdf
 
KIT-601 Lecture Notes-UNIT-4.pdf Frequent Itemsets and Clustering
KIT-601 Lecture Notes-UNIT-4.pdf Frequent Itemsets and ClusteringKIT-601 Lecture Notes-UNIT-4.pdf Frequent Itemsets and Clustering
KIT-601 Lecture Notes-UNIT-4.pdf Frequent Itemsets and Clustering
 
Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2Research Methodolgy & Intellectual Property Rights Series 2
Research Methodolgy & Intellectual Property Rights Series 2
 
"United Nations Park" Site Visit Report.
"United Nations Park" Site  Visit Report."United Nations Park" Site  Visit Report.
"United Nations Park" Site Visit Report.
 
internship exam ppt.pptx on embedded system and IOT
internship exam ppt.pptx on embedded system and IOTinternship exam ppt.pptx on embedded system and IOT
internship exam ppt.pptx on embedded system and IOT
 
KIT-601 Lecture Notes-UNIT-3.pdf Mining Data Stream
KIT-601 Lecture Notes-UNIT-3.pdf Mining Data StreamKIT-601 Lecture Notes-UNIT-3.pdf Mining Data Stream
KIT-601 Lecture Notes-UNIT-3.pdf Mining Data Stream
 
Top 13 Famous Civil Engineering Scientist
Top 13 Famous Civil Engineering ScientistTop 13 Famous Civil Engineering Scientist
Top 13 Famous Civil Engineering Scientist
 

Production of-n-propyl-acetate-by-reactive-distillation-experimental-and-theoretical-study 2007-chemical-engineering-research-and-design

  • 1. PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION Experimental and Theoretical Study M. Brehelin1 , F. Forner2 , D. Rouzineau1 , J.-U. Repke2 , X. Meyer1 , M. Meyer1Ã and G. Wozny2 1 Laboratoire de Ge´nie Chimique, INP ENSIACET, Toulouse, France. 2 Institute of Process and Plant Technology, Technische Universita¨t Berlin, Germany. Abstract: First steps of the development of a catalytic reactive distillation process for the pro- duction of n-propyl acetate based on experiments and simulations are proposed. The kinetics for homogeneously (sulphuric acid) and heterogeneously (Amberlyst 15) catalysed reaction were investigated and the constants for a pseudo-homogeneous model are presented. Pilot plant experiments were performed using a homogeneous strong acid catalyst in a packed column. A top-column decanter is used to withdraw the aqueous phase and to reflux the organic phase. Simulation results are in good agreement with experimental data. Thermodynamics non- idealities are taken into account using VLE and LLE NRTL interaction parameters. Alcohol conver- sion and n-propyl acetate purity may be dramatically increased just by adding to the pilot plant a stripping section in an additional column: six different configurations are identified to achieve such a production. The startup is studied in order to determine the best strategy to achieve steady-state conditions. The strong influence of the composition of the initial charging in the decanter can be seen and an initial charging of the two-phase top product leads to the fastest startup. Keywords: reactive distillation; kinetics; startup; experimental data; n-propyl acetate. INTRODUCTION The combination of reaction and distillation in one reactive distillation (RD) unit can lead to significant reduction of investment and oper- ational cost. However, interactions between reactions and separation increase complexity and require a better understanding of the pro- cess. Whereas RD is intensively studied for years, only few works have been published on vapour–liquid–liquid (VLL) systems, thus our investigations focus on VLL reactive systems. Experimental investigations have already been published for the production of butyl acetate (Steinigeweg and Gmehling, 2002) or amyl acetate (Chiang et al., 2002) but, to the best of our knowledge, none for n-propyl acetate reactive distillation; thus, information about n-propyl acetate synthesis can hardly be found in literature. Investigations on reac- tive membrane separation have led Huang et al. (2005) to publish data on chemical equi- librium. Bart et al. (1996) have published a kinetic modelling of the n-propyl acetate syn- thesis catalysed by an ion exchange resin (Dowex) as well as Steinigeweg (2003) with the Amberlyst-15. Unfortunately, uncertainty remains on the chemical equilibrium constant value. Finally, no data about reactive distilla- tion pilot-plant experiments is published. However, reactive distillation column design requires reliable kinetic data in addition to other crucial information such as phase equilibria. Methods carried out to assess the reactive distillation process feasi- bility are often based on residue curve maps (RCM) for non-reactive distillation, and on reactive residue curve maps (rRCM) for reac- tive distillation. In addition to the four pure components, three azeotropes are the singu- lar points of the n-propyl acetate synthesis reaction rRCM (Table 1). Two of these azeotropes (PrOH/PrOAc and PrOH/H2O) are remaining from the non-reactive RCM although two others involving reactive species (PrOAc/H2O and PrOH/PrOAc/H2O which are both heterogeneous) disappear; the last singular point of this rRCM is a quaternary reactive azeotrope. The chemical equilibrium constant value has a great influence on the liquid phase stability of this point (but not on its topological stability since it is always an unstable node): when a value of 20 (Huang et al., 2005) is used, the reactive azeotrope is homogeneous, although a value of 48.8 (Okasinski and Doherty, 2000) leads to a heterogeneous reactive azeotrope. 109 Vol 85 (A1) 109–117 Ã Correspondence to: Dr M. Meyer, Laboratoire de Ge´nie Chimique, INP ENSIACET, Toulouse, France. E-mail: michel.meyer@ ensiacet.fr DOI: 10.1205/cherd06112 0263–8762/07/ $30.00 þ 0.00 Chemical Engineering Research and Design Trans IChemE, Part A, January 2007 # 2007 Institution of Chemical Engineers
  • 2. These effects are clearly illustrated in Figure 1 for which rRCM have been computed using a previous work on the rRCM calculation taking into account rigorous VLLE equilibria (Brehelin et al., 2006); the transformed variable coordinates (Ung and Doherty, 1995) are used with n-propyl acetate as the reference component. For the sake of better represen- tation of the chemical and phase equilibria complexity, rRCM in molar composition are represented too in Figure 1 (bottom). Table 1. Singular points of n-propyl acetate synthesis reaction rRCM. System T (8C) Stability Molar compositionTopological Physical HOAc 117.9 SN PrOAc 101.5 Sa H2O 100.0 Sa PrOH 97.2 SN PrOH/PrOAc 94.7 Sa Homogeneous 0.625/0.375 PrOH/H2O 87.8 Sa Homogeneous 0.432/0.568 HOAc/PrOH/PrOAc/H2O 85.4 UN Homogeneous 0.069/0.309/0.146/0.476 84.0 Heterogeneous 0.059/0.245/0.262/0.435 Figure 1. n-Propyl acetate synthesis reaction rRCM in transformed (top) and molar (bottom) compositions showing the miscibility gap for two values of the chemical equilibrium. Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 110 BREHELIN et al.
  • 3. As a consequence, great attention has to be paid to the chemical and phase equilibria determination in order to perform the design of such a process. Therefore, this article presents the first steps of an approach to the development of a RD process for the production of n-propyl acetate: the catalysed reaction kinetics are firstly investigated in order to employ a simulation environment based on an equilibrium stage model; in parallel pilot plant experiments are carried out in order to validate this simulation tool. On the basis of these results, an optimal startup strategy is proposed and several process configurations are studied in order to propose the best setup for high purity n-propyl acetate production. EXPERIMENTAL SECTION Materials and Analysis The chemicals used for the kinetic study were of analytical grade (rectapur); Amberlyst 15 and concentrated sulphuric acid were selected as catalysts. For the reactive distillation experiments, the chemicals were of reaction grade (99.5%) and sulphuric acid was the catalyst. All samples were analy- sed by gas chromatography (Varian 3800) using cross-linked polyethylene glycol CP-WAX 52 CB 30 m  0.32 mm column with FID detector. For the homogeneously catalyzed kinetic runs, the acetic acid concentration has been determined using titration; additional titrations were performed with a METTLER DL 35 Karl–Fischer-Titrator to measure the quan- tity of water formed during the reaction. A data reconciliation process was led in order to appre- ciably increase reliability and consistency of results: the elimination of aberrant measurements is performed on the base of the atomic conservation (Fillon et al., 1995). An accu- racy of molar fractions of 3% can thus be expected. Kinetic Experiments Kinetic experiments were conducted in a thermostated three-neck glass reactor with a volume of 500 cm3 . One neck was connected to a thermometer and another was devoted to the GC sampling; a cold trap, in addition to a reflux condenser, was installed in the third neck to avoid any loss of volatile components. The temperature of the heat- ing jacket was kept constant within +0.1 K. Before each kinetic experiment was started, both reactants were brought to reaction temperature separately; when the desired temp- erature was reached, reactants and catalyst were added to the reactor and time measurement was started. Liquid samples of about 1 cm3 volume, without any catalyst in case of Amberlyst 15, were taken using a syringe and cooled to 253 K to avoid any further reaction. The contents were stirred by a magnetic stirrer. The stirrer speed is fixed in such a manner that it has no influence on the reaction rate (Steinigeweg and Gmehling, 2002); thus, the external mass transfer resistance seems not to be significant (Pitochelli, 1980) and the internal mass transfer resistance was neglected on the basis of a Weisz and Hicks criterion calculation (Weisz and Hicks, 1962). Reactive Distillation Experiments Setup Experiments were performed in a 4-m of glass column with an inner diameter of 80 mm (Figure 2): the reactive section is packed with 8 mm  8 mm glass Raschig rings and the recti- fying section contained Sulzer-CY packing; the liquid phase total hold-up estimated by the Engel’s method (Engel et al., 2001) is of 80 ml. Each reactant is fed separately in the column. Reaction was catalysed by homogeneous sulphuric acid fed at the top of the reactive section with a weight amount of 1.3% in the acetic acid flow. The reboiler, with a liquid hold-up of approximately 10 dm3 was heated by an oil-boiler and insulated by mineral wool whereas the column was equipped with two couples of heating mantles; the maxi- mal heating power measured was of 3 kW. Distillate was collected in a decanter operated at ambient temperature and all organic phase was refluxed at the top of the column. Feed and product flows were measured by balances. Temperatures were measured by Pt-100 thermometers installed at the bottom of each section, as well as in the feed stream, the reboiler and the condenser. Startup The startup strategy was to start with empty reboiler and decanter; the preheated feeds and a gradual heat duty at the reboiler led the mixture to boiling point. When a sufficient amount of distillate was collected to the decanter the reflux pump was turned-on and the set-point was adjusted in order to reflux all the organic phase. Results Although this pilot plant has not been built especially for the n-propyl acetate production by reactive distillation—its design is far from optimal since an acetic acid conversion of 79% with a molar purity of n-propyl acetate of 64% has been achieved—dynamic and steady-state simulation can be per- formed on the base of these experimental results. Thus the aim of these runs was to obtain experimental data useful to validate our simulations. The material balances of a Figure 2. Conceptual schema of the pilot-plant used to perform reactive distillation experiments. Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 111
  • 4. typical reactive distillation run summarized in Table 2 show the reliability of a typical steady-state. RESULTS AND ANALYSES Kinetic Parameter Identification A pseudo-homogeneous model [equation (1)] is used to describe the reaction catalysed both with Amberlyst 15 or sul- phuric acid. This choice is theoretically supported by the fact that the heterogeneous catalyst swells in contact with polar solvents making the sulfonic groups of the catalyst readily accessible to reactants; in such a model, the ion-exchange resin is thought to act as solvated protons; consequently, rate constants can be expressed as a function of proton amount; its temperature dependency is expressed by Arrhenius’ law [equation (2)], so that four adjustable para- meters (Kf,o, Ka,o, Ef, Ea) have to be fitted (Table 3). r ¼ 1 ni dni dt ¼ nHþ Kf aHOAcaPrOH À aPrAcaH20 Ka (1) Ki ¼ Ki,0: exp À Ei R:T (2) The molar fractions of acetic acid, computed by the numeri- cal integration of the kinetic equation, are compared with the experimental ones. The activity coefficients were calculated using the NRTL equation of which interaction parameters are given in Table 4(a). Parameters are fitted by minimization of the criterion defined by equation (3) using a least square method. The mean value of this criterion, i.e., the mean rela- tive deviation—assumed to be representing the accuracy of the model—is inferior to 5%. crit ¼ min X all samples (xcalc HOAc À xexp HOAc)2 (3) The theoretical dependence on temperature of the equili- brium constant is shown in Figure 3: the obvious agreement between our chemical equilibrium data and Huang’s ones [almost 20 on the range temperature (80–1108C)] reinforce the validity of our identification. Moreover experimental data are related to the standard enthalpy of reaction by the van’t Hoff equation: experimental and theoretical values of the standard energy of reaction are concordant (21.02 and 21.69 kJ mol21 ). Simulation Steady state Steady-state simulations based on an equilibrium model were carried out using the commercial software ProSim Plus; the activity coefficients are derived from NRTL model [Table 4(a)] and used for phase equilibriums as well as for the kinetic model. It will be shown later that distillate stream is free of acetic acid so specific liquid–liquid NRTL parameters are used to compute phase equilibria into the decanter [Table 4(b)]. Vapour phase association due to the presence of acetic acid in the system has been accounted for with the chemical theory (Marek and Standard, 1954). Table 2. Material balances of a typical reactive distillation run. HOAc Feed PrOH feed Aqueous phase Bottoms I–O (%) 46.63 41.00 38.57 53.50 4.9 F (mol h21 ) Molar composition Dn (mol) HOAc 1 0 0.018 0.173 236.7 PrOH 0 1 0.008 0.131 233.7 PrOAc 0 0 0.004 0.638 þ34.3 H2O 0 0 0.970 0.058 þ40.6 Table 3. Parameters of the kinetic model for sulphuric acid and Amberlyst 15 as catalysts. Ko (mol s21.molHþ 21 ) Ea (J mol21 ) Esterification (H2SO4) 7.699 Â 104 37,084 Esterification (A15) 2.576 Â 106 51,540 Chemical equilibrium 17.52 2366 Table 4. (a) NRTL parameters for the calculation of activity coeffi- cients and phase equilibria (Huang et al., 2005; Okasinski and Doherty, 2000). I J aij (cal mol21 ) bij(cal mol21 ) a HOAc PrOH 2 327.52 256.90 0.3044 HOAc PrOAc 2 410.39 1050.56 0.2970 HOAc H2O 2 342.20 1175.72 0.2952 PrOH PrOAc 2 340.02 111.74 0.3005 PrOH H2O 2 152.51 1866.34 0.3747 PrOAc H2O 2 667.45 3280.60 0.2564 Table 4. (b) Liquid–liquid NRTL parameters for the ternary PrOH/ PrOAc/H2O (Sørensen and Arlt, 1982). I J aij (cal mol21 ) bij(cal mol21 ) a PrOH PrOAc 286.17 467.23 0.2 PrOH H2O 21102.18 2949.20 0.2 PrOAc H2O 2602.82 3023.92 0.2 Figure 3. Natural logarithm of the equilibrium constant against reci- procal temperature. This study (B) Huang et al. (A), Bart et al. (4), Steinigeweg (...) and theoretical estimation based on enthalpy of for- mation (– – –). Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 112 BREHELIN et al.
  • 5. Figure 4 illustrates the good agreement observed between experimental data and simulation (mean deviation of 3.5% on molar composition). Startup analysis The startup of a reactive distillation column is a very com- plex, time and energy consuming process. The dynamic pro- cess behaviour becomes even more complex in case of a miscibility gap at the column top which is influenced by the reaction. Special attention has to be paid to the behaviour of the decanter during startup in order to rapidly reach con- ditions leading to the phase split. Reepmeyer et al. (2004) have analysed the startup of homogeneously catalysed reac- tive distillation in tray columns and have shown that initial charging of product to the column trays can lead to significant reduction of startup time. Tran (2005) have studied non-reac- tive three phase distillation in a tray column with a decanter. They have shown both by experiment and simulation that the startup time depends strongly on the composition of the initial hold-up in the decanter. Surprisingly, for the system ethanol/ water/cyclohexane, initial charging of ethanol (bottom pro- duct) to the decanter led to a reduction of about 60% of startup time. On the other hand charging with a two phase mixture close to steady state composition extended the startup time (Tran 2005). The startup process which is considered in this contribution begins when feed is entering the initially cold and empty column. The dry packing is wetted and the column bottom is filled up until the reboiler is switched on. Vapour starts rising in the column and partly condenses on the cold internals. When it reaches the top, reflux from the condenser enters the column and finally all variables reach the steady-state values. During this process the hydraulic and thermodynamic variables underlie large changes. The hold-up in the column and the flow rates change from zero to their steady state values. The temperatures rise from ambient temperature to the boiling temperature. Due to these transitions the startup cannot be described by a standard dynamic equilibrium stage model which is only valid for conditions close to the operating point. Different sets of equations are therefore needed for the different distinguishable phases of the startup, requiring a switching between these model equations at cer- tain points for every single stage j separately. This procedure eventually leads to an equilibrium stage model for the operat- ing range consisting of the MESH equations, reaction kinetics and hydraulic correlations [pressure drop and liquid hold-up correlations from Engel et al. (2001) for the stages, Toricelli type correlations for the liquid hold-up in reboiler and decan- ter]. During the startup simulation, switching concerning liquid and vapour flow rates assures that the hydraulic corre- lations are only considered if a certain amount of liquid has accumulated on a stage (liquid hold-up correlation) or a cer- tain pressure drop has built up (pressure drop correlation). The corresponding flow rates are otherwise set to zero. At low temperatures liquid and vapour are not in equilibrium, and the pressure is therefore first set to a constant initial value, the temperature is calculated independently from the VLE and the vapour phase is not considered. When during the heating-up the temperature calculated by the startup model reaches the additionally calculated bubble point temp- erature at this initial pressure and the current liquid compo- sition, then the vapour phase is included in the model and temperature and pressure correspond to the VLE [switching according to equation (4)]. In the decanter model the possible appearance of a second liquid phase is considered during the whole startup by continuously checking the liquid composition. If Tj ! TLV j then Tj ¼ TLV j Vsection ¼ HUL j vL þ HUV j vV Else Tj = TLV j pj ¼ const: HUV j ¼ 0 (4) The developed simulation model has been validated with startup data from the presented experiment. Figure 5 shows the comparison between the concentrations in the reboiler and the organic phase in the decanter during startup. The startup begins with the start of the feeding and the reboiler duty ramp (t ¼ 0). Since reboiler and decanter are initially empty, the first liquid samples are taken after 1–1.5 hs. The simulation describes the dynamic behaviour very well, includ- ing the instant formation of a miscibility gap when the first sub- cooled condensate enters the decanter. In the following the startup process with different initial charging of the decanter is analysed using the validated model. During all the simulation runs the reboiler is initially empty and filled by the feeds entering the column. When the level is high enough the reboiler duty is increased step- wise to the steady-state value. In case of an initially charged Figure 4. Liquid composition profile along theoretical stages. HOAc (5), PrOH (A), PrOAc (4), H2O (W). Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 113
  • 6. decanter, the reflux is turned on as soon as the first vapour condenses, otherwise (case 0) the decanter is first filled with the condensate. As in the experiments the organic phase is completely returned to the column as the reflux, the aqueous phase is withdrawn from the system. Specifica- tions are taken from the experimental setup. Figure 6 shows the comparison of the trends of the MX function (calculated for the organic phase in the decanter) which gives the deviation of the current from the steady state composition for the different startup strategies. The strong influence of the composition of the initial charging can be seen. As expected, initial charging of the two-phase top product leads to the fastest startup (case 3) and reduces startup time compared to an initially empty decanter (case 0). The composition of the organic phase deviates only very little and only for a short period of time from the steady-state values. All other initial charging results in a slower startup. The prolonging of startup time due to charging of the two phase product to the decanter as described by Tran (2005) could therefore not be observed for this process. Conceptual Design The conceptual design of a high-purity n-propyl acetate production process is initiated on the basis of the method developed by Tang et al. (2005) assuming that any acetic acid esterification reactive distillation process includes both rectifying and stripping sections: for n-propyl acetate system, when most of acetic acid is consumed in the reactive section, composition at the top of the column is close to the minimum boiling ternary azeotrope located inside the two liquid phases region (Figure 7); thus a decanter is placed to remove the aqueous phase of the distillate whereas organic phase is further processed in a stripping column to obtain high-purity n-propyl acetate: Figure 8 regroups all studied flowsheets for this esterification. The process design has been developed using a sequen- tial approach based on steady-state simulations. The feasible values of different parameters leading to a PrOAc molar purity of 98% for each configuration [except (e) for which water content in product stream is remaining slightly higher than 2%)] are related in Table 5. The qualitative analysis of all possible flowsheets gives the following conclusions: . the optimal feed tray for the acetic acid (high boiling reac- tant) is the column base unlike the usual double-feed con- figuration (the low-boiling reactant is fed in the reboiler and Figure 7. RCM of the PrOH/PrOAc/H2O non-reactive mixture show- ing the miscibility gap, the distillation boundaries, the unstable (W) and stable nodes (†) and the saddle points (A). Figure 5. Dynamic composition trend in the reboiler (top) and of the organic phase in the decanter (bottom) from experiment and simu- lation. HOAc (5,–dash-dot line–), PrOH (A, dashed line), PrOAc (D, solid line), H2O (o, dotted line). Figure 6. Comparison of the MXtop-function in the organic phase in the decanter during startup for different simulation runs. (0): initially empty decanter, (1): initial charging of n-propanol, (2): initial charging of n-propyl acetate, (3): initial charging with two phase mixture (steady state composition), (4): initial charging of acetic acid. Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 114 BREHELIN et al.
  • 7. Figure 8. Six possible flowsheets for the n-propyl acetate production by catalytic reactive distillation (single/multiple feeds). (a) (b) Partial withdrawal of the decanter organic phase towards an additional stripping column; (c) (d) total withdrawal of the decanter organic phase towards an additional stripping column; (e) (f) total reflux of the decanter organic phase in a single reactive distillation column with an inter- mediate PrOAc withdrawal. Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 115
  • 8. high-boiling above the reactive section); this can be explained since the reboiler holdup is larger than the column holdup; . the bottom products may not always be neglected since depending on PrOAc stream specifications (purity and recovery rate), unreacted HOAc has to be withdrawn; . a configuration with total withdrawal of the decanter organic phase [(c) and (d)] is more suitable to produce high-purity PrOAc; the last column in Table 5(c’) shows that a quasi-pure PrOAc stream can be produced when the stripper NTS and the reboiler duties are increased; . the single-column configuration with an intermediate PrOAc withdrawal (e) is powerful in terms of conversion and recovery, but seems not to be adapted to pure PrOAc production: water is still remaining in the product stream (2%) which may not be suitable. Moreover, an addition of 10 stages in the rectifying section results in only a slight improvement of PrOAc purity (,0.1%). Note that the number of theoretical stages in the rectifying section, the reflux ratio of organic phase of the decanter and the feed ratio are the main sensitive parameters. The rectify- ing section is required to decrease the acetic acid amount in the distillate (see liquid composition profiles Figure 9), but its concentration in the bottom products depends on the reaction conversion only: that’s why a small alcohol excess is set in the feed. Some attempts of fractionating the alcohol feed were carried out in order to increase reaction rate on the upper stages of the reactive section, but no significant effect was found. CONCLUSIONS This study is focused on the conceptual design of high- purity n-propyl acetate production by reactive distillation. The first part of this article is devoted to experimental results obtained on two kinds of device: stirred reactor and RD pilot plant. A kinetic model for both homogeneously and heteroge- neously catalysed reaction has been developed and chemi- cal equilibrium data have been successfully compared with previous publications. The second part deals with the Table 5. Feasible setups of the three single-feed configurations. (a) (c) (e) (c’) Design variables NTS reactive section 15 15 16 15 NTS rectifying section 15 15 10 þ 10 15 NTS stripping section 15 10 — 20 Operating variables FHOAc (mol h21 ) 83.26 83.26 83.26 83.26 FPrOH (mol h21 ) 84.03 84.03 84.03 84.03 Decanter temperature (8C) 20 20 20 20 Decanter reflux ratiob 0.923 0 1 0 RD reboiler duty (kW) 4.468 5.025 6.220 6.205 Stripper reboiler duty (kW) 1.819 1.115 — 2.000 Stripper reflux ratio 4.985 1.000 — 1.000 Products PrOAc flow rate (mol h21 ) 82.82 83.22 84.61 82.83 xHOAc 0.0018 0.0003 0.0002 0.0005 xPrOH 0.0182 0.0197 0.0023 0.0005 xPrOAc 0.9800 0.9800 0.9770 0.9990 xH2O 0.0000 0.0000 0.0202 0.0000 H2O flow rate (mol h21 ) 82.24 82.11 82.33 84.22 xHOAc 0.0004 0.0001 0.0000 0.0000 xPrOH 0.0055 0.0023 0.0099 0.0109 xPrOAc 0.0029 0.0015 0.0031 0.0032 xH2O 0.9912 0.9961 0.9870 0.9859 Bottoms flow rate (mol h21 ) 2.22 1.97 0.35 0.24 xHOAc 0.6551 0.6937 0.7402 0.7397 xPrOH 0.1918 0.1702 0.1463 0.1466 xPrOAc 0.1084 0.0972 0.0816 0.0817 xH2O 0.0447 0.0389 0.0319 0.0320 Molar conversion (%) 98.0 98.3 99.7 99.7 Recovery rate (%) 99.4 99.6 99.7 99.7 Performance indexb (%) 97.4 98.0 99.3 99.4 a Ratio of the organic stream refluxed in the RD column on the withdrawn stream sent to the stripper. b Ratio of molar PrOAc production on molar HOAc feed. Figure 9. Liquid phase composition profiles and temperature along reactive distillation for type (c) process (PrOAc purity of 99.9%). HOAc (5), PrOH (A), PrOAc (4), H2O (W) and temperature (solid line). Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 116 BREHELIN et al.
  • 9. simulation of the reactive distillation pilot plant. Model vali- dation has been carried out with steady-state and dynamic data and an optimal startup strategy has been recommended using dynamic simulations. On this basis, six different setups have been identified as being able to product high-purity n- propyl acetate; a single-feed configuration with a decanter at the top of the column and the further purification of all the decanter organic withdrawal has been noticed to be the most suitable design to achieve such a production. In the near future, an optimal design of a catalytic distilla- tion column will be achieved based on our model and using MINLP approach; finally, the performances of this new design will be experimentally checked in an heterogeneous catalysed column located in TU Berlin. NOMENCLATURE a liquid phase activity Dn molar transformation, mol E energy of activation, J mol21 F flow rate, mol h21 K equilibrium constant Ko preexponential factor HU hold-up, mol n molar amount, mol p pressure, bar Q reboiler duty, kW r rate of reaction, mol s21 T temperature, K v molar volume, m3 / mol V volume, m3 x liquid phase composition Subscripts and superscripts a activity f forward (esterification) i component index j stage index L liquid LV boiling point o reference or initial state V vapour REFERENCES Bart, H.J., Kaltenbrunner, W. and Landschutzer, H., 1996, Kinetics of esterification of acetic acid with propyl alcohol by heterogeneous catalysis, Int J Chem Kinetics, 28: 649–656. Brehelin, M., Rouzineau, D., Thery, R., Meyer, X. and Meyer, M., 2006, Calcul des courbes de re´sidu re´actif he´te´roge`ne de la reaction de synthe`se de l’ace´tate de n-propyle, Proceedings of SIMO’06. Chiang, S.F., Kuo, C.L., Yu, C.C. and Wong, D.S.H., 2002, Design alternatives for the amyl acetate process—coupled reactor column and reactive distillation, Ind Eng Chem Res, 41: 3233–3246. Engel, V., Stichlmair, J. and Geipel, W., 2001, Fluid dynamics of pack- ings for gas-liquid contactors, Chem Eng Technol, 24(5): 459–462. Fillon, M., Meyer, M., Pingaud, H. and Joulia X., 1995, Data reconci- liation based on elemental balances applied to batch experiments, Comput Chem Eng, 19S: 293–299. Huang, Y.S., Sundmacher, K., Tulashie, S. and Schlu¨nder, E.U., 2005, Theoretical and experimental study on residue curve maps of propyl acetate synthesis reaction, Chem Eng Sci, 60: 3363–3371. Marek, J.and Standart, G., 1954, Vapor–liquid equilibria in mixtures containing an associating substance. II. Binary mixtures of acetic acid at athmospheric pressure, Collect Czech Chem Comm, 19: 1074–1084. Okasinski, M.J. and Doherty, M.F., 2000, Prediction of hetero- genenous reactive azeotropes in esterification systems, Chem Eng Sci, 55: 5263–5271. Pitochelli, A.R., 1980, Ion exchange catalysis and matrix effects (Rhom and Haas Co., Philadelphia, PA). Reepmeyer, F., Repke, J.-U. and Wozny, G., 2004, Time optimal start-up strategies for reactive distillation columns, Chem Eng Sci, 59(20): 4339–4347. Sørensen, J.M. and Arlt, W., 1982, Liquid-Liquid Equilibrium Data Collection. Chemistry Data Series, (Dechema). Steinigeweg, S., 2003, Zur Entwicklung von Reaktivrektifikationspro- zessen am Beispiel gleichgewichtslimitierter Reaktionen (Shaker Verlag Aachen Germany). Steinigeweg, S. and Gmehling, J., 2002, n-Butyl acetate synthesis via reactive distillation thermodynamic aspects, reaction kinetics, pilot-plant experiments, and simulation studies, Ind Eng Chem Res 41:, 5483–5490. Tang, Y.T., Chen, Y.W., Huang, H.P., Yu, C.C., Hung, S.B. and Lee, M.J., 2005, Design of reactive distillation for acetic acid esterifica- tion, AIChE J, 51(6): 1683–1699. Tran, T.K., 2005, Analyse des Anfahrens von Dreiphasenkolonnen (Shaker Verlag, Aachen, Germany). Ung, S. and Doherty, M.F., 1995, Vapor–liquid phase equilibrium in systems with multiple chemical reactions, Chem Eng Sci, 50: 23–48. Weisz, P.B. and Hicks, J.S., 1962, The behaviour of porous catalyst particles in view of internal mass and heat diffusion effects, Chem Eng Sci, 17: 265–275. The Manuscript was received 18 July 2006 and accepted for publication after revision 24 October 2006. Trans IChemE, Part A, Chemical Engineering Research and Design, 2007, 85(A1): 109–117 PRODUCTION OF n-PROPYL ACETATE BY REACTIVE DISTILLATION 117