Sei sulla pagina 1di 13

Chemical Engineering and Processing 96 (2015) 113

Contents lists available at ScienceDirect

Chemical Engineering and Processing:


Process Intensication
journal homepage: www.elsevier.com/locate/cep

Comparison of different reactive distillation schemes for ethyl acetate


production using sustainability indicators
Miguel A. Santaella, Alvaro Orjuela* , Paulo C. Narvez
Grupo de Procesos Qumicos y Bioqumicos, Departamento de Ingeniera Qumica y Ambiental, Universidad Nacional de Colombia Sede Bogot, Cra 30 # 4503, Edicios 412, Bogot, Colombia

A R T I C L E I N F O

A B S T R A C T

Article history:
Received 30 March 2015
Received in revised form 3 July 2015
Accepted 30 July 2015
Available online 3 August 2015

An assessment of different intensied processes for ethyl acetate production by direct esterication was
performed. After the selection and validation of the adequate thermodynamic and kinetic models, a
comparison among different reported processing technologies (i.e., conventional, reactive distillation,
reactive distillation with pressure swing, dividing wall column with reactive reboiler), and a novel
proposed conguration using a reactive dividing wall column was carried out. Specications of raw
materials and products used as constrains during simulations were dened according with market
requirements. To perform a fair assessment among different alternatives, each process was optimized
using a mixed strategy of sensitivity analysis coupled with sequential quadratic programming algorithm.
Total annual cost was selected as the optimization variable. Conversion, productivity, mass intensity,
mass productivity, Sheldons Factor (E-Factor), water-free Sheldons factor (EW-Factor), and energy
intensity were used as sustainability indicators. According with results, the novel reactive dividing wall
column conguration proposed in this work ends up being the more energy efcient and cost effective
(46% energy and 26% cost savings compared with the traditional process) and also it is characterized by
the best sustainability indicators.
2015 Elsevier B.V. All rights reserved.

Keywords:
Reactive distillation
Dividing wall column
Ethyl acetate
Esterication
Sustainability

1. Introduction
Over the last decades, there has been an increasing interest in
products and processes that meet sustainable development
criteria. After the broad denition of sustainability accounting
for the triple bottom line dimensions (environmental, social and
economic), different attempts have been done to transform such a
concept into basic green engineering principles, and furthermore
into sustainability indicators to measure how green processes or
products can be. Despite the sustainability indicators can be used
to evaluate existing processes, its major value can be encountered
when applied in the analysis of new processes that need to be more
cost-effective and environmentally friendly than existing ones; for
example, when developing processes that switch from fossil to biobased feedstock. Due to the increase use of biomass as a feedstock,
and taking into account its higher costs compared with fossil
feedstock, it is necessary to develop more efcient technologies to
reach cleaner, safer, and more cost-effective production of biobased derivatives [1].

* Corresponding author.
E-mail address: aorjuelal@unal.edu.co (A. Orjuela).
http://dx.doi.org/10.1016/j.cep.2015.07.027
0255-2701/ 2015 Elsevier B.V. All rights reserved.

Among the different types of chemicals currently produced


from non-renewable feedstock that can be replaced by bio-based
alternatives, solvents stand out because their intensive consumption in a large variety of processes and products. In addition to the
fossil resources depletion, the environmental impacts of using
traditional solvents are evident, for instance in the contamination
of surface and underground waters and also in the air pollution
near urban areas. From the variety of traditional solvents, ethyl
acetate (EtAc) distinguishes as a major industrial commodity,
which is mainly used as solvent for paints, coatings, resins, inks,
and in decaffeination. It is also a main ingredient in different
fragrances and avors for consumer products.
There are several chemical routes for EtAc production
implemented at the industrial scale, using both fossil-based and
bio-based raw materials (Table 1). EtAc is mainly produced by
esterication of acetic acid and ethanol in liquid or vapor phase (1),
by acetylation of ethylene (2), and by ethanol dehydrogenation (3).
The safety indicators [2] and atom economies of these routes are
also presented in Table 1.
The presence of ethylene in the olen acetylation (route 2), the
production of hydrogen in the Tishchenko pathway (route 3)
together with the extreme conditions of both processes affect
negatively their safety indicators. This reduces the benign-ness of

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

Nomenclature

BRD
C
Ea
EtAc
EtOH
H2O
H2SO4
HAc
k
Kc
PS
RD
RDWC
Wcat
x
y
TAC

Batch reactive distillation


Concentration (mol/L)
Activation energy (J/mol)
Ethyl acetate ()
Ethanol ()
Water ()
Sulfuric acid ()
Acetic acid ()
Kinetic constant (mol/s)
Equilibrium constant
Pressure swing
Reactive distillation ()
Reactive dividing wall column ()
Catalyst mass fraction (kgcatalyst/kgsolution)
Mole fraction liquid phase ()
Mole fraction vapor phase ()
Total annual cost (million Usd/year)

these routes despite the fact both have high atom economy and
that hydrogen co-produced in the second route could be of interest.
Comparatively, direct esterication (route 1) has the lowest atom
economy from all three reactions; however it has the lower value
of the safety indicator being the most benign chemical route
according with metrics described by Srinivasan et al. [2].
Additionally, in a recent study it was found that among the
chemical routes of Table 1, direct esterication of bio-based acetic
acid and ethanol appears to be the more cost-effective alternative
as well as the one with the lower global warming potential [3].
Despite Fisher esterication to EtAc is a classic example used in
the academia to introduce basic concepts on chemical reaction
engineering and thermochemistry, this system is still under active
investigation. In recent years, different processing technologies
applied to EtAc production have been reported, most of which
focus on a process intensication (PI) approach. Notwithstanding
the proposed technologies have claimed to be cost-effective and
environmentally friendly, no sustainability assessment of such
technologies has been carried out. Furthermore, as there is room
for future processes development by introducing more complexity
during PI (for instance by implementing heat-integrated reactive
separations), there is need for including also the sustainability
evaluation in the early stages of the process synthesis.
Benets of PI are associated with the enhanced safety, space and
waste reductions, energy savings, higher efciency and productivity, and consequently better economic and environmental performance. As an effort to classify most PI attempts reported in the
literature, Van Gerven and Stankiewick [4] described four
approaches commonly used according with the studied domain:
thermodynamic, functional, spatial, and temporal.
From a general point of view, the thermodynamic approach for PI
consists of energy integration of streams, stages or operations, to
replace the required amount of heating and/or cooling provided with

utilities. The use of preheating exchangers, pump-arounds, vapor


recompression, and dividing wall columns are examples of such
energy integrations. The functional domain deals with the integration of two or more operations or technologies within a single unit.
This is the case of hybrid and reactive separations such as reactive
distillation (RD), catalytic distillation, reactive adsorption, membrane distillation, and membrane reactors among others.
Considering all the above, this work studies and develops a novel
processing technology for EtAc production by using a reactive
dividing wall column. Although RD and dividing wall column with
reactive reboiler has already been proposed separately [5,6], the
proposed conguration of using a reactive dividing wall column
(RDWC) is innovative for EtAc production, and has not been reported
in the open literature. The performance of the proposed technology
was compared with other intensied processes by considering
quantitative economic and sustainability indicators.
Because most reports on EtAc production by intensied
processes have been developed using different modeling basis
(thermodynamics, kinetics, models, etc.), specications (raw
materials and product purities, catalysts, etc.), and assumptions;
a direct comparison among them is difcult and anyhow
subjective. In this sense, it was necessary to carry out such
comparison under equivalent conditions with the optimized
processes. To the knowledge of the authors, this comparison of
current intensied technologies (including economic and sustainability indicators) has not been previously accomplished.
2. Methodology
In order to objectively compare the different intensied
processing alternatives reported in the literature, a complete
revision of the fundamentals, constrains, and specications of the
process was necessary. To begin with, a proper thermodynamic
model that adequately describes phase equilibria of the reactive
systems was selected. Regarding reaction kinetics, there was the
need to select a specic catalyst and to explore and validate the
corresponding kinetic model. Afterwards, a review of the reported
intensied processes for EtAc production was performed, paying
special attention to reactive distillation sequences. Detailed
described processing alternatives were selected for simulations,
and further sustainability analysis was performed. Simulation of
the different schemes was performed by mean of steady-state
equilibrium-stage simulations using Aspen plus 7.31. Raw
materials and product specications used during simulations
were the same for all intensied processes under study. In this
sense, specications were established according with commercial
technical datasheets. In some cases, some adjustments over the
original reported processes were necessary to achieve the
specications of the commercial product. The design pressure
for all columns was selected to be 1 bar, except for the pressured
column in the pressure swing process. Temperature in the
condensers and decanters was veried to guarantee the use of
water from cooling towers, which was assumed to be at 25  C.
After all simulations met the design criteria, the total annual
costs (TAC) were computed for each conguration. To do so, a 3year period for return on investment was used. Since worlds

Table 1
Industrial chemical routes for ethyl acetate production and some sustainability indicators.
Chemical route

Reaction

Atom economy (%)

Safety indicatora

1
2
3

CH3 CH2 OH CH3 COOH$CH3 COOCH2 CH3 H2 O


CH2 CH2 CH3 COOH ! CH3 COOCH2 CH3
2CH3 CH2 OH ! CH3 COOCH2 CH3 2H2

83,0
100
95,6

1.7
2.6
2.2

Calculated as a cumulative index of the four chemical safety parameters (Toxicity, Reactivity, Explosiveness and ammability) as mentioned in [2].

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

market for EtAc is around 3 million tons per year, and large plants
have a capacity for manufacturing around 100.000 t per year [7],
this was selected as the base case production capacity. For the xed
costs calculation, the methodology proposed by Douglas [8] was
used (Eqs. (1)(4). To calculate the variable costs, updated
averaged raw material and utility prices were consulted. Natural
gas was used as fuel and 85% efciency was assumed for heating
loop.
TAC Fixed Costs Variable Costs

Fixed Costs

Installed Cost
3yr

Installed Costs Base CostCost indexIF Fc  1

(1)

(2)
(3)

here, IF corresponds to installation factor, and Fc is a correction


factor for material, pressure, etc. [8]. Operating costs were
calculated based upon utilities consumption, specically heating
costs.


USD

EnergyCosts
yr
 
 

h
Heat duty kW
 Natural gas price USD
 8160 yr
h
m3
 
(4)
0; 85  Natural gas energy kW
m3
After the preliminary economic evaluation of the base cases,
optimization of each conguration was performed by running a
sensitivity analysis over the discrete variables, coupled with a
sequential quadratic programing (SQP) algorithm for the continuous variables to minimize TAC. Thereafter, sustainability indicators
were calculated to complement the economical comparison. First,
traditional indicators such as conversion, recovery and productivity were calculated according to Eqs. (5) and (7). These indicators
contribute to evaluate the inherent safety and therefore the
sustainability of a given process since conversion and recovery, and
hence productivity and yield, are also directly related to the
inventory of reactants and recycle streams ow rates.[2]
ConversionX

RecoveryRc

moles of HAC Converted


moles of HAC Fed

moles of EtAc in product Stream


moles of HAC Converted

ProductivityP X  Rc
moles of EtAc in product Stream

moles of HAC Fed

(5)

(6)

(7)

Then, energy intensity (EI), Sheldons E-factor, water-free


Sheldons EW-Factor mass intensity (MI), and mass productivity
(MP) [9,10], were calculated according to Eqs. (8)(12).
Energy intensityEI

Energy usedW
mass of product kg

Total waste streams kg


mass of product kg

EW

(8)

mass IntesityMI

Total mass fed kg  Water kg


mass of product kg

Mass productivityMP

1
 100
MI

(11)

(12)

The EI factor measures the amount of energy used per kilogram


of product. In this case, we added up the major sources of power
consumption (i.e., distillation) and then dividing it by the mass
ow of pure EtAc. Sheldons E-Factor is a measure of the adequate
or inadequate utilization of resources; it quanties the mass of
waste generated per kilogram of product. For some authors, all
efuents from the process different from the main product stream
will be considered waste while other authors consider water
should not be included in the total waste calculation. We have
included both calculations since water is used in some processes to
facilitate phase splitting, but it is also generated as a byproduct of
the reaction. An analysis regarding the differences between both
Sheldon's factors (water-free and not water-free) will be carried
out in the discussion section with the calculated values from each
alternative. The MI factor is a ratio of the total mass required per kg
of product; a MI value of 1 means that all of the mass consumed in
the process is being transformed into useful product. The greater
the value of the MI factor, the greater the amount of byproducts
and waste. MI does not take into account the amount of water
being used since it is considered as a carrier or solvent rather than
as a reactant, likewise it is also not taken into account as part of a
product. The MP, is the mass percentage that is actually being
transformed into useful product, it may be calculated as the inverse
of the MI.[10]
3. Modelling fundamentals
3.1. Thermodynamics
The quaternary reactive system under study (HAcEtOHEtAc
H2O) exhibits non-ideal behavior with the formation of three
homogeneous azeotropes and one heterogeneous azeotrope [11].
Additionally, H2O and EtAc are partially immiscible; therefore,
various activity coefcient models have been proposed to address
the non-ideality of the liquid phase. Regarding the vapor phase,
acetic acid has been found to dimerize, and this behavior has been
successfully modeled with equations of state corrected by the
chemical theory (e.g., Hayden OConnellHOC). Nevertheless,
some previous researchers have assumed ideality in the vapor
phase to simplify the VLLE model. In a previous work, Tang et al. [5]
validated a set of parameter for an NRTL-HOC model with
experimental data of pure, binary and ternary mixtures. Because
the goodness of the model, this has been applied in subsequent
investigations [1214], and will be used in this work to describe
phase equilibrium of the reactive system. Parameters of the
thermodynamic model are summarized in the supplementary
material.
3.2. Kinetic models

(9)

Total waste streams kg  Water in waste streamskg


mass of product kg
(10)

A large amount of kinetic studies on Fisher esterication to EtAc


have been reported using acid catalysts. Traditionally, H2SO4 has
been widely used as catalyst because is more active than most
commercial homogeneous (e.g., p-toluene sulfonic acid, Methanesulfonic acid) and heterogeneous (e.g., ion exchange resins)
catalysts. H2SO4 loadings in commercial applications range from
0.2 to 1% wt of the reactive mixture [15]. However, there are some

Table 2
Kinetic expressions reported for acetic acidethanol esterication with different catalyst.
Catalyst

Rate

Parameters

H2SO41

H2SO42

HCl

rA

dC AcA
p
q
k1 C nAcA C m
EtA  k2 C EtAc C H2 O
dt

r k1 c1 c2  k2 c3 c4


C A C B  CEKCW
2:6  1014 e 104129
RT
dC
rA  A
dt
C B 3:7C W

Amberlyst
152



 
Ea;3
xEtAc xH2 O
r3 wcat rsol ka;3 exp
 xAcAc xEtOH 
RT
K a;3

Amberlyst
153

 
dC
k
 A kC A C B 
CE CW
Ke
dt

Amberlyst
35 wet

Amberlyst
36

Amberlyst
70
Purolite1
CT179

Auto cat.
1

Auto cat.
2

r mcat k1 C HAc C EtOH  k1 C EtAc C H2 O

r wcat k C HAc C EtOH  k1 C EtAc C H2 O



6500:1
k1 4:195C k 0:08815exp 
T

kc 7:558  0:012T


310
k1 4:86  104 0:496exp 
0:205C 0:85
T


220
k2 3::715  104 0:01572exp 
0:09785C 0:85
T
n = m = 0.5, p = q = 1
k1 = 4.76  104
k2 = 1.63  104
T = 100  C
keq 4:0  0:2

Ka,3 = 1320
Ka,3 = 4550
Ka,3 = 2.71
K = 0.0026
Ke = 3.2
T = 343K (70  C)
Wcat = 100gcat/L



6105:6
k1 1:24  109 exp 
T



5692:1
k1 1:34  108 exp 
T


57960
k 131137exp 
RT



 
Ea;3
xEtAc xH2 O
ka,3 = 12,500  200
r3 wcat rsol ka;3 exp
 xAcAc xEtOH 
RT
K a;3
Ea,3 = 56,700  200
Ka,3 = 3.0  0.1
r k1 xEtOH xm
m = 1.5
AcOH  k2 xEtOAc xH2 O

r k1 c1 c2  k2 c3 c4



CE CW
E
r A1 e RT C A C B 
Ke



60550
k 81389exp 
RT



48300
k1 4:24  106 exp 
RT


66200
k2 4:55  108 exp 
RT


59774
k1 0:485exp 
RT


59774
k2 0:123exp 
RT

A1 = 4.85  102
E = 14,300 cal/mole
Ke = 4

Ref

Units

Comments

Modications

[23] r:mol/sm3, C:mol/


m3
Ck:%vol
[19]

C:Sulphuric acid
conc (wt%)
rA:kmol/m3min
Ci:kmol/m3

[21]

C:mol/L
r:gmol/(Lmin)
k:L/(mingmol)
[22] Ci:mol/dm3
Time :min
Cinetic factor
(m3)2kmol2
s1

kmol
1
[24]
ka;3 :
kgcat
s
Ea,3:kJ/kmol
[20] k:L/(molmin)

[14]

K is non temperature dependent

Authors used mole


Forward frequency factor units
(m3)2 kmol2 s1 are not consistent with fraction to t kinetic
model. Ci replaced by xi
concentration based kinetics
Validated but non peer reviewed

k units are not mentioned, although it


should be in L/(molmin). It is non
temperature dependent

m:gcat
r:mol/min

[13]

R: Jmol/K
C: mol/L
r:mol/Lgmin
k:L/molming
wcat:g

 
kmol
1
[17]
ka;3 :
kgcat
s
Ea,3:kJ/kmol
[25] k1,k2:mol/kgcats
R: J/mol

[23] Ci: mol/dm3


r: mol/sm3
k: m3/mols

[26]

C units changed to weight


fraction

A1 :

1
mole:s

C:mol/L
r:mol/Ls

R should be in J/(molK) instead of Jmol/


K
r includes Wcat in expression, therefore
units should be mol/Lmin instead of mol/
Lgmin

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

Amberlyst
15 1

k1
r 1 k1 C 1 C 2  C 3 C 4
Kc

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

reaches a conversion up to 20% at high residence times. This


indicates that the auto-catalytic reaction rate has to be included
when modeling the non-catalyzed reactive stages in RD columns,
mainly at high temperatures. A kinetic model reported for
Amberlyst 15 [22] seems to be inconsistent because it reaches
equilibrium condition in less than one hour, and exceeds the rate of
reaction obtained with sulfuric acid.
Another important outcome of this validation is the unexpected
behavior observed on the Alejski and Duprats kinetics [23] using
sulfuric acid as catalyst (H2SO4-1 in Table 2). This prole does not
show signicantly faster kinetics than most resin catalyzed
reactions, which does not agree with literature reports or with
the previous experience of the authors. It is interesting to point out
that this model (apparently incorrect), has been widely used in
most studies of EtAc production via RD using sulfuric acid as
catalyst [5,12,2729]. In comparison, the kinetic prole obtained
for H2SO4 catalyzed reaction using Atalays model [19] (H2SO4-2
in Table 2) does show the expected behavior, reaching equilibrium
condition much faster than heterogeneous catalysts.
In order to compare different processing alternatives for EtAc
production under the best reaction conditions, sulfuric acid was
selected as catalyst for further simulations. Taking into account the
above, simulations were performed using the kinetic model
reported by Atalay [19], and also the auto-catalytic kinetic model
[23] was included when needed. For simplication, the catalyst
loading was assumed constant along the reactive stages within the
columns.
Unlike mass or molar-fraction, the molarity-based kinetic
expressions used in this work are temperature dependent. To
ensure good kinetic predictions, it is necessary to have good
representation of the temperature dependence of densities in the
reactive system. In Figs. S1 and S2 of the supplementary material it
is presented a comparison of experimental and predicted densities
of pure components and reactive mixtures respectively at different
temperatures (Data also included in Supplementary material). As
veried, liquid density predictions are in good agreement with

disadvantages when using H2SO4, including the darkening of


products, equipment corrosion, required neutralization, and
infeasibility of reusing and recycling the catalyst. Despite these
disadvantages are avoided when using solid catalysts, almost none
commercial heterogeneous catalyst, including resins, has been
found to increase reaction rates to EtAc better than sulfuric acid
under the same operating conditions and acid equivalent loadings.
In Table 2, a summary of reaction rate expressions reported
using different catalysts is presented, together with the corresponding parameters. Since some kinetic expressions have been
obtained including different thermodynamic models than the one
selected here (activity-based kinetics [16,17]), only the reports
including concentration-based or molar fraction-based kinetics are
listed. Expressions with no registry of experimental validation
were excluded [18]. In order to evaluate reliability of the kinetic
expressions, a comparison of these models was performed under
the same operating conditions that are feasible to reach within a
reactive distillation (RD) column. Despite rate equations were used
as published, modications in some models were necessary to
reproduce experimental data reported on the original articles. For
instance, using mass fraction instead of mass percentage to
calculate the rate constant in the work by Atalay [19] reproduced
the exact values of the kinetic constants reported in the original
paper. In some extreme cases, reported rate expressions were
inconsistent taking into account the units of the parameters and
the variables of the model. Some comments over these rate models
are also included in Table 2.
Liquid phase esterication proles obtained with kinetic
expressions from Table 2 are presented in Fig. 1. Comparison
was performed under batch operation using an equimolar feed of
acid and alcohol, at 80  C, and with the same amount of acid
equivalents corresponding to a loading of 0.5% wt of H2SO4.
Expressions reported at a single temperature (i.e., Amberlyst 15
[20] and hydrochloric acid [21]) were excluded from Fig. 1. As
expected, most proles show an equilibrium conversion around
64%, and a large improvement when using catalysts compared with
non-catalytic reaction. However, the non-catalyzed reaction

HAc Conversion

0.8
0.7

Amb15-1

0.6

Amb15-2
Amb35

0.5

Amb36
0.4

Amb70
Purolite CT179

0.3

H2SO4 -1

0.2

H2SO4 -2
0.1
0

Auto cat -1
0

200

400

600

800

1000

Auto cat -2

Time (min)
Fig. 1. Effect of catalysts on conversion of acetic acid (HAc) during esterication with ethanol (EtOH) at 80  C and molar ratio 1:1.

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

experimental reports, hence, the molarity-based kinetic model can


be used with condence.

and raw materials specications, catalysts, kinetics, and a variety of


thermodynamic models to describe phase equilibrium.

3.3. Processing alternatives for ethyl acetate production

3.3.1. Base caseconventional process


A scheme of the conventional continuous process to produce
EtAc is presented in Fig. 2. In this process, the esterication
reaction is carried out in a stirred tank reactor (Rx) at 80  C with a
residence time of 200 min where excess HAc is used to shift the
reaction towards the ester production. The out coming stream
from the reactor is directed to the principal distillation column
(DC) to separate EtAc from the remaining HAc. The distillate from
the principal column is cooled down and mixed with pure water to
trigger phase separation in a decanter. The organic phase from the
decanter, containing EtAc above 90% wt, is puried in a recovery
column (RC) to obtain pure EtAc in the bottoms. The distillate from
the recovery column is once again returned to the main column
(DC). Although not included in [44], an azeotropic distillation
column (AD) which uses EtAc as an entrainer is generally used to
recover unreacted HAc and to recycle it back to the reactor. The
distillate from the azeotropic column, containing mainly H2O and
EtAc is cooled down to trigger phase separation and the organic
phase is returned to the principal column. In addition to the
complexity and large energy consumption of this process, there are
raw materials and product loss within the aqueous outlet streams.

The liquid esterication of acetic acid (HAc) with ethanol


(EtOH) is a reversible reaction. Equimolar amounts of acid and
alcohol reach an equilibrium conversion of around 64% at 80  C.
Therefore, to increase conversion any of the following alternatives
may be applied: the use of excess of acid or alcohol, the reduction
of temperature as the reaction is slightly exothermic, or the use of
simultaneous reaction and separation techniques to displace
equilibrium towards product formation. The temperature reduction in this case is not recommended since the equilibrium
constant barely increases (71% conversion at 25  C, with a molar
basis equilibrium constant of 6,5), and there is a more pronounced
negative effect caused by a reduced reaction rate, increasing the
required reactor volume. Likewise, the use of an excess of reactant
will require recovery and recycle. If excess of alcohol is used,
ternary separation of EtOHEtAcH2O is challenging because the
azeotropic behavior of the mixture. When an excess of HAc is used,
equipment has to be corrosion-resistant and it is necessary to
accomplish HAc separation from water, which is also difcult and
generally requires entrainers in distillation. In this sense, the
preferred alternative for process intensication is to carry out
simultaneous reaction and separation, for example by using
reactive distillation (RD). Table 3 summarizes main characteristics
of some recent attempts to improve EtAc production by using RD
under different congurations (e.g., conventional, pressure swing,
multiple reactions, thermally couple distillation, dividing wall
column). As observed, these investigations use different product

3.3.2. Reactive distillation


Various reactive distillation congurations have been studied
for the production of EtAc [5,12,14,41,42,29,4547]. However, most
processes can be summarized in a scheme that includes a RD
column, a decanter and a purication stripper (RC), as shown in
Fig. 3. HAc inlet stream is mixed with the recovered bottoms from

Table 3
Reports on intensied processes involving reactive distillation for ethyl acetate production.
Technology

Catalyst

Thermodynamic
model

Reactants purity (%)

EtAc purity (wt%)

Ref.

RD
RD + pressure swing
RD
RD
BRD
RD
RD + coproduction of butanol
Heat integrated RD
RDWC
RD
R and RD

H2SO4
H2SO4

H2SO4

Purolite
Potassium Butanolate
H2SO4
Equilibrium
H2SO4
Purolite

UNIFAC:IDEAL
WILSON:IDEAL
WILSON:IDEAL
KOMATSU
Constant a
UNIQUAC-HOC
NRTL:IDEAL
NRTL:IDEAL
NRTL:IDEAL
NRTL:IDEAL
NRTL:IDEAL

82,2a
99,5
98,6c
83,3a

82,4a
99,7a
99,7a
82,6a
80,4a
92,8

[27]
[16]
[30]
[23]
[18]
[31]
[32]
[28]
[33]
[34]
[35]

BRD
RDWC
RD
SBRD
RD
RD
RD
RD
RD
RD + coproduction of butyl acetate
RD

Equilibrium
Non-cat
Purolite
H2SO4
H2SO4 and Purolite
H2SO4 and Purolite
Amberlyst 15 and Purolite
H2SO4
Amberlyst 15
Amberlyst 35

NRTL:IDEAL
NRTL:IDEA
NRTL:IDEAL
NRTL:IDEAL
NRTL:PR:IDEAL
NRTL:HOC
NRTL:HOC
NRTL:HOC
NRTL:HOC
NRTL:HOC
NRTL:HOC

97,3b
83,4a
73,0a
63,3b
99,6a
99,5a
99,5a
99,5a
99,5
99,5
99,95

[36]
[37]
[38]
[39]
[40]
[12]
[41]
[42]
[5]
[13]
[14]

RD

Purolite

NRTL:HOC

99,5a

[43]

RD
RDWC

H2SO4
H2SO4 (at equilibrium)

NRTL:HOC
NRTL:RK

Pure
Pure
Pure
HAc:95,2 mol EtOH:82,2 mol

Pure
Pure
Pure
Pure
Pure
HAc:glacial 100wt
EtOH: p.a. grade 96 vol
Pure
Pure
HAc:49,63 mol EtOH48,08 mol H2O:2,29 mol
Pure
EtOH:88,17 mol HAc:100 mol
EtOH:87 mol HAc 95 mol
EtOH:87 mol HAc:95 mol
Pure
ETOH:95,2 mol HAc:82,2 mol
Pure
EtOH:86,98 wt
HAc:96,75 wt
EtOH:87 mol
HAc:95 mol
EtOH:82,2 mol HAc:95,2 mol
Pure

99,5
99,5

[29]
[6]

D: distillation, RD: reactive distillation, BRD: batch reactive distillation, SBRD: semi batch reactive distillation, RDWC:reactive distillation dividing wall, a; relative volatility.
a
Calculated from reported molar composition.
b
Calculated from molar purity assuming EtOH as other compound.
c
Calculated from transformed molar coordinate.

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

H2O

RC
HAC
DC
ETOH

H2O

Rx

ETAC
H2O
AD

Fig. 2. Conventional production processadapted from Kirk-Othmer [44] and completed (dotted line) with water-acid azeotropic distillation column for acid recovery and
recycle.

EtAc is the heavier component, and can be obtained as a bottom


product in a subsequent recovery column (RC). The distillate of this
column is returned to atmospheric condition and recycled to the
RD column. Unlike the previous alternatives, pressure swing
scheme has only two outlet streams, assuring total conversion and
complete use of reactants. For this reason, the bottom product of
the atmospheric RD column is nearly pure water. In general, this RD
conguration would be preferable among others, as both reaction
products are removed simultaneously from the reactive media
through different outlet streams.

Fig. 3. Reactive distillation processadapted from Tang et al. [5] (shaded region
corresponds to catalytic section).

the RD before entering the column. The distillate of the RD column


is introduced to a decanter where water is added to facilitate phase
splitting. The organic phase is partially reuxed to the RD and
partially driven to an additional stripper (RC) for nal purication
where EtAc is removed in the bottoms product. If cost-effective, the
aqueous phase produced in the decanter that contains mainly H2O
(82% wt) and roughly equal amount of alcohol and ester, is puried
to be recycled, otherwise treated before disposal. Because some
reactant and product are lost within the aqueous phase, excess
EtOH is required and EtAc recovery is reduced.
3.3.3. Pressure swing reactive distillation
As discussed in some studies [16,28], it is feasible to use a
pressure swing strategy to produce high purity EtAc running the
RD at atmospheric pressure and the recovery column at higher
pressure (350 kPa). The reported scheme is presented in Fig. 4.
While bottoms from RD are mainly water, composition of the
distillate approaches the ternary azeotrope. Using a pressure
swing, the azeotrope moves into a distillation region where pure

3.3.4. Dividing wall column with reactive reboiler


The use of a dividing wall column with a reacting reboiler
(DWC) to reduce energy consumption in the synthesis of EtAc has
been recently reported [6,33,48]. In this case, acid and alcohol are
fed into the reboiler of the DWC where the reaction takes place,
and a side stream is withdrawn to remove water. Though high
energy savings over the conventional RD column were claimed
(30%), the purity of the EtAc product remained below 75% wt
since no further recovery was implemented. In addition to the loss

Fig. 4. Pressure swing reactive distillation processadapted from Seferlis and


Grieving [16]

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

Fig. 5. Dividing wall column process with reactive reboileradapted from


Hernandez et al. [33] (dotted area) by adding EtAc rening column. (shadowed
condenser was replaced by a two phase decanter).

of raw materials within the product, the side stream from the main
column also contained large amount of HAc, which cannot be
recycled before water removal (difcult separation as mentioned
in Section 3.3.1).
To perform a fair comparison with other technologies, the
original DWC scheme [33] was coupled to a side stripper (RC) in
order to obtain the required high purity EtAc, as presented in Fig. 5.
As in the original report, the catalyzed reaction is assumed to take
place in the reboiler only. The distillate from the DWC is directed to
a decanter where water is added to induce phase splitting. A
fraction of the organic phase is reuxed to the DWC and the
remaining fraction is directed into a side stripper for purication.
Maintaining the original concept, a side stream is withdrawn from
the main column containing over 75% wt water.
3.3.5. novel proposalreactive dividing wall column
Taking into account the improvements reported for the
intensied schemes presented above, a novel process was
conceptually constructed. Analyzing the RD conguration presented by Tang et al. [5] and the DWC presented by Barroso-Muoz
et al. [6], it was expected that a conguration of a reactive dividing
wall column (RDWC) like the presented in Fig. 6a, would have
resulted in higher conversions and energy savings. This was
expected because the continuous removal of water would help
reducing the remixing effect into the reactive section. This kind of
technology is far from new, and one of the rst descriptions of a
RDWC was reported by Keibel in 1984 [49].
After a preliminary evaluation of a proposed RDWC scheme
described in Fig. 6a, large losses of EtAc were observed within the
aqueous stream leaving the overhead decanter (7% w/w). An
attempt to improve EtAc recovery led to the modied conguration
presented in Fig. 6b with water removal as a side stream. According
with a preliminary analysis of this design, EtAc losses were
minimized dramatically. However, there was either some HAc
losses in the side stream or an internal water recycle to the reboiler
and to the reactive stages, all resulting in lower conversions.
Finally, a new RDWC conguration presented in Fig. 6c was
proposed and the preliminary analysis indicated improved
performance compared with the previous ones. This RDWC
conguration was evaluated in further detail in this work, and it
consists of a reactive dividing wall column that uses two reboilers.
In a basic sense, the proposed scheme represents a main RD
column with a side stripper, both located inside the same shell. The

Fig. 6. Different proposed RDWC congurations. (a) Side stream to remove water.
Organic phase used as reux. (b) Side stream to remove water. Both organic
(partially) and aqueous (totally) phases reuxed back to the column. (c) Additional
reboiler (right) and additional bottoms stream (right) to remove water. Both organic
(partially) and aqueous (totally) phases reuxed back to the column.

dividing wall goes from the last stage up to the rectifying section of
the main column. As in the other schemes, this new conguration
also requires a decanter and an additional recovery column. This
scheme can be modeled as a thermally coupled reactive distillation
sequence with side stripper, so an additional reboiler was required.
In this case the catalyst is fed and contained within the main
reactive column where the appropriate catalytic kinetics was used.
In the other side of the wall, self-catalytic reaction can occur, and
the corresponding kinetic model was included. The organic phase
withdrawn from the decanter was partially reuxed to the dividing
wall column and the remaining fraction is sent to the recovery
column (RC) for further purication. The aqueous phase from the
decanter was directed to the non-reactive section of the dividing
wall column in order to remove pure water from the bottom

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

stream. The base case considered a number of reactive stages equal


to those for the reported RD conguration [5]. A preliminary
sensitivity analysis was carried out to determine the position of the
diving wall of this conguration.
3.4. Simulation and optimization
During simulations raw material feed streams and EtAc nal
product were specied according with technical datasheets
information [5052]. Such specications were selected to represent characteristics of commonly processed feedstock at industrial
facilities (Ethanol at 92.2 wt% and acetic acid at 98.5 wt%). Side
reactions and byproduct formation were neglected; however when
using sulfuric acid as catalyst, ethanol etherication and some
oxidation reactions might occur mainly at high temperatures.
Sulfuric acid was assumed to remain in the liquid phase, and
therefore its concentration in all reactive stages was dened as 1%
wt. The esterication reaction was considered in all sections of the
column: catalytic in the reactive stages and self-catalytic in the
non-reactive sections.
As there is no RDWC model within Aspen Plus software, these
schemes where simulated using two coupled distillation modules

Base Case

Base Case
Simulaon

Modify
Design

Base case meets


desired EtAC
purity?

No

Yes
M=Mo-Ones
i=1

with stream exchange between them, using RadFrac model


(equilibrium stage model). In the reported DWC, gas and liquid
exchange was implemented in the top and in the bottom of the
prefactionator column. In this new process, vapor and liquid
interconnection was only implemented in the top of the prefractionator. During simulating the conventional process [44], a
water-acid azeotropic distillation column was added for acid
recuperation. In the DWC simulation, a head decanter and a
recovery column were added to achieve the EtAc desired purity.
Raw materials, products, and fuel prices were obtained from
updated reports [5355], and were used for the calculation of TAC
(Eq. 1). Cooling water costs were considered negligible compared
with xed costs and heating costs, and heat integration was not
considered in this study. As the catalyst is homogeneous, bubble
cap tray type columns were used in the economic evaluation. In the
specic case of the RDWC, the diameter of column was computed
based upon the total transversal area in both sides of the dividing
wall, by adding the area required in the reactive side and that of the
side stripper. The cost of the dividing wall was consider negligible.
A complete description of the design specications of the
distillation columns from each conguration is listed in Table 4.
Both continuous and discrete variables were considered during
optimization of each conguration, and the owchart presented in
Fig. 7 describes the design-optimization algorithm. In the
owchart Nvars is the number of discrete variables, Mo is a
vector of Nvars size with the base case values, Ones is a vector of
Nvars size with the value 1 in each position, M is a vector of Nvars
size that holds the values of the discrete values to be evaluated. The
seed values of variables used to initiate the optimization process
were established according with the reports of the original papers
(where congurations were obtained from). The continuous
variables adjusted during optimization were: molar feed of EtOH,
molar feed of HAc, either total reux ratio (when condenser was
used) or organic phase reux ratio (when decanter was used),
bottoms molar ow rate, liquid and vapor molar ow rate across
the two sections in the dividing wall columns (when applicable),
reboiler heavy duty, water ow to decanter, decanter temperature
(when applicable), distillate to feed ow rate (when applicable).
Discrete variables evaluated in the sensitivity analysis were: feed
stage location, column number of stages and number of stages
above and below dividing wall (when applicable).
4. Results and discussion

j=-1

M(i)=Mo(i)+j
Minimize: TAC
Method: SQP
Constrains: Purity>99.5wt%

j>1?

No

j=j+1

Yes

i>Nvars?

No

i=i+1

Yes
End
Fig. 7. Design-optimization procedure for each case study.

After achieving convergence for all the different base-cases


using the operating conditions reported in literature, the
computational time required to achieve an optimal solution for
each conguration was c.a 24 h (using an average laptop). Main
results for the optimized conventional and intensied processes
are presented in Table 5, and the energy savings compared with the
traditional process are presented in Fig. 8. According with these
results heating duty is proportional to the cooling duty, so energy
costs might be calculated merely from either heating or cooling
duty as suggested in the methodology.
As reported in literature, the RDWC and DWC show high energy
savings with respect to the conventional process (Fig. 2) of 47% and
39% respectively. On the other hand, it is surprising to note that the
traditional RD scheme results in large columns with the highest
energy demand of all congurations (even larger than the
conventional process). Besides, the RD scheme needs a large
reux ratio in the main column compared with the recovery
column where most mass ow of streams is processed. This
behavior explains the poor sustainability indicators of the RD
conguration as we shall later analyze. All other intensied
congurations have in the recovery column a higher reux ratio
than in the main column.

10

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

Table 4
Design specications values and ranges for each conguration.
Design specication
(intervals*)
Column

RD

Conventional
DC

AD

0,01-4
Reux Ratio
Reboiler Duty (kW)
Reboiler Duty right side (kW)
Material
SS
Pressure (bar)
1
Bubble
Tray type
cap
1416
Number stages
Feed stage for reactive
1315
mixture
Feed stage for ethanol stream
Residence time (min)
2
D/F
0,60,8
B (kmol/h)
B Ratio

RD + PS

DWC

RDWC

RC

RD

RC

RD

RC

DWC

0,01-4

0,001-10**

N/A

0,001-10

0-10

0,5-1**

RC

RDWC

RC

0,001-4

N/A
05000

SS
1
Bubble
cap
1921
1416

CS
1
Bubble
cap
911
4

SS
1
Bubble
cap
2830
911

CS
1
Bubble
cap
810
1

SS
1
Bubble
cap
2224
57

CS
3,5
Bubble
cap
1517
79

SS
1
Bubble
cap
3941
Reb

CS
1
Bubble
cap
10
1

40220
SS
1
Bubble
cap
2830
911

CS
1
Bubble
cap
9
1

2
0,30,9

2
0,60,8

2628
2
0,91

2
0,60,7

1517
2
0,10,85

2
0,010,85

Reb
2
1

2729
2

2002000

050
4,374

Stages above wall


Stages below wall
Lateral stream stage
Lateral stream ow
L liq (kmol/h)
V vap

810
1012
1921
100200
170200
350380

13
0

6080

SS: stainless steel CS: carbon steel reb: reboiler.


Some design specications are not xed because an optimization ranges was used instead.
**
Reux Ratio from the organic phase leaving the decanter.
*

Fig. 8. Energy savings over conventional process.

Fixed, variable and total costs of evaluated processes are


summarized in Table 6. As it can be seen, more than 90% of the TAC
are concentrated in the raw materials while energy costs and xed
costs only represent c.a 5% and 1% respectively. From this
distribution it can be expected that the more cost effective
alternative to be the one with the higher productivity rather than

the one with the lower energy consumption or smaller equipment.


Therefore, RDWC, RD + PS, and RD congurations ended up with
high economic savings with respect to the conventional process,
mainly because of the enhanced conversion and the consequent
higher productivity promoted by the continuous displacement of
the equilibrium reaction and secondly due to the lower energy
consumption in the separations stages. Surprisingly and although
the DWC scheme achieves the high energy savings mentioned
above, the use of this processing alternative increases total costs
mainly due to the lower conversion. Raw material losses in the side
stream are the major cause of such inefcient performance. To
maintain the specied production rate under DWC conguration,
large feedstock consumption and large processing units are
required. As we shall analyze below, material losses do not only
affect the TAC, they also have a radical effect over many
sustainability indicators.
The complete set of results from the optimized processes can be
consulted in Figs. S3S7 in the supplementary material. The
sustainability indicators obtained under these optimal operating
conditions for the different congurations are summarized in
Table 7.
As reported before, energy consumption in most intensied
processes is substantially reduced, except in the case of the RD
conguration. This conguration requires a large reux ratio to
achieve high conversions. As the acid inlet is located at the top of

Table 5
Main characteristics of the optimized reactive distillation processes.
Process

EtOH
stream
(kg/h)

HAC
stream
(kg/h)

Heating
duty
(kW)

Cooling
duty
(kW)

Number of
stages main
column

Number of
reactive
stages

Reux ratio
Reux ratio
main column recovery
column

Fresh water
inlet (kg/h)

D (m)
main
column

D (m)
recovery
column

Conventional
RD
RD+PS
DWC
RDWC

10219
7465
7397
1357
6989

10079
8680
8601
15936
8541

26582
29213
19302
16217
14098

26664
29704
19203
16565
14155

15
29
24
40
29

1
18
18
1
19

0.93
5,14a
0.25
0,78
0,03

19309
1247

29575
1516

4,3
7,4
4,1
4,3
4,5

3,1
2,7
3,4
3,1
2,4

Reux Ratio from the organic phase leaving the decanter.

0.67
1,87
1.5
1.75
2,08

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

11

Table 6
Fixed, variable and total annual costs.
Process

Energy costs MM.USD/yr

Raw Materials costs MM.USD/yr

Variable costs MM.USD/yr

Fixed costs MM.USD/yr

TAC MM.USD/yr

Conventional
RD
RD + PS
DWC
RDWC

6,5
7,1
4,7
3,9
3,4

123,5
97,3
96,5
177,5
93,4

130
104,4
101,1
181,4
96,8

2,3
0,8
0,5
0,6
0,5

132,3
105,2
101,6
182
97,3

Table 7
Sustainability indicators of optimized reactive distillation processes for EtAc production.

Conventional
RD
RD + PS
DWC
RDWC

Conversion

Recovery

Productivity

MI

EW

MP

EI

0.98
1
0.98
0.71
0.99

0.86
0.97
1
0.75
1

0.84
0.97
0.98
0.53
0.99

1.58
1.26
1.25
2.3
1.21

2.23
0.42
0.31
3.82
0.39

0.34
0.05
0.04
1.02
0.01

0.63
0.79
0.8
0.44
0.83

2.17
2.38
1.69
1.32
1.15

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29

Catalytic stages

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9

Mole Fraction

the column. In Fig. 9, the composition proles of the most costeffective conguration (RDWC) are presented. While the most of
the HAc is maintained in the reactive side, EtOH is separated from
water in the side column and sent to the reactive stages assuring
complete conversion. Almost pure water is removed from bottoms
of the side stripper and the azeotrope is send to EtAc purication
column. In the modeling specications, it was established that
both the reactive and side columns had the same number of stages.
However, as observed in the composition proles of Fig. 9, the side
stripper is oversized. In this case, the use of different column
internals or a lower number of trays in the side column might
decrease even more the capital costs of the RDWC. It is interesting
to notice how the intensied congurations (RD, RD + PS and
RDWC) have similar conversion to that of the conventional scheme
but different productivity. The overall high conversion from the
conventional process is achieved by recycling almost all of the
limiting reagent of the process, which is the HAc from the AD
column. However, the low recovery capacity of the decanters in the
conventional scheme, where a large amount of fresh water is
needed to trigger phase separation explain the low productivity.
On the other hand, the productivity of DWC scheme is very low

Stage Number

Stage Number

the reactive zone of the column, some acid is pulled to the top and
lost within the distillate. To avoid such losses, and similarly to the
AD column of the conventional conguration (Fig. 2), a high reux
ratio of the organic phase is needed. With this, the EtAc rich stream
acts as an extracting agent that helps pushing the acid back to the
reactive zone improving conversion. This large reux explains why
RD presents the worst performance according with the EI indicator,
being even 10% more energy intensive than the conventional
process. Comparatively, and despite the fact that a higher pressure
is required, the RD+PS conguration requires a lower reux ratio
than the RD process (i.e., less energy), since the recycle stream
from the high pressure column helps keeping the acid content low
in the condenser from the reactive column. The herein proposed
RDWC conguration offers the lowest energy intensity and the
higher savings from all intensied processes since it avoids
remixing during multicomponent distillation, and also allows
selective product separation during esterication inside the same
unit. In this case, the low organic phase reux ow rate (see
Table 5) is possible due to the recycle of the aqueous stream from
the decanter back to the column, since it lowers the top stage
temperature and increases the actual amount of liquid reuxed to

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
0

0,1 0,2 0,3 0,4 0,5 0,6 0,7 0,8 0,9

Mole Fraction

Fig. 9. Liquid phase composition proles in a reactive dividing wall column conguration for ethyl acetate production using sulfuric acid as catalyst. Leftmain reactive
column, rightside stripper. () HAc, (--) EtOH, () EtAc, ( . . . ) H2O.

12

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113

mainly because of its low conversion. Which can be explained from


the DWC conguration, where the limiting reagent (HAc)
withdrawn within the side stream is never recycled back to the
reactive reboiler.
Regarding the MI factor for the different technologies evaluated
in this work, differences from RD, RD + PS and RDWC, can be
explained by the different excess reagent amounts that are fed to
each system. While RD and RD+PS required a 5% excess of ethanol,
the RDWC optimum solution uses nearly equimolar feed, which
combined with high productivity, makes mass utilization (MI) very
efcient. With respect to the E-Factor and E-W-Factor, DWC
produces more waste than the conventional process and other
intensied alternatives once again because of the unreacted HAc
leaving the process form the side stream, which in this case is
assumed as waste. Furthermore, the E-W factor shows that the
RDWC scheme generates 5 times less waste than other intensied
and protable congurations (RD and RD + PS) and 40 times less
waste than the conventional scheme. In this case the RDWC
produces extremely low waste amounts because this novel design
does not allow unreacted HAc nor EtOH to leave the process, as
well as the fact that there is no EtAc lost within the waste water
stream.
Although the RD + PS conguration has low E-factor because no
additional water is needed, the RDWC shows a better performance
because it uses the water obtained as product of the side column to
trigger phase separation in the decanter. This, together with the
lower TAC, low energy and materials intensity, makes RDWC
alternative to be the more attractive from an economical and
sustainable point of view.
5. Conclusions
An assessment of different intensied processes for the
production of ethyl acetate through Fisher esterication has been
carried out. The analysis allowed to select reliable thermodynamic
models to describe the multicomponent phase equilibria involved
in the process, as well as a homogeneous kinetic model that
correctly describes the reaction rate using sulfuric acid as catalyst.
The thermodynamic and kinetic models selected together with
commercial specications of raw materials and products were
implemented for the simulation of different intensied processes
proposed in the literature. Conventional process, reactive distillation, reactive distillation with pressure swing, dividing wall
column with reactive reboiler, and a novel process consisting of
a reactive dividing wall column, congurations were considered.
The optimization of those congurations was carried out by
minimizing total annual costs, using sensitivity analysis and SQP
approach. Finally, conversion, productivity, mass intensity, mass
productivity, E-Factor, EW-Factor, and energy intensity were
selected as sustainability indicators of the technologies evaluated.
According with results, reactive distillation offers lower costs and
better sustainability although having higher energy intensity than
the conventional process. DWC although less energy intensive, has
the higher costs and the worst sustainability indicators. Reactive
distillation with pressure swing offers energy savings, lower costs
and better sustainability than the conventional process. Amongst
all the evaluated alternatives, the novel reactive dividing wall
column conguration developed here offers the highest energy
savings, the lower annual costs, and the overall better sustainability indicators.
Acknowledgement
This work is supported by Colombian Administrative Department of Science, Technology and Innovation (COLCIENCIAS) under
the project: Produccin de plasticantes a partir de acido ctrico

usando procesos hbridos de reaccin y separacin simultanea


cod. 1101-569-33201.
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.cep.2015.07.027.
References
[1] J.P.M. Sanders, J.H. Clark, G.J. Harmsen, H.J. Heeres, J.J. Heijnen, S.R.a. Kersten,
et al., Process intensication in the future production of base chemicals from
biomass, Chem. Eng. Process. Process Intensif. 51 (2012) 117136, doi:http://
dx.doi.org/10.1016/j.cep.2011.08.007.
[2] R. Srinivasan, N.T. Nhan, A statistical approach for evaluating inherent benignness of chemical process routes in early design stages, Process Saf. Environ.
Prot. 86 (2008) 163174, doi:http://dx.doi.org/10.1016/j.psep.2007.10.011.
[3] N.T.H. Thuy, Y. Kikuchi, H. Sugiyama, M. Noda, M. Hirao, Techno-economic and
environmental assessment of bioethanol-based chemical process: a case study
on ethyl acetate, Environ. Prog. Sustain. Energy 30 (2011) 675684, doi:http://
dx.doi.org/10.1002/ep.
[4] T. Van Gerven, A. Stankiewicz, Structure, energy, synergy, timesthe
fundamentals of process intensication, Ind. Eng. Chem. (2009) 24652474.
[5] Y. Tang, H.-P. Huang, I.-L. Chien, Design of a complete ethyl acetate reactive
distillation system, J. Chem. Eng. Jpn. 36 (2003) 13521363.
[6] F.O. Barroso-Muoz, S. Hernndez, J.G. Segovia-Hernndez, H. HernandezEscoto, A.F. Aguilera-Alvarado, Thermally coupled distillation systems: study
of an energy-efcient reactive case, Chem. Biochem. Eng. Q. 21 (2007)
115120.
[7] L. Kelley, F. Mirasol, Ethyl acetate, ICIS Chem. Bus. 281 (2012) 4344.
[8] J.M. Douglas, DouglasConceptual Design Of Chemical Processes, McGrawHill, 1988.
[9] R.a Sheldon, Atom utilisation, E factors and the catalytic solution, Comptes
Rendus lAcadmie Des. Sci. - Ser. IIC - Chem. 3 (2000) 541551, doi:http://dx.
doi.org/10.1016/S1387-1609(00)01174-9.
[10] C. Jimenez-Gonzalez, D.J.C. Constable, Green Chemistry and Engineering: A
Practical Design Approach, Wiley, New Jersey, 2011.
[11] J. Gmehling, J. Menke, J. Krafczyk, K. Fischer, Azeotropic Data, 2nd ed., WileyVCH, Weinheim, 2004.
[12] H. Huang, I. Chien, H. Lee, Plantwide Control of a Reactive Distillation Process,
in: G.P. Rangaiah, V. Kariwala (Eds.), Plantwide Control Recent Dev. Appl, First
John Wiley & Sons Inc., Taipei, 2012, pp. 319338.
[13] H. Tian, H. Zheng, Z. Huang, T. Qiu, Y. Wu, Novel procedure for coproduction of
ethyl acetate and n-butyl acetate by reactive distillation, Ind. Eng. Chem. Res.
51 (2012) 55355541, doi:http://dx.doi.org/10.1021/ie202154x.
[14] I.-K. Lai, Y.-C. Liu, C.-C. Yu, M.-J. Lee, H.-P. Huang, Production of high-purity
ethyl acetate using reactive distillation: experimental and start-up procedure,
Chem. Eng. Process. Process Intensif. 47 (2008) 18311843, doi:http://dx.doi.
org/10.1016/j.cep.2007.10.008.
[15] W. Riemenschneider, H.M. Bolt, Esters, Organic, Ullmanns Encycl. Ind. Chem.
(2005) 86768694, doi:http://dx.doi.org/10.1002/14356007.a09.
[16] P. Seferlis, J. Grievink, Optimal design and sensitivity analysis of reactive
distillation units using collocation models, Ind. Eng. Chem. Res. 40 (2001)
16731685, doi:http://dx.doi.org/10.1021/ie0005093.
[17] A. Orjuela, A.J. Yanez, A. Santhanakrishnan, C.T. Lira, D.J. Miller, Kinetics of
mixed succinic acid/acetic acid esterication with Amberlyst 70 ion exchange
resin as catalyst, Chem. Eng. J. 188 (2012) 98107, doi:http://dx.doi.org/
10.1016/j.cej.2012.01.103.
[18] L.S. Balasubramhanya, F.J. Doyle III, Nonlinear model-based control of a batch
reactive distillation column, J. Process Control 10 (2000) 209218, doi:http://
dx.doi.org/10.1016/S0959-1524(99)00024-4.
[19] F.S. Atalay, Kinetics of the esterication reaction between ethanol and acetic
acid, Dev. Chem. Eng. Miner. Process 2 (1994) 181184, doi:http://dx.doi.org/
10.1002/apj.5500020210.
[20] K. Tanaka, R. Yoshikawa, C. Ying, H. Kita, K. Okamoto, Application of zeolite
membranes to esterication reactions, Catal. Today 67 (2001) 121125, doi:
http://dx.doi.org/10.1016/S0920-5861(01)00271-1.
[21] J.M. Smith, Chemical Engineering Kinetics, McGraw-Hill, 1956.
[22] S.I. Kirbaslar, Z.B. Baykal, U. Dramur, Esterication of acetic acid with ethanol
catalysed by an acidic ion-exchange resin, Turk. J. Eng. Environ. Sci. 25 (2001)
569577.
[23] K. Alejski, F. Duprat, Dynamic simulation of the multicomponent reactive
distillation, Chem. Eng. Sci. 51 (1996) 42254237.
[24] A. Orjuela, A.J. Yanez, C.T. Lira, D.J. Miller, Ethyl Acetate Esterication Kinetics
Using Amberslyst 15 as Catalyst, Internal Research Report, Michigan State
University, 2014.
[25] G. Hangx, G., Kwant, H., Maessen, P., Markusse, I., Urseanu, P. Moritz, et al.
REACTION KINETICS OF THE ESTERIFICATION OF ETHANOL AND ACETIC ACID
TOWARDS ETHYL ACETATE Name of Partner: DSM Authors: Intelligent Column
Internals for Reactive Separations (INTINT), (2001) 15.
[26] H.J. Arnikar, T.S. Dao, A.A. Bodhe, A gas chromatographic study of the kinetics
of the uncatalysed esterication of acetic acid by ethanol, J. Chromatogr. A 47
(1970) 265268.

M.A. Santaella et al. / Chemical Engineering and Processing 96 (2015) 113


[27] P.V.S.R. Chandra, C. Venkateswarlu, Multistep model predictive control of ethyl
acetate reactive distillation column, Indian J. Chem. Technol. 14 (2007)
333340.
[28] S.V. Mali, A.K. Jana, A partially heat integrated reactive distillation: feasibility
and analysis, Sep. Purif. Technol. 70 (2009) 136139, doi:http://dx.doi.org/
10.1016/j.seppur.2009.09.006.
[29] H.-Y. Lee, H.-P. Huang, I.-L. Chien, Control of reactive distillation process for
production of ethyl acetate, J. Process Control 17 (2007) 363377, doi:http://
dx.doi.org/10.1016/j.jprocont.2006.10.002.
[30] A. Avami, W. Marquardt, Y. Saboohi, K. Kraemer, Shortcut design of reactive
distillation columns, Chem. Eng. Sci. 71 (2012) 166177, doi:http://dx.doi.org/
10.1016/j.ces.2011.12.021.
[31] M. Klker, E.Y. Kenig, A. Hoffmann, P. Kreis, A. Grak, Rate-based modelling and
simulation of reactive separations in gas/vapour-liquid systems, Chem. Eng.
Process. Process Intensif. 44 (2005) 617629, doi:http://dx.doi.org/10.1016/j.
cep.2003.12.011.
[32] S.-J. Wang, H.-P. Huang, C.-C. Yu, Plantwide design of transesterication
reactive distillation to co-generate ethyl acetate and n-butanol, Ind. Eng.
Chem. Res. 49 (2010) 750760, doi:http://dx.doi.org/10.1021/ie901413c.
[33] S. Hernndez, R. Sandoval-Vergara, F.O. Barroso-Muoz, R. Murrieta-Dueas,
H. Hernndez-Escoto, J.G. Segovia-Hernndez, et al., Reactive dividing wall
distillation columns: simulation and implementation in a pilot plant, Chem.
Eng. Process. Process Intensif. 48 (2009) 250258, doi:http://dx.doi.org/
10.1016/j.cep.2008.03.015.
[34] E.Y. Kenig, H. Bader, A. G, Investigation of ethyl acetate reactive distillation
process, Chem. Eng. Sci. 56 (2001) 61856193.
[35] Q. Smejkal, J. Kolena, J. Hanika, Ethyl acetate synthesis by coupling of xed-bed
reactor and reactive distillation columnprocess integration aspects, Chem.
Eng. J. 154 (2009) 236240, doi:http://dx.doi.org/10.1016/j.cej.2009.04.022.
[36] E. Re, M. Meyer, A generic feasibility study of batch reactive distillation in
hybrid congurations, AIChE J. 55 (2009) 11851199, doi:http://dx.doi.org/
10.1002/aic.
[37] R. Delgado-Delgado, S. Hernndez, F.O. Barroso-Muoz, J.G. SegoviaHernndez, A.J. Castro-Montoya, From simulation studies to experimental
tests in a reactive dividing wall distillation column, Chem. Eng. Res. Des. 90
(2012) 855862, doi:http://dx.doi.org/10.1016/j.cherd.2011.10.019.
[38] H. Bock, M. Jimoh, G. Wozny, Analysis of reactive distillation using the
esterication of acetic acid as an example, Chem. Eng. Technol. 20 (1997) 182
191, doi:http://dx.doi.org/10.1002/ceat.270200305.
[39] Z.M. Shakor, K.A. Sukkar, Dynamic simulation of semi-batch catalytic
distillation used for estercation reaction, Eng. Technol. (2008) 26.

13

[40] Y. Tavan, S.H. Hosseini, Design and simulation of a reactive distillation process
to produce high-purity ethyl acetate, J. Taiwan Inst. Chem. Eng. 44 (2013) 577
585, doi:http://dx.doi.org/10.1016/j.jtice.2012.12.023.
[41] H. Lee, H., Huang, I. Chien, Design and control of homogeneous and
heterogeneous reactive distillation for ethyl acetate process, 16th Eur. Symp.
Comput. Aided Process Eng. 9th Int. Symp. Process Syst. Eng. (2006) 1045
1050.16th European Symposium on Computer Aided Process Engineering and
9th International Symposium on Process Systems Engineering.
[42] Y.-T. Tang, Y.-W. Chen, H.-P. Huang, C.-C. Yu, S.-B. Hung, M.-J. Lee, Design of
reactive distillations for acetic acid esterication, AIChE J. 51 (2005) 1683
1699, doi:http://dx.doi.org/10.1002/aic.10519.
[43] I.-K. Lai, S.-B. Hung, W.-J. Hung, C.-C. Yu, M.-J. Lee, H.-P. Huang, Design and
control of reactive distillation for ethyl and isopropyl acetates production with
azeotropic feeds, Chem. Eng. Sci. 62 (2007) 878898, doi:http://dx.doi.org/
10.1016/j.ces.2006.10.019.
[44] Kirk-Othmer, Encyclopedia of Chemical Technology, 8 (n.d.) 331338.
[45] H. Zheng, H. Tian, W. Zou, Z. Huang, X. Wang, T. Qiu, et al., Residue curve maps
of ethyl acetate synthesis reaction, J. Cent. South Univ. 20 (2013) 5055, doi:
http://dx.doi.org/10.1007/s11771-013-1458-2.
[46] S. Hung, M. Lee, Y. Tang, Y. Chen, I. Lai, W. Hung, et al., Control of different
reactive distillation, AIChE J 52 (2006) 14231440, doi:http://dx.doi.org/
10.1002/aic.
[47] S.-J. Wang, D.S.H. Wong, S.-W. Yu, Design and control of transesterication
reactive distillation with thermal coupling, Comput. Chem. Eng. 32 (2008)
30303037, doi:http://dx.doi.org/10.1016/j.compchemeng.2008.04.001.
[48] F.O. Barroso-muoz, S. Hernndez, B. Ogunnaike, Analysis of Design and
Control of Reactive Thermally Coupled Distillation Sequences, 17th Eur, Symp.
Comput. Aided Process Eng. - ESCAPE 17 (2007) 16.
[49] G. Kaibel, Method of carrying out chemical reactions and for the simultaneous
fractionation of a mixture into several fractions by a distillation column,
EP126288 A2, 1984.
[50] Eastman Chemical Company, Eastman Ethyl Acetate, Urethane Grade Product
Data Sheet, Eastman Chemical Company, 2001.
[51] Eastman Chemical Company, Eastman Dilute Acetic Acid, 95% Product Data
Sheet, Eastman Chemical Company, 2000.
[52] Recochem Inc, Ethanol 95 SG (SVR 95%), Recochem Inc, 2013. http://www.
recochem.com.au/index.php/products/industrial_products/ethanols/item/
ethanol_95_sg_svr_95
[53] F. Mirasol, Ethanol, ICIS Chem. Bus. 280 (2011) 71.
[54] E. Burridge, Chemical Prole: Acetic acid, ICIS Chem. Bus. May 31- Ju (2010) 36.
[55] G.D. Ulrich, P.T. Vasudevan, How to estimate utility costs, Chem. Eng. 6 (2006)
6669.

Potrebbero piacerti anche