Chemical Engineering Science 59 (2004) 5839 – 5845

www.elsevier.com/locate/ces

Analysis of the environmental impact of butylacetate process through

the WAR algorithm
C.A. Cardonaa,∗ , V.F. Marulandaa , D. Youngb
a National University of Colombia at Manizales. Cra. 27 No. 64-60 Of. F-505, Caldas, Columbia
b National Risk Management Research Laboratory, EPA. Cincinnati, OH 45268, USA

Received 5 July 2003; received in revised form 17 May 2004; accepted 30 June 2004
Available online 2 September 2004

Abstract
The WAR algorithm, a methodology for determining the potential environmental impact (PEI) of a chemical process, is applied to
the case study of n-butyl acetate production. Two different schemes are considered: the conventional process, in which the reaction and
separation are carried out in separated devices, and the reactive distillation process in which the reaction and separation are carried out
in a packed column simultaneously. The WAR algorithm allows determining which process is better from the point of view of generation
of PEI into the process, based on the output wastes and energetic consumption.
䉷 2004 Elsevier Ltd. All rights reserved.

Keywords: Reaction engineering; Environment; Product design; Pollution control; WAR

1. Introduction 2. Esterification of acetic acid with n-butanol

Reactive distillation has been used as a “clean technology” We study the esterification of acetic acid with n-butanol
for chemical synthesis in the industry because of advan- to produce n-butylacetate.
tages like saving of energy cost and reduced capital invest-
H+
ment. Comparison of the reactive distillation process with HAc + BuOH ⇔ BuAc + H2 O.
conventional process are poorly mentioned in the literature
(Solokhin et al., 1997; Paiva and Malcata, 2000; Cardona and This reversible reaction needs to be catalyzed by strong
Cubides, 2003). All these comparisons are based on energy acids. Sulfuric acid, p-toluenesulfonic acid or solid acidic
consumption. We focus our work to an integral comparison catalysts are usually used for this purpose (Zhicai et al.,
from the point of view of outlet and generation of poten- 1998). The influence of temperature of the equilibrium
tial environmental impact (PEI) from the process based on constant for the production of butylacetate is shown in
output wastes and energy consumption. The combination of Fig. 1.
shortcut methods, rigorous calculations and the subsequent Experimental values of the equilibrium constant are re-
modeling and simulation of generated PEI make it possible ported by Loning et al. (2000). Fig. 1 confirms a good agree-
to assess the real impact of reactive distillation on butyl ac- ment between the theoretical and experimental results.
etate production. Because of the stringent purity requirements on n-butyl
acetate (less than 50 ppm acetic acid) the reaction mixture
should be fed with a little excess of n-butanol, so that the
∗ Corresponding author. Tel.: +57-68-866-766; fax: +57-68-863-220. acetic acid becomes limit reactive (Venimadhavan et al.,
E-mail address: [email protected] 1999). The equilibrium conversion can be found by means
(C.A. Cardona). of the thermodynamic relation between the equilibrium
0009-2509/$ - see front matter 䉷 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ces.2004.06.043
5840 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845

3. Reaction kinetics of n-Butylacetate process

The reaction is heterogeneously catalyzed, for such a re-

action; Loning et al. (2000) report the following parameters
for a Langmuir–Hinshelwood–Hougen–Watson (LHHW)
model:
aBuOH · aAcAc − 1
ka · aBuAc · aH2 O
r = khet , (3)
1 + KAcAc · aAcAc + KH2 O · aH2 O

where khet = 32 eq·s

mol
(reaction rate constant based on active
sites), KAcAc = 9 (sorption equilibrium constant of acetic
acid), and KH2 O = 1.4(sorption equilibrium constant of wa-
ter) and ai is the activity of component Ai in the liquid phase.
The parameters were determined for ambient pressure and
a range of temperature between 90 and 120 ◦ C. Amberlyst
15 was used as the catalyst.

Fig. 1. Equilibrium constant vs. temperature.

4. Vapor–liquid–liquid equilibrium

In the reaction mixture there are four binary azeotropes

and two ternary azeotropes of minimum boiling point. The
binary azeotropes water–butanol and water–butylacetate and
the ternary azeotrope butanol–butylacetate–water are hetero-
geneous (Table 1).
The system contains two stable nodes, the two highest
boiling points (n-butylacetate and azeotrope 6) and one un-
stable node, the lowest boiling azeotrope (Azeotrope. 1).
Then, one can conclude that the residue curves end at the
highest (stable node) and begin at the lowest boiling compo-
nent (unstable node) for every degree of conversion reached
in the equilibrium in Fig. 2, this means that the conversion
reached in the reactor defines the products which will be
obtained in the distillation column. The theoretical residue
curve map is shown in Fig. 3.

5. Conventional process
Fig. 2. Equilibrium conversion in the butylacetate system.
According to the state of art in conventional process de-
velopment (Hanika et al., 1999; Levering et al., 1999), a
constant and the activities of each chemical involved in the butanol–acetic acid mixture (49% mol acetic acid) is fed
reaction. to a PBR reactor which is packed with strongly acidic ion


exchange resin (Hanika et al., 1999). The inlet temperature
K= ai i , (1) is maintained at 90 ◦ C, which provides the optimal condi-
i tions for the resin. Because of the exothermic reaction, the
outlet temperature is 110 ◦ C, low enough to avoid the deac-
ai = i · xi . (2) tivation of the resin. The equilibrium mixture of the reac-
tor is separated in a distillation column to obtain a mixture
For the calculation of the activity coefficients the UNIFAC with a composition near to ternary azeotrope n-butanol-n-
model was used. butylacetate–water. This heterogeneous azeotrope condenses
With a feed comprising of 49% acetic acid and 51% bu- into a water- and organic-rich phase that contains ∼56% mol
tanol, the equilibrium conversion is shown in Fig. 2. of n-n-butyl acetate, the remnant being a mixture of water
The conversion does not vary strongly with temperature as and n-butanol, as is shown in Table 2.
shown in Fig. 2. It only decreased by 0.09 over a temperature The amount of the reaction water is not sufficient to dis-
range of 335–420 K. till all n-butyl acetate in the mentioned form; in addition, it
 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845 5841

Table 1
Singular points in the butylacetate system (Karpilovski, 1997)

Name Composition T [◦ C] xBuOH xHAc xBuAc

Azeotrope 1 BuOH/BuAc/H2 O 90.66 0.0895 0 0.2041

Azeotrope 2 BuAc/H2 O 91.19 0 0 0.2766
Azeotrope 3 BuOH/H2 O 92.96 0.2334 0 0
Water H2 O 100 0 0 0
Azeotrope 4 BuOH/BuAc 116.15 0.7004 0 0.2996
Ácetic acid HAc 117.77 0 1 0
n-butanol BuOH 117.97 1 0 0
Azeotrope 5 BuOH/BuAc/HAc 121.3 0.2 0.46 0.34
Azeotrope 6 BuOH/HAc 122.69 0.5161 0.4839 0
n-butylacetate BuAc 126.17 0 0 1

This overhead product is recycled by means of a separa-

tion train comprised of two distillation columns in which
the formation of the heterogeneous azeotrope between wa-
ter and n-butanol is used to accomplish the recuperation
and recycling of n-butanol to the feed of the system, while
n-butyl acetate of purity usually up to 98 % mass is with-
drawn as a bottom product (Fig. 6).
The principal disadvantages of the conventional process
are the high energetic consumption of the product distilla-
tion column because of the additional water added and the
complexity of purifying the reactants and products in the
separation system because of the strongly non-ideality of the
system. A general flowsheet for the conventional process is
shown in Fig. 6.
The specifications of the feed, effluent streams and the en-
ergetic consumption for the conventional process are shown
in Table 3. These results were obtained in Hysym (Hypro-
tech).
The energetic consumption for the conventional process
is calculated by summing the duties of the columns in the
whole process described in Fig. 7. The water phase from
the decanter is practically pure water, so it can be treated
Fig. 3. Residue curve map for the system water, n-BuOH and n-BuAc.
as an outlet waste. In the second column is easy obtain-
ing n-butyl acetate as a product stream 11, meanwhile, the
top mixture, which contains n-butanol, n-butyl acetate and
is almost impossible obtaining n-butyl acetate from the bot- water is purified in a distillation system comprised of two
tom mixture of the distillation column of the first column columns.
because of the strong non-ideality of the system as shown
in the residue curve maps of Fig. 3, therefore additional wa-
ter must be added. The volume of waste water to be subse- 6. Reactive distillation process
quently treated is thus increased, which is one of the main
disadvantages of the conventional process (Fig. 4). The flow sheet for the reactive distillation process is
The simulations were carried out in Hysym (Hyprotech). shown in Fig. 7 (Karpilovski, 1997). The gray section in
As shown in Fig. 5, it is possible to obtain an organic phase Fig. 7 represents the reactive section of the packed col-
from the decanter with 55% mol n-BuAc as almost pure umn with an acidic resin similar to Amberlyst 15. The
water is withdrawn as an outlet waste. localization of the reactive section is done by means of an
Separation of the organic phase from the water phase in analysis of statics (Cardona, 2001; Pisarenko et al., 2001).
the heterogeneous azeotrope gives conditions for subsequent The conversions of n-butanol and acetic acid are 99.64 and
separation of n-butyl acetate from n-butanol by distillation 96.83, respectively. The column removes as distillate the
of the organic phase in another distillation column from heterogeneous binary azeotrope between n-butyl acetate
which a mixture containing n-butanol, dilute water and a and water, that is, additional water must be added to the
small amount of n-butyl acetate is withdrawn as distillate. process (Epifanova, 1988).
5842 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845

Table 2
Mole fractions of the heterogeneous azeotrope containing n-butyl acetate n-butanol and water (Karpilovski,
1997; Venimadhavan et al., 1999)

Component Overall composition (mol/mol) Aqueous phase (mol/mol) Organic phase (mol/mol)

Water 0.712 0.995 0.232

n-BuOH 0.085 0.004 0.206
n-BuAc 0.203 0.001 0.562

The principal disadvantages of the reactive distillation

process are the high energetic consumption because of the
additional water added and the high reflux numbers (5–8).
The energetic consumption of the reactive distillation pro-
cess, as same as the conventional process, is based on the
energetic consumption of the reboiler of the reactive dis-
tillation column, which is high because of the high reflux
numbers in the condenser. Karpilovski (1997) reports a con-
sumption of 431507 KJ/Kg mol n-BuAc.

7. Evaluation of potential environmental impact of

both process

The PEI of a given quantity of material and energy can be

Fig. 4. Conventional Process for producing butylacetate. generally defined as the effect that this material and energy
would have on the environment if they were to be emitted
into the environment (Young and Cabezas, 1999; Mallick
et al., 1999) . The WAR algorithm is a tool to perform the
206.896 ibmol/h organic phase
HAc 0.0004 76.1473 ibmol/h
HAc 0.0009
evaluation of the PEI of a process based on the product
BuAc 0.203
BuOH 0.085
Water 0.712
BuAc 0.55
BuOH 0.22 streams, outlet wastes and the feed to the process.
150 ibmol/h Water 0.23
HAc 0.21
BuAc 0.23
The software used to evaluate the PEI of both processes
BuOH 0.28
T=109.85˚C is WARGUI (waste reduction algorithm graphical user inter-
T=15˚C
water phase face) from the Environmental Protection Agency. Two dif-
Water 105.3 ibmol/h 130.749 ibmol/h
column T=30˚C HAc 0.00001 ferent analyses can be performed using the WAR algorithm,
BuAc 0.000296
R=3
stages=50 bottom product
BuOH 0.00392
Water 0.995
product and non-product analysis, depending on whether the
feed=stage 30
water=stage 10
48.4146 ibmol/h
HAc 0.6505
product stream of n-butyl acetate is included in the analysis
BuAc 0
BuOH 0.35 (Young and Cabezas, 1999).
Water 0
The overall PEI of a chemical is determined by summing
Fig. 5. Simulation of the product column from the reactor.
the specific PEI of the chemical over all the possible impact
categories (Eq. (4)).

k = l kl (4)
organic phase
76.1473 ibmol/h l
HAc 0.000009
BuAc 0.55
BuOH 0.22 kl represents every impact category and l is the weighting
Water 0.23
factor, which is used to emphasize the particular areas of
concern. These categories fall into two general areas of con-
Top product cern with four categories in each area: global atmospheric
T=15˚C HAc 0.000001
BuAc 0.043 and local toxicological. The four global atmospheric im-
BuOH 0.4677
Water 105.3 ibmol/h Water 0.489 pact categories are global warming potential (GWP), ozone
T=30˚C T=17.16˚C
depletion potential (ODP), acidification or acid-rain poten-
Bottom product
48.4146 ibmol/h Product
HAc 0
tial (AP), and photochemical oxidation or smog formation
HAc 0.6505
BuAc 0 BuAc 0.983
BuOH 0.017
potential (PCOP). The four local toxicological impact cat-
BuOH 0.35
Water 0 Water 0
T=125.09˚C
egories are human toxicity potential by ingestion (HTPI),
human toxicity potential by either inhalation or dermal
Fig. 6. Distillation column—organic phase. exposure (HTPE), aquatic toxicity potential (ATP), and
 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845 5843

Table 3
Specifications of the conventional process

Butanol (1) Acetic (2) water (26) Water phase (9) Product (11) waste (14)

Flow (Kg/h) 1419.79 1128.97 859.75 1090.544 2154.7202 156.3216

Mass fractions
BuOH 1 0 0 0.0011932 0.010915 0
Hac 0 1 0 0.00003266 0 0.00002683
BuAc 0 0 0 0.0247634 0.98908 0.125723
Water 0 0 1 0.9740107 0 0.874

Energetic consumption:18172.493 MJ/h

BuOH 0.001 Total output rate of PEI*

HAc 0.009
BuAc 0.324 14000
Water 0.666 organic phase
T=92˚C 41.2755 ibmol/h 12000
BuAc 0.98

Impact (PEI/hr)
HAc 0.017 10000
BuOH 0.003
8000
water 6000
water phase
42.69722 ibmol/h
83.57018 ibmol/h
Water 0.995 4000
HAc 0.00745 2000
BuOH 0.000027
0
HTPI HTPE TIP ATP GWP ODP PCOP AP TOTAL**
HAc Impact Categories
41.77424 ibmol/h
R=5 - 8 Convensional Reactive distillation
BuOH 0.096 P=0.33 atm
BuOH
HAc 0.207 stages=40
40.596 ibmol/h
BuAc 0.691 Fig. 8. PEI—product analysis.
Water 0.006
T=115˚C

Fig. 7. Reactive distillation process for the production of n-butylacetate Total output rate of PEI*
[4]. 450
400
350
Impact (PEI/hr)

300
terrestrial toxicity potential (TTP) (Young and Cabezas, 250
1999; Mallick et al., 1999). 200
150
100
50
8. Analysis of results 0
HTPI HTPE TIP ATP GWP ODP PCOP AP TOTAL**
Impact Categories
The total output rate of environmental impact for both Convensional Reactive distillation
process is shown in Fig. 8, which corresponds to the anal-
ysis of products. The total output rate of PEI allows one to Fig. 9. Total output rate of PEI—non-products analysis.
compare alternative processes in terms of the potential effect
on the environment external to the process. The total output
rate of PEI is more useful in assessing whether a particular est value for this category; in fact, it has a value of 6 for
site is able to accommodate a given process plant (Young this category. The other chemicals like n-butanol and acetic
and Cabezas, 1999). acid have a value of zero. As a result, the PEI/h is almost
From Fig. 8 is possible to see that the PEI/h for both pro- the unique contribution of the PCOP category because of
cesses is almost the same, 13400 PEI/h for the conventional the stream products.
and 13100 PEI/h for the reactive distillation process. These The non-products analysis is shown in Fig. 9. From this
values do not allow comparing both processes from the point picture it is possible to accomplish a better analysis of the en-
of view of the PEI since the methods used to calculate the vironmental performance of both processes. The total PEI/h
categories of PEI are empirical and can vary depending on for the conventional process is 406 PEI/h vs. 56 PEI/h for
the chosen method. The category of PEI that has the highest the reactive distillation process. That means the conventional
value is the photochemical oxidation potential PCOP. This process releases 7 times more contamination to the envi-
is consistent with the fact that n-butyl acetate has the high- ronment than the reactive distillation process. The highest
5844 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845

Total generation rate of PEI* Total generation rate of PEI*

400 14000
200
12000
0
10000
Impact (PEI/hr)

-200

Impact (PEI/hr)
-400 8000
-600 6000
-800
4000
-1000
-1200 2000
-1400 0
-1600 -2000
HTPI HTPE TIP ATP GWP ODP PCOP AP TOTAL** HTPI HTPE TIP ATP GWP ODP PCOP AP TOTAL**
Impact Categories Impact Categories
Convensional Reactive distillation Convensional Reactive distillation

Fig. 10. Total generation rate of PEI—non-product analysis. Fig. 11. Total generation rate of PEI—product analysis.

value for the PCOP for the conventional process is because The product analysis for the generation of PEI is kind
of the n-butyl acetate that is leaving the system in the stream of different as shown in Fig. 11. There are two contradic-
wastes from the decanter and the recuperation column of tory effects: the consumption of PEI in the HTPI and TTP
water of the conventional process. Although a considerable categories and the generation of PEI because of the high
amount of acetic acid is released in the reactive distillation PCOP for the n-butyl acetate. That means the total gener-
process, 37.8 Kg/h, the contribution of this chemical to the ation of PEI is less than the PCOP category, which could
impact categories, by comparison with the n-butyl acetate give a wrong interpretation about both process. In this case
in the output waste is despicable. the real dimension of the PEI must be considered like the
The product analysis shown in Fig. 8 hides this effect be- PCOP category.
cause of the inclusion of the stream products. It shows a
high value of the acidification potential for the conventional
process, 109 PEI/h vs. 47 PEI/h for the reactive distillation
process in Fig. 9. This PEI is because of the energetic con- 9. Conclusions
sumption of the processes and it is consistent with the val-
ues reported for the flowsheets, 18,000 MJ/h for the con- The total rate of PEI for the non-products analysis allows
ventional process vs. 8000 MJ/h for the reactive distillation determining the real dimension of the environmental perfor-
process. mance of reactive distillation processes when these are com-
The generation of PEI for both processes, non-products pared to the conventional processes. Indexes like the total
analysis, is shown in Fig. 10. The total generation rate of generation rate of PEI for the non-product analysis shows
PEI allows comparing different processes in terms of their that the reactive distillation process improves conversion of
generation of new PEI within the process. The generation reactants, which is shown for a total generation index of
rate of PEI is important because once it is created; it will around −1400 PEI/h vs. −1200 PEI/h for the conventional
likely take resources to keep the PEI from becoming actual process.
impacts in the environment. The total generation rate of PEI
allows one to determine if the chemical reactions occurring
within the process produce products that are less environ- Notations
mentally friendly that the feed stock chemicals.
For both processes, the categories HTPI(human toxicity AP acidification potential
potential by ingestion) and TTP (Terrestrial toxicity poten- ATP aquatic toxicity potential
tial) oscillate around −700 PEI/h, that is it, the process is BuAc n-butyl acetate
carried out towards the consumption of PEI when the stream BuOH n-butanol
products is not included in the analysis. The total genera- GWP global warming potential
tion rate of PEI/h is more negative for the reactive distilla- HTPE human toxicity potential by exposure
tion process, what means the RD process consumes more HTPI human toxicity potential by ingestion
PEI/h than the conventional process. For the conventional K equilibrium constant
process, categories like the PCOP (Photochemical oxidation Ka equilibrium constant
potential) and AP (Acidification potential) generation make Khet reaction rate constant based on active sites
it unsuitable for regions with a high level of atmospheric ODP ozone depletion potential
pollution (greenhouse effect) or acid rains. PCOP photochemical oxidation potential
 C.A. Cardona et al. / Chemical Engineering Science 59 (2004) 5839 – 5845 5845

PEI potential environmental impact Epifanova, O.A., 1988. Steady states of continuous combined reactive
TTP terrestrial toxicity potential distillation processes (n-butyl acetate technology). Ph. D. Dissertation,

i stoichiometric coefficient of component i Moscow Institute of Fine Chemical Technology, Moscow.
Hanika, J., Kolena, J., Smejkal, Q., 1999. n-Butyl acetate via reactive
WAR waste reduction algorithm
distillation modelling and experiment. Chemical Engineering Science
xi mole fraction of component i in the liquid phase 54, 5205–5209.
Karpilovski, O.I., 1997. Design of variants for reactive distillation
Greek letters processes. Ph.D. Thesis. Moscow Institute of Fine Chemical
Technology, Moscow.
Levering, W., Bailer, O., Hanika, J., Kolena, J., Moravek, P., Smejkal, Q.,
i weighting factor for impact category l (dimen-
Lederer, J., Macek, V., 1999. Process and apparatus for the production
sionless) of n-butyl acetate and isobutyl acetate. Patent WO9948855.
i activity coefficient of component i Loning, S., Horst, C., Hoffmann, U., 2000. Theoretical Investigations on
k overall PEI of chemical k the quaternary System n-Butanol, n-Butyl acetate, Acetic Acid and
kl specific PEI of chemical k for impact category l Water. Chemical Engineering Technology 23 (9), 789–794.
Mallick, S., Cabezas, H., Bare, J., 1999. Pollution prevention with
chemical process simulators: the generalized waste reduction (WAR)
algorithm-full version. Computers and Chemical Engineering 23,
Acknowledgements 623–634.
Paiva, A.L., Malcata, X., 2000. Comparison of the performance of
integrated and sequential reaction and separation units in terms of
The authors wish to thank Todd Martin for providing a re-
recovery of a desired product. Chemical Engineering Science 55,
vised version of the software WARGUI . Finally the authors 589–599.
wish to thank the Investigation Department of the National Pisarenko, Y.A., Serafimov, L.A., Cardona, C.A., Efremov, D.L.,
University of Colombia at Manizales for financial support Shuwalov, A.S., 2001. Reactive distillation design: analysis of the
of this project. process statics. Reviews in Chemical Engineering 17 (4).
Solokhin, A.W., Blagov, S.A., Pimoseew, W.S., 1997. Comparative
analysis of reactor column system and the reactive distillation based
on the energetic consumption. Theoretical Foundations of Chemical
References Engineering 31 (1), 49–54.
Venimadhavan, G., Malone, M.F., Doherty, M.F., 1999. A novel distillate
Cardona, C.A., 2001. Design of reactive distillation processes for policy for batch reactive distillation with application to the production
producing alkylacetates by esterification of acetic anhydride. of n-Butyl Acetate. Industrial & Engineering Chemistry Research 38,
Ph.D. Thesis, Lomonosov Academy of Fine Chemical Technology. 714–722.
Moscow. Young, D.G., Cabezas, H., 1999. Designing sustainable process with
Cardona, C.A., Cubides, D.C., 2003. Comparison of reactive distillation simulation: the waste reduction (WAR) algorithm. Computers and
and the sequential process based on energy consumption. (ref. 7.2- Chemical Engineering 23, 1477–1491.
0080). 4th European Congress in Chemical Engineering, Granada, Zhicai, Y., Xianbao, C., Jing, G., 1998. Esterification-distillation of butanol
Spain, September 21–25. and acetic acid. Chemical Engineering Science 53, 2081–2088.