You are on page 1of 13

Separation and Purification Technology 272 (2021) 118887

Contents lists available at ScienceDirect

Separation and Purification Technology


journal homepage: www.elsevier.com/locate/seppur

Novel energy-saving methods to improve the three-column extractive


distillation process for separating ethyl acetate and ethanol using furfural
Cong Duan , Chunli Li *
National-Local Joint Engineering Laboratory for Energy Conservation in Chemical Process Integration and Resources Utilization, School of Chemical Engineering and
Technology, Hebei University of Technology, Tianjin 300130, China

A R T I C L E I N F O A B S T R A C T

Keywords: A mixture of ethyl acetate and ethanol is difficult to separate because of a minimum-boiling homogeneous
Vapor–liquid coupling azeotrope. The characteristics of the ethyl acetate/ethanol system are analyzed. The energy-saving potential of a
Heat integration three-column extractive distillation process to separate the system is explored. The effects of the distillate
Partial condensation
composition of pre-concentrator, the temperature of solvent feed and operating pressures are studied. Energy-
Extractive distillation
Energy saving
saving processes including heat pump distillation, heat integration and partial condensation are proposed. A
Ethyl acetate/ethanol novel way of partial condensation is proposed and compared with the conventional way. An innovative method
of vapor–liquid coupling between the per-concentrator and the extractive column is then proposed. It is more
economical and energy-efficient than the above processes. Besides, based on the vapor–liquid coupling process,
two further improvements are proposed through heat integration and through compressors. All processes are
optimized using the sequential iterative optimization procedure with the minimum total annual cost as the
target. Compared with the conventional process operated at 101.3 kPa, the two improved processes can reduce
total annual cost by more than 52%, reduce energy consumption by more than 60%, and reduce CO2 emissions
by more than 61%.

emissions and high total annual cost (TAC) [15]. Therefore, it is


1. Introduction necessary to explore energy-saving processes for the TCED.
In the literature, many methods and techniques have been used to
Ethyl acetate (EA) is not only an important raw material and solvent reduce energy consumption and intensify the process for the ED process,
[1], but a popular product in organic synthesis [2]. During the synthesis including optimization process improvement [16,17], heat integration
process of EA, a mixture of EA and ethanol is difficult to separate [18,19], heat pump [20,21], partial condensation [22,23], column
because of a minimum-boiling homogeneous azeotrope [1] with a combination [9,24,25], etc.
composition of 53.5 mol % EA at 101.3 kPa as given in Fig. 1. Many Among all the optimization variables, distillate composition of the
methods have been studied to separate EA and ethanol including C1 (xD1) is an important one [13,26]. Operating pressure of a column is
pressure-swing distillation [3–5], extractive distillation (ED) [1,6,7], getting more and more attention. It has an influence on the phase
and liquid–liquid extraction [8]. In this work, ED with furfural as the equilibrium [16,17,27], entrainer selection [10], and the diameter of the
entrainer [9–11] is studied. columns.
ED is widely used to separate binary close-boiling/azeotropic mix­ There are two cases of heat integration in the ED process, one of
tures by introducing a solvent to change the relative volatility between which is double-effect heat integration. The double-effect heat integra­
the feed components [12]. When the feed composition is dilute and far tion includes two columns with different pressures. The overhead vapor
from the azeotropic composition, a three-column extractive distillation of the high-pressure column is used as the heat source in the reboiler of
(TCED) process is usually used. The three columns are pre-concentrator the low-pressure column [28]. The other case is to use economizers to
(C1), extractive column (C2) and solvent recovery column (C3) [13]. As preheat the feed streams with the hot solvent from the C3 base [29]. A
a kind of special distillation [14], ED not only has the advantages in partial heat-integrated TCED scheme to separate acetonitrile/water
operation and control, but also has the disadvantage of high energy proposed by Qi et al. [19] could reduce energy consumption by 36.92%,
consumption. High energy consumption is accompanied by high CO2 and reduce TAC by 27.26% compared with the conventional TCED

* Corresponding author.
E-mail address: lichunli_hebut@126.com (C. Li).

https://doi.org/10.1016/j.seppur.2021.118887
Received 11 January 2021; Received in revised form 20 April 2021; Accepted 29 April 2021
Available online 9 May 2021
1383-5866/© 2021 Elsevier B.V. All rights reserved.
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Nomenclature P electric power, kW


P1 operating pressure for the pre-concentrator, kPa
Abbreviations P2 operating pressure for the extractive column, kPa
COP coefficient of performance P3 operating pressure for the solvent recovery column, kPa
EA ethyl acetate PC-1 the process in which a partial condenser is installed in the
ED extractive distillation pre-concentrator.
NRTL non-random two-liquid PC-2 the process in which a partial condenser is installed in the
TCED three-column extractive distillation extractive column
TAC total annual cost, k$/y Q energy consumption, kW
VLE vapor–liquid equilibrium QW reboiler duty, kW
QW1 the reboiler duty of the pre-concentrator, kW
Symbols QW2 the reboiler duty of the extractive column, kW
C1 pre-concentrator QW3 the reboiler duty of the solvent recovery column, kW
C2 extractive column QD condenser duty, kW
C3 solvent recovery column QD1 the condenser duty of the pre-concentrator, kW
C% carbon content QD2 the condenser duty of the extractive column, kW
D distillate rate, kmol/h QD3 the condenser duty of the solvent recovery column, kW
S solvent flowrate, kmol/h Qh the heat rejected at high temperature, kW
hProc the enthalpy of the utility steam used in the process, kJ/kg RR reflux ratio
hwater,373.15K the enthalpy of water at 373.15 K, kJ/kg Ts the temperature of solvent feed, K
HP-1 heat pump distillation process for the pre-concentrator TFTB the theoretical flame temperature, K
HI-12 heat integration process between the pre-concentrator and Tstack the stack temperature, K
the extractive column T0 ambient temperature, K
Nt1 the number of the theoretical stages of the pre- ΔT temperature difference, K
concentrator ΔTm the minimum temperature difference, K
Nt2 the number of the theoretical stages of the extractive VLC-12 the vapor–liquid coupling process between the pre-
column concentrator and the extractive column
Nt3 the number of the theoretical stages of the solvent recovery VLC-HI the improved VLC-12 process by heat integration
column VLC-C the improved VLC-12 process by a compressor
Nf1 the feed stage location of the pre-concentrator W the work input for the compressor, kW
Nf2 the feed stage location of the extractive column xD1 distillate composition of the pre-concentrator
Nfs the feed stage location of solvent x liquid molar fraction of ethyl acetate
Nf3 the feed stage location of the solvent recovery column y vapor molar fraction of ethyl acetate
NHV the net heating value of fuel, kJ/kg
Opt the process that optimizing more variables

without heat integration. Another two-column ED process with optimal


partial heat integration to separate acetone and methanol [30] could
reduce energy consumption by 32.2% and reduce TAC by 24.4%
compared with the base case without heat integration.
The energy can also be reduced by upgrading low-temperature waste
heat to a higher temperature with a heat pump [31]. CO2 emissions can
be greatly reduced in this way. A partial vapor recompression heat pump
process to separate acetone and water by two-column ED can reduce
CO2 emissions by 55.92%, despite an 18.12% increase in TAC (3-year
payback period) compared with the case without heat pump [30]. Cui
et al. [32] shown that using heat pumps in the ED process to separate
EA/isopropanol/water allowed for a 37.16% reduction in CO2 emis­
sions, and a 34.78% reduction in TAC.
Partial condensers can also be used for the TCED to provide a vapor
feed to reduce the energy. Errico et al. [22] proposed various alternative
sequences for bioethanol production and showed the energy-saving
potential of partial condensers. Zhang et al. [23] replaced the C1 and
its partial condenser with a pre-stripper. Using a pre-stripper in TCED
could save TAC by 23.6% and 14.4% compared with using a pre-
concentration column with a total condenser for acetonitrile/water/
(ethylene glycol) system and N-propanol/water/(N-methyl-2-pyrroli­
done) system respectively (note that the brackets indicate the solvent).
Column combination [9,24,25] is also very popular in the ED pro­
cess. For the TCED, C1 and C2 can be combined [9], or C1 and C3 can be
Fig. 1. T–xy diagram for EA/ethanol at 101.3 kPa. combined [24]. Liang et al. [24] combined C1 and C3 for a conventional
TCED process and showed that the column combination could save TAC
by 12.1% and 14.7% for separating acetonitrile/water/(ethylene glycol)

2
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 2. The sequential iterative optimization procedure.

system and isopropanol/water/(dimethyl sulfoxide) system respec­ obtained, energy consumption, CO2 emissions and coefficient of per­
tively. An et al. [9] proposed two conventional TCED processes at at­ formance (COP, if there is a heat pump in the process) will be also
mospheric pressure and their corresponding energy-saving processes in calculated to evaluate the processes.
which C1 and C2 were combined. They showed that the column com­
bination could save TAC by 15.2% and 17.6% for separating EA/ 2.1. Thermodynamic model
ethanol/(furfural) system and N-propanol/water/(N-methyl-2-pyrroli­
done) system respectively. Thermodynamic model is very important in predicting vapor–liquid
It is worth mentioning that for many systems such as ethanol/water/ equilibrium (VLE) in the simulation, which affects the accuracy of the
(ethylene glycol) system and acetonitrile/water/(dimethyl sulfoxide) calculation results [33]. According to the work by An et al. [9], non-
system, the separation in the extractive column becomes easier with a random two-liquid (NRTL) model is selected for this system. The bi­
decreasing pressure. Moreover, the light key component is the overhead nary parameters of NRTL model have been slightly modified based on
product of the extractive column. However, the separation of EA and the literature [9]. The source of the binary parameters for EA/furfural
ethanol using furfural is different, which will be discussed in this work. system was changed from VLE-IG to NIST-IG. The Table A1 in Supple­
Based on an above conventional TCED process to separate the EA/ mentary material shows the three sets of binary parameters used for
ethanol system [9], the purpose of this work is to explore its energy- NRTL model. The accuracy of the binary parameters is verified by the
saving potential. experimental data [34–38]. Fig. A1 (a)− (c) shows that the experimental
The effects of some key variables are analyzed. The characteristics of data fits well with the model for the three binary systems.
the EA/ethanol/(furfural) system are obtained. Several energy-saving
processes including a heat-pump distillation process, a heat-integrated 2.2. Optimization procedure
process, a conventional partial-condensation process, a novel partial-
condensation process and an innovative vapor–liquid coupling process During the optimization, the objective function is TAC and the target
are proposed. Based on the innovative vapor–liquid coupling process, is to minimize the TAC. Optimization variables include three operating
further improvements through heat integration and vapor recom­ pressures (P1, P2, P3), three total numbers of the theoretical stages (Nt1,
pression are explored. All the processes are optimized using the Nt2, Nt3), three feed stage locations (Nf1, Nf2, Nf3), the feed stage
sequential iterative optimization procedure with minimum TAC as the location of solvent (Nfs), the solvent flowrate (S), xD1, and the solvent
target. After the optimal processes are obtained, the CO2 emissions and feed temperature (Ts).
energy consumption are also calculated for the processes. The sequential iterative optimization procedure [9,19,24] is used in
this work. Fig. 2 shows the procedure, where the three columns are
2. Methods of simulation, optimization and evaluation optimized in the sequence of C1, C2, C3, C1… until the TAC is the
minimum. The detailed steps and some special cases are in the Supple­
In this work, all the simulations and optimizations are carried out mentary material.
using Aspen Plus V11 software. The sequential iterative optimization
procedure is selected for process optimization. The optimization target is 2.3. Economic evaluation
to minimize TAC. Therefore, TAC is the objective function for every
process. After the optimal processes with the minimum TAC are TAC is calculated from:

3
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 3. Base case: the conventional TCED process at 101.3 kPa proposed by An et al.[9]

capital cost
TAC = + operating cost (1) [CO2 ]Emiss = QFuel ⋅FuelFact (3)
payback period
Eq. (3) shows that the CO2 emissions are related to the amount of fuel
The capital cost, payback period and operating cost are calculated
burnt QFuel and the effect of the fuel FuelFact.
according to the formulas listed in Table A2, assuming an annual
operating time of 8000 h. The costs of pumps, valves and reflux drums QProc ( ) TFTB − T0
QFuel = hProc − hwater,373.15K (4)
are ignored at the conceptual design stage [39]. The coefficient M&S is λProc TFTB − TStack
1638.2 (2018) [40,41].
The formulas to calculate the vacuum system costs are from the book where λProc (kJ/kg) and hProc (kJ/kg) represent the latent heat and the
of Seider et al. [42]. Liquid-ring pumps are selected due to the lower cost enthalpy of the utility steam used in the process, respectively. The
than that of steam-jet electors. An isentropic compressor model is used theoretical flame temperature (TFTB) of 2073.15 K, the stack tempera­
in the simulation with an isentropic efficiency of 0.4 and a mechanical ture (TStack) of 433.15 K and the ambient temperature of 293.15 K are
efficiency of 0.92 to calculate the power of the liquid-ring pump. The used in Eq. (4). QProc is the heat duty required by the process. The boiler
amount of air leakage is estimated from the equation in the Table A1. feedwater is assumed to be at 373.15 K with an enthalpy of 419 kJ/kg.
When the air leakage is added to the overhead vapor, a flash model is ( α )( C% )
used in the simulation. It gives a vapor leaving the reflux drum and FuelFact = (5)
NHV 100
entering the vacuum system. The flash is set at a temperature of 323 K
and a pressure drop of 5 kPa. where α (=3.67) is the ration of molar masses of CO2 and C, while NHV
represents the net heating value of a fuel with a carbon content of C%.
2.4. Energy consumption evaluation Heavy fuel oil is used in this work. The NHV and C% are 39771 kJ/kg
and 86.5, respectively [30].
Energy consumption (Q, kW) is mainly composed of thermal energy The CO2 emissions of electricity are 51.1 kg CO2/GJ according to the
consumption (reboiler duty (QW, kW)) and electrical energy consump­ article [46]. Using enthalpy data from Aspen V11 and using the equa­
tion. Considering the thermoelectric efficiency, it is assumed that elec­ tions from (3) to (5), the calculated CO2 emissions of LP steam, MP steam
tric power (P, kW) can be converted into the equivalent thermal energy and HP steam are 98.44 kg CO2/GJ, 104.53 kg CO2/GJ and 135.75 kg
by a factor of 3 [43,44]. So, the energy consumption is calculated from: CO2/GJ, respectively.

Q = QW + 3P (2)
2.6. Heat pump performance evaluation

2.5. CO2 emissions evaluation The heat pump performance is evaluated by the COP, which is pro­
posed by Bruinsma and Spoelstra [47]. It is the ratio of the heat rejected
CO2 emissions are considered for environmental impact [23]. The (Qh) at high temperature to the work input of the compressor (W):
hot utility of the reboiler is the steam which is produced by steam boilers Qh
from the combustion of fuel. The CO2 emissions from the steam boilers COP = (6)
W
are calculated from [45]

Table 1
Optimal design results.
Optimization results Base case Opt HP-1 HI-12 PC-1 PC-2 VLC-12 VLC-HI VLC-C

Capital cost (k$) 3315 3059 4855 2975 2558 2856 2400 2547 2567
Energy cost (k$/y) 3685 2476 1402 1698 2585 2189 1662 1425 1293
Q (kW) 8531 6176 4211 4092 6500 5387 4004 3221 3393
CO2 (kg/h) 3359 2220 1136 1520 2326 1961 1485 1293 1138
TAC (k$/y) 4790 3496 3021 2690 3437 3141 2462 2274 2148
Q saving* – 27.6% 50.6% 52.0% 23.8% 36.9% 53.1% 62.2% 60.2%
CO2 reduction* – 33.9% 66.2% 54.7% 30.8% 41.6% 55.8% 61.5% 66.1%
TAC saving* – 27.0% 36.9% 43.8% 28.2% 34.4% 48.6% 52.5% 55.2%
*
Compared with the base case, which is from the reference article [9].

4
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 4. x–y diagram for EA/ethanol at 101.3, 50, and 200 kPa. Fig. 6. x–y diagram for ethanol/furfural at 101.3, 50, and 200 kPa.

Fig. 5. x–y diagram for pseudo EA/ethanol system at 101.3, 50, and 200 kPa Fig. 7. x–y diagram for EA/furfural at 101.3, 50, and 200 kPa.
with furfural as the solvent (the composition of furfural is 75%).
comparison in this work.
Plesu et al. [48] proposed that when the COP was higher than 10, the
heat pump was clearly recommended; when the COP was between 5 and
10, a detailed evaluation should be carried out. 3.2. VLE analysis

3. Results and discussion Fig. 4 shows the x–y diagram for the EA/ethanol system at different
pressures. It can be clearly seen that the pressure has effects on the
3.1. Base case relative volatility and azeotropic composition. Since the feed in this
work is 0.25 mol% EA, which is less than the azeotropic composition, the
Fig. 3 shows the conventional flowsheet for the TCED process pro­ C1 bottom product will be ethanol while the distillate product will be a
posed by An et al. [9]. The product specifications are 99.5 mol% EA and mixture of ethanol and EA. Moreover, in the region where x is less than
99.5 mol% ethanol. The feed is a saturated liquid with a composition of the azeotropic point, the VLE curve deviates more from the diagonal line
25 mol% EA and a flow rate of 300 kmol/h. The operating pressure of at a lower pressure, and the azeotropic composition is greater at a lower
each column is set at 101.3 kPa. The xD1 is specified to be 45.0 mol% EA, pressure. Therefore, a lower pressure is more helpful for the separation
which near the azeotropic composition. The Ts is specified to be 343.15 in the C1.
K. Using the above conditions for simulation, the recalculated results are Fig. 5 shows the entrainer-free basis equilibrium curves at 101.3, 50
shown in Column base case in Table 1, which are used as the basis for and 200 kPa. The molar composition of the furfural in the feed stream is

5
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 8. The optimal three-column process without heat integration (Opt).

0.75. In Fig. 4, x is the liquid molar fraction of EA without computing 3.3. Process improvement by optimizing more variables (Opt)
furfural while y is the vapor molar fraction of EA without computing
furfural. The Fig. 5 can reflect most of the separation process (extractive Among all the variables in the design and optimization of the TCED,
section and stripping section) in the C2. The VLE curves are below the xD1, Ts and operating pressures are important. However, in the base
diagonal line, which means the volatility of ethanol is greater than that case, optimization is performed with the fixed xD1, Ts, P1, P2 and P3,
of EA. Therefore, the distillate of the C2 is ethanol. Moreover, the VLE and without calculating the pressure drop. According to the first feature
curve deviates more from the diagonal line under a higher pressure, of the system in Section 3.2, the pressure has effects on the three col­
indicating a higher pressure is more helpful for the separation. However, umns and these effects are different. Therefore, the base case can be
higher pressures mean higher bottom temperatures and more expensive improved by varying P1, P2, P3 as well as xD1, Ts. The pressure drop for
energy sources [49]. Therefore, there is a trade-off on selecting an each stage is added and set at 0.6 kPa for the columns at atmospheric
operating pressure. pressure and high pressure. The tray pressure drop in a vacuum column
The separation process of the solvent recovery section in the C2 can is set at 0.2 kPa according to the paper by Luyben [27].
be reflected in Fig. 6, which shows the x–y diagram for the ethanol/ As the pressure increases, the condenser temperature and the
furfural system at different pressures. Fig. 7 shows the x–y diagram for reboiler temperature both increase. In this work, a minimum reflux-
the EA/furfural system at different pressures. It can reflect the separa­ drum temperature of 323 K is selected so that the heat sink used in
tion process in the C3. The separation is very easy for the two systems the condenser can be cooling water [27]. However, the minimum tem­
due to the very fat curves. Moreover, as the pressure decreases from 200 perature difference (ΔTm) for reboilers is set to 20 K so that thermosi­
kPa to 50 kPa, the VLE curve becomes fatter, so the separation becomes phon reboilers can be used. The bottom temperature should be
easier. controlled so that heat source used in the reboilers can be steam.
From the VLE analysis and Fig. 1, three important features of the Fig. 8 shows the process in which P1, P2, P3, xD1 and Ts are included
system are obtained. (1) The separation becomes easier as the pressure in the optimization variables. This process is named Opt process. During
increases when using furfural as the solvent in the C2, while in the other the optimization (mainly when optimizing Nt2), it is not conducive to
two columns, the separation becomes easier as the pressure decreases. practical application due to too many trays. Moreover, as the Nt2 in­
(2) At 101.3 kPa, the difference of boiling points between the two creases, the impact of it on TAC is small after more than 70 stages.
components is only 1.1 K, and the difference between the boiling points Therefore, the upper limit of the Nt2 is set at 70 in this work. The
and the minimum azeotropic point is less than 6.5 K. (3) The C1 bottom optimal results are listed in Column Opt in Table 1. The effects of the
product is the same with the C2 overhead product when the feed variables on TAC are studied as follows.
composition is lower than the azeotropic composition. For the C1 and C3, the separation is favored by a lower pressure as

Fig. 9. (a) Effects of P1 on capital cost, energy cost and TAC; (b) effects of P3 on capital cost, energy cost and TAC.

6
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 10. (a) Effects of P2 on QW2 and QW3; (b) effects of P2 on capital cost, energy cost and TAC.

Fig. 11. (a) Effects of xD1 on heat duty distribution; (b) effects of xD1 on energy cost, capital cost and TAC.

shown in Figs. 4 and 7. Therefore, as the pressure increases, energy costs causes the reduction in energy as shown in Fig. 10 (a). This is because
will increase while the column diameter and the vacuum pump costs the thermal conditions of the C2 bottom stream have changed, leading to
(capital cost and operating cost) will decrease. As the pressure increase changes in the feed conditions of C3, which affects the C3 reboiler duty
causes the temperature difference (ΔT) in the reboiler to decrease, the (QW3). Meanwhile, the equipment cost of the reboiler increases due to
capital cost of the reboiler will increase. Meanwhile, as the pressure the higher bottom temperature. The column diameter decreases due to
increases, the ΔT in the condenser increases and the capital cost of the the reduced vapor density. The total capital cost increases due to the
condenser will decrease. All the effects are added together as shown in overall effects. In addition, when the pressure is greater than 393 kPa,
Fig. 9. The optimal P1 is 48 kPa, and the optimal P3 is 52 kPa. the bottom temperature exceeds 437 K. In order to ensure a ΔT of 20 K,
For the C2, as analyzed in Section 3.2, the separation is more favored the heat source (middle-pressure steam, $14.19/GJ, 457 K) needs to be
by a higher pressure. Therefore, as P2 increases, the energy cost de­ replaced with the high-pressure steam ($17.70/GJ, 537 K). Therefore,
creases. The important thing is that not only the C2 but also the C3 that the energy cost suddenly increases, while the equipment cost suddenly

Fig. 12. (a) Effects of Ts on the C2; (b) effects of Ts on energy cost, capital cost and TAC.

7
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 13. Flowsheet of heat pump distillation (HP-1).

decreases due to an increased ΔT in the reboiler. As a result, the TAC resulting in little effect on the energy cost as shown in Fig. 12 (b).
increases suddenly. Fig. 10 (b) shows the effects of P2 on the costs. Meanwhile, the Ts has a major impact on the condenser duty. Therefore,
Fig. 11 (a) shows the effects of xD1 on heat duty distribution. As xD1 as Ts increases, the capital cost of the condenser increases. However, the
increases, the reboiler duty of the C1 (QW1) increases because the capital cost of the cooler decreases. The decrease is more than the in­
separation becomes more difficult when xD1 approaches the azeotropic crease. As a result, the capital cost and TAC decrease slowly. The TAC is
composition (0.613 at 48 kPa). However, the reboiler duty of C2 (QW2) the minimum when the Ts is the maximum. Therefore, the cooler is
decreases. This is because the composition of EA in the C2 feed is getting removed and the high-temperature solvent is directly fed to the C2 as
higher and the feed flowrate is getting smaller. The QW3 remains un­ shown in Fig. 8.
changed. Therefore, the total QW of the three columns first decreases After the optimization, TAC could be reduced by 27.0% compared
and then increases. The change in the condenser duty is similar to the with the base case. Meanwhile, energy consumption could be reduced by
change in the reboiler duty as shown in Fig. 11 (a). When xD1 equals 27.6% and CO2 emissions could be reduced by 33.9%. This process is
0.55, QW has the minimum values. Fig. 11 (b) shows the effects of xD1 on even more energy-efficient (6176 kW) compared with the energy-
the energy cost, capital costs and TAC. The change trend of the TAC is efficient combined-column process (7177 kW) in the literature [9].
similar to the change trend of the energy cost, and is similar to the trend The results show that adding an economizer and performing a
of QW in Fig. 11 (a). Therefore, when xD1 equals 0.55, the TAC is also the comprehensive optimization are very effective energy-saving measures
minimum. for this system.
Fig. 12 (a) shows the effects of Ts on the C2 in terms of the condenser
duty and the reboiler duty. As Ts increases, the QW2 decreases hardly, 3.4. Process improvement by heat-pump distillation
but the C2 condenser duty (QD2) increases rapidly. The heat sink in the
condenser is inexpensive cooling water (0.354 $/GJ) which is much Based on the Opt process, it is considered to use heat pump distil­
cheaper than middle-pressure steam (14.19 $/GJ) used as the heat lation for the C1 (HP-1) due to a low overall column temperature [30] in
source in the reboiler. The Ts has little effect on the reboiler duty, C1 (the second feature of the system from Section 3.2). The HP-1 process

Fig. 14. Flowsheet of heat integration between the C1 and the C2 (HI-12).

8
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 15. Flowsheet of no condensation for the C1 (PC-1).

is shown in Fig. 13. The C1 overhead vapor is fed to a compressor to 3.5. Process improvement by heat integration between the C1 and the C2
increase the pressure and the temperature. Therefore, it could provide (HI-12)
heat for the C1 reboiler. At the same time, it is condensed. After reducing
pressure through a valve, part of it is refluxed to the top of the column, In the Opt process in Fig. 15, both the C1 and the C3 are operated at a
while the other part is taken out as the overhead product. Usually, an vacuum pressure, while the C2 is at a high pressure. According to Luy­
auxiliary reboiler or an auxiliary condenser needs to be installed when ben’s paper, if the C2 condenser temperature is sufficiently larger
the heat is not so matched [50]. (20–30 K) than the C1 reboiler temperature, a heat exchanger can serve
The COP value for this heat pump is 7.5, which indicates that as both a condenser and a reboiler [51].
whether a heat pump should be used requires further evaluation. The Due to the coupling of QD2 and QW1, the parameters that affect QD2
other detailed results are shown in Table 1. It achieves a 36.9% reduc­ and QW1 are important. All the variables should be analyzed and
tion in TAC, a 50.6% reduction in energy consumption, and a 66.2% optimized with the minimum TAC as the target. During the optimiza­
reduction in CO2 emissions compared with the base case. Therefore, the tion, if |QD2| is greater than QW1, an auxiliary condenser is needed for
heat-pump distillation process is recommended for the system. the C2. If |QD2| is less than QW1, an auxiliary reboiler is necessary for
the C1. After the comparison, it is found that the former is more
economical. However, the auxiliary condenser duty is very small (0.1
kW), which means an almost complete coupling.
The optimal results are listed in Column HI-12 in Table 1. The

Fig. 16. Flowsheet of installing a partial condenser for the C2 (PC-2).

9
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 17. Flowsheet of the vapor–liquid coupling between the C1 and the C2 (VLC-12).

flowsheet is shown in Fig. 14. TAC could be reduced by 43.8%, energy During the optimization process, it is found that it is more economical to
consumption could be reduced by 52.0%, and CO2 emissions could be remove the condenser when the reflux ratio (RR) is very small. The
reduced by 54.7% compared with the base case. absence of a condenser is considered an extreme case of partial
Heat integration between C1 and C2 is very suitable for this system condensation. So, in order to describe this method, even if there is no
because of the natural ΔT. condenser, the name of the method is still called partial condensation.
The process of installing a partial condenser for the C1 to produce a
vapor product which is fed to the C2 (Fig. 15) is considered first. In this
3.6. Process improvement by partial condensation
way, the capital cost of the C1 condenser and the energy cost of the C2
reboiler are greatly reduced.
3.6.1. Partial condensation (or no condensation) for the C1 (PC-1)
The vapor feed makes the concentration of the entrainer in liquid
In a partial condenser, only the reflux stream is condensed, while the
phase more constant which is beneficial to the separation in the C2.
distillate stream is in the vapor phase. Energy can be saved when this
The partial condenser for the C1 can be removed due to the small RR.
vapor product is fed to the next column. Using a partial condenser in­
Therefore, the C1 is replaced by a pre-stripper [23]. This process can
volves the problem of vapor feed. To avoid using a compressor, the
reduce TAC by 28.2%, reduce energy consumption by 23.8%, and
pressure difference between the vapor stream and the feed stage is at
reduce CO2 emissions by 30.8% compared with the base case.
least 15 kPa considering the resistance loss of pipelines and valves.

Fig. 18. Flowsheet of vapor–liquid coupling between the C1 and the C2 and heat integration between the C1 and the C3 (VLC-HI).

10
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

Fig. 19. Flowsheet of vapor–liquid coupling between the C1 and the C2 and installing a compressor between the C1 and the C2 (VLC-C).

3.6.2. A novel way of partial condensation (PC-2) overhead product is lower, resulting in a lower energy consumption.
Because the third feature analyzed in Section 3.2 is that the C2 Therefore, VLC-12 is better than HI-12.
overhead product and the C1 bottom product are both ethanol, another
way to install a partial condenser is innovatively proposed. Based on the 3.7.2. Further improvement based on the VLC-12 process through heat
direct steam injection method [52], a partial condenser can be installed integration (VLC-HI)
for the C2 to produce a vapor stream which could be directly fed to the In Section 3.7.1, we have found the most economical process, namely
C1 bottom to provide some direct heat input. In this way, the capital cost VLC-12. However, energy saving potential can still be explored for this
of the C2 condenser and the energy cost of the C1 reboiler are greatly process. The first way is heat integration. As shown in Fig. 18, the C3
reduced. condenser is thermally integrated with the C1 reboiler. The high-
Fig. 16 shows the optimal PC-2 process and Table 1 shows the results. temperature solvent from the C3 bottom is used as a heat source to
Installing a partial condenser for the C2 can reduce TAC by 34.4%, preheat the C2 feed.
reduce energy consumption by 36.9%, and reduce CO2 emissions by In the thermal integration process, when QW1<|QD3|, an auxiliary
41.6% compared with the base case. condenser is installed in the C3. When QW1>|QD3|, an auxiliary
The TAC, energy consumption and CO2 emissions of PC-2 are all reboiler is installed in the C1. It is found that the former scheme is more
reduced compared to those of the PC-1 process Therefore, installing a economical after the calculation. However, just like the HI-12 process,
partial condenser for C1 is not so good as installing a partial condenser the auxiliary condenser is also vary small (0.03 kW). In this way, TAC is
for C2. This is mainly because the optimum pressures of the C1 and the reduced by 7.6%, energy consumption is reduced by 19.6%, and CO2
C2 are too different. The C1 and the C2 are operated at their respective emissions are reduced by 12.9% compared with those of the VLC-12
optimum pressures in the PC-2, but not in the PC-1. process.
The latent heat recovered by the PC-2 process is less than that of HI-
12 due to the reflux in the C2. So, HI-12 is more energy efficient, more 3.7.3. Further improvement based on the VLC-12 process through a
ecological and more economical than PC-2. compressor (VLC-C)
In addition to heat pump distillation, there are other ways to use
compressors. Base on the VLC-12 process in Section 3.7.1, a partial
3.7. Process improvement by a novel energy-saving method of condenser is installed in the C1 to produce a vapor stream. This vapor
vapor–liquid coupling stream then enters a suitable location of C2 after being pressurized by a
compressor as shown in Fig. 19. In this way, capital cost of C1 condenser
3.7.1. Vapor-liquid coupling between the C1 and C2 (VLC-12) and costs of C2 reboiler are all reduced despite a large increase in the
In order to increase the amount of latent heat recovered in PC-2, the equipment cost and operating cost of the compressor. Therefore, TAC is
C2 condenser is removed so that all the vapor can be fed to the C1 reduced by 12.8%, energy consumption is reduced by 15.3%, and CO2
bottom. Part of the C1 bottom product is refluxed to the C2. The amount emissions are reduced by 23.4% compared with those of the VLC-12
of the reflux is also one of the optimized variables. In this way, C2 process. Compared with the HP-1, the VLC-C process can save TAC by
overhead vapor can be fully utilized without partial condensation. 28.9%, and save energy by 19.4%, despite a very small increase (0.2%)
During the optimization process, it is found that as QW1 decreases, TAC in CO2 emissions.
decreases. So, the reboiler is finally removed. The optimal process is As shown from Fig. 8 and Figs. 13–19, all the processes have different
shown in Fig. 17, and the optimal results are shown in Table 1. It ach­ optimal values of xD1, Ts and pressures. Therefore, these parameters
ieves a 48.6% reduction in TAC, a 53.1% reduction in energy con­ should not be fixed but should be re-analyzed and re-optimized in
sumption, and a 55.8% reduction in CO2 emissions compared with the different processes. Table 1 lists the optimal results of all the processes.
base case. Comparing the above processes, it is found that VLC-12 is the As shown in Table 1, the most economical process is VLC-C. VLC-C is not
most economical process. as good as VLC-HI in terms of energy saving, but better than VLC-HI in
Compared with the HI-12 process, it is found that in the VLC-12 terms of reducing carbon emissions.
process, a condenser and a reboiler are saved, and the purity of the C2

11
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

4. Conclusions Sep. Purif. Technol. 239 (2020), 116571, https://doi.org/10.1016/j.


seppur.2020.116571.
[6] T. Shi, W. Chun, A. Yang, Y. Su, S. Jin, J. Ren, W. Shen, Optimization and control of
Based on the VLE analysis, the characteristics of using furfural to energy saving side-stream extractive distillation with heat integration for
separate EA and ethanol in a TCED process are obtained. Eight energy- separating ethyl acetate-ethanol azeotrope, Chem. Eng. Sci. 215 (2020), 115373,
saving processes are proposed and optimized with the minimum TAC as https://doi.org/10.1016/j.ces.2019.115373.
[7] Z. Zhu, Y. Ri, H. Jia, X. Li, Y. Wang, Y. Wang, Process evaluation on the separation
the target. The optimal results show that they are all more energy- of ethyl acetate and ethanol using extractive distillation with ionic liquid, Sep.
efficient than the energy-saving combined-column process (7177 kW) Purif. Technol. 181 (2017) 44–52, https://doi.org/10.1016/j.seppur.2017.03.011.
in the reference article [9]. Among the optimization variables, the [8] D.L. Zhang, Y.F. Deng, C.B. Li, J. Chen, Separation of ethyl acetate-ethanol
azeotropic mixture using hydrophilic ionic liquids, Ind. Eng. Chem. Res. 47 (2008)
operating pressures, xD1 and Ts are all key variables for the TCED pro­ 1995–2001, https://doi.org/10.1021/ie070658m.
cess especially for the processes that use energy-saving methods. [9] Y. An, W. Li, Y. Li, S. Huang, J. Ma, C. Shen, C. Xu, Design/optimization of energy-
Heat pump distillation, heat integration, two kinds of partial saving extractive distillation process by combining preconcentration column and
extractive distillation column, Chem. Eng. Sci. 135 (2015) 166–178, https://doi.
condensation, and vapor–liquid coupling methods are all very suitable org/10.1016/j.ces.2015.05.003.
for the EA/ethanol/(furfural) system. The novel way of partial [10] X. Zhang, X. Li, G. Li, Z. Zhu, Y. Wang, D. Xu, Determination of an optimum
condensation is better than the conventional way in terms of TAC, en­ entrainer for extractive distillation based on an isovolatility curve at different
pressures, Sep. Purif. Technol. 201 (2018) 79–95, https://doi.org/10.1016/j.
ergy and CO2 emissions. The novel vapor–liquid coupling energy-saving seppur.2018.03.007.
method is more economical than the others. [11] Z. Zhu, M. Yu, W. Zhang, Y. Liu, P. Cui, J. Yang, Y. Wang, J. Gao, Exploration of the
The VLC-12 process can be further improved by additional heat effects of pressure on the controllability of extractive distillation for separating
pressure-sensitive azeotropes, Sep. Purif. Technol. 227 (2019), 115681, https://
integration or vapor recompression. The VLC-C process, where a
doi.org/10.1016/j.seppur.2019.115681.
compressor is installed for the C1 in the VLC-12 process to generate a [12] V. Aniya, D. De, A. Singh, B. Satyavathi, Design and operation of extractive
vapor feed for the C2, is the most economical process. It can achieve a distillation systems using different class of entrainers for the production of fuel
55.2% reduction in TAC. It can also achieve a 60.2% reduction in energy grade tert-butyl Alcohol: a techno-economic assessment, Energy. 144 (2018)
1013–1025, https://doi.org/10.1016/j.energy.2017.12.099.
consumption and a 66.1% reduction in CO2 emissions. [13] W.L. Luyben, Control comparison of conventional extractive distillation with a new
The PC-2, VLC-12, VLC-HI, and VLC-C processes proposed in this split-feed configuration, Chem. Eng. Process. 107 (2016) 29–41, https://doi.org/
work are applicable to other systems whose characteristic is that the 10.1016/j.cep.2016.06.004.
[14] Y. Zhao, K. Ma, W. Bai, D. Du, Z. Zhu, Y. Wang, J. Gao, Energy-saving thermally
distillate product of C2 and bottom product of C1 are the same. coupled ternary extractive distillation process by combining with mixed entrainer
for separating ternary mixture containing bioethanol, Energy 148 (2018) 296–308,
CRediT authorship contribution statement https://doi.org/10.1016/j.energy.2018.01.161.
[15] A. Yang, S. Jin, W. Shen, P. Cui, I.L. Chien, J. Ren, Investigation of energy-saving
azeotropic dividing wall column to achieve cleaner production via heat exchanger
Cong Duan: Conceptualization, Data curation, Investigation, Meth­ network and heat pump technique, J. Clean. Prod. 234 (2019) 410–422, https://
odology, Software, Writing - original draft, Writing - review & editing. doi.org/10.1016/j.jclepro.2019.06.224.
[16] X. You, J. Gu, V. Gerbaud, C. Peng, H. Liu, Optimization of pre-concentration,
Chunli Li: Supervision, Validation, Funding acquisition, Writing - re­ entrainer recycle and pressure selection for the extractive distillation of
view & editing. acetonitrile-water with ethylene glycol, Chem. Eng. Sci. 177 (2018) 354–368,
https://doi.org/10.1016/j.ces.2017.11.035.
[17] C. Cui, J. Sun, Rigorous design and simultaneous optimization of extractive
Declaration of Competing Interest distillation systems considering the effect of column pressures, Chem. Eng. Process.
139 (2019) 68–77, https://doi.org/10.1016/j.cep.2019.04.001.
[18] J.P. Knapp, M.F. Doherty, Thermal integration of homogeneous azeotropic
The authors declare that they have no known competing financial distillation sequences, AIChE J. 36 (1990) 969–984, https://doi.org/10.1002/
interests or personal relationships that could have appeared to influence aic.690360702.
[19] J. Qi, Y. Li, J. Xue, R. Qiao, Z. Zhang, Q. Li, Comparison of heterogeneous
the work reported in this paper.
azeotropic distillation and energy-saving extractive distillation for separating the
acetonitrile-water mixtures, Sep. Purif. Technol. 238 (2020), 116487, https://doi.
Acknowledgements org/10.1016/j.seppur.2019.116487.
[20] H. Luo, C.S. Bildea, A.A. Kiss, Novel heat-pump-assisted extractive distillation for
bioethanol purification, Ind. Eng. Chem. Res. 54 (2015) 2208–2213, https://doi.
This work was supported by the National Key R&D Program of China org/10.1021/ie504459c.
[grant number 2017YFB0602500], and the Central Committee Guides [21] Z. Feng, W. Shen, G.P. Rangaiah, L. Dong, Design and control of vapor
recompression assisted extractive distillation for separating n-hexane and ethyl
Local Science and Technology Development Special Project [grant
acetate, Sep. Purif. Technol. 240 (2020), 116655, https://doi.org/10.1016/j.
number 19944507G]. seppur.2020.116655.
[22] M. Errico, B.G. Rong, G. Tola, M. Spano, Optimal synthesis of distillation systems
for bioethanol separation. Part 1: Extractive distillation with simple columns, Ind.
Appendix A. Supplementary material Eng. Chem. Res. 52 (2013) 1612–1619, https://doi.org/10.1021/ie301828d.
[23] X. Zhang, J. He, C. Cui, J. Sun, A systematic process synthesis method towards
Supplementary data to this article can be found online at https://doi. sustainable extractive distillation processes with pre-concentration for separating
the binary minimum azeotropes, Chem. Eng. Sci. 227 (2020), 115932, https://doi.
org/10.1016/j.seppur.2021.118887.
org/10.1016/j.ces.2020.115932.
[24] K. Liang, W. Li, H. Luo, M. Xia, C. Xu, Energy-efficient extractive distillation
References process by combining preconcentration column and entrainer recovery column,
Ind. Eng. Chem. Res. 53 (2014) 7121–7131, https://doi.org/10.1021/ie5002372.
[25] D. Han, Y. Chen, Combining the preconcentration column and recovery column for
[1] Q. Zhang, M. Liu, C. Li, A. Zeng, Design and control of extractive distillation
the extractive distillation of ethanol dehydration with low transition temperature
process for separation of the minimum-boiling azeotrope ethyl-acetate and ethanol,
mixtures as entrainers, Chem. Eng. Process. 131 (2018) 203–214, https://doi.org/
Chem. Eng. Res. Des. 136 (2018) 57–70, https://doi.org/10.1016/j.
10.1016/j.cep.2018.08.005.
cherd.2018.04.043.
[26] A.A. Kiss, R.M. Ignat, Optimal economic design of an extractive distillation process
[2] I.P. Semenov, Y.A. Pisarenko, A.K. Frolkova, Separation of a reaction mixture of
for bioethanol dehydration, Energy. Technol. 1 (2013) 166–170, https://doi.org/
ethyl acetate production via ethanol dehydrogenation, Theor. Found. Chem. EN+
10.1002/ente.201200053.
51 (2017) 418–431, https://doi.org/10.1134/s0040579517030113.
[27] W.L. Luyben, Distillation column pressure selection, Sep. Purif. Technol. 168
[3] K. Wang, J. Li, M. Lian, Z. Li, T. Du, Pressure swing distillation for separation of
(2016) 62–67, https://doi.org/10.1016/j.seppur.2016.05.015.
ethyl acetate and ethanol in sub-plateau region, China Pet Process Pe. 22 (2020)
[28] A.K. Jana, Heat integrated distillation operation, Appl. Energ. 87 (2010)
104–110.
1477–1494, https://doi.org/10.1016/j.apenergy.2009.10.014.
[4] Y. Wang, K. Ma, M. Yu, Y. Dai, R. Yuan, Z. Zhu, J. Gao, An improvement scheme for
[29] W.L. Luyben, Control of heat-integrated extractive distillation processes, Comput.
pressure-swing distillation with and without heat integration through an
Chem. Eng. 111 (2018) 267–277, https://doi.org/10.1016/j.
intermediate connection to achieve energy savings, Comput. Chem. Eng. 119
compchemeng.2017.12.008.
(2018) 439–449, https://doi.org/10.1016/j.compchemeng.2018.09.012.
[30] X. You, I. Rodriguez-Donis, V. Gerbaud, Reducing process cost and CO2 emissions
[5] Q. Zhang, A. Zeng, Y. Ma, X. Yuan, J. Gao, Dynamic control analyses of eco-
for extractive distillation by double-effect heat integration and mechanical heat
efficient partially heat-integrated side-stream pressure-swing distillation processes,

12
C. Duan and C. Li Separation and Purification Technology 272 (2021) 118887

pump, Appl. Energ. 166 (2016) 128–140, https://doi.org/10.1016/j. [41] C. Wang, Y. Zhuang, L. Liu, L. Zhang, J. Du, Design and comparison of energy-
apenergy.2016.01.028. saving double column and triple column reactive-extractive hybrid distillation
[31] L. Hegely, P. Lang, Reduction of the energy demand of a second-generation processes for ternary multi-azeotrope dehydration, Sep. Purif. Technol. 259 (2021),
bioethanol plant by heat integration and vapour recompression between different 118211, https://doi.org/10.1016/j.seppur.2020.118211.
columns, Energy 208 (2020), 118443, https://doi.org/10.1016/j. [42] W. Seider, J. Seader, D. Lewin, Product and process design principles: synthesis,
energy.2020.118443. analysis, and evaluation, 2004.
[32] P. Cui, F. Zhao, D. Yao, Z. Ma, S. Li, X. Li, L. Wang, Z. Zhu, Y. Wang, Y. Ma, D. Xu, [43] G.U. Bhaskar Babu, A.K. Jana, Reducing total annualized cost and CO2 emissions in
Energy-saving exploration of mixed solvent extractive distillation combined with batch distillation: dynamics and control, AIChE J. 59 (2013) 2821–2832, https://
thermal coupling or heat pump technology for the separation of an azeotrope doi.org/10.1002/aic.14076.
containing low-carbon alcohol, Ind. Eng. Chem. Res. 59 (2020) 13204–13219, [44] Z. Feng, W. Shen, G.P. Rangaiah, L. Lv, L. Dong, Process development, assessment,
https://doi.org/10.1021/acs.iecr.0c02225. and control of reactive dividing-wall column with vapor recompression for
[33] Y. Wu, D. Meng, D. Yao, X. Liu, Y. Xu, Z. Zhu, Y. Wang, J. Gao, Mechanism producing n-propyl acetate, Ind. Eng. Chem. Res. 58 (2019) 276–295, https://doi.
analysis, economic optimization, and environmental assessment of hybrid org/10.1021/acs.iecr.8b05122.
extractive distillation-pervaporation processes for dehydration of n-propanol, Acs. [45] M.A. Gadalla, Z. Olujic, P.J. Jansens, M. Jobson, R. Smith, Reducing CO2 emissions
Sustain. Chem. Eng. 8 (2020) 4561–4571, https://doi.org/10.1021/ and energy consumption of heat-integrated distillation systems, Environ. Sci.
acssuschemeng.0c00263. Technol. 39 (2005) 6860–6870, https://doi.org/10.1021/es049795q.
[34] C.-H. Tu, Y.-S. Wu, F.-C. Ou, Effect of 1,2-propanediol on the vapor-liquid [46] M.A. Waheed, A.O. Oni, S.B. Adejuyigbe, B.A. Adewumi, D.A. Fadare, Performance
equilibria of the ethyl acetate + ethanol system at 101.3 kPa,, Fluid Phase. enhancement of vapor recompression heat pump, Appl. Energ. 114 (2014) 69–79,
Equilibr. (1997). https://doi.org/10.1016/j.apenergy.2013.09.024.
[35] M. Kato, H. Konishi, M. Hirata, New apparatus for isobaric dew and bubble point [47] O.S.L. Bruinsma, S. Spoelstra, Heat pump in distillation. Eindhoven, Netherlands,
method. Methanol-water, ethyl acetate-ethanol, water-1-butanol, and ethyl 2010, pp. 21–28. ISBN 978-90-386-2215-6.
acetate-water systems, J. Chem. Eng. Data 15 (1970) 435–439, https://doi.org/ [48] V. Pleşu, A.E. Bonet Ruiz, J. Bonet, J. Llorens, Simple equation for suitability of
10.1021/je60046a021. heat pump use in distillation, in: J.J. Klemeš, P.S. Varbanov, P.Y. Liew (Eds.),
[36] N. Calvar, A. Dominguez, J. Tojo, Vapor-liquid equilibria for the quaternary Computer Aided Chemical Engineering, vol. 33, Elsevier, 2014, pp. 1327–1332.
reactive system ethyl acetate plus ethanol plus water plus acetic acid and some of [49] W.L. Luyben, Methanol/trimethoxysilane azeotrope separation using pressure-
the constituent binary systems at 101.3 kPa, Fluid, Phase. Equilibr. 235 (2005) swing distillation, Ind. Eng. Chem. Res. 53 (2014) 5590–5597, https://doi.org/
215–222, https://doi.org/10.1016/j.fluid.2005.07.010. 10.1021/ie500043c.
[37] S.E. Kharin, V.M. Perelygin, K.K. Polyanski, Liquid-vapor phase equilibrium in [50] A. Kazemi, A. Mehrabani-Zeinabad, M. Beheshti, Recently developed heat pump
water-furfural and ethanol-furfural systems, in: Section Title: Cellulose, Lignin, assisted distillation configurations: a comparative study, Appl. Energ. 211 (2018)
Paper, and Other Wood Products, vol. 23, 1970, pp. 15–16. 1261–1281, https://doi.org/10.1016/j.apenergy.2017.12.023.
[38] M.G. Myles, R.E. Wingard, Calculating activity coefficients, Ind. Eng. Chem. 53 [51] W.L. Luyben, Comparison of extractive distillation and pressure-swing distillation
(1961) 219–222, https://doi.org/10.1021/ie50615a029. for acetone/chloroform separation, Comput. Chem. Eng. 50 (2013) 1–7, https://
[39] W.L. Luyben, Distillation Design and Control using Aspen Simulation, second ed., doi.org/10.1016/j.compchemeng.2012.10.014.
AlChE: Wiley, Hoboken, New Jersey, 2013. [52] W.C. Silva, E.C.C. Araújo, C.E. Calmanovici, A. Bernardo, M. Giulietti,
[40] Q. Li, Z. Feng, G.P. Rangaiah, L. Dong, Process optimization of heat-integrated Environmental assessment of a standard distillery using aspen plus®: Simulation
extractive dividing-wall columns for energy-saving separation of CO2 and and renewability analysis, J. Clean. Prod. 162 (2017) 1442–1454, https://doi.org/
hydrocarbons, Ind. Eng. Chem. Res. 59 (2020) 11000–11011, https://doi.org/ 10.1016/j.jclepro.2017.06.106.
10.1021/acs.iecr.0c00666.

13

You might also like