a r t i c l e i n f o a b s t r a c t
Article history: Levulinic acid (LA), which is one of the top twelve value-added chemicals from biomass
Received 18 June 2015 feedstock, has been recognized in a large number of applications. Nevertheless, its capability
Received in revised form 31 August on an industrial scale has been limited by the high-cost of the raw materials and the lack of
2015 detailed process design. This paper reports the simulation, detailed design and optimization
Accepted 17 September 2015 of an industrial process recovering LA from biomass, a renewable and inexpensive feedstock.
Available online 28 September 2015 The aqueous feed mixture from the acidic hydrolysis of biomass was fed to the separation
process to recover LA as the main product, along with furfural and formic acid as valuable
Keywords: byproducts. Furfural was then used as an extracting solvent, resulting a signicant decrease
Lignocellulosic biomass in total cost. Detailed analysis was conducted to assess the feasibility of the separation
Levulinic acid method. The purication process was intensied by applying an innovative top dividing
Furfural wall column with a decanter conguration (TDWC-D). The response surface methodology
Energy efciency improvement (RSM), which allows the interactions between variables to be identied and quantied, was
Dividing wall column used to optimize the TDWC-D structure. The predictions by the RSM showed satisfactory
Response surface methodology agreement with the rigorous simulation results. The results showed that the innovative
TDWC-D conguration could save up 16.4% and 20.6% of the overall energy requirements
and total annual cost, respectively, compared to the conventional sequence.
2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
Corresponding author. Tel.: +82 53 810 2512.
E-mail address: mynlee@yu.ac.kr (M. Lee).
1
These authors contributed equally to this work.
http://dx.doi.org/10.1016/j.cherd.2015.09.013
0263-8762/ 2015 The Institution of Chemical Engineers. Published by Elsevier B.V. All rights reserved.
chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136 127
Fig. 1 Dividing wall columns (DWCs): (a) conventional DWC; (b) top DWC; (c) bottom DWC.
materials and energy-intensive processes (Seibert, 2010). of the separation being performed twice, which is the main
Several processes have been used to produce LA from an contributor to their superior energy efciency over other col-
inexpensive feedstock, lignocellulosic biomass. Dunlop and umn congurations (Poth et al., 2004). Additional benets arise
Wells (1957) proposed the earliest continuous process for pro- from the fact that only one reboiler and condenser are involved
ducing LA from corn cob waste. The aqueous feed mixture instead of two reboilers and two condensers in a conventional
obtaining from acidic hydrolysis is introduced to an extractor distillation sequence (Premkumar and Rangaiah, 2009).
to extract LA and then fed to a distillation column to produce To design and optimize DWC, Halvorsen and Skogestad
pure LA. Ghorpade and Hanna (1999) presented the LA pro- (2003) proposed the use of a minimum vapor ow, or the so
cess based on the reactive extrusion from corn starch. These called Vmin diagram method, which is a graphical visualiza-
two processes only considered LA as a product. Recently, tion of the minimum energy. On the other hand, it cannot be
Sadhukhan et al. (2014) presented a good example of a LA applied to the separation of azeotropic mixtures. Several rig-
production process from corn cob, which considered both the orous methods have proposed, such as a sequential quadratic
reaction and separation stages. Seibert (2010) reported a pro- programming-based method implemented in Aspen Plus (Kiss
cess that uses furfural to extract LA and formic acid from and Suszwalak, 2012a), and a MILP approach based on the
the acidic hydrolysis of biomass. The advantage of this pro- use of two superstructure levels (Caballero and Grossmann,
cess is the use of furfural, one of the products, as a solvent. 2014).
Interestingly, these reports showed that a combination of an Recently, Long and Lee (2012) reported a practical and ef-
extractor and a series of distillation columns is used to purify cient method based on the response surface methodology
LA, showing great potential to improve the energy efciency, (RSM). The RSM is a general technique that is used for an
economic performance and environmental impact. Further- empirical study of the relationships between the measured
more, none of the previous processes considered the detailed responses and independent input variables (Box and Wilson,
design methodology and advanced techniques in the case 1951). By allowing the interactions between variables to be
of commercial scale production. A comprehensive design by identied and quantied, RSM can predict reliable results with
intensied congurations with economic evaluation is needed a little computational effort. Therefore, it can be used to opti-
to expand the capacity to an industrial scale as well as to mize the design and operation conditions of a process (Long
reduce the LA price and environmental impact. and Lee, 2012). On the other hand, when the behavior of the
Many studies have examined the relative advantages of a function is highly nonlinear in terms of the design variables,
dividing wall column (DWC) (as shown in Fig. 1) to improve the RSM encounters some difculties in approximating the
the energy efciency and reduce the consumption of utilities sufcient quality on the entire design space (Merval et al.,
(Amminudin and Smith, 2001; Long and Lee, 2013a,b). These 2006). The optimum may not be dened if the range of fac-
reports have shown that DWC systems can achieve energy tors is too narrow or too wide. Furthermore, critical factors
savings of up to 30% over conventional direct and indirect dis- may not be dened or specied correctly. In the case of many
tillation sequences (Kiss and Ignat, 2012; Long et al., 2013). factors, the problem becomes complex and the RSM cannot be
This is because DWCs can allow reversible splits with no part predicted well.
128 chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136
Fig. 2 Ternary diagram for the furfuralwaterformic acid system (mass basis).
This paper proposes an economical and environmental concentrations of the mixture ranged from 3 to 8 wt.% LA, from
process for producing LA from an inexpensive raw mate- 1 to 5 wt.% furfural, from 1 to 5 wt.% formic acid, and the bal-
rial, lignocellulosic biomass, along with the detailed design ance being water (Reunanen et al., 2013). With this component
variables and operating conditions. The simulation work was mixture, although LA did not form an azeotrope with the other
conducted using the commercial simulator, Aspen Plus. A components, two azeotrope mixtures still formed. Fig. 2 shows
range of sequences and intensied congurations applying a ternary diagram for the furfuralwaterformic acid system.
TDWC-D are presented with respect to their energy consump- As shown in this gure, the waterfurfural mixture has a
tion and economic performance. Furthermore, the RSM was minimum-boiling heterogeneous azeotrope with a composi-
used to optimize the structure of the proposed TDWC-D. tion of 64.5 wt.% furfural at 97.8 C. Another maximum-boiling
homogeneous azeotrope forms between formic acid and water
with a higher azeotropic temperature of 106.8 C. In that
2. LA separation process ternary map, there are two distillation regions divided by the
distillation boundary.
2.1. Separation feasibility Fig. 3 presents a simplied feasible ow sheet for producing
LA from lignocellulosic biomass (Seibert, 2010). The conven-
The feed mixture considered was obtained from the acidic tional process sequence consists of one extractor and three
hydrolysis process of biomass and was an aqueous solution distillation columns. The feed mixture and the furfural solvent
containing LA, furfural, formic acid, and water. The typical are introduced into the extractor (EX) to generate a water-rich
Fig. 3 Schematic diagram of the optimal conventional sequence for producing LA from biomass.
chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136 129
Table 1 Conditions of the feed mixture. Table 2 Design parameters of a conventional sequence
for the production of LA from biomass.
Feed conditions
C1 C2 C3
Component Mass fraction (wt.%)
Number of trays 22 27 28
Water 86% Tray type Sieve Sieve Sieve
Formic acid 3% Column diameter (m) 3.6 1.7 1.8
Furfural 4% Number of ow paths 1 1 1
Levulinic acid (LA) 7% Tray spacing (mm) 610 610 610
Temperature ( C) 25 Max ooding (%) 87.83 80.36 85.93
Pressure (atm) 2 Condenser duty (kW) 17,220 2699 3694
Mass owrate (kg/h) 90,000 Reboiler duty (kW) 22,930 4250 3770
k
k
Y = 0 + i Ni + ii Ni2 + ij Ni Nj + (2)
i=1 i=1 i<j
A Box Behnken design was used under the response sur- In the C2 of the conventional sequence, there was an
face methodology to determine how the variables interact amount of furfural loss (2175 kg/h) in the C2-top due to a
and to optimize the system in terms of the TAC saving. Box heterogeneous azeotrope between water and furfural (see
and Behnken (1960) proposed some three-level designs for Fig. 6). As shown in Fig. 2, a signicant liquid phase split
tting the response surfaces. These designs are formed by envelope is clearly shown in the ternary diagram of the
combining 2k factorials with incomplete block designs. The furfuralwaterformic acid system so that a decanter can be
resulting designs are generally very efcient in terms of the applied to reduce the furfural loss and make the distillation
number of required runs, and they are either rotatable or more efcient. In particular, the decanter or gravity settling
nearly rotatable. Rotability is a very important property in the tank, are the simplest form of equipment used to separate
selection of the RSM. Because the purpose of the RSM is opti- liquidliquid mixtures. Several studies have reported the ben-
mization and the location of the optimum is unknown prior ets of a decanter in solving a heterogeneous azeotrope (Kiss
to running the experiment or simulation, it makes sense to and Suszwalak, 2012b; Wu et al., 2014). In this case, furfural
use a design that provides equal estimation precision in all and formic acid are collected as the bottom stream of the
Fig. 9 Schematic diagram of the proposed conventional sequence for producing LA from biomass.
columns, C2 and C3, of the CS sequence. The rigorous sim- Note that medium pressure steam was used on P-C1 and P-
ulation results show that the optimized TDWC-D can save C2, whereas P-C3 used high pressure steam. The simulation
up 30.6% and 28.1% in terms of the operating cost and TAC, results showed that the energy requirement in the reboiler
respectively, compared to the two columns, C2 and C3, of using HP columns is reduced from 22,930 kW in C1 of the CS
the base sequence. Compared to the whole CS sequence, the sequence to 13,850 kW in C3 of the PS sequence. As a result, the
innovative conguration combining TDWC and a decanter can PS sequence that changes the order of the distillation columns
allow operating cost and TAC savings of 7.0% and 7.2%, respec- can reduce the annual operating cost and TAC by up to 16.4%
tively. and 16.0%, respectively, compared to the base case. Note that
these heuristics are not novel but they can be applied easily
3.3. Improving conventional sequence by changing and efciently, which leads to signicant TAC savings in the
sequence proposed sequence (PS) LA process. This process is particularly interesting to many
companies that prefer to use conventional columns.
Multicomponent mixtures can be separated by different dis-
tillation column sequences. Many studies have examined the 3.4. Improving the proposed sequence by decanter
optimal sequence of simple columns that do not involve (PS-D)
heat integration (Nishida et al., 1981; Smith, 2005; Westerberg,
1985). The number of possible distillation sequences increases In the PS sequence, the main role of the rst column (P-C1) is
with increasing number of products (King, 1980). Interestingly, to remove water as much as possible. On the other hand, the
the rule of thumb is the most common method used for nd- minimum-boiling heterogeneous azeotrope mixture between
ing the optimal sequence of simple distillation columns (King, furfural and water forms, causing a furfural loss of 3196 kg/h
1980). The sequences that remove the lightest components in the top of this column. Therefore, a decanter, which can
alone one by one in the column overheads should be favored split the liquidliquid mixture naturally, can be integrated in
(Smith, 2005), i.e., favor the direct sequence. In addition, in the this column as same as the second column in the base case
base case, the rst distillation column C1 requires the largest to reduce the loss of furfural. Fig. 10 shows the simplied
amount of energy and highest level of steam (high pressure ow sheet of the P-C1 integrated with a decanter. To improve
steam) compared to the other columns. This column is used the separation efciency in decanter, the overhead of P-C1
to vaporize a large number of furfural and light components as was cooled to 60 C before being introduced to the decanter.
well as purify LA, which has the highest boiling point (257 C) Not only can the furfural loss be reduced by up to 87.9%, but
compared to the other components, furfural (162 C), formic also the amount of water removed increases from 2960 kg/h
acid (101 C), water (100 C). Based on the well-known heuris- to 3157 kg/h, which is comparable to the column without a
tics for separation sequence synthesis, a proposed sequence decanter. The simulation results showed that not only can
for producing LA was used to enhance the energy efciency the energy requirement in the reboiler duty in P-C1-D column
(see Fig. 9). be reduced slightly (1.7%), but there is also less furfural loss
An aqueous mixture from the acidic hydrolysis reaction of compared to the column without a decanter.
biomass and the furfural solvent was fed to the extractor (P-EX)
to generate a water-rich phase as the top stream and solvent- 3.5. Intensifying proposed sequence with innovative
rich phase containing mainly furfural, LA and formic acid as TDWC-D (PS-TDWC-D)
the bottom stream. In contrast to the CS sequence, the bot-
tom stream of P-EX was fed to the proposed C1 (P-C1), which In the CS sequence, although the C1 requires most energy,
delivered a light boiling azeotrope of furfural and water as however, TDWC-D can only be used to replace the C2 and
the top stream, whereas the bottom stream contained formic C3. On the other hand, in the PS sequence, both the P-C1
acid, furfural and LA. The bottom of P-C1 was then introduced and P-C3 are energy-intensive columns so that intensication
to the proposed C2 (P-C2) to collect formic acid as the distil- either the P-C1 or P-C3 can bring more benet. In particular,
late and the bottom stream comprising furfural and LA, which to intensify the proposed conventional sequence, the TDWC
was subjected to the proposed C3 (P-C3). The role of P-C3 is was used to integrate the rst two columns. To optimize the
to separate levulinic acid as the bottom product and furfural TDWC-D, some preliminary simulation runs were carried out
as the top product. Here, LA is recovered at the third col- to determine the main optimizing variables and their levels.
umn so the proposed sequence is called the direct sequence. Corresponding to the changes in the factor value, the response
chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136 133
values (TAC saving) were recorded. The presence of curvature As shown in Fig. 12, the TAC saving was predicted to be
suggests that the simulation region is close to the optimum. highest (15.3%) with the coded levels of the top section (N1),
Table 4 lists the factors and levels used in the RSM-based opti- bottom section (N2) and feed rectifying section (N3) of 0.6970,
mization. The TDWC-D parameters were optimized over 15 0.5152 and 0.9798, respectively. The resulting determination
simulation runs. For each run, the internal vapor ow to the coefcient was R2 = 95.7, which indicates that the model is
prefractionator was varied to maximize the TAC savings while suitable for prediction within the range of simulated vari-
obtaining the required product purity and recovery. Fig. 11 ables. The natural values of the variables can be derived from
presents a contour plot of the interactions between the design the coded levels. Recycling of the vapor was then optimized
variables N1, N2 and N3. In each case, the one remaining to maximize the TAC savings while achieving the product
parameter was set automatically as its center point value by purity requirements and recovery. Fig. 13 compares the three-
the software. The resulting second-order polynomial model dimensional response surface plots of the interaction between
was as follows: N1 and N2 evaluated by the RSM and by rigorous simulation.
The gure shows that the three-dimensional response surface
Y = 10.3743 + 5.44663N1 + 3.99257N2 4.11248N3 plots by the RSM are similar to the actual ones. Fig. 14 shows a
schematic diagram illustrating the optimal TDWC-D from the
= 3.01056N12 5.15933N22 1.55725N32 + 1.2084N1 N2
conventional sequence. The TDWC-D combining a decanter,
+ 1.96654N1 N3 0.582208N2 N3 (3) which can increase the furfural recovery and improve the
Fig. 11 Contour plots of the interactions between the top section (N1), bottom section (N2) and feed rectifying section (N3).
134 chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136
Fig. 13 Three-dimensional response surface plots of the interaction between N1 and N2 predicted by the RSM (a) and by
the rigorous simulation (b).
Fig. 14 Schematic diagram of the proposed sequence intensied by the innovative TDWC-D.
heterogeneous azeotrope separation, with a TDWC, which can (15.3%). In addition, using a TDWC-D could save 15.4% of the
reduce energy requirement and TAC, can bring many benets energy requirements in terms of the reboiler duty compared
in terms of energy saving and cost reduction. The simulation to P-C1 and P-C2 of the PS sequence.
showed that the TDWC-D can save up to 13.2% in terms of the Table 5 summarizes the key results, while Fig. 15 con-
TAC compared to the C2 and C3 of the conventional sequence. veniently illustrates the annual operating cost and TAC for
This rigorous simulation agreed well with the predicted value all congurations considered. The results suggest that the
Table 5 Comparison of the different structural alternatives for improving the performance of process for producing LA
from biomass.
Structural alternative CS CS-D CS-TDWC-D PS PS-D PS-TDWC-D
Annual operating cost (US $) 16,332,077 16,234,472 15,151,046 13,646,300 13,729,322 12,812,024
Annual operating cost saving (%) 0.6% 7.2% 16.4% 15.9% 21.6%
Total annual cost (US $) 17,308,411 17,214,679 16,103,366 14,541,721 14,640,722 13,736,949
Total annual cost saving (%) 0.5% 7.0% 16.0% 15.4% 20.6%
chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136 135
(6)
where
L D
BC is the bare cost, for vessels : BC = BC0
Fig. 15 Comparison of the different LA recovery processes L0 D0
in terms of the annual operating cost and TAC. (8)
4. Conclusions Q
Area of the heat exchanger, A = (10)
UT
A levulinic acid recovery process from abundant and inexpen-
where MPF is the material and pressure factor; MF is the mod-
sive lignocellulosic biomass was designed and optimized to
ule factor (typical value), which is affected by the base cost. D,
be more energy-efcient. This process was then intensied
L and A are the diameter, length and area, respectively.
efciently using an innovative TDWC-D. After a preliminary
design, the TDWC-D structure was then optimized and imple-
Updated bare module cost for the tray stack (BMC)
mented simply and efciently by the RSM using Aspen Plus
and Minitab. The simulated results were in good agreement = UF BC MPF (11)
with the predicted values. The use of innovative TDWC-D can
have many benets, such as increasing the furfural recovery,
increasing the separation efciency and decreasing the TAC The material and pressure factor : MPF = Fm + Fs + Fm (12)
signicantly. The incorporation of a decanter into the TDWC
is a very attractive option for improving the separation ef- (b) Operating cost (Op):
ciency and energy savings, especially for the separation of
heterogeneous azeotrope mixtures. The proposed sequence Op = Csteam + CCW (13)
including TDWC-D showed 21.6% and 20.6% decreases in the
energy requirement and TAC, respectively, compared to a con- where Csteam is the cost of the steam and CCW is the cost of
ventional column sequence. Based on these results, the use cooling water.
of TDWC-D for LA production, which has the limitations in (c) Total annual cost (TAC) (Smith, 2005):
commercial scale of existing processes currently, is technically n
i(1 + i)
feasible and particularly interesting in constructing a new lev- TAC = capitalcost n + Op (14)
(1 + i) 1
ulinic plant. The process compactness is an additional benet
from TDWC-D applications.
where i is the fractional interest rate per year and n is the
number of years.
Acknowledgments
References
The authors wish to acknowledge the nancial support from
the R&D Convergence Program of NST (National Research
Amminudin, K.A., Smith, R., 2001. Design and optimization of
Council of Science & Technology) of Republic of Korea and fully thermally coupled distillation columns: Part 2:
KITECH (Korea Institute of Industrial Technology) (ES150001). Application of dividing wall columns in retrot. Chem. Eng.
Res. Des. 79, 716724.
Appendix. Column cost correlations Aspen Technology, 2010. Aspen Plus: User Guide, vols. 1 & 2.
Biegler, L.T., Grossmann, I.E., Westerberg, A.W., 1997. Systematic
Methods of Chemical Process Design. Prentice Hall Inc., New
(a) Capital cost: Guthries modular method was applied (Biegler Jersey.
et al., 1997). The investment cost for conventional distilla- Box, G.E.P., Behnken, D.W., 1960. Some new three level designs for
tion is the total cost of the column and auxiliary equipment, the study of quantitative variables. Technometrics 2, 455476.
such as reboilers and condensers. For the DWC, it includes Box, G.E.P., Wilson, K.B., 1951. On the experimental attainment of
the additional dividing wall cost. In this study, the Chemical optimum conditions. J. R. Stat. Soc. Ser. B 13, 145.
Burke, R.A., 2013. Hazardous Materials Chemistry for Emergency
Engineering Plant Cost Index of 585.7 (2011) was used for cost
Responders, third ed. CRC Press.
updating. Caballero, J.A., Grossmann, I.E., 2014. Optimal synthesis of
thermally coupled distillation sequences using a novel MILP
Tray stack = (N 1) tray spacing (4) approach. Comput. Chem. Eng. 61, 118135.
136 chemical engineering research and design 1 0 7 ( 2 0 1 6 ) 126136
Corma, A., Iborra, S., Velty, A., 2007. Chemical routes for the and deisobutanizing fractionation of NGL. Korean J. Chem.
transformation of biomass into chemicals. Chem. Rev. 107, Eng. 30, 286294.
24112502. Long, N.V.D., Kwon, Y., Lee, M.Y., 2013. Design and optimization of
Dunlop, A.P., Wells, P.A., 1957. Process for Producing Levulinic thermally coupled distillation schemes for the trichlorosilane
Acid. US Patent 2,813,900. purication process. Appl. Therm. Eng. 59, 200210.
Ghorpade, V.M., Hanna, M.A., 1999. Method and Apparatus for Merval, A., Samuelides, M., Grihon, S., 2006. Application of
Production of Levulinic Acid via Reactive Extrusion. US Patent response surface methodology to stiffened panel
5,859,263. optimization. In: 47th conference on
Grand View Research, 2014. Levulinic Acid Market Analysis and AIAA/ASME/ASCE/AHS/ASC Structures, Structural Dynamics,
Segment Forecasts to 2020., ISBN 978-1-68038-068-2. and Materials Conference, Newport, Rhode Island.
Halvorsen, I.J., Skogestad, S., 2003. Minimum energy 2010. Minitab 16 Statistical Software (Computer Software).
consumption in multicomponent distillation. 2: Minitab, Inc., State College, PA (www.minitab.com).
Three-product Petlyuk arrangements. Ind. Eng. Montgomery, D.C., 2001. Design and Analysis of Experiments, 5th
Chem. Res. 42, 605615. ed. John Wiley & Sons, New York, pp. 427510.
Hayes, D.J., Fitzpatrick, S., Hayes, M.H.B., Ross, J.R.H., 2008. The Nishida, N., Stephanopoulos, G., Westerberg, A.W., 1981. A review
Bione Process Production of Levulinic Acid, Furfural, and of process synthesis. AIChE J. 27, 321351.
Formic Acid from Lignocellulosic Feedstocks Petrus, L., Noordermeer, M.A., 2006. Biomass to biofuels a
Bioreneries-Industrial Processes and Products. Wiley-VCH chemical perspective. Green Chem. 8, 861867.
Verlag GmbH, pp. 139164. Poth, N., Stichlmair, J., Brusis, D., 2004. Minimaler Energiebedarf
Huber, G.W., Iborra, S., Corma, A., 2006. Synthesis of von Trennwandkolonnen. Chem. Ing. Tech. 76, 18111814.
transportation fuels from biomass: chemistry, catalysts, and Premkumar, R., Rangaiah, G.P., 2009. Retrotting conventional
engineering. Chem. Rev. 106, 40444098. column systems to dividing-wall columns. Chem. Eng. Res.
King, C.J., 1980. Separation Processes, second ed. McGraw-Hill, Des. 87, 4760.
New York. Reunanen, J., Oinas, P., Nissinen, T., 2013. A Process for Recovery
Kiss, A.A., Ignat, R.M., 2012. Enhanced methanol recovery and of Formic Acid. US Patent 8,530,695 B2.
glycerol separation in biodiesel production DWC makes it Sadhukhan, J., Siew, N.K., Martinez, E.H., 2014. Bioreneries and
happen. Appl. Energy 99, 146153. Chemical Processes, Design Integration and Sustainability
Kiss, A.A., Suszwalak, D.J.P.C., 2012a. Innovative dimethyl ether Analysis, rst ed. John Wiley & Sons, United Kingdom.
synthesis in a reactive dividing-wall column. Comput. Chem. Seibert, F., 2010. A Method of Recovering Levulinic Acid. WIPO
Eng. 38, 7481. Patent WO/2010/030617 A1.
Kiss, A.A., Suszwalak, D.J.P.C., 2012b. Enhanced bioethanol Smith, R., 2005. Chemical Process Design and Integration. John
dehydration by extractive and azeotropic distillation in Wiley & Son, West Sussex, England.
dividing-wall columns. Sep. Purif. Technol. 86, Turton, R., Bailie, R.C., Wallace, B.W., Joseph, A.S., Debangsu, B.,
7078. 2012. Analysis, Synthesis, and Design of Chemical Processes,
Koh, L.P., Ghazoul, J., 2008. Biofuels, biodiversity, and people: fourth ed. Prentice Hall.
understanding the conicts and nding opportunities. Bio. Westerberg, A.W., 1985. The synthesis of distillation-based
Conserv. 141, 24502460. separation systems. Comput. Chem. Eng. 9, 421429.
Long, N.V.D., Lee, M.Y., 2012. Dividing wall column structure Werpy, T., Petersen, G., 2004. Top Value Added Chemicals from
design using response surface methodology. Comput. Chem. Biomass Volume I Results of Screening for Potential
Eng. 37, 119124. Candidates from Sugars and Synthesis Gas. National
Long, N.V.D., Lee, M.Y., 2013a. Optimal retrot of a side stream Renewable Energy Laboratory (NREL).
column to a dividing wall column for energy efciency Wu, Y.C., Lee, H.Y., Huang, H.P., Chien, I.L., 2014. Energy-saving
maximization. Chem. Eng. Res. Des. 91, 22912298. dividing-wall column design and control for heterogeneous
Long, N.V.D., Lee, M.Y., 2013b. Design and optimization of heat azeotropic distillation systems. Ind. Eng. Chem. Res. 53 (4),
integrated dividing wall columns for improved debutanizing 15371552.