Applied Energy 99 (2012) 146–153

Contents lists available at SciVerse ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Enhanced methanol recovery and glycerol separation in biodiesel

production – DWC makes it happen
Anton A. Kiss a,⇑, Radu M. Ignat b
a
AkzoNobel Research, Development & Innovation, Process Technology ECG, Zutphenseweg 10, 7418 AJ Deventer, The Netherlands
b
University ‘‘Politehnica’’ of Bucharest, Department of Chemical Engineering, Polizu 1-7, 011061 Bucharest, Romania

a r t i c l e i n f o a b s t r a c t

Article history: Biodiesel is a renewable fuel that consists of fatty acids methyl esters – currently produced by trans-
Received 31 January 2012 esterification of glycerides with methanol. After the biodiesel synthesis, the downstream processing steps
Received in revised form 30 March 2012 involve the purification of crude glycerol, as well as the separation of excess methanol (recyclable), glyc-
Accepted 12 April 2012
erol by-product and water (from washing and pre-treatment steps). The separation of the ternary mix-
Available online 4 June 2012
ture methanol–water–glycerol is carried out in a conventional direct sequence that requires two
distillation columns and rather high amounts of energy.
Keywords:
This study proposes an efficient process intensification method for this ternary separation, namely the
Dividing-wall column
Biodiesel purification
use of a dividing-wall column (DWC) that is able to separate all products at high purity, in only one
Design and control equipment unit. AspenTech Aspen Plus and Aspen Dynamics were used as computer aided process engi-
Process intensification neering tools to perform the rigorous steady-state and dynamic simulations, as well as the optimization
Capital and energy savings of the new DWC separation alternative. In order to allow a fair comparison, all designs analyzed here
were optimized using the sequential quadratic programming (SQP) method. Remarkable, the results
show that the proposed DWC system requires 27% less energy and 12% lower investment costs, thus hav-
ing a significant contribution towards inexpensive biodiesel production.
Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction
Renewable energy sources are developed to tackle the shortage
of fossil fuels, the volatile oil prices and the increased greenhouse
gas emissions. At present, employing waste and non-edible raw
materials is mandatory to comply with the ecological and ethical
requirements for biofuels. Biodiesel is a renewable fuel consisting
of fatty acid methyl esters (FAME), currently produced from green
sources such as vegetable oils, animal fat or waste cooking-oils [1–
4]. The most widespread manufacturing technologies use homoge-
neous catalysts, in processes where both reaction and separation
steps can create bottlenecks. Although reactive separations were
recently proposed for the esterification of free fatty acids (FFA) to
FAME [5–9], the vast majority of existing biodiesel plants are based
on trans-esterification of tri-glycerides leading to biodiesel and
glycerol by-product [10–12]. The biodiesel production by trans-
esterification (Fig. 1) requires several expensive downstream pro-
cessing steps such as the purification of biodiesel, as well as the
separation of excess methanol, glycerol by-product and water –
from the washing and FFA pre-treatment steps [13–17]. In practice,
the separation of the ternary mixture methanol–water–glycerol is
⇑ Corresponding author. Tel.: +31 26 366 9420.
E-mail addresses: Tony.Kiss@akzonobel.com, tonykiss@gmail.com (A.A. Kiss).
carried out in a conventional direct sequence that requires two dis-
tillation columns and plenty of energy (Fig. 2 top). Consequently,
reducing the energy requirements is not only a challenge but also
a great opportunity for improving the operation and economics of
biodiesel processes [18,19].
Distillation is still the most important thermal separation tech-
nology, but in spite of its widespread use and major benefits, a key
drawback is its high energy demand generating over 50% of plant
operating costs [20]. The last decades led to energy efficient distil-
lation solutions based on process integration and intensification
techniques, such as: cyclic distillation, heat-integrated distillation
column, reactive distillation, and thermally coupled columns
[20–23,7,8,24,25].
Dividing-wall column (DWC) – used for ternary separations – is
a practical implementation of the Petlyuk configuration, consisting
of a pre-fractionator and a main distillation column. A vertical wall
is inserted in a DWC at a position such that it splits the column
shell into two sections [23,24]. The feed stream is introduced into
the pre-fractionator (feed side of the column) while a side stream
is removed from the main column (side stream section). The light-
est component goes overhead as distillate and the heaviest compo-
nent goes out as bottoms, while the side stream contains only the
intermediate boiling component. At the top of the wall the refluxed
liquid is split between the two sides of the column, while at the
bottom the vapor flow is split according to the pressure drop on

0306-2619/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.apenergy.2012.04.019
 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153
Oil/fat
Alcohol/ Reactor
catalyst Separator
Alcohol
Alcohol
recovery
column
Glycerol + Water
Fig. 1. Process flow diagram of a conventional biodiesel production process based on trans-esterification of tri-glycerides.
both sides of the wall (Fig. 2 btm). The DWC technology is very
appealing to industry as it confers major benefits over conven-
tional distillation: up to 30% reduced investment costs and up to
40% energy savings [24]. Remarkable, these DWC benefits are not
limited to ternary separations alone, but they can be present also
in azeotropic and extractive distillation [26–28], as well as reactive
distillation [27–31].
Thermally coupled distillation columns and DWC were re-
cently proposed for the biodiesel production, but mainly for the
synthesis of FAME by esterification of FFA in reactive distillation
systems [30,32,33] and not for the downstream processing steps.
In order to reduce the number of distillation equipment units and
the energy requirements for the current biodiesel purification,
this study proposes the use of a DWC for a single step methanol
recovery and glycerol separation. Aspen Plus and Aspen Dynamics
were used to perform rigorous steady-state and dynamic simula-
tions, as well as the optimization of the new DWC separation
alternative. In order to allow a fair comparison, all designs were
optimized using the sequential quadratic programming (SQP)
method.
Based on a solid theoretical and computational foundation, the
sequential quadratic programming (SQP) method has become the
most successful method for solving nonlinearly constrained opti-
mization problems [34–39]. Nowadays, there are a large variety
of SQP techniques, from the initial approaches introduced in the
early 1960s, to the more state-of-the-art techniques, e.g. based
on the multiple shooting techniques. These SQP approaches have
already proved their advantages in many areas of applications in
chemical engineering, such as model predictive control and dy-
namic optimization, among others [40–46].
2. Problem statement
The biodiesel production by trans-esterification requires several
expensive downstream processing steps such as the purification of
crude glycerol, as well as the separation of excess methanol (recy-
clable stream), glycerol by-product and water (from the washing
and FFA pre-treatment steps). The problem is that the separation
of the ternary mixture methanol–water–glycerol is still carried
147
Biodiesel
Dryer
Biodiesel
Wash
column
Water
Glycerol
+ Alcohol
+ Water
Water
Glycerol
recovery
Glycerol column
+ Alcohol
+ Water
Glycerol
out in a conventional direct sequence that requires two distillation
columns (Fig. 2 top), a relatively large footprint and significantly
high amounts of energy. To solve this problem, this study proposes
an efficient process intensification method for this ternary separa-
tion, namely the use of a dividing-wall column (DWC) that is able
to separate all products at high purity, in only one equipment unit
(Fig. 2 btm).
3. Results and discussion
In this study we consider the separation of the ternary mixture
methanol–water–glycerol as described in the literature [13–16].
For the base case (direct distillation sequence) and the DWC alter-
native the feed basis is 2900 kg/h – equivalent to a biodiesel pro-
duction rate of about 100 ktpy. The target product purity
considered for each product cut is min. 99.5 %wt. Note that the
reactants and product amounts for trans-esterification are in the
mass ratio of 100:22:11 for biodiesel:methanol:glycerol, meaning
that 11 kg of glycerol is synthesized per each 100 kg of biodiesel
produced [13]. Moreover, the amount of methanol introduced in
the reactor typically exceeds 50% of the required stoichiometric ra-
tio, thus this must be recovered and recycled. It is also important
noting that the composition of the feed mixture can vary, depend-
ing on the type of the biodiesel production process – alkali or acid
catalyzed – and the amount of water produced in the FFA pre-
treatment step and also used for the biodiesel washing [13–17].
Typically, the weight amounts of methanol and glycerol are simi-
lar, while water is present in lower quantities [13,16].
Steady-state simulations were carried out in AspenTech Aspen
PlusÒ using the rigorous RADFRAC unit enhanced with the RateSep
(rate-based) model. The well-known MESH equations are govern-
ing the process. Note that MESH is an acronym referring to the type
of equations: M – mass balance, E – equilibrium relationships, S –
summation equations, and H – enthalpy balance. Due to the nature
of the components involved in the separation, UNIQUAC was se-
lected as the most accurate property model [30,9]. It is worth not-
ing that the mixture does not form any azeotropes, and there is no
liquid phase splitting possible at any given composition. Conse-
quently, the residues curve maps and the ternary diagrams are
148 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153

Table 1
Design parameters of an optimal direct sequence of two columns.

Methanol Water Design parameters Value Unit

Flowrate of feed stream 2900 kg/h
Feed composition (mass fractions)
Methanol:Water:Glycerol 0.473:0.054:0.473 –
Temperature of feed stream 60 °C
Pressure of feed stream 1.2 bar
Feed Operating pressure (DC1) 0.5 bar
DC 1 DC 2
Operating pressure (DC2) 0.5 bar
Column diameter (DC1) 0.9 m
Column diameter (DC2) 0.4 m
Total number of stages (DC1) 14 –
Total number of stages (DC2) 12 –
Feed stage (DC1) 9 –
Feed stage (DC2) 7 –
DC1-Btm Glycerol Reflux ratio (DC1) 1.4 kg/kg
Reflux ratio (DC2) 0.2 kg/kg
Distillate to feed ratio (DC1) 0.473 kg/kg
Distillate to feed ratio (DC2) 0.1 kg/kg
Methanol product purity 99.9/99.9 %wt/%mol
Water product purity 99.6/99.8 %wt/%mol
Glycerol product purity 99.9/99.9 %wt/%mol
Reboiler duty DC1 1080 kW
Methanol Condenser duty DC1 1029 kW
Reboiler duty DC2 250 kW
rL Condenser duty DC2 120 kW

3.1. Direct sequence

Water
Fig. 2 (top) shows the direct distillation sequence as the conven-
Feed tional industrial practice for methanol recovery and glycerol sepa-
ration [13–16]. Table 1 provides the optimal design parameters for
this base case, while Fig. 3 conveniently plots the temperature and
liquid composition profiles. Methanol is collected as top distillate
rV from the first distillation column (DC1), while the bottom product
– consisting of water and glycerol – is fed to the second distillation
DWC
column (DC2). The recovered methanol is typically recycled back to
the trans-esterification reactor. High purity glycerol is obtained as
Glycerol
bottom product, and water as distillate product of DC2, respec-
tively. Both distillation columns are operated at 0.5 bar in order
to keep the temperature in the reboiler sufficiently low to prevent
Fig. 2. Methanol–water–glycerol separation in a direct sequence (top) and a
the degradation of glycerol.
dividing-wall column (btm).

3.2. DWC alternative

not included here since they do not provide any added value
information.
Both configurations (base case and DWC alternative) described
hereafter were optimized in terms of minimal energy demand
using the sensitivity analysis and sequential quadratic program-
ming (SQP) tools from Aspen Plus [47,39]. Note that the SQP opti-
mization method implemented in Aspen Plus is a state-of-the-art,
quasi-Newton nonlinear programming algorithm – while the SQP-
Biegler variant is the SQP implementation developed by Biegler
et al. [48,49]. The default optimization procedure in Aspen Plus is
to converge tear streams and the optimization problem simulta-
neously, using the SQP method. When the SQP method is used to
converge tears and optimization problems simultaneously, the
algorithm is a hybrid of an infeasible path method where the tears
are not converged at each iteration but are converged at the opti-
mum, and a feasible path method where the tears are converged at
each iteration of the optimization [47]. In practice, minimizing the
total heat duty of the distillation sequence, constrained by the re-
quired purity and recovery of the products, was carried out com-
bining the sensitivity analysis and SQP tools from Aspen Plus.
Several optimization variables were used: total number of stages,
feed-stage location, side-stream location, location and length of
dividing wall, reflux ratio, liquid and vapor split [27,28].
The DWC was designed according to the rules described in the
literature [50,23,51,24,52]. In order to design of the most energy
efficient DWC we use the concept of minimum energy or minimum
vapor flow rate (Vmin). Plotting the Vmin diagrams pioneered by
Halvorsen and Skogestad [53,52] contributes to the understanding
of the optimal operating point with the minimum energy require-
ments. Sections with infinite number of stages are assumed, thus
the numbers given in the Vmin diagram are independent of the de-
tailed stage design. In practice, the real number of stages is se-
lected based on the required separation purity, while taking into
account the trade-off between the equipment cost and the energy
cost. Clearly, the real energy input is slightly higher than the min-
imum requirements. Nevertheless, Vmin is a straightforward and
common measure used to compare the energy consumption in dif-
ferent distillation arrangements [53,52]. Fig. 4 presents the Vmin
diagram for the ternary mixture methanol–water–glycerol – noted
here as A, B, C for convenience. The minimum amount of energy re-
quired for the separation in a DWC is given by the highest peak in
the diagram (PA/B).
A sequence of two rigorous RADFRAC units was used to model
the DWC as no off-the-shelf DWC unit is available in the current
commercial process simulators. This configuration is thermally
equivalent to a DWC, as long as the temperature difference on both
 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153
DC1
DC2
Fig. 3. Temperature and composition profile in DC1 and DC2 units of the direct
distillation sequence.
sides of the wall indicates that there is no heat transfer between
the two sides. The SQP optimization method and the efficient sen-
sitivity analysis tool from Aspen Plus were also employed. The
objective of the optimization is to minimize the total reboiler duty
required, as follows:
MinðQ Þ ¼ f ðNT ; NF ; NSS ; NDWS ; N DWC ; V; RR; r V ; rL Þ
ð1Þ
Subject to ~ym P ~ xm
where the optimization parameters used here are: total number of
stages (NT), feed location (NF), side-draw stage (NSS), wall size
(NDWS) and location (NDWC), boilup rate (V), reflux ratio (RR), liquid
149
Fig. 4. Vmin diagram of the methanol–water–glycerol system.
Fig. 5. Temperature and composition profile in the dividing-wall column.
and vapor split (rL and rV), while ym and xm are the vectors of the
obtained and required purities for the m products [27,28]. Note that
in order to determine the optimal ratio between the energy cost and
the number of stages, an additional objective function was used,
Min NT (RR + 1), which approximates very well the minimum of to-
tal annualized cost of a conventional distillation column [54].
Since both distillation columns of the direct sequence (Fig. 2
top) are operated at the same pressure, DWC (Fig. 2 btm) can be
an attractive alternative for the industrial implementation. Metha-
nol is recovered as top distillate, glycerol as bottom product and
water is withdrawn as side stream of the main column. As starting
point of the DWC simulation, the results obtained from the direct
sequence Aspen Plus model were used. In fact, these results
150 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153

Table 2 3.3. Dynamics and control

Design parameters of an optimal dividing-wall column.

Design parameters Value Unit DWC technology offers indeed key economic benefits, but the
Flowrate of feed stream 2900 kg/h controllability of the process is just as important as the savings
Feed composition (mass fractions) in the capital and operating costs. Fig. 6 illustrates the proposed
Methanol:Water:Glycerol 0.473:0.054:0.473 – control structure: two multi loops are needed to stabilize the col-
Temperature of feed stream 60 °C umn and another three to maintain the set points specifying the
Pressure of feed stream 1.2 bar
Operating pressure 0.5 bar
product purities. The regulatory control focuses on maintaining
Column diameter 1.1 m the specified product purities, while the inventory control aims to
Number of stages pre-fractionator side 10 – stabilize the column and to control the level in the reflux tank
Total number of stages DWC 30 – and the level in the reboiler. The level of the reflux drum and the
Feed stage pre-fractionator 7 –
reboiler can be controlled by the manipulated variables D (distil-
Side stream withdrawal stage 22 –
Wall position (from/to stage) 15–25 – late), and B (bottoms), respectively. The composition of the free
Distillate to feed ratio 0.473 kg/kg product streams is controlled by the remaining variables: L (liquid
Reflux ratio 0.83 kg/kg reflux), S (side-stream) and V (vapor boil-up). Note that the control
Liquid split ratio (rL) 0.42 kg/kg structure DB/LSV was reported as the best option for a DWC in our
Vapor split ratio (rV) 0.25 kg/kg
previous studies [55,56,7,8].
Methanol product purity 99.9/99.8 %wt/%mol
Water product purity 99.6/99.8 %wt/%mol An additional optimization loop is used to manipulate the liquid
Glycerol product purity 99.9/99.9 %wt/%mol split, in order to control the heavy component composition in the
Reboiler duty 975 kW top of pre-fractionator (PF), and implicitly achieving minimization
Condenser duty 793 kW
of the energy requirements [53,52,57,56,58]. Note that any heavy
component (glycerol) going out the top of the wall will end up also
in the liquid flowing down in the main column and thus strongly
affecting the purity of the side stream (S). Since the side stream
is collected as a liquid product, it means that any small amounts
of light impurity in the vapor phase will not significantly affect
LC
Methanol its composition. However, even tiny amounts of heavy impurity
in the liquid phase will greatly affect the composition of the side
rL
CC stream.
CC
At the bottom of the dividing-wall section, the vapor flow is
YG split proportionally to the cross-sectional area of each side and
the hydrodynamic conditions (e.g. flow resistance). The cross-sec-
Water
tional area of each side is fixed by the physical location of the wall,
Feed and this is already set at the design stage hence it cannot be chan-
CC
ged later on, during operation [23]. Because the location of the wall
fixes how the vapor flow splits between the two sides of the col-
rV umn, the vapor split (rV) variable is not adjustable during operation
for control purposes [7,8]. Therefore, the vapor split ratio is not
CC used as a manipulated variable in the dynamic simulations pre-
DWC
sented hereafter.
LC Glycerol Table 3 lists the tuning parameters of the PID controllers. The
control loops were tuned by a simple version of the direct synthe-
sis method [59]. Fig. 7 presents the results of the dynamic simula-
Fig. 6. Control structure of the DWC unit (4-point control: DB/LSV + rL). tions for various industrially relevant disturbances in the set point
(SP), the feed flow rate (F) and feed composition. The mass frac-
provide the initial estimates for the total number of trays, the feed tions of all components are returning to their set point within rea-
tray location, as well as the liquid and vapor split. The optimal de- sonable short times, thus proving that the system can successfully
sign of the DWC was obtained according to the method described reject the disturbances. Moreover, the dynamic response of the
earlier. Fig. 5 plots the temperature and liquid composition profiles proposed control structure is clearly characterized by low over-
in the DWC unit, while the key parameters of the optimal design shooting and short settling times.
are presented in Table 2. Remarkable, the temperature difference
between the two sides of the wall is very low – less than 25 °C 3.4. Process comparison
– such conditions being easily achievable in the practical applica-
tion with little heat transfer expected and negligible effect on the In order to perform a fair comparison of the two process alter-
column performance [23,24]. natives, the total investment cost (TIC), total operating cost (TOC)

Table 3
Tuning parameters of the PID controllers.

Controlled variable Manipulated variable Gain P (%/%) Int. time I (min) Drv. time D (min) Control direction
xmethanol (in D) L 0.8 40 0 
xwater (in S) V 0.6 20 0 
xglycerol (in B) S 0.6 40 0 
yG (at top PF) rL 1 20 0 +
Level reflux drum D 1 100 0 +
Level reboiler B 1 100 0 +
 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153 151

Table 4
Head-to-head comparison of conventional vs. DWC alternative.

Key performance indicators Conventional DWC

process alternative
Total investment cost (TIC) $563,429 $499,087
Total operating costs (TOCs) $280,491 $208,402
Total annual costs (TACs) $336,834 $258,310
Specific energy requirements 967.5 709.3
(kW h/ton glycerol)
CO2 emissions (kg CO2/h ton glycerol) 135.31 99.19

and total annual cost (TAC) were calculated. The equipment costs
are estimated using correlations from the Douglas textbook to
the price level of 2010, as described by Dejanović et al. [54]. The
Marshall and Swift equipment cost index (MSI) considered here
has a value of 1468.6. For a carbon steel column, the estimated cost
in US$ is given by the relation:
1:066 0:802
C shell ¼ fp ðMSI=280Þdc hc ð2Þ

where fp is the cost factor (equal to 2981.68 in this case), dc is the

column diameter (calculated using the internals-sizing procedure
from Aspen Plus) and hc its height (tangent-to-tangent) considering
a tray-spacing of 0.6 m. For heat exchangers (e.g. condensers and
reboilers) the next expression was used to calculate the equipment
cost (US$):

C hex ¼ ðMSI=280Þcx A0:65 ð3Þ

where cx = 1609.13 for condensers and 1775.26 for kettle reboilers,

while A is the heat transfer area (m2). Also, a price of 600 US $/m2
was used for the sieve trays cost calculations.
The following utility costs were considered: US $0.03/t cooling
water and US $13/t steam. For the TAC calculations, a plant lifetime
of 10 years was considered. While the accuracy of the correlations
is in the range of acceptable and realistic ±30%, this level of accu-
racy is less important when comparing design alternatives since
the error is consistent in all cases.
Table 4 provides a head-to-head comparison of the key perfor-
mance economic indicators, while Fig. 8 conveniently illustrates
the costs and the CO2 emissions of the two processes considered.
Remarkable, the DWC alternative is the most efficient in terms of
energy requirements allowing energy savings of 27% while also
being the least expensive in terms of capital investment and oper-
ating costs, leading to 25% lower total annual costs.
The energy requirements are closely linked to the CO2 emis-
sions, but only when no heat integration is considered. When part
Methanol of the process heat is reused instead of primary energy, then the
Water CO2 emissions are lower as compared to the figure expected from
Glycerol the energy data. The CO2 emissions were calculated according to
method described by Gadalla et al. [60]:
½CO2 emissions ¼ ðQ fuel =NHVÞðC%=100Þa ð4Þ

where a = 3.67 is the ratio of molar masses of CO2 and C, NHV is the
net heating value, and C% is the carbon content – dependent on the
fuel. For natural gas, NHV is 48,900 kJ/kg and the carbon content is
0.41 kg/kg. Hence the amount of fuel used can be calculated as
follows:

Q fuel ¼ Q proc =kproc  ðhproc  419ÞðT FTB  T 0 Þ=ðT FTB  T stack Þ ð5Þ

where kproc (kJ/kg) and hproc (kJ/kg) are the latent heat and enthalpy
Fig. 7. Results of the dynamic simulations at various disturbances in the set point of the steam, TFTB (K) and Tstack (K) are the flame and stack temper-
(change from 0.995 to 0.999), feed flowrate (+10% F) and feed composition ature, respectively.
(±10%  water). The hourly rate of CO2 emissions for the conventional and DWC
152 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153
Fig. 8. Comparison of conventional and DWC alternatives in terms of key
performance indicators: CO2 emissions, total investment, operating and annual
costs.
alternatives are 183.19 kg/h and 136.34 kg/h, respectively. Table 4
also lists the specific amount of CO2 emissions per ton of glycerol.
As these emissions are closely linked to the amount of energy re-
quired, it comes as no surprise that the DWC alternative exhibits
also the lowest carbon footprint.
4. Conclusions
This study successfully demonstrated a novel application of
DWC technology for a single step methanol recovery and glycerol
separation in biodiesel production. Compared to an optimized con-
ventional direct sequence of two distillation columns, the novel
proposed DWC alternative reduces the energy requirements by
27% and the equipment costs by 12%, thus leading to 25% savings
in the total annual costs. Notably, the novel separation scheme also
requires less equipment units and reduced plant footprint. Based
on these results, the use of dividing-wall column in biodiesel pro-
duction is especially interesting in case of building new large bio-
diesel plants, but also in the case of revamping existing plants –
where the equivalent Petlyuk configuration should be certainly
considered.
Acknowledgments
The financial support from the Sector Operational Program Hu-
man Resources Development 2007-2013 of the Romanian Ministry
of Labor, Family and Social Protection through the Financial Agree-
ment POSDRU/88/1.5/S/61178 is gratefully acknowledged.
References
[1] Van Gerpen J. Biodiesel processing and production. Fuel Process Technol
2005;86:1097–107.
[2] Bowman M, Hilligoss D, Rasmussen S, Thomas R. Biodiesel: a renewable and
biodegradable fuel. Hydrocarbon Process 2006;85:103–6.
[3] Lam MK, Lee MT, Mohamed AR. Homogeneous, heterogeneous and enzymatic
catalysis for transesterification of high free fatty acid oil (waste cooking oil) to
biodiesel: a review. Biotechnol Adv 2010;28:500–18.
[4] Leung DYC, Wu X, Leung MKH. A review on biodiesel production using
catalyzed transesterification. Appl Energy 2010;87:1083–95.
[5] Kiss A. Separative reactors for integrated production of bioethanol and
biodiesel. Comput Chem Eng 2010;34:812–20.
[6] Kiss AA. Heat-integrated reactive distillation process for synthesis of fatty
esters. Fuel Process Technol 2011;92:1288–96.
[7] Kiss AA, Bildea CS. Integrated reactive absorption process for synthesis of fatty
esters. Bioresour Technol 2011;102:490–8.
[8] Kiss AA, Bildea CS. A control perspective on process intensification in dividing-
wall columns. Chem Eng Process 2011;50:281–92.
[9] Kiss AA, Bildea CS. A review on biodiesel production by integrated reactive
separation technologies. J Chem Technol Biotechnol 2012;87. http://dx.doi.org/
10.1002/jctb.378.
[10] Meher LC, Vidya Sagar D, Naik S. Technical aspects of biodiesel production by
transesterification: a review. Renew Sustain Energy Rev 2006;10:248–68.
[11] Demirbas A. Comparison of transesterification methods for production of
biodiesel from vegetable oils and fats. Energy Convers Manage
2008;49:125–30.
[12] Qiu ZY, Zhao LN, Weather L. Process intensification technologies in continuous
biodiesel production. Chem Eng Process 2010;49:323–30.
[13] Dunford NT. Biodiesel production techniques, Robert M. Kerr Food &
agricultural products center. Food technology fact sheet. FAPC-150; 2007. p.
1–4.
[14] Fjerbaek L, Christensen KV, Norddahl B. A review of the current state of
biodiesel production using enzymatic transesterification. Biotechnol Bioeng
2009;102:1298–315.
[15] Helwani Z, Othman MR, Aziz N, Fernando WJN, Kim J. Technologies for
production of biodiesel focusing on green catalytic techniques. Fuel Process
Technol 2009;90:1502–14.
[16] Chemstations Inc. Biodiesel in ChemCAD. White paper; 2010. p. 1–7.
[17] Atadashi IM, Aroua MK, Abdul Aziz AR, Sulaiman NMN. Refining technologies
for the purification of crude biodiesel. Appl Energy 2011;88:4239–51.
[18] Balat M, Balat H. Progress in biodiesel processing. Appl Energy
2010;87:1815–35.
[19] Lin L, Zhou C, Saritporn V, Shen X, Dong M. Opportunities and challenges for
biodiesel fuel. Appl Energy 2011;88:1020–31.
[20] Olujic Z, Jodecke M, Shilkin A, Schuch G, Kaibel B. Equipment improvement
trends in distillation. Chem Eng Process 2009;48:1089–104.
[21] Asprion N, Kaibel G. Dividing wall columns: fundamentals and recent
advances. Chem Eng Process 2010;49:139–46.
[22] Harmsen J. Process intensification in the petrochemicals industry: drivers
and hurdles for commercial implementation. Chem Eng Process 2010;49:70–3.
[23] Dejanović I, Matijašević L, Olujić Ž. Dividing wall column – a breakthrough
towards sustainable distilling. Chem Eng Process 2010;49:559–80.
[24] Yildirim O, Kiss AA, Kenig EY. Dividing wall columns in chemical process
industry: a review on current activities. Sep Purif Technol 2011;80:403–17.
[25] Maleta VN, Kiss AA, Taran VM, Maleta BV. Understanding process
intensification in cyclic distillation systems. Chem Eng Process
2011;50:655–64.
[26] Bravo-Bravo C, Segovia-Hernández JG, Gutiérrez-Antonio C, Duran AL, Bonilla-
Petriciolet A, Briones-Ramírez A. Extractive dividing wall column: design and
optimization. Ind Eng Chem Res 2010;49:3672–88.
[27] Kiss AA, Suszwalak D. J-P. C., Enhanced bioethanol dehydration by extractive
and azeotropic distillation in dividing-wall columns. Sep Purif Technol
2012;86:70–8.
[28] Kiss AA, Suszwalak DJ-PC. Innovative dimethyl ether synthesis in a reactive
dividing-wall column. Comput Chem Eng 2012;38:74–81.
[29] Kiss AA, Pragt H, van Strien C. Reactive dividing-wall columns – how to get
more with less resources? Chem Eng Commun 2009;196:1366–74.
[30] Kiss AA, Segovia-Hernandez JG, Bildea CS, Miranda-Galindo EY, Hernandez S.
Reactive DWC leading the way to FAME and fortune. Fuel 2012;95:352–9.
[31] Hernandez S, Sandoval-Vergara R, Barroso-Munoz FO, Murrieta-Duenasa R,
Hernandez-Escoto H, Segovia-Hernandez JG, et al. Reactive dividing wall
distillation columns: simulation and implementation in a pilot plant. Chem
Eng Process 2009;48:250–8.
[32] Gomez-Castro FI, Rico-Ramirez V, Segovia-Hernandez JG, Hernandez S.
Feasibility study of a thermally coupled reactive distillation process for
biodiesel production. Chem Eng Process 2010;49:262–9.
[33] Gomez-Castro FI, Rico-Ramirez V, Segovia-Hernandez JG, Hernandez-Castro S.
Esterification of fatty acids in a thermally coupled reactive distillation column
by the two-step supercritical methanol method. Chem Eng Res Des
2011;89:480–90.
[34] Wilson RB. A simplicial method for convex programming. PhD dissertation,
Harvard University; 1963.
[35] Murray W. An algorithm for constrained minimization. In: Fletcher R, editor.
Optimization. London and New York: Academic Press; 1969. p. 247–58.
 A.A. Kiss, R.M. Ignat / Applied Energy 99 (2012) 146–153
[36] Gill PE, Murray W, Wright MH. Practical optimization. London: Academic
Press; 1981.
[37] Powell MJD. Variable metric methods for constrained optimization. In:
Bachem A, Grotschel M, Korte B, editors. Mathematical programming: the
state of the art. Springer Verlag; 1983. p. 288–311.
[38] Boggs PT, Tolle JW. Sequential quadratic programming. Acta Numerica
1995;4:1–51.
[39] Bartholomew-Biggs M. Nonlinear optimization with engineering applications.
Springer Optimiz Appl 2008;19:1–14.
[40] Storen S, Hertzberg T. The sequential linear-quadratic programming algorithm
for solving dynamic optimization problems – a review. Comput Chem Eng
1995;19:S495–500.
[41] Nagy ZK, Mahn B, Franke R, Allgower F. Efficient output feedback nonlinear
model predictive control for temperature control of industrial batch reactors.
Control Eng Pract 2007;15:839–59.
[42] Nagy ZK. Model based robust control approach for batch crystallization
product design. Comput Chem Eng 2009;33:1685–91.
[43] Simon LL, Nagy ZK, Hungerbuehler K. Model based control of a liquid swelling
constrained batch reactor subject to recipe uncertainties. Chem Eng J
2009;153:151–8.
[44] Roman R, Nagy ZK, Cristea MV, Agachi SP. Dynamic modelling and nonlinear
model predictive control of a fluid catalytic cracking unit. Comput Chem Eng
2009;33:605–17.
[45] Mesbah A, Huesman AEM, Kramer HJM, Nagy ZK, van den Hof PMJ. Real-time
control of a semi-industrial fed-batch evaporative crystallizer using different
direct optimization strategies. AIChE J 2011;57:1557–69.
[46] Alsumait JS, Sykulski JK, Al-Othman AK. A hybrid GA–PS–SQP method to solve
power system valve-point economic dispatch problems. Appl Energy
2010;87:1773–81.
[47] Aspen Technology. Aspen Plus: user guide, vols. 1 & 2; 2010.
153
[48] Biegler LT, Cuthrell JE. Improved infeasible path optimization for sequential
modular simulators. Part II: the optimization algorithm. Comput Chem Eng
1985;9:257–67.
[49] Lang YD, Biegler LT. A unified algorithm for flowsheet optimization. Comput
Chem Eng 1987;11:143–58.
[50] Rangaiah GP, Ooi EL, Premkumar R. A simplified procedure for quick design of
dividing-wall columns for industrial applications. Chem Prod Process Model
2009;4(1) [article 7].
[51] Rong BG. Synthesis of dividing-wall columns (DWC) for multicomponent
distillations – a systematic approach. Chem Eng Res Des 2011;89:1281–94.
[52] Halvorsen IJ, Skogestad S. Energy efficient distillation. J Nat Gas Sci Eng
2011;3:571–80.
[53] Halvorsen IJ, Skogestad S. Optimizing control of Petlyuk distillation:
understanding the steady-state behaviour. Comput Chem Eng
1997;21:249–54.
[54] Dejanović I, Matijašević L, Olujić Ž. An effective method for establishing the
stage and reflux requirement of three-product dividing wall columns. Chem
Biochem Eng Q 2011;25:147–57.
[55] Diggelen van RC, Kiss AA, Heemink AW. Comparison of control strategies for
dividing-wall columns. Ind Eng Chem Res 2010;49:288–307.
[56] Kiss AA, Rewagad RR. Energy efficient control of a BTX dividing-wall column.
Comput Chem Eng 2011;35:2896–904.
[57] Ling H, Luyben WL. New control structure for divided-wall columns. Ind Eng
Chem Res 2009;48:6034–49.
[58] Rewagad RR, Kiss AA. Dynamic optimization of a dividing-wall column using
model predictive control. Chem Eng Sci 2012;68:132–42.
[59] Luyben WL, Luyben ML. Essentials of process control. New-York,
USA: McGraw-Hill; 1997.
[60] Gadalla M, Olujic Z, de Rijke A, Jansens PJ. Reducing CO2 emissions of internally
heat-integrated distillation columns for separation of close boiling mixtures.
Energy 2006;31:2409–17.