G Model

CACE-5212; No. of Pages 18 ARTICLE IN PRESS

Computers and Chemical Engineering xxx (2015) xxxxxx

Contents lists available at ScienceDirect

Computers and Chemical Engineering

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

Simulation and process integration for tert-amyl-methyl ether

(TAME) synthesis
Olivia Ana Perederic a , Valentin Plesu a, , Petrica Iancu a , Gheorghe Bumbac a ,
Alexandra-Elena Bonet-Ruiz a,b , Jordi Bonet-Ruiz b , Bertram Muchan c
a
University POLITEHNICA of Bucharest, Centre for Technology Transfer in the Process Industries, 1, Gh. Polizu Street, RO-011061 Bucharest, Romania
b
University of Barcelona, Department of Chemical Engineering, 1, Mart i Franqus Street, 6th Floor, E-08028 Barcelona, Spain
c
OMV Petrom, PETROBRAZI Renery, 65, Trandarilor Street, RO-100316 Brazi, Prahova County, Romania

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

Article history: This paper proposes an extended approach to develop a new sustainable process to produce tert-amyl-
Received 30 October 2014 methyl ether (TAME) using as feedstock enriched C5 fraction (LCN light cracking naphtha) from uid
Received in revised form 18 May 2015 catalytic cracking (FCC). To the best of our knowledge, up to now, different authors developed the sep-
Accepted 23 May 2015
aration section without considering all possible options. The main contribution is to bring together for
Available online xxx
comparison different separation techniques of the given mixture and to develop new congurations
for the separation section of the plant. In this respect, pressure swing is combined with liquidliquid
Keywords:
separation. Existing technologies consider methanol (MeOH) separation from reactor efuent only by
TAME synthesis and separation
Methanol recovery
water extraction, combined with distillation. Conceptual design based on residual curve maps (RCM)
Environmental impact analysis, considered in this paper, reveals new possibilities to use pressure swing, eventually combined
Pressure swing distillation with liquidliquid separation. Thus, compared to other results reported in literature, new separation
Liquidliquid equilibrium sequences are proposed for TAME synthesis reactor efuent separation, in the frame of an extended and
Techno-economical evaluation detailed analysis for the whole process.
To underline process characteristics, three case studies, with those different congurations are pre-
sented and analysed using Aspen HYSYS v8.4. Main details are obtained using process simulation,
process integration and environmental impact computer tools. In the rst case study, classical MeOH
separation using water extraction is considered. The second case study is based only on pressure swing
distillation to separate the azeotropes between hydrocarbons and methanol. In the third case study,
pressure swing distillation is combined with separation based on hydrocarbonmethanol liquidliquid
phase equilibrium. Using process simulation results, setup with Aspen HYSYS v8.4, heat integration
analysis, performed with SPRINT v2.8, is accomplished to exploit energy savings. Environmental impact
calculations are performed using WAR algorithm, considering different fuel types for utilities generation.
Results show that the elimination of water in separation section and the use of liquidliquid phase sep-
aration ensure lower energy consumption (overall heat recovery in case study 3 is 9.87 MW, compared
to 7.47 MW for case study 2) and better environmental performance. Economic indicators calculated
with Aspen Process Economic Analyzer allow identication of attractive process changes, for the new
proposed process conguration.
2015 Elsevier Ltd. All rights reserved.

1. Introduction Feedstock for TAME production is represented by isoamylenes

Tert-amyl methyl ether (TAME), along with tert-amyl-ethyl
ether (TAEE) and ethyl-tert-butyl ether (ETBE), are used in gaso-
line composition as additives to boost performance: higher octane
number and lower environmental impact (Su and Chang, 2000).
Corresponding author. Tel.: +40 21 4023916; fax: +40 21 3185900.
E-mail address: v plesu@chim.upb.ro (V. Plesu).
enriched C5 fraction, from uid catalytic cracking unit (FCC), e.g.
light cracking naphtha.
TAME production process consists of two main parts: reaction
section and separation section, with recycle loops between them.
Reaction section is represented by one or more reactors. TAME
formation reaction is catalysed by cationic exchange resin at
around 8 bar pressure, which ensures liquid phase reaction, avoid-
ing evaporation. Separation section aim is to produce required
TAME purity and to recover unreacted MeOH from the mixture with

http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
0098-1354/ 2015 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
2 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

hydrocarbons. In newer technologies, reaction section and

Nomenclature column for TAME separation are combined in a reactive distil-
lation column. All commercial technologies (NExTAME Neste
Abbreviations
Jacobs, CDEtherol CDTech, Ethermax UOP/Huls, Catacol
2M1B 2 methyl-1-butene
Axens) propose separation schemes based on MeOH extraction
2M2B 2 methyl-2-butene
with water (European Commission, 2013). However, this process
ASME American Society of Mechanical Engineers
created the possibility to contaminate gasoline with water. It is
BFW boiler feed water
also to mention that MeOHwater mixture separation involves
C1 column C1
high-energy consumption. Energy efciency of the process is
C2 column C2
directly connected to CO2 emissions and environmental impact
C3 column C3
(Bonet-Ruiz, 2012).
CEPCI chemical engineering plant cost index
To the best of our knowledge, separation sequences for TAME
CS1 case study 1
system reported in literature are not considering liquidliquid
CS2 case study 2
equilibrium in MeOHhydrocarbon system as a potential basis for
CS3 case study 3
separation process. Industrial processes tend to evolve in time
CW cooling water
when better technologies become available at an affordable cost.
HEN heat exchanger network
The industry has a standard practice called best available tech-
HI heat integration
nique (BAT) referring to the better technology affordable for a
Iout impact output, PEI/kg
certain process at a certain time. The European Union created
Igen impact generated, PEI/kg
the Integrated Pollution Prevention and Control Bureau (IPPCB)
iC5 isopentane
that publishes reports about the BATs available for industries
L-L liquidliquid
established in EUs members. Directive 2010/75/EU (Integrated Pol-
MeOH methanol
lution Prevention and Control) establishes that all EU industries
MER maximum energy recovery
must use the reported techniques or better ones. Best available
NPV net present value
technique for TAME production in the BAT Reference Document
NRTL non-random two liquids
Integrated Pollution Prevention and Control Bureau (2015) is taken
PEI potential environmental impact
as reference for process selection. This process follows the conven-
QHmin minimum hot utility consumption, MW
tional scheme, including a chemical reactor, separator and recycle
QCmin minimum cold utility consumption, MW
streams. In this process, the isoamylenes are separated from the
R1 reactor R1
light cracked spirit stream (LCCS) from the FCC unit and catalyti-
R2 reactor R2
cally reacted with methanol in the presence of hydrogen to produce
RCM residue curve map
TAME. The main stages for TAME production are pentane removal,
SW shaft water
scavenging, reaction and purication. C5 removal is achieved by
Tmin,opt minimum optimum temperature difference, C
distillation (depentaniser) of the LCCS feedstock. Overheads are
TS supply temperature, C
condensed and the hydrocarbons are returned as reux while gases
Tsat saturation temperature, C
go to the renery ue-gas system. A C5 side-stream is withdrawn
TT target temperature, C
from the column as feed to the TAME unit. Column bottom liq-
TAME tert amyl methyl ether
uids (C6+) are routed to re-blending with the eventual product
P1 TAME (nal product cooling)
from the TAME unit. The C5 stream is then scavenged to remove
P2 Gasoline (nal product cooling)
catalyst poisons by passing it through an ion exchange resin to
V1 vessel V1
remove basic nitrogen compounds, e.g. ammonia and any metallic
V2 vessel V2
contamination. A hydrogen feed is also scavenged to remove any
acidic components. The feedstock, containing injected methanol
Notations
and hydrogen, is fed to the reactor section. Hydrogen is used to con-
aj activity of j species
vert dienes into monoolens and prevent gum formation during
A pre-exponential factor, kmol/m3 uid/s
the reaction. This takes place over a palladium-impregnated ion-
Ai pre-exponential factor for direct reaction,
exchange resin and the isoamylenes are converted to TAME. The
kmol/m3 uid/s
TAME product stream is puried by fractional distillation, wash-
Ai pre-exponential factor for indirect reaction,
ing and phase separation. Fractionator overheads pass to a reux
kmol/m3 uid/s
drum with the gaseous phase of low-boiling hydrocarbons (C1, C2,
E activation energy, kJ/kmol
C4, etc.) together with unreacted hydrogen before being vented
Ei activation energy for direct reaction, kJ/kmol
to renery fuel gas or are. The bottom product of TAME gaso-
Ei activation energy for indirect reaction, kJ/kmol
line with some methanol is cooled and mixed with recycled water
i/j stream labelled i to stream labelled j from PFD
from the methanol recovery plant, then routed to a settler for phase
when entering data to obtain process integration
separation. The TAME gasoline fraction from this is blended with
ki kinetic constant for direct reaction, kmol/(m3 uid s)
the depentaniser bottoms C6+ stream and passed to storage. The
ki kinetic constant for indirect reaction,
methanol/water fraction is recycled to the methanol recovery plant
kmol/(m3 uid s) feed drum. Methanol is recovered by distillation in a common still,
R ideal gas constant, kJ/(kmol K) with the overheads methanol being condensed and passed to buffer
vi reaction rate, kmol/(m3 uid s) storage for recycling to the TAME plant or other uses. The bot-
toms are essentially water with some contaminants and are mainly
Indexes
recycled with a purge to efuent treatment to avoid formic acid
i 1, 2, 3 indexes for reactions R1, R2, and R3
buildup.
j 2M1B, 2M2B, MeOH, TAME
However, alternative approaches have been proposed for TAME
production. Subawalla and Fair (1999) studied a reactive distil-
lation system for TAME production, presenting the steady-state

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx
design of a prereactor and a reactive distillation column. Al Arfaj
and Luyben (2004) extended this work to design a complete plant
and to provide the plantwide control of the entire process, consider-
ing a reactive distillation column for TAME synthesis and methanol
recovery using a pressure-swing azeotropic separation method.
Luyben (2005) presents a quantitative steady state and dynamic
comparison of the pressure-swing process with an extractive dis-
tillation process, using water as extractive agent. In this case, the
extractive distillation process is found to be much more economical
(40% lower capital investment and 60% lower energy cost). Further-
more, Wu et al. (2009) consider a third alternative owsheet for the
separation of the C5/methanol azeotropes that takes advantage of
the heterogeneity of the azeotrope. The set-up considered involves
two stripping columns and a decanter, obtaining the total annual
cost of this owsheet 4 times less than that of the pressure swing
system.
Different approaches for reactive distillation including heat
integration (Gao et al., 2014) and optimisation problems (Lyu et al.,
2014) are presented in the literature. Using the heat integrated dis-
tillation column (HIDiC) conguration for reactive distillation, the
heat of reaction can improve the performance of HIDiC. Exother-
mic reactions, such as the ones involved in TAME synthesis, can
be a source of heat for liquid vaporisation in stripping section, and
the endothermic reactions can cool the vapor stream in rectica-
tion and generate more liquid reux in the column. This would lead
to more heat integration energy saving. Vanaki and Eslamloueyan
(2012) investigated the synthesis of TAME by a reactive internally
heat integrated distillation column (R-HIDiC). The performance of
the proposed R-HIDiC is compared to the conventional reactive
distillation column, the simulation results showing that the total
energy consumption of the proposed R-HIDiC is about 22% less
than that of the conventional reactive distillation column. Pulido
et al. (2011) performed a simulation study to compare the perfor-
mance of reactive distillation column and R-HIDiC column in the
TAME synthesis, achieving a better performance of R-HIDiC with
energy saving around 8%. They also demonstrated the possibility
of R-HIDiC column to operate at different pressures between the
rectifying and stripping sections, which permits to overcome sepa-
ration limitations resulting from the presence of azeotropes. Plesu
et al. (2008) applied the / analysis to evaluate new alternatives
for TAME synthesis and to propose energy-efcient alternatives
by process integration, such as the kinetically controlled reactive
distillation including hybrid systems. Energy savings obtained by
using reactive distillation are evaluated comparing with the tra-
ditional system consisting of reactor and distillation columns. The
solution proposed, subsequent to the validation by rigorous simu-
lation, offers a 20% decrease in the number of stages and a reboiler
energy saving of 10%. Rashed (2009) compares two PFDs for TAME
separation: rst one uses a water extractive distillation column
for MeOH separation and one distillation column for MeOHwater
separation. In the second one, a liquidliquid extraction column
replaces the extractive distillation.
Non-ideal behaviour of reactor efuent determined by the
presence in the same mixture of three types of compounds
Fig. 1. Simplied PFD of CS1.
Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
3
(hydrocarbons, MeOH and TAME) is high enough, determining the
formation of two azeotropes: TAMEMeOH and iC5 MeOH. It is
interesting to mention that the composition of these azeotropes
varies consistently with pressure, allowing the use of pressure
swing distillation in the separation scheme (Doherty and Malone,
2001).
In this paper, three different case studies are presented and ana-
lysed trying to nd favourable separation scheme for this plant,
comparing the results with those reported in literature or applied
in industry. Similar approaches were made for other systems: iso-
propyl alcohol-di-isopropyl ether (Haotao et al., 2014), di-n-propyl
ether and n-propyl alcohol (Llandosa et al., 2011) and acetone-
chloroform (Luyben, 2013). Consequently, each case study has the
same reaction section, but different separation section topologies
(Turton et al., 2014). Reaction section consists of a series of two
adiabatic reactors with intermediate cooling, for a higher nal con-
version of reactants (Ferreira et al., 2007). The role of separation
section is to obtain acceptable purity of main components: TAME,
MeOH and hydrocarbons (unreacted gasoline). In this respect, for
each case study different techniques are combined (Perederic et al.,
2014). The topology of distillation columns sequence is determined
by Residual Curve Maps (RCM) analysis performed with SIMULIS
Thermodynamics tool.
First case study (CS1) (Fig. 1) considers classical process for
separation (Rashed, 2009), i.e. liquidliquid extraction of MeOH
in water. First distillation column (C1) aims to collect TAME in
bottoms product, while all other components are majority of distil-
late product. Then, MeOH contained in this stream is separated by
liquidliquid extraction, using water, in C2 column. In this case,
C5 fraction (gasoline) is the main component in the light phase
product. Heavy phase product is MeOHH2 O mixture, which is sep-
arated in C3 distillation column. Two recycles are implemented
for C3 column products: MeOH stream is recycled to rst reactor
feed (R1) and H2 O stream is recycled to C2 liquidliquid extraction
column feed.
Case study 2 (CS2) (Fig. 2) separation scheme is based on
pressure swing distillation, given the variation of azeotropes com-
position with pressure, as shown by RCM analysis in Section 2.1. The
topology is based on indirect separation scheme: in C1 distillation
column TAME is the main component of bottoms stream. Column
C2 operates at a different pressure resulted from RCM analysis.
C5 hydrocarbons (gasoline) is the main component in C2 column
bottoms stream, while MeOHC5 hydrocarbons nearly azeotrope
mixture is separated as C2 column distillate (Luyben, 2011). Third
distillation column (C3) is designed to another pressure, allowing
separation of pure MeOH as main component of bottoms stream
and MeOHC5 hydrocarbons nearly azeotrope mixture as distillate.
MeOH stream is recycled back to reactor R1 feed, while distillate
stream is recycled to C2 column feed for further MeOH recovery.
The potential to separate two liquid phases of MeOHC5 hydrocar-
bons azeotropic mixtures is not exploited in this case study.
In case study 3 (CS3), similar technique is mainly taken into
account as in CS2, i.e. pressure swing distillation. Additionally, the
potential to form two liquid phases of MeOHC5 azeotropes is also
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
4 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

Fig. 2. Simplied PFD of CS2.

Fig. 3. Simplied PFD of CS3.

considered. In this respect, stream cooling and liquidliquid sepa-
rators are implemented for both C2 and C3 distillate products, i.e.
vessels V1 and V2. MeOH enriched phase from both V1 and V2 is
summed up to form C3 column feed, while light phases, enriched
in hydrocarbons from those vessels are recycled to C2 column feed.
Simplied PFD for CS3 is presented in Fig. 3.
In conclusion, the results presented in literature until now
insisted more on reaction system, eventually considering reaction-
separation conguration based on reactive distillation. Less
attention is given to the whole separation section of the TAME
synthesis plant. In this respect, the present paper investigates dif-
ferent potential techniques for TAME reactor efuent separation,
aiming to nd a good conguration including also the separation
of methanolhydrocarbons mixture. Consequently, different tools
are used to investigate deeply this issue, nding that combination
of pressure swing with liquidliquid extraction as attractive.
2. Methodology
Development of a new separation scheme for TAME produc-
tion uses several analyses starting from physical and chemical
behaviour of the involved components and their mixture to
plant simulation along with other aspects: possibilities of heat
integration, environmental impact, and maybe the most impor-
tant one, economical performances. Fig. 4 presents each step of the
entire study on which basis is selected the case study with the best
performances.
2.1. RCM analysis
Separation section sequence for CS2 and CS3 is designed
based on RCM analysis (Doherty and Malone, 2001) with
SIMULIS Thermodynamics (SIMULIS Thermodynamics, 2013).
Taking into account reactor section efuent composition, two
ternary systems are investigated: TAMEiC5 MeOH, and respec-
tively 2M2BiC5 MeOH. For both systems, thermodynamic models
used for RCM generation are: NRTL activity model for liquid
phase and PengRobinson equation of state for vapour phase,
MHV2 for mixing rules, LeeKeslerPloker for liquid molar volume,
ElyHanley model for transport properties, standard conditions for
mixture rules, and pure liquid fugacity standard state. Binary inter-
action coefcients estimation is made using NRTL activity model as
implemented in computer tools used in this paper. It is right that in
literature there are publications aimed to determine binary interac-
tion coefcients (Yang et al., 2015), but in our case this was beyond
the purpose of this paper. RCM at different pressures are gener-
ated and analysed to best placement of distillation columns, for the
separation section.
2.2. Process simulation
Aspen HYSYS v8.4 is used for owsheet simulation. The
feedstock, enriched C5 fraction from FCC, is simplied: all inert
hydrocarbons are represented by isopentane (iC5 ) without con-
sideration of C6 and C7 components. The reactive alkenes taken
into account are only 2-methyl-1-butene (2M1B) and 2-methyl-2-
butene (2M2B). NRTL is the model selected to describe non-ideality
of the liquid phase and PengRobinson equation of state is chosen
to describe vapour phase behaviour. Binary interaction param-
eters are taken as built-in values from Aspen HYSYS v8.4.
Main reactions taken into account are: reactions between each
isoamylene and MeOH (R1, R2) and the isomerisation reaction of
isoamylenes (R3) (Syed et al., 2000). EleyRideal kinetic model pro-
posed by Rihko and Krause (1995) is used. Transformation from
mole/(kgcat s) (usually given in literature for reaction rate expres-
sion), to kmol/(m3 uid s) (as implemented in process simulator) is
achieved to meet the simulator equation (Eq. (1)). Different stud-
ies in literature propose industrial data for reactor feed (Lyu et al.,
2014) along with some side reactions: etherication of MeOH to
DME and water addition to olens with tertiary alcohol formation
(Mao et al., 2008), isoamylenes dimerization (Libert et al., 2010).
Those aspects are not tackled in this paper, as well as the congu-
ration of reaction section.
2M1B + MeOH TAME (R1)
2M2B + MeOH TAME (R2)
2M1B 2M2B (R3)
Kinetic model based on EleyRideal mechanism, presented by
Rihko and Krause (1995) utilise components activity to describe

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 5

Fig. 4. Methodology algorithm.

Table 1
Kinetic parameters implemented in Aspen HYSYS v8.4 (Rihko and Krause, 1995).

Reaction Ai , kmol/(m3 uid s) Ei , kJ/kmol Ai , kmol/(m3 uid s) Ei , kJ/kmol A, kmol/(m3 uid s) E, kJ/kmol

Reaction R1 3.20 10 10
7.68 10 4
2.01 10 14
1.11 10 5

Reaction R2 4.00 1013 9.97 104 1.52 1017 1.26 105 4.01 105 24,212
Reaction R3 7.50 1010 8.17 104 1.02 1011 8.89 104

the reaction kinetics. Only TAME and MeOH adsorption on cata- Reaction section is represented by two reactors (R1 and R2)
lyst is taken into account, neglecting hydrocarbon adsorption. The implemented as plug ow reactors (PFR). Reactors geometrical
model parameters presented by the authors are modied, to meet details are chosen: 0.7 m diameter, 3 m length for R1, and respec-
Aspen HYSYS implemented kinetic equation. Amberlyst 35 wet, tively same diameter and 4 m (R2) length. PFR model implemented
from Rohm and Haas (division of Dow Chemicals), is taken into con- in Aspen HYSYS with a series of 20 adiabatic continuously stirred
sideration as catalyst. Modied model is described by Eqs. (1)(6). tank reactors (CSTR) of equal volume is chosen. Catalyst data spec-
Reaction rate for TAME formation, v1 and v2 (Eqs. (1) and (2)) is ications for Amberlyst 35 wet are: 0.8 mm mean diameter, unity
given by direct reaction, indirect reaction and adsorption of MeOH sphericity factor, solid density 800 kg/m3 , and solid heat capac-
and TAME on catalyst surface. Isomerisation reaction rate, v3 (Eq. ity 250 kJ/(kgK). To reach higher conversion, inlet temperature for
(3)) considers only direct and indirect reactions. The rate constants each reactor is set to 75 C. For both reactors, the inlet pressure
for direct and indirect reaction, ki and ki (Eqs. (4) and (5)) and the is considered 8 bar, and pressure drop is calculated with Ergun
adsorption constant, K (Eq. (6)) are described by an Arrhenius type equation, implemented in Aspen HYSYS PFR reactor model. This
equation. Table 1 presents the implemented kinetic parameters. pressure conditions ensures that the chemical reactions take place
in liquid phase and the temperature observes industrial data pro-
k1 a2M1B aMeOH k1 aTAME
v1 =  aTAME n
 (1) visions (Kiviranta-Pkknen et al., 1998).
1+K aMeOH First column of the separation section is TAME separation col-
umn (C1). For all three case studies C1 has the same conguration
k2 a2M2B aMeOH k2 aTAME and operation pressure (6 bar). Pressure drop along the column is
v2 =  aTAME n
 (2)
1+K aMeOH
0.2 bar for upper and respectively for lower zone. The value of oper-
ating pressure is chosen from RCM analysis, based on separation
v3 = k3 a2M1B k3 a2M2B (3) performance and reboiler temperature. A higher value is possible,
 Ei
 with higher energy consumption and higher reboiler temperature,
ki = Ai exp (4) but without a signicant gain in separation performance (Luyben,
RT
2011). This distillation column is modelled with a total condenser
 
Ei and once-through reboiler. Feed-tray selection is made considering
ki = Ai exp (5) feed composition and thermal condition, to obtain a smooth tem-
RT
perature column prole. The column C1 has 10 theoretical trays,
 E

the feed tray is the 4th tray. To complete the simulation, column
K = A exp (6)
RT specications are: 0.99 TAME molar fraction in bottoms and 2 reux

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
6 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx
Table 2
Proposed hot and cold utilities for heat integration, Tmin,opt calculation.
Cold utilities TS , C TT , C
Cooling water (CW) 28 38
Shaft water (SW) 5 15
Boiling feed water (BFW) 50 85
ratio. Simultaneous correction solver is used with azeotropic and
adaptive damping factor to keep into account the highly non-ideal
nature of mixtures. Error tolerance for equilibrium is 105 and error
tolerance for heat and specications is 5 104 .
In the CS1, the next element of separation section is the
liquidliquid extraction column (C2) designed to separate MeOH
from the mixture with hydrocarbons, modelled with ve theoreti-
cal trays. Light phase, C1 column distillate, enters the bottom stage,
while the extracting agent, water, enters on the top stage. Total
water ow rate (fresh and recycled water) is selected taking into
account separation performance (i.e. high methanol recovery).
Last column of CS1 separation section is the MeOHwater distil-
lation column (C3) used to separate MeOH from C2 extract. For this
column 12 theoretical trays are setup, with feed stream entering the
6th tray. The feed condition is boiling point temperature. Feed tray
selection is made based on smoothness of both temperature and
liquid composition proles, inside the column. Feed tray pressure
is established at 1.3 bar. The reason for selection of this pressure
value is to allow using cooling water at 28 C in the condenser. The
whole column pressure drop is considered of 0.2 bar. The congura-
tion is completed with total condenser and once-through reboiler.
HYSIM inside-out solver with azeotropic adaptive damping factor
and aggressive acceleration of K value and H model parameters are
used to ensure simulation convergence. Column specications are:
0.99 MeOH recovery ratio in distillate and 0.99 water recovery ratio
in bottoms.
To complete CS1 owsheet, two recycle operations are imple-
mented to ensure MeOH return to reactor section feed, and water
to be recycled to liquidliquid extraction column. Solver congu-
ration is simultaneous solution with Wegstein acceleration mode.
Sensitivities for recycle calculations are 1 for vapour fraction, pres-
sure, temperature, enthalpy and entropy and 0.1 for composition
and ow rate.
In CS2, separation section consists of three columns: C1 to sep-
arate TAME, and a sequence of two distillation columns (C2 and
C3) aimed to exploit azeotropes composition modication with
pressure for separation of MeOHhydrocarbon mixture (pressure
swing). These columns are congured with total condenser and
once-through reboiler. Modied HYSIM inside-out solver proved to
be most performing for taking into account non-ideality of involved
mixtures. Other solver parameters (as azeotropic and adaptive
dumping factor with conservative acceleration of K value and H
model parameters) are carefully tested to obtain best conditions for
solver operation. From RCM analysis, C2 column operation pressure
is 5 bar (considering whole column pressure drop of 0.4 bar). This
column is congured with six theoretical trays; feed is located on
3rd theoretical tray. Feed condition is boiling point temperature.
The two specications needed are: 2.4 for reux ratio and 0.2563
MeOH fraction in distillate, to approach MeOHiC5 azeotrope
composition. Reux ratio value is selected considering column sep-
aration performance and lower reboiler duty. On the other hand,
RCM analysis indicated that an improved separation is obtained
imposing a lower pressure in third distillation column (C3). In this
respect, operating pressure for C3 column is 2 bar (considering a
global pressure drop of 0.4 bar). This selection allows, as well, the
use of cooling water (28 C) in the condenser. This column is
congured with eight theoretical trays, feed is placed on 4th theo-
retical tray. Column feed condition is at boiling point temperature.
Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
Price, D /kWh Hot utilities p, bar Tsat , C Price, D /kWh
0.001136 HP 30 234.18 0.009653
0.001136 MP 15 128.29 0.007722
0.00182 LP 4 143.62 0.005148
Necessary column specications, established after more tests, are:
0.99 MeOH molar fraction in bottoms and reux ratio value of 3.
For this case study two recycle operations are implemented. One
recycle is aimed to ensure MeOH recovery from C3 bottoms to the
reaction section. Second recycle recovers C3 distillate adding it to
C2 feed to improve MeOHhydrocarbon separation. Numerical con-
guration parameters of those two recycle operations are: solver
choice, simultaneous solution with Wegstein acceleration, and iter-
ations sensitivities of 1 or lower than 1, depending on the variable
considered as presented above.
CS3 owsheet simulation is quite similar to CS2, keeping similar
specications for all three distillation columns. C2 and C3 dis-
tillation columns aim is to separate MeOHhydrocarbon mixture
by pressure swing distillation to exploit azeotropes composition
change with pressure. The difference between CS2 and CS3 is that
in CS3 L-L equilibrium is exploited to increase the effectiveness
of separation by implementing two phase separation vessels: V1
and V2. In CS3 both C2 and C3 distillates are split into two liquid
phases: one enriched in MeOH (heavy phase) and one enriched in
hydrocarbons (light phase). V1 and V2 are operated at 35 C, allow-
ing to using 25 C cooling water. The heavy phase streams (MeOH
rich) from both vessels are summed up to form C3 distillation col-
umn feed. The light phase streams (hydrocarbon rich) from both
V1 and V2 vessels are recycled to C2 distillation feed. C2 and C3
have the same conguration: total condenser and once-through
reboiler. The convergence of both columns is obtained reasonably
by the same solver modied HYSIM inside-out. C2 distillation column
is congured with six theoretical trays and operation pressure is
5 bar (considering 0.4 bar global pressure drop). Feed is located on
the 3rd tray with condition at boiling point temperature. To com-
plete needed data, column specications are: 2.4 reux ratio and
0.2563 MeOH molar fraction in distillate, to approach azeotrope
MeOHiC5 composition. The specications value is selected taking
into account separation performance. Operating pressure of C3 dis-
tillation column is lower (2 bar), compared to C2 pressure. Global
pressure drop is 0.4 bar. The column is congured with eight the-
oretical trays, feed stream condition is boiling point temperature.
Feed stream is located on the 4th tray. Following column specica-
tions are chosen: iC5 recovery ratio in distillate 0.9895 and bottoms
temperature is set to 83.2 C. Three recycle operations are imple-
mented in CS3: C3 bottoms stream to R1 feed, V1 and V2 light
phase to C2, and respectively, V1 and V2 heavy phase to C3. The
parameters are selected as in CS2.
2.3. Heat integration
Heat integration of the three case studies is performed with
SPRINT software tool. Process streams are extracted from corre-
sponding simulations, observing well known rules (Klemes, 2013).
Proposed cold and hot utilities are presented in Table 2 consid-
ering industrial data (Perederic, 2013). The production of boiler
feed water (BFW) is proposed as well as cold utility for lower
quality heat recovery. Cold and hot streams from simulation are
extracted from Aspen HYSYS simulation. Necessary energy for
products cooling at 30 C is considered in the analysis. Minimum
temperature difference for each case is considered from capital and
operation costs trade-off, according to supertargeting procedure
(Klemes et al., 2014). Capital cost is estimated considering shell
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx
Table 3
Raw materials and products price considered for economic analysis.
Raw material Price Product
Feed C5, D /kg 0.58861 TAME, D /kg
MeOH, D /kg 0.35000 Gasoline, D /kg
and tube exchanger cost model: 772.17A0.7 , $/(kWm2 ). Original
exchanger cost model (Seider et al., 2004) is modied to meet actual
price level by indexing it with CEPCI factor report from year when
the model is proposed and July 2014. Operation cost accounts only
for the utility cost. For steam price values an industrial platform
with classical steam generation (no co-generation) is taken into
account (Turton et al., 2014). These prices and cost model are only
used for minimum optimum temperature difference (Tmin,opt )
calculation.
2.4. Environmental impact
Environmental performance evaluation of the three case stud-
ies is based on potential environmental impact (PEI), using Waste
Reduction Algorithm (Cabezas et al., 1999) implemented in WAR
GUI v1.0.17 software tool, developed by US Environmental Protec-
tion Agency (EPA, 2013). WAR GUI contains physical properties
and toxicology data, requiring information about raw materials
streams, products streams, and waste streams (Barret et al., 2011).
For these case studies, waste streams are not considered. From
owsheet simulation, inlet streams are Feed C5, Feed MeOH and
Feed H2O. Product streams are TAME and Gasoline, who are cooled
at storage temperature, when needed. For each case study, three
different fuels are proposed for energy generation: coal, oil and nat-
ural gas (Bonet-Ruiz et al., 2010). Eight indicators: human toxicity
(ingestion and inhalation), aquatic and terrestrial toxicity, global
warming, ozone depletion, photochemical oxidation, and acidi-
cation potential (Sikdar and El-Halwagi, 2001) are calculated from
PEI value.
2.5. Techno-economic evaluation
Techno-economic analysis is performed with Aspen Process
Economic Analyzer v8.4 (Aspen, 2013), only for the best case study
(CS3), selected according to results obtained by previous analyses
types (Kim et al., 2011). The input data for equipment, operating,
and capital cost estimation are the results from Aspen HYSYS sim-
ulations. Analysis basis of the scenario is built using metric system,
ASME standard and Euro currency. Scenario assumptions are as
follows (Murat Sen et al., 2012):
Fig. 5. RCM diagram for TAMEiC5 MeOH system at 6 bar and column C1 position (D1 distillate, F1 feed, W1 bottom).
Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
7
Price Utility Price
0.80626 Electricity, D /kWh 0.0775
0.65500 Fuel, D /MWh 26.5
Process is considered as redesigned, since similar industrial pro-
cesses are implemented with a typical complexity. Digital control
of the process is assumed.
Since the process is continuous, and feedstock is intermediary
product within an oil renery, plant construction is considered
grass roots, being included into an existing industrial site in
Europe.
Economic life of the project is considered to be 13 years with 1
year period for construction.
A straight line depreciation method is assumed along with 40%
tax rate, 20% ROR, and 25% interest rate. Values for escalation
parameters are: project capital 5%, products 3%, operating and
maintenance labour 2.5%, utilities 2.3%. A value of 20% initial
capital cost is considered as salvage value.
Capital cost values are taken from Aspen Process Economic
Analyzer cost data base: 1Q12 v21.0.0. Price for raw materials
and products, presented in Table 3 are taken from industrial data.
TAME price is for 80% purity, while purity considered for evalu-
ation is 99% (Perederic, 2013). Underestimation of TAME price
makes the results to be more realistic.
3. Results and discussions
3.1. RCM analysis
Separation section topology for CS2 and CS3 results from
RCM analysis. For TAME reactor efuent composition, two sys-
tems need to be taken into account, as previously mentioned:
TAMEiC5 MeOH and 2M2BiC5 MeOH. The systems are chosen
based on majority components from the system: TAME, iC5 and
MeOH for the rst system and 2M2B, iC5 and MeOH for the second
system.
Fig. 5 presents ternary diagram of TAMEiC5 MeOH sys-
tem at 6 bar, observing industrial conditions. The diagram
shows two binary azeotropes: iC5 MeOH (0.7260:0.2740) and
TAMEMeOH (0.1703:0.8297). Distillation boundary generated by
these azeotropes divides the RCM diagram into two zones. Typi-
cal efuent composition from TAME reactors is placed in the upper
distillation zone, represented in Fig. 5 as point F1. TAME has the
lowest volatility, and its representative node in RCM diagram is a
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
8 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

Fig. 6. RCM diagram for 2M2BiC5 MeOH system at 5 bar and column C2 position (D2 distillate, F2 feed, W2 bottom).

Table 4 Fig. 8, a distillation column feed with D2 mixture, operating at 2 bar

Azeotrope composition and temperature for 2M2BiC5 MeOH system (SIMULIS
can separate nearly pure MeOH in the bottoms, and consequently
Thermodynamics ).
MeOHiC5 mixture, point D3 , in the distillate. These results allow
Pressure, Azeotrope Composition, molar fr. Boiling point for conceptual design of CS2 separation section. If L-L equilibrium is
bar type temperature, C
exploited, Fig. 7, as in CS3, cooling down a mixture placed near the
2M2B iC5 MeOH distillation boundary (as point D2), two liquid phases are formed.
Unstable 0.8103 0.1897 42.38 Heavy liquid phase (as point L1) is placed in MeOH distillation zone,
2 and light liquid phase (as point L2) is placed in iC5 distillation zone.
Saddle 0.7376 0.2624 50.71

Unstable 0.7587 0.2413 64.10

4
Saddle 0.6882 0.3118 72.21

Unstable 0.7409 0.2591 71.67 3.2. Process simulation

5
Saddle 0.6721 0.3279 79.70

Unstable 0.7260 0.2740 78.07
6
Saddle 0.6589 0.3411 86.04
stable node. iC5 has the highest volatility among the pure compo-
nents, and it is a saddle node, while MeOH is a stable node. Residual
curves position reveal the possibility to separate nearly pure TAME
in the bottoms of the rst distillation column, point W1, and a
nearly binary mixture iC5 MeOH in column distillate, D1 point.
Taking into account composition of the remaining mixture
(distillate from rst column), mainly 2M2B, iC5 , and MeOH, the sec-
ond system, 2M2BiC5 MeOH, is analysed for hydrocarbonMeOH
separation. More RCM diagrams are obtained for each system
at different pressures (2 bar, 4 bar, 5 bar, and 6 bar) to under-
line azeotrope shifting and residue curves shape. In Table 4,
azeotropes composition and corresponding temperature are pre-
sented. These data allow choosing pressure swing in corresponding
columns: C2 pressure 5 bar and C3 pressure 2 bar. This sys-
tem forms two binary azeotropes: 2M2BMeOH (saddle point),
iC5 MeOH (unstable point). Consequently the distillation bound-
ary appears and connects these azeotropes. Distillate stream from
C1 column, represented in Fig. 6 by point F2, is separated into
a mixture of hydrocarbons (W2) in bottoms of the column, and
MeOHhydrocarbons mixture, with composition close to binary
azeotrope, in distillate stream (D2). Then, changing the pres-
sure from 5 bar in C2 to 2 bar in C3, a decrease of 0.0694 MeOH
units of molar fraction, in MeOHiC5 azeotrope, and a decrease of
0.0787 MeOH units of molar fraction in MeOH2M2B azeotrope
are obtained. This variation leads to a boundary shifting to the
iC5 zone, increasing the MeOH distillation zone. Consequently, the
composition azeotrope (and any other mixture placed on the dis-
tillation boundary D2) from 5 bar diagram is placed in the MeOH
distillation zone in the 2 bar diagram. In these new conditions,
Process simulation results for reaction section and C1 distilla-
tion column performances for all three case studies are presented
in the rst part of this section. Then, the focus is on separation
MeOHhydrocarbon mixtures for each case study. Binary interac-
tion parameters used in simulation are presented in Table 5. All Bij
parameter values are set to 0.
R1 and R2 reactor models are easily simulated. Temperature
variation within R1 is 17 C and within R2 is 4 C, respectively.
Isoamylenes conversion is 45.9% in R1 and additionally 19% in R2.
Composition proles of R1 and R2 illustrated in Figs. 9 and 10, indi-
cating a good selection of reactor length. Lower conversion and
temperature variation within R2 are a result of approaching equi-
librium conditions.
C1 column solver convergence is achieved easily with param-
eters selected as described in Section 2. The performance is
presented in Figs. 1113. Temperature prole and composition pro-
le show a good selection of the feed tray position, as they are
smooth. Bottom stream composition, along with liquid ow in the
column upper zone, reveal that both specications are satised.
Molar fraction of iC5 is higher on the rst and second stage of the
column (Fig. 12), underlining the possibility to have a side stream,
but this is beyond the scope of this paper. Liquid ow within the
column presents a slow increase on the top stages (13) given by
the reux, a sharp increase under the feed tray (4th tray) and a slow
decrease in the bottom zone, given by the mass transfer into vapour
phase. Vapour ow decreases on the 1st and 2nd stage. Another
decrease is noticed in stages 5th to 9th, given by feed stream
inuence. Bottom product contains traces of iC5 and 2M2B. Cal-
culated TAME recovery ratio in bottoms is 0.9920. Reboiler duty is
5.34 MW and condenser duty is 5.38 MW. Details for the streams
connecting reactors (R1 and R2) and C1 column are quite similar
for all case studies. Table 6 presents the results for these streams
from CS3.

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 9

Fig. 7. L-L equilibrium diagram for 2M2BiC5 MeOH system at 35 C.

Fig. 8. RCM diagram for 2M2BiC5 MeOH system at 2 bar and column C3 position (D3 distillate, F3 feed, W3 bottom).

Table 5
NRTL binary interaction parameters from Aspen HYSYS v8.4.

Aij , kJ/kg ij

iC5 2M1B 2M2B TAME MeOH iC5 2M1B 2M2B TAME MeOH

iC5 139.8489 526.06 852.225 4978.868 0.3 0.3277 0.3 0.4394

2M1B 15.51469 646.097 1081.47 3586.28 0.3 0.3176 0.3 0.4003
2M2B 864.3525 1346.832 878.605 4267.797 0.3277 0.3176 0.3 0.4331
TAME 1845.697 1878.18 1627.712 1960.541 0.3 0.3 0.3 0.3
MeOH 6587.423 5361.315 5537.629 1705.088 0.4394 0.4003 0.4331 0.3

For CS1 some comments for liquidliquid extraction column recovery in C2 column. As a consequence, the molar fraction of
simulation results are given. The distillate stream of C1 column is MeOH in the heavy phase stream is 0.23. Rafnate, gasoline stream,
cooled and enters at the bottoms of C2 column with 0.13 MeOH contains less than 0.005 molar fraction of water and MeOH. Stream
molar fraction. Total water feed is 120 kmol/h for MeOH better details obtained from simulation are presented in Table 7. Composi-
tion prole of MeOH in light and heavy phase (Fig. 14) shows a good
Table 6 operation of liquidliquid extraction column, i.e. nearly all MeOH
Process stream data for reaction section and C1 column for CS3. is extracted in heavy phase. C3 distillation column convergence is
Stream name 1 7 8 TAME
achieved easily, achieving specications values based on parame-
ters presented above (Section 2). Column performance is presented
Temperature, C 25 92 79 160
in Figs. 1517. Composition prole for C3 (Fig. 16) shows a slight
Pressure, bar 8.0 7.4 5.8 6.2
iC5 , kmol/h 213.9 214.0 213.7 0.3 decrease in MeOH fraction in the vapour phase at the 6th stage
2M1B, kmol/h 17.9 2.7 2.7 0.0 given by the feed (liquid phase). Both temperature and composi-
2M2B, kmol/h 36.4 21.0 20.9 0.1 tion proles present a smooth variation. MeOH molar fraction in
TAME, kmol/h 0.0 30.5 0.2 30.3 C3 distillate stream is 0.97, while in C3 bottoms stream is less than
MeOH, kmol/h 65.0 34.5 34.5 0.0

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
10 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

70 1.0
2M1B 2M2B
0.9
60 TAME MeOH
0.8

Composition, molar fraction

50 0.7
Molar flow, kmole/h

iC5
0.6
40 2M1B
0.5
2M2B
30 0.4
TAME
0.3
20
0.2 MeOH

10 0.1

0.0
0
0 1 2 3 4 5 6 7 8 9 10 11
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Stage
Lenght, m
Fig. 12. C1 column liquid composition prole for CS1.
Fig. 9. R1 reactor composition prole for CS1.

900
45
800
2M1B 2M2B TAME MeOH
40
700
35
600

Flow rate, kmole/h

Molar flow, kmole/h

30
500
25 400

20 300

15 200
Vapour
10 100
Liquid
5 0
0 1 2 3 4 5 6 7 8 9 10 11
0 Stage
0.0 1.0 2.0 3.0 4.0
Lenght, m Fig. 13. C1 column vapour and liquid ow for CS1.

Fig. 10. R2 reactor composition prole for CS1. 0.25

165 MeOH - light liquid

0.20 MeOH - heavy liquid
Composition, molar fraction

155

145
0.15
135
Temperature, C

125 0.10
115

105 0.05

95

85 0.00
1 2 3 4 5
75 Stage
0 1 2 3 4 5 6 7 8 9 10 11
Stage Fig. 14. L-L extraction column C2 composition prole of MeOH for CS1.

Fig. 11. C1 column temperature prole for CS1.

distillate stream, while 2M2B is concentrated in column bottoms
product. Isoamylenes distribution is different in bottoms and dis-
0.003. This good performance allows for a high separation of water tillate streams: more than half quantity of 2M1B is separated in
and MeOH, according to specication values given for simulation. distillate stream, while more than half quantity of 2M2B is sep-
Reboiler duty is 5.34 MW, while condenser duty is 5.38 MW. arated in bottoms stream. Vapour and liquid ow proles show
CS2 simulation results are presented in Table 8. C2 distillation a sharp increase in liquid ow under the feed tray vapour ow
column convergence is easily achieved, and its performance is pre- presents a decrease in the upper zone of the column. Reboiler and
sented in Figs. 18 (temperature prole), 19 (composition prole condenser duties are around 9.3 MW. C3 distillation column spec-
in liquid phase), and 20 (vapour and liquid ow proles). Com- ications are easily achieved. These variations can be explained by
position prole shows that nearly all MeOH is separated in the modication of mixtures latent vaporisation heat along the column,

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 11

Table 7
Process stream data for CS1 for PFD.

Stream name Gasoline 9 11 MeOH H2 O Make up H2 O

Temperature, C 35 35 109 51 109 35

Pressure, bar 4.0 4.2 1.4 1.2 1.4 4.0
iC5 , kmol/h 213.66 0.9 0.0 0.9 0.0
2M1B, kmol/h 2.72 0.02 0.00 0.02 0.00
2M2B, kmol/h 20.80 0.11 0.00 0.11 0.00
TAME, kmol/h 0.24 0.01 0.00 0.01 0.00
MeOH, kmol/h 0.07 35.89 0.21 35.53 0.36
H2O, kmol/h 0.39 119.70 119.44 0.02 119.67 0.47

110 250
Vapour
100 Liquid
200

90

Flowrate, kmole/h
Temperature, C

150

80

100
70

50
60

50 0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 0 1 2 3 4 5 6 7 8 9 10 11 12 13
Stage Stage

Fig. 15. C3 column temperature prole for CS1. Fig. 17. C3 column vapour and liquid ow for CS1.

1.0
88

0.9
MeOH 86
H2O
0.8 84
Composition, molar fraction

0.7 82
Temperature, C

0.6 80
0.5 78
0.4 76
0.3
74
0.2
72
0.1
70
0.0 0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7 8 9 10 11 12 13 Stage
Stage
Fig. 18. C2 column temperature prole for CS2.
Fig. 16. C3 column liquid composition prole for CS1.

Table 8
Process stream data for CS2 from PFD.
due to composition change. It should be stressed that MeOH latent Stream name 10 Gasoline 11 14 MeOH
vaporisation heat is very different of that of hydrocarbons. Temper-
Temperature, C 78 86 72 41 83
ature prole (Fig. 21) and liquid composition prole (Fig. 22) show Pressure, bar 5.8 5.2 4.8 1.8 2.2
that the separation is more intensive in column bottoms. Here, iC5 , kmol/h 492.5 214.0 278.5 278.5 0.1
MeOH molar fraction specication of 0.99 gives a MeOH recovery 2M1B, kmol/h 6.0 2.7 3.3 3.3 0.0
ratio of 0.3337. A higher recovery is obtained for lower bottoms 2M2B, kmol/h 38.1 20.9 17.2 17.1 0.1
TAME, kmol/h 0.3 0.2 0.0 0.0 0.0
purity, or a higher reux ratio. From simulation conditions, reboiler
MeOH, kmol/h 103.1 0.0 103.1 68.7 34.4
duty is 10.61 MW and condenser duty is 10.71 MW. Both C2 dis-
tillate stream composition and C3 distillate stream composition
show good agreement between azeotropes composition estimated CS3 simulation results are quite similar regarding C2 distilla-
with SIMULIS Thermodynamics . Slight difference in both compo- tion column performance. Temperature prole (Fig. 24) presents a
sition and temperature proles is created by the 2M1B presence smooth shape showing a good selection of the feed tray. MeOH
and pressure prole throughout the column (Fig. 23). molar fraction specication is achieved, nearly all MeOH being

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
12 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

1.0

0.9

0.8

Composition, molar fr.

0.7

0.6 iC5
0.5
2M2B
0.4

0.3 MeOH
0.2

0.1

0.0
0 1 2 3 4 5 6 7 8 9
Fig. 19. C2 column liquid composition prole for CS2. Stage

Fig. 22. C3 column liquid composition prole for CS2.

800

700
1600
600
1400
Flowrate, kmole/h

500
1200
400
Flow rate, kmole/h

1000
300
800

200
600
Vapour Liquid
100
400

0 Vapour Liquid
0 1 2 3 4 5 6 7 200
Stage
0
Fig. 20. C2 column vapour and liquid ow for CS2. 0 1 2 3 4 5 6 7 8 9
Stage

Fig. 23. C3 column vapour and liquid ow for CS2.

85

80

75 88

70 86
Temperature, C

65 84

82
Temperature, C

60
80
55
78
50
76
45
74
40
72
35
70
0 1 2 3 4 5 6 7 8 9
0 1 2 3 4 5 6 7
Stage Stage

Fig. 21. C3 column temperature prole for CS2. Fig. 24. C2 column temperature prole for CS3.

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 13

1.0 85
80
0.8 iC5 75
Composition, molar fr.

2M1B 70

Temperature, C
0.6
2M2B 65

0.4 60
MeOH
55
0.2 50
45
0.0
0 1 2 3 4 5 6 7 40
Stage 35
0 1 2 3 4 5 6 7 8 9
Fig. 25. C2 column liquid composition prole for CS3.
Stage

1200 Fig. 27. C3 column temperature prole for CS3.

1000 1.0
0.9 iC5 2M1B 2M2B MeOH
Molar flow, kmole/h

800
0.8

Composition, Molar fr.

0.7
600
0.6
400 0.5
0.4
200 Vapour
0.3
Liquid 0.2
0
0 1 2 3 4 5 6 7 0.1
Stage 0.0
0 1 2 3 4 5 6 7 8 9
Fig. 26. C2 column vapour and liquid ow for CS3.
Stage

Fig. 28. C3 column liquid composition prole for CS3.

separated in the distillate (Fig. 25). Vapour ow decreases under
the feed tray (Fig. 26). Analysing C2 feed stream (Stream 10) from
CS2and CS3 (Table 8 and Table 9), C2 feed ow for CS3 is lower than 250
the one for CS2, for both hydrocarbons and MeOH. This is reected
in reboiler and condenser duty lower values compared to CS2:
200
4.66 MW for reboiler and 4.64 MW for condenser. The difference of
C2 feed ow is given by the presence and performance of the two L-
Molar flow, kmole/h

L separation vessels and C3 performance. Aspen HYSYS prediction 150

for L-L equilibrium is different than SIMULIS Thermodynamics
prediction, and it is reected in V1 and V2 vessels separation per-
formances. MeOH molar fraction in the two liquid phases, for both 100
V1 and V2, is 0.50 in Stream 13 (equivalent to L1 phase) and 0.10 in
Stream 12 (equivalent to L2 phase). C3 feed is represented by MeOH
50
enriched phases, feed ow is lower than C3 feed from CS2. This is Vapour
reected in reboiler and condenser duty which are around 10 times Liquid
smaller than those for C3 from CS2: 1.53 MW (reboiler duty) and 0
1.54 MW (condenser duty). C3 temperature prole (Fig. 27) and liq- 0 1 2 3 4 5 6 7 8 9
uid composition (Fig. 28) prole show a more intensive separation Stage
in the bottoms. MeOH recovery ratio in bottom stream is 0.7777,
Fig. 29. C3 column vapour and liquid ow for CS3.
nearly two times higher compared to C3 from CS2. Vapour ow
within the column presents a small decrease under the 4th tray
while the liquid ow increases under the 4th tray (Fig. 29). 3.3. Heat integration
Comparing C2 and C3 feed, products, and recycle from CS2 and
CS3, it can be noticed that V1 and V2 presence lead to a lower Hot and cold stream data extracted from Aspen HYSYS v8.4 for
quantity of material recycled. This has a positive impact not only each case study are presented in Tables 1012. Energy perform-
on operation costs (i.e.: lower reboiler and condenser duty involve ances of each case study are compared based on process Shifted
lower energy demand for pumping), but also in capital cost (i.e.: Composite Curves and Balanced Grand Composite Curve (Klemes
smaller column, smaller pumps, etc.). et al., 2007).

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
14 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

Table 9
Process stream data for CS3 from PFD.

Stream name 10 Gasoline 11 12 13 15 16 MeOH 18 19

Temperature, C 79 86 72 35 35 53 40 83 35 35
Pressure, bar 5.8 5.2 4.8 4.8 4.8 2.7 1.8 2.2 2.7 2.7
iC5 , kmol/h 352.60 213.55 139.05 35.87 103.18 40.81 40.38 0.08 4.95 35.44
2M1B, kmol/h 4.34 2.73 1.61 0.43 1.18 0.49 0.48 0.00 0.06 0.42
2M2B, kmol/h 27.52 20.88 6.64 1.78 4.85 2.04 1.99 0.04 0.26 1.73
TAME, kmol/h 0.25 0.25 0.00 0.00 0.00 0.00 0.00 0.00 0.00 0.00
MeOH, kmol/h 50.73 0.00 50.73 38.63 12.10 43.97 9.51 34.38 5.34 4.17

Table 10
CS1 stream details for heat integration.
(a) 180
Stream for HI PFD stream TS , C TT , C H, MW
150
1 1/2 39.90 75.00 0.53
2 3/4 92.27 75.00 0.27
3 6/7 78.89 92.24 0.21 120

Temperature, C
4 C1 reboiler 156.80 160.08 5.39
5 C1 condenser 84.72 78.71 5.42
6 8/9 78.71 35.00 0.56
90
7 C3 reboiler 107.43 108.80 1.33
8 C3 condenser 68.83 50.52 1.17
9 12/H2 O 108.82 35.00 0.20
60
10 TAME/P1 160.08 30.00 0.27

30 Hot Composite
Table 11
CS2 stream details for heat integration. Cold Composite

0
Stream for HI PFD stream TS , C TT , C H, MW 0 20 40 60 80 100
1 1/2 29.69 75.00 0.68 Enthalpy, MW
2 3/4 92.38 75.00 0.27
3 6/7 78.85 92.33 0.21 (b) 240
4 C1 reboiler 156.60 160.00 5.29 Process Grand Composite
5 C1 condenser 84.88 78.74 5.33
6 9/10 57.61 78.43 0.63 200 UtilityGrand Composite
7 C2 reboiler 85.63 85.95 9.30
8 C2 condenser 71.70 71.60 9.35
Temperature, C

9 C3 reboiler 57.31 82.76 10.20 160

10 C3 condenser 40.75 40.63 10.71
11 TAME/P1 160.00 30.00 0.27
12 Gasoline/P2 85.95 30.00 0.65 120

80
Table 12
CS3 stream details for heat integration.

Stream for HI PFD stream TS , C TT , C H, MW 40

1 1/2 29.40 75.00 0.68

2 3/4 92.38 75.00 0.27 0
3 6/7 78.85 92.33 0.21 0 10 20 30 40 50
4 C1 reboiler 156.60 160.00 5.29 Enthalpy, MW
5 C1 condenser 84.88 78.74 5.33
6 9/10 62.96 78.71 0.33
Fig. 30. Shifted Composite Curve (a) and Balanced Grand Composite Curve (b) for
7 C2 reboiler 85.90 86.00 4.66
CS1.
8 C2 condenser 71.70 71.64 4.64
9 C3 reboiler 64.86 83.17 1.53
10 C3 condenser 40.75 40.57 1.54
value for cold utilities temperature is chosen to 77.5 C. In this
11 V1 (11/12, 11/13) 71.64 35.00 0.32
12 TAME/P1 160.00 30.00 0.27
respect, BFW target temperature is modied to 75 C. Grand Com-
13 Gasoline/P2 86.00 30.00 1.94 posite Curve, Fig. 30b, is used for utility selection: thus, only BFW,
shaft water, LP steam and MP steam are selected. Two utility Pinch
point appear in the sink zone at 54.61 C (i.e. 49.61 C for cold
Minimum optimum temperature difference for HEN design is streams and 59.61 C for hot streams) and in the source zone at
determined from total cost value, which accounts for investment 139.52 C (i.e. 134.52 C for cold streams and 144.52 C for hot
and operational cost. SPRINT software utilises specic method streams).
for minimum optimum temperature difference estimation (Kemp, For CS2, minimum optimum temperature is Tmin,opt = 13 C.
2007). Minimum optimum temperature, determined for CS1 is Process Pinch point is located at 65.30 C (i.e. 58.80 C for cold
10 C. This temperature difference allows to recover 0.77 MW, as streams and 71.80 C for hot streams). Process to process heat
given by Shifted Composite Curves, Fig. 30a. Process Pinch point is recovery is 7.02 MW, as given by Shifted Composite Curves,
located at 87.27 C (i.e. 82.72 C for cold streams and 92.27 C for Fig. 31a. Balanced Grand Composite Curves give the utility selec-
hot streams). To avoid heat exchange across the Pinch, a maximum tion, placement (Fig. 31b) satisfying minimum heat and cold utility

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 15

(a) 180 (a) 180

Hot Composite Hot Composite

150 Cold Composite 150 Cold Composite

120

Temperature, C
Temperature, C

120

90 90

60 60

30 30

0 0
0 10 20 30 40 50 0 4 8 12 16 20 24 28

Enthalpy, MW Enthalpy, MW

(b) 200 (b) 240

Process Grand Process Grand Composite
Composite Utility Grand Composite
Utility Grand 200
160
Composite

Temperature, C
160
Temperature, C

120
120

80
80

40 40

0 0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14
Enthalpy, MW
Enthaply, MW
Fig. 32. Shifted Composite Curve (a) and Balanced Grand Composite Curve (b) for
Fig. 31. Shifted Composite Curve (a) and Balanced Grand Composite Curve (b) for
CS3.
CS2.

consumption. Assigned utilities for CS2 are: SW, LP and MP steam.
One utility pinch point is present in the source zone at 137.22 C
(i.e. 130.72 C for cold streams and 143.72 C for hot streams).
For CS3, minimum optimum temperature difference is
Tmin,opt = 9 C. Process Pinch point is located at 80.38 C (i.e.
75.88 C for cold streams and 84.88 C hot streams). Process to
process heat recovery is 2.19 MW, as given by Shifted Composite
Curves, Fig. 32a. Grand Composite Curve allows to selected utilities
(shaft water, BFW, LP and MP steam), as shown in Fig. 32b. Ini-
tially proposed BFW target temperature 85 C has to be modied
to a target temperature lower than 69.88 C, to avoid heat transfer
across the Pinch point. BFW selected temperature is 68 C, it gen-
erates a utility Pinch point located at 54.56 C in BGCC (i.e. 50.06 C
for cold streams and 59.06 C for hot streams). Another two util-
ity pinch points appear in BGCC: one in the sink zone at 36.14 C
(i.e.: 31.64 C for cold streams and 40.64 C for hot streams) and
the other one in the source zone at 139.22 C (i.e. 134.72 C for cold
streams and 143.72 C for hot streams). HEN for CS3 is presented
in Fig. 33. Initial HEN design consists of 31 heat exchangers. Study-
ing heat exchangers area, duty and connections, ve of them are
eliminated during the retrot stage, having very low values. Pro-
posed HEN allows recovery of 1.72 MW with a small increase in LP
and SW consumption, compared to minimum requirements, as pre-
sented in Table 13. More information regarding exchangers design
is presented in supporting information section of the paper.
In Table 14 the results for all three case studies are summ-
arised. Minimum hot and cold utilities required are higher for
CS2, decreasing then for CS3, and respectively for CS1. Maxi-
mum heat recovery is achieved for CS2. Maximum heat recovery
through BFW is achieved in CS3, while in CS2, BFW is not cho-
sen as utility. Overall heat recovery for CS3, considering also the
recovery through BFW, is 9.87 MW compared to 7.47 MW from
CS2.
3.4. Environmental impact
Results from WAR GUI are presented in Table 14. From all indices
needed for PEI determination, acidication potential (AP) has the
highest inuence, while ozone depletion potential (ODP) has the
lowest contribution. For each case study, Iout value decreases when
considering different fuels: when oil is compared to coal, the
decrease is 34.8% for CS3, while for CS1 and CS2 the difference
is around 75%. Differences between Iout for coal and gas are higher:
78% for CS2, 83% for CS3 and 93% for CS4. Lowest values are achieved
when natural gas is used as fuel for all three case studies. Com-
paring Iout values of natural gas for each case study, the decrease
is as follows: CS2 to CS3 (48%) and CS3 to CS1 (49%). The low-
est environmental impact is achieved for CS1 when using natural
gas as fuel for utility generation, Iout = 0.0118 PEI/kg, followed by
CS3 and CS2, Iout = 0.0405 PEI/kg. Regarding the generated impact,
this has negative values, meaning that the process products have
a lower environmental impact than the feedstock. When coal is

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
16 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx

Fig. 33. HEN design at MER for CS3.

Table 13
Heat integration calculations results for CS1, CS2 and CS3.

Case QCmin , MW QHmin , MW Process heat recovery, MW Utility load, MW

SW CW BFW LP MP

CS1 7.13 6.69 0.77 0.81 6.31 1.31 5.39

CS2 19.56 19.30 7.01 19.56 14.00 5.29
CS3 12.13 10.52 2.19 1.62 0.89 9.62 5.23 5.29
CS3 (non-integrated) 14.33 6.35 5.29
CS3 (HEN design) 1.72 1.67 0.87 9.62 5.42 5.29

Table 14
Output and generated environmental impact for CS1, CS2 and CS3.

Case Fuel type

Coal Oil Natural gas

Iout , PEI/kg Igen , PEI/kg Iout , PEI/kg Igen , PEI/kg Iout , PEI/kg Igen , PEI/kg

CS1 0.0476 1.3720 0.0118 1.4080 0.0030 1.4160

CS2 0.0476 1.3720 0.0405 1.3790 0.0104 1.4090
CS3 0.0318 1.3860 0.0207 1.3970 0.0053 1.4130

used as fuel, CS3 has a slightly better environmental indicator,
Igen = 1.386 PEI/kg, compared with CS1 and CS2 which have nearly
same value, Igen = 1.372 PEI/kg. For the other fuels, CS1 has slightly
better performances compared with CS2 and CS3. The values for
all indices, used for environmental impact calculation, are given in
supporting information of the paper.
includes main and auxiliary equipment cost, project engineering
(basic and detailed engineering). A key role in project feasibility
is played by raw materials and product prices. Very small varia-
tion of these values leads to better or worse project performance.
Should be reminded that TAME price is considered for a lower
purity. Project total cost includes direct and indirect costs for both

3.5. Techno-economic evaluation

A summary of techno-economic evaluation results, as given by

Aspen Economic Analyzer , is presented in Table 15. Operating cost
includes: raw material cost, utilities cost, operating labour, main-
tenance, management, staff and supervision salaries. Capital cost

Table 15
Techno-economic analysis indicators for CS3.

Capital cost, MD 11.99

Total operating cost, MD /year 102.05
Total raw material cost, MD /year 93.35
Total product sales, MD /year 109.77
Project net present value (NPV)a , MD 6.16 Fig. 34. Project total cost for CS3. Project total cost include direct cost (equipment,
piping, instrumentation, electrical, civil, and insulation) and indirect cost (engineer-
a
NPV value is considered at the end of the project. ing, other costs, and contingency).

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx 17

10

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14
NPV, M -5

-10

-15

-20

-25

-30

-35
Project life, year

Fig. 35. NPV variation through project life.

capital and operating costs, Fig. 34. Project NPV value variation is
presented in Fig. 35. Positive values of NPV are achieved starting
10th year of the project life.
4. Conclusions
The paper presents a systematic analysis for TAME synthesis
process by simulation, process integration, environmental impact
and techno-economic analysis. The methodology has novelty fea-
tures from two points of view: it is extended from thermophysical
bases (study of feasibility with Residual Curve Maps) to nali-
sation of process conguration. Then, the analysis is performed
with multitude of state-of-the-art software tools (thermophysical
data base server, process simulator, process integration, environ-
mental impact assessment and economic evaluation). These tools
allowed the investigation in parallel of three techniques for the
separation of TAME synthesis reactor efuent, which generated
three case studies. The proposed separation sequence presents an
original solution, based on combination of pressure swing with
liquidliquid equilibrium. To underline the importance of this new
separation sequence compared with other existing or proposed
alternatives three case studies are presented. The rst case study
(CS1) represents classical process, with hydrocarbonMeOH sep-
aration by L-L water extraction. The other two case studies are
based on pressure swing, given the thermophysical behaviour
of hydrocarbonMeOH mixture azeotropes, as described by RCM
build with SIMULIS Thermodynamics . In the second case study
(CS2) separation of hydrocarbonMeOH mixture is based on
hydrocarbonMeOH azeotrope composition signicant variation
with pressure. In the third case study (CS3), additionally to this
technique, MeOHhydrocarbon L-L equilibrium, as highlighted by
RCM build in SIMULIS Thermodynamics is exploited, generating
an original separation sequence. Consequently, Aspen HYSYS v8.4
simulation results indicate that the combination of these two tech-
niques in CS3 provides best separation performance. The reactor
and the separation column for TAME have nearly the same cong-
uration for all three case studies, some differences appear due to
numerous recycles. Those recycles enable better recovery of com-
ponents. The dimension of equipment is comparable in the case of
CS1 and CS3.
The environmental impact performance for some fuels (as
coal) CS3 has better indicators (Igen = 1.386 PEI/kg) compared
with CS1 (Igen = 1.372 PEI/kg). Best results are achieved for CS3,
when natural gas is used as fuel for utility generation. For other
fuels CS1 and CS3 have similar performance as indicated in
Table 14. Techno-economic evaluation of CS3, considering the
non-integrated scheme, shows the possibility to implement this
solution. Plant total cost, accounting for above-mentioned assump-
tions, is 12 MD , which represents approximately 10% from annual
operating cost. Based on these results, the work will continue in
future considering CS3 optimisation aspects.
Considering all these aspects, this paper deals for the rst time
with the combination of pressure swing with liquidliquid equilib-
rium separation, providing the best conguration for the separation
of the hydrocarbonMeOH mixture. Careful quantitative evaluation
is based on process simulation, process integration, environmental
impact assessment and economical evaluation of performances, for
each case study.
Acknowledgements
The authors acknowledge the nancial support to POSCCE
Romanian Government - EU programme, project ID 652.
For one of the authors, POA, this research was nan-
cially supported by Sectoral Operational Programme Human
Resources Development, nanced from the European Social Fund
and by the Romanian Government under the contract number
POSDRU/156/1.2/G/135764 Improvement and implementation of
universitary master programmes in the eld of Applied Chemistry and
Materials Science ChimMaster.
Appendix A. Supplementary data
Supplementary data associated with this article can be found,
in the online version, at http://dx.doi.org/10.1016/j.compchemeng.
2015.05.020
References
Al Arfaj MA, Luyben WL. Plantwide control for TAME production using reactive
distillation. AIChE J 2004;50(7):146273.
Aspen process economic analyzer V8. 0. Users guide. Burlington, USA: Aspen Tech-
nology; 2013.
Barret WB, van Batenb J, Martina T. Implementation of the waste reduction (WAR)
algorithm utilizing owsheet monitoring. Comput Chem Eng 2011;35:26806.
Bonet-Ruiz AE. [Ph.D. thesis] Modelling and simulation of continuous catalytic dis-
tillation processes [Ph.D. thesis]. Bucharest, Romania: University POLITEHNICA
of Bucharest; 2012 [Online] Available: http://romdoc.upb.ro/record/623?ln=fr
[accessed 01.10.13].
Bonet-Ruiz AE, Bonet-Ruiz J, Plesu V, Bozga G. Environmental performance assess-
ment for reactive distillation processes. Resour Conserv Recycl 2010;54:31525.

Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
G Model
CACE-5212; No. of Pages 18 ARTICLE IN PRESS
18 O.A. Perederic et al. / Computers and Chemical Engineering xxx (2015) xxxxxx
Cabezas H, Bare C, Mallick K. Pollution prevention with chemical process simulators:
the generalized waste reduction (WAR) algorithm-full version. Comput Chem
Eng 1999;23:62334.
Doherty MF, Malone MF. Conceptual design of distillation systems. New York, USA:
McGraw-Hill; 2001.
Ferreira MV, Riberibo AM, Loureiro JM. Experimental and simulation studies of TAME
synthesis in a xed bed reactor. Ind Eng Chem Res 2007;46(4):110513.
Gao X, Wang F, Li H, Li X. Heat-integrated reactive distillation process for TAME
synthesis. Sep Purif Technol 2014;132:46878.
Haotao L, Kai L, Weisong L, Ye L, Ming X, Chunjian X. Comparison of pressure-swing
distillation and extractive distillation methods for isopropyl alcohol/diisopropyl
ether separation. Ind Eng Chem Res 2014;53:1516782.
Integrated pollution prevention and control bureau best available techniques
(BAT) reference document for the rening of mineral oil and gas; 2015. p.
756, http://eippcb.jrc.ec.europa.eu/reference/BREF/REF BREF 2015.pdf [con-
sulted 13.03.15].
Kemp IC. Pinch analysis and process integration. A user guide on process integration
for the efcient use of energy. 2nd ed. Amsterdam, Netherlands: Elsevier; 2007.
Kim J, Henao CA, Johnson TA, Dedrick DE, Miller JE, Stecheld EB, et al. Methanol
production from CO2 using solar-thermal energy process development and
techno-economic analysis. Energy Environ Sci 2011;4:312232.
Kiviranta-Pkknen P, Struckmann L, Krause A. Comparison of the various kinetic
models of TAME formation by simulation and parameter estimation. Chem Eng
Technol 1998;21:3216.
Klemes JJ. 36 Epilogue: the importance of problem formulation and data extrac-
tion in process integration in Woodhead publishing series in energy. In: Klemes
JJ, editor. Handbook of Process Integration (PI). Philadelphia, USA: Woodhead
Publishing Limited; 2013.
Klemes JJ, Fridler F, Bulatov I, Varbanov PS. Sustainability in the process industry.
Integration and optimization. New York, USA: McGraw-Hill; 2007.
Klemes JJ, Varbanov PS, Wan Alwi SR, Manan ZA. Process integration and intensi-
cation. Saving energy, water and resources. Berlin, Germany: Walter de Gruyter
Ed; 2014.
Llandosa E, Monton JB, Burguet M. Separation of di-n-propyl ether and n-propyl
alcohol by extractive distillation and pressure-swing distillation: computer sim-
ulation and economic optimization. Chem Eng Process 2011;50:126674.
Libert V, Hector T, Gmehling J. Chemical equilibrium conversion of the tert-
Amyl-methyl ether synthesis in the presence of n-pentane, tetrahydrofuran, or
benzene. Ind Eng Chem Res 2010;49:44129.
Lyu W, Li C, Li J, Xin F. Simulation and optimization on etherication of light gasoline
in a process of two xed-bed reactors and one distillation column. Comput Chem
Eng 2014;68:4755.
Luyben WL. Comparison of pressure-swing and extractive-distillation methods for
methanol-recovery systems in the TAME reactive-distillation process. Ind Eng
Chem Res 2005;44(15):571525.
Luyben WL. Comparison of extractive distillation and pressure-swing distillation for
acetone/chloroform separation. Comput Chem Eng 2013;50:17.
Luyben WL. Principles and case studies of simultaneous design. New Jersey, USA:
John Wiley & Sons Inc; 2011.
Mao W, Wang X, Wang H, Chang H, Zhang X, Han J. Thermodynamic and kinetic study
of tert-amyl methyl ether (TAME) synthesis. Chem Eng Process 2008;47:7619.
Murat Sen S, Henao CA, Braden DJ, Dumesic JA, Maravelias CT. Catalytic conversion
of lignocellulosic biomass to fuels process development and technoeconomic
evaluation. Chem Eng Sci 2012;67:5767.
Perederic OA. TAME production plant from enriched C5 fraction from uid cat-
alytic cracking (FCC) [Graduation Project]. Bucharest, Romania: University
POLITEHNICA of Bucharest; 2013 [Online] Available: http://romdoc.upb.ro/
record/624/les/ [accessed 03.10.13].
Perederic OA, Plesu V, Iancu P, Bumbac G, Bonet-Ruiz AE, Bonet-Ruiz J, et al. Simu-
lation and process integration for TAME synthesis. In: Klemes JJ, Varbanov PS,
Liew PY, editors. Proceedings of the 24th European Symposium on Computer
Aided Process Engineering Elsevier, vol. 33; 2014. p. 148994.
Plesu AE, Bonet J, Plesu V, Bozga G, Galan MI. Residue curves map analysis
for tert-amyl methyl ether synthesis by reactive distillation in kinetically
Please cite this article in press as: Perederic OA, et al. Simulation and process integration for tert-amyl-methyl ether (TAME) synthesis.
Computers and Chemical Engineering (2015), http://dx.doi.org/10.1016/j.compchemeng.2015.05.020
controlled conditions with energy-saving evaluation. Energy 2008;33(10):
157289.
Pulido JL, Martnez EL, Maciel MR, Filho RM. Heat integrated reactive distillation
column (r-HIDiC): implementing a new technology distillation. Chem Eng Trans
2011;24:13038.
Rashed J. [Ph.D. thesis] Model-based retrot design and analysis of petrochemi-
cal processes [Ph.D. thesis]. Lyngby, Denmark: Danmarks Tekniske Universitet;
2009.
Rihko LK, Krause OA. Kinetics of heterogeneously catalyzed tert-amyl methyl ether
reactions in the liquid phase. Ind Eng Chem Res 1995;34:117280.
SIMULIS Thermodynamics v1. 4, thermodynamic properties and phase equilibria
calculation server. Labge, France: ProSim S.A; 2013.
Subawalla H, Fair JR. Design guidelines for solid-catalyzed reactive distillation sys-
tems. Ind Eng Chem Res 1999;38(10):3696709.
Syed FH, Egleston C, Datta R. Tert-amyl methyl ether (TAME). Thermodynamic anal-
ysis of reaction equilibria in the liquid phase. J Chem Eng Data 2000;45:31923.
Seider WD, Seader JD, Lewin D. Process design principles. Synthesis, analysis and
evaluation. New York, USA: John Wiley & Sons Inc; 2004.
Sikdar SK, El-Halwagi M, editors. Process design tools for the environment. London,
UK: Taylor & Francis; 2001.
Su WB, Chang JR. Modelling and simulation of tubular reactor in the tert-amyl methyl
ether synthesis process. Ind Eng Chem Res 2000;39:41407.
Turton L, Bailie R, Whiting W, Shaeiwitz JA, Bhattacharyya D. Analysis, synthesis
and design of chemical processes. 4th ed. New Jersey, USA: Pearson Education
International; 2014.
Vanaki A, Eslamloueyan R. Steady-state simulation of a reactive internally heat inte-
grated distillation column (R-HIDiC) for synthesis of tertiary-amyl methyl ether
(TAME). Chem Eng Process: Process Intensif 2012;52:217.
Wu YC, Chien I-L, Luyben WL. Two-stripper/decanter owsheet for methanol
recovery in the TAME reactive-distillation process. Ind Eng Chem Res
2009;48(23):1053240.
Yang Z, Cui X, Jie H, Yu X, Zhang Y, Feng T, et al. Kinetic study and process simulation
of transesterication of methyl acetate and isoamyl alcohol catalyzed by ionic
liquid. Ind Eng Chem Res 2015;54:120415.
Further reading
Amberlyst 35 wet characteristics. The Dow Chemical Company.
[Online]. Available: www.dow.com/assets/attachments/business/
process chemicals/amberlyst/amberlyst 35wet/tds/amberlyst
35wet.pdf [accessed 08.03.13].
European Commission Institute for Prospective Techno-
logical Studies, Sustainable Production and Consumption Unit,
European IPPC Bureau, Best Available Techniques (BAT) Ref-
erence Document for the Rening of Mineral Oil and Gas.
[Online]. Available: http://eippcb.jrc.ec.europa.eu/reference/BREF/
FD REF July 2013online.pdf [accessed 10.10.13].
Environmental Protection Agency of USA. Chemical Process Sim-
ulation for Waste Reduction: WAR Algorithm. [Online]. Available:
www.epa.gov/nrmrl/std/war/sim war.htm [accessed 13.03.13].
Plant Cost Index, Chemical Engineering [Online] Available:
www.chemengonline.com/pci-home [accessed 15.08.14].
Plant Cost Index, Chemical Engineering 2002 [Online] Available:
www.chemengonline.com/Assets/File/CEPCI 2002.pdf [accessed
15.08.14].
SPRINT. [Online]. Available: www.ceas.manchester.ac.uk/
research/centres/cpi/software [accessed 10.10.13].