See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/271196739

Multiobjective Optimization of Methanol Synthesis Loop from Synthesis Gas

via a Multibed Adiabatic Reactor with Additional Interstage CO2 Quenching

Article  in  Energy & Fuels · December 2014

DOI: 10.1021/ef502073b

CITATIONS READS

3 7,709

4 authors, including:

Saad Ali Al-Sobhi A. Elkamel

Qatar University University of Waterloo
18 PUBLICATIONS   52 CITATIONS    483 PUBLICATIONS   5,319 CITATIONS   

SEE PROFILE SEE PROFILE

Eric Croiset
University of Waterloo
136 PUBLICATIONS   4,752 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Modeling of pipeline networks View project

Optimal storage and renewable distributed energy resources planning in distribution network View project

All content following this page was uploaded by Saad Ali Al-Sobhi on 05 February 2018.

The user has requested enhancement of the downloaded file.

 Article

pubs.acs.org/EF

Multiobjective Optimization of Methanol Synthesis Loop from

Synthesis Gas via a Multibed Adiabatic Reactor with Additional
Interstage CO2 Quenching
Abdulaziz Alarifi,* Saad Alsobhi, Ali Elkamel, and Eric Croiset
University of Waterloo, Waterloo, Ontario N2L 3G1, Canada

ABSTRACT: The conversion of syngas derived from natural gas into methanol has been considered a relatively clean and
environmentally friendly process. However, carbon dioxide is emitted as a result of using natural gas as fuel in the reformer
furnace combustion zone to supply the heat required for endothermic reforming reactions. Carbon dioxide is a primary
greenhouse gas emitted as flue gas from the reformer and has been contributing to global warming over the past few decades.
Thereby, environmental regulations for new and existing industrial facilities have been enforced to mitigate the adverse effects of
carbon dioxide emission. In this research, multiobjective optimization is applied for the operating conditions of the methanol
synthesis loop via a multistage fixed bed adiabatic reactor system with an additional interstage CO2 quenching stream to
maximize methanol production while reducing CO2 emissions. The model prediction for the methanol synthesis loop at steady
state showed good agreement against data from an existing commercial plant. Then, the process flowsheet was developed and
fully integrated with the Genetic Algorithms Toolbox that generated a set of optimal operating conditions with respect to upper
and lower limits and several constraints. The results showed methanol production was improved by injecting shots of carbon
dioxide recovered from the reformer at various reactor locations.

■ INTRODUCTION
The demand for alternative cleaner energy is growing rapidly,
more CO2 emission and requires a relatively larger catalyst
volume as the temperature profile reaches the equilibrium
leading to an interest in methanol production and this demand maximum path.9 Environmental regulations are aiming at
is expected to increase further in the next decades. Methanol is reducing greenhouse gases especially carbon dioxide and
extensively used as a raw material for formaldehyde production, nitrous oxide. The production of methanol from natural gas
methyl tert-butyl ether (MTBE), and acetic acid. It is also used via steam methane reforming is considered a relatively clean
in a direct methanol fuel cell (DMFC), where methanol is process. However, a major source of CO2 is generated in the
directly oxidized with air to water and carbon dioxide while reformer furnace combustion zone as result of burning fuel gas
producing electricity.1,2 Methanol is expected to play a major to supply the heat required for the reforming reactions. As a
role in our independence from the traditional gasoline derived consequence, CO2 ends up being flue gas vented from the
from fossil fuel as an environmentally sustainable fuel. It has reformer to the atmosphere. Hence, a carbon dioxide recovery
been reported by the U. S. Environmental Protection Agency system is implemented to recover vented gas and utilize the
(EPA) that switching gasoline to methanol would reduce 90% CO2 recovered as a feed to the upstream or downstream side of
of the incidents caused by fuel.3,4 Moreover, methanol can be the methanol synthesis to obtain the desired syngas molar ratio
used as a diesel replacement that does not produce particulates H2/CO of 2.1 for producing methanol.
or soot when combusted. It also burns at a lower temperature The modeling and optimization of methanol process have
than diesel; consequently, very low NOx emission occurs.5 The been addressed in the literature. For example, Lie et al.16
first synthetic methanol plant was introduced by Badische presented a mixed integer programming (MIP) model of
Anilin-und-Soda-Fabrik (BASF) in 1923;6,7 the process polygeneration technology (producing both electricity and
synthesized methanol at 400 °C and 200 atm. During the methanol). Their case study considers China with the
1960s, Imperial Chemical Industries (ICI) developed a low- expectation that methanol would substitute 1−5% of the
pressure methanol process using a copper-based catalyst that projected oil consumption from 2010 to 2035, and they
completely altered the technology of methanol synthesis.8 This concluded that polygeneration technologies are superior to
process operates at 220−300 °C and 50−100 atm. The ICI conventional stand-alone technologies. Julian-Duran et al.17
quench-cooled converter consists of a number of adiabatic performed a techno-economic and environmental impact study
catalyst beds installed in series in one pressure vessel.9,10 A of methanol produced from shale gas. Although a CO2 recovery
quench reactor is the most common methanol converter design system was implemented and considered in their flowsheet
that has been on line for more than 40 years and has proven its analysis, CO2 was either vented to the atmosphere or recycled
robustness and reliability for capacities ranging from 50 up to back to the reforming section. In our analysis, the captured CO2
3000 metric tons per day (MTPD).11,12 In comparison with
other pseudoisothermal converter types such as tubular (Lürgi), Received: September 15, 2014
double-tube heat-exchange (Mitsubishi),13−15 and radial Revised: December 11, 2014
isothermal (Methanol Casale), the quench converter produces Published: December 29, 2014

© 2014 American Chemical Society 530 DOI: 10.1021/ef502073b

Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

Figure 1. Process flow diagram of methanol synthesis loop via natural gas reforming.

is directly sent to methanol reactor to investigate methanol to-carbon is required to suppress the coke formation (3) over
production enhancement. A multiobjective optimization frame- the catalyst and to increase hydrogen production by the water
work is used to propose new operating conditions of low- gas shift reaction.18
pressure methanol synthesis loop via a multistage fixed bed
adiabatic reactor system with additional interstage CO2 CO + H 2O ⇌ CO2 + H 2 (2)
quenching stream to maximize production while reducing
CO2 emissions. 2CO(g) → CO2 (g) + C(s) (3)

■ PROCESS DESCRIPTION
A typical methanol process includes desulfurization of natural
The SMR process consists of two major parts: a packed
catalyst in tubes and a furnace to heat the reformer tubes. The
desired syngas molar ratio for methanol synthesis is 2.1
gas, steam reforming, methanol synthesis, and methanol according to the following formula: stoichiometric ratio (SR) =
refining. (H2 − CO2)/(CO + CO2). As illustrated in Figure 1, the
syngas produced in the reformer is passed through a heat
exchanger, generating medium pressure steam, while the syngas
is cooled down and then fed into a heat recovery system for
further cooling to a mild temperature. The steam associated
with syngas is condensed and used as boiler feedwater. Finally,
dry syngas is compressed at 50−100 atm and then heated up to
200−250 °C in order to prepare syngas for the methanol
synthesis reactions.
Methanol Synthesis. A commercial methanol synthesis
unit is generally licensed by various process and catalyst
suppliers. Each licenser has developed its own schematic
diagram and catalyst structure. The heart of the ICI low
pressure process is a quench converter packed with activated
catalyst (Cu−ZnO/Al2O3) and operated under 10 MPa and
200−300 °C. The ICI’s low pressure methanol process with a
quench reactor system is widely used, and therefore, many
methanol plants still use this technology.8 This converter
consists of multiple adiabatic beds in series where the synthesis
reaction is quenched by additional cooled synthesis gas
Figure 2. ICI Low pressure quench converter.25 between the beds. Johnson Matthey Catalyst (formally ICI)
collaborated with Methanol Casale to develop a new version of
the converter that has the same quenching concept and is called
Syngas Production. Production of syngas is traditionally axial radial converter (ARC). The ARC is a quench type
performed through steam reforming of natural gas (methane) converter but with separate catalyst bed rather than a single
as follows: continuous bed in order to improve gas distribution. The
CH4 + H 2O → CO + 3H 2 (1) methanol reactions in the low pressure synthesis loop can be
summarized as
Steam reforming of methane and water gas shift reactions (1)
and (2) are taking place simultaneously. A high ratio of steam- CO + 2H 2 ⇌ CH3OH (4)

531 DOI: 10.1021/ef502073b

Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

CO2 + 3H 2 ⇌ CH3OH + H 2O (5) dT

Nrxn
−uρg Cpg − ρb ∑ (−ΔHr ,n)ηnrn = 0
CO2 + H 2 ⇌ CO + H 2O dx (12)
(6) n=1

Many methanol kinetic models proposed in the literature where Fi is the flow rate of component i and T is the
have been derived based on Langmuir−Hinshelwood mecha- temperature of the reactants inside the converter. The Ergun
nisms. However, Vanden Busshce and Froment19 experiments equation is used to calculate pressure drop through the catalyst
involved a commercial Cu/ZnO/Al2O3 catalyst under the packed beds
typical industrial conditions. The following expressions were ⎛ ⎞
1 dP G ⎛ 1 − φ ⎞⎜ 150(1 − φ)μg
obtained for the rate of methanol synthesis and the reverse =− ⎜ 3 ⎟⎜
+ 1.75G⎟⎟
water gas shift reaction: L dx ρg gc dp ⎝ φ ⎠⎝ dp ⎠
⎡ (13)
1 PH2OPCH3OH ⎤
k1PCO2PH2⎢1 − keq1 PH3 2PCO2 ⎥
i
The initial conditions are x = 0, F (0) = Fiinlet,
T(0) = Tinlet,
⎣ ⎦
rCH3OH = and P(0) = Pinlet. The methanol conversion reactions are
⎛ PH2O ⎞ 3
exothermic; thereby, the heat of reaction is removed in each
⎜1 + k + kad2PH0.52 + kad3PH2O⎟
⎝ ad1 PH2 ⎠ (7) stage to increase the conversion per pass through the reactor.
Methanol Refining. The crude methanol from methanol
⎡ PH OPCO ⎤ synthesis is purified by the use of distillation columns.
k 2PCO2⎢1 − keq2 P 2 P ⎥
⎣ H2 CO2 ⎦ According to U. S. Federal grades of methanol, there are two
rRWGS = methanol grades: grade AA and grade A. The methanol
⎡ PH2O 0.5 ⎤
⎢⎣1 + kad1 PH2 + kad2PH2 + kad3PH2O⎥⎦ (8)
minimum weight of grade AA and grade A are the same
99.85%, whereas water maximum weights are 0.1% and 0.15%
The equilibrium correlations are given as follows: respectively. Typically grade A of crude methanol requires two
distillation columns, whereas grade AA requires three
3066 distillation columns. The reactor effluent is cooled by heat
log10(Keq1) = − 10.592
T (9) exchange with the fresh syngas stream and boiler feedwater, and
then unreacted syngas/crude methanol separation is carried out
2073 in a vessel under pressure. The unconverted syngas is recycled
log10(Keq2) = − + 2.029
T (10) to the converter through a circulator compressor. A small purge
of unreacted syngas is taken from the recycled gas to maintain a
The reaction rate and adsorption constants are of the
minimum level of impurities and hydrogen in the loop and used
Arrhenius form, with more details given elsewhere.20−23 The
as fuel in the reformer. The condensed product (water/
reactants (CO, CO2, and H2) are supplied by makeup syngas.
methanol) then proceeds to the crude methanol storage tank
The converter consists of a pressure vessel with five catalyst
after being flashed at a pressure above atmospheric pressure to
beds, with interstage cooling shots accomplished by quenching
remove gases physically dissolved at the separation pressure.
as shown in Figure 2. Part of the fresh syngas enters the top
The crude methanol is sent to a distillation column system
bed, and the residue is entered as an interstage quench. This
normally consisting of two columns: the first column operates
interstage quench gas is added to the gas reacting within the
at elevated pressure and removes syngas and hydrocarbon
converter by means of gas mixing system. All the five beds are
impurities from the methanol/water mixture; the second
loaded with Haldor Topsøe methanol synthesis catalyst MK-
column separates methanol from water at atmospheric pressure.
101. By reduction of the catalyst, a theoretical amount of 125
The crude methanol is pumped from the storage tank into a
kg of H2O is produced from one metric ton of catalyst. The
topping column for removal of light impurities, such as methyl
catalyst activity will decrease slowly during normal operation
formate, in the overhead. This column is operated under
although during a short period, approximately 1−3 months
elevated pressure, slightly higher than atmospheric pressure.26
after initial start-up, the activity will decrease more rapidly than
The bottom stream from the topping column is fed into the
in the rest of the period. The catalyst has an expected life of
refining column where the purified methanol is separated as a
four years provided that it is properly protected against poisons
top stage distilled product and the wastewater is discharged
such as sulfur and chlorine.24
from the system.

■
Subsequently, part of the syngas stream enters the reactor
after preheating by the reactor effluent. The remaining syngas is METHODOLOGY AND SOLUTION STRATEGY
used as quenched shot gas. The reactor effluent is cooled by
heat exchange with the feed and boiler feedwater, and then An end-user interface is developed to employ a simulation-
syngas/crude methanol separation is carried out in a vessel based multiobjective optimization scheme (a combination of a
under pressure. The gas is recycled after purging a small part of Microsoft Excel (MS) Visual Basic for Applications (VBA)
the feed to keep the level of impurities and hydrogen in the macro and an optimization technique developed in Matlab).
loop within limits. The purge gas is used as fuel for the Then, it is used to optimize the methanol process by means of
reforming section. Aspen Plus flowsheet simulation software. To accomplish this
A mathematical model for adiabatic converter associated with task, a rigorous reactor model is developed in Aspen Custom
interstage quenching is developed as follows. The mass and Modeler and imported into Aspen Plus. The Soave−Redlich−
energy balances for the fluid phase are Kwong (SRK) equation-of-state is implemented to improve
phase equilibrium calculations in water−hydrocarbon systems
Nrxn and equation based calculations by using composition
1 dFi
− + Ac ρ b ∑ vn,iηnrn = 0 independent fugacity.24 Most of the units involved in the
L dx n=1 (11) process are successfully described by the available built-in
532 DOI: 10.1021/ef502073b
Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

Table 1. Parameters of Gamultiobj Function −Multiobjective

Optimization Using Genetic Algorithm
name of parameters value
population type double vector
population size 200
crossover fraction 0.8 (default)
elite count
fitness functions intermediate, ratio = 1
mutation fraction use constraint dependent default
generation use default: 200 × number of variables.
ParetoFraction use default: 0.35
DistanceFcn crowding distance

Figure 3. Sketch of the programs’ integration and data exchange. in the Matlab command window and saved in Matlab files for
further analysis. The complete process of methanol synthesis
involving catalytic reaction, separation, and purification has
models in the Aspen Plus model library. A nonrandom two- been simulated in Aspen Plus as shown in Figure 4. Although
liquid (NRTL) thermodynamic package is used for predicting all these steps are important, the conversion of syngas to
the VLE of the mixture of light hydrocarbon and methanol/ methanol is the most critical step because all of the existing
water with high accuracy. equipment are being used to their fullest advantage by
The Aspen Plus user interface is an ActiveX Automation examining operating data to identify equipment bottlenecks.
Server, formerly known as object linking and embedding Observing the importance of the reactor unit, this study
(OLE) automation, which enables VBA macro to interact with investigates the benefits of adding an extra CO2 stream to
the process flowsheet by sending inputs and receiving results. quench more CO2 gas along the reactor length. The
This feature of integrating Aspen Plus with the Microsoft optimization objectives are defined as follows:
application allows the Matlab optimization toolbox to
maximize f1 = FS20,MeOH (14)
manipulate operational variables and receive the response
from Aspen Plus through an Excel spreadsheet. An error may maximize f2 = (FS1,CO2 + FS31,CO2 + FS25,CO2) − FS14,CO2
happen in calling or accessing Aspen Plus object. Thus, it is
important to write error handling statements for the code to (15)
resume the process of communicating an automation inter- where streams (S20, S1, S31, S25, and S14) are shown in
face.27 Figure 4. The functions for optimization are the methanol
The structure of the end-user interface and the links between production rate and CO2 emission reduction. The reduction of
MS Excel, Aspen Plus, and Matlab are illustrated in Figure 3. In CO2 emission is defined as the net of CO2 consumed inside the
the first run, the input data can be provided manually including reactor, considering the extra amount of CO2 quenched into
the bound on each variable and the constraints. the reactor stages. For the optimization of methanol synthesis
Then, it passes to MS Excel to be implemented in the Aspen via multibed adiabatic reactor, the decisions were made based
Plus flowsheet. The outcome of this run provides new input on the operating variables of an existing plant and the
data for the next one. The model is kept running until the possibility of adding an extra quenching stream to inject
solver’s stopping criteria is achieved. The results are displayed more CO2 along the reactor. The methanol plant under

Figure 4. Aspen Plus flowsheet of methanol synthesis loop with additional interstage CO2 quenching.

533 DOI: 10.1021/ef502073b

Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

Figure 5. Temperature and conversion profiles versus dimensionless reactor length.

investigation has a production capacity of 2682 MTPD.28 The

decision variables with their corresponding bounds are
0 < FS25,CO2 < 5000 kmol/hr (16)

50 < TS25 < 200 °C (17)

0 ≤ X1 ≤ 1 split fraction to stream S26 (18)

0 ≤ X2 ≤ 1 split fraction to stream S27 (19)

0 ≤ X3 ≤ 1 split fraction to stream S27 (20)

4
∑ Xi = 1
i=1 (21)
These variables are subjected to one path constraint stating
that the temperature of the catalyst beds must be less than 543
K to avoid severe catalyst deactivation and low selectivity
Tmax < 543 K Figure 6. Mole percent of methanol versus temperature diagram in
(22)
Quench Converter.
The optimization runs were performed on a computer with
2.5 GHz dual-core Intel Core i5 and 16 GB memory to
facilitate accessing and storing large amounts of data and to
prevent “out of memory” error messages appearing due to the a
large number of runs of multiobjective genetic algorithm solver
“gamultiobj” in the Matlab optimization toolbox,29 which uses a
controlled elitist genetic algorithm (a variant of NSGA-II).30,31
Some important options must be set suitably in the genetic
algorithm (GA) solver. These include the number of variables,
fitness functions, constraints, and two elitism options
“ParetoFraction” and “DistanceFcn” which are used to limit the
number of individuals on the Pareto front (elite members) and
to maintain diversity by controlling the elite members of the
population as the algorithm progresses, respectively. Table 1
shows the parameters that were tuned experientially to improve
the smoothness and diversity of the Pareto optimal solutions.
Other options were defined as the default.

■ RESULTS AND DISCUSSION

A generic ICI’s methanol synthesis process flow diagram has Figure 7. Pareto optimal set obtained from dual maximization of
been developed, including an adiabatic quench type reactor methanol production and CO2 consumption.
with five catalytic beds separated by four interstage cold shot
distributes. These distributors are designed to give a good through the reactor. The simulation model was validated
mixing between injected cold shots and the hot gas flows against real industrial plant data that provides only the
534 DOI: 10.1021/ef502073b
Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

Figure 8. (a)−(f) Decision variables corresponding to Pareto optimal set in Figure 7; X1, X2, X3, and X4 are the split fractions of stream S25 to
streams S26, S27, S28, and S29, respectively.

temperature between beds and the compositions at the end of The optimization of methanol synthesis has been carried out
the reactor. Figures 5a and 5b show that both temperature and with the elitist nondominated sorting genetic algorithm
methanol conversion results are in good agreement with real (NSGA-II). The design and operating conditions of the
plant data. synthesis loop are taken from published industrial data and
The methanol concentration is reduced while the fresh used to validate the results of the model to ensure reliability
synthesis gas is injected through the distributors and and robustness. The additional CO2 injected to the system and
consequently more catalyst being required to achieve a high the splitting ratios are used as decision variables, whereas the
methanol yield per bed. In order to demonstrate this reactor design, including the injection location, and design of
characteristic, methanol concentration along the reactor length other units are fixed.
and the equilibrium concentration of methanol were plotted Figure 7 clearly shows that there is no solution that is the
versus the corresponding reaction temperature in Figure 6. The best with respect to methanol production and carbon dioxide
path of methanol concentration revolves around the equili- consumption. Instead, a set of solutions that are superior to the
brium-temperature trajectory to ensure the maximum methanol best of solutions in the search space, these solutions are known
yield is achieved in each stage of the reactor. as Pareto optimal solutions or nondominated solutions.
535 DOI: 10.1021/ef502073b
Energy Fuels 2015, 29, 530−537
Energy & Fuels Article

The Pareto optimal results reveal clearly the trade-off rRWGS Rate of water gas shift reaction, kmol hr−1 kg−1 cat−1
between the carbon dioxide consumed and the methanol G Superficial mass velocity(G = ρgμ), kg m−2 s−1
produced. If the operator desires to operate the reactor with L Total length of the reactor, m
conditions that minimize the CO2 emission (in other words, u Gas velocity, m s−1
consuming more CO2 inside the synthesis loop), then more Xx Dimensionless reactor length
wastewater will be produced and almost no extra methanol can T Gas phase temperature, K
be produced. However, the highest methanol generated by the
quench converter is ∼2778 MTPD when ∼5010 MTPD of Greek Letters and Symbols
carbon dioxide is consumed. This leads to an improvement of ΔHr,n Heat of reaction n, kJ kmol−1
the plant production by 3% as well as preventing 3430 MTPD dp Diameter of particle in the bed, m
of CO2 from being released to the atmosphere. Figure 8 (in a, vn,i Stoichiometric coefficient of component i in reaction n
b, c, d, and f) shows the values of the decision variables ηn Effectiveness factor
corresponding to the Pareto domain of Figure 7 plotted versus μg Viscosity of the gas mixture, unit
the objectives. These decision variables have a great influence ρb Bulk density of catalytic bed, kg m−3
on several factors that control the reactor performance; more ρg Molar density of gas, mol m−3
CO2 consumed in the production of methanol reactions Φ Porosity (void fraction)
increased the temperature and produced CO, which in turn was
consumed in the production of methanol. Abbreviations

■
EPA U. S. Environmental Protection Agency
CONCLUSIONS MTBE Methyl tert-butyl ether (fuel additive)
MTPD Metric tons per day
The multiobjective optimization problem of ICI’s low pressure
NRTL Nonrandom two liquid
methanol process operating conditions has been performed for
RWGS Reverse water gas shift reaction
efficient methanol production with reduced CO2 emission. To
SRK Soave−Redlich−Kwong
tackle this problem, the elitist nondominated sorting genetic

■
algorithm (NSGA-II) is adopted to determine the optimum
amount of captured CO2 that can be utilized, the favorable REFERENCES
temperature and the splitting factor of each quenching stream.
(1) Olah, G. A.; Surya, G. K. P. Efficient and selective chemical
The potential possibilities of improving the process were
recycling of carbon dioxide to methanol, dimethyl ether and derived
analyzed using a rigorous model integrated with a generic products. U. S. Patent 8,212,088, July 3, 2012.
process flowsheet to obtain necessary comparative results. In (2) Olah, G. A. Beyond Oil and Gas: The Methanol Economy.
contrast to the base case of industrial plant, the results have Angew. Chem., Int. Ed. 2005, 44, 2636−2639.
illustrated different operating conditions to operate the plant at (3) U. S. Environmental Protection Agency. http://www.epa.gov
different carbon dioxide consumption levels and production (accessed Dec 3, 2014).
rates. It is shown that the highest Methanol generated by the (4) Chapel, D. G.; Ernest, J.; Mariz, C. L.; Ernest, J. Recovery of CO2
quench converter is ∼2778 MTPD when ∼5010 MTPD of from Flue Gases: Commercial Trends; Fluor Daniel Inc.: Aliso Viejo,
carbon dioxide is consumed. This leads to an improvement of CA, 1999; originally presented at the Canadian Society of Chemical
the plant production by 3% as well as preventing 3430 MTPD Engineers 1999 annual meeting in Saskatoon, Saskatchewan, Canada.
(5) Olah, G. A.; Goeppert, A.; Prakash, G. K. S. Chemical Recycling
of CO2 from being released to the atmosphere. The main
of Carbon Dioxide to Methanol and Dimethyl Ether: From
advantage of applying multiobjective optimization is the ability Greenhouse Gas to Renewable, Environmentally Carbon Neutral
to choose one of the optimal solutions based on a good Fuels and Synthetic Hydrocarbons. J. Org. Chem. 2009, 74, 487−498.
knowledge of the process and the preferred condition.

■
(6) Natta, G. Synthesis of methanol. Catalysis 1955, 3, 349−411.
(7) Bakemeier, H.; Laurer, P. R.; Schroder, W. Development and
AUTHOR INFORMATION application of a mathematical model of the methanol synthesis. Chem.
Corresponding Author Eng. Prog. 1970, 66, 1−10.
(8) Cheng, W. H. Methanol production and use; CRC Press: New
*E-mail: aalarifi@uwaterloo.ca.
York, 1995; pp 73−98.
Notes (9) Al-Fadli, A. M.; Soliman, M. A.; Froment, G. F. Simulation of
The authors declare no competing financial interest. transients in a multi bed adiabatic methanol synthesis reactor; King

■ NOMENCLATURE
Parameters
Abdulaziz University: Jeddah, Saudi Arabia, 1995, 121−127; presented
at the Fourth Saudi Engineering conference in Riyadh, Saudi Arabia.
(10) Methanol plant design choices affect operations, costs, other
equipment, Oil Gas J. 1993, 91, 53−57.
Ac The cross-sectional area of the reactor, m2 (11) Beckett, P. C.; Evans, T. I.; Ryan, W. E.; Akgerman, A.
Cpg Heat capacity of gas mixture, kJ kmol−1 K−1 Designing a multistage adiabatic reactor for minimum catalyst volume.
dp The particle diameter, m Chem. Eng. Commun. 1989, 76, 107−124.
gc Conversion factor (equals one for the metric system) (12) Lange, J. P. Methanol synthesis: a short review of technology
Nrxn Number of reaction improvements. Cat. Today 2001, 64, 3−8.
rn Rate of reaction n, mol kg−1 s−1 (13) Makihara, H.; Niwa, K.; Nagai, H.; Morita, K.; Horizoe, H.;
Kobayashi, K.; Kuwada, C. Characteristics of a new methanol synthesis
Kad,j Adsorption equilibrium constant of component j reactor. Energy Prog. 1987, 7, 51−58.
Keq,i Equilibrium constant of reaction i (14) Matsumoto, H.; Nagai, H.; Watanabe, H.; Morita, K.; Makihara,
Pj Partial pressure of component j H. Advanced technology for large scale methanol plant; Technical Review,
ki Rate constant of reaction i Mitsubishi Heavy Industries: Nagasaki, Japan, 1997, 34, 58−62.
rCH3OH Rate of methanol synthesis reaction, kmol hr−1kg−1 (15) Kobayashi, K.; Osora, H.; Nagai, H.; Ohira, H. Apparatus for
cat−1 methanol synthesis. U. S. Patent 7,189,379, March 13, 2007.

536 DOI: 10.1021/ef502073b

Energy Fuels 2015, 29, 530−537
 Energy & Fuels Article

(16) Liu, P.; Gerogiorgis, D. I.; Pistikopoulos, E. N. Modeling and

optimization of polygeneration energy systems. Catal. Today 2007,
127, 347−359.
(17) Julián-Durán, L. M.; Ortiz-Espinoza, A. P.; El-Halwagi, M. M.;
Jiménez-Gutiérrez, A. Techno-economic assessment and environ-
mental impact of shale gas alternatives to methanol. ACS Sustainable
Chem. Eng. 2014, 2, 2338−2344.
(18) Herdem, M. S.; Farhad, S.; Dincer, I.; Hamdullahpur, F.
Thermodynamic modeling and assessment of a combined coal
gasification and alkaline water electrolysis system for hydrogen
production. Int. J. Hydrogen Energy 2014, 39, 3061−3071.
(19) Bussche, K. M.; Froment, G. F. A Steady-State Kinetic Model
for Methanol Synthesis and the Water Gas Shift Reaction on a
Commercial Cu/ZnO/Al2O3 Catalyst. J. Catal. 1996, 161, 1−10.
(20) Alarifi, A.; Elkamel, A.; Croiset, E. (2013). Steady-State
Simulation of a Novel Annular Multitubular Reactor for Enhanced
Methanol Production. Ind. Eng. Chem. Res. 2013, 52, 15387−15393.
(21) Froment, G. F.; Bischoff, K. B.; De Wilde, J. Chemical reactor
analysis and design; John Wiley & Sons: New York, 1979; 2, 44−79.
(22) Zahedi, G.; Elkamel, A. ; Lohi, A. Enhancing CO2 conversion to
methanol using dynamic optimization, applied on shell temperature
and inlet hydrogen during four years operation of methanol plant.
Energy Sources 2007, Part A 29, 1385−1396.
(23) Zahedi, G.; Elkamel, A.; Lohi, A. Dynamic optimization
strategies of a heterogeneous reactor for CO2 conversion to methanol.
Energy Fuels 2007, 21, 2977−2983.
(24) Methanol 07/08−2; Nexant, New York, 2008.
(25) Ilyukhin, N. A.; Mikhailov, A. Collaboration between JSC
Metafrax and Methanol Casale S.A. for Revamping Methanol Production
Plant; Casale Third Customer Symposium: Lugano, Switzerland, 2011.
(26) Løvik, I. Modelling, estimation and optimization of the
methanol synthesis with catalyst deactivation. Ph.D. Dissertation,
Norwegian University of Science and Technology, Trondheim,
Norway, 2001.
(27) Aspen Plus, version 10.2; User Guide Version 10.2; Aspen
Technology Inc., Cambridge, USA, 2000.
(28) Al-Fadli, A. M.; Soliman, M. A.; Froment, G. F. Steady State
Simulation Of a Multi-Bed Adiabatic Reactor for Methanol
Production. J. King Saud Univ. Eng. Sci. 1995, 7, 101−133.
(29) Matlab, version R2011b; Global Optimization Toolbox;
MathWorks Inc.: Natick, MA, 2011.
(30) Deb, K. Multi-objective optimization using evolutionary algorithms;
John Wiley & Sons: New York, 1979; 16, 101−150.
(31) Nandasana, A. D.; Ray, A. K.; Gupta, S. K. Applications of the
non-dominated sorting genetic algorithm (NSGA) in chemical
reaction engineering. Int. J. Chem. React. Eng. 2003, 1, No−pp.

537 DOI: 10.1021/ef502073b

Energy Fuels 2015, 29, 530−537

View publication stats