Journal of Natural Gas Science and Engineering 19 (2014) 221e227

Contents lists available at ScienceDirect

Journal of Natural Gas Science and Engineering

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

Selecting different raw materials for methanol production using an

MINLP model
c Kralj*
Anita Kova

Faculty of Chemistry and Chemical Engineering, University of Maribor, Smetanova 17, Maribor, Slovenia

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

Article history: This paper aims at selecting different raw materials, using an MINLP (mixed-integer nonlinear pro-
Received 12 March 2014 gramming) model, when considering the usage of more favourable raw materials for methanol
Received in revised form production.
7 May 2014
The best selection of raw material alternatives was sought for methanol production. Methanol is
Accepted 9 May 2014
produced from synthesis gas that is produced from different raw materials e natural gas or biogas. The
Available online 3 June 2014
basic starting point when comparing them is the same mass inlet flow rate for both raw materials.
Methanol production was simulated for both e natural gas or biogas as the raw material, using an Aspen
Keywords:
Methanol production
Plus simulator with a real chemical thermodynamic. Methanol production can be enlarged by simulta-
Natural gas neous structuring such as selecting the usage of more favourable raw materials, and parameter opti-
Biogas misation using the MINLP. The selection of raw methanol with optimal parameters has the greatest
Mathematical model impact on higher methanol production.
Mixed-integer nonlinear programming Optimal methanol conversion can take place during this operation, by applying optimal parametric
data within a reformer unit (temperature ¼ 840  C and pressure ¼ 8 bar), using natural gas. The optimal
production of methanol from natural gas was 17 510 kg/h under optimal parameters for 8.4% higher
production than under existing parameters.
© 2014 Elsevier B.V. All rights reserved.

1. Introduction in which the biodegradation of organic matter occurs in the

Methanol is an industrially important chemical, used mainly as a
feedstock during the production of bulk chemicals such as form-
aldehyde and acetic acid. It has also been used, for example, in the
manufacture of methyl tertiary-butyl ether, methyl methacrylate,
and as a solvent (McKetta and Cunningham, 1985). The first com-
mercial process for the production of methanol by the destructive
distillation of wood was already operational in the 19th century.
The first industrial plant producing methanol from synthesis gas
was built by BASF in 1923. This process used zinc oxide/chromium
oxide catalysts operating at 300  C and 200 atm. This process, also
known as high-pressure methanol, was carried out by synthesis.
Later a low-pressure process (p < 100 atm) was developed. Today,
this process is well-established in several companies, including ICI,
Lurgi and Mitsubishi (commercial technology solutions) (McKetta
and Cunningham, 1985).
Methanol can also be produced from biogas using
commercially-technical solutions. Anaerobic digestion is a process
* Tel.: þ386 02 2294 454; fax: þ386 02 2527 774.
E-mail address: anita.kovac@uni-mb.si.
absence of dissolved oxygen. It is a well-established and interna-
tionally applied technology for stabilising municipal sewage sludge,
treating organic wastes, products and wastewaters from industries,
households, and farms (DeBaere, 2000). The resulting methane gas
is a highly energetic biogas that is used in combined heat and po-
wer generators. The development of biogas technology took place
at the beginning of the 19th century. However, owing to the energy
crises of the 1970s, anaerobic digestion technology underwent
significant development (Alvarez, 2003; Strik et al., 2005).
Anaerobic digestion systems for the fermentation of organic
matters are widely used within commercial digesters of
70e5000 m3, small units are used mainly for heating, whilst large
units are used for electricity generation. Much of the technology is
based in Europe, with Germany and Denmark leading the field
(Nickolas et al., 2004). According to Nacke and co-authors, by the
end of 2005 there were more than 2000 biogas plants in Germany,
of different sizes (Nacke et al., 2005).
The time during which the mixture of wastewater stays in the
digester, with the microbial population, in order to produce the
biogas is called the ‘hydraulic retention time’ (HRT). This time is
very important when designing of biogas digesters (Jatinder and

http://dx.doi.org/10.1016/j.jngse.2014.05.008
1875-5100/© 2014 Elsevier B.V. All rights reserved.
222 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227

Sarbjit, 2004). Anaerobic bacteria (the methanogens), are sensitive
to those acid concentrations where the optimum pH value is found
to be within a range of 6.5e8.5 (Nickolas et al., 2004). In addition,
according to Nickolas, the concentration of ammonia increases at
the ends of the processes, so the pH level will be between 7.2 and
8.2.
Temperature is an important factor that determines the rate of
digestion. Most digesters are operated within the mesophilic range
(30e35  C), but it is possible to operate digesters within the ther-
mophilic range (approximately 55  C) but with higher operating
costs, lower process stability, and more structural requirements
(Rubia et al., 2002). The carbon to nitrogen (C/N) ratio for optimal
biogas production should be within the range 25e30. When a batch
of waste is received with high C/N ratio, this implies a complex
organic matter that is difficult to biodegrade; therefore, some
adjustment is required such as adding a high content of nitroge-
nous waste. On the other hand, low C/N ratio needs straw or crop
residues to be added, so as to adjust the carbon content (El-Hinnawi
and Biswas, 1981).
According to Koelsch and co-authors, the total solids content
range is about 8e13% and 80% of the solids are volatile solids
(Koelsch et al., 2001). One-half of the volatile solids (the biode-
gradable ones) are converted into methane and carbon dioxide.
Typical solid separation of the effluent will remove 4% of the solids
from the effluent. About one-third of the solids are converted to
gas, one-third can be separated out mechanically, and one-third
remains in the separated liquid effluent (Mattocks and Mark, 2000).
This paper presents the selecting of different raw materials for
methanol production (chapter 2) including the parameters' opti-
misation, by using an MINLP model (chapter 3), which includes the
real technical limitations and conditions, therefore providing a
solution near of the optimum.
2. Methanol production
The case study is based on a Lurgi methanol process (McKetta
and Cunningham, 1985) (Fig. 1). The methanol process is
composed of three subsystems:
 production of synthesis gas
 production of crude methanol and
 purification of methanol (F301, D301D304).
The possibility of better conversion was studied regarding syn-
thesis gas when retrofitting the Lurgi process for low-pressure
crude methanol production (without purification) from natural
gas or biogas.
Raw material (natural gas or biogas) is first desulphurised
(D101) and then heated up within a steam reformer (REA-1), where
synthesis gas is produced from raw material (natural gas or biogas)

and steam, at 825 C 15 bar:
3C2H6 þ 6.5H2O / 2CO þ 12H2 þ 1.75CH4 þ 2.25CO2 DrH298
¼ 196.17 kJ/mol (R1)

purge gas purge gas

fuel
flue gas
from F101
REA-1
E107
air
natural S1
to F101
gas E103 E104 E102
SEP
E109 D101 steam
turbine furnace HEA E312
E105 E101B E101A
EA202 HEW
F101
from D201 to E107 G202 from F201
to E301/2/4 from E107 VEN1
E110 G201-I G201-II HEPR to E103

F1 E111 F2 EA101 F107 E112 F108 F203 M2 REA-2

E201 to SH1
condensate
F102
water F301 D301
to F201
G301
D302
EA301
5 bar steam
E301 D303
E302
MeOH
E304
LEGEND: E305E303
reactor to E110

separator

heat exchanger

flash D304

water

Fig. 1. Process flow diagram of a low-pressure Lurgi methanol plant.

 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227 223

3C3H8 þ 10H2O / 3.5CO2 þ 17H2 þ 3CO þ 2.5CH4DrH298 Table 1

¼ 277.88 kJ/mol (R2) The composition of natural gas.

Component in n.g. Mass flow/(kg/h)

3C4H10 þ 13.5H20 / 4.75CO2 þ 22H2 þ 4CO þ 3.25CH4 DrH298 CH4 10 361
¼ 361.48 kJ/mol (R3) C2H6 86
C3H8 38
CH4 þ H2O $ CO þ 3H2DrH298 ¼ 206.08 kJ/mol (R4) nC4H10 8
iC4H10 7
CO2 25
CO þ H2O $ CO2 þ H2DrH298 ¼ 41.17 kJ/mol (R5)

The hot stream of the synthesis gas is cooled in an E107 boiler, in

E109, E110, E111 heat exchangers in an EA101 air cooler, and in an
E112 water cooler. The condensate is expanded in flashes: F1, F2,
F107 and F108. The synthesis gas is compressed in G201I and G201II
two-stage compressors.
In the second subsystem, methanol is produced by the catalytic
hydrogenation of carbon monoxide and/or carbon dioxide within a
REA-2 reactor using three main reactions (r ¼ R6, R7, R8):
CO þ 2H2 $ CH3OHDrH298 ¼  90.77 kJ/mol
CO2 þ 3H2 $ CH3OH þ H2ODrH298 ¼  49.58 kJ/mol
CO2 þ H2 $ CO þ H2ODrH298 ¼ 41.19 kJ/mol
The high-pressure reactor REA-2 is operated within the existing
parameters and un-converted gas is recycled. The inlet stream of
the reactor is heated by a process stream (HEPR). Furthermore, the
stream is cooled using air (HEA) and water (HEW) coolers before
entering the flash (SEP). The liquid stream of the separation is the
product and the recycled gas stream is compressed to 51 bar in a
new, two-stage compressor (COMP1, 2) with intermediate water
cooling (HEW1). The purge gas is separated from the crude meth-
anol in the flash F301. The purification includes the distillation
columns (D301D304).
Most parameters' effects on material balance were studied, by using
an Aspen Plus simulator to determine the material balance of the
synthesis gas, the crude methanol mass flow (mMeOH), and the
possible production of steam heat flow rate (Qsteam). The compo-
sition of synthesis gas has a strong effect on the composition of
crude methanol mass flow (mMeOH) and the possible production of
steam heat flow rate (Qsteam). The more important is the conversion
of methane within the reformer. Optimal methane conversion
could take place by an operation using the optimal parametric data
(R6)
in a reformer unit. The best methane conversion is under lower
pressure and higher temperature. The reaction of synthesis gas was
(R7)
created by using an equilibrium reactor model (RGIBBS).
The pressure and temperature effects on synthesis gas conver-
(R8)
sions from natural gas (as raw material ¼ rm ¼ N) were determined
with changing parameters and variables by using an Aspen Plus
simulator and were modelled using equations (1e14), and included
within the MINLP model (Chapter 3). The some possible changes of
temperature and pressure were tested within the real process,
therefore equations presented the real relationships between pro-
cess parameters and variables.
Equation (1) presents the dependence on the composition of
methane in the synthesis gas, as a function of pressure (p) in the
reformer:
mN 2
CH4 ;p ¼ 0:242p þ 143:64p  361:09 (1)

2.1. Methanol production from natural gas Equation (2) presents the dependence on the composition of
carbon dioxide in the synthesis gas, as a function of pressure (p) in
Methanol production using natural gas or biogas as the raw the reformer:
material was simulated using an Aspen Plus simulator (with real
chemical thermodynamic and selected processing units, which are mN 2
CO2 ;p ¼ 0:951p þ 10:914p þ 8965:3 (2)
very comparable to existing material and energy flow rates) (Aspen
Equation (3) presents the dependence on the composition of
Technology, 2002). The thermodynamic properties of gases can be
water in the synthesis gas, as a function of pressure (p) in the
calculated using different methods and models. These models can
reformer:
estimate fugacity coefficients, enthalpies, densities, entropies, and
free energies. The Grayson-Streed model of corresponding states
mN 2
H2 O;p ¼ 0:121p þ 156:72p þ 17445 (3)
correlation is used for synthesis gas. Fugacity coefficients, entropies
and free energies can be estimated using this model. The Grayson- Equation (4) presents the dependence on the composition of
Streed model is a generalized method for predicting vapoureliquid carbon monoxide in the synthesis gas, as a function of pressure (p)
equilibria in gaseous hydrocarbon and hydrogen mixtures. This in the reformer:
method is needed for simulation. The thermodynamic model of
component properties is very important for approaching a good mN 2
CO;p ¼ 1:020p  257:73p þ 13037 (4)
simulation. The thermodynamic properties of synthesis gas were
estimated by using the Grayson-Streed model with an Aspen Plus Table 2
simulator, and the results were in relatively good agreement with The composition of synthesis gas from natural gas and other pa-
the experimental data within an existing methanol production rameters under the existing condition T ¼ 825  C and p ¼ 15 bar.
plant (Kova
c Kralj et al., 2000). Methanol production using natural Component in s.g. Mass flow/(kg/h)
gas (Table 1) as the raw material was simulated using an Aspen
CH4 1737
Plus simulator, and thus the existing crude methanol CO2 8915
ðmNMeOH;ex ¼ 16 146 kg=hÞ, and steam production CO 9404
N
H2 3648
ðQsteam;ex ¼ 13 309 kWÞ could be produced (Table 2) under stan- H2O 19 821
dard conditions. Others parameters
The composition of synthesis gas has a strong effect on the mMeOH (kg/h) 16 146
Qsteam (kW) 13 309
composition of crude methanol mass flow and steam production.
224 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227

Equation (5) presents the dependence on the composition of B

(Qsteam;ex ¼12 646 kW) could be produced (Table 3) under standard
hydrogen in the synthesis gas, as a function of pressure (p) in the conditions.
reformer: The pressure and temperature effects on synthesis gas conver-
sions from biogas were determined by using an Aspen Plus simu-
mN 2
H2 ;p ¼ 0:054p  53:767p þ 4441:6 (5) lator and were modelled using equations (15e28), and included
within the MINLP model (Chapter 3).
Equation (6) presents the dependence on the composition of
Equation (15) presents the dependence on the composition of
methane in the synthesis gas, as a function of temperature (T) in the
methane in the synthesis gas, as a function of pressure (p) in the
reformer:
reformer:
mN 2
CH4 ;T ¼ 0:071T  137:11T þ 66778 (6)
mBCH4 ;p ¼ 1:145p2 þ 54:2p  150 (15)
Equation (7) presents the dependence on the composition of
carbon dioxide in the synthesis, gas as a function of temperature (T) Equation (16) presents the dependence on the composition of
in the reformer: carbon dioxide in the synthesis gas, as a function of pressure (p) in
the reformer:
mN
CO2 ;T ¼ 13:777T þ 20272 (7)
mBCO2 ;p ¼ 0:626p2  5:521p þ 9713:3 (16)
Equation (8) presents the dependence on the composition of
water in the synthesis gas, as a function of temperature (T) in the Equation (17) presents the dependence on the composition of
reformer: water in the synthesis gas, as a function of pressure (p) in the
reformer:
mN 2
H2 O;T ¼ 0:076T  143:33T þ 86001 (8)
mBH2 O;p ¼ 1:569p2 þ 62:551p þ 21131 (17)
Equation (9) presents the dependence on the composition of
carbon monoxide in the synthesis gas, as a function of temperature Equation (18) presents the dependence on the composition of
(T) in the reformer: carbon monoxide in the synthesis gas, as a function of pressure (p)
in the reformer:
mN
CO;T ¼ 29:101T  14368 (9)
mBCO;p ¼ 1:614p2  90:916p þ 9403 (18)
Equation (10) presents the dependence on the composition of
hydrogen in the synthesis gas, as a function of temperature (T) in Equation (19) presents the dependence on the composition of
the reformer: hydrogen in the synthesis gas, as a function of pressure (p) in the
reformer:
mN 2
H2 ;T ¼ 0:026T þ 50:558T  20131 (10)
mBH2 ;p ¼ 0:477p2  20:548p þ 3576:3 (19)
Equation (11) presents the dependence on the composition of
crude methanol in the methanol reactor (REA-2), as a function of Equation (20) presents the dependence on the composition of
pressure (p): methane in the synthesis gas, as a function of temperature (T) in the
reformer:
mN 2
MeOH;p ¼ 0:32p  299:67p þ 20565 (11)
mBCH4 ;T ¼ 0:047T 2  89:742T þ 42990 (20)
Equation (12) presents the dependence on the composition of
crude methanol in the methanol reactor (REA-2), as a function of Equation (21) presents the dependence on the composition of
temperature (T): carbon dioxide in the synthesis gas, as a function of temperature (T)
in the reformer:
mN 2
MeOH;T ¼ 0:152T þ 295:4T  124050 (12)
mBCO2 ;T ¼ 12:61T þ 19894 (21)
Equation (13) presents the dependence on the production of
steam flow rate during the methanol process, as a function of Equation (22) presents the dependence on the composition of
pressure (p): water in the synthesis gas, as a function of temperature (T) in the
reformer:
N
Qsteam;p ¼ 0:007p2  11:306p þ 13480 (13)
mBH2 O;T ¼ 0:052T 2  94:338T þ 64845 (22)
Equation (14) presents the dependence on the production of
steam flow rate, as a function of temperature (T):

N
Qsteam;T ¼ 38:227T  18223 (14) Table 3
The composition of synthesis gas from biogas and other parameters
under the existing condition T ¼ 825  C and p ¼ 15 bar.

2.2. Methanol production from biogas Component in s.g. Mass flow/(kg/h)

CH4 919
Methanol could be produced by biogas that contains 75% CO2 9491
CO 7678
methane (7895 kg/h), 23% carbon dioxide (2420 kg/h), and 2% of
H2 3163
hydrogen (210 kg/h) (Kova
c Kralj and Kralj, 2010). Methanol pro- H2O 22 421
duction using biogas (as raw material ¼ rm ¼ B) was simulated Others parameters
using an Aspen Plus simulator and the existing crude methanol mMeOH (kg/h) 14 615
Qsteam (kW) 12 646
ðmBMeOH;ex ¼ 14 615 kg=hÞ; and thus steam production
 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227
Equation (23) presents the dependence on the composition of
carbon monoxide in the synthesis gas, as a function of temperature
(T) in the reformer:
mBCO;T ¼ 0:083T 2 þ 166:49T  73188
Equation (24) presents the dependence on the composition of
hydrogen in the synthesis gas, as a function of temperature (T) in
the reformer:
mBH2 ;T ¼ 0:017T 2 þ 32:943T  12084
Equation (25) presents the dependence on the composition of
crude methanol in the methanol reactor (REA-2), as a function of
pressure (p):
mBMeOH;p ¼ 2:739p2  110:73p þ 16890
Equation (26) presents the dependence on the composition of
crude methanol in the methanol reactor (REA-2), as a function of
temperature (T):
mBMeOH;T ¼ 0:004T 2 þ 197:11T  77982
Equation (27) presents the dependence on the production of
steam flow rate during the methanol process, as a function of
pressure (p):
B
Qsteam;p ¼ 0:137p2  3:908p þ 12735
Equation (28) presents the dependence on the production of
steam flow rate, as a function of temperature (T):
B
Qsteam;T ¼ 35:882T  16956
3. Mathematical model of methanol production from
different raw materials
Retrofitting an existing plant focuses on the general mathe-
matical MINLP method including many result assumptions by the
Aspen Plus simulator.
Retrofitted methanol production from different raw materials
by optimal condition can be increased by synthesis gas conversion
and, therefore, the crude methanol conversion. Methanol produc-
tion from different raw materials, including process parameters, are
optimised using an MINLP (mixed-integer nonlinear programming)
model (Biegler et al., 1997).
The MINLP approach deals with both continuous and discrete
variables, simultaneously. Whilst continuous variables are defined
for the continuous optimisation of parameters (mass flow, heat
flow rate, temperatures, pressures, costs, etc.), discrete 0e1 vari-
ables are used to express discrete decisions, i.e. usually the exis-
tence (1) or non-existence (0) of structural elements within the
defined structure. Different raw materials may also be defined as
discrete alternatives. As the discrete optimisations are carried out
simultaneously, together with the continuous optimisation, the
MINLP approach additionally determines the optimal continuous
parameters.
The MINLP model contains continuous variables of the process
parameters: molar heat capacities, material flow rates, heat flow
rates, pressures and temperatures, which are limited by real con-
straints. The MINLP model contains equations that enable meth-
anol and steam productions. The most important is the conversion
of methane in the reformer. Optimal methane conversion could
take place by an operation using the optimal parametric data in a
225
reformer unit. Mathematical problems could include those equa-
tions that present synthesis gas composition (equations 110),
crude methanol production (equations 11 and 12), and steam
production (equations 1314) from natural gas. Mathematical
(23) problems could include those equations which also present syn-
thesis gas composition (equations 1524), crude methanol pro-
duction (equations 2526), and steam production (equations
2728) from biogas.
For the MINLP algorithm, a simplified model for the use of
different raw materials (e.g. rm ¼ B, Z ¼ RM) between natural gas
(24) (N) and biogas (B), which are denoted by discrete (or binary)
variables.
Many constrained engineering and industrial optimisation
problems can be modelled as mixed integer nonlinear program-
ming (MINLP) problems (Biegler et al., 1997).
(25) The handling of binary (yt ¼ 0,1) variables allows for the spec-
ification of those constraints that are relevant for synthesising a
practical flow-sheet structure, in our case, the selection of different
raw materials for methanol production.
If we want to select only one type of raw material (our case),
equation (29) can be used:
(26) RM
X
yrm ¼ 1 (29)
When selecting at least one type of raw material, equation (30)
can be used:
(27) RM
X
yrm  1 (30)
When selecting, at most, one type of raw material, equation (31)
(28) can be used:
RM
X
yrm  1 (31)
In addition the binary variables can be related to activating or
deactivating continuous variables, inequalities or equations: As an
example, consider the conditions for the continuous variable e x, in
our case, the mass flow of methanol production and the heat flow
rate of steam production:
if yrm ¼ 1/L  x  U; if yrm ¼ 0/Lx ¼ 0;
which can be modelled through the constraint: L yrm  x  U yrm
Where L is the lowest value (in our case LP ¼ 0 and LF ¼ 0) and U
is the highest value of parameters (in our case UP ¼ mMeOH and
UF ¼ Fsteam) for both raw materials - natural gas (N) and biogas (B).
The objective function (eq. (32)) of the MINLP model maximised
the additional annual profit of the retrofit and included income and
depreciation (Table 4). Incomes include the additional methanol
(mN N N N
MeOH  mMeOH;ex ) and steam ðQsteam  Qsteam;ex Þ productions
from natural gas depending on pressure (p) and temperature (T)
functions. Incomes include the additional methanol
(mBMeOH  mBMeOH;ex ) and steam ðQsteam B B
 Qsteam;ex Þ productions
from biogas, also depending on pressure (p) and temperature (T)
Table 4
Cost items for the example process.
Price of methanol, CMeOH/(EUR/t): 115
Price of HPsteam, Csteam/(EUR/kWa): 106
Cost of natural gas, CN/(EUR/t): 60
Cost of biogas, CB/(EUR/t): 60
Cost of additional heating, Cheating/(EUR/K a): 3000
Operating hours, Otime/(h/a): 8000
226 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227

1200 Methanol production from natural gas could be increased by 8.4%

using processed operational and parametric optimisation using
MINLP. The profit was 0.741 MEUR/a depending on the optimised
1000
conditions. The resulting MINLP problem for methanol production
from different raw materials can be solved by using a
800 commercially-available optimisation tool, GAMS (Brooke et al.,
Profit, kEUR/a

600
400
200
0
0 20 40 60
Cost of raw material, EUR/t
Fig. 2. The profit depending on the cost of raw material.
functions. Temperature and pressure could have an affect on
additional methanol and steam productions, therefore it could be
divided into two objective functions. The existing methanol mass
flow from natural gas (mN MeOH; ex ) is 16 146 kg/h under existing
unchanged process conditions. The existing steam heat flow from
N
natural gas (Qsteam; ex ) is 13 309 kW under existing unchanged
process conditions. The existing methanol mass flow from biogas
(mBMeOH; ex ) is 14 615 kg/h under existing unchanged process con-
ditions. The existing steam heat flow from biogas (Qsteam; B
ex ) is
12 646 kW under existing unchanged process conditions. Depre-
ciation includes the cost of additional heating within the reformer if
the temperature is higher than the existing temperature
(Tex ¼ 825  C). CN and CB are the costs of natural gas and biogas by
including the same mass flow of raw material (mN ¼ mB ¼ 10.5 t/h)
and operating time (Otime ¼ 8000 h/a). The retrofitted modification
cost can only account for one-tenth (1/10) of the total raw material
cost (Table 4). The existing income from methanol production using
natural gas or biogas is included as additional annual profit.
Maximal additional annual profit (Vmax) for retrofit:
1992).
Nature gas The methanol production from natural gas is more effective
than from biogas but only the low cost of biogas can be replaced by
Biogas
natural gas. Biogas can be more favourable, only if the price of
biogas is 40% lower than the natural gas. The cost of the raw ma-
terials (natural gas and biogas) has the most important effect on the
selecting of raw material (Fig. 2).
The research was based on developing mathematical model that
could present the real process situation, including the real testing of
parameters within the existing process.
80
4. Conclusions
This paper aimed at selecting different raw materials using an
MINLP (mixed-integer nonlinear programming) model, whilst
considering the usage of more favourable raw materials for meth-
anol production.
Methanol is produced from synthesis gas that is produced from
raw materials e natural gas or biogas. The MINLP model contains
equations that enable optimal methanol and steam productions by
using simultaneous parametric optimisation and different raw
materials. The more important is the conversion of methane within
the reformer. Optimal methane conversion could take place by an
operation using the optimal parametric data in a reformer unit
using different raw materials. Mathematical problems could
include equations that represent synthesis gas composition, and
crude methanol and steam productions. The primary objective of
retrofit is the uses of different raw materials with a minimum of
additional costs, and thus maximising the production of methanol
and steam. The optimal raw material was natural gas. Optimal
production of methanol from natural gas was 8.4% higher
(17 510 kg/h) under optimal parameters (temperature ¼ 840  C and
pressure ¼ 8 bar) than under existing parameters within the
reformer. Optimal steam production was 13 635 kW. Biogas can be

     
mN N
MeOH;p  mMeoH;ex N mN N
MeOH;T  mMeOH;ex N
N
Qsteam;p N
 Qsteam;ex N
Vmax ¼ CMeOH Otime þ CMeOH Otime þ Csteam
2 2 2
     
N
Qsteam;T N
 Qsteam;ex N mBMeOH;p  mBMeoH;ex B mBMeOH;T  mBMeOH;ex B
þ Csteam þ CMeOH Otime þ CMeOH Otime
2 2 2
   
B
Qsteam;p B
 Qsteam;ex B B
Qsteam;T B
 Qsteam;ex B
þ Csteam þ Csteam  ðT  Tex ÞCheatin  CN mN Otime N=10  CB mB Otime B=10
2 2
(32)

The primary objective of retrofit is to select the optimal raw more favourable under the same optimal condition, but only if the
material (rm ¼ N) with a minimum of additional costs, and to price of biogas is 40% lower than natural gas.
maximise the production of methanol and steam. The optimal raw
material was the natural gas. The optimal production of methanol
was 17 510 kg/h under optimal parameters (temperature ¼ 840  C References
and pressure ¼ 8 bar) in the reformer. Optimal steam production
was 13 635 kW. The additional cost included additional heating of Alvarez, J.M., 2003. Biomethanization of the Organic Fraction of Municipal Solid
Wastes, first ed. IWA Publishing.
the reformer and one-tenth of the cost of the raw material with Aspen Technology, 2002. ASPEN PLUS User Manual Release 11.1. Aspen Technology
0.55 MEUR/a. The total annual income was 1.291 MEUR/a. Inc., Cambridge, MA 02139, USA.
 A.K. Kralj / Journal of Natural Gas Science and Engineering 19 (2014) 221e227 227

Biegler, L.T., Grossmann, I.E., Westerberg, A.W., 1997. Systematic Methods of Chemical Nickolas, J., Themelis, S., Verma, S., 2004. Anaerobic digestion of organic waste in
Process Design. Prentice Hall, Upper Saddle River, New Jersey, pp. 1e408. MSW. Waste Manag. World, JanuaryeFebruary, 41e47.
Brooke, A., Kendrick, D., Meeraus, A., 1992. GAMS: a User's Guide. Scientific Press, Rubia, M.A., Perez, M., Romero, L.I., Sales, D., 2002. Anaerobic mesophilic and
Palo Alto. thermophilic municipal sludge digestion. Chem. Biochem Eng. 16, 119e124.
DeBaere, L., 2000. State-of-the- art of anaerobic digestion of solid waste in Europe. Strik, D., Domnanovich, A., Zani, L., Braun, R., Holubar, P., 2005. Prediction of trace
Water Sci. Technol. 41, 283e290. compounds in biogas from anaerobic digestion using the MATLAB Neural
El-Hinnawi, E., Biswas, A.K., 1981. Renewable Sources of Energy and the Environ- Network Toolbox 20. Environ. Model. Softw., 803e810.
ment, Natural Resources and the Environmental Series, first ed., vol. 6
Jatinder, S.K., Sarbjit, S.S., 2004. Comparative study of economics of different models
of family size biogas plants for state of Punjab, India. Energy Convers. Manage
Nomenclatures
45, 1329e1341.
Abbreviations:
Koelsch, R.K., Fabian, E.E., Guest, R.W., Campbell, J.K., 2001. Anaerobic Digesters for
Dairy Farms, Agricultural and Biological Engineering Extension Bulletin 458.
Cornell University, Ithaca, NY, 14853. HP: high pressure
Kova
c Kralj, A., Kralj, D., 2010. Methanol production from biogas. Int. J. Math. MINLP: nonlinear programming
Comput. Simul. 4 (2). N: natural gas
Kova
c Kralj, A., Glavi
c, P., Kravanja, Z., 2000. Retrofit of complex and energy B: biogas
intensive processes II: stepwise simultaneous superstructural approach. Com-
put.. Chem. Eng. 24 (1), 125e138. Variables:
Mattocks, R.P., Mark, A.M., 2000. Fate of Incoming Solids to Measure Manure
Digester Performance, Animal, Agricultural and Food Processing Wastes Pro-
ceedings ASAE. 2950 Niles Rd., St. Joseph, MI 49085-9659, pp. 187e193. m: mass flow, kg/h
McKetta, J.J., Cunningham, W.A., 1985. Encyclopedia of Chemical Processing and p: pressure, bar
Design, vol. 29. Marcel Dekker, New York, pp. 418e474. T: temperature, K
Nacke, T., Brückner, K., Go € ller, A., Kaufhold, S., Nakos, X., Noack, S., 2005. New type Q: heat flow rate, W
of dry substances content meter using microwaves for application in biogas
plants. Anal. Bioanal. Chem. 383, 752e757.