Developing your own
distributed model
gPROMS Process Training
Developing your own distributed model
Developing your own distributed model
Methanol synthesis
Background
Cooled multi-tubular reactors are well-established and widely used in chemical industry. These reactors
are cooled by a counter-current stream on the shell side. The tubes of the packed-bed reactor are filled
with spherical catalyst particles while those of the monolithic reactor contain several pieces of cylindrical
monoliths with square channels as depicted below. Monoliths are either made of inert material that is
coated with the catalytic phase, or full extruded from the catalytic material.
Objectives
The aim of this hands-on session is to simulate the operation of a catalytic tubular reactor system in
which a gas phase reaction takes place.
Exercise 1
Reducing CPU time without compromising on accuracy for a
2D gas-phase PFR
Objectives
The aim of this hands-on session is to:
• Complete the 2D model, taking into account variations along the radial direction of the reactor
• Reduce the CPU time of the simulation without compromising on accuracy
Things to do
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 2
Developing your own distributed model
Constructing the model
1. Open the file ‘Starter 203 Exercise 1.gPJ’ provided. This model represents the behaviour of a gas-
phase PFR model for the synthesis of methanol
2. Save the project file under a different name.
3. Open the model gas_phase_PFR and observe the following list of PARAMETERs
Symbol Description Units gPROMS identifier Size Type
NC Components - components - ORDERED_SET
Molecular weight of component
MWi kg/kmol molecular_weight components REAL
i
𝛎𝐢 Stoichiometry of component i - reaction_stoichiometry components INTEGER
∆𝑯 Enthalpy of reaction kJ/kmol reaction_enthalpy - REAL
Pre-exponential factor in
𝐤 𝐫,𝟎 kmol/hr/kgcat pre_constant_r - REAL
𝑘𝑟 reaction kinetic expression
Activation energy for
𝐄𝐚𝐫 kJ/kmol activation_energy_r - REAL
𝑘𝑟 reaction
Pre-exponential factor in
𝐤 𝐢,𝟎 reaction kinetic expression for kPa-1 pre_constant components REAL
each component i
Activation energy for reaction of
𝐄𝐚𝐢 kJ/kmol activation_energy components REAL
each component i
Equilibrium constants for 𝐾𝑒𝑞
𝐤 𝟏𝐞𝐪 , 𝐤 𝟐𝐞𝐪 - equilibrium_constants 2 REAL
kinetic expression
V Volume of reactor m3 volume - REAL
L Length of reactor m reactor_length - REAL
R Radius of reactor m reactor_radius - REAL
𝛒𝐛 Density of catalytic bed kg/m3 bulk_density - REAL
3
𝛒𝐟 Density of stream kg/m fluid_density - REAL
𝐂𝐩,𝐟 Specific heat of stream J/kg/K specific_heat_capacity - REAL
𝐃𝐳 Diffusion with flow m2/s axial_diffusion - REAL
𝐤𝐳 Thermal conductivity with flow W/m/K axial_thermal_conductivity - REAL
2
𝐃𝐫 Diffusion normal to flow m /s radial_diffusion - REAL
Thermal conductivity normal to
𝐤𝐜 W/m/K radial_thermal_conductivity - REAL
flow
𝛆 Void fraction - void - REAL
𝐡𝐰 Local heat transfer coefficient W/m2/K local_heat_transfer_coeff - REAL
3
𝛒𝐜 Density of coolant kg/m coolant_density - REAL
𝐂𝐩,𝐜 Specific heat of coolant J/kg/K coolant_heat_capacity - REAL
𝐕𝐜 Volume of jacket m3 jacket_volume - REAL
2
U Overall heat transfer coefficient W/m /K overall_heat_transfer_coeff - REAL
2
S Surface area of heat transfer m heat_transfer_area - REAL
and notice the following list of VARIABLEs
Symbol Description Units gPROMS identifier Size Variable type
components, molar_concentration_
𝐂𝐢 Molar concentration kmol/m3 molar_concentration
axial, radial kmol_per_m3_gML
components molar_concentration_
𝐂𝐢,𝐢𝐧 Inlet molar concentration kmol/m3 inlet_molar_concentration
kmol_per_m3_gML
Inlet pressure of pressure_partial_kPa_
𝐩𝐢,𝐢𝐧 kPa inlet_partial_pressure components
component i gML
𝐓𝐢𝐧 Inlet stream temperature K inlet_temperature - temperature_gML
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 3
Developing your own distributed model
𝐓𝐰 Wall temperature K wall_temperature - temperature_gML
T Reactor temperature K temperature axial, radial temperature_gML
kmol/hr/ reaction_rate_kmol_per_
Rrxn Molar reaction rate molar_reaction_rate axial, radial
kgcat s_kg_gML
keq Equilibrium constant - k_eq axial, radial no_type_gML
Kinetic parameters for -1 components,
ki kPa k no_type_gML
component i axial, radial
Kinetic parameter for main kmol/hr/
kr kr axial, radial no_type_gML
reaction kgcat
Fugacity coefficients for components,
fi - fugacity_coeff no_type_gML
component i axial, radial
u Velocity of stream m/s velocity - velocity_gML
components, pressure_partial_kPa_
pi Pressure of component i kPa partial_pressure
axial, radial gML
components,
wi Mass fraction of component i - mass_fraction mass_fraction_gML
axial, radial
components,
xi Molar fraction of component i - molar_fraction molar_fraction_gML
axial, radial
Q Energy rate W/m2 energy_flux - energy_flux_gML
FC mass flowrate for coolant kg/s coolant_mass_flowrate - mass_flowrate_gML
Tc,in Inlet coolant temperature K coolant_in_temperature - temperature_gML
Tc Coolant temperature K coolant_temperature - temperature_gML
and notice how spatial discretisation is incorporated into the model through the
DISTRIBUTION_DOMAIN, VARIABLE, BOUNDARY and EQUATION sections. For example,
notice how the ideal gas relation is written in the gPROMS language, as well as how the distributed
variables involved are declared:
VARIABLE
molar_concentration AS DISTRIBUTION(components,axial,radial)
OF molar_concentration_kmol_per_m3_gML
temperature AS DISTRIBUTION(axial,radial) OF temperature_gML
partial_pressure AS DISTRIBUTION(components,axial,radial) OF no_type_gML
EQUATION
(…)
# Ideal Gas Relation
FOR i IN components DO
inlet_partial_pressure(i) =
inlet_molar_concentration(i)*ideal_gas_constant*inlet_temperature ;
FOR z := 0 TO reactor_length DO
FOR r := 0 TO reactor_radius DO
partial_pressure(i,z,r) =
molar_concentration(i,z,r)*ideal_gas_constant*temperature(z,r) ;
END
END
END
4. Using the table of PARAMETERs and VARIABLEs above, complete the EQUATION section for:
a. The molar balance as follows:
𝜕𝐶𝑖 (𝑧, 𝑟) 𝜕𝐶𝑖 (𝑧, 𝑟) 𝜕 2 𝐶𝑖 (𝑧, 𝑟) 𝜕 2 𝐶𝑖 (𝑧, 𝑟) 1 𝜕𝐶𝑖 (𝑧, 𝑟)
= −𝑢 + 𝜀𝐷𝑧 + 𝜀𝐷𝑟 +𝜀𝐷𝑟 + 𝜌𝑏 𝜈𝑖 𝑅𝑟𝑥𝑛 (𝑧, 𝑟)
𝜕𝑡 𝜕𝑧 𝜕𝑧 2 𝜕𝑟 2 𝑟 𝜕𝑟
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 4
Developing your own distributed model
EQUATION
# Mass balance
FOR i IN components DO
FOR z := 0|+ TO reactor_length|- DO
FOR r := 0|+ TO reactor_radius|- DO
$molar_concentration(i,z,r) = # COMPLETE
END
END
END
b. The energy balance as follows:
𝜕𝑇(𝑧, 𝑟) 𝜕𝑇(𝑧, 𝑟) 𝜕 2 𝑇(𝑧, 𝑟) 𝜕 2 𝑇(𝑧, 𝑟)
𝜌𝑓 𝐶𝑝,𝑓 = −𝜌𝑓 𝐶𝑝,𝑓 𝑢 + 𝑘𝑧 + 𝑘 𝑐
𝜕𝑡 𝜕𝑧 𝜕𝑧 2 𝜕𝑟 2
1 𝜕𝑇(𝑧, 𝑟)
+ 𝑘𝑐 + 𝜌𝑏 (−∆𝐻)𝑅𝑟𝑥𝑛 (𝑧, 𝑟)
𝑟 𝜕𝑟
EQUATION
# Energy balance
FOR z := 0|+ TO reactor_length|- DO
FOR r := 0|+ TO reactor_radius|- DO
fluid_density*specific_heat_capacity*$temperature(z,r) =
# COMPLETE
END
END
Simulating the model
5. Open the process Simulate_gas_phase_PFR.
6. Complete the SET section to use [BFDM,1,80] and [OCFEM,3,5] to discretise the axial and radial
domain, respectively.
7. Complete the INITIAL section according to the table below. Keep in mind that the domain of the
initial conditions must match domain of the differential variables in the EQUATION section.
Variable gPROMS identifier Value Units
Molar concentration of carbon monoxide molar_concentration('CARBON MONOXIDE',z,r) 0.291 kmol/m3
Molar concentration of hydrogen molar_concentration(‘HYDROGEN’,z,r) 0.037 kmol/m3
Molar concentration of methanol molar_concentration(‘METHANOL',z,r) 0.009 kmol/m3
Reactor temperature temperature(z,r) 473.15 K
Coolant temperature coolant_temperature 473.15 K
8. Run the simulation.
9. Plot the reactor temperature and the molar concentration of the components at the end of the
simulation along the reactor length, in the centre of the reactor (r=0).
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 5
Developing your own distributed model
10. Observe and take note of the CPU time.
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 6
Developing your own distributed model
11. Try different discretisation methods, orders and number of elements for both domains and observe
the results and CPU time.
Reducing CPU time
12. Copy and paste the process Simulate_gas_phase_PFR and name it
Simulate_gas_phase_PFR _NUG.
13. Go to the discretisation method definitions in the SET section and replace the axial description from
[BFDM, 1, 80] to [BFDM, 1, 80, TRANSFORM(LOG,4)].
14. Simulate the process and compare the temperature plot with the one from the previous simulation.
How does the plot change? Compare the CPU time with the previous ones.
_________________________________________________________________________________
_________________________________________________________________________________
15. Reduce the number of elements for the axial distribution from 80 to 30 and simulate the process.
16. Compare the temperature plot with the two previous ones. How does the plot change?
_________________________________________________________________________________
_________________________________________________________________________________
17. Increase alpha from 4 to 10 and simulate the process.
18. Compare the temperature plot with the ones from the previous simulations. How does it change?
_________________________________________________________________________________
_________________________________________________________________________________
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 7
Developing your own distributed model
COMMERCIAL IN CONFIDENCE
© Siemens Industry Software Ltd 8
