Main objective of this project is to build an open source tool for modeling Solid Oxide Fuel Cells (SOFC). I wanted to upload the MVP (minimum liable product) for the part zero-dimensional models with hydrogen fuel while I work on other modules.
Code contains lots of comments and explanations. My purpose is to provide a working example for other researchers to build/modify their own code on similar topics.
A zero-dimensional (lumped parameters) model of a Solid Oxide Fuel Cell can be simulated by writing a script, or using graphical user interface.
Just write a python script with following lines and run. GUI window will show up.
from pysofc import startGUI
startGUI()
Watch the video of an example usage case of graphical user interface. It provides much more information than pages long documentation.
sofc.mp4
Second option is to write Python script.
from pysofc import HydrogenSOFC # Physical representation of an SOFC
from pysofc import LumpedModelHydrogenSOFC # Zero-Dimensional (Lumped parameters) simulation
cell_file = r"Parameter Files\cell_parameters_jung2005.json" # Journal of Power Sources 155 (2006) 145–151
fuel_cell = HydrogenSOFC("SOFC-1", load_from_json= cell_file)
model = LumpedModelHydrogenSOFC(fuel_cell)
model.input["X_fuel_in"] = {"H2": 0.97, "H2O": 0.03} # Define fuel gas inlet molar composition
model.input["X_oxygen_in"] = {"O2": 0.21, "N2": 0.79} # Define air/sweep gas inlet molar composition
model.input["molar_flowrate_fuel"] = 0.0001859 # Fuel gas feed rate [mol/s]
model.input["molar_flowrate_oxygen"] = 0.000495733 # Air/sweep gas feed rate [mol/s]
model.input["T_operation"] = 923 # Operating temperature [K]
model.input["pressure"] = 1 # Operating pressure [bar]For most of the cases these operational inputs would be sufficient. If you want to implement a modified version, please see the all available operation parameters of class LumpedModelHydrogenSOFC in source file pysofc/models_zero_dim.
To generate a voltage-current (V-I) curve at given operation conditions you should supply a list of current densities. If related experimental data are present, you can compare simulated V-I curve with experiments.
...
# Define current density points to be evaluated as list with unit [A/cm^2]
current_list = [0.064, 0.13, 0.188, 0.248, 0.375, 0.502, 0.626, 0.753, 0.874, 0.998, 1.125, 1.246, 1.373, 1.496, 1.623, 1.744, 1.868]
# If experimental data are present, give them as two seperate lists for current density and voltage
experiment_current = [0.064, 0.13, 0.188, 0.248, 0.375, 0.502, 0.626, 0.753, 0.874, 0.998, 1.125, 1.246, 1.373, 1.496, 1.623, 1.744, 1.868]
experiment_voltage = [1.017, 0.931, 0.872, 0.826, 0.753, 0.695, 0.644, 0.597, 0.553, 0.517, 0.479, 0.439, 0.404, 0.369, 0.331, 0.296, 0.257]
sim = model.isothermal_voltage_current_curve(J_list=current_list, experiment_J=experiment_current, experiment_V=experiment_voltage, experiment_name='Jung-2005')It returns the following dictionary of results and shows plots by default. If you don't want to have plots to be shown, set function parameter show_plot=False.
>>> sim
{"J":Current Density, "V":Voltage, "ASR":Area Specific Resistance,
"FU":Fuel Utilisation, "POWER":Electrical Power, "ERROR":Error relative to experiments}
# Each item contains results as list in the same order of current density.
# ERROR is calculated only if experimental data are provided, otherwise retturns None.Screenshot of generated plots
You can also calculate operation voltage for a single current density point or vice versa.
...
model.calculate_operation_voltage(current_density= 0.5)
>>> 0.71975723
model.calculate_current_density(operation_voltage= 0.71975723)
>>> 0.500009In example above all required parameters of a Solide Oxide Cell are read from the file cell_parameters_jung2005.json. Please see the structure of a cell parameter file pysofc/Parameter Files/cell_parameters_jung2005.json.
Alternatively, these parameters can be defined or altered by changing corresponding variables.
...
fuel_cell = HydrogenSOFC("SOFC-1") # Note that load_from_json is removed
model = LumpedModelHydrogenSOFC(fuel_cell)
# Define or alter parameters one by one
fuel_cell.dimensions["cell_width"] = 0.005
# Or replace with dictionary as whole.
# If it is not relevant for the simulation, you can define it as None.
fuel_cell.dimensions = {
"channel_height": None,
"channel_width": 0.0005,
"channel_length": 0.01,
"total_electroactive_area": 0.0016,
"thickness_cathode_electrode": 30e-6,
"thickness_anode_electrode": 20e-6,
"thickness_electrolyte": 8e-6,
}
...Please see the exact structure of parameters of HydrogenSOFC class in pysofc/solid_oxide_cell.py. Using GUI program makes this process easier, and is the recommended method to create/modify parameter .json files.
Functions in pysofc/utility.py module can be used stand-alone (static methods).
from pysofc import BasicThermo
from pysofc import PolarizationFunctions
# Available species: "H2", "H2O", "O2", "CO", "CO2", "CH4", "N2"
BasicThermo.enthalpy_gas("H2", T= 600) # Calculates the gas phase enthalpy of Hydrogen at 600 K
>>> 8804.468 # J/mol
BasicThermo.entropy_gas("H2O", T= 600) # Calculates the gas phase entropy of Hydrogen at 600 K
>>> 151.086 # J/mol.K
# Binary diffusion coefficient [m^2/s] of Water-Hydrogen at 800 K and 1 bar
PolarizationFunctions.binary_diffusion_explicit(specie_1="H2", specie_2="H2O", temperature=800, pressure=1)
>>> 0.0005156019See pysofc/utility.py for other possible usages.
Ongiong works:
- Simulation of SOFC running on reformate fuels containing Methane, Carbon monoxide, Carbon dioxode
- One-Dimensional SOFC modeling both hydrogen and reformate fuels.
- Integration of surface reaction kinetics for reformate fuels using Cantera
- New GUI that covers other chemical components and models
Possible improvements:
- Implementing a One-Dimensional model using SIMPLE algorithm.
- Publishing all as Python package