JOURNAL of ELECTROCHEMICAL

SCIENCE and TECHNOLOGY

RESEARCH ARTICLE

Kinetic and Thermodynamic

Parameter Optimization of
an Anode-Supported Solid Oxide
Fuel Cell: A Computational Fluid
Dynamics Study
Huma Hussain1 ⏐ M.W. Tahir1* ⏐ Adil Mehmood1 ⏐ Muhammad Yousaf Arshad2

1
Department of Chemical Engineering, ABSTRACT
University of Engineering and Tech-
nology, 54890 G.T. Road Lahore In recent years, Anode-Supported Solid Oxide Fuel Cell (SOFC) technology
2
School of Chemical Engineering, The gains greater attention. Kinetic and thermodynamic parameters play a pivotal
University of Adelaide, Adelaide role in influencing the overall efficiency and functionality of SOFCs. In present
5005, SA, Australia study, 3D model of a single cell anode-supported SOFC is developed using
ANSYS fluent. Kinetic and thermodynamic parameters are optimized by com-
parison with experimental data. Optimized parameters are subsequently
used to investigate the impact of pressure and fuel utilization on the over-
potential. Simulation results indicate that optimized parameters give rise to
an increased fuel consumption and hence water formation. This can be
explained by the fact that optimized parameters give rise to an increased rate
of reaction. Optimized parameters lead to a reduction in cell overpotential
thereby increasing current and power densities at elevated pressure.
Increased fuel utilization correlates with a decrease in power and current
densities.

KEYWORDS
Fuel cell, Computational fluid dynamics, Solid oxide fuel cell, Modeling and
simulation

CORRESPONDENCE
T: +92-322-4032321
E: [Link]@[Link]
[Link]

EPUB AHEAD OF PRINT RECEIVED 13 March 2024 ACCEPTED 4 November 2024

This is an open-access article distributed under the terms of the Creative Commons Attribution Non-Commercial License ([Link]
which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
2 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

INTRODUCTION
Last few decades have observed an exponential growth in world population that puts extra burden on the
energy supply chain. Present day energy demand is met mostly through the conventional sources like fossil
based fuels that produce undesired by-products causing air pollution and greenhouse gas (GHG) emissions.
Fossil-fuel-based environmental pollution is the major cause of smog formation and global warming [1].
Human interventions caused by the abundant and unprotected use of fossil-based fuels are the major fac-
tors upsetting the global ecosystem and energy balance of the Planet. In addition to the above factors, fluc-
tuations in global oil market and depletion of oil reserves are the other vulnerabilities associated with the
use of fossil fuels. These issues paved the way for the scientists and researchers to explore alternative
sources of energy that are clean and sustainable [2]. These alternate sources of energy include fuel cells
(FCs) and electrolyzers that are electrochemical energy conversion devices used to convert chemical energy
of gaseous fuels e.g., hydrogen into electrical energy [3]. FCs are classified based on the operating tempera-
ture (high, medium, and low), operating mode (active and passive), and electrolyte membrane type and
concentration. Platinum (Pt) and Pt-alloy catalysts used in low-temperature FCs (e.g., Proton exchange
membrane FC or PEMFC) are not only expensive but also susceptible to contaminants like CO. Solid Oxide
Fuel Cells (SOFCs) fall under the high-temperature category. Transition metals are used in both the elec-
trodes (i.e., cathode and anode) enabling the SOFCs to handle a variety of different fuels such as biogas and
syngas [4,5]. SOFCs are cost-effective, exhibit high productivity, and appear to have no adverse environmen-
tal impact. A number of advantages are associated with the high temperature operability (700oC–1000oC) of
SOFCs. For example, gas leaving the anode (anode-off-gas or AOG) still contain significant amount of ther-
mal energy due to high temperature operation of SOFCs. This could be used to feed endothermic reforming
reactions in a cogeneration or CRP (combustion-reforming-power) system. Hence, efficiency of the SOFC-
based CRP systems could go as high as 90% [6]. Numerous design configurations of SOFC have been
reported in literature. Anode-supported SOFC (AS-SOFC) design is popular amongst the other designs. In AS-
SOFC, anode acts as the support-structure for cell components. This allows a thinner electrolyte layer (EL)
giving rise to lower ohmic resistance across the electrodes [7]. Four major components or layers constitut-
ing AS-SOFC are cathode functional layer (CFL), electrolyte layer (EL), anode functional layer (AFL), and
anode support layer (ASL). Ni/Yttria-stabilized zirconia (YSZ), Strontium-doped lanthanum manganite (LSM)
with YSZ, and YSZ are the most popular materials serving as the anode, cathode, and electrolyte respec-
tively of AS-SOFC. Fuel cell efficiency has direct impact on the energy production cost of the cell. Fuel cell
efficiency can be improved through optimization of design and operating parameters [8]. Design parame-
ters are associated with the geometry and dimensions of the fuel and oxidizer flow channels. In addition,
structural parameters like thickness, porosity, and permeability of the CFL, EL, and ASL have substantial
impact on the performance and efficiency the fuel cell. Operating parameters impacting the cell perfor-
mance include operating temperature and pressure (fuel and oxidizer) [9].
Experimental characterization and parameter optimization of SOFCs is laborious, expensive, and time-
consuming. Besides, high operating temperature in SOFCs further limits the experimental measurement
and evaluation of these parameters [10–12]. Mathematical modeling, on the other hand, has shown prom-
ise in characterizing fuel cells and components in a time-efficient and cost-effective manner. Different mod-
eling approaches have been used to improve fuel cell performance through parameter optimization [13].
Tubular and planar design configurations have been the focus of modeling and simulation oriented research
of SOFCs [14]. Nonetheless, planar architecture offers numerous advantages like simple construction, easy
to fabricate, and higher power density [7,15]. Numerous studies corresponding to SOFC parameter optimi-
zation using different numerical approaches are reported in literature. Effect of design and structural
parameters on the performance of different SOFC configurations has assiduously been studied and reported
in literature [10,16–23]. Similarly numerous studies corresponding to the effect of operational parameters
on SOFC performance have also been reported [8,9,24–27]. In addition to the design and operating param-
eters, SOFC performance is directly linked to the rate of electrochemical reactions taking place within the
electrodes and voltage drop across the electrodes. Voltage drop is caused mainly by the polarization (activa-
tion and concentration) and ohmic resistances. SOFC power is a function of current density (a measure of
electrochemical reaction rate) and cell potential. Hence reduction in either of the two parameters i.e., cur-
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 3

rent density and potential negatively impacts the power deliverance capabilities of the fuel cell. Rate of
electrochemical reaction and cell potential are regulated by the kinetic and thermodynamic parameters
respectively. Hence, in addition to the operational and design parameters, optimization of kinetic and ther-
modynamic parameters could significantly improve performance and maximize efficiency of the SOFCs.
Contrary to the widely reported numerical studies regarding structural and operational parameters of
SOFCs, research articles corresponding to the SOFC reaction kinetics and thermodynamics are seldom found
in the published literature. Thence, some of the recently published research articles devoted exclusively to
the numerical studies focusing SOFC reaction kinetics and thermodynamics are presented.
Electrochemical kinetics and polarization related losses in SOFCs are generally studied using semi-empiri-
cal models. Semi-empirical models require expensive experimental data for building the correlations and
validation of simulation results. Physics-based models involving mass, momentum, and energy balances
show promise in this regard. Such models are computationally intensive but more accurate [28]. Model
dimensionality is another important aspect impacting accuracy and computational burden. In 0D models
the cell domain is represented by a single point in space used to evaluate global performance of the cell. 3D
models predict the cell local responses and give a clear picture of spatial gradients occur within the cell
domain. 1D and 2D models represent trade-offs between computational speed of 0D models and prediction
accuracy of 3D models [29]. Bruno Conti et al. [30] proposed a simplified semi-empirical kinetic model for
an intermediate-temperature SOFC. Model assumed linear dependence of parameters on ohmic and activa-
tion overpotentials while neglecting concentration variations. Detailed experiments were carried out to
measure temperature and gas composition for model validation. Simulation results were found in good
agreement with the measured data. F. R. Bianchi et al. [31] extended the model proposed in [29] for a but-
ton AS-SOFC into a more generic formulation. 0D and 1D steady-state mass balance equations were used
simulate the cell global responses and optimize concentration overpotential respectively. Optimized con-
centration overpotential was subsequently used to evaluate the effect of diffusion across the cell thickness
on cell voltage. Y. Sahli et al. [32] proposed a thermodynamic model to optimize SOFC performance. Effect
of losses due to ohmic and polarization resistances and species concentrations were investigated. SOFC
power output was found to be directly proportional to oxidizer O2 concentration and operating tempera-
ture while inversely proportional to the thickness of AFL, EL, and CFL. It was found that temperature has
more pronounced effect on SOFC power density than the thickness of functional layers. Min, Park and Hong
[33] proposed a 1D thermodynamic model for SOFC stack to optimize operating conditions. Simulations
supplant tedious, costly, and time-consuming experiments required to find optimum operating conditions
for SOFC stack. Due to the involvement of Multiphysics phenomena and complex dynamics of SOFC sys-
tems, lower model dimensionalities often prove insufficient in adequately capturing the local responses and
spatial gradients within the cell. 3D CFD models involving coupled physics on a 3D level have shown promise
in resolving these shortcomings. Fewer research articles are found in literature focusing on the comprehen-
sive evaluation of electrochemical kinetic and thermodynamic parameters using CFD simulations with
higher model dimensionality. Li et al. [34] developed a 3D non-isothermal model to optimize the perfor-
mance of a planer AS-SOFC as a function of interconnect land-width and thickness of cathode. Transport
equations were solved using volumetric reaction and kinetic source terms. They concluded that reduction in
local temperature as well as thermal gradient is crucial for a highly uniform distribution of current density
indicating a strong link between current density and temperature variation within the cell. A. N. Celik [35]
implemented a 3D Multiphysics model into COMSOL Multiphysics software package coupling electrochemi-
cal model equations with computational fluid dynamics (CFD). This coupled approach gives a comprehen-
sive understanding of the system behavior through different operating and boundary conditions. Tabish et
al. [36] developed a 2D CFD model of anode supported SOFC using triple-phase boundary (TPB) based kinet-
ics. Effect of operating temperature and TPB density on the cell performance was investigated.
Reference exchange current density, transfer coefficients, reaction enthalpy, and entropy are the major
kinetic and thermodynamic parameters having pronounced impact on the performance and efficiency of
SOFCs. Due to the inter-dependence of Multiphysics phenomena, 3D simulations using Multiphysics CFD
software are required to optimize these parameters. In view of the above cited literature, such studies have
not been reported that justifies the efforts behind present study. Present work addresses this gap by opti-
mizing the above-mentioned kinetic and thermodynamic parameters using a Multiphysics CFD simulation
4 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

software ANSYS Fluent. The aim is to contribute to the ongoing efforts in SOFC research supporting the goal
for clean and sustainable energy future.

MODEL DEVELOPMENT
A 3D model of a planer AS-SOFC was developed accounting for electrochemical and thermodynamic
properties of the system. The model was implemented into a CFD simulation software package ANSYS Flu-
ent. ANSYS Fluent is a Multiphysics CFD software capable of simultaneously solving transport equations
(mass, momentum, heat, and species) coupled with electrochemical reaction kinetics using computational
algorithms. A built in SOFC Add-on module of ANSYS Fluent was used to compute the electric parameters
across 3D cell domain.

Model Geometry
A 13-channel planar AS-SOFC designed at the Chinese academy of science Ningbo Institute of Material
Technology and Engineering (NIMTE) has been selected for this study. Owing to the symmetric construction
of the squared-shaped cell, simulation domain was reduced to a single channel to reduce computational
cost. The actual domain consisted of 13 rectangular-shaped air (oxidizer) and fuel (hydrogen) channels.

Table 1. Geometrical design parameters used to develop CAD model [37]

Parameter Numeric values Units
Cell length 40 mm
Cell width 40 mm
Channel width 2 mm
Channel height 1 mm
Thickness of anode diffusion layer 0.38 mm
Thickness of anode functional layer 0.02 mm
Thickness of electrolyte 0.01 mm
Thickness of cathode diffusion layer 0.05 mm
Thickness of cathode functional layer 0.02 mm
Thickness of interconnect 1.3 mm

Fig. 1. (a) Cross-section of single-channel AS-SOFC elucidating different componenets; and (b) CAD model of the
single channel AS-SOFC with co-flow configuration.
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 5

Reduced single channel geometry model shown in Fig. 1b was developed using ANSYS Design Modeler. Geo-
metrical design parameters derived from [37] are presented in Table 1.
Schematic diagram of a planar AS-SOFC cross-section is shown in Fig. 1b. The model consisted of anode
support (diffusion layer), anode catalyst (functional layer), electrolyte layer, cathode support (diffusion
layer), cathode catalyst (functional layer), flow channels, and interconnects. Diffusion and catalyst layers are
also referred to as the gas-diffusion-layer or GDL and triple-phase-boundary or TPB.

Mesh Independence
Accuracy of the simulation results have strong
dependence on quality of the mesh generated as a
result of discretization of the domain. Hence, a high-
quality rectangular or mapped mesh (with hexahedral
elements) instead free mesh (with tetrahedral ele-
ments) was used across the whole domain. Prior to
conducting numerical simulations, mesh quality
check was performed using mesh quality indices e.g.,
element aspect ratio, orthogonal quality etc. to
ensure the accuracy of model predictions. Mesh size
and density significantly impacts simulation results;
hence mesh size independency study was conducted Fig. 2. Effect of mesh density or number of
to further ensure accuracy. The analysis revealed that discretization elements on H2 mass fraction.
increasing mesh density above 93000 elements, simu-
lations exhibited negligible variations in H2 mass frac-
tion illustrated in Fig. 2. Hence, a mesh density of
93000 elements was used for model validation and
subsequent simulation experiments. Simulation
domain with different mesh densities is shown in Fig.
3. Optimum mesh resolution ensures reliability of the
simulation outcomes of the present study.

Governing Equations
During operation, multiple physical and chemical
processes occur simultaneously within different com-
ponents of SOFC. These processes include transport
of species, momentum, mass, and heat, fluid flow,
and charge transport (ions and electrons). Hence a
fully coupled Multiphysics model is developed and
implemented into ANSYS Fluent. The transfer pro-
cesses mentioned above along with the electrochemi-
cal reaction were incorporated in the model. These
processes occur simultaneously and are mutually
influential influencing impact each other [39]. Elec-
trochemical reactions taking place at the anode and
cathode sides are given as:

Anode: 2H2 + 2O2– ↔ 2H2O + 4e–

Cathode: O2 + 4e– ↔ 2O2–

The cell potential is given as a function of open cir-

cuit potential and overpotentials due to activation, Fig. 3. Domain with different mesh resolutions with
concentration, and ohmic losses [40]. corresponding number of elements.
6 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

Ecell = EOCV – ηconc – ηact – ηoh mic (1)

Momentum Conservation
In SOFC, gases flow through channels and diffuse through the porous electrodes. The fluid flow is gov-
erned by Navier Stokes equation. This equation describes the flow in the channels; however, friction is cre-
ated when the gases pass through the porous electrode structures. The additional pressure loss induced by
the movement of gas through the porous media can be described by Darcy’s law. The momentum loss of a
fluid moving through a porous media is represented by the Eq. 2. Incompressible flow is assumed in both
the channels and within the porous electrodes while neglecting the effect of gravity.
μ
F = – ----- V (2)
B0

The Brinkman equation describes the fluid flow through the porous media by introducing Darcy’s law into
Navier Stokes equation.

ρ
--- ( V ⋅ ∇ )V = – ε∇P + ∇ μ ⎛ ϕ – 2
--- ∇ ⋅ V⎞ + εF
⎝ ⎠ (3)
ε 3

Where μ is the dynamic viscosity of gaseous mixture; P is the total gas pressure in the mixture; ε is poros-
ity of the porous material; V is the velocity vector; ϕ is the viscous stress tensor and B0 is permeability of
the porous structures [39].

Species Transport
Transport of gas components in the porous electrodes and gas channels is affected by diffusion and con-
vection respectively. A modified Fick’s law is used to better describe the diffusion behavior of gases in flow
channels and porous media. incorporation of Knudsen diffusion. For each chemical species ι the conserva-
tion equation is given as:

∇ ⋅ ( ρω i V ) + ∇ ⋅ J ι = R i (4)

∇ ⋅ ( ρω i V ) and ∇ ⋅ J ι represent convection and diffusion the terms. Ri accounts for the rate of consump-
tion and production of species per unit volume due to chemical and electrochemical reactions. ωi is the
mass fraction of component i; ρ is the density of the multi-component mixture; V is the velocity vector; J ι
is the diffusion flux of component i given by the diffusion model.
MO
R O2 = – ----------2 J c (5)
2F
MH
R H2 = – ---------2 J a (6)
2F
M H2 O
R H2 O = – ------------
-J (7)
2F a
Where F is Faraday’s constant; M is the molar mass; J is the exchange current density and subscript c and
a represent cathode and anode respectively. Exchange current density can be expressed as:

RT ( – E act )
J 0 = -------- k e exp ---------------- (8)
ne F RT

ke is pre-exponential factor and Eact is the activation energy.

Charge Conservation
The ions and electron are in motion during an electrochemical reaction and their mobility with ionic and elec-
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 7

tronic potentials follows Ohm’s law. Ohm’s law can be used to formulate the equation of charge conservation.

eff ⎧ – S a ⋅ i trans,an ⎫
∇ ⋅ ( – σ elec ∇ϕ elec ) = ⎨ ⎬ (9)
⎩ – S a ⋅ i trans,ca ⎭

eff ⎧ S a ⋅ i trans,an ⎫
∇ ⋅ ( – σ ion ∇ϕ ion ) = ⎨ ⎬ (10)
⎩ S a ⋅ i trans,ca ⎭

Where specific surface area is represented by Sa; in addition, itrans,an and itrans,ca represent the transfer cur-
rent density produced by electrochemical process at anode and cathode respectively. ϕion and ϕelec are ionic
and electronic potentials respectively. The following expression can be used to compute effective ionic and
electronic conductivities in Eq. 9 and 10 [41].
eff eff
σ ion = σ ion ( 1 – θ ) ( 1 – ε ) ; σ elec = σ elec θ ( 1 – ε ) (11)

In the porous electrode, the relative volume fraction of electron carrier particles is denoted by θ.
Whereas σion for anode and cathode (σion,an, σion,ca) and σelec are given by [42]:
7
9.5 × 10 – 1150
σ ion,an = ---------------------- ⋅ exp ⎛⎝ ---------------⎞⎠ (12)
T T
7
4.2 × 10 – 1200
σ ion,ca = ---------------------- ⋅ exp ⎛⎝ ---------------⎞⎠ (13)
T T
– 10300
σ elec = 3.34 × 10 ⋅ exp ⎛⎝ ------------------⎞⎠
4
(14)
T

Mass Conservation
Despite the fact that species generate and degenerate during the ongoing reaction within the SOFC
anode and cathode, the total mass of materials remains conserved. The following is the mass conservation
equations [36].
At gas channels: ∇ ⋅ ( ρ ⋅ V ) = 0 (15)
At Electrodes: ∇ ⋅ ( ρ ⋅ V ) = W· (16)
· is the mass source term described by the following
Where V gas flow velocity; ρ is the density; W
expressions:
At anode: W· = (M – M ) S i ⁄ 2F (17)
H2 O H2 an an
· = M S i ⁄ 2F
At cathode: W (18)
O 2 an an

Energy Conservation
The temperature field across the entire domain described by the energy conservation equation. In porous
electrodes, the local thermal equilibrium concept indicates that the fluid and solid phases are at the same
temperature. In SOFC convection, gas and solid phase heat conduction, heat transport by the gas species,
and heat source term are incorporated in the thermal energy equation [43].

∇ ( ρ e c P VT ) = ∇ ( k eff ΔT ) + S T (19)

Where cP and keff are specific heat capacity and effective thermal conductivity terms given by:

c P = εc g + ( 1 – ε )c s (20)
8 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

k eff = εk g + ( 1 – ε )k s (21)
cg and cs represent the specific heat capacity of gas and solid while kg and ks denotes the thermal conduc-
tivity of gas and solid respectively. ST is the heat source term which is equal to the sum of heat generated
due to polarization overpotential, joule heating, and heat due to electrochemical reaction [34].
S T = S act + S J + S r (22)
2 2
ie is dU
S T = Jη + -------
- + -------- + JT ---------0 (23)
κ
eff
σ
eff dT
2 eff 2 eff
Where Jη represents polarization overpotential, i e ⁄ κ and i s ⁄ σ are ohmic and joule heating terms,
and JT ( dU 0 ⁄ dT ) is the heat losses due reaction (entropy change).
Extended form of the model equations with detailed description of different terms can be found in
[12,37,38].

Boundary Conditions
ANSYS Fluent was used to solve the model equations. User-defined scalars (UDS) were used for the
implantation of charge transport equations and the source terms. At the anode and cathode flow channel
inlets, mass flow rate, species mass fractions, and temperature were applied as inlet boundary conditions.
Pressure-outlet type boundary condition is used at the outlet of flow channel. SOFC anode was used as
grounded and hence an electric potential of 0 volt was applied. Whereas, the electric potential at the cath-
ode was used as Vcell. Adiabatic boundary conditions were used at the cell walls to simulate thermal insula-
tion. Detailed description of the boundary and operating conditions is presented in Table 2. User Defined
Function (UDF) was used for the implementation of momentum and energy source terms.

Table 2. Operating and boundary conditions used in the simulations

Parameters value Reference
Thermal conductivity k, at anode (W/m·K) 11
Thermal conductivity k, at cathode (W/m·K) 6
[44]
Thermal conductivity k, at electrolyte (W/m·K) 2.7
Thermal conductivity k, at interconnect (W/m·K) 20
2 3
Active surface to volume ratio (m /m ) 108000 [37]
Specific heat cp, at anode (J/kg·K) 495
Specific heat cp, at cathode (J/kg·K) 573
[45]
Specific heat cp, at electrolyte (J/kg·K) 606
Specific heat cp, at interconnect (J/kg·K) 550
Porosity 0.3 [44]
Density ρ, at anode (kg/m ) 3
3030
Density ρ, at cathode (kg/m3) 3310
[45]
Density ρ, at electrolyte (kg/m ) 3
5160
Density ρ, at interconnect (kg/m3) 8030
Activation energy at anode Eact,a (kJ/mole) 140
[37]
Activation energy at anode Eact,c (kJ/mole) 137
Ref. exchange current density at anode, J0,a (A/m2) 3500
2
[46]
Ref. exchange current density at cathode, J0,c (A/m ) 2013
Transfer coefficient α, anode/cathode 0.5 [37]
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 9

Table 2. Continued
Parameters value Reference
3
Reference H2 concentration, (mol/m ) 10.78
3
[44]
Reference O2 concentration, (mol/m ) 2.38
Faraday’s constant F, (C/mol) 9.65×107
Pre-exponential factor ke, at cathode (1/Ω·m2) 2.35×1011
Pre-exponential factor ke, at anode (1/Ω·m2) 6.54×1011
Operating temperature T (K) 1073
2
Electrolyte projected area (m ) 1.2×10–4 [37]
–8
Mass flow rate of fuel (H2) at anode (kg/s) 2×10
Mass flow rate at of air (O2) at cathode (kg/s) 8×10–7
Mass fraction at inlet anode ( M H2 / M H2 O ) 0.95/0.05
Mass fraction at inlet cathode ( M O2 / M N2 ) 0.233/0.767

RESULTS AND DISCUSSION

SOFC simulation domain was discretized using finite volume method (FVM). ANSYS Fluent Add-on Fuel
Cell Module was used for the model implementation. Convection and diffusion coupling was implemented
using the second order upwind differencing scheme (UDS). To implement source terms in the governing
equations, user-defined functions (UDF) were used. UDS was written using a C-based code which was com-
piled using Visual Studio 17.0 compiler. Prior to carrying out the detailed simulation study and parameter
optimization, it is important to validate the model for authenticity of the predictions. Hence, the following
section includes the validation results and description.

Model Validation
Model validation is an essential step in proving the authenticity of the subsequent predictions resulting from
detailed simulations. To validate the model, polarization curve resulting from numerical simulation was com-
pared to the experimental data derived from [37]. The Chinese Academy of Sciences, Ningbo Institute of Material
Technology and Engineering (NIMTE). NIMTE produced Ni-YSZ anode-supported SOFCs, which were used for
both modeling and testing. The data presented in [37] used for model validation was the outcome of experiments
carried out at NIMTE. Details of the experimental setup and procedure have not been provided in reference [37].
Nevertheless, details regarding operating conditions e.g., flow rates, temperature etc. were provided. Anode and
cathode were supplied with humidified hydrogen and air respectively. The operating temperature used in the
experiments was 800°C (1073 K). Fuel intake mass flow rate per duct was 2×10−8 kg/s, whereas air inlet mass flow
rate per duct was 0.8×10−6 kg/s. Operating voltage of 0.7 V was used applied at the cathode keeping the anode
grounded (0 V). Simulations were carried out using the same operating conditions used in the experiments. A
previous research article [38] from the same institute i.e., NIMTE, Chinese Academy of Sciences presents a similar
study with details of the experimental setup procedure included. Component dimensions of the AS-SOFC used in
the study were quite the same with minor differences. Hence reference [38] could be referred for details of the
experimental procedure adopted in [37].
Experimental and simulated polarization curves are shown in Fig. 4. A reasonably good agreement is found
between the simulated and measured polarization curves. Validation results prove authenticity of the
model for subsequent investigations based on detailed simulation experiments. Slight deviation at the mid-
dle and towards the end of the polarization curve could be attributed to the assumptions and simplifica-
tions made in an attempt to reduce the model complexity and computational load of the simulations.
Nonetheless, deviation is well within the allowable range for most of the model predictions. Furthermore,
constant exchange current density and entropy values were used in the model, but in reality, they vary as a
function of other parameters. This could also be the contributing factor to the deviation.
10 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

Fig. 4. Comparison between experimental and simulated polarization curve.

Effect of Reference Exchange Current Density

Different values of reference exchange current densities (J0) have been adopted from previous studies.
Simulations were carried out at different applied voltage using different values of reference exchange cur-
rent densities. In Fig. 5, V–I curves with different values of reference exchange current density are plotted
along with the experimental curve. Exchange current density has pronounced effect on the local current
density as can be seen in Fig. 5. V–I curve corresponding to the anodic and cathodic reference exchange cur-
rent densities of 3500 A/m2 and 2013 A/m2 respectively fall within the closest vicinity of experimental V–I
curve. Reference exchange current densities is an important kinetic parameter that significantly impacts
local current density and hance the power output of SOFC.

Fig. 5. Applied cell potential versus current density as a function of reference exchange current density (J0).
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 11

Fig. 6. SOFC power output as a function of anodic and cathodic reference exchange current densities.

Fig. 7. Hydrogen fraction as a function of anodic and cathodic exchange current densities at applied voltage 0.7 V.

Effect of reference exchange current density on the local current density of SOFC is shown in Fig. 5. It can
be seen that the maximum current density is delivered with anodic and cathodic reference exchange cur-
rent densities corresponding to 7460 A/m2 and 10090 A/m2 respectively. Power output of SOFC as a func-
tion of applied current density is shown in Fig. 6. Maximum power is delivered at the anodic and cathodic
reference exchange current densities corresponding to 7460 A/m2 and 10090 A/m2 respectively. So, the
optimized values of reference exchange current density correspond the maximum power output of SOFC.
Hence in this case, 7460 A/m2 and 10090 A/m2 are the optimized reference exchange current densities for
anode and cathode respectively.
In SOFC during operation, hydrogen enters the anode side, diffuses through porous electrode i.e., GDL
and reacts with oxygen to form water. So, major part of hydrogen entering the anode gets consumed with
some left unreacted and leaves at the channel outlet. Fig. 7 shows contour plots of hydrogen mass fraction
at different exchange current densities of the anode and cathode with 0.7 V applied voltage. It can clearly
12 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

be seen in Fig. 7 that more of the hydrogen gets consumed as we move from top to bottom. Maximum
hydrogen consumption is achieved at J0,a = 7460 A/m2 and J0,c = 10090 A/m2 while fraction of H2 leaving the
anode is as little as 0.0006. This indicates an increases reaction rate and favorable kinetics. The lowest con-
sumption of hydrogen is observed at J0,a = 2100 A/m2 and J0,c = 700 A/m2 with unreacted mass fraction of
0.18. In this case the amount of current obtained is also minimum as compared to the other cases. This sug-
gests slow reaction rate and unfavorable kinetics.
As reaction proceeds, hydrogen react with oxygen and the water formed at anode side. Fig. 8 Shows con-
tours of water mass fraction corresponding to different exchange current densities at 0.7 V applied voltage.
Higher water formation indicates higher reaction rate and favorable reaction kinetics. Maximum rate of
reaction indicated by higher water formation is observed at J0,a = 7460 A/m2 and J0,c=10090 A/m2. Water for-
mation starts very close to the inlet implying that reaction starts immediately as the gas (hydrogen) enters
the channel again indicating the faster kinetics. The lowest water formation is observed at J0,a = 2100 A/m2
and J0,c = 700 A/m2. It can be seen that water starts forming far from the inlet indicating slower kinetics.
Oxygen or air is used as oxidizer in SOFCs that enters the inlet of cathode channel and leaves at the outlet.
In the simulations used in this study air was used as the oxidizer but only oxygen considered while analyzing
the results since only oxygen in the air reacts with hydrogen during the reaction. Fig. 9 shows that O2 fol-
lows the same pattern as that of the hydrogen. Again, different reference exchange current densities were

Fig. 8. Water formation as a function of anodic and cathodic exchange current densities at applied voltage of 0.7 V.

Fig. 9. Oxygen fraction as a function of anodic and cathodic exchange current densities at applied voltage 0.7 V.
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 13

used in the simulation experiments applying 0.7 V as the operating voltage. Minimum reaction rate was
observed at J0,a = 2100 A/m2 and J0,c = 700 A/m2 that corresponds to oxygen fraction of 0.182 at the cathode
outlet. On the contrary maximum reaction rate was observed at J0,a = 7460 A/m2 and J0,c = 10090 A/m2 indi-
cated by the immediate consumption of oxygen as it enters the cathode. Maximum reaction rate is also
indicated by the lower fraction of oxygen at the cathode outlet i.e., 0.056.

Effect of Transfer Coefficient

In addition to exchange current density, transfer coefficient (α) is another important kinetic parameter
that influence the reaction kinetics and hence the power deliverance capabilities of SOFC. Hence, optimiza-

Fig. 10. Effect of transfer coefficient (α) on V–I behavior of anode supported SOFC.

Fig. 11. SOFC power output as a function of anodic and cathodic transfer coefficients.
14 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

tion of this important kinetic parameter is also worthwhile and opted in this study. From Fig. 10, it can be
seen that changing α, significantly effects the characteristics of V–I curve which is the indicator of SOFC output
power. Maximum power is achieved at α=0.5 (anodic), 0.5 (cathodic) as indicated by the graph in Fig. 11.

Effect of Entropy
Entropy (∆S) is an important thermodynamic parameter influencing the reaction rate and hence the per-
formance of SOFC. In this case, entropy of the reaction is varied to see its impact on the V–I curve and
power output of the SOFC. Heat source term in energy conservation equation contain entropy change (a
thermodynamic term), which represents heat production due to reaction. A user define function (UDF) was

Fig. 12. Effect of entropy change on V–I behavior of the anode supported SOFC.

Fig. 13. Effect of entropy change on the power output of the anode-supported SOFC used.
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 15

used to define energy source term by compiled in Visual Studio. For each of the simulation experiments, a
new value of ∆S was used in the UDF. Fig. 12 shows V–I curves with different reaction entropies. It can be
seen that at higher current densities and lower applied voltages effect of entropy becomes more dominant.
It can be explained by the increase reaction rate at higher currents that gives rise to higher polarization and
ohmic resistances. Fig. 13 shows the power curve under different values of entropy change. At peak power
i.e., higher current, effect of entropy becomes more dominant. At peak power maximum current density
corresponds to maximum reaction rate, since, current density is the direct measure of the rate of electro-
chemical reaction. Hence, the effect of entropy is dominant and visible more clearly.
Water formation occur at anode side due to the reaction between hydrogen and oxygen. The higher
water formation implies reaction rate between the reacting species is higher. Water formation at the anode
side during reaction between hydrogen and oxygen is elucidated by the contours of water mass fraction
shown in Fig. 14. As mentioned above and illustrated in Fig. 12 and Fig. 13 that entropy has little impact on
the cell power and hence performance except for the sufficiently higher currents or reaction rates. The
same is supported by the contours of water fraction shown in Fig. 14. There is negligibly small variation in
contours of water fraction produced at different entropy values. This can be explained by the fact that at
low to moderate power of SOFC i.e., < 0.5 W/cm2 entropy has negligibly small impact on its performance.
But at higher power output entropy starts impacting the SOFC power performance. Results presented in Fig.
12–14 show that maximum power is delivered by the cell at ∆S = –0.056 J/(mol∙K) whereas minimum power
output corresponds to ∆S = –54.9 J/(mol∙K). Mass fraction of water at the outlet corresponding to maximum
and minimum power output of the cell is 0.93 and 0.90 again indicating little impact of entropy.
Another important parameter that determines the performance of SOFC is the overpotential. After iden-
tifying the impact of entropy change on the power output of the cell, it is important to study how does
entropy change influence the cell overpotential. In anode supported SOFC, the overpotential has largest
contribution to overall losses of cell voltage and hence the efficiency. The overall overpotential comprised
by the activation, ohmic and concentration overpotentials. These individual overpotentials correspond to
the resistances due to activation, electronic and ionic flow, and concentration changes. Fig. 15 shows the
effect of entropy on overpotential in terms of V–I curve. During operation in SOFC, overpotential increases
with an increase in current density due to increased reaction rate. In other words, higher power output of
SOFC is associated with increased losses due to an increase in overpotential at higher reaction rate as indi-
cated in Fig. 15. It can be inferred that effect of entropy change on the power output of the cell is negligible
(Fig. 13) but this effect is quite marked in case of overpotential (Fig. 15). Despite a significant difference in

Fig. 14. Contours of water fraction depicting water formation in SOFC as a function of entropy change.
16 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

Fig. 15. Overpotential as a function of current density at different values of entropy.

the magnitude of impact in both the cases, trend is the same. That means ∆S = –0.056 J/(mol∙K) is favorable
in both the cases i.e., it is associated with higher power (Fig. 13) and low overpotential (Fig. 15).

Effect of Operating Pressure

Pressure is an important operating parameter that has pronounced impact of SOFC performance. After
the optimization of kinetic and thermodynamic parameters, operating parameters like pressure and fuel
utilization were also investigated. Fig. 16 shows the effect of operating pressure on the power output and
V–I characteristics of the cell. As mentioned before that V–I curve is a useful way to calibrate operating volt-
age and current to optimize the cell power output.
Operating pressure is varied between 1 bar and 4 bar. It is clear from Fig. 17 that as operating pressure
increases; current and power density increases. Current and power densities increase as the operating pres-
sure increases. As pressure increases, activation and concentration overpotentials drop causing an increase

Fig. 16. Effect of operating pressure on SOFC power output and voltage.
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 17

Fig. 17. Effect of fuel utilization on SOFC current and power densities.

in the power output. At higher pressures, the activation overpotential drops due to increase in fuel concen-
tration in the porous electrode. The gas diffusion rate increases at the same time, resulting in a drop in the
concentration overpotential, which consequently raises the power performance of the cell.

Effect of Fuel Utilization

As mentioned above that fuel utilization is an important operating parameter. In Fig. 17, the effect of fuel
utilization (UF) on current and power curve can be seen. At higher fuel utilization, both current and power
density decreases. This is due to the fuel depletion near the exit of anode assembly. Although high FU is
advantageous in order to attain high electrical efficiency, the main disadvantage is its lower power density.
This implies that a larger cell surface area is required for optimum cell operation. As a result, FU should be
carefully chosen in order to achieve high electrical efficiency at a tolerable power density.

CONCLUSIONS
Kinetic parameters i.e., reference exchange current density and transfer coefficient and thermodynamic
parameter i.e., entropy are studied simultaneously to optimize SOFC performance. First the model is vali-
dated with the experimental voltage–current (V–I) plot. Simulations were carried out using the same
parameters and conditions used in the experiments. The model was found in a reasonably good agreement
with the experimental data. Subsequently SOFC performance optimization study was conducted by carrying
out simulation experiments. The main optimization parameters selected included kinetic and thermody-
namic parameters i.e., reference exchange current density (anodic and cathodic), transfer coefficients, and
reaction entropy. In addition, SOFC operational parameters including operating pressure and fuel utilization
were also studied. Optimized kinetic parameters including J0,a, J0,c, αa, and αc (anodic and cathodic reference
exchange current densities and transfer coefficients) correspond to 7460 A/m2, 10090 A/m2, 0.5 and 0.5
respectively. Optimized entropy (∆S) corresponds to –0.056 J/(mol∙K). Following the kinetic and thermody-
namic parameter optimization, SOFC operating pressure was optimized ranging between 1 bar and 4 bar. It
was found that increasing operating pressure decreases cell overpotential thereby improving the cell per-
formance. Furthermore, cell current and power densities increase at higher pressures giving rise to an
improved cell performance. Finally, the effect of fuel utilization on current and power density was investi-
gated. Simulation results indicate that higher fuel utilization lowers the power density of the cell. Although
different important parameters have been identified and optimized to improve the overall cell perfor-
mance. Nonetheless, further research on the optimization of other physical and chemical parameters like
diffusivity, enthalpy, specific heat, and thermal conductivity is required.
18 JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print

NOMENCLATURE
η Overpotential [V] J Exchange current density R Universal gas constant
α Transfer coefficient [A/cm2] F Faraday constant
σ Conductivity [Ω∙m] τ Tortuosity FU Fuel Utilization
γ Coefficient for exchange ω Mass fraction ΔS Entropy change [J/mol∙k]
current density ε Porosity

SUBSCRIPT
conc Concentration g Gas ref Reference
act Activation e Electrolyte r Reaction
ohmic Ohmic eff Effective s Solid
an Anode Kn Knudsen
ca Cathode trans Transfer

REFERENCES
[1] G. Akhmat, K. Zaman, T. Shukui, and F. Sajjad, Renew. Sustain. Energy Rev., 2014, 36, 123–134.
[2] A. Mehmood, M. W. Tahir, M. A. Saeed, M. Y. Arshad, H. Hussain, J. Mularski, and L. Niedzwiecki, Fire, 2023, 6(9), 361.
[3] S. A. Kashmiri, M. W. Tahir, and U. Afzal, Energies, 2020, 13(19), 5186.
[4] G. N. Prodromidis and F. A. Coutelieris, Renew. Energy, 2017, 108, 1–10.
[5] M. A. Abdelkareem, W. H. Tanveer, E. T. Sayed, M. E. H. Assad, A. Allagui, and S. W. Cha, Renew. Sustain. Energy Rev.,
2019, 101, 361–375.
[6] F. Ramadhani, M.A. Hussain, and H. Mokhlis, Processes, 2019, 7(12), 950.
[7] K. A. Kuterbekov, A. V. Nikonov, K. Zh. Bekmyrza, N. B. Pavzderin, A. M. Kabyshev, M. M. Kubenova, G. D. Kabdrakhimova,
and N. Aidarbekov, Nanomaterials, 2022, 12(7), 1059.
[8] A. Raj, A. P. Sasmito, and T. Shamim, Energy Convers. Manag., 2015, 90, 138–145.
[9] J. Liao, H. Jie, C. Zhang, and W. Hong, J. Solid State Electrochem., 2021, 25, 2321–2332.
[10] M. Andersson, J. Yuan, and B. Sundén, Fuel Cells, 2014, 14(2), 177–188.
[11] J. Kupecki, K. Motylinski, A. Zurawska, M. Kosiorek, and L. Ajdys, Int. J. Hydrogen Energy, 2019, 44(38), 21148–21161.
[12] Z. Zhang, D. Yue, G. Yang, J. Chen, Y. Zheng, H. Miao, W. Wang, J. Yuan, and N. Huang, Int. J. Heat Mass Transfer, 2015,
84, 942–54.
[13] A. M. Nassef, A. Fathy, E. T. Sayed, M. A. Abdelkareem, H. Rzk, W. H. Tanveer, and A. G. Olabi, Renew. Energy, 2019, 138,
458–464.
[14] B. Timurkutluk, C. Timurkutluk, M. D. Mat, and Y. Kaplan, Renew. Sust. Energy Rev., 2016, 56, 1101–1121.
[15] W. G. Wang and M. Mogensen, Solid State Ion., 2005, 176(5–6), 457–462.
[16] S. Altindal, E. G. Erol, and B. Gurel, Int. J. Hydrogen Energy, 2024, 52, 1475–1490.
[17] E. Khodabandeh, O. A. Akbari, S. Akbari, A. Taghizadeh, M. S. Pour, M. Ersson, and P. G. Jönsson, J. Therm. Anal. Calorim.,
2021, 144, 245–256.
[18] X. Lu, A. Bertei, T. M. M. Heenan, Y. Wu, D. JL. Brett, and P. R. Shearing, J. Power Sources, 2019, 434, 226744.
[19] D. Cui, Y. Ji, C. Chang, Z. Wang, X. Xiao, and Y. Li, J. Power Sources, 2020, 460, 228092.
[20] H. Feng, L. Chen, Z. Xie, and F. Sun, J. Power Sources, 2015, 286, 406–413.
[21] I. Khazaee and A. Rava, Energy, 2017, 119, 235–244.
[22] B. Timurkutluk and E. Ucar, Int. J. Energy Res., 2021, 45(15), 20634–20650.
[23] W. Kong, Z. Han, S. Lu, X. Gao, and X. Wang, Int. J. Hydrogen Energy, 2020, 45(39), 20329–20338.
[24] F. M. Cimen, M. Ilbas, and S. Yalcin, Int. J. Hydrogen Energy, 2023, 48(60), 23136–23145.
[25] M. Ilbas, M. A. Alemu, and F. M. Cimen, Int. J. Hydrogen Energy, 2022, 47(5), 3416–3428.
[26] S. Huang, C. Yang, H. Chen, N. Zhou, and D. Tucker, Case Stud. Therm. Eng., 2022, 31, 101868.
[27] W. Kong, W. Zhang, H. Huang, Y. Zhang, J. Wu, and Y. Xu, J. Mater. Sci. Technol., 2018, 34(8), 1436–1440.
[28] F. Ramadhani, M. A. Hussain, H. Mokhlis, and S. Hajimolana, Renew. Sust. Energy Rev., 2017, 76, 460–480.
[29] M. W. Tahir and C. Merten, Energy Convers. Manag., 2022, 258, 115490.
JOURNAL of ELECTROCHEMICAL SCIENCE and TECHNOLOGY Epub ahead of print 19

[30] B. Conti, B. Bosio, S. J. McPhail, F. Santoni, D. Pumiglia, and E. Arato, Catalysts, 2019, 9(1), 36.
[31] F. R. Bianchi, B. Bosio, A. Baldinelli, and L. Barelli, Catalysts, 2020, 10(1), 104.
[32] Y. Sahli, B. Zitouni, and H. Ben-Moussa, U.P.B. Sci. Bull., Ser. B, 2018, 80(2), 159–170.
[33] G. Min, Y. J. Park, and J. Hong, Energy Convers. Manag., 2020, 209, 112614.
[34] Y. Li, H. Yan, Z. Zhou, and W.-T. Wu, Int. J. Energy Res., 2019, 43(13), 6907–6921.
[35] A. N. Celik, Int. J. Hydrogen Energy, 2018, 43(42), 19730–19748.
[36] A. N. Tabish, L. Fan, I. Farhat, M. Irshad, and S. Z. Abbas, J. Power Sources, 2021, 513, 230564.
[37] Z. Zhang, J. Chen, D. Yue, G. Yang, S. Ye, C. He, W. Wang, J. Yuan, N. Huang, Energies, 2014, 7(1), 80–98.
[38] Z. Zhang, D. Yue, G. Yang, J. Chen, Y. Zheng, H. Miao, W. Wang, J. Yuan, and N. Huang, Int. J. Heat Mass Transf., 2015, 84,
942–954.
[39] S. Fan, J. Ye, Q. Yao, D. Wang, L. Yang, and Z. Jin, Int. J. Energy. Res., 2021, 46(2), 876–890.
[40] A. Yahya, S. Hammouda, S. Slimene, and H. Dhahri, Int. J. Therm. Sci., 2022, 174, 107437.
[41] J. M. Park, D. Y. Kim, J. D. Baek, Y.-J. Yoon, P.-C. Su, and S. H. Lee, Energies, 2018, 11(3), 473.
[42] K. Tseronis, I. Bonis, I. K. Kookos, and C. Theodoropoulos, Int. J. Hydrogen Energy, 2012, 37(1), 530–547.
[43] J. Kim, E. Ali, M. Kim, H. Lim, and H. Park, Heat Mass Transf., 2020, 56, 617–625.
[44] Q. Shen, S. Li, G. Yang, N. Huang, J. Yuan, and B. Sundén, Numer. Heat Transf.; A: Appl., 2019, 75(8), 509–522.
[45] D. L. Damm and A. G. Fedorov, J. Power Sources, 2006, 159(2), 956–967.
[46] M. I. Radaideh and M. I. Radaideh, Int. J. Energy Res., 2020, 44, 2517–2534.