Applied Energy 312 (2022) 118788

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Comparative study of alkaline water electrolysis, proton exchange

membrane water electrolysis and solid oxide electrolysis through
multiphysics modeling
Kewei Hu a, Jiakun Fang a, *, Xiaomeng Ai a, Danji Huang a, Zhiyao Zhong a, Xiaobo Yang b,
Lei Wang b
a
State Key Laboratory of Advanced Electromagnetic Engineering and Technology, School of Electrical and Electronic Engineering, Huazhong University of Science and
Technology, Wuhan 430074, China
b
Hitachi ABB Power Grids Research, Beijing, China

H I G H L I G H T S

• A unified multiphysics model for AE, PEM and SOEC is derived to bridge the gap among different PtH researches.
• Various definitions on cell efficiency are collected and evaluated to unify the definitions of efficiency.
• Cell performances are compared and explained from multiphysics process.
• Main drawbacks of the cell performances are reviewed to explain the parametric gaps.

A R T I C L E I N F O A B S T R A C T

Keywords: This paper provides a comparative study of alkaline water electrolysis (AE), proton exchange membrane water
Multiphysics modeling electrolysis (PEM) and solid oxide electrolysis cells (SOEC) through multiphysics modeling. A unified multi­
Power-to-hydrogen physics model for the three types of electrolyzers, considering current density distribution, single-phase and
Cell efficiency
diphasic flow, mass transfer and temperature dependence in the electrolyzers, is derived to make a thorough
Drawback analysis
interpretation of their differences in terms of I-V characteristics, cell efficiency and part-load performances.
Besides, various definitions of electrolysis efficiency from different researches on PtH are compared and eval­
uated from thermodynamics to unify the definitions of efficiency. Main drawbacks of the cell performances are
then reviewed respectively to explain the parametric gap among three types of water electrolysis.

1. Introduction scientific research including mechanics, electromagnetism, electronics

and material science has been carried out. On micro and nanometer
Power-to-hydrogen (PtH) has been recognized as an essential scale, efforts have been made to improve the cell performance by
element in future renewable energy systems for grid balancing and microstructure optimization, e.g. nano- or porous structured catalysts
sector coupling [1–4]. It can not only play the role of energy storage that [14,15]. In the studies at the power device level, there are comparative
contribute to renewable energy accommodation avoiding temporal studies on DC/DC converter topologies [16] and AC/DC converters [17]
power fluctuation that may cause shortage or surplus in energy supply for electrolysis applications. The impact of the fluctuating power supply
on different time scales [5–9], but also possesses the potential to act as has been studied with the comparison of the thyristor and insulated-gate
an energy router coupling electricity and gas systems in future energy bipolar transistor-based power supply topologies [18]. At the system
networks [10–12]. level, a production-curve-based management strategy is proposed to
Among the PtH technologies, water electrolysis is carbon-free and, optimize the intermodular load allocation in a high-temperature PtH
therefore, draws broad research interests [13]. Interdisciplinary plant [19]. As for the hydrogen transport and utilization, the hydrogen-

* Corresponding author.
E-mail address: [email protected] (J. Fang).

https://doi.org/10.1016/j.apenergy.2022.118788
Received 1 August 2021; Received in revised form 16 February 2022; Accepted 18 February 2022
Available online 28 February 2022
0306-2619/© 2022 Elsevier Ltd. All rights reserved.
K. Hu et al. Applied Energy 312 (2022) 118788

enriched natural gas system was proposed and analyzed in [20] where and SOEC through fundamental multiphysics explanations, whose con­
the annual maximum and minimum energy and exergy efficiencies of tributions are summarized as follows:
the system were given theoretically.
The aforementioned literature aims at improving PtH performances – A unified multiphysics model for the three types of electrolyzers is
from various aspects. In fact, the optimization of the electrolyzer is also derived to bridge the gap among different PtH researches.
of high importance since it is the core component of PtH. Currently, – Various definitions on cell efficiency are collected and evaluated
there are three types of electrolyzers, namely alkaline water electrolysis from the perspective of thermodynamics to unify the definitions of
(AE), proton exchange membrane water electrolysis (PEM) and solid efficiency in different research fields.
oxide electrolysis cell (SOEC). They are supposed to be applied to – Cell performances including I-V characteristics, efficiency and part-
different occasions according to their performances on efficiency, part- load range are compared and explained from the multiphysics pro­
load range, dynamic response, cost, etc. An in-depth review on three cess to point out potential scenarios of the three types of power-to-
types of electrolysis technologies is from Alexander Buttler, et.al, in hydrogen technologies.
which the performance of each technology on commercial and research
as well as the performance relevant for PtG applications was summa­ The rest of the paper is organized as follows. Starting from the basic
rized [21]. Another generalized review is proposed by Alfredo Ursua, principle of water electrolysis, various definitions of cell efficiency are
et al., whose investigation focused on the state-of-the-art water elec­ collected in Section 2 to underline the significance of thermal energy
trolysis integration with renewable energy sources [22]. input for water electrolysis, especially for SOEC. In Section 3, governing
Moreover, there are overviews on a certain type of electrolyzer. For equations and boundary conditions of physical fields are listed including
instance, Marcelo Carmo, et al. provided a comprehensive introduction momentum, mass and heat transfer and electrochemical reaction.
on PEM electrolysis and its advantages and disadvantages compared Coupled multiphysics models of AE, PEM and SOEC are then introduced
with the other technologies [23]. the historical development and the respectively. The results derived from multiphysics simulation are dis­
research trends of AE is presented in [24], where the challenges like cussed and validated in Section 4, followed by the analysis on cell effi­
electrolyte improvement and bubble management were also collected. ciency and industrial challenges for further development.
As for SOEC, Lei Bi, et al. provided a broad overview of the research
progresses made for SOECs and pointed out potential development di­ 2. Basic principle of water electrolysis
rections [25].
However, these papers are majorly based on statistics of experi­ The electrolyzer is the device where electric and thermal energies are
mental results with the lack of multiphysics explanations. In fact, water input and transformed into chemical energy stored in hydrogen. The
electrolysis is a multiphysics process where the electrochemical reac­ energy conversion happens at the electrodes of the electrolyzer, namely
tion, the momentum transfer, the mass transfer and the heat transfer are anode and cathode, and can be quantified by the voltage added to them.
coupled. The multiphysics fields interact with one other and determine Therefore, the electrode characteristics play a significant role in water
the performance of the water electrolyzers. Therefore, multiphysics electrolysis, and it is where the multiphysics fields in the electrolyzer are
modeling is an effective way to unveil the intrinsic characteristics of the coupled. Starting from the electrode kinetics, various definitions of
electrolyzer and is widely used in different research fields. For instance, electrolysis efficiency are going to be evaluated from the perspective of
the solid oxide fuel cell is modeled in [26] where the ionic, mass, mo­ thermodynamics in this section.
mentum and heat transfers are coupled. The field distributions including
current density, hydrogen fraction, velocity and temperature are
analyzed to illustrate the interaction of the multiphysics. Besides, the 2.1. Electrode kinetics
Lithium-ion battery is also a system where multiphysics fields are
coupled [27]. The multiphysics including the charge transfer, heat In water electrolysis, the electric potential between the anode and
transfer and solid mechanics are proposed by Qiushi Wang, et al. where the cathode, which is quantified by the reversible voltage Urev , is given
a new strategy to improve the electrochemical stability of the Li-ion by Nernst equation [28]:
battery is designed based on multiphysics modeling [28]. They also RT αH O
propose a facile and scalable method engineering consecutive chemical
0
Urev = Urev − ln √2 ̅̅̅̅̅̅̅ (1)
2F pH2 pO2
bonds to stabilize silicon anode material of the lithium-ion battery
through multiphysics modeling [29]. where R is the gas constant, T the temperature, F the Faraday’s constant,
Nevertheless, the researches on multiphysics modeling of electro­ pi the partial pressure of specific species i, αH2 O the thermodynamic
lyzers are yet quite limited. A comprehensive comparison of the existing 0
activity of water and Urev the reversible potential in standard conditions.
modeling work on AE and PEM technologies, including mass, mo­ Empirical estimation of the relationship of U0rev and the temperature T
mentum and heat transfer as well as electrochemical reactions, was according to [34] is shown as:
concluded in [12]. One rather elaborate PEM electrolysis model inte­
grating diphasic flow field is proposed by S.A. Grigoriev, et.al. where the 0
Urev = 1.253 − 2.4516 × 10− 4 T (2)
gas crossover mechanism was considered [30]. Other models, like those
When the electric energy is supplied, both anodic and cathodic
for PEM electrolysis [31,32] and AE [33], are mostly zero-dimensional
electrodes will be activated and the potential of the electrodes will
or one-dimensional so that the flow field can be neglected. To
deviate, which can be quantitatively described by Butler-Volmer
conclude, there exist several problems in the current PtH researches:
equation:
– a unified multiphysics model suitable for AE, PEM and SOEC is still RT i
η= sinh− 1 ( ) (3)
missing so that different PtH researches are at present poorly linked. nF 2i0
– the definitions of cell efficiency in different researches differ from
one another, making it confusing to compare the results in different where i is the current density, defined positive when the current is
literatures. injected into the electrolyzer; i0 is the exchange current density; n is the
– the reason why the industrial performances of the three types of number of transferred electrons in 1 mol reaction; f is the function of
electrolyzers varies needs unveiling from multiphysics explanations. temperature; η is the overpotential, defined as:
ηa = Ea − φa (4)
Therefore, this paper provides an in-depth comparison of AE, PEM

2
K. Hu et al. Applied Energy 312 (2022) 118788

for the anode and, whose relationship is given by [35]:

ΔG
ηc = Ec − φc (5) Urev = (8)
nF
for the cathode, where E is the electric potential of the electrode, φ the The reversible voltage is, therefore, the minimum voltage required
reversible potential of a single electrode: for the electrochemical process to take place. In standard conditions, Urev
equals 1.23 V. Another critical voltage is the thermal neutral voltage,
Urev = φa − φc (6)
defined as:
As the cell voltage Ucell is supplied to the electrolyzer, the current is
ΔH
injected at the anode and outflows at the cathode. According to (3), the Uth = (9)
nF
anodic overpotential ηa is positive while the cathodic overpotential ηc is
negative. In other words, electrode activation pulls down the potential where ΔH is the enthalpy change of the reaction. If the thermal neutral
of the cathode and elevates the potential of the anode. Fig. 1 shows the voltage is supplied to the cell, the thermal energy absorbed is exactly
electrode potential influenced by activation according to (3). equivalent to the energy losses during water electrolysis. In standard
It can be inferred from Butler-Volmer equation (3) that excess conditions, Uth equals 1.48 V and is rarely temperature dependent [35].
voltage is supposed to be supplied for electrode activation so that the Based on the voltages introduced above, there are at present several
electrochemical reaction could take place [24]. The overpotential for different definitions for hydrogen production efficiency.
electrode activation is represented by activation overvoltage Uact . Be­ The first definition is to compare the minimum energy needed for the
sides, the ohmic overvoltage on the equivalent resistance of the cell, reaction with the supplied electric energy straightforward, that is, the
which is proportional to the current density, exists in the electrolyzer ratio of the reversible voltage Urev to the cell voltage Ucell :
and the cell voltage can be described as:
Urev
Ucell = Urev + Uact + Uohm εelec = (10)
(7) Ucell
= Urev + |ηa | + |ηc | + iRe
which is, according to Fig. 2, the efficiency of electric energy.
where Re is the equivalent resistance of the electrolyzer, caused by the The second definition is widely used in water electrolysis research,
noninfinite conductivity of the electrodes and the membrane. It can be which adds the absorbed thermal energy, represented by TΔS in Fig. 2
observed from Fig. 1 that the minimum cell voltage for the water elec­ and is evaluated by Uth − Urev , to the numerator of (10) and is given by:
trolysis to happen is the reversible voltage Urev . When the cell voltage
exceeds the reversible voltage, the current density i is positive, which εV =
Urev + (Uth − Urev )
=
Uth
(11)
determines the hydrogen production rate. Meanwhile, the activation Ucell Ucell
overvoltage Uact and the ohmic overvoltage Uohm are generated, result­
where Uth is the thermal-neutral voltage. From the perspective of ther­
ing in energy losses and efficiency decline. As the supplied voltage Ucell
modynamics, this definition is the ratio of the absorbed thermal and
increases, the current density is elevated at the expense of higher
electric energy to the total electric energy. Since the thermal energy is
overvoltage. In other words, the selection of the cell voltage is, in
not regarded as the input, this definition is supposed to be avoided. An
essence, the trade-off between the current density and the overvoltage,
example to illustrate the irrationality of the definition is that the effi
that is, the hydrogen production rate and the cell efficiency. Therefore,
the key to promoting the efficiency of the electrolyzer is to reduce the
overvoltage under a certain current density.

2.2. The efficiency of water electrolysis

Efficiency is a major concern of power-to-hydrogen. As discussed

above, the reversible voltage Urev is the effective part of the supplied
voltage while various kinds of overvoltage result in efficiency decline.
From the perspective of thermodynamics, the reversible voltage Urev
corresponds to the Gibbs free energy change ΔG of the water splitting,

Fig. 2. Energy consumption of water electrolysis and cell voltage breakdown

Fig. 1. Electrode activation characteristics of electrolysis. (a)Ucell > Uth (b)Urev < Ucell < Uth .

3
K. Hu et al. Applied Energy 312 (2022) 118788

ciency would exceed 100% when Urev < Ucell < Uth with (11), as shown electrolysis, as shown in Fig. 3(a) and (b). Besides, the concentration of
in Fig. 2(b). the electrolyte in AE will also affect the electrochemical process by the
Therefore, it is may be appropriate to derive another definition electrode potential and the ionic conductivity. On the other hand, the
considering the thermal energy input as: flow pattern in SOEC becomes different since the water is vaporized and
the gas shielding effect disappears. Nevertheless, the inhomogeneous
Urev + (Uth − Urev ) Uth
εcell = = (12) distribution of the water vapor and the products in the transverse di­
Ucell + (Uth − Urev ) Ucell + (Uth − Urev )
rection will result in concentration activation of the electrode, as pre­
By comparing (10) with (12), it is obvious that the thermal energy sented in Fig. 3(c). According to the physical process introduced above,
evaluated by Uth − Urev is added to both the numerator and the denom­ coupled multiphysics in each type of electrolyzers, including electro­
inator. However, (12) is based on the assumption that the thermal ef­ chemical process, mass transfer, diphasic/laminar flow and tempera­
ficiency equals to 100%. As a matter of fact, thermal losses usually exist ture, are modeled in Comsol Multiphysics, as displayed in Fig. 4, Fig. 5
due to the incomplete adiabaticity of the devices and the limited re­ and Fig. 6.
covery rate of waste heat. Hence, based on (12), the definition consid­ While the models of AE and PEM feature the two-way coupling of the
ering the thermal efficiency εth is given as: diphasic flow and the electrochemical process, the mass transfer is the
pivot of the multiphysics model of SOEC since the concentration of the
Uth
εcell = (13) substances would interact with the electrochemical process and the flow
Ucell + (Uth − Urev )/εth
field. The governing and coupling equations of each physics field of AE,
Therefore, (13) evaluates the efficiency of both the electric and PEM and SOEC are introduced in detail in this section.
thermal energy conversions, and is called the cell efficiency. In Section
4, the efficiencies of AE, PEM and SOEC are going to be compared based
on the different definitions.

3. Multiphysics modeling of three types of electrolyzers

Different types of PtH technologies have similar working principles.

First, the electrolyte is heated to the proper temperature before being
pumped into the electrolyzers. As the electrolyte circulates in the sys­
tem, the water is consumed and the hydrogen and oxygen are acquired
through gas-liquid separators. Besides, DC voltages are added to the
electrodes to activate the electrochemical reaction. Therefore, the
electrolyzer is a multiphysics system in which mass, momentum and
energy transports are coupled with electrochemical reaction.
However, the multiphysics modeling of different types of electro­
lyzers differs from one other. For instance, the diphasic flow field in AE
and PEM cells plays a significant role since the coverage rate of
hydrogen and oxygen bubble on electrodes will bring a nonnegligible
influence on electrochemical reactions in low-temperature water Fig. 4. Multiphysics in AE.

Fig. 3. Cathodic boundary layers of AE (a), PEM (b) and SOEC (c).

4
K. Hu et al. Applied Energy 312 (2022) 118788

→
where ρ is the density of the mixture; V is the velocity of the fluid; p is
→ →
the pressure; Fμ is the drag force caused by viscosity; Fσ is the surface
tension; g is the gravitational acceleration. The conservation of mass is
described as:
→
∇∙(ρ V ) = 0 (19)
AE&PEM – Low-temperature electrolysis features the diphasic flow
field where the bubble effect brings a nonnegligible influence on cell
performances [37]. In AE and PEM cells, the parameters in the gov­
erning equations are modified through the mixture model [38]:
ρ = ϕg ρg + (1 − ϕg )ρl (20)
Fig. 5. Multiphysics in PEM.
μ = μg (1 − ϕg )− 2.5[(μg +0.4μl )/(μg +μl )]
(21)

where the subscripts l and g denote the liquid and the gas, respectively.
The governing equations of the mixture model is given by:
̅→ →
∇⋅(ϕg Vg + (1 − ϕg )Vl ) = 0 (22)

̅→
∇⋅(ϕg Vg ) = 0 (23)

for mass conservation of each phase and

̅→
ρ→
q ∇⋅(→
q ) = − ∇p + (ρ − ρl )→
g + Fμ (24)

for momentum conservation, where →

q is the mass mean velocity of the
mixture, defined as:
Fig. 6. Multiphysics in SOEC.
̅→ →
ρ→
q = ρg α Vg + ρl (1 − α)Vl (25)
3.1. Electrochemical process
As the boundary conditions, the Stefan velocity is defined at the
SOEC – The current density distribution in the electrolyzer is gov­ electrode where chemical reactions take place:
erned by the current conversion equation and Ohm’s Law: →
̅→ il Mg
→ Vg = (26)
∇⋅ il = 0 (14) nF ρg

→
il = − σmem ∇φl (15) where Mg is the molar mass of the gas.

→
where il is the current density vector; φl is the electric potential; σmem is 3.3. Mass transfer
the ionic conductivity of the membrane. Besides, the boundary condi­
tions at the electrode is given by Bulter-Volmer equation (3). SOEC – In SOEC, the concentration of water vapor and product de­
AE&PEM – Low-temperature water electrolysis is remarkably termines the equilibrium potential of the electrodes. Maxwell-Stefan
affected by the bubble effect. When the gas bubble is deformed at the equation is adapted to describe the transport of water vapor and
electrode surface, the effective area of the electrode decreases. There­ hydrogen:
fore, the Butler-Volmer equation in AE and PEM is modified as:
zFc ∑1 → →
RT i − c∇xi − xi ∇ϕ = (xj Ji − xi Jj ) (27)
η = sinh− 1 ( ) (16) RT j∕
=i
Dij
nF 2i0 (1 − ϕg )
where c is the total concentration of substances; xi and xj are molar
where ϕg represents the void fraction. Besides, in AE, the conductivity of
fractions of substance i and j, respectively; Dij is the diffusion coefficient;
the membrane is given by: → →
Ji and Jj are relative fluxes of substance i and j [39]. In the electrode
σ mem = σ KOH εmem (17) area, the source is defined by Faraday’s Law:

where εmem is the porosity of the diaphragm. The Bruggeman model is Ni =

i
(28)
integrated considering the bubble effect [36]. zi F
AE – The potassium hydroxide (KOH) solution is applied in AE. The
3.2. Momentum transfer concentration of the lye (pH) will influence the cell performance
through the conductivity of the electrolyte [40]:
SOEC – The momentum transfer in the electrolyzer is realized by
liquid-gas flow in low-temperature electrolyzers and gas-phase flow in
σKOH = K1 (100w) + K2 T + K3 T 2 + K4 (T⋅100w)
+ K5 (T 2 (100w)K6 ) + K7 (T/(100w)) (29)
SOEC. In SOEC, laminar flow model is adopted, governed by Navier-
+ K8 (100w/T)
Stokes equation:
→→ ̅→ ̅→ where w is the mass fraction of the solute, T the temperature,
∇∙(ρ V V ) = − ∇p + Fμ + ρ→
g + Fσ (18)
Ki (i = 1⋯8) the coefficients.
However, the pH of the lye will not influence the reversible voltage of

5
K. Hu et al. Applied Energy 312 (2022) 118788

the electrolyzer [35]. As the pH changes, the equilibrium potentials of

the anode and the cathode φa and φc are shifted synchronously ac­
cording to the Nernst Equations:
φa ≈ 1.23 − 0.06pH (30)

φc ≈ − 0.06pH (31)
Consequently, the reversible voltage is pH-independent according to
(6), while the ohmic overvoltage will be affected by pH of the lye.
PEM– The deionized water is applied in PEM electrolyzers in place of
the lye in AE. The dissolved hydrogen across the membrane is modeled
in PEM cells, which is one of the most concerning issue in PEM cells.
Maxwell-Stefan equation can be simplified into Nernst-Planck equation
in liquid phase [41]:
→ →
Ni = − Di ∇ci + ci V − zi um,i Fci ∇φl (32)
Fig. 7. Schematic diagram of the electrolyzer and the two-dimensional model
(red rectangle). (For interpretation of the references to colour in this figure
where mass transfer mechanisms including diffusion, convection and
legend, the reader is referred to the web version of this article.)
electro-osmotic drag are expressed orderly. In the electrode, the super­
saturation model is adopted in the model based on the theory of bubble
formation, where the hydrogen flux in the liquid phase is given by Table 1
[42,43]: Characteristic parameters of AE, PEM and SOEC ([44,45]).
i AE PEM SOEC
NH2 (l) = − kL (cH2 − SH2 (p − pH2 O )) (33)
2F Membrane thickness 2[mm] 250[μm] 10[μm]
Exchange current density of anode 0.048[A/m2] 0.1[A/m2] 0.01[A/m2]
where kL is transfer coefficient of hydrogen content. Exchange current density of cathode 1.9[A/m2] 10[A/m2] 0.1[A/m2]
Temperature 70[◦ C] 50[◦ C] 800[◦ C]
Concentration of the electrolyte (aq) 30[%]
3.4. Temperature dependence

As mentioned above, temperature rise could improve the cell per­ that in AE and PEM due to far higher working temperature. On the one
formance in two aspects. Firstly, less electric energy is required when the hand, high temperature results in the reduction of the reversible voltage
temperature rises and more thermal energy is supplied. From the according to (34). On the other hand, the electrodes are better activated
perspective of thermodynamics, the Gibbs free energy of the reaction is in SOEC according to (35).
reduced, so as to the reversible voltage. Besides, although the current density of all the electrolyzers tends to
SOEC – The temperature dependence of the equilibrium potential of decline along the channel, the mechanism is different. In AE and PEM,
the cathode and anode are given according to thermodynamics: the adhering bubbles on the electrode will reduce the effective area of
the electrode according to Fig. 9(a) and Fig. 9(b). As the bubbles accu­
ΔH − TΔS mulate near the outlet, the current density will decrease. On the other
φ= (34)
nF hand, the bubble effect in PEM is more prominent since the current
Although the changing of the equilibrium potential could be density of PEM is higher than that of AE, according to Fig. 8.
neglected in low-temperature electrolysis, a remarkable variation of the As for SOEC, the reduction of the water vapor and the increase of the
reversible potential occurs in SOEC when the temperature reaches products near the outlet of the electrolyzer lead to current density
higher than 700 ◦ C. The reversible potential of SOCE at 800 ◦ C drops to decline in Fig. 8(c). According to Nernst equation (1), this phenomenon
1.01 V, approximately 18% less than that at standard conditions. will lead to the elevation of the anodic potential and the descend of the
Meanwhile, the electrochemical process is better activated by cathodic potential. As a result, the cell potential increases and thus the
elevated temperature. The reaction rate, represented by exchange cur­ current density drops to a large extent. Fig. 9(c) and Fig. 9(d) display the
rent density, increases along with temperature based on Arrhenius concentration distribution of the water vapor and the cathodic
equation [37]: hydrogen.
Eexc 1 1
(35)
[− R (T − Tref )]
i0 = i0,ref e 4.2. I-V characteristics
PEM – the proton conductivity of the solid polymer electrolyte in
The I-V curves of three types of electrolyzers are acquired by
PEM is also temperature dependent [23]:
parameter scanning in COMSOL Multiphysics, as shown in Fig. 10. Given
the voltage of the electrolyzer, SOEC has the highest current density
Epro 1 1
(36)
[− R (Tref − T )]
σ PEM = σref e
among the three solutions. This is due to its high temperature. On the
According to the multiphysics introduced above, the models of the one hand, high temperature decreases the activation energy of the
three types of electrolyzers are built in COMSOL Multiphysics based on electrochemical reaction so that the hydrogen production rate rises. On
Fig. 7. Characteristic parameters for multiphysics simulation are shown the other hand, Gibbs free energy is reduced at a higher temperature,
in Table 1. leading to lower external energy needed. PEM cell takes second place in
terms of cell performance due to better ionic conductivity of the solid
4. Results and discussions polymer electrolyte compared with AE. The results meet well with the
industrial statistics collected in [23,46].
4.1. Field analysis
4.3. Efficiency comparison
The current distribution of the three types of electrolyzers under 1.6
V are presented in Fig. 8. The current density of SOEC is far higher than As discussed above, the efficiency of the electrolyzer is determined

6
K. Hu et al. Applied Energy 312 (2022) 118788

Fig. 8. Current density distribution (A/m2) of (a) AE, (b) PEM and (c) SOEC when.Ucell = 1.6V.

Fig. 9. Void fraction of AE (a) and PEM (b), Concentration distribution (mol/m3) of water vapor (c) and hydrogen (d) of SOEC when.Ucell = 1.6V.

by the amplitude of the supplied voltage. Therefore, the efficiency due to the introduction of the thermal energy absorbed in the numerator
curves of different electrolyzers can be derived from I-V characteristics yet the ignoration of thermal energy input in the denominator. There­
in Section 4-B, which are shown in Fig. 11, Fig. 12 and Fig. 13. fore, the definition of voltage efficiency is supposed to be avoided,
The voltage efficiency of the electrolyzers is given by (11). The re­ especially in SOEC where huge quantities of thermal energy are supplied
sults of the voltage efficiency based on LHV and HHV are shown in in place of electric energy.
Fig. 11 and Fig. 12, respectively. The simulation results are in accor­ If the definition of cell efficiency is adopted, given by (12), the results
dance with the statistical data in [47]. In both cases, SOEC possesses are shown in Fig. 13. Since both thermal and electrical energies are
absolute predominance over PEM and AE. This is due to its high tem­ taken into consideration in the denominator, the efficiency of SOEC is no
perature. This is due to high operating temperature reduces electric longer superior to PEM and AE. Moreover, the efficiency of the heat
energy demand as well as activates the electrochemical process. In management of SOEC is discussed. If the system were absolutely
electrolysis near ambient temperature, PEM cells beat AE for less ohmic isothermal and capable of 100% waste heat recovery, the efficiency of
losses and higher conductivity of the membrane. However, it is obvious the heat management should be defined as 100%. Unfortunately, SOEC
that the efficiency exceeds 100% based on either LHV or HHV. This is would show no advantage over PEM even if the ideal heat management

7
K. Hu et al. Applied Energy 312 (2022) 118788

Fig. 10. The I-V characteristics of three types of electrolyzers. Fig. 13. Cell efficiency comparison considering thermal efficiency (Eq. (12)).

4.4. Researches on multiphysics constraints

AE—The concentration and the temperature of the KOH solution are

investigated in AE, as shown in Fig. 14. When the mass fraction of the
KOH is lower than 30%, the current density rises with the increase of the
mass fraction. This is because a higher concentration of the ions pro­
motes the conductivity of the lye. Therefore, the resistance of the elec­
trolyzer decreases thus resulting in current density elevation. However,
the current density will decline when the mass fraction exceeds 30%.
This is due to the lower dissociation degree and slower movement of the
ions. Therefore, the KOH solution with a 30% mass fraction is most
commonly used in modern alkaline water electrolyzers.
Contrarily, the temperature dependence of the current density is
monotonous, as shown in Fig. 14. The current density increases along
with the temperature rise due to the higher conductivity of the elec­
Fig. 11. Voltage efficiency comparison based on LHV (Eq. (11)). trolyte and better-activated electrodes. However, in order to avoid
vaporization of the water, the temperature of AE is usually below 90 ◦ C.
The main operating parameters of AE are analyzed as follows. The
limitation for the hydrogen production rate of AE, proportional to cur­
rent density, is due to the poor properties of I-V characteristics. The
employment of liquid electrolyte, as well as the insulating diaphragm,
results in far higher resistivity and thickness of membrane than PEM and
SOEC. Up to 2.2 V is required for AE to reach 1000 A/m2 when the ef­
ficiency of the cell already drops to 60% [49].
Besides, there exists a lower limit for the current density of AE, which
represents the capability of part-load operation. Although the
decreasing of the hydrogen load contributes to a lower voltage and thus
benefits the cell efficiency according to (12), it also brings problems
including lower Faraday efficiency and safety issues. This is due to the
fact that the oxygen production rate at the anode is decreased more than
the hydrogen crossover through the membrane. When the load is cut

Fig. 12. Voltage efficiency comparison based on HHV (Eq. (11)).

were achieved. When the efficiency of the heat management drops to

60% in operation, PEM beats SOEC in terms of energy conversion effi­
ciency. Therefore, it can be concluded that SOEC possesses no advantage
over low-temperature electrolyzers in the aspect of energy conversion
efficiency, and might even be inferior on condition of poor heat
management.
Nevertheless, since the total electric energy needed for a certain
amount of hydrogen production in SOEC is less than that in low-
temperature water electrolysis, SOEC is supposed to be economically
efficient when cheap or even free heat sources are integrated, like the
nuclear power plant or the gas turbine. Besides, the reverse process of
water electrolysis occurring in SOFC is strongly exothermic so that it can
potentially provide sufficient thermal energy for SOEC. This makes the
development of the reversible solid oxide cell promising [48]. Fig. 14. The concentration of KOH(aq) and temperature dependence of AE
(Ucell = 2V).

8
K. Hu et al. Applied Energy 312 (2022) 118788

down to 40%, the contamination of hydrogen at the anode might exceed

the lower explosive limit (LEL). On the other hand, the decline of the
hydrogen crossover rate cannot catch up with that of the hydrogen
production rate at the cathode. As a result, the Faraday efficiency of the
cell declines. It has already been pointed out that the Faraday efficiency
drops drastically when the current density is as low as 0.1 A/cm2 [50].
PEM—The upper bound of the current density in PEM is 2 A/cm2,
quite close to the maximum tolerant current density of Nafion equiva­
lent to 2.5 A/cm2. Apart from that, the efficiency of PEM would drop to
60% if the current density exceeds 2 A/cm2, according to the simulation
results. In other words, the current density is also dragged by the sup­
plied voltage, the same as the situation in AE.
Different from the poor selectivity of the diaphragm in AE, the proton
exchange membrane in the PEM cell features the excellent segregation
of anode and cathode. Hence, pressurized water electrolysis is prevalent Fig. 16. Temperature dependence of SOEC (Ucell = 1V).
in PEM water-splitting technology in order to cut down the energy
consumption of the downstream compressors, since hydrogen is, in most Faraday efficiency. The range of current density of SOEC is 0.3–1.0 A/
cases, pressurized before storage and transportation to guarantee suffi­ cm2. It is obvious that the extreme current density of SOEC cannot reach
cient energy density. Since excessive pressure promotes the hydrogen the level of PEM cells. This is due to the fact that overrated current
crossover through the membrane, hydrogen crossover tends to occur in density would cause a sharp potential gradient near boundary layers
pressurized PEM cells, resulting in reduced Faraday efficiency and between the anode and solid polymer electrolyte, leading to drastic
higher proportion of hydrogen at the anode in part-load operation. degradation of the adjacent materials. As a matter of fact, SOEC is
Based on the simulation results of hydrogen crossover in PEM (shown in supposed to work with thermoneutral voltage when the thermal stress
Fig. 15), the current density of PEM cells is supposed to be maintained caused by temperature gradient is eliminated, thus extending the life­
above 0.8 A/cm2 under 20 bar to guarantee a secure ratio of anodic time of cell materials [52]. As for the minimum load rate of SOEC
hydrogen. Obviously, further increase of the lower bound of the current research, DONITZ, et.al. have reported their experiments where 0.3 A/
density can be inferred on condition of higher operating pressure. Extra cm2 is achieved with the supplied voltage was as low as 1.07 V [53]. It
investment in safety is also necessary for pressurized operations. has also been said that the Faraday efficiency of the test cell has reached
Therefore, the pressure is supposed to be confined below 50 bar in 100%.
current technology [42]. Table 2 shows the operating parameters of the three types of elec­
SOEC—SOEC features relatively high temperatures. The temperature trolyzers. The reason why the energy conversion efficiency of SOEC is
has great influences on cell performances, as shown in Fig. 16. As the the highest is that the cell voltage is far lower than those of AE and PEM.
temperature rises, the reversible voltage drops due to higher thermal In other words, the high efficiency of SOEC is at the expenses of the
energy input according to Fig. 2. Therefore, the current density under hydrogen production rate, as is shown in the table. Therefore, it can be
Ucell = 1V increases with higher temperature. This is the main reason predicted that the energy conversion efficiency of SOEC would drop
why SOEC outperforms AE and PEM in terms of current density and around the level of PEM when voltage elevation is achieved, as is shown
power-to-hydrogen efficiency. in Fig. 13. However, SOEC would still beats the other two technologies
However, the high working temperature in SOEC is a two-fold impact in terms of power-to-hydrogen efficiency, making it promising to inte­
on cell performances. On the one hand, both better-activated catalysts grate SOEC with infrastructures containing cheap thermal sources, like
and sufficient thermal energy are fulfilled with the introduction of high nuclear power plants.
temperature so that the I-V characteristics show an absolute advantage
over PEM and AE. On the other hand, the efforts for cell performance 5. Conclusions
improvement are constrained by exceeding temperature. Component
alternation and structural failure of the materials under extreme work­ In this paper, a unified multiphysics model for the AE, PEM and
ing conditions are the dominant factors of SOEC degradation. Currently, SOEC is derived to bridge the gap among different PtH researches, based
the operating temperature of SOEC could be as high as 1000 ◦ C [51]. The on which their differences in terms of I-V characteristics, cell efficiency
excessive temperature would create a drastic decline in the material and part-load performances are explained from multiphysics theories.
lifetime, as well as aggravate hydrogen crossover rate, thus reducing Different definitions of efficiency are collected and evaluated from the
perspective of thermodynamics and electrode kinetics. Conclusions are
that the voltage efficiency, which is widely adopted in current research,
is proved to be invalid since both electric and thermal energies are

Table 2
Operating parameters of AE, PEM and SOEC ([21,23,44,45,51,53]).
AE PEM SOEC

Current density (A/cm2) 0.25–0.45 1.0–2.0 0.3–1.0

Cell voltage (V) 1.8–2.5 1.8–2.2 0.9–1.2
Operating pressure (bar) 10–30 20–50 1–15
Lower partial load range (%) 20–40 0–10 –
Power-to-hydrogen efficiency (%) ( 50–60 65–70 90–95
Fig. 11)
Energy conversion efficiency (%) ( 50–65 65–75 75–81(εth =
Fig. 13) 100%)
65–70(εth =
60%)
Operating temperature (◦ C) 60–90 50–80 700–1000
Fig. 15. Simulation of the anodic hydrogen content of PEM cells.

9
K. Hu et al. Applied Energy 312 (2022) 118788

necessary for water electrolysis to take place. For the multiphysics Trans. Sustain. Energy 2015;6(4):1234–44. https://doi.org/10.1109/
TSTE.2015.2424885.
modeling of AE, PEM and SOEC, the uniformity and the differences of
[7] Bassano C, Deiana P, Lietti L, Visconti CG. P2G movable modular plant operation
the models are analyzed. In PEM and AE, the diphasic flow field is on synthetic methane production from CO2 and hydrogen from renewables
considered since the bubble coverage effect reduces the effective elec­ sources. Fuel 2019;253(May):1071–9. https://doi.org/10.1016/j.
trode area as well as elevates the electrolyte resistance. In SOEC, the fuel.2019.05.074.
[8] Padhy S, Panda S. Simplified grey wolf optimisation algorithm tuned adaptive
liquid-gas two-phase flow disappears while the concentration activation fuzzy PID controller for frequency regulation of interconnected power systems. Int
is included in the model with Maxwell-Stefan equation. J Ambient Energy 2021;4. https://doi.org/10.1080/01430750.2021.1874518.
The result shows that the current density of SOEC reaches the top [9] Badal FR, Das P, Sarker SK, Das SK. A survey on control issues in renewable energy
integration and microgrid. Prot Control Mod Power Syst 2019;4(1). https://doi.
since high temperature reduces the electric demand and activates the org/10.1186/s41601-019-0122-8.
electrodes. PEM comes second for the excellent performance of the solid [10] Kreuter W, Hofmann H. Electrolysis: the important energy transformer in a world
polymer electrolyte. Next, the efficiency of the three types of electro­ of sustainable energy. Int J Hydrogen Energy 1998;23(8):661–6. https://doi.org/
10.1016/S0360-3199(97)00109-2.
lyzers is compared based on different efficiency definitions. The result [11] Dou X, Zhao W, Lang Y, Li Y, Gao C. A Review of Operation of Natural Gas-
shows SOEC possesses less advantage over PEM cells in terms of energy Electricity Coupling System Considering Power-to-Gas Technology. Dianwang
conversion efficiency, and might even be inferior on condition of poor Jishu/Power Syst. Technol. 2019;43(1):165–73. https://doi.org/10.13335/j.1000-
3673.pst.2017.2964.
heat management. Notwithstanding, we can still speculate that SOEC is [12] Erdiwansyah, Mahidin, Husin H, Nasaruddin, Zaki M, Muhibbuddin. A critical
outperforming the others in terms of cost-efficiency if cheap or even free review of the integration of renewable energy sources with various technologies.
heat sources are integrated. This makes the development of reversible Prot Control Mod Power Syst 2021;6(1). https://doi.org/10.1186/s41601-021-
00181-3.
solid oxide electrolyzers and the integration of SOEC with calorific de­
[13] Balat M. Potential importance of hydrogen as a future solution to environmental
vices promising. and transportation problems. Int J Hydrogen Energy 2008;33(15):4013–29.
Finally, the industrial operating parameters of AE, PEM and SOEC https://doi.org/10.1016/j.ijhydene.2008.05.047.
are analyzed considering multiphysics constraints. For AE, the promo­ [14] Siracusano S, Hodnik N, Jovanovic P, Ruiz-Zepeda F, Šala M, Baglio V, et al. New
insights into the stability of a high performance nanostructured catalyst for
tion of cell performances is limited by cell resistance where the lye acts sustainable water electrolysis. Nano Energy 2017;40:618–32. https://doi.org/
as the electrolyte. For PEM cells, the tolerant current density of the 10.1016/j.nanoen.2017.09.014.
Nafion membrane is the main drawback for current density elevation. [15] Liu H-b, Xu H, Pan L-M, Zhong D-H, Liu Y. Porous electrode improving energy
efficiency under electrode-normal magnetic field in water electrolysis. Int J
Moreover, both types of technology face the challenge of excessive Hydrogen Energy 2019;44(41):22780–6. https://doi.org/10.1016/j.
hydrogen crossover in part-load operation, which could be eliminated ijhydene.2019.07.024.
by improving the selectivity of the membrane. For SOEC, high temper­ [16] Guilbert D, Collura SM, Scipioni A. DC/DC converter topologies for electrolyzers:
State-of-the-art and remaining key issues. Int J Hydrogen Energy 2017;42(38):
ature provides more thermal energy as well as activates the electrodes 23966–85. https://doi.org/10.1016/j.ijhydene.2017.07.174.
more effectively, but, at the same time, becomes the bottleneck of the [17] Yodwong B, Guilbert D, Phattanasak M, Kaewmanee W, Hinaje M, Vitale G. AC-DC
technology due to degradation issues. Converters for Electrolyzer Applications: State of the Art and Future Challenges.
Electronics 2020;9(6):912. https://doi.org/10.3390/electronics9060912.
[18] Koponen J, Poluektov A, Ruuskanen V, Kosonen A, Niemelä M, Ahola J.
CRediT authorship contribution statement Comparison of thyristor and insulated-gate bipolar transistor -based power supply
topologies in industrial water electrolysis applications. J Power Sources 2021;491:
229443. https://doi.org/10.1016/j.jpowsour.2020.229443.
Kewei Hu: Conceptualization, Writing – original draft, Methodol­ [19] Xing X, Lin J, Song Y, Song J, Mu S. Intermodule Management within a Large-
ogy, Software, Investigation, Writing – review & editing. Jiakun Fang: Capacity High-Temperature Power-to-Hydrogen Plant. IEEE Trans Energy Convers
2020;35(3):1432–42. https://doi.org/10.1109/TEC.2020.2978552.
Methodology, Writing – review & editing, Supervision. Xiaomeng Ai:
[20] Ozturk M, Dincer I. Development of renewable energy system integrated with
Methodology, Writing – review & editing, Supervision. Danji Huang: hydrogen and natural gas subsystems for cleaner combustion. J Nat Gas Sci Eng
Conceptualization, Software. Zhiyao Zhong: Conceptualization, Inves­ 2020;83:103583. https://doi.org/10.1016/j.jngse.2020.103583.
tigation. Xiaobo Yang: Writing – review & editing. Lei Wang: Writing – [21] Buttler A, Spliethoff H. Current status of water electrolysis for energy storage, grid
balancing and sector coupling via power-to-gas and power-to-liquids: a review.
review & editing. Renew Sustain Energy Rev 2018;82(February 2017):2440–54. https://doi.org/
10.1016/j.rser.2017.09.003.
[22] Ursúa A, Gandía LM, Sanchis P. Hydrogen production from water electrolysis:
Declaration of Competing Interest Current status and future trends. Proc IEEE 2012;100(2):410–26. https://doi.org/
10.1109/JPROC.2011.2156750.
The authors declare that they have no known competing financial [23] Carmo M, Fritz DL, Mergel J, Stolten D. A comprehensive review on PEM water
electrolysis. Int J Hydrogen Energy 2013;38(12):4901–34. https://doi.org/
interests or personal relationships that could have appeared to influence 10.1016/j.ijhydene.2013.01.151.
the work reported in this paper. The paper is funded by (1) National [24] Zeng K, Zhang D. Recent progress in alkaline water electrolysis for hydrogen
Natural Science Foundation of China, Grant ID: 52177089; (2) ABB production and applications. Prog Energy Combust Sci 2010;36(3):307–26.
https://doi.org/10.1016/j.pecs.2009.11.002.
Power Grids Research, Contract No. ABB20171127REU-CTR. [25] Bi L, Boulfrad S, Traversa E. Steam electrolysis by solid oxide electrolysis cells
(SOECs) with proton-conducting oxides. Chem Soc Rev 2014;43(24):8255–70.
References https://doi.org/10.1039/c4cs00194j.
[26] Celik AN. Three-dimensional multiphysics model of a planar solid oxide fuel cell
using computational fluid dynamics approach. Int J Hydrogen Energy 2018;43
[1] Gyawali N, Ohsawa Y. Integrating fuel cell/electrolyzer/ultracapacitor system into
(42):19730–48. https://doi.org/10.1016/j.ijhydene.2018.08.212.
a stand-alone microhydro plant. IEEE Trans Energy Convers 2010;25(4):1092–101.
[27] Wang Q, Li Y, Meng T, Huang B, Hu L, Su H, et al. Engineering Heterostructure-
https://doi.org/10.1109/TEC.2010.2066977.
Incorporated Metal Silicates Anchored on Carbon Nanotubes for Highly Durable
[2] Gahleitner G. Hydrogen from renewable electricity: an international review of
Lithium Storage. ACS Appl Energy Mater 2021;4(2):1548–59. https://doi.org/
power-to-gas pilot plants for stationary applications. Int J Hydrogen Energy 2013;
10.1021/acsaem.0c0273810.1021/acsaem.0c02738.s001.
38(5):2039–61. https://doi.org/10.1016/j.ijhydene.2012.12.010.
[28] Wang Q, Yang H, Meng T, Yang J, Huang B, Gu FL, et al. Boosting Electron Transfer
[3] Götz M, Lefebvre J, Mörs F, McDaniel Koch A, Graf F, Bajohr S, et al. Renewable
with Heterointerface Effect for High-Performance Lithium-Ion Storage. Energy
Power-to-Gas: a technological and economic review. Renew. Energy 2016;85:
Storage Mater 2021;36:365–75. https://doi.org/10.1016/j.ensm.2021.01.003.
1371–90. https://doi.org/10.1016/j.renene.2015.07.066.
[29] Wang Q, Meng T, Li Y, Yang J, Huang B, Ou S, et al. Consecutive chemical bonds
[4] Schiebahn S, Grube T, Robinius M, Tietze V, Kumar B, Stolten D. Power to gas:
reconstructing surface structure of silicon anode for high-performance lithium-ion
Technological overview, systems analysis and economic assessment for a case
battery. Energy Storage Mater 2021;39:354–64. https://doi.org/10.1016/j.
study in Germany. Int J Hydrogen Energy 2015;40(12):4285–94. https://doi.org/
ensm.2021.04.043.
10.1016/j.ijhydene.2015.01.123.
[30] Grigoriev SA, Kalinnikov AA, Millet P, Porembsky VI, Fateev VN. Mathematical
[5] Al-Badi AH, Yousef H, Alaamri O, Alabdusalam M, Alshidi Y, Alharthy N.
modeling of high-pressure PEM water electrolysis. J Appl Electrochem 2010;40(5):
Performance of a stand-alone renewable energy system based on hydrogen energy
921–32. https://doi.org/10.1007/s10800-009-0031-z.
storage. IEEE Trans Energy Convers 2014;19(3):356–9. https://doi.org/10.1109/
[31] García-Valverde R, Espinosa N, Urbina A. Simple PEM water electrolyser model
ISCCSP.2014.6877887.
and experimental validation. Int J Hydrogen Energy 2012;37(2):1927–38. https://
[6] Clegg S, Mancarella P. Integrated Modeling and Assessment of the Operational
doi.org/10.1016/j.ijhydene.2011.09.027.
Impact of Power-to-Gas (P2G) on Electrical and Gas Transmission Networks. IEEE

10
K. Hu et al. Applied Energy 312 (2022) 118788

[32] Choi P, Bessarabov DG, Datta R. A simple model for solid polymer electrolyte (SPE) [44] Shiva Kumar S, Himabindu V. Hydrogen production by PEM water electrolysis – A
water electrolysis. Solid State Ionics 2004;175(1–4):535–9. https://doi.org/ review. Mater Sci Energy Technol 2019;2(3):442–54. https://doi.org/10.1016/j.
10.1016/j.ssi.2004.01.076. mset.2019.03.002.
[33] Ulleberg Ø. Modeling of advanced alkaline electrolyzers a system. Hydrog. Energy [45] Rashid MM, Al Mesfer MK, Naseem H, Danish M. Hydrogen Production by Water
2003;28:21–33. Electrolysis: A Review of Alkaline Water Electrolysis, PEM Water Electrolysis and
[34] Zeng S, Zhang X, Song Chen J, Li T, Andersson M. Modeling of solid oxide fuel cells High Temperature Water Electrolysis. Int J Eng Adv Technol, 2015;3:2249–8958.
with optimized interconnect designs. Int J Heat Mass Transf 2018;125:506–14. [46] Zhang W, Yu B. Development Status and Prospects of Hydrogen Production by
https://doi.org/10.1016/j.ijheatmasstransfer.2018.04.096. High Temperature Solid Oxide Electrolysis. J Electrochem 2020;26(2):212–29.
[35] Godula-Jopek A. Hydrogen production: by electrolysis. John Wiley & Sons; 2015. [47] Buttler A, Spliethoff H. Current status of water electrolysis for energy storage, grid
[36] Akita K, Yoshida F. Bubble Size, Interfacial Area, and Liquid-Phase Mass Transfer balancing and sector coupling via power-to-gas and power-to-liquids: A review.
Coefficient in Bubble Columns. Ind Eng Chem Process Des Dev 1974;13(1):84–91. Renew Sustain Energy Rev 2018;82(February):2440–54. https://doi.org/10.1016/
https://doi.org/10.1021/i260049a016. j.rser.2017.09.003.
[37] Aubras F, Deseure J, Kadjo J-J-A, Dedigama I, Majasan J, Grondin-Perez B, et al. [48] Srikanth S, Heddrich MP, Gupta S, Friedrich KA. Transient reversible solid oxide
Two-dimensional model of low-pressure PEM electrolyser: Two-phase flow regime, cell reactor operation – Experimentally validated modeling and analysis. Appl
electrochemical modelling and experimental validation. Int J Hydrogen Energy Energy 2018;232(September):473–88. https://doi.org/10.1016/j.
2017;42(42):26203–16. https://doi.org/10.1016/j.ijhydene.2017.08.211. apenergy.2018.09.186.
[38] Wedin R, Dahlkild AA. On the transport of small bubbles under developing channel [49] Ogawa T, Takeuchi M, Kajikawa Y. Analysis of Trends and Emerging Technologies
flow in a buoyant gas-evolving electrochemical cell. Ind Eng Chem Res 2001;40 in Water Electrolysis Research Based on a Computational Method: A Comparison
(23):5228–33. https://doi.org/10.1021/ie001073u. with Fuel Cell Research. Sustainability 2018;10(2):478. https://doi.org/10.3390/
[39] Tseronis K, Kookos IK, Theodoropoulos C. Modelling mass transport in solid oxide su10020478.
fuel cell anodes: a case for a multidimensional dusty gas-based model. Chem Eng [50] Buttler A, Spliethoff H. Current status of water electrolysis for energy storage, grid
Sci 2008;63(23):5626–38. https://doi.org/10.1016/j.ces.2008.07.037. balancing and sector coupling via power-to-gas and power-to-liquids: a review.
[40] See DM, White RE. Temperature and concentration dependence of the specific Renew Sustain Energy Rev 2018;82(September 2017):2440–54. https://doi.org/
conductivity of concentrated solutions of potassium hydroxide. J Chem Eng Data 10.1016/j.rser.2017.09.003.
1997;42(6):1266–8. https://doi.org/10.1021/je970140x. [51] Laguna-Bercero MA. Recent advances in high temperature electrolysis using solid
[41] Wang S, Worek WM, Minkowycz WJ. Performance comparison of the mass transfer oxide fuel cells: A review. J Power Sources 2012;203:4–16. https://doi.org/
models with internal reforming for solid oxide fuel cell anodes. Int J Heat Mass 10.1016/j.jpowsour.2011.12.019.
Transf 2012;55(15–16):3933–45. https://doi.org/10.1016/j. [52] Schefold J, Brisse A, Poepke H. Long-term Steam Electrolysis with Electrolyte-
ijheatmasstransfer.2012.03.024. Supported Solid Oxide Cells. Electrochim Acta 2015;179:161–8. https://doi.org/
[42] Trinke P, Haug P, Brauns J, Bensmann B, Hanke-Rauschenbach R, Turek T. 10.1016/j.electacta.2015.04.141.
Hydrogen Crossover in PEM and Alkaline Water Electrolysis: Mechanisms, Direct [53] Dönitz W, Erdle E. High-temperature electrolysis of water vapor-status of
Comparison and Mitigation Strategies. J Electrochem Soc 2018;165(7):F502–13. development and perspectives for application. Int J Hydrogen Energy 1985;10(5):
https://doi.org/10.1149/2.0541807jes. 291–5. https://doi.org/10.1016/0360-3199(85)90181-8.
[43] Trinke P, Bensmann B, Hanke-Rauschenbach R. Current density effect on hydrogen
permeation in PEM water electrolyzers. Int J Hydrogen Energy 2017;42(21):
14355–66. https://doi.org/10.1016/j.ijhydene.2017.03.231.

11