International Journal of Hydrogen Energy 28 (2003) 2133

www.elsevier.com/locate/ijhydene
Modeling of advanced alkaline electrolyzers: a system
simulation approach
Iystein Ulleberg
1
Institute for Energy Technology, P.O. Box 40, N-2027 Kjeller, Norway
Abstract
A mathematical model for an advanced alkaline electrolyzer has been developed. The model is based on a combination
of fundamental thermodynamics, heat transfer theory, and empirical electrochemical relationships. A dynamic thermal model
has also been developed. Comparisons between predicted and measured data show that the model can be used to predict the
cell voltage, hydrogen production, eciencies, and operating temperature. The reference system used was the stand-alone
photovoltaic-hydrogen energy plant in J ulich. The number of required parameters has been reduced to a minimum to make
the model suitable for use in integrated hydrogen energy system simulations. The model has been made compatible to a
transient system simulation program, which makes it possible to integrate hydrogen energy component models with a standard
library of thermal and electrical renewable energy components. Hence, the model can be particularly useful for (1) system
design (or redesign) and (2) optimization of control strategies. To illustrate the applicability of the model, a 1-year simulation
of a photovoltaic-hydrogen system was performed. The results show that improved electrolyzer operating strategies can be
identied with the developed system simulation model. ? 2002 International Association for Hydrogen Energy. Published by
Elsevier Science Ltd. All rights reserved.
Keywords: Alkaline electrolyzer; Hydrogen systems; Stand-alone power; Renewable energy; Modeling; System simulation
1. Introduction
1.1. Background
Hydrogen is often referred to as the energy carrier of the
future because it can be used to store intermittent renewable
energy (RE) sources such as solar and wind energy. The idea
of creating sustainable energy systems lead over the past
decade to several hydrogen energy demonstration projects
around the world [1]. The main objectives of these hydrogen
projects was to test and develop components, demonstrate
technology, and perform system studies on two categories of
systems: (1) stand-alone power systems and (2) hydrogen
refueling stations. In the latter category, the most notable
1
Temporary address until 31.12.2002: Murdoch University,
South Street, WA, Perth 6150, Australia.
E-mail address: [email protected] (I. Ulleberg).
project is the hydrogen refueling station at Munich Airport
[2]. Most of the previous RE=H
2
-projects have been based
on solar energy from photovoltaics (PV). However, lately
also wind energy conversion systems (WECS) have been
considered to be a possible power source, particularly for
weak-grid applications.
In all of the cases mentioned above the electrolyzer is a
crucial component, and the technical challenge is to make
it operate smoothly with intermittent power from renewable
energy sources. Up until now most of the R&D on water
electrolysis related to RE=H
2
-projects have focused on alka-
line systems, although there have been some major research
eorts on proton exchange membrane (PEM) electrolyzers
as well, particularly within the Japanese WE-NET program
[3]. However, the costs associated with PEM-electrolysis are
still too high, and the market for small-scale H
2
-production
units is at present day still relatively small.
Institute for Energy Technology (IFE) has since the early
1990s been carrying out theoretical and practical research
0360-3199/02/$ 22.00 ? 2002 International Association for Hydrogen Energy. Published by Elsevier Science Ltd. All rights reserved.
PII: S 0360- 3199( 02) 00033- 2
22 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
Nomenclature
Acronyms
AC alternating current
DC direct current
EES Engineering Equation Solver
FZJ ForschungsZentrum J ulich
HYSOLAR HYdrogen SOLAR
IEA International Energy Agency
IFE Institute for Energy Technology
JANAF Joint Army-Navy-Air Force (database
for thermochemical properties)
KOH potassium hydroxide
MPPT maximum power point tracker
PEM proton exchange membrane
PHOEBUS PHOtovoltaik-Elektrolyse-Brennstozelle
Und Systemtecknik
PV photovoltaic
R&D research and development
RE renewable energy
RMS root mean square
SAPS stand-alone power system
SIMELINT SIMulation of Electrolyzers in INTer-
mittent operation
TRNSYS TRaNsient SYstem Simulation program
WECS wind energy conversion system
WE-NET World Energy-Network
Symbols
A area of electrode, m
2
aq water based solution
C
cw
thermal capacity of cooling water, J K
1
C
t
overall thermal capacity of electrolyzer,
J K
1
emf electromotive force, V
f
1
parameter related to Faraday eciency,
mA
2
cm
4
f
2
parameter related to Faraday eciency
g gas
h
cond
parameter related to conduction heat
transfer, W K
1
h
conv
parameter related to convection heat
transfer, W K
1
A
1
I current, A
l liquid
LMTD log mean temperature dierence,

C
n
c
number of cells in series per stack
p pressure, bar
r parameter related to ohmic resistance of
electrolyte, m
2
R
t
overall thermal resistance of electrolyzer,
W
1
K
s coecient for overvoltage on electrodes, V
t coecient for overvoltage on electrodes,
A
1
m
2
SOC state of charge (battery), 0 : : : 1
T temperature, K or

C
U voltage, V
UA
HX
overall heat transfer coecient-area prod-
uct for heat exchanger, W
1
K
G change in Gibbs energy, J mol
1
H change in enthalpy, J mol
1
S change in entropy, J K
1
mol
1
n molar ow rate, mol=s

Q heat transfer rate, W

t time interval, s
Subscripts
a ambient
cool cooling (auxiliary)
cw cooling water
gen generated
H
2
pure hydrogen
H
2
O pure water
i, o inlet, outlet
ini initial
loss loss to ambient
O
2
pure oxygen
rev reversible
Constants
F 96 485 C mol
1
or As mol
1
Faraday con-
stant
z 2 number of electrons transferred per reac-
tion
R 8:315 J K
1
mol
1
universal gas constant
v
std
0:0224136 m
3
mol
1
volume of an ideal
gas at standard conditions
in the area of stand-alone power systems (SAPS) based
on RE sources and H
2
-technology [46], and joined in
1999 the IEA Hydrogen Program Annex 13 [7]. The elec-
trolyzer modeling eorts performed in this context focused
on alkaline electrolysis, as this was the technology chosen
for the relevant applications. It is this modeling eort that
is being reported in this paper. However, it should be noted
that IFE is currently in the process of acquiring a small-scale
PEM-electrolyzer unit for testing in a laboratory setup,
which hopefully will give valuable system performance data
over the next 23 years. The theory and modeling philoso-
phy presented here could be applied to the PEM-technology.
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 23
Fig. 1. Principle of a monopolar electrolyzer design.
Over the past decade there has been an increasing interest
in system analysis of integrated RE=H
2
-systems, especially
among those energy and utility companies that are trying
to position themselves in the future markets of distributed
power generation and alternative fuels. Therefore, there is a
need for an electrolyzer model that is suitable for dynamic
simulation of such systems. There is also a need for a model
with a relatively high level of detail. One particularly im-
portant requirement is that the technical model can be cou-
pled to economic models that account for both investment
and operational costs.
1.2. Technology
The electrolyte used in the conventional alkaline water
electrolyzers has traditionally been aqueous potassium
hydroxide (KOH), mostly with solutions of 2030 wt% be-
cause of the optimal conductivity and remarkable corrosion
resistance of stainless steel in this concentration range [8].
The typical operating temperatures and pressures of these
electrolyzers are 70100

C and 130 bar, respectively.

Physically an electrolyzer stack consists of several cells
linked in series. Two distinct cell designs does exist:
monopolar and bipolar [9]. In the monopolar design the
electrodes are either negative or positive with parallel elec-
trical connection of the individual cells (Fig. 1), while in
the bipolar design the individual cells are linked in series
electrically and geometrically (Fig. 2). One advantage of
the bipolar electrolyzer stacks is that they are more compact
than monopolar systems. The advantage of the compactness
of the bipolar cell design is that it gives shorter current
paths in the electrical wires and electrodes. This reduces the
losses due to internal ohmic resistance of the electrolyte,
and therefore increases the electrolyzer eciency. How-
ever, there are also some disadvantages with bipolar cells.
One example is the parasitic currents that can cause cor-
rosion problems. Furthermore, the compactness and high
Fig. 2. Principle of a bipolar electrolyzer design.
pressures of the bipolar electrolyzers require relatively so-
phisticated and complex system designs, and consequently
increases the manufacturing costs. The relatively simple and
sturdy monopolar electrolyzers systems are in comparison
less costly to manufacture. Nevertheless, most commercial
alkaline electrolyzers manufactured today are bipolar.
In new advanced alkaline electrolyzers the operational
cell voltage has been reduced and the current density in-
creased compared to the more conventional electrolyzers.
Reducing the cell voltage reduces the unit cost of electrical
power and thereby the operation costs, while increasing the
current density reduces the investment costs [8]. However,
there is a conict of interest here because the ohmic resis-
tance in the electrolyte increases with increasing current due
to increasing gas bubbling. Increased current densities also
lead to increased overpotentials at the anodes and cathodes.
Three basic improvements can be implemented in the de-
sign of advanced alkaline electrolyzers: (1) new cell con-
gurations to reduce the surface-specic cell resistance de-
spite increased current densities (e.g., zero-gap cells and
low-resistance diaphragms), (2) higher process tempera-
tures (up to 160

C) to reduce the electric cell resistance in

order to increase the electric conductivity of the electrolyte,
and (3) new electrocatalysts to reduce anodic and cathodic
overpotentials (e.g., mixed-metal coating containing cobalt
oxide at anode and Raney-nickel coatings at cathode). In
the zero-gap cell design the electrode materials are pressed
on either side of the diaphragm so that the hydrogen and
oxygen gases are forced to leave the electrodes at the rear.
Most manufacturers have adopted this design [9].
1.3. Modeling
Most of the relevant electrolyzer modeling found in the
literature is related to solar-hydrogen demonstration projects
from the past decade. The most detailed model to date
is probably the SIMELINT-program, developed as part of
the Saudi ArabianGerman HYSOLAR-project [10]. This
24 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
program, which was validated against measured data, accu-
rately predicts the thermal behavior, cell voltage, gas puri-
ties, and eciencies for any given power or current prole.
Other empirical models have also been developed [1115],
but these have either been less detailed or not tested and
veried against experimental data.
The objective of the work described in this paper has
been to develop a model that accurately predicts the elec-
trochemical and thermal dynamic behavior of an advanced
alkaline electrolyzer. The model is primarily intended for
use in integrated renewable energy systems simulations stud-
ies that comprise subsystems such as PV-arrays, WECS,
electrolyzers, fuel cells, and hydrogen storage. A few key
requirements were placed upon the model; it needed to be
numerically robust, versatile and practical to use. Hence, the
model needed to be a trade-o between simple and com-
plex modeling. For instance, empirical relations are used to
model the most complex electrochemical processes. At the
same time, a signicant eort has been made to minimize
the number of required parameters required by the empirical
relations. In order to make the model as generic as possi-
ble, fundamental thermodynamics and heat transfer theory
is used where appropriate.
The electrolyzer model presented is written as a FOR-
TRAN subroutine primarily designed to run with the simu-
lation programs TRNSYS and EES, but the model has also
been designed so that it readily can be integrated into other
simulation programs (e.g., MATLAB
J
Simulink
J
). TRN-
SYS is a transient systems simulation program with a mod-
ular structure [16]. The TRNSYS library includes many of
the components commonly found in thermal and electrical
renewable energy systems, as well as component routines to
handle input of weather data or other time-dependent forcing
functions. The modular structure of TRNSYS gives the pro-
gram the desired exibility, as it facilitates for the addition
of mathematical models not included in the standard library.
The program is well suited to perform detailed analyses of
systems whose behavior is dependent on the passage of time.
EES, an engineering equation solver, has built-in functions
for thermodynamic and transport properties of many sub-
stances, including steam, air, refrigerants, cryogenic uids,
JANAF table gases, hydrocarbons and psychrometrics [17].
Additional property data can be added, and the program
allows user-written functions, procedures, modules, and tab-
ular data. In this study EES was used to perform parameter
sensitivity analyses and to test and verify the model against
measured data, while TRNSYS was used to perform inte-
grated system simulations.
2. Model description
The decomposition of water into hydrogen and oxygen
can be achieved by passing an electric current (DC) between
two electrodes separated by an aqueous electrolyte with
good ionic conductivity [9]. The total reaction for splitting
Fig. 3. Operation principle of alkaline water electrolysis.
water is
H
2
O(l) + electrical energy H
2
(g) +
1
2
O
2
(g): (1)
For this reaction to occur a minimum electric voltage must
be applied to the two electrodes. This minimum voltage, or
reversible voltage, can be determined from Gibbs energy
for water splitting (described below). In an alkaline elec-
trolyzer the electrolyte is usually aqueous potassiumhydrox-
ide (KOH), where the potassium ion K
+
and hydroxide ion
OH

take care of the ionic transport. The anodic and ca-

thodic reactions taking place here are
Anode : 2OH

(aq)
1
2
O
2
(g) + H
2
O(l) + 2e

; (2)
Cathode : 2H
2
O(l) + 2e

H
2
(g) + 2OH

(aq): (3)
In an alkaline solution the electrodes must be resistant to
corrosion, and must have good electric conductivity and cat-
alytic properties, as well as good structural integrity, while
the diaphragm should have low electrical resistance. This
can, for instance, be achieved by using anodes based on
nickel, cobalt, and iron (Ni, Co, Fe), cathodes based on
nickel with a platinum activated carbon catalyst (Ni, CPt),
and nickel oxide (NiO) diaphragms. Fig. 3 illustrates the
operation principle of alkaline water electrolysis.
2.1. Thermodynamic model
Thermodynamics provides a framework for describing
reaction equilibrium and thermal eects in electrochemical
reactors. It also gives a basis for the denition of the driving
forces for transport phenomena in electrolytes and leads to
the description of the properties of the electrolyte solutions
[18]. Details on the fundamental equations for electrochemi-
cal reactors, or electrolyzers, are found in the basic literature
[19]. Below is a brief description of the thermodynamics
of the low-temperature hydrogenoxygen electrochemical
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 25
reactions used in the electrolyzer model. (A maximum elec-
trolyzer temperature of 100

C was assumed in this study.)

The following assumptions can be made about the water
splitting reaction: (a) hydrogen and oxygen are ideal gases,
(b) water is an incompressible uid, and (c) the gas and
liquid phases are separate. Based on these assumptions the
change in enthalpy H, entropy S, and Gibbs energy G
of the water splitting reaction can be calculated with refer-
ence to pure hydrogen (H
2
), oxygen (O
2
), and water (H
2
O)
at standard temperature and pressure (25

C and 1 bar). The

total change in enthalpy for splitting water is the enthalpy
dierence between the products (H
2
and O
2
) and the re-
actants (H
2
O). The same applies for the total change in
entropy. The change in Gibbs energy is expressed by
G = H TS: (4)
At standard conditions (25

and 1 bar) the splitting of water

is a non-spontaneous reaction, which means that the change
in Gibbs energy is positive. The standard Gibbs energy for
water splitting is G

=237 kJ mol
1
. For an electrochem-
ical process operating at constant pressure and temperature
the maximumpossible useful work (i.e., the reversible work)
is equal to the change in Gibbs energy G. Faradays law
relates the electrical energy (emf) needed to split water to
the chemical conversion rate in molar quantities. The emf
for a reversible electrochemical process, or the reversible
cell voltage, is expressed by
U
rev
=
G
zF
: (5)
The total amount of energy needed in water electrolysis
is equivalent to the change in enthalpy H. From Eq.
(4) it is seen that G includes the thermal irreversibil-
ity TS, which for a reversible process is equal to the
heat demand. The standard enthalpy for splitting water is
H

= 286 kJ mol
1
. The total energy demand H is
related to the thermoneutral cell voltage by the expression
U
tn
=
H
zF
: (6)
At standard conditions U
rev
= 1:229 V and U
tn
= 1:482,
but these will change with temperature and pressure. In the
applicable temperature range U
rev
decreases slightly with
increasing temperature (U
rev@80

C; 1 bar
= 1:184 V), while
U
tn
remains almost constant (U
tn@80

C; 1 bar
=1:473 V). In-
creasing pressure increases U
rev
slightly (U
rev@25

C; 30 bar
=
1:295 V), while U
tn
remains constant.
2.2. Electrochemical model
The electrode kinetics of an electrolyzer cell can be mod-
eled using empirical currentvoltage (I U) relationships.
Several empirical I U models for electrolyzers have been
suggested [11,13,14,20]. The basic form of the I U curve
0 50 100 150 200 250 300 350
1.0
1.2
1.4
1.6
1.8
2.0
2.2
Current Density, mA/cm
2
V
o
l
t
a
g
e
,

V
/
c
e
l
l
T = 80C
U
rev@T = 80C
T = 20C
U
rev@T = 20C
Overvoltage
U
tn@T=20-80C
Fig. 4. Typical I U curves for an electrolyzer cell at high and low
temperatures.
used in this study is, for a given temperature
U = U
rev
+
r
A
I + s log

t
A
I + 1

: (7)
Fig. 4 shows the cell voltage versus the current density at
a high and low operation temperature for a typical alkaline
water electrolyzer. As seen, the dierence between the two
I U curves is mainly due to the temperature dependence of
the overvoltages.
In order to properly model the temperature dependence
of the overvoltages (Eq. (7)) can be modied into a more
detailed I U model, which takes into account the tempera-
ture dependence of the ohmic resistance parameter r and the
overvoltage coecients s and t. A temperature dependent
I U model has been proposed by the author [21]:
U = U
rev
+
r
1
+ r
2
T
A
I
+s log

t
1
+ t
2
=T + t
3
=T
2
A
I + 1

: (8)
The Faraday eciency is dened as the ratio between the ac-
tual and theoretical maximum amount of hydrogen produced
in the electrolyzer. Since the Faraday eciency is caused
by parasitic current losses along the gas ducts, it is often
called the current eciency. The parasitic currents increase
with decreasing current densities due to an increasing share
of electrolyte and therefore also a lower electrical resistance
[20]. Furthermore, the parasitic current in a cell is linear to
the cell potential (Eq. (8)). Hence, the fraction of parasitic
currents to total current increases with decreasing current
densities. An increase in temperature leads to a lower re-
sistance, more parasitic currents losses, and lower Faraday
eciencies. An empirical expression that accurately depicts
these phenomena for a given temperature is

F
=
(I=A)
2
f
1
+ (I=A)
2
f
2
: (9)
According to Faradays law, the production rate of hydro-
gen in an electrolyzer cell is directly proportional to the
26 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
transfer rate of electrons at the electrodes, which in turn is
equivalent to the electrical current in the external circuit.
Hence, the total hydrogen production rate in an electrolyzer,
which consists of several cells connected in series, can be
expressed as
n
H
2
=
F
n
c
I
zF
: (10)
The water consumption and oxygen production rates are sim-
ply found from stoichiometry (Eq. (1)), which on a molar
basis
2
is
n
H
2
O
= n
H
2
= 2 n
O
2
: (11)
The generation of heat in an electrolyzer is mainly due to
electrical ineciencies. The energy eciency can be calcu-
lated from the thermoneutral voltage (Eq. (6)) and the cell
voltage (Eq. (8)) by the expression

e
=
U
tn
U
: (12)
For a given temperature, an increase in hydrogen production
(i.e., an increase in current density) increases the cell voltage
(Fig. 4), which consequently decreases the energy eciency.
For a given current density, the energy eciency increases
with increasing cell temperature. It should be noted here that
Eq. (12) is only valid for systems where no auxiliary heat
is added to the system. (If auxiliary heat is added, the volt-
age may, at very low current densities, drop into the region
between the reversible and thermoneutral voltage, and the
eciency would be greater than 100%). In low-temperature
electrolysis, the cell voltage will during normal operation
(5080

C and 40300 mA cm
2
) always be well above the
thermoneutral voltage, as observed in Fig. 4. However, some
initial heating may be required during start-up if the elec-
trolyzer has been allowed to cool down to ambient temper-
ature ( 20

C).
In order to calculate the overall performance of an elec-
trolyzer system, information about the number of cells in
series and=or parallel per stack and the number of stacks per
unit is needed. The rated voltage of an electrolyzer stack is
found fromthe number of cells in series, while the number of
cells in parallel yields the rated current (and H
2
-production).
The total power is simply the product of the current and
voltage.
2.3. Thermal model
The temperature of the electrolyte of the electrolyzer can
be determined using simple or complex thermal models,
depending on the need for accuracy. Assuming a lumped
thermal capacitance model [22], the overall thermal energy
2
Gas ow rates are commonly given on a normal cubic meter
per hour (Nm
3
h
1
) basis (Eq. (A.1)).
balance can be expressed as
C
t
dT
dt
=

Q
gen

Q
loss

Q
cool
; (13)
where

Q
gen
=
c
(U U
tn
)I = n
c
UI (1
e
); (14)

Q
loss
=
1
R
t
(T T
a
); (15)

Q
cool
= C
cw
(T
cw; i
T
cw; o
) = UA
HX
LMTD (16)
and
LMTD =
(T T
cw; i
) (T T
cw; o
)
ln[(T T
cw; i
)=(T T
cw; o
)]
: (17)
The rst term on the right-hand side of Eq. (13) is the in-
ternal heat generation, the second term the total heat loss
to the ambient, and the third term the auxiliary cooling de-
mand. The overall thermal capacity C
t
and resistance R
t
for
the electrolyzer, and the UA-product for the cooling water
heat exchanger are the constants that need to be determined
analytically or empirically prior to solving the thermal equa-
tions. It should be noted that the thermal model presented
here is on a per stack basis.
A simple method to calculate the electrolyzer tempera-
ture is to assume constant heat generation and heat transfer
rates for a given time interval. If the time steps are chosen
suciently small, the result is a quasi steady-state thermal
model. Using Eq. (13) as the basis, a quasi steady-state ther-
mal model can be expressed as
T = T
ini
+
t
C
t
(

Q
gen

Q
loss

Q
cool
): (18)
A more complex method is to solve the dierential equation
analytically and calculate the temperature directly. How-
ever, for this to be possible an expression for the outlet cool-
ing water must rst be found. If a constant temperature in
the LMTD-expression above is assumed, Eq. (16) can be
rewritten to
T
cw; o
= T
cw; i
+ (T T
cw; i
)

1 exp

UA
HX
C
cw

: (19)
Using Eq. (13) as the basis and inserting Eqs. (14)(16)
and (19), it can be demonstrated that the overall ther-
mal energy balance on the electrolyzer can be expressed
by the linear, rst-order, non-homogeneous dierential
equation
dT
dt
+ aT b = 0 (20)
with solution
T(t) =

T
ini

b
a

exp(at) +
b
a
; (21)
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 27
where
a =
1

t
+
C
cw
C
t

1 exp

UA
HX
C
cw

; (22)
b =
n
c
UI (1
e
)
C
t
+
T
a

t
+
C
cw
T
cw; i
C
t

1 exp

UA
HX
C
cw

: (23)
One advantage of having an analytical expression for the
temperature is that it facilitates the determination of the ther-
mal time constant (
t
= R
t
C
t
). It also provides a means to
double check a numerical solution of the dierential equa-
tion.
In an alkaline electrolyzer with a stationary electrolyte,
it has been observed that the overall UA-product for the
cooling water heat exchanger is (indirectly) a function of
the electrical current required by the electrolyzer (Fig. 10).
Hence, an empirical expression that accounts for both con-
duction and convection heat transfer phenomena is pro-
posed:
UA
HX
= h
cond
+ h
conv
I: (24)
The physical explanation for this behavior is that since the
electrolyte is stationary, and no pump is being used, the
convection heat transfer increases as a result of more mixing
of the electrolyte. An increase in mixing occurs because the
volume of the gas bubbles in the electrolyte increases with
increasing current density. Similarly, the ohmic resistance
in the electrolyte increases with increasing currents due to
increasing gas bubbling. Hence, this behavior is accounted
for in Eq. (24) and in the ohmic resistance term of Eq. (8).
3. Testing and verication of model
The alkaline electrolyzer analyzed in this study is the
one installed at the PHOEBUS plant in J ulich [23]. It is
a so-called advanced alkaline electrolyzer that operates at
a pressure of 7 bar and at temperatures up to about 80

C.
The cells are circular, bipolar (Fig. 1), have a zero spac-
ing geometry, and consist of NiO diaphragms and activated
electrodes (Fig. 3), which make them highly ecient. The
electrolyte is a stationary 30 wt% KOH solution. Each cell
has an electrode area of 0:25 m
2
and there are 21 cells con-
nected in series. This gives an operation voltage in the range
3040 V.
The hydrogen production and water cooling ow rates for
the PHOEBUS electrolyzer was not logged and collected on
a regular basis, along with the minutely collected operational
data. However, an experiment, where this and other pertinent
data was sampled for every 5 min, was performed on June
17, 1996 [24]. It is this 1-day experiment that forms the
basis for the comparisons between simulated and measured
data presented below.
0 50 100 150 200 250 300 350
1.2
1.4
1.6
1.8
2.0
2.2
Current Density, mA/cm
2
V
o
l
t
a
g
e
,

V
/
c
e
l
l
Predicted
T = 30C
T = 40C
T = 50C
T = 60C
T = 70C
T = 80C
Fig. 5. Predicted versus measured electrolyzer cell voltage.
3.1. Electrochemical model
In order to nd the six parameters needed in the proposed
empirical I U relationship (Eq. (8)), a systematic strategy
for obtaining the best possible curve t was developed (Ap-
pendix A). A comparison between simulated and measured
values for current and voltage for various operation tem-
peratures are presented in Fig. 5. The current, voltage, and
temperature data base (317 data points) used in Fig. 5 was
derived from 3 months (MayJuly 1996) of operational data
for the PHOEBUS electrolyzer.
The results show to which degree the ohmic resistance
parameter r is linearly dependent on temperature. Further-
more, the results show that the overvoltage coecient s can
be assumed constant, while the proposed expression for the
overvoltage coecient t can be used. That is, only six pa-
rameters are needed to model the I U curve. Fig. 5 demon-
strates that the predictability of the proposed I U model in
Eq. (8) is excellent; the RMS error is about 2:5 mV cell
1
.
Detailed measurements of the hydrogen production at
various current densities for the PHOEBUS electrolyzer
(26 kW; 7 bar) were only available for an operation temper-
ature of 80

C. However, detailed experiments on the tem-

perature sensitivity of the Faraday eciency were performed
on a very similar electrolyzer (10 kW; 5 bar) installed at the
HYSOLAR test and research facility for solar hydrogen pro-
duction in Stuttgart, Germany [20]. A comparison between
these two electrolyzers is given in Fig. 6, which shows the
data points from the HYSOLAR experiments (performed at
temperatures of 40

C; 60

C, and 80

C), the data points for

PHOEBUS (80

C), and the corresponding curve ts. Fig. 6

illustrate that the form of the Faraday eciency expression
proposed in Eq. (9) is suitable. A more detailed analysis of
the results shows that the coecients f
1
and f
2
vary lin-
early with temperature (Appendix A).
In a system simulation study it is important to re-
duce the number of parameters required. Having this in
mind, a simplied Faraday eciency expression with
non-temperature-dependent coecients (Eq. (9)) was
tested. The measured hydrogen production (recorded at
28 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
0 50 100 150 200
0
20
40
60
80
100
Current Density, mA/cm
2
E
f
f
i
c
i
e
n
c
y
,

%
T = 40C
T = 80C
T = 60C
HYSOLAR:
T = 80C
PHOEBUS:
Predicted
Fig. 6. Predicted versus measured faraday eciency for two ad-
vanced alkaline electrolyzers: (1) PHOEBUS (26 kW; 7 bar) and
(2) HYSOLAR (10 kW; 5 bar).
0 1 2 3 4
0
1
2
3
4
Measured, Nm
3
/hr
P
r
e
d
i
c
t
e
d
,

N
m
3
/
h
r
RMS error = 0.053 Nm
3
/hr
Fig. 7. Predicted versus measured hydrogen production.
5 min intervals over a 15-h time period) was compared
to the model. The results (Fig. 7) show that for system
simulations it suces to model the Faraday eciency with
a simple non-temperature-dependent expression. This can
be explained by the fact that in an actual system the elec-
trolyzer is always operating above a minimum protective
(idling) current (Fig. 9). Hence, the electrolyzer is usu-
ally operating in a region where the Faraday eciency is
not signicantly aected by a change in temperature. A
closer look at the hydrogen production for a typical day
with variable electrical current (Figs. 8 and 9) show that
there is good agreement between the measured data and the
model.
3.2. Thermal model
The cooling of the electrolyzer is crucial to prevent over-
heating. The most convenient cooling method is usually to
use regular tap water, as was the case with the PHOEBUS
electrolyzer. The average tap water ow rate and inlet tem-
perature for the specic day discussed here was 0:6 Nm
3
h
1
and 14:5

C, respectively. In order to estimate the cooling

eect an overall UA-product for the heat transfer between
0 4 8 12 16 20 24
0
1
2
3
4
Time, h
H
2

F
l
o
w

R
a
t
e
,

N
m
3
/
h
r
Electrolyzer switched ON
Predicted Predicted
Measured Measured
Fig. 8. Predicted versus measured H
2
-production for a typical day
with variable electrolyzer current input. (For clarity, only every
fourth measured data point was plotted).
0 50 100 150 200 250 300 350 400
0
1
2
3
4
Current, A
H
2
F
l
o
w
R
a
t
e
,
N
m
3
/
h
r
Measured
Predicted
Fig. 9. Predicted and measured H
2
-production as a function of
measured electrical current. (The same day as shown in Fig. 8, but
with all of the data points included).
0 4 8 12 16 20 24
0
100
200
300
400
500
0
10
20
30
40
50
Time, h
C
u
r
r
e
n
t
,

A
Current
Electrolyzer switched ON
U
A
-
p
r
o
d
u
c
t
,

W
/

C
Measured
Measured
Predicted
Predicted
Fig. 10. Overall UA-product for the heat exchanger between the
cooling water and the electrolyzer.
the cooling water and the electrolyzer was proposed in Eq.
(24). A comparison between the measured and predicted
UA-product is given in Fig. 10. The results clearly shows
that there exists an indirect relationship between the elec-
trolyzer current and the UA-product, and that this can, to
a good approximation, be accounted for by the proposed
empirical equation.
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 29
0 4 8 12 16 20 24
0
20
40
60
80
100
0
100
200
300
400
500
Time, h
T
e
m
p
e
r
a
t
u
r
e
,

C
Current
T
meas
T
pred
Electrolyzer switched ON
C
u
r
r
e
n
t
,

A
Fig. 11. Predicted versus measured electrolyzer temperature.
The electrolyzer temperature for a day with variable
power input (solar energy minus user load) is depicted in
Fig. 11. The initial temperature (at midnight) was 56:4

C
and the temperature at start-up of the electrolyzer (04:00
AM) was 51:7

C. This initial decrease in temperature is

only due to natural cooling to the ambient, with a tempera-
ture of about 20

C. The values for the thermal capacitance

and for the overall thermal resistance were found by inves-
tigating the cooling pattern for electrolyzer for a number
of dierent days: C
t
= 625 kJ

C
1
and R
t
= 0:167

C W
1
(equal to
t
= 29 h). The heat generation was calculated
from the energy eciency (Eq. (14)), where the electrical
current input was based on measurements depicted in Fig.
11, and the auxiliary cooling was based on measured tap
water conditions (0:6 Nm
3
h
1
and 14:5

C).
The result of this 1-day simulation shows that the model
slightly underpredicts the temperature. There might be
several explanations for this. One possible reason is mea-
surement error, where the main source of uncertainty is the
measured temperature of the electrolyte. Another possible
reason is simply the lack of detail in the thermal model,
where the main deciency of the model is that the tem-
perature of the electrolyte is assumed to be homogeneous.
However, in general, the thermal dynamic behavior of the
electrolyzer is predicted quite accurately.
The importance of modeling the electrolyzer temperature
accurately depends on the purpose of the models. A com-
parison between predicted and measured electrolyzer volt-
age and the corresponding power shows that the slight un-
derprediction of the temperature has relatively little signi-
cance from an energy system simulation point of view. For
instance, in the 1-day simulation presented above, the error
between the total simulated and measured energy demand
was less than 2%.
4. System simulation results
The intention of the electrolyzer model presented in this
study is to integrate it with other renewable and hydrogen en-
ergy models, and perform system simulation studies. Thus,
the usefulness of the model can essentially be divided into
two distinct modeling areas: (1) system design (or redesign)
and (2) optimization of control strategies.
The applicability of the model is best illustrated by show-
ing some results from integrated system simulations. The
reference system (Fig. 12) used in the simulations pre-
sented here is the PHOEBUS demonstration plant at the
Research Center in J ulich, Germany [23]. At the time of
the study this system consisted of four dierently oriented
PV-arrays with maximum power point trackers (MPPTs),
a pressurized advanced alkaline electrolyzer, hydrogen and
oxygen storage pressure vessels, an alkaline fuel cell, power
conditioning equipment (two DC=DC-converters and one
DC=AC-inverter), and a lead acid battery bank.
The level of detail of the electrolyzer model makes it pos-
sible to investigate a number of important system perfor-
mance parameters such as the number of electrolyzer starts,
H
2
-production, operating time, and standby (idling) time.
Statistical data such as minimum, maximum, and average
electrolyzer current, voltage, power, ow rates and temper-
atures can also be analyzed. Comparisons with the reference
system show that the model has a suitable level of detail
[5].
One of the key system control parameters is the opera-
tional mode of the electrolyzer, which determines whether
the electrolyzer is to operate in a xed or variable current
mode. In the constant current mode the battery is charged
during periods of excess current on the busbar and dis-
charged during periods with decit current. The battery state
of charge (SOC) in this case will mainly depend on two
uncontrollable variables, the solar radiation and user load,
and one controllable variable, the xed current (or power)
setting of the electrolyzer.
In the variable current scenario only excess current avail-
able on the busbar is fed to the electrolyzer, hence the
battery SOC remains constant. It is important to note that
most alkaline electrolyzers, even advanced ones specically
designed to manage uctuating input current, can only
operate down to about 20% of their rated power, and an
idling current needs to be maintained. Table 1 summarizes
the result from an analysis made on the inuence of alterna-
tive electrolyzer control strategies on system performance.
Electrolyzer mode of operation (xed or variable): The
benet of operating the electrolyzer in a variable current
mode rather than at a xed current is illustrated by com-
paring Sim ASim B. In Sim B the electrolyzer was oper-
ating at a xed current of 550 A ( 21 kW), which gave
a much more frequent electrolyzer on=o-switching than in
Sim A. This is also reected in the low average run time.
The net result is less hydrogen production, yielding a lower
nal pressure level in the H
2
-storage at the end of the year.
In Sim A, on the other hand, a large fraction of the energy
from the PV-arrays was used directly to run the electrolyzer.
As a consequence the use of the battery was minimized. A
comparison between Sim A and Sim B shows that the bat-
tery discharging energy increased by about 50%. Other xed
current set points have been investigated [5] and the trend
30 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
Fig. 12. Reference system (PHOEBUS).
Table 1
Inuence of alternative electrolyzer control strategies on system performance
Component=system Units Electrolyzer variations
A B C
Control set points
Electrolyzer Fixed current set point (if any) A Variable 550 Variable
Battery SOC level for on-switching % 90 90 80
Battery SOC level for o-switching % 80 80 70
Key performance indicators
H
2
-storage Initial H
2
storage pressure 0 : : : 1 0.45 0.45 0.45
Final H
2
storage pressure 0 : : : 1 0.24 0.11 0.20
Performance parameters
Electrolyzer Energy consumption MW h 10.56 9.94 10.83
Average power kW 7.48 21.01 7.37
Number of starts 156 273 162
Average run time h 9.0 1.7 9.1
H
2
production Nm
3
2,719 2,473 2,788
Battery Energy, discharging MW h 9.74 15.09 9.54
Energy, charging MW h 8.90 13.68 8.85
Italics = change in set point or majoreect due to change in set point.
is the same: variable electrolyzer operation mode gives a
better system performance than xed current mode.
Basic control strategy for electrolyzer: The inu-
ence of reducing the upper and lower thresholds for the
on=o-switching of the electrolyzer can be seen by compar-
ing Sim ASim C. In Sim C the on=o-switching set points
were lowered by 10% compared to Sim A, which required
the battery to operate much more frequently at medium high
SOC (6080%), and less frequently at high SOC levels.
Hence, the higher threshold settings used in Sim A gave a
better utilization of the installed battery capacity. In general,
there is a tradeo between battery capacity utilization on
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 31
one hand, and the need to dump energy (at SOC100%)
on the other. This needs to be incorporated into the control
strategy.
5. Conclusions
A mathematical model for an advanced alkaline elec-
trolyzer has been developed based on a combination funda-
mental thermodynamics, heat transfer theory, and empirical
electrochemical relationships. A lumped capacitance ther-
mal dynamic model with a special empirical relationship
for the overall heat transfer between a stationary electrolyte
and a cooling water loop has also been proposed. Data from
a reference system, the stand-alone photovoltaic-hydrogen
energy plant (PHOEBUS) in J ulich, was collected. Compar-
isons between predicted and measured data show that the
electrochemical part of the model accurately predicts the cell
voltage, hydrogen production, and eciencies. The results
also show that the thermal model can be used to predict the
transient behavior of the electrolyzer temperature.
The number of required parameters has been reduced to a
minimum so that the model is suitable for use in integrated
hydrogen energy systemsimulations. The electrolyzer model
has, along with several other hydrogen energy models such
as fuel cells and hydrogen storage, been made compatible
to a transient system simulation program (TRNSYS), which
makes it possible to integrate hydrogen energy component
model with a standard library of thermal and electrical re-
newable energy components. The model can be particularly
useful for (1) system design (or redesign) and (2) optimiza-
tion of control strategies. To illustrate the applicability of
the model, a 1-year simulation of a photovoltaic-hydrogen
system was made. The results show that several improved
electrolyzer-operating strategies can be identied with the
developed system simulation model.
The technical electrolyzer model presented in this paper
is suitable for dynamic simulation of RE=H
2
-systems. The
level of detail in the model is relatively high, but not too
high. This means that the model readily can be coupled to
economic models that account for both investment and oper-
ational costs. Detailed techno-economical system optimiza-
tion of RE=H
2
-systems is the next research topic that will
be studied by this author.
Acknowledgements
The work presented in this paper is based a Ph.D.-study
carried out at the Institute for Energy Technology (IFE)
in conjunction with the Norwegian University of Science
and Technology (NTNU). The author would like to thank
the Norwegian Research Council (NFR) for the nancing
(Ph.D. and post.doc.), IFE for providing the necessary re-
search facilities, and the Research Center in J ulich (FZJ) for
making the experimental data accessible. A special thanks
goes to J urgen Mergel (FZJ) for being so helpful. A humble
thought also goes to my late advisor Professor Odd Andreas
AsbjHrnsen.
Appendix A.
A.1. Curve tting
The following systematic seven-step procedure that facil-
itates the curve tting of the six parameters needed in the
proposed empirical I U relationship (Eq. (8)) is recom-
mended:
(1) Collect experimental or operational data for current I ,
voltage U, and temperature T.
(2) Organize the measured values for I and U in sets with
respect to constant values for T.
(3) Perform individual curve ts of the three coecients
r; s, and t in Eq. (7).
(4) Repeat step (3) for a few other temperatures (e.g., T =
20

C80

C).
(5) Perform intermediate curve ts on the temperature-
dependent coecients r and t.
(6) Verify that the temperature-dependent coecients in
Eq. (8) behave according to the expressions:
r(T) = r
1
+ r
2
T and t
1
+ t
2
=T + t
3
=T
2
:
(7) Perform an overall curve t on the entire data set, using
the values for the parameters r
i
; s, and t
i
found from
steps (1) to (6) as initial values for the regression.
This systematic procedure is illustrated graphically in Fig.
13, which shows the results of individual curve ts at xed
temperatures (data points), intermediate curve ts of these
data points (thin lines), and nally the overall curve t (solid
lines). The step-by-step strategy for nding the unknown
parameters in Eq. (8) proves to be very robust. This indicates
that the approach is not only limited to the curve tting of
the I U characteristics of an electrolyzer cell, but can also
be used in other situations where coecients in a model are
sensitive to inputs such as temperature, pressure, or other
governing conditions.
A.2. Parameters
Tables 24 list the parameters found and used in this
study.
A.3. Equations
Gas ow rates are commonly given on a normal cubic
meter per hour (Nm
3
h
1
) basis. For ideal gases the con-
version from mol s
1
to Nm
3
h
1
is given by

V = nv
std
3600; (A.1)
32 . Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133
20 40 60 80 100
0x10
0
2x10
-5
4x10
-5
6x10
-5
8x10
-5
10
-4
0.0
0.2
0.4
0.6
0.8
1.0
Temperature, C
R
e
s
i
s
t
a
n
c
e

P
a
r
a
m
.

(
r
)
,

m
2
step 7
step 3
step 5
O
v
e
r
v
o
l
t
a
g
e

C
o
e
f
f
.

(
t
)
,

m
2
/
A
step 7
step 3
step 5
Resistance, r:
Overvoltage, t:
Fig. 13. Electrolyzer I U curve tting; illustration of a step-by-step procedure to determine the temperature sensitivity of the resistance
parameter r and the overvoltage coecient t in Eq. (8).
Table 2
I U curve parameters (Eq. (8))
r
1
8:05e 5 m
2
r
2
2:5e 7 m
2

C
1
s 0:185 V
t
1
1:002 A
1
m
2
t
2
8:424 A
1
m
2

C
t
3
247:3 A
1
m
2

C
2
Table 3
Faraday eciency parameters (Eq. (9))
PHOEBUS HYSOLAR
T 80 40 60 80

C
f
1
250 150 200 250 mA
2
cm
4
f
2
0.96 0.990 0.985 0.980 0 : : : 1
Table 4
UA
HX
parameters (Eq. (24))
h
cond
7 W

C
1
h
conv
0:02 W

C
1
per A
where

V and n are the volumetric and molar ow rates,
respectively, and v
std
is the volume of an ideal gas at standard
conditions (0

C and 1 bar).
The root mean square (RMS) error is calculated according
to the equation:
RMS error =

n
i=1
( y
i
y
i
)
2
n 1
; (A.2)
where y
i
and y
i
are the predicted and measured values,
respectively, and n is the number of data points evaluated.
References
[1] Schucan TH. Case studies of integrated hydrogen energy
systems. Report, IEA=H2=T11=FR1-2000, International
Energy Agency Hydrogen Implementing Agreement Task
11Integrated Systems. Operating agent: National
Renewable Energy Laboratory, Golden, Colorado, 1999.
[2] The Hydrogen project at munich international airport.
http://www.hydrogen.org/h2muc/, October 2001.
[3] Yamaguchi M, Shinohara T, Taniguchi H, Nakanori T,
Okisawa K. Development of 2500 cm
2
solid polymer
electrolyte water electrolyser in WE-NET. Proceedings of the
12th World Hydrogen Energy Congress, Buenos Aires, 2126
June 1998. p. 74755.
[4] Galli S, Stefanoni M, Borg P, Brocke WA, Mergel
J. Development and testing of a stand-alone small-size
solar photovoltaic-hydrogen power system (SAPHYS).
Report, JOU2-CT94-0428, JOULE II-Programme, Directorate
General XII: Science, Research and Development, European
Commission, Brussels, 1997.
[5] Ulleberg I. Stand-alone power systems for the future: optimal
design, operation and control of solar-hydrogen energy
systems. Ph.D. thesis, Norwegian University of Science and
Technology, Trondheim, 1998.
[6] Eriksen J, Aaberg RJ, Ulleberg I, Ingebretsen F. System
analysis of a PEMFC-based stand alone power system
(SAPS). First European PEFC Forum, Lucerne, Switzerland,
36 July 2001. p. 44758.
[7] International energy agency (IEA) hydrogen program.
Annex 13design and optimization of integrated systems.
http://www.eren.doe.gov/hydrogen/iea/, October 2001.
[8] Wendt H, Plzak H. Hydrogen production by water electrolysis.
Kerntechnik 1991;56(1):228.
[9] Divisek J. Water electrolysis in low- and medium-temperature
regime. In: Wendt H, editor. Electrochemical hydrogen
technologies-electrochemical production and combustion of
hydrogen. Oxford: Elsevier, 1990.
[10] Hug W, Bussmann H, Brinner A. Intermittent operation and
operation modeling of an alkaline electrolyzer. Int J Hydrogen
Energy 1993;18(12):9737.
[11] Griesshaber W, Sick F. Simulation of HydrogenOxygen
Systems with PV for the Self-Sucient Solar House. FhG-ISE,
Freiburg im Breisgau, 1991 (in German).
. Ulleberg / International Journal of Hydrogen Energy 28 (2003) 2133 33
[12] Ulleberg I, Morner SO. TRNSYS simulation models for
solar-hydrogen systems. Solar Energy 1997;59(46):2719.
[13] Havre K, Borg P, Tommerberg K. Modeling and control of
pressurized electrolyzer for operation in stand alone power
systems. Second Nordic Symposium on Hydrogen and Fuel
Cells for Energy Storage, Helsinki, January 1920, 1995. p.
6378.
[14] Vanhanen J. On the performance improvements of small-scale
photovoltaic-hydrogen energy systems. PhD thesis, Helsinki
University of Technology, Espoo, Finland, 1996.
[15] Meurer C, Barthels H, Brocke WA, Emonts B, Groehn HG.
PHOEBUSan autonomous supply system with renewable
energy: six years of operational experience and advanced
concepts. Solar Energy 1999;67(13):1318.
[16] Klein SA, Beckman WA, Mitchell JW, Due JA, Due NA,
Freeman TL, Mitchell JC, Braun JE, Evans BL, Kummer
JP, Urban RE, Fiksel A, Thornton JW, Blair NJ, Williams
PM, Bradley DE. TRNSYSa transient system simulation
program. Manual v15, Solar Energy Laboratory, University
of Wisconsin, Madison, 2000.
[17] Klein SA, Alvarado FL. EESengineering equation solver.
Manual v6.315, F-Chart Software. Middleton: Wisconsin,
2001.
[18] Rousar I. Fundamentals of electrochemical reactors. In:
Ismail MI, editor. Electrochemical reactors: their science and
technology part A. Amsterdam: Elsevier, 1989.
[19] Pickett DJ. Electrochemical reactor design, 2nd ed. New York:
Elsevier, 1979.
[20] Hug W, Divisek J, Mergel J, Seeger W, Steeb H. Highly
ecient advanced alkaline electrolyzer for solar operation. Int
J Hydrogen Energy 1992;17(9):699705.
[21] Ulleberg I. Simulation of autonomous PV-H2 systems:
analysis of the PHOEBUS plant design, operation and energy
management. ISES 1997 Solar World Congress, Taejon,
August 2430, 1997.
[22] Incropera FP, DeWitt DP. Fundamentals of heat and mass
transfer, 3rd ed. New York: Wiley, 1990.
[23] Barthels H, Brocke WA, Bonho K, Groehn HG, Heuts
G, Lennartz M, Mai H, Mergel J, Schmid L, Ritzenho
P. PHOEBUS-J ulich: an autonomous energy supply system
comprising photovoltaics, electrolytic hydrogen, fuel cell. Int
J Hydrogen Energy 1998;23(4):295301.
[24] Mergel J. Personal communication, PHOEBUS electrolyzer:
detailed operational data. FZJForschungszentrum, J ulich,
February 1997.