Energy 118 (2017) 156e171

Contents lists available at ScienceDirect

Energy
journal homepage: www.elsevier.com/locate/energy

Hydrogen generator characteristics for storage

of renewably-generated energy
Janusz Kotowicz*, Łukasz Bartela, Daniel We˛ cel, Klaudia Dubiel
Institute of Power Engineering and Turbomachinery, Silesian University of Technology, Konarskiego 18, 44-100 Gliwice, Poland

a r t i c l e i n f o a b s t r a c t

Article history: The paper presents a methodology for determining the efficiency of a hydrogen generator taking the
Received 24 May 2016 power requirements of its auxiliary systems into account. Authors present results of laboratory experi-
Received in revised form ments conducted on a hydrogen generator containing a PEM water electrolyzer for a wide range of device
4 November 2016
loads. On the basis of measurements, the efficiency characteristics of electrolyzers were determined,
Accepted 7 November 2016
including that of an entire hydrogen generator using a monitored power supply for its auxiliary devices.
Available online 18 December 2016
Based on the results of the experimental tests, the authors have proposed generalized characteristics of
hydrogen generator efficiency. These characteristics were used for analyses of a Power-to-Gas system
Keywords:
Hydrogen generator
cooperating with a 40 MW wind farm with a known yearly power distribution. It was assumed that
Efficiency nightly-produced hydrogen is injected into the natural gas transmission system. An algorithm for
Power-to-Gas determining the thermodynamic and economic characteristics of a Power-to-Gas installation is pro-
Economic analysis posed. These characteristics were determined as a function of the degree of storage of the energy pro-
duced in a Renewable Energy Sources (RES) installation, defined as the ratio of the amount of electricity
directed to storage to the annual amount of electricity generated in the RES installation. Depending on
the degree of storage, several quantities were determined.
© 2017 Elsevier Ltd. All rights reserved.

1. Introduction large, un-adapted coal-fired units working in the largest centrally

It is forecast that the share of renewable energy sources in the
world's electricity production structure will increase [1,2]. In
Poland, wind turbines and photovoltaic cells are among the most
dynamically developing RES technologies [3,4]. The most important
advantage of use these sources is lack of emission of harmful sub-
stances. In contrast to use of carbon sources the use of renewable
energy does not contribute to the emission of dust, sulfur com-
pounds, nitrogen oxides, mercury and carbon dioxide. Common-
ness of occurrence and renewability of energy potential make that
these energy sources are safe and most commonly cheap in
exploitation. The popularity of RES is stimulated by support
mechanisms regulated by national laws [5,6]. The environmentally-
friendly nature of these technologies is a feature that deserves
rational support. The basic mechanism of this support, in addition
to a system of tradable certificates, is guaranteed priority access to
the electricity grid. This mechanism creates a situation in which the
* Corresponding author.
E-mail address: janusz.kotowicz@polsl.pl (J. Kotowicz).
organized systems take on a role of regulation system. The Polish
energy system has a relatively low level of elasticity. In addition,
various renewable energy sources are characterized by consider-
ably differing potentials for electricity generation, which rarely
correlate with end-user electricity demand, particularly in the case
of wind energy [7e12]. Mentioned features and the necessity to
ensure energy security causes that the development of renewable
energy sources is limited by the operator of the electricity system.
A further increase in installed wind farm capacity can lead to a
situation in which coal-fired units will reach the limits of their
regulatory capabilities, and the power system will become unstable
[13,14]. Therefore further increase in installed capacity of wind
farms will be possible by improving operational features of coal-
fired power plants [15e17]. This will be the result of the startup
of new units with supercritical parameters [18,19]. It is also
important that the increase of installed capacity in power plants
will be based on gas turbines [20e22]. However, experts from the
Polskie Sieci Elektroenergetyczne S.A. predict that the dynamic of
startups of new wind farms will contribute to insecurity of the
system by 2025. In addition to the investment in modern power
plant units, it is also very important to implement energy storage

http://dx.doi.org/10.1016/j.energy.2016.11.148
0360-5442/© 2017 Elsevier Ltd. All rights reserved.
 J. Kotowicz et al. / Energy 118 (2017) 156e171
systems to avoid this scenario [23e27]. Energy storage systems
would draw power during off-peak demand periods and produce
during peak periods [28,29].
Energy storage is important for a wide range of potential ben-
eficiaries of technology - from individual users of electric equip-
ment, automotive users, and industry to network operators.
Depending on the scale, energy storage can be classified into five
groups by storage method: mechanical, electrical, chemical, elec-
trochemical and thermal. In the case of cooperation with the large
power system, the most favorable characteristics have chemical
and mechanical solutions, including pumped storage and com-
pressed air energy storage (CAES) [30,31]. A significant develop-
ment potential have also electrochemical systems. Wide discussion
on the potential of the various solutions were carried out in
Refs. [32,33]. This article presents the results of analyses for Power-
to-Gas technology, which is classified in the chemical group. Ac-
cording to authors' assumptions, the analyzed system is used in a
daily cycle and enables the production of hydrogen during the so-
called night valleys, i.e., at night. The hydrogen produced is accu-
mulated in buffer tanks and directed to the gas transmission grid.
2. Power-to-Gas technology
The Power-to-Gas technology allows to produce the gas of high
potential energy using electricity taken directly from a generation
system or the network. The transformation process within a Power-
to-Gas installation of electricity to the chemical energy of gas fuel
can be justified during periods of higher electricity production and
lower electricity demand. This overproduction occurs mainly in
systems based on renewable sources (mainly solar and wind
sources). A positive economic effect for investments in energy
storage systems may occur in a situation in which the price of en-
ergy in a period of overproduction is very low [34,35]. The over-
production usually occurs during low demand periods (the night
valley, the weekend valley). Such scenarios commonly occur in
countries where the share of renewable energy is significant (e.g.,
Denmark, Spain, and Germany). This is also a justified investment
from the point of view of the power system operator. In this case,
the Power-to-Gas installation can adopt a function of regulator and
adapt the available capacity of the system to the demand [36e38].
The strong interest in energy storage systems, due to their ability to
regulate supply to accommodate demand, is observed particularly
in countries where a growing share of RES is accompanied by a
significant share of coal-fired power plants [12,39e42]. In Poland
for example, increasing installed wind farm capacity with priority
access to the grid is in force, resulting in an increase in forced
shutdowns of power units. Shutdowns of power units are required
by the system operator's instructions; they may result from high
outputs from wind-power sources and low electricity demand
[43,44]. Increased coal-fired power in off-peak periods could be
directed to storage; this could contribute to a decrease of the
number of the forced shutdowns of units, and thus reduce the
related cost [45e47].
The criterion for classifying Power-to-Gas technology can be
based on the type of gas produced. The basic element of the system,
irrespective of the type of gas, is the electrolyzer, whose function is
to produce hydrogen. If the installation contains no additional gas
conversion equipment, hydrogen is the final product of the process.
In this case, the hydrogen can be injected into the natural gas grid,
forming part of the gas transportation infrastructure. Additionally,
the product of the electrolysis process can be delivered by tank
trucks to industry (e.g., chemical, metallurgical, electronic, refining
and fats industries). It can also be used for energy purposes at the
place of production (Power-to-Gas-to-Power). It should be
157
remembered, however, that hydrogen is difficult to transport or use
for energy generation [48,49]. The literature often emphasizes the
problems connected with the use of hydrogen-rich fuels in gas
turbines and piston engines [50,51]. Synthetic Natural Gas (SNG)
can be a more convenient fuel for these purposes. Synthetic Natural
Gas can be produced from carbon dioxide and hydrogen and can be
the end-product of a Power-to-Gas installation; however, it re-
quires installing a methanation reactor. A Power-to-Gas installation
can be of interest in power plants whose emission of carbon dioxide
generates higher operations cost. Carbon dioxide can be a product
of gasification of biomass [52,53]. The advantage of these systems is
the possibility of efficient use of oxygen, which is the by-product of
the electrolysis process. Optional ways of the use of oxygen, and
their potential has been described in Ref. [54]. Worldwide, pilot
installations currently use methanation reactors based both on
biological and chemical processes. The Power-to-Gas technology
can be also implemented successfully in other sectors of the in-
dustry e.g. in pulp mill, which is a large producer of wood-based
CO2, so the hydrogen can be used for the methanation process
and the oxygen in the pulp and paper mill, replacing the existing O2
production facility [55]. Fig. 1 shows possible connections in a
Power-to-Gas system.
3. Production of hydrogen
Hydrogen is a product of an electrolysis installation, constituting
of battery of currently available compact electrolyzers. The most
popular types of electrolyzers for pilot installations worldwide are
electrolyzers with a polymer electrolyte membrane (PEM)
[45,56,57] and alkaline electrolyzers (AEL) with aqueous alkaline
solutions as electrolytes (KOH or NaOH) [58e60]. Other types of
electrolyzers include devices with a ceramic membrane (SOECs)
[61e66]. PEM and alkaline electrolyzers operate at a process tem-
perature between 70 and 100
C. SOEC electrolyzers operate be-
tween 700 and 1000
C. These types of electrolyzers require a
supply of heat in the form of steam generated in an external process
[67]. This limits the potential points of use. Because of the high
process temperature, the efficiency of SOEC electrolyzers (defined
as the ratio of the chemical energy of fuel obtained to input electric
energy) may exceed 100%. SOEC electrolyzers are rarely considered
in installation planning; this may serve to regulate power systems
because of the long heating and cooling periods of system ele-
ments. In this respect, the PEM and alkaline electrolyzers are more
favorable. In addition, the times required to change their loads are
from 10% to 100% of the nominal power per second and from 10% to
25% of the nominal power per second, respectively. The PEM
electrolyzers, despite lower technological maturity, are a perfect
alternative to alkaline electrolyzers because of their wide load
range (5%e100% of nominal power), very high purity of produced
hydrogen (>99.999%) and lower operating cost. Unfortunately, the
lower technological maturity of PEM electrolyzers is reflected in
lower membrane lifetime and higher restoration costs [68,69]. An
important aspect of the installation, irrespective of the type of
electrolyzer, is that the water used for electrolysis must comply
with the strictest standards; therefore, it is necessary to install
water treatment in Power-to-Gas systems. Especially, the removal
of minerals and ions is required before begin electrolysis process
[70,71].
The efficiency of the hydrogen generators used is important for
the efficiency of the entire Power-to-Gas storage system. There are
highly diverse methods to determine this quantity. Knowledge of
the efficiency characteristics as a function of electrolyzer load is the
basis for a correct analysis of an energy storage system cooperating
with a generation system.
158 J. Kotowicz et al. / Energy 118 (2017) 156e171
Fig. 1. Diagram of possible connections of a Power-to-Gas system.
3.1. The efficiency of hydrogen generators
The efficiency of a water electrolysis process may be defined as
the ratio of the theoretical amount of energy Et to the actual
amount of energy Er required to split 1 mol of water into hydrogen
and oxygen, according to the relationship:
Et
hec ¼ ; (1)
Er
In real systems operating in the low-temperature technology,
the electrolysis process proceeds according to chemical reaction:
1 where: I e current flowing through the electrolyzer, t e time.
H2 OðlÞ ¼ H2 ðgÞ þ O2 ðgÞ; (2)
2
where the water is in a liquid state (l), and hydrogen and oxygen are
gaseous (g).
To maintain the electrolysis process it is necessary to supply
electricity and heat. The minimum amount of electrical energy
required to split of 1 mol of water is equivalent to Gibbs free energy
DG0, which is related to the “free energy” voltage E0. For splitting
liquid water, this energy is:
DG0 ¼ n$F$E0 ¼ 237:22 kJ=mol; (3)
where n e the number of electrons transferred during the elec-
trochemical reaction of splitting one molecule of water (n ¼ 2), F e
Faraday constant (F ¼ 96,485 C/mol), E0 e standard electrolysis
potential (at temperature 298 K and pressure 1 bar this voltage is
equal E0 z 1.23 V) [71e73].
The heat required is the result of the entropy change and the
temperature at which water is supplied. The entropy change for the
liquid water is DS ¼ 163,15 J/(mol$K). Hence, the total energy
required to split 1 mol of water is the sum of the Gibbs free energy
and the heat required, which corresponds to the enthalpy change
DH0 for the hydrogen:
DH0 ¼ DG0 þ T$DSðTÞ ¼ 285:84 kJ=mol; (4)
With alkaline and PEM electrolyzers, the heat needed for the
reaction is generated by current flow through the electrolysis cell.
However, this requires supply of a voltage higher than E0 to the
electrolysis cell. The theoretical amount of energy Et can therefore
be defined by the thermoneutral voltage V0, which is related to the
enthalpy change for the liquid water:
DH 0
V0 ¼ z1:48 V; (5)
n$F
The theoretical amount of energy can be calculated from the
relationship [74]:
Et ¼ V 0 $I$t; (6)
The thermoneutral voltage V0 specified above is related to the
higher heating value (HHV) of hydrogen, which is equal to the
enthalpy change DH0 for hydrogen [74]. When calculating the ef-
ficiency of the electrolysis process, it is usual practice to use the
lower heating value (LHV) of hydrogen. Manufacturers of electro-
lyzers often use specific energy consumption in kWh/m3n or kWh/
kg, to describe their effectiveness.
The actual energy Er needed to split water into hydrogen and
oxygen is mainly dependent on the voltage U supplied to the
electrolyzer, which is higher than the thermoneutral voltage V0
because of losses related to the electrolytic resistance, activation of
the electrolysis process and mass transport through the membrane
[75]:
Er ¼ U$I$t; (7)
The actual energy is thus the electric energy supplied to the
electrolyzer. From equations (1), (6) and (7), it follows that the ef-
ficiency of the electrolyzer is equal to the ratio of the thermoneutral
voltage V0 and the voltage supplied the electrolyzer U.
The voltage values given above refer to standard conditions
(temperature 298 K and pressure 1 bar). However, the voltages
depend on temperature and pressure, and in the case of the voltage
U, on the current density in the electrolytic cell; thus the theoretical
efficiency of the electrolysis depends on the ratio of voltages, the
temperature, the pressure and the electrolyzer load [75]. Due to the
prediction of an electrolyzer's operation with varying loads, the
changes in values of current flowing through the electrolyzer must
be taken into account. Generally, the efficiency of an electrolyzer
 J. Kotowicz et al. / Energy 118 (2017) 156e171
increases with a decrease in electrolyzer output compared to the
nominal output. This is because losses are reduced with a smaller
supply current. The reduction in current value causes a nonlinear
reduction of losses, mainly related to ohmic resistance in the
electrolyzer, which is independent of the load [75,76]. The elec-
trolyzer operating temperature has a significant impact on its
hydrogen production efficiency; if it increases, it causes an increase
of efficiency. The electrolyzer operating pressure has less impor-
tance [74,75,77].
Because the electrolyzer is direct current (DC) powered, and
considering equations (1), (6) and (7), as well as using the higher
heating value (HHV) of hydrogen, the electrolyzer efficiency will be
calculated with the formula:
V_ H2 $HHV
hEL ¼ ; (8)
NDC
where: V_ H2 e volume flow rate of produced hydrogen, NDC e DC
electrical power supplied to the electrolyzer (9).
Electrical power is determined by the relationship:
NDC ¼ U$I: (9)
Hydrogen generators that use electrolysis for hydrogen pro-
duction are usually AC powered. This increases the energy re-
quirements of the entire system, mainly due to AC/DC converters,
which operate with a specified efficiency. In addition, generators
are equipped with a series of auxiliary devices, normally powered
from the same source. Therefore, the efficiency of the hydrogen
generator is determined by:
V_ H2 $HHV
hHG ¼ ; (10)
NAC
where: NAC e AC electric power supplied to the hydrogen generator.
Some analyses additionally account for hydrogen losses in a
hydrogen purification system and in some safety devices in the
hydrogen flow path. This is important, especially in small units with
capacities of several m3n/h in which the flow of hydrogen produced
by the generator can be even a few percent lower compared to the
theoretical capacity that results from the value of the current
[78e80]. Because the hydrogen flow is measured at the outlet of the
hydrogen generator, all possible losses occurring within the entire
generator are included in determining its efficiency.
Commonly available hydrogen generators are composed of
several basic systems:
 electrolyzer e a basic element of the generator in which water is
split into hydrogen and oxygen; it is usually in the form of a
stack of electrolytic cells supplied with direct current. A single
generator may comprise several stacks.
 monitoring and control system - regulates the proper supply of
the electrolyzers and controls the electrolysis process; controls
the operation of power systems and auxiliary equipment to
maintain operating parameters of the generator.
 auxiliary equipment - pumps, fans, cooling systems, heaters,
hydrogen purification systems [81].
A block diagram of a hydrogen generator system, along with the
direction of power flow, is shown in Fig. 2.
An electrolyzer is normally powered from the AC mains through
an AC/DC converter, which is a source of relatively high energy
losses. The efficiency of an AC/DC converter is load-dependent and
in the case of small converters fluctuates near 90% at its nominal
output power; at low load it may drop below 70% [75,82]. The
159
converter nominal power and the type of elements used in the
rectifier circuit are of great importance [80]. In some solutions it is
possible to connect the electrolyzer directly to a DC power source,
which may be, e.g., photovoltaic cells (PV). It is required to match
the voltage and current levels of the PV system to the electrolyzer
or use adequate DC/DC converters. In every case, power losses of
approximately 10% are to be expected when operating in the range
of 50%e100% of nominal power, with losses increasing as electro-
lyzer power decreases [75,83].
The auxiliary equipment and the monitoring and control sys-
tems of hydrogen generators are usually supplied from the same
energy source as the electrolyzer. However, often they require
different voltage values. In this case, it is necessary to apply voltage
converters other than those in the supply system of the electro-
lyzers. It is also possible to supply, e.g., pumps or fans directly from
the AC mains. Power required for these devices is constant over the
entire capacity range of the most frequently used hydrogen gen-
erators. Therefore, the applied AC/DC converter will always work
with the same efficiency. With the above assumptions, the power
NAC can be determined according to the relationship:
NAC ¼ NDC þ Naux ; (11)
where: Naux e sum of the auxiliary power and losses.
NAU þ NCS þ Np
Naux ¼ DN þ ; (12)
hauxAC=DC
where: DN e power losses in AC/DC converter of electrolyzer, NAU e
electric power of auxiliary equipment, NCS e electric power of the
monitoring and control systems, Np e electric power of pumps,
hauxAC/DC e efficiency of any AC/DC converter supplying auxiliary
systems (not a function of the electrolyzer power).
The efficiency of a hydrogen generator also depends on the size
of the device. The construction of larger units with higher elec-
trolyzer capacities allows reducing losses. Based on the literature
data and technical data of currently available PEM electrolyzers, the
impact of nominal capacity on the achievable efficiency of elec-
trolyzers and the entire system for hydrogen production can be
determined (Fig. 3) [47,78,84e89]. The tendency of increasing ef-
ficiency with increasing capacity is clear. However, there is a sig-
nificant difference between the efficiency of the cell itself and that
of the entire system for hydrogen production (hydrogen generator);
this efficiency difference may even reach 20%. It should be noted
that an increase in nominal capacity has less impact on the effi-
ciency of electrolyzers than on that of the hydrogen generator. This
is related to a lower auxiliary power index d in large hydrogen-
production units, which takes a value of approximately 10%. This
index is defined by:
Naux
d¼ ; (13)
NAC
The largest units with PEM technology currently offered are
characterized by efficiencies (HHV) of 70%. All these values are
obtained at the nominal power. A partial electrolyzer load results in
a change in its hydrogen production efficiency.
3.2. Experimental determination of hydrogen generator efficiency
In order to determine the efficiency characteristics of an actual
hydrogen generator and electrolyzer, a device manufactured by the
TsvetChrom company was tested. The subject of experimental
research was a hydrogen generator with four PEM electrolyzers
whose total rated capacity as specified by the manufacturer was
160 J. Kotowicz et al. / Energy 118 (2017) 156e171
Fig. 2. Diagram of hydrogen generator.
Fig. 3. The efficiencies of commercially available PEM electrolyzers.
0.108 m3n/h. Nominal parameters of a single electrolyzer were:
Izn ¼ 13 A, Uzn ¼ 14 V, which translates into a total power of all
electrolyzers of NDC ¼ 728 W [90]. Electrolyzers in the hydrogen
generator were electrically connected in series and were connected
to an external DC power source that allowed changing the power
supplied; they all contributed to a common hydrogen outlet. A
diagram of the tested circuit is shown in Fig. 4, which presents all
electrolyzers schematically as a single electrolytic cell. The points at
which measurements were made are indicated also.
During testing, the entire system should be in steady state
operation to maintain constant values: current and voltages on
electrolyzer stacks, operation temperature and pressure of elec-
trolyzers. It was assumed that all of the auxiliary equipment
operated stably with constant power consumption. In determining
the auxiliary power periodic run the pumps for example refilling
water tanks, periodic opening the valves in order to clean hydrogen
tubes from residual water, switch on the signaling lights was
omitted. DC voltage and current was measured directly on the
electrolyzer. The operating temperature was determined based on
temperature measurements of water flowing out of the electro-
lyzer. The hydrogen pressure was controlled at the outlet of
hydrogen from the generator. The electricity demand of the
hydrogen generator was determined based on measurement of the
AC power, which supplied all circuits of the generator.
Accurate measurement of hydrogen flow rate obtained from the
electrolyzer was very important for the proper determination of
efficiency. This measurement directly out of the electrolyzer is very
difficult because some water and oxygen penetrate the electrolyzer
membrane and flow together with the hydrogen. It is necessary to
use a purification system, which dries the hydrogen and removes
impurities. Theoretically, these systems produce hydrogen with a
purity of more than 99.999%. Only this purified hydrogen is fed to
the hydrogen generator outlet and measured using a thermal flow
meter.
During the experiments, the DC current supplying the electro-
lyzers was varied in the range of approximately 15%e108% of the
rated value (2e14 A), and the operational temperature kept be-
tween 22 and 32
C. Steady state operation of the electrolyzers was
difficult to maintain because of the periodically opening water
purge valve in the hydrogen purification system; this resulted in
momentary changes in electrolyzer voltage and fluctuations in
hydrogen flow. All measurements were performed with a hydrogen
outlet pressure equal to atmospheric pressure. Auxiliary equipment
was powered from the AC power source. The measured power
consumed by these devices (the second part of the formula (12))
was independent of the electrolyzer current and was equal to 57 W,
also accounting for the AC/DC conversion losses. Based on this
study, the characteristics of the hydrogen generator were
determined.
Fig. 5 shows the characteristics of the sum of all electrolysis
voltages and the total DC power supplied to the electrolyzers versus
DC current. The relationship between voltage and current is loga-
rithmic [91,92].
The electrolyzer efficiency hEL was determined based on
 J. Kotowicz et al. / Energy 118 (2017) 156e171 161

Fig. 4. Diagram of tested hydrogen generator, where points of the measured values are:V_ e hydrogen volume flow rate, t1 e electrolyzer temperature, t2 e cooling water tem-
perature, p1 e hydrogen pressure, p2 e water pressure, A e current, V e voltage; DC e DC power source, AC e AC power source.

Fig. 5. Basic characteristics of the electrolyzers.

Fig. 6. Characteristics of the efficiency and capacity of a hydrogen generator.

equation (8), while the calculation of the hydrogen generator effi-

ciency hHG takes into account only the auxiliary equipment power.
Because the electrolyzers were supplied from an external DC
source, losses in the AC/DC converters suppling electrolyzers were
not taken into account. Based on the measurements, it was found
that the power consumed by the auxiliary equipment of the
hydrogen generator was constant, and at nominal capacity was 7.2%
of the nominal power of hydrogen generator (NAC)nom. As the
production of the hydrogen generator decreased, the auxiliary
power system index d changed significantly, and at a load of
0,2Nnom, the auxiliary power system index was approximately 36%.
Fig. 6 shows the efficiency characteristics of the electrolyzer and
hydrogen generator as a function of the AC power, referred to the
nominal power. The figure shows also the change of the relative
volume flow of hydrogen produced as a function of the relative
power change.
On the basis of these studies, it was found that for the range of
measured capacity of the hydrogen generator, the flow rate of
hydrogen produced was linearly dependent on the supplied AC
power. The efficiency of the tested electrolyzer hEL increased with
decreasing capacity; however, this relationship was not linear; it
can be described by the equation:

0:134
NAC
hEL ¼ 0:49$ ; (14)
ðNAC Þnom
The equation has been fit to the minimum of the square of the
correlation coefficient, which in this case is 0.925.
This relationship has been determined based on several mea-
surement series, at which generated repeatability of measurements
was about 2%. An important factor that affects on the obtained
results is the temperature of the electrolysis process, which was
maintained at the level 22e32
C. Operation at the higher tem-
perature causes reduction of thermoneutral voltage V0 and AC
power NAC, and causes an increase of the electrolyzer efficiency hEL.
162 J. Kotowicz et al. / Energy 118 (2017) 156e171

However, at the observed temperature fluctuations, the voltage efficiency of the electrolyzer and the auxiliary power index through
varies not more than 0.5%. Reached temperatures are relatively low equation (16):
for the PEM electrolyzers, what can also explain the fact of low
efficiency. The second factor that affects on the obtained efficiency hHG ¼ hEL $ð1  dÞ; (16)
is the measurement accuracy of hydrogen flow-rate V_ H2 . The used
thermal mass flow meter is characterized by a accuracy of 0.5%, so It should be noted that this is similar to describing the net ef-
the measurement accuracy of hydrogen flow-rate has no significant ficiency of the power plant, i.e., using the gross efficiency and
influence on the overall shape of the efficiency characteristic. auxiliary power index. If we assume Naux ¼ constant, the auxiliary
The equation for the electrolyzer efficiency can therefore be power index of hydrogen generator can be written as:
represented in general form using the coefficients a and m by the
relation: Naux Naux ðN Þ ðN Þ
d¼ ¼ $ AC nom ¼ dnom $ AC nom ; (17)
NAC ðNAC Þnom NAC NAC
a
hEL ¼ h im ; (15)
NAC where: dnom e auxiliary power system index at nominal AC power.
ðNAC Þnom
Substituting equations (15) and (17) into equation (16) yields:
The efficiency hEL of the tested electrolyzer at nominal power
 
was 49%, while at the lowest power it achieved better than 60%. a ðNAC Þnom
This range of efficiencies is consistent with values described for
hHG ¼ h im $ 1  dnom $ ; (18)
NAC NAC
other small-capacity electrolyzers [93]. The low values obtained for ðNAC Þnom
the tested electrolyzer may result from the relatively high electrical
The effects of the parameters a, m and dnom on the hydrogen
losses in the electrolytic cell, which translates into a high supply
generator efficiency hHG, are shown in Figs. 7e9, according to
voltage. Low efficiency of the electrolyzer is characterized by low
equation (18).
power units (Fig. 3).
To obtain the maximum value of efficiency, we differentiate
Modern PEM electrolyzers are characterized by efficiency above
equation (18):
70% at nominal capacity [73,77,79,89,92]. Most of the large power
electrolyzers currently built are composed of many small capacity
vhHG a$m a$dnom $ðm þ 1Þ
electrolyzer stacks which are connected in parallel [94e96]. This
NAC
¼ h imþ1 þ h imþ2 ; (19)
solution does not change the operating characteristics of a large vðNAC Þnom NAC NAC
ðNAC Þnom ðNAC Þnom
unit in relation to a small electrolyzer, it is only scaled depending on
the amount of used electrolyzer stacks. Therefore, the efficiency of From the condition:
the electrolyzer is less dependent on the nominal capacity of this
unit. While greater importance on the shape of efficiency charac- vhHG
teristics of the hydrogen generator will be auxiliary power index, NAC
¼ 0; (20)
vðNAC Þ
which clearly decreases with increasing nominal power of the nom

generator hydrogen. The correct determination of this index is

the generator load relationship was calculated where the efficiency
possible when is known the own needs power of hydrogen
hHG reaches an extremum:
generator.
On the shape of the electrolyzer efficiency characteristics affects
the voltage efficiency and the so-called Faraday efficiency. The NAC mþ1
¼ dnom $ ; (21)
voltage efficiency, defined as the ratio of thermoneutral voltage V0 ðNAC Þnom m
to the voltage U supplied to the electrolyzer, is closely related to the Thus, the maximum value of efficiency is equal to:
current-voltage characteristics of the cell electrolyzer, and it de-
pends on the occurring losses in the electrolyzer. Whereas the
a 1
Faraday efficiency, defined as the ratio of the volume of gas pro- hHGðmaxÞ ¼ ·  : (22)
m þ 1 dnom mþ1 m
duced over a given time interval to the theoretical volume that m
should be produced during that time in accordance with Faraday's
laws, is constant in a wide range of electrolyzer capacity and very
close to 100%. Accordingly, a current-voltage characteristic has the
decisive influence on the efficiency characteristics. General the
dependence of voltage U supplied to the electrolyzer versus current
is described by logarithmic function. Considering the formula on
efficiency hEL (8) and above information, theoretical dependence of
efficiency versus the AC power can be described as a power func-
tion in a fairly wide range of electrolyzer capacity.
The proposed relation (15) can be used in a range of the
hydrogen generator nominal power from 20 to 100%. When the
power is less than 20% of nominal power, the relation contain un-
certainty, because of the high non-linearity of current-voltage
characteristics and less Faraday efficiency in this area.
The efficiency of the hydrogen generator is at least a few per-
centage points lower than the efficiency of electrolyzer and is
almost constant over a fairly large power range. When the power
supply is less than 30% of nominal power, efficiency hHG falls
rapidly, so the production of hydrogen becomes ineffective. The Fig. 7. The efficiency characteristics of a hydrogen generator (hHG) for different values
efficiency of the hydrogen generator is proposed to depend on the of coefficient a for m ¼ 0.2 and dnom ¼ 0.075.
 J. Kotowicz et al. / Energy 118 (2017) 156e171
Fig. 8. The efficiency characteristics of a hydrogen generator (hHG) for different values
of parameter m for a ¼ 0.725 and dnom ¼ 0.075.
Fig. 9. The efficiency characteristics of a hydrogen generator (hHG) for different values
of auxiliary power system (dnom) for a ¼ 0.725 and m ¼ 0.2.
4. The storage of hydrogen
In Power-to-Gas technology, a hydrogen generation system is
often integrated with hydrogen storage. Installations of energy
storage can be classified as short- and long-term storage, depend-
ing on their roles. Storage tanks, depending on the amount of gas
stored, can be constructed as above-ground cylindrical or spherical
tanks or can be underground. Natural formations, e.g., rock caverns
created by extraction of gas or oil, coal or salt mines, can be adopted
as underground tanks. The tank can also be caverns created spe-
cifically for this purpose by leaching salt deposits. Salt caverns
require large financial resources (construction of caverns of
750,000 cubic meter capacity is estimated to cost 20e30 M V), but
they are suitable for storage tanks because the salt is inert to the
hydrogen. Because the hydrogen is stored only in gaseous form, the
capacity of energy storage is linked to the storage pressure. For
large salt caverns with a capacity of 500,000 m3, hydrogen can be
stored at pressures in the range of 60e180 bar. The highest pres-
sures of hydrogen storage are achieved in above-ground tanks.
Cylinder tanks used for hydrogen storage to supply fuel-cell vehi-
cles achieve a high pressure of 700 bar.
When integrating hydrogen production into the natural gas grid
and the grid receives the produced hydrogen, the gas tanks in a
Power-to-Gas installation are optional. In this option, the tank
function is to keep the required amount of hydrogen injected into
the natural gas pipeline. The flow rate of hydrogen can be deter-
mined by the maximum acceptable hydrogen concentration in
natural gas. In the literature, this maximum concentration,
163
depending on the location of the energy storage system, ranges
widely from 5% to 20%. The concentration level of hydrogen in
Polish conditions is 5%. The concentration limit may result from
quality assurance of gas delivered to consumers or may be imposed
by the technical requirements of components of the gas network
such as pipes, compressors, etc. In addition to the maximum
pressure, other basic design values for determining hydrogen tank
capacity include the maximum storage temperature, the injection
and outflow velocities, and the maximum quantity of hydrogen
delivered to a gas network in a single operation cycle. Finally, a
minimum pressure of storage is important in the determination of
storage capacity. This pressure level depends on the minimum in-
jection pressure of hydrogen into the gas pipeline, which in turn
depends on the pressure in the natural gas transmission network.
5. A case study of a Power-to-Gas installation
Fig. 10 presents a diagram of a simple Power-to-Gas installation;
it is supplied with electricity produced by production systems using
renewable energy sources. Hydrogen injected into the gas grid is
one final product of its process. This system concept, the subject of
analysis presented in this paper, is also of interest to agencies in
Poland concerned with electricity and gas transmission and elec-
tricity production. There is great interest in this subject, associated
mainly with an increase the share of renewables in electricity
production year over year. The structure of power-generation
sources in the Polish power system, dominated by centralized
systems based on coal with limited flexibility, appears to support
this technology. The advantage of the system shown in Fig. 10 is its
simple structure, which includes devices currently available
commercially. The basic elements of the system are the electro-
lyzers and the storage tanks for the electrolysis products: hydrogen
and oxygen. The hydrogen is directed to a gas grid pipeline, while
the oxygen is transported away by tank trucks. The energy-storage
system is equipped with a water treatment installation to ensure
the quality of the water used in the electrolysis process. It is
assumed that the system is integrated with a renewable power
source whose generation capacity varies in time.
5.1. Selection of nominal power of electrolyzers
The basic input data necessary for determining the nominal
power of electrolyzers are the power characteristics of the RES
installation as a function of time and the assumed operating times
of the Power-to-Gas installation. The Power System is characterized
by variable load throughout the day, resulting from the changing
electricity demand. Within the day there are two main periods:
peak load and minimum load for system. Due to the specificity of
the National Power System (NPS), it seems reasonable to power the
electrolysis process during periods of lower energy demand (during
the so-called valley of the night). This solution is also economically
justifiable because the electricity market price is lower in this
period. According to the Polish Power Exchange yearly average
price of electricity in the period from 10pm until 6am in 2015
amounted around 29 V/MWh. For comparison, the average price of
electricity between 10am and 6pm was more than 56 V/MWh.
Similar considerations, concerning supplying of the Power-to-Gas
installation in the valley of the electricity demand, also apply in
the systems with predominance of nuclear energy which are
characterized by minimal flexibility in load change [97,98]. The
Polish power generation system based on coal is also inflexible in
this field. Continuity of production in large centralized production
systems is essential to maintain high national energy security. For
the simultaneous development of RES the investments in energy
storage systems are necessary [99].
164 J. Kotowicz et al. / Energy 118 (2017) 156e171

Fig. 10. Power-to-Gas installation.

The basic decision in the design process of energy storage sys-
tem is the determination of its size. The selected size of the storage
system intended for cooperate with a renewable resource should
enable to store a sufficiently large quantity of produced energy.
Oversizing of the storage system, however, can carry to adverse
economic consequences. For determination of size of the energy
storage system, it is proposed to use the so-called degree of storage
gs, which is defined as the ratio of the annual amount of electricity
directed to storing, Eel_s, to the annual total amount of electricity
generated in the RES installation, Eel_a:
Eel s
gs ¼ : (23)
Eel a
The hourly production values of a RES installation in the annual
working cycle are the basis of the analysis. The production used in
the present analysis was taken from a wind farm. The electricity
production varies both daily and annually. The annual power
characteristic of the RES installation is shown in Fig. 11. For sample
input data, the assumptions were a 15% degree of storage and an 8-
h work cycle of the storage installation in the night valley (between
10pm and 6am). With these assumptions, the nominal power of the
hydrogen generators, calculated iteratively, was 5.96 MW. The
nominal power may be the sum of the nominal powers of several
devices, forming a battery. The power supply of the hydrogen
generators is additionally outlined in grey in Fig. 11. Electricity
produced by the wind farm outside the night valley was directed to
the electrical grid. Between 10pm and 6am, the electrolysis process
is powered first. If RES production exceeds the nominal power of
the hydrogen generators, the surplus is directed into the power
grid. Therefore, the input power of the installation is limited by the
nominal power of hydrogen generator (5.96 MW) or the momen-
tary potential of the wind farm. As the power of the wind farm
varies, additional units can be turned on or off.
The similar characteristics of a randomly-selected week of the
RES installation are presented in Fig. 12. The nigh valleys are
marked by vertical dotted lines, while the production of the RES
installation directed to Power-to-Gas is shown in grey. The over-
capacity during night valleys and production outside these periods
are directed to the power grid.
The literature appears a number of other algorithms for sup-
plying hydrogen generators in collaboration with renewable energy
sources. In Ref. [100] the authors analyze the system with small-
scale wind turbine supplying generator hydrogen in off-grid
configuration and in grid-integrated mode and assume a constant
power of electrolysis system. The excess electricity is connected to
dump load or sold to the grid. In the case of less production of the
turbine than the nominal power of electrolyzer it is switched off. In
Ref. [101] a large scale wind-hydrogen plant is analyzed for the oil
sands industry. The capacity of wind is 563 MW. The authors

Fig. 11. A sample annual cycle characteristic of a power of RES installation with an outline of the input power of hydrogen generators (grey).
 J. Kotowicz et al. / Energy 118 (2017) 156e171 165

Fig. 12. The characteristics of the power of RES installation in selected week with marked areas of production directed to the supply of electrolysis installation (grey) and to the
power grid.

considered a few plant configuration with different models and

number of electrolyzer units. The simulation includes 8760 h to
yield and presents among others the annual hydrogen flow rate,
annual sold electricity, capacity factor and economic results of the
minimum hydrogen production cost.
The indicator for the assessment of a technical energy storage
system may be the rate of utilization of the working time of an
installation dHG, defined by the formula:

tE
dHG ¼ ; (24)
tN

where: tE - number of hours of working electrolyzers of installation

within a year, tN - number of hours provided for the operation of
the installation (duration of the night valleys throughout the year,
i.e., 365  8 h ¼ 2920 h).
One of the criteria that could serve for evaluating the benefits of
hydrogen generators with a specific nominal power is the utiliza-
tion rate dHG of their nominal power, defined by the equation:
ðNHG Þav
dHG ¼ ; (25)
ðNHG Þnom
where: ðNHG Þav - average power of installed hydrogen generators
achieved in the annual cycle, ðNHG Þnom - nominal power of the
installed generators.
For the analyzed case study, the rate of utilization of working
time of the installation, assuming a degree of storage of 15%,
calculated in accordance with the formula (24), was 92.2%.The
utilization rate of nominal power calculated on the basis of the
formula (25) was 74.1%.
The nominal input power of electrolyzer installations increases
with increasing degree of storage, while the utilization rate of
nominal power decreases, which is illustrated by Fig. 13. For the
analyzed case, assuming the Power-to-Gas system operates during
the night valleys, the maximum value of the degree of storage was
32.8%. It occurred when the power of the hydrogen generator was
equal to the maximum night valley power achieved by the wind
farm in the analyzed year.
The utilization rate of nominal power defined by equation (25)
applies to the annual cycle. This indicator can also be applied to
the daily cycle. Sorting the value of this rate for all days of the year
allows making an ordered chart. This type of graph for seven
selected degrees of storage is presented in Fig. 14. Comparison of
Fig. 13. The characteristics of nominal power of electrolyzers: installation and utili-
zation rate of nominal power as a function of the degree of storage.
their characteristics shows that increasing the share of energy
directed to the Power-to-Gas installation relative to total RES pro-
duction decreases the utilization rate of nominal power of
hydrogen generators. High degrees of storage are synonymous with
an adverse decrease in the number of days per year in which spe-
cific values of utilization rates of nominal power are achieved by an
installation. This is due to the high fluctuations in the power from
renewable energy sources, both in the daily and annual cycles.
5.2. Determination of the amount of hydrogen produced
On the bases of the calculated nominal input power of a Power-
to-Gas installation and the assumed rated power of a single
hydrogen generator, the number of devices can be determined.
ðN Þ
nHG   HG nom n2N; (26)
NHGj nom
where: nHG - number of electrolyzers of the installation, ðNHG Þnom -
the calculated nominal input power ðNHGj Þnom - the nominal input
power of a single hydrogen generator.
For our example, the number of required devices was deter-
mined assuming the nominal power of a single hydrogen generator
of 1.5 MW. Similarly, the number of hydrogen tanks was selected to
provide storage for the maximum daily production of hydrogen.
The assumed capacity of the tank was 100 m3. The tanks were
selected assuming that the maximum and minimum pressures of
166 J. Kotowicz et al. / Energy 118 (2017) 156e171
Fig. 14. The ordered chart of daily utilization rate of nominal power of hydrogen generators for different degrees of storage.
hydrogen storage were 3.5 MPa and 1.6 MPa. The numbers of
hydrogen generators and tanks as a function of the degree of
storage are shown in Fig. 15.
Due to the variability in the power-on time of the RES installa-
tion and in the amount of energy directed to storage, the load on
one of the electrolyzers may be changed. The power input to the
hydrogen generator system at a given moment is:
   
NHG ¼ ðnHG Þnom $ NHGj nom þ NHGj D ;
where: nHG nom - the number of hydrogen generators working with
nominal power (ðnHG Þnom  nHG ), ðNHGj Þnom - the nominal input
power to a single hydrogen generator,ðNHGj ÞD - the input power of a
hydrogen generator working under a load lower than nominal
value (0  ðNHGj ÞD <ðNHGj Þnom ).
The efficiency of a hydrogen generator working at partial load
changes with input power. On this basis, the chemical energy flux of
hydrogen produced at a given moment is determined by the
formula:
   
E_ chH2 ¼ ðhHG Þnom $ðnHG Þnom $ NHGj nom þ hHG $ NHGj D :
Fig. 15. The number of hydrogen generators and tanks of the installation as functions
of the degree of storage.
The efficiency of the generator at loads other than the nominal
hHG changes with the load in accordance with (18). For analysis the
authors adopted a ¼ 0.725; m ¼ 0.2 and dnom ¼ 0.075.
The value a has been selected on the basis of the technical data
of large power units, collected from several manufacturers
[94e96,102,103]. Efficiency of these hydrogen generators are fluc-
tuated around 70%. Therefore it is assumed that the newly designed
system should be characterized by high efficiency of 72.5% at
(27) nominal capacity of hydrogen generator.
The flux of hydrogen chemical energy ðV_ H2 Þ in m3n/s is then
determined using the formula:
E_ chH2
V_ H2 ¼ ; (29)
HHV DrH2
where: HHV e the higher heating value of hydrogen (141.8 MJ/kg),
and rH2 e hydrogen density in normal conditions (0.0899 kg/m3n).
A graph of the maximum volume flow of hydrogen ðV_ H2 Þmax ,
received in system operation with the calculated nominal power, as
a function of the degree of storage (gs), is shown in Fig. 16.
(28)
The annual production volume of hydrogen (VH2year) can be
Fig. 16. The characteristics of the maximum volume flow of hydrogen as a function of
the degree of storage.
 J. Kotowicz et al. / Energy 118 (2017) 156e171
determined by calculating the integral over a year:
Z
8760
VH2 year ¼ V_ H2 dt; (30)
0 There is no doubt that Power-to-Gas technologies can be an
In the case of the rated input power of the installation ðNHG Þnom
of 5.96 MW and the input power of a single gas generator ðNHGj Þnom
of 1.5 MW, the installation has 4 hydrogen generators. The annual
hydrogen production was 2 252 875 m3n. The relationship between
the annual production and the degree of storage is shown in Fig. 17.
5.3. Economic analysis
Power production in Poland is still based on coal- and lignite-
fired power plants [19]. Simultaneously, sources of power genera-
tion based on renewable resources have priority access to the po-
wer network. Despite the positive environmental impact of
promoting use of energy from renewable sources, this situation
contributes to adverse impacts on the power system. The priority of
the RES thrusts large conventional coal-fired units into the role of
marginal regulatory units that must adjust of their generation ca-
pacity to the current demand for electricity. The dynamically
changing relationship between the production capacity and the
size of the electricity demand contributes to the need for frequent
changes of load and shutdowns of coal-fired plants, which
adversely impacts their economic characteristics. This situation
presents a serious risk of disrupting the energy security of the
country. The power-system operator is obliged to create solutions,
including prevention mechanisms. It is important that such
mechanisms do not limit the potential of RES. Organization of
market mechanisms to enable economically justified operation of
energy storage systems can solve the problem of low elasticity of
coal blocks while enabling further growth in installed capacity from
renewable sources.
The Power-to-Gas installation can be beneficial for the elec-
tricity grid in a few fields.
It can meet the regulatory function by receiving excess energy
from the system. The energy storage, including Power-to-Gas may
contribute to reducing the number of shutdowns of the coal-fired
plants and thus to increase their average efficiency, reduce the
frequency of repairs and reduce costs associated with restarting of
units.
In addition, the storage of energy can help to the unloading of
power grids, which cooperate with the wind farm characterized by
fluctuation of electricity production. Thus, the modernization of the
network to increase its capacity may become less costly.
Natural gas with small amount of hydrogen (about 5% by vol-
ume) can be successfully used in combined cycle gas turbines. The
Fig. 17. The annual production of hydrogen depending on the degree of storage.
167
natural gas even with hydrogen is very flexible in terms of location,
the time of demand or energy use [104]. The enriched gas also can
be use in the autonomic gas turbines, working as a peak sources,
which perform a regulatory function in the system.
important mechanism in preventing these adverse effects of the
growth of RES. At the current stage of development of the tech-
nology, Power-to-Gas is characterized by relatively low efficiencies.
Its main advantages are few requirements for built-up space or for
site conditions (the suitable height differences or the geological
conditions that are required, e.g., in storage systems such as CAES or
pumped-storage power plants). At this writing, energy storage
technologies are of interest to the potential beneficiaries, such as
representatives of manufacturing industries, of the natural gas
transmission system, and entities specializing in electric energy
trading under the principles of safe and secure operation of the
National Power System. Technologies with a high utilitarian po-
tential during the market test phase have options of receiving
financial support. The technologies with high importance for na-
tional energy security can be covered by future financial support
mechanisms, such as adequate tariffs or certificates. Another
mechanism put forward for discussion is a link between the
amount of co-financing of construction and the declared profile of
the power of unstable energy sources such as wind farms. It should
be noted that in spite of great importance for the energy security
the financial support for storage systems is not guaranteed by law.
Obtaining additional income required to economically justified
operation of the energy storage systems is only possible by periodic
provision of services for the operator of system. Investment risk in
this case appears to be high. The results of the analysis, which was
carried out in this section indicate that support mechanisms are
essential to enable the development of energy storage systems.
The economic analyses of the energy storage system were
conducted, similarly to the technical analysis, for the full range of
possible values of the degree of storage. Within the scope of the
analyses, the break-even price of hydrogen was calculated from the
condition:
NPV ¼ 0: (31)
This value was determined from the relationship:
Pt¼N ½ðJDÞþKSO2 þTAL
t¼0
be ð1þrÞt
CH 2
¼ Pt¼N VH2 year $HHV ; (32)
t¼0 ð1þrÞt
where: J - the investment cost, D - the size of the grants, K- the
operating costs, SO2 - revenue from sales of oxygen, T - taxes (on
income and property), Aedepreciation, L - the liquidation value,
rethe discount rate, t - another year of the analyses, from t ¼ 0 (the
start of construction) to t ¼ N (the last year of operation).
The most important assumptions used for the economic analysis
were shown in Table 1.
The results of the economic analysis for the three levels of
project co-financing of 0%, 50% and 100% are shown in Figs. 18e20.
To assess the economic efficiency of this energy storage system,
the values obtained for the break-even prices of hydrogen can be
referred to the current market price of natural gas. In both cases,
the price must be understood as a cost for the purchase of units of
energy contained in the gas. The investment may be economically
effective, assuming a current market price of natural gas of 8 V/GJ,
only for the small storage systems (gs < 0:17) with 100% financing
and free electricity supplied to the system during the off-peak
periods. Note, however, that the hydrogen at its conversion into
electricity is not burdened with carbon dioxide emission. Due to the
168 J. Kotowicz et al. / Energy 118 (2017) 156e171

Table 1
Assumptions data for economic analysis.

Parameter Value

Grant variants 0/50/100%

Unit investment cost (for all system referred to the nominal 1350 V/kW
power of hydrogen generators)
Construction time 2 years
Construction cost distributed 30/70%
Time of the system operation 20 years
Discount rate 5%
The purchase price of electricity in off-peak demand period 0/10/20/30V/MWh
Price of water 1.1 V/m3
Operating cost 3.5 V/GJH2
Yearly maintenance costs 0.3% of the investment costs for the first 4 years of operation, 0.6% for the next four years,
0.9% during next four years, 1.2% during the next four years and 1.5% for the last four
years of operation
Price of oxygen 0.06 V/m3n
Value of the depreciation rate 6.7%
Property tax 2% of the capital expenditures incurred on the property while it is subject to depreciation,
and 2% of the market value of the property on 1 January of the last year of depreciation
The liquidation value of the installation 20% of the investment

Fig. 20. Break-even price of hydrogen for four different electricity purchase prices in
Fig. 18. Break-even price of hydrogen, for four different electricity purchase prices in the off-peak demand period as a function of the degree of storage in the absence of
the off-peak demand period as a function of the degree of storage with 100% financing. financing.

appearance of negative prices within the markets of some Euro-

pean countries contributed to the stimulation of interest in storage
technologies. Simplified analysis was conducted for determined
economic conditions - constant electricity prices during the night
valleys and peak periods. In real market conditions, where elec-
tricity prices are shaped by a number of macroeconomic factors, the
use of large storage systems, which can allow for storage of energy
over long periods, allows to conduct an optimal energy trade
strategy and for this purpose the use of the mechanism of price
arbitration. It is also to be noted that the further development of the
Power-to-Gas technology and renewable energy sources can lead to
a reduction in the investment costs. In literature [105] the lower
cost levels in an 85% renewable energy scenario are also analyzed:
1000, 750 and 500 V/kW.

Fig. 19. Break-even price of hydrogen for four different electricity purchase prices in
the off-peak demand period as a function of the degree of storage with 50% co-
financing.
existence Trading Emission System production of electricity from
hydrogen fuel can be cheaper than from natural gas. In an extreme
case of the analysis the price of electricity which occurs during the
night valley was assumed equal to 0. It should however be consider
the possibility of occurrence of negative prices, which can signifi-
cantly improve the economic efficiency of investment. The
6. Summary
Currently, there has been a significant increase in the popularity
of hydrogen production by electrolysis. Hydrogen generators with
PEM and alkaline electrolyzers have experienced the greatest in-
terest, as indicated by pilot installations. The greatest interest in
storage systems is in countries that have experienced the most
dynamic increase in installed capacity of wind or solar systems in
recent years. The core function of storage systems should be sta-
bilization of power systems while maintaining the full potential of
 J. Kotowicz et al. / Energy 118 (2017) 156e171
renewable energy sources and providing reassurance of the effec-
tive functioning of conventional generation.
The important issue for the analyses of energy storage systems
using electrolysis is the ability to determine the efficiencies of
hydrogen generators under changing loads. The literature often
presents analyses focusing on electrolyzers but ignores auxiliary
equipment such as pumps, fans, converters, transformers or control
systems. Energy analyses taking these elements into account allow
the determination of the potential efficiency of the work of
hydrogen generators belonging to the storage systems provided for
cooperation with unstable energy sources. The experimental
studies conducted by the present authors have shown that the total
auxiliary power for a test hydrogen generator under nominal load
constituted 7.2% of the power supply of the system. The reduction
of load leads to increasing electrolyzer efficiency. A constant level of
auxiliary power ensures that the reduction of load corresponds to
decreased efficiency of the hydrogen generator. The auxiliary po-
wer index at the load of 0.2(NHG)nom is approximately 36%. Will-
ingness to conduct analyses of the system, including the
multivariate analyses to determine the energy efficiency of storage
systems, creates a need for suitable models of hydrogen generators.
Through the generalized characteristics of the generator efficiency
proposed in this paper, the efficiency can be determined for any
load on the device. The proposed equation may constitute a basis
for the analysis of energy storage systems cooperating with un-
stable energy sources such as wind farms, where determination of
the efficiency of a hydrogen generator is needed in a whole range of
work with changing input power. Having obtained parameteriza-
tion, it is possible to match the form of equation to the character-
istics of devices available on the market.
A very important issue is the selection of the nominal power of
hydrogen generators cooperating with a specific RES installation. In
the paper, this quantity was determined as a function of the degree
of storage. The degree of storage was defined as the ratio of the
amount of electricity directed to storage to the total annual amount
of electricity generated in RES installation. The authors proposed, as
the criterion for evaluation of the legitimacy of installing hydrogen
generators with a specific nominal power, the utilization rate dHG of
their nominal power. The paper shows that the nominal power of
hydrogen generators, the utilization rate of their nominal power,
and the economic characteristics of the Power-to-Gas installation
are dependent on the degree of storage gs .
The use of an energy storage system in Power-to-Gas technology
certainly delivers positive effects in terms of the impact on the
electricity system. This is mainly because it mitigates a possible
adverse impact on the National Electricity System caused by
changes of RES production. Although hydrogen generators are
already available in the market, they are still characterized by
relatively low efficiencies. Despite its imperfections however, this
technology could become more popular.
Stimulating factors could be:
- the expected development of the technology of electrolyzers,
- the expected decrease in electrolyzer prices,
- the emergence of considerable decreases in energy prices in off-
peak periods, which will continue to experience an increasing
share of renewable energy in the structure of primary energy
consumption,
- introduction of support mechanisms, that increase the flexibility
in terms of available power for the network (subsidies, tariffs,
certificates).
The results of this economic analysis confirm the general view
that there is a need for the introduction of market mechanisms
supporting energy storage systems.
169
Acknowledgement
This research was financed within the project commissioned by
the National Center for Research and Development under the
GEKON program: "The energy storage in the form of hydrogen in
salt caverns", implemented in 2015e2016.
References
[1] REN21. Renewables 2014 global status report. Paris. ISBN 978-3-9815934-2-
6. 2014.
[2] Wyrwa A, Szurlej A, Gawlik L, Suwała W. Energy scenarios for Poland e a
comparison of PRIMES and TIMES e PL modeling results. J Power Technol
2015;95:100e6.
 ski B, Buczkowski R, Iglin
[3] Iglin  ska A, Cichosz M, Plaskacz-Dziuba M. SWOT
analysis of the renewable energy sector in Poland. Case study of Wielko-
polskie region. J Power Technol 2015;95(2):143e57.
[4] Gawlik L, Szurlej A, Wyrwa A. The impact of the long-term EU target for
renewables on the structure of electricity production in Poland. Energy
2015;92:172e8.
[5] Paska J, Surma T. Electricity generation from renewable energy sources in
Poland. Renew Energy 2014;71:286e94.
[6] Haas R, Resch G, Panzer C, Busch S, Ragwitz M, Held A. Efficiency and
effectiveness of promotion systems for electricity generation from renewable
energy sources e Lessons from EU countries. Energy 2011;36:2186e93.
[7] Shafiullah GM, Amanullah MTOo, Shawkat Ali ABM, Wolfs P. Potential
challenges of integrating large-scale wind energy into the power grideA
review. Renew Sustain Energy Rev 2013;20:306e21.
[8] Papaefthymiou G, Dragoon K. Towards 100% renewable energy systems:
uncapping power system flexibility. Energy Policy 2016;92:69e82.
[9] Shivashankar S, Mekhilef S, Mokhlis H, Karimi M. Mitigating methods of
power fluctuation of photovoltaic (PV) sources e a review. Renew Sustain
Energy Rev 2016;59:1170e84.
[10] Manditereza PT, Bansal R. Renewable distributed generation: the hidden
challenges e a review from the protection perspective. Renew Sustain En-
ergy Rev 2016;58:1457e65.
[11] Technology roadmap, hydrogen and fuel cells. France: International Energy
Agency; 2015.
[12] Tomczyk P. Prospects and obstacles of hydrogen economy development.
Polityka Energ 2009;12(2):593e607.
[13] Brouwer AS, van den Broek M, Seebregts A, Faaij A. Operational flexibility
and economics of power plants in future low-carbon power systems. Appl
Energy 2015;156:107e28.
[14] Huber M, Dimkova D, Hamacher T. Integration of wind and solar power in
Europe: assessment of flexibility requirements. Energy 2014;69:236e46.
[15] Rahmani M, Jaramillo P, Hug G. Implications of environmental regulation and
coal plant retirements in systems with large scale penetration of wind po-
wer. Energy Policy 2016;95:196e210.
[16] Hong L, Lund H, Mo € ller B. The importance of flexible power plant operation
for Jiangsu's wind integration. Energy 2012;41(1):499e507.
[17] Spicker S, Weber C. The future of the European electricity system and the
impact of fluctuating renewable energy e a scenario analysis. Energy Policy
2014;65:185e97.
[18] Atsonios K, Violidakis I, Sfetsioris K, Rakopoulos DC, Grammelis P, Kakaras E.
Pre-dried lignite technology implementation in partial load/low demand
cases for flexibility enhancement. Energy 2016;96:427e36.
[19] Kamin  ski J, Kudełko M. The prospects for hard coal as a fuel for the Polish
power sector. Energy Policy 2010;38:7939e50.
[20] Badyda K. Trends, conditions and prospects for the construction of new gas
turbine combined cycle plants in Poland. Rynek Energii 2013;5:26e33.
[21] Wołowicz M, Badyda K. Gas turbine selection for feedwater repowering.
J Power Technol 2015;95(4):302e8.
[22] Chmielniak T, Lepszy S, Czaja D. Economic assessment of gas-steam systems
taking account of variable loads. J Power Technol 2015;95:54e62.
[23] Despre s J, Mima S, Kitous A, Criqui P, Hadjsaid N, Noirot I. Storage as a
flexibility option in power systems with high shares of variable renewable
energy sources: a POLES-based analysis. Energy Econ 2016. http://dx.doi.org/
10.1016/j.eneco.2016.03.006.
[24] Kondziella H, Bruckner T. Flexibility requirements of renewable energy
based electricity systems e a review of research results and methodologies.
Renew Sustain Energy Rev 2016;53:10e22.
[25] Steffen B, Weber C. Efficient storage capacity in power systems with thermal
and renewable generation. Energy Econ 2013;36:556e67.
[26] Bertsch J, Growitsch C, Lorenczik S, Nagl S. Flexibility in Europe's power
sector d an additional requirement or an automatic complement? Energy
Econ 2016;53:118e31.
[27] Gahleitner G. Hydrogen from renewable electricity: an international review
of power-to-gas pilot plants for stationary applications. Int J Hydrogen En-
ergy 2013;38:2039e61.
[28] Guandalini G, Campanari S, Romano MC. Power-to-gas plants and gas tur-
bines for improved wind energy dispatchability: energy and economic
assessment. Appl Energy 2015;147:117e30.
170 J. Kotowicz et al. / Energy 118 (2017) 156e171
[29] Bussar C, Sto€ cker P, Cai Z, et al. Large-scale integration of renewable energies
and impact on storage demand in a European renewable power system of
2050dsensitivity study. J Energy Storage 2016;6:1e10.
[30] Bouman EA, Øberg MM, Hartwich EG. Environmental impacts of balancing
offshore wind power with compressed air energy storage (CAES). Energy
2016;95:91e8.
[31] Kotowicz J, Jurczyk M. Efficiency of diabatic CAES installation. Rynek Energii
2015;119(4):49e56.
[32] Chen H, Cong TN, Yang W, Tan C, Li Y, Ding Y. Progress in electricity energy
storage system: a critical review. Prog Nat Sci 2009;19:291e312.
[33] Luo X, Wang J, Dooner M, Clarke J. Overview of current development in
electrical energy storage technologies and the application potential in power
system operations. Appl Energy 2015;137:511e36.
[34] Lund H, Salgi G, Elmegaard B, Andersen AN. Optimal operation strategies of
compressed air energy storage (CAES) on electricity spot markets with
fluctuating prices. Appl Therm Eng 2009;29:799e806.
[35] Auer H, Haas R. On integrating large shares of variable renewables into the
electricity system. Energy 2016. http://dx.doi.org/10.1016/
j.energy.2016.05.067.
[36] He H, Chang M, Chen A, Ma L, Lu J, Li C. Enhanced hydrogen evolution
process of water electrolysis assisted by photocatalysis. Materials for
renewable energy & environment (ICMREE). In: International conference,
Shanghai, 20e22 May 2011; 2011. http://dx.doi.org/10.1109/
ICMREE.2011.5930608.
[37] Jiang W, Wu YK, Yang T, Yu FY, Wang W. Identification and Power Electronic
Module Design of a solar powered hydrogen electrolyzer. In: Asia-pacific
power and energy engineering conference; Shanghai, 27e29 March 2012;
2012. http://dx.doi.org/10.1109/APPEEC.2012.6307202.
[38] Paska J, Pawlak K, Surma T. Support systems as an essential element of the
impact optimization of new, green sources of electricity on the power sys-
tem. Rynek Energii 2013;2(105):48e53.
[39] Leading the energy transition. Fact book: electricity storage. SBC Energy
Institute; September 2013.
[40] Mirek P, Nowak W. OZE w systemie i co dalej? Potencjał i ewolucja układo w
magazynowania energii cz. 1. Energ Ciepl i Zawodowa 2015;4:8e12.
[41] Mirek P, Nowak W. W ogonie za Stanami. Potencjał i ewolucja układo w
magazynowania energii cz. 2. Energ Ciepl i Zawodowa 2015;5:64e6.
[42] Zoulias E, Varkaraki E, Lymberopoulos N, Christodoulou CN, Karagiorgis GN.
A review on water electrolysis. Project report “Cluster pilot project for the
integration of renewable energy sources into European energy sectors using
hydrogen”. Pikermi, Greece: Centre for Renewable Energy Sources (CRES);
2006.
[43] Bhattacharyya S., Cobben S., Consequences of poor power quality e an
overview, Power quality, Andreas Eberhard (Ed.), InTech, ISBN: 978-953-
307-180-0, Available from: http://www.intechopen.com/books/power-
quality/consequences-of-poor-power-quality-an-overview.
[44] We˛ cel D, Ogulewicz W, Kotowicz J, Jurczyk M. Dynamics of electrolysers
operation during hydrogen production. Rynek Energii 2016;122(1):59e65.
[45] Keatley P. Cost modelling of coal power plant start-up in cyclical operation.
Coal Power Plant Materials and Life Assessment. Elsevier; 2014. p. 358e88.
[46] Hentschel J, Babi c U, Spliethoff H. A parametric approach for the valuation of
power plant flexibility options. Energy Rep 2016;2:40e7.
[47] van den Bergh K, Delarue E. Cycling of conventional power plants: technical
limits and actual costs. Energy Convers Manag 2015;97:70e7.
[48] Singh S, Jain S, Venkateswaran PS, Tiwari AK, Nouni MR, Pandey JK, et al.
Hydrogen: a sustainable fuel for future of the transport sector. Renew Sus-
tain Energy Rev 2015;51:623e33.
[49] Taamallah S, Vogiatzaki K, Alzahrani FM, Mokheimer EMA, Habib MA,
Ghoniem AF. Fuel flexibility, stability and emissions in premixed hydrogen-
richgas turbine combustion: technology, fundamentals, and numerical sim-
ulations. Appl Energy 2015;154:1020e47.
[50] Chiesa P, Lozza G, Mazzocchi L. Using hydrogen as gas turbine fuel. J Eng Gas
Turbines Power 2005;127(1):73e80.
[51] Ulleberg Ø, Nakken T, Ete A. The wind/hydrogen demonstration system at
Utsira in Norway: evaluation of system performance using operational data
and updated hydrogen energy system modeling tools. Int J Hydrogen Energy
2010;35:1841e52.
[52] Gassner M, Mare chal F. Thermo-economic optimisation of the integration of
electrolysis in synthetic natural gas production from Wood. Energy 2008;33:
189e98.
[53] Martínez I, Romano MC. Flexible sorption enhanced gasification (SEG) of
biomass for the production of synthetic natural gas (SNG) and liquid bio-
fuels: process assessment of stand-alone and power-to-gas plant schemes
for SNG production. Energy 2016;113:615e30.
[54] Kato T, Kubota M, Kobayashi N, Suzuoki Y. Effective utilization of by-product
oxygen from electrolysis hydrogen production. Energy 2005;30:2580e95.
[55] Breyer C, Tsupari E, Tikka V, Vainikka P. Power-to-Gas as an emerging
profitable business through creating and integrated value chain. Energy
Procedia 2015;73:182e9.
[56] Carmo M, Fritz DL, Mergel J, Stolten D. A comprehensive review on PEM
water electrolysis. Int J Hydrogen Energy 2013;38(12):4901e34.
[57] Briguglio N, Brunaccini G, Siracusano S, et al. Design and testing of a compact
PEM electrolyzer system. Int J Hydrogen Energy 2013;38(26):11519e29.
[58] Milewski J, Guandalini G, Campanari S. Modeling an alkaline electrolysis cell
through reduced-order and loss-estimate approaches. J Power Sources
2014;269:203e11.
[59] Ziems C, Tannert D, Krautz HJ. Project presentation: Design and installation
of advanced high pressure alkaline electrolyzer-prototypes. Energy Procedia
2012;29:744e53.
[60] Hammoudi M, Henao C, Agbossou K, et al. New multi-physics approach for
modelling and design of alkaline electrolyzers. Int J Hydrogen Energy
2012;37(19):13895e913.
[61] Milewski J, Szczesniak A, Lewandowski J. Dynamic characteristics of auxiliary
equipment of SOFC/SOEC hydrogen peak power plant. IERI Procedia 2014;9:
82e7.
[62] Sanz-Barmejo J, Mun ~ oz-Anto n J, Gonzalez-Aguilar J, Romero M. Part load
operation of a solid oxide electrolysis system for integration with renewable
energy sources. Int J Hydrogen Energy 2015;40(26):8291e303.
[63] Stempien JP, Sun Q, Chan SH. Performance of power generation extension
system based on solid-oxide electrolyzer cells under various design condi-
tions. Energy 2013;55:647e57.
[64] Odukoya A, Naterer GF, Roeb M, Mansilla C, Mougin J, Yu B, et al. Progress of
the IAHE nuclear hydrogen division on international hydrogen production
programs. Int J Hydrogen Energy 2016;41(19):7878e91.
[65] Kupecki J, Motylin  ski K, Ferraro M, Sergi F, Zanon N. Use of NaNiCl battery for
mitigation of SOFC stack cycling in base-load telecommunication power
systemda preliminary evaluation. J Power Technol 2016;96(1):63e71.
[66] Szcze˛ sniak A, Milewski J. The reduced order model of a proton-conducting
solid oxide fuel cell. J Power Technol 2014;94(2):122e7.
[67] Mansilla C, Sigurvinssona J, Bontemps A, Mare chal A, Werkoffa F. Heat
management for hydrogen production by high temperature steam electrol-
ysis. Energy 2007;32:423e30.
[68] Ayers KE, Anderson EB, Capuano CB, Carter BD, Dalton LT, Hanlon G, et al.
Research advances towards low cost, high efficiency PEM electrolysis. ECS
Trans 2010;33(1):3e15. http://dx.doi.org/10.1149/1.3484496 [© The Elec-
trochemical Society].
[69] Smolinka T. Cost break down and analysis of PEM electrolysis systems for
different industrial and Power to Gas applications. Stuttgart (DE):
Fraunhofer-Institut für Solare Energiesysteme ISE, World of Energy Solu-
tions; October 12, 2015.
[70] Final Report: Systems Analyses Power to Gas, Deliverable 1: Technology
Review DNV KEMA energy & sustainability. Jun. 2013.
[71] Ursua A, Gandia LM, Sanchis P. Hydrogen production from water electrolysis:
current status and future trends. IEEE February 2012;100(2). 0018e9219.
[72] Basile A, Iulianelli A. Advances in hydrogen production, storage and distri-
bution. Woodhead Publishing; 2014. ISBN 978-0-85709-768-2.
[73] Tsiplakides D., PEM water electrolysis fundamentals. Available: http://
research.ncl.ac.uk/sushgen/docs/summerschool_2012/PEM_water_
electrolysis-Fundamentals_Prof._Tsiplakides.pdf.
[74] Harrison KW, Remick R, Martin GD, Hoskin A. Hydrogen production: fun-
damentals and gas study summaries. In: The 18th world hydrogen energy
conference, essen, Germany; May 16e21, 2010.
[75] Godula-Jopek A. Hydrogen production by electrolysis. Wiley-VCH, Verlag
GmbH & Co. KGaA; 2015.
[76] Zhang H, Su S, Lin G, Chen J. Efficiency calculation and configuration design
of a PEM electrolyzer system for hydrogen production. Int J Electrochem Sci
2012;7:4143e57.
[77] Selamet OF, € Becerikli F, Mat MD, Kaplan Y. Development and testing of a
highly efficient proton exchange membrane (PEM) electrolyzer stack. Int J
Hydrogen Energy 2011;36(17):11480e7.
[78] Stolten D. Hydrogen and fuel cells: fundamentals, technologies and appli-
cations. John Wiley & Sons; 2010.
[79] Mori M, Mr zljak T, Drobni c B, Sekav cnik M. Integral characteristics of
hydrogen production in alkaline electrolysers. J Mech Eng 2013;59:585e94.
[80] Ursua A, Marroyo L, Gubia E, Gandia LM, Dieguez PM, Sanchis P. Influence of
the power supply on the energy efficiency of an alkaline water electrolyser.
Int J Hydrogen Energy 2009;34:3221e33.
[81] User manual, Helicentris, Rev. 3.0 Hydrogen generator EL 500/dryer. October
2015.
[82] Blyth P. Understanding AC/DC power supply efficiency e the hot topic. The
XPerts in Power. January 2010.
[83] We˛ cel D, Ogulewicz W. Identification influence of the electrolyser power
supply method on efficiency hydrogen fuel production. Rynek Energii
2010;89(4):77e82.
[84] Smolinka T. Water electrolysis: status and potential for development. Brus-
sels (BE): Joint NOW GmbH e FCH JU, Water Electrolysis Day; April 03, 2014.
[85] Smolinka T, Garche J, Hebling C, Ehret O. Overview on water electrolysis for
hydrogen production and storage. In: SYMPOSIUM - water electrolysis and
hydrogen as part of the future renewable energy system Copenhagen/
Denmark; May 10, 2012.
[86] Bertuccioli L, Chan A, Hart D, Lehner F, Madden B, Standen E. Study on
development of water electrolysis in the European Union. Final Report. Fuel
Cells and Hydrogen Joint Undertaking. February 2014.
[87] Mazloomi K, Sulaiman N, Moayedi H. Electrical efficiency of electrolytic
hydrogen production. Int J Electrochem Sci 2012;7:3314e26.
[88] Millet P., Hydrogen production by polymer electrolyte membrane water
electrolysis. University of Paris-Sud, Orsay Cedex, France
[89] Smolinka T, Gunther M, Garche J. Stand und Entwicklungspotenzial der
Wasserelektrolyse zur Herstellung von Wasserstoff aus Regenerativen
Energien. Berlin: Fraunhofer ISE, NOW-Studie; 2011.
 J. Kotowicz et al. / Energy 118 (2017) 156e171
[90] Ogulewicz W, We˛ cel D, Wiciak G, Łukowicz H, Kotowicz J, Chmielniak T.
Pozyskiwanie energii z ogniw paliwowych typu PEM chłodzonych ciecza˛ .
Gliwice: Wydawnictwo Politechniki Sla  ˛ skiej; 2010.
[91] Kotowicz J, Ogulewicz W, We˛ cel D, Jurczyk M. Analysis of hydrogen elec-
trolyzer work. In: The 19th annual International conference energy and
environment, Ostrava; 2015. p. 44e7.
[92] Siracusano S, Baglio V, Briguglio N, Brunaccini G, Di Blasi A, Stassi A, et al. An
electrochemical study of a PEM stack for water electrolysis. Int J Hydrogen
Energy 2012;37:1939e46.
[93] Reichert F. Wind-to-Gas-to-Money? Economics and perspectives of the
power-to-gas technology. Aalborg University; 2012.
[94] PEM* Electrolysers, AREVA H2Gen, http://www.arevah2gen.com/en/
products-services.
[95] High Capacity Hydrogen Systems M Series PEM Electrolysers, Proton OnSite,
http://protononsite.com/.
[96] HGas Hydrogen Generator Energy Storage Solutions, product information
sheet, ITM Power, http://www.itm-power.com/.
[97] Walker SB, Fowler M, Ahmadi L. Comparative life cycle assessment of power-
to-gas generation of hydrogen with a dynamic emissions factor for fuel cell
vehicles. J Energy Storage 2015;4:62e73.
[98] Ryazantsev E, Chabak A. Hydrogen production, storage, and use at nuclear
power plants. At Energy 2006;101(6):876e81.
171
[99] Barelli L, Desideri U, Ottaviano A. Challenges in load balance due to renew-
able energy sources penetration: the possible role of energy storage tech-
nologies relative to the Italian case. Energy 2015;93:393e405.
[100] Go€kçek M. Hydrogen generation from small-scale wind-powered electrolysis
system in different power matching modes. Int J Hydrogen Energy 2010;35:
10050e9.
[101] Olateju B, Monds J, Kumar A. Large scale hydrogen production from wind
energy for upgrading of bitumen from oil sands. Appl Energy 2014;118:
48e56.
[102] Power-to-Gas: how to convert green power into green gas via water elec-
trolysis, Denis Thomas, Hydrogenics. Green Gas Research Outlook Sweden;
2015. http://www.hydrogenics.com/home.
[103] SILYZER 200 High-pressure efficiency in the megawatt, Siemens, ran-
gehttp://www.industry.siemens.com/topics/global/en/pem-electrolyzer/
silyzer/Pages/silyzer.aspx.
[104] Walker SB, Mukherjee U, Fowler M, Elkamel A. Benchmarking and selection
of Power-to-Gas utilizing electrolytic hydrogen as an energy storage alter-
native. Int J Hydrogen Energy 2016;41:7717e31.
[105] Jentsch M, Trost T, Sterner M. Optimal use of Power-to-Gas energy storage
systems in an 85% renewable energy scenario. Energy Procedia 2014;46:
254e61.