i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.elsevier.com/locate/he

Control strategy of electrolyzer in a wind-hydrogen

system considering the constraints of switching
times

Ruiming Fang*, Yin Liang

Department of Electrical Engineering, Huaqiao University, Xiamen, 361021, China

article info abstract

Article history: In order to improve the operational performance of alkaline electrolyzers powered by wind
Received 3 February 2019 power, the influences of the fluctuating wind power on alkaline electrolyzers must be
Received in revised form taken in accounts. In pursuing this goal, the influences of fluctuating wind power on both
23 February 2019 the hydrogen production and the self-safety of alkaline electrolyzer is discussed firstly.
Accepted 4 March 2019 And a wind-hydrogen integrated energy system (WHIES) integrated supercapacitor is
Available online 26 March 2019 designed to smooth the fluctuation of wind power. In which, the fluctuation of wind power
is divided into two kinds, instantaneous fluctuation and wide power fluctuation, the
Keywords: former is absorbed by supercapacitors, the latter is overcomed by adopting a modular
Wind-hydrogen system adaptive control strategy to optimize the operation mode of alkaline electrolyzer. A
Alkaline electrolyzer simulation has been developed for a specific wind farm located in Northeast China. The
Fluctuation of wind power simulation results show that under the condition of fluctuating wind power, the WHIES
Optimization control with the proposed control strategy can reduce the switching times of electrolyzers by 93.5%
Supercapacitor and increase hydrogen production by more than 44.18% when compared with other control
strategies.
© 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

renewable energy with energy storage system [8e10].

Introduction Hydrogen energy storage is a large-scale energy storage
method developed by utilizing the interdependence of electric
With the increasingly prominent problems of energy, envi- power and hydrogen energy. It is considered as one of the key
ronment and climate change, it is of great significance to technologies to solve the energy crisis and to absorb renew-
vigorously develop renewable energy [1e3]. Renewable energy able energy. In recent years, it has attracted much attention
generation, especially wind power, has the characteristics of [11e16]. The hydrogen energy storage system mainly includes
randomness and intermittence, it will bring about power the production, storage, transportation and utilization of
quality problems such as voltage fluctuation and frequency hydrogen. Hydrogen storage system with renewable energy
fluctuation when directly connected to power grid, and may is the main utilization mode at present. Storing and down-
even affect the safety and stability of the power grid [4e7]. In stream application of stochastic fluctuating renewable energy
order to ensure the safe and economic operation of power grid generation through electrolysis hydrogen production can
with large-scale renewable energy, it is necessary to cooperate

* Corresponding author.
E-mail address: [email protected] (R. Fang).
https://doi.org/10.1016/j.ijhydene.2019.03.033
0360-3199/© 2019 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1 25105

absorb excess renewable energy such as wind energy, which is

an effective solution to alleviate the power supply pressure of Influence of fluctuation wind power in
power grid [17]. Hydrogen produced by electrolytic hydrogen electrolyzer
production is widely used as an energy carrier and it can be
used directly in the downstream hydrogen industry chain, for The types of electrolyzer mainly include proton exchange
example, it can be transported into the hydrogenation station membrane electrolyzer (PEME) [24], alkaline electrolyzer (AE)
to supply fuel cell vehicles, or be converted into electricity and [25], solid oxide electrolyzer (SOE) [26], etc. Ideally, the elec-
heat energy. trolyzer can operate at rated power with a power fluctuation
Electrolytic hydrogen production is a mature, efficient tolerance of 0e100%. With the progress of science and tech-
and clean hydrogen production method, but its cost is nology, the minimum operation power of alkaline electrolyzer
expensive. In areas rich in renewable energy such as wind is limited to 5%e10% of the total electric power, and the speed
power, the phenomenon of wind power abandonment is of regulation reaches milliseconds, which is more suitable for
always serious. However, when wind power is combined large-scale wind-hydrogen coupling system. Therefore, alka-
with water electrolysis hydrogen production, it can not only line electrolyzer is taken as the research object of this paper.
solve the problem of wind power absorption, but also reduce
the cost of water electrolysis hydrogen production. However,
Working principle of alkaline electrolyzer
wind power has the characteristics of randomness and
Alkaline electrolyzer uses KOH, NaOH aqueous solution as
instability. When it is directly connected to the electrolyzer
electrolyte, asbestos cloth as diaphragm. Under the action of
as power source, its input fluctuating current and voltage
direct current, water is electrolyzed to generate hydrogen and
will cause the random drift of the working point of the
oxygen. The produced gas needs desalination fog treatment.
electrolyzer, which will affect the hydrogen production, the
Alkaline electrolyzer was industrialized in the mid-20th cen-
life and efficiency of the electrolyzer [18e20]. Therefore, it is
tury. The technology is mature, and its service life can be over
one of the key technologies for renewable energy integrating
15 years. The structure of alkaline electrolyzer is character-
with hydrogen storage to study the influence of fluctuating
ized by liquid electrolyte and porous separator [27], as shown
wind power on hydrogen production and electrolyzers life,
in Fig. 1.
and explore how to keep electrolyzers safe, stable and
The voltage model of alkaline electrolyzer can be expressed
efficient running under fluctuating input power. In Ref. [21],
as follows [28]:
Ulleberg proposed a mathematical model for an advanced
alkaline electrolyzer which considers the operational tem-

r1 þ r2 T t1 þ t2 =T þ t3 =T2
perature as main variable. Based on the Ulleberg's model, Uel ¼ Urev þ Iel þ s log Iel þ 1 (1)
A A
Die'guez et al. evaluated the thermal performance of an
alkaline electrolyzer under different operating conditions Where Uel , the working voltage of alkaline electrolyzer; A，
and found that the power supply of alkaline electrolyzer has electrode area; T, the operating temperature of alkaline elec-
a significant impact on the temperature of electrolyzer [22]. trolyzer, K ; Iel , the working current of alkaline electrolyzer; r1 ;
The above researches are mainly related to solar-hydrogen r2 , internal resistance of alkaline electrolyzer; s; t1 ; t2 ; t3 , rele-
coupling system. Valenciaga & Evangelista developed a vant parameters of overvoltage caused by polarization of
control scheme for a stand-alone wind based hydrogen electrodes and electrolytes in electrolyzer; Urev , reversible
production system to increase the hydrogen production with open circuit voltage, which is calculated as follows [28]:
less impact on the operative life [23]. DG DH  TDS
Nevertheless, the situation is become more complicated Urev ¼ ¼ (2)
2F 2F
for grid-connected wind-hydrogen coupled system, and there
is an urgent need for a control strategy that is suitable for grid-
connected wind-hydrogen system. Thus, this study provides a
grid-connected wind-hydrogen system integrated with ca-
pacitors which is useful in smoothing the influences of fluc-
tuating wind power on alkaline electrolyzers. On this basis, a
modular adaptive strategy is given to control alkaline elec-
trolyzers which can reduce the switching times of electro-
lyzers while increasing hydrogen production. The rest of the
paper is structured as follows. The influence of wind power on
alkaline electrolyzers under instantaneous fluctuation and
wide power fluctuation is discussed respectively in section
Influence of fluctuation wind power in Electrolyzer. Section
Control Strategy of Hydrogen Production by Wind Power
provides a detailed description of the structure and the control
strategy of a grid-connected wind-hydrogen system. In Sec-
tion Simulation and discussion, the simulation results based Fig. 1 e Schematic diagram of working principle of alkaline
on different control strategies are compared to valid the electrolyzer.
proposed approach, and finally Section Conclusions expounds
the main conclusions.  The model of alkaline electrolyzer
25106 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1

where DG, the Gibber's change value, J=mol; DH, the enthalpy's into heat. It is difficult for the electrolyzer to maintain the
change value, J=mol; DS, the entropy 's change value, J=kmol; F, rated working temperature, resulting in the deviation of the
Faraday constant, F ¼ 96485C=mol. electrolyzer from the rated working state, and the reduction of
The hydrogen production of alkaline electrolyzer is as hydrogen production and hydrogen production efficiency.
follows [28]:
" # Influence of wind power fluctuation on the safety of alkaline
a2 þ a3 T a4 þ a5 T Nel Iel electrolyzer
QH2 el ¼ a1 exp þ (3)
Iel =A ðIel =AÞ2 zF
In this works, the KZDQ alkaline electrolyzer is selected as an
where ai ; i ¼ 1; 2; /; 5is the relative coefficient of Faraday ef-
example. Owing to the small amount of hydrogen and oxygen
ficiency; Nel , the number of electrolytic cells in series; z, the
permeating through the diaphragm, there are oxygen purity
number of electrolytic cells in series.
analyzer and oxygen analyzer in hydrogen in alkaline elec-
Usually, the working current density of alkaline electro-
trolyzer. In view of the explosion limit of hydrogen in oxygen
lyzer is about 0.25 A/cm2, and the energy efficiency is usually
is 4%e95% [29], its monitoring range is 0%e2%. When
about 60%. In the liquid electrolyte system, the alkaline elec-
exceeding the set range, the hydrogen production system will
trolyte (such as KOH) used reacts with CO2 in the air to form
automatically protect and shut down. Oxygen purity in
insoluble carbonates, such as K2CO3, under alkaline condi-
hydrogen is generally superior to hydrogen purity in oxygen
tions. These insoluble carbonates will block the porous cata-
based on the characteristics of water electrolysis hydrogen
lytic layer, hinder the transfer of products and reactants, and
plant and the strong permeability of hydrogen. Therefore,
greatly reduce the performance of the electrolytic cell. On the
hydrogen purity in oxygen is used as the most important
other hand, alkaline liquid electrolyte cell is difficult to shut
index to measure the safety of electrolyzer in the industry.
down or start up quickly, and the speed of hydrogen produc-
The relationship between hydrogen purity in oxygen of alka-
tion is difficult to adjust quickly, because the pressure on both
line electrolyzer and continuous reduction of wind power is
sides of the anode and cathode of the cell must be kept
tested and shown in Fig. 2.
balanced at all times to prevent hydrogen and oxygen gas
Some conclusions can be drawn from the Fig. 2:
from mixing through porous asbestos membrane, thus
causing explosion. Therefore, it is necessary to optimize the
 Under rated power, the purity of hydrogen in oxygen is
control strategy of alkaline electrolyzer to prevent rapidly and
stable at about 1.05%.
frequently switching when cooperating with renewable en-
 Without changing other process parameters, the input
ergy sources with fast fluctuation characteristics.
power of alkaline electrolyzer is adjusted to 75% of the
When wind power is connected to alkaline electrolyzer as
rated power, and the purity of hydrogen in oxygen
power source, its power fluctuation characteristics can be
decreases by about 0.35% points with power reduction
divided into instantaneous fluctuation and wide power fluc-
within 1 h, and then tends to be stable.
tuation. The instantaneous fluctuation makes the input cur-
 Without changing other process parameters, the input
rent fluctuate frequently, and the diaphragm will move
power of alkaline electrolyzer is adjusted to 50% of the
repeatedly in the neutral region. The change of relaxation and
rated power, the purity of hydrogen in oxygen decreases to
tightness of the diaphragm results in the change of stress,
about 2% within 2 h with power reduction, reaching the
which tends to tighten under low current and high liquid level
upper limit of safe operation of hydrogen production
pressure difference, and the decrease of its porosity and
system.
permeability may lead to the damage of the diaphragm per-
 Without changing other process parameters, the input
formance and affect the operation life of the electrolyzer.
power of alkaline electrolyzer is adjusted to 25% of the
Furthermore, frequent changes in current will increase the
rated power, the purity of hydrogen in oxygen decreases to
mass concentration of KOH electrolyte. Once the mass con-
about 2% within less 1 h with power reduction, which is
centration is greater than the critical mass concentration
dangerous for alkaline electrolyzers.
(about 140 g/L), the side effects of the alkaline electrolyzer will
increase, the current efficiency will decrease, and the opera-
tion life of the electrolyzer will be further shortened. In addi-
tion, with the fluctuation of input power, the output of
hydrogen and oxygen on both sides of the electrolyzer will
change, resulting in the change of the liquid level on both
sides of the electrolyzer, which causes the frequent start-up of
the alkali liquid level balance device in the electrolyzer,
thereby affecting the service life of components such as
pneumatic valves or solenoid valves, increasing the power
consumption of auxiliary machines, and reducing the
efficiency of hydrogen production by electrolysis.
Under the condition of wide power fluctuation input, that
is, when the input power continuously increases or decreases
over a long period of time, most of the electric energy beyond Fig. 2 e The relationship between hydrogen concentration
the rated input range of alkaline electrolyzer will be converted in oxygen and input power.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1 25107

The relationship between hydrogen purity in oxygen of lies in the purpose and requirements for the electrolyzer. For
alkaline electrolyzer and operating temperature of alkaline grid-connected wind power, coupled hydrogen energy system
electrolyzer under different input power is tested and shown is only an auxiliary system, the use of grid-connected wind
in Fig. 3. power to produce hydrogen is to maximize the use of waste
Some conclusions can be drawn from the Fig. 3: wind power to produce hydrogen, while optimizing the qual-
ity of grid-connected wind power. And for off-grid wind
 Input rated power without changing other process pa- power, to produce hydrogen is the main purpose of hydrogen,
rameters, and then adjust the operating temperature of it can also be used as an energy storage link in off-grid energy
hydrogen production system within the allowable range, system. The scale of hydrogen production by grid-connected
the purity of hydrogen in oxygen is stable within 1.05%. wind power is generally large, because the grid can supply
 Adjust the input power to 75% rated power, and keep the power in reverse when the wind power cannot output useful
other parameters unchanged, then adjust the operating power, the electrolyzer can work in a better state. So, in this
temperature of alkaline electrolyzer within the allowable paper, the technical feasibility of grid-connected wind-
range. Temperature rises by 10 o c, while the change of hydrogen coupled energy system is analyzed only.
hydrogen concentration in oxygen does not exceed 0.02%. The architecture of a grid-connected wind-hydrogen
coupled energy system is shown in Fig. 5.

Influence of wind power fluctuation on the hydrogen Main operation mode of the proposed system
production of alkaline electrolyzer
The main operation modes include power demand retention
Continuously reduce the input power, and the relationship mode and power demand balance control mode.
between the input power of alkaline electrolyzer and hydrogen
production and power consumption are shown in Fig. 4. 1) Power demand balance control mode
As shown in Fig. 4, with the decrease of the input Gong rate
of the electrolyzer, the hydrogen production of the electro- The purpose of this model is to maintain the stability of
lyzer decreases and the unit power consumption decreases grid-connected wind power and improve the quality of grid-
dramatically. connected wind power. In this mode, the order of the moni-
The above experiments show that under the condition of toring system is to require wind turbines to be connected to
wide power fluctuation input, controlling power fluctuation the grid according to the given power of grid operators, that is,
within a certain range can significantly reduce the energy wind turbines have idle wind power. As shown in Fig. 6. When
consumption of alkaline electrolyzer. The main reasons the idle wind power is small, it is used to produce hydrogen by
include two aspects: one is the influence of overvoltage, which electrolyzer. When the idle wind power is large, the electro-
is because the overvoltage of hydrogen and oxygen decreases lyzer operates at the maximum rated power, and the
with the decrease of current density, and is linearly related to remaining wind power is absorbed by supercapacitors. When
the logarithm of current density; the other is resistance voltage the output of wind turbines is not larger than the power given
effect, that is, with the decrease of current density, the ohmic by grid operators, the operation mode is the same as the
voltage of electrolyte, diaphragm and metal conductor de- maximum power generation mode required by grid. Wind
creases, and the bubble effect in electrolyte plays a leading role. turbines try their best to transmit the generated energy to the
network. Super capacitor discharges to ensure that the elec-
trolyzer can operate in the minimum current mode and
Control strategy of hydrogen production by wind prevents frequent switching of alkaline electrolyzer.
power
2) Demand Power Retention Model of Power Grid
Wind-hydrogen coupled energy system can be divided into
two types: grid-connected and off-grid. The main difference When the penetration power of wind power exceeds 10%,
grid operators hope that wind farms can help balance and
regulate the grid when the grid breaks down, which requires
wind farms to operate in a certain capacity retention mode.
The larger the reserve, the better the emergency safety of
power grid, but the larger the reserve, the waste of wind
energy will be caused. By coupling the hydrogen energy sys-
tem, some or all of the reserved amount is used to produce
hydrogen from electrolyzer. When the power grid fails and the
control system automatically limits the power input of the
electrolyzer, supercapacitor discharges to ensure the safety of
the electrolyzer.
On the one hand, this model can increase the reserve to
enhance the balance adjustment ability of wind power to
Fig. 3 e The relationship between hydrogen concentration power grid emergency, it can greatly reduce energy waste. As
in oxygen and operating temperature. shown in Fig. 7, when the reserve amount D1 is less than the
25108 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1

Fig. 4 e The influences of input power on the hydrogen production and power consumption.

DC role in smoothing wind power output and electrolyzer

BUS operation.
Power In order to further optimize the start-up and shutdown
PMSG AC/DC DC/AC
Gird
control of the electrolyzer, the modular control strategy of the
electrolyzer is discussed below.
Super
Capacitor
DC/DC The electrolyzes in the wind-hydrogen coupled system are
divided into N modules with equal specifications, each
Alkaline
module is named as EL1 ; EL2 ; /; ELN , where Nis the number of
DC/DC
Electrolyzer modules.
This strategy divides each cell module into two power
Hydrogen
stages. Parameter i is defined as the starting sequence of the
Controller Controller Controller Controller
storage electrolyzer, and PM is the upper power limit of each electro-
lyzer module, PM ¼ 0:8Pel ðNÞ. When Ps is larger than the min-
Fig. 5 e Architecture of a grid-connected wind-hydrogen imum starting powerPel:min of EL1 , wherePel:min ¼ 0:5Pel ðNÞ, the
coupled energy system. electrolyzer EL1 starts; when Ps increases beyond the PM of the
electrolyzer EL1 , the excess wind power will be distributed to
the electrolyzer EL2 ; when Ps continue to increase and exceed
rated power Pel ðNÞ, the reserve amount is used for hydrogen the PM of the electrolyzer EL2 , the excess Ps power will be
production by electrolyzer. When the reserve amount D2 is distributed to the electrolyzer EL3 , and so on, until the PM
bigger than the rated power Pel ðNÞ, the electrolyzer operates at power increases to all the electrolyzer start-up. At this time,
the rated power, and the excess energy is absorbed by the the electrolyzer completes the start-up of stage 1. If the output
supercapacitor. of Ps continues to increase, the redundant power will be
redistributed to the electrolyzer EL1 to reach the rated power
Modular control strategy of electrolyzer Pel ðNÞ, and then to the electrolyzer EL2 ,EL3 , and so on.

The control flow chart of the proposed system is shown in

Fig. 8. As drawn from Fig. 8, supercapacitors play an important Simulation and discussion

In this section, the proposed control strategy is verified taking

the 2017's output data of a wind farm in Northeast China [30].
The wind farm consists of 33 wind turbines with a capacity of
Wind power (MW)

Psc
Δ1<Pel(N) Δ 2>Pel(N)
Wind power (MW)

Psc
Pel
Pel(N)
PGrid
PGrid PGrid

Time /h Time /h

Fig. 6 e Schematic diagram of demand balance control Fig. 7 e Schematic diagram of power retention mode for
mode for power grid. power grid demand.
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1 25109

Fig. 8 e Control flow chart of the proposed system.

1.5 MW and a total capacity of 49.5 MW. The monthly power supercapacitor bank. Then the simulation of the proposed
limit data for the wind farm in 2017 is given in Fig. 9. It is drawn control strategy of electrolyzer under the situation of Ps shown
from Fig. 9 that there are sever wind abandoning phenomenon in Fig. 10 is carried out in MATLAB 7.0 programming environ-
in the wind farm every month. The months of severe power ment on i7-6600U 2.6 GHz, 20.0 GB RAM computer system.
limitation are JanuaryeMay and NovembereDecember, the In actual operation, considering the balanced utilization of
months of less power limitation are JuneeOctober, February each module, the strategy of "start first, close first" is adopted.
and March are the most serious months, the average power In all operation strategies, when Ps increase, the first step is to
limits in February and March are 106473 kW and 14715 kW start electrolyzer EL1 ; followed by electrolyzer EL2 and elec-
respectively. And September and October are the least months, trolyzer EL3 ;/;EL10 ; when Ps decrease, the first step is to close
the average power limits in September and October were electrolyzer EL1 ; followed by electrolyzer EL2 ; andEL3 ; /; EL10 .
1530 kW and 1581 kW respectively. The simulation results are shown in Table 1.
According to the system architecture shown in Fig. 5, the
power of the electrolysis hydrogen production unit is 30% of the
installed capacity of wind power, so the total power is 15 MW,
which is consists of 5 modules, each module has a capacity of
3 MW. Supercapacitors’ capacity is chosen as 5500 F, rated
voltage is 2.7 V, 330 in series with 17 in parallel form a

Fig. 9 e Monthly power limit of the wind farm in 2017. Fig. 10 e Daily output of Ps in the wind farm.
25110 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1

Table 1 e Performance comparison of electrolyzer modules under different control strategies.

contro Strategy Hydrogen production/kg Switching times of the electrolyzer modules
EL1 EL2 EL3 EL4 EL5 Total times
Simple start-stop 1128.75 35 76 109 137 164 521
Slow start 1523.47 13 26 39 51 69 198
Segment start 1627.39 3 5 5 8 13 34

In order to make a further comparison, the effects of the respectively from the aspects of instantaneous power fluctu-
two other control strategies are also simulated. ation and wide power fluctuation, and a grid-connected wind-
hydrogen system integrated supercapacitors is constructed
 the simple start-stop strategy: when Ps is greater than the and its control strategy is proposed. The simulation results
rated power of the electrolyzer EL1 , the EL1 is started; when prove the following conclusions:
Ps increases and reaches the rated power of the electro-
lyzer EL2 , the EL2 is started, and so on.  Frequent fluctuation of wind power will lead to frequent
 Slow start strategy: the minimum starting power is set for start-up and shutdown of alkaline electrolyzers, which will
each electrolyzer. When Ps is greater than the minimum lead to the reduction of hydrogen production, and even
starting power of the electrolyzer EL1 , EL1 is started; then degrade the safety of alkaline electrolyzers.
the output power of the electrolyzer EL1 increases with the  For a grid -connected wind-hydrogen system, the integra-
increase of Ps, and when Ps continues to increase and Ps tion of supercapacitors is useful to smooth the fluctuation
exceeds the sum of the rated power of the electrolyzer EL1 of wind power on electrolyzers.
and the minimum starting power of the electrolyzer EL2 ,  By adapting a modular adaptive strategy to control the
the electrolyzer EL2 is started; and so on. switching of electrolyzers, the switching times of alkaline
electrolyzers can be greatly reduced while the hydrogen
Some conclusions can be drawn by comparing the results production will be greatly increased. This is beneficial to
of the above three control strategies: the operation of alkaline electrolyzers.

 By adapting the simple start-stop strategy, the total

switching times of electrolyzers reached 521 times, of
which the switching times of EL5 reached 164 times, that
Acknowledgements
means the EL5 will start-up and shut-down about every
8.78 min in the 24 h, which is obviously not conducive to
This work has been supported by the Natural Science Foun-
the stability of the electrolyzer.
dation of Fujian Province under the Grant 2019J01054.
 By adapting the slow start-up strategy, the switching times
of electrolyzers is lower than that of the simple start
strategy. The total switching times is 198, of which the references
switching times of EL5 is 69, that means the EL5 will start-
up and shut-down about every 21 min in the 24 h.
 By adapting the proposed segment start strategy, the [1] Turner JA. A realizable renewable energy future. Science
switching times of electrolyzers have been improved to a 1999;285(5428):687e9.
varying degree. Among them, the total switching times of [2] Lund H. Renewable energy strategies for sustainable
development. Energy 2007;32(6):912e9.
all electrolyzers is 34 times, which is only 6.5% of the
[3] Panwar NL, Kaushik SC, Kothari S. Role of renewable energy
simple start-stop control strategy. The switching times of
sources in environmental protection: a review. Renew
EL5 is 13, which is 92% lower than that of the simple start- Sustain Energy Rev 2011;15(3):1513e24.
stop control strategy. [4] Sutherland PE. Ensuring stable operation with grid codes: a
 Compared with the first two control strategies, Hydrogen look at canadian wind farm interconnections. IEEE Ind Appl
production of the system by adapting the proposed segment Mag 2011;22(1):60e7.
start strategy has also been significantly improved, and it [5] Mohseni-Bonab SM, Rabiee A, Mohammadi-Ivatloo B.
Voltage stability constrained multi-objective optimal
increases by 44.18% compared with that of the simple start-
reactive power dispatch under load and wind power
stop strategy, which means that the efficiency and life of uncertainties: a stochastic approach. Renew Energy
hydrogen production by electrolysis have been improved to 2016;85:598e609.
a considerable extent. [6] Baloch MH, Jie W, Kaloi GS. Stability and nonlinear controller
analysis of wind energy conversion system with random
wind speed. Int J Electr Power Energy Syst 2016;79(11):75e83.
[7] Schaab K, Hahn J, Wolkov M, Stursberg O. Robust control for
voltage and transient stability of power grids relying on wind
Conclusions
power. Contr Eng Pract 2017;60:7e17.
[8] Ge B, Wang W, Bi D, Rogers CB, Peng FZ, De Almeida Anı́bal T.
In this paper, the influences of the fluctuating wind power on Energy storage system-based power control for grid-
the performances of alkaline electrolyzers are discussed
 i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 4 4 ( 2 0 1 9 ) 2 5 1 0 4 e2 5 1 1 1 25111

connected wind power farm. Int J Electr Power Energy Syst wind energy systems. Int J Hydrogen Energy
2013;44(1):115e22. 2015;40(7):2927e38.
[9] Olabi AG. Renewable energy and energy storage systems. [20] Bergen A, Pitt L, Rowe A, Wild P, Djilali N. Transient
Energy 2017;136(1):1e6. electrolyser response in a renewable-regenerative energy
[10] Peng Y, Nehorai A. Joint optimization of hybrid energy system. Int J Hydrogen Energy 2009;34(1):64e70.
storage and generation capacity with renewable energy. IEEE [21] Ulleberg O. Modeling of advanced alkaline electrolyzers: a
Trans Smart Grid 2017;5(4):1566e74. system simulation approach. Int J Hydrogen Energy
[11] Amrouche SO, Rekioua D, Rekioua T. Ba. Overview of energy 2003;28(1):21e33.
storage in renewable energy systems. Int J Hydrogen Energy [22] guez PM, Ursúa A, Sanchis P, Sopena C, Guelbenzu E,
Die
2016;41(45):20914e27. Gandı́a LM. Thermal performance of a commercial alkaline
[12] Hosseini SE, Wahid MA. Hydrogen production from water electrolyzer: experimental study and mathematical
renewable and sustainable energy resources: promising modeling. Int J Hydrogen Energy 2008;33(24):7338e54.
green energy carrier for clean development. Renew Sustain [23] Valenciaga F, Evangelista CA. Control design for an
Energy Rev 2016;57:850e66. autonomous wind based hydrogen production system. Int J
[13] Nowotny J, Hoshino T, Dodson J, Atanacio AJ, Ionescu M, Hydrogen Energy 2010;35(11):5799e807.
Peterson V. Towards sustainable energy. generation of [24] Cheng W, Waje M, Xin W, Tang JM, H RC, Yan Y. Proton
hydrogen fuel using nuclear energy. Int J Hydrogen Energy exchange membrane fuel cells with carbon nanotube-based
2016;41(30):12812e925. electrodes. Nano Lett 2004;4(2):345e8.
[14] Nikolaidis P, Poullikkas A. A comparative overview of [25] Miles MH. Evaluation of electrocatalysts for water electrolysis
hydrogen production processes. Renew Sustain Energy Rev in alkaline solutions. J Electroanal Chem 1975;60(1):89e96.
2017;67:597e611. [26] Meng N, Leung MKH, Leung DYC. Technological
[15] Alrafea K, Fowler M, Elkamel A, Hajimiragha A. Integration development of hydrogen production by solid oxide
of renewable energy sources into combined cycle power electrolyzer cell (SOEC). Int J Hydrogen Energy
plants through electrolysis generated hydrogen in a new 2008;33(9):2337e54.
designed energy hub. Int J Hydrogen Energy [27] Marini S, Pesenti R, Villa M. Advanced alkaline water
2016;41(38):16718e28. electrolysis. Electrochim Acta 2012;82(82):384e91.
[16] Chi J, Yu H. Water electrolysis based on renewable energy for [28] Haug P, Kreitz B, Koj M, Turek T. Process modelling of an
hydrogen production. Chin J Catal 2018;39(3):390e4. alkaline water electrolyzer. Int J Hydrogen Energy
[17] Zhang F, Zhao P, Niu M, Maddy J. The survey of key 2017;42:15689e707.
technologies in hydrogen energy storage. Int J Hydrogen [29] China Shipbuilding Heavy Industry Group Company 718
Energy 2016;41(33):14535e52. Research Institute. Installation and operation instructions
[18] Troncoso E, Newborough M. Implementation and control of for KZDQ series hydrogen generation units [Z]. 2015.
electrolyzers to achieve high penetrations of renewable [30] Dawei H, Deqing Q, Na Y, Guowei C. Capacity allocation
power. Int J Hydrogen Energy 2007;32(13):2253e68. method of hydrogen production system consuming
[19] Sarrias-Mena R, Ferna  ndez-Ramı́rez LM, Garcı́a-Va
 zquez CA, abandoned wind power. Acta Energiae Solaris Sin
Jurado F. Electrolyzer models for hydrogen production from 2017;38(6):1517e25.