Energy Conversion and Management 173 (2018) 167–178

Contents lists available at ScienceDirect

Energy Conversion and Management

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

Techno-economic analysis of hydrogen production by solid oxide T

electrolyzer coupled with dish collector
⁎
Amin Mohammadi, Mehdi Mehrpooya
Renewable Energies and Environment Department, Faculty of New Sciences and Technologies, University of Tehran, Tehran, Iran

A R T I C LE I N FO A B S T R A C T

Keywords: Hydrogen is considered as one of the best alternatives for fossil fuels, especially when it is produced using
Thermodynamic analysis renewable energy resources. In this paper, dish collector is used to provide the required energy of a solid oxide
Economic analysis electrolyzer cell (SOEC) to produce hydrogen. Since dish collectors operate at high temperature, they would be
Solid oxide electrolyzer an ideal match for high temperature electrolyzers. These electrolyzer cells need both thermal and electrical
Dish collector
energy to produce hydrogen. A compressed air energy storage (CAES) system is used to produce electricity. To
Compressed air energy storage system
reduce its fuel consumption, it is combined with a dish collector. Another dish collector is also used to produce
thermal energy. To analyze the system, both thermodynamic and economic analyses are conducted. The results
showed that the system could produce 41.48 kg/day hydrogen. It is shown that efficiency of the power cycle and
the electrolyzer cell is equal to 72.69% and 61.70%, respectively and levelized cost of hydrogen is 9.1203 $/kg H2 .
To study the effect of key parameters on the system performance, sensitivity analysis is performed. It was
concluded that maximum and minimum pressure of air cavern in the CAES system have the highest effect on the
levelized cost of hydrogen. Also higher operating temperature of the electrolyzer cell benefits the system both
thermodynamically and economically.

1. Introduction decomposition, etc. Also chemical looping hydrogen production sys-

Currently, fossil fuels are the most dominant source of energy as
they supply almost 80 percent of world total primary energy.
Utilization of these fuels have led to huge environmental problems such
as global warming, acidification, etc. To alleviate the situation, re-
searchers tried to find new sources of energy which do not produce any
contaminant. The fact that fossil fuels are an exhaustible source of
energy and one day they would finish, increased the importance of
finding a new energy carrier. In this respect, hydrogen is suggested as
an alternative for fossil fuels. The most important advantages of hy-
drogen are its high energy conversion efficiency, different forms of
storage, ability of transportation [1], its near zero emission combustion
[2], etc. Hydrogen is converted to electrical power in the fuel cells with
high efficiency [3].
There are different methods which could be used for hydrogen
production. A comparison between different methods is performed by
Holladay et al. [4]. Currently, natural gas steam reforming (48%), oil
reforming (30%), coal gasification (18%) and water electrolysis (3.9%)
are the most common methods for hydrogen production [5]. It should
be mentioned that there are some other methods such as dark fer-
mentation, thermochemical conversion cycles, plasma arc
tems as one the methods have been developed [6,7]. But most of these
methods are at development stages and none of them are yet practical.
Since hydrogen is going to be used as an alternative for fossil fuels, it is
suggested to shift toward the electrolysis process, because it does not
use fossil fuels to produce hydrogen. Also the purity of hydrogen pro-
duced in this method is very high [8].
Electrolysis process is exactly the opposite of the fuel cell operation.
While in fuel cells, hydrogen is consumed to produce water and elec-
tricity, in an electrolysis process, electricity and water are used to
produce hydrogen. The reaction which occurs in electrolysis to produce
hydrogen is shown below:
1
H2 O → H2 + O2
2
Bhandari et al. showed that the main environmental concern in
electrolysis process is related to the electricity supply [9]. Consequently
the required electricity should be provided by renewable sources. Dif-
ferent kind of renewable energy sources could be used to produce the
required electricity such as wind energy [10], solar energy [11,12],
geothermal energy [13], etc. The problem with most renewable energy
sources is that they could produce energy in specific locations. But solar

⁎
Corresponding author.
E-mail address: mehrpoya@ut.ac.ir (M. Mehrpooya).

https://doi.org/10.1016/j.enconman.2018.07.073
Received 21 April 2018; Received in revised form 11 July 2018; Accepted 23 July 2018
0196-8904/ © 2018 Published by Elsevier Ltd.
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

Nomenclature W power [kW]

Z cost [$]
A area [m2]
Bg flow permeability Greeks
Cp specific heat capacity [kJ/kg K]
da anode thickness [m] α thermal diffusivity [m2/s]
dc cathode thickness [m] γ ratio of specific heats
Ė energy [kW h] η efficiency
Eact , i activation energy for electrodes κi pre-exponential factor for electrodes
ENernst nernst potential μ dynamic viscosity [Pa s]
F faraday constant [sA/mol] ν kinematic viscosity [m2/s]
g gravitational acceleration [m/s2] ρ density [kg/m3]
h specific enthalpy [kJ/kg], convective heat transfer coeffi- σ Stefan-Boltzmann constant
cient [W/m2 K] ∊ radiative emissivity
H total annual working hour [h]
I direct normal insolation per unit of collector area [W/m2] Subscripts
i interest rate
J current density [A/m2] a air
k thermal conductivity of the ambient air [W/mK] comp compressor
L length [m] w water
LCOH levelized cost of hydrogen [$/kg] f fuel
LHV lower heating value [kJ/kg] CC combustion chamber
ṁ mass flow rate [kg/s] g flue gas
Ṅ molar flow [mol/s] GT gas turbine
Nu Nusselt number elec electricity
P pressure [bar] u useful
Pr Prandtl number amb ambient
Q̇ heat [kW] cond conduction
R universal gas constant [J/mol K] conv convection
Ra Reighley number rad radiation
Rp cavern inlet pressure act activation
rp pressure ratio of compressor conc concentration
RTE round trip efficiency ohm ohmic
S entropy [kJ/K] O&M operation and maintenance
T temperature [K] cw cooling water
t time [h]
V volume [m3], volt [V]

energy is the most promising option as it could be used almost every-
where. Different types of electrolyzer based on the renewable energy
sources are investigated [14].
There are three different technologies for dissociation of water in
electrolysis process including alkaline, polymer membrane (PEM) and
solid oxide [15]. The first two technologies only need electricity and
their operating temperature is low. They are also quite mature and most
of the current electrolysis plants are based on one of these two tech-
nologies. Carmo et al. performed a review on PEM electrolyzer cells
[16]. El-Emam et al. [17] proposed a multi-generation system based on
solar energy to produce power, hot water, cooling and hydrogen si-
multaneously. They used PEM electrolyzer in their system.
Khalilnejad et al. [18] coupled an alkaline electrolyzer with a hy-
This feature makes it easier to integrate this type of electrolyzer
with solar energy. Because solar energy could be used to produce both
thermal and electrical energy for the electrolyzer. Usually photovoltaic
(PV) modules are used to produce electricity, while solar thermal col-
lectors are used for providing thermal energy. Ngoh et al. [20] analyzed
a large scale hydrogen production system based on solar energy. They
used PV modules for producing electricity and parabolic trough
Total energy demand Electrical energy demand
Thermal energy demand
300
Energy demand (kJ/mol H2)

250
brid photovoltaic-wind system to produce hydrogen. They optimized
the system using imperial competitive colony algorithm to maximize 200
hydrogen production with minimum excess power usage. Kiaee et al.
[19] analyzed the same system for a hydrogen fueling station in 150
Norway.
Unlike PEM and alkaline cells, solid oxide electrolyzer needs both 100
thermal and electrical energy and its operating temperature is very 50
high. Also its efficiency is higher. Since its operating temperature is
very high, SOEC receives a part of the required energy for water 0
splitting in the form of heat. Consequently, its required electrical energy 400 600 800 1000 1200 1400
is lower. In other words, total energy demand is constant and with Temperature (K)
increasing temperature, required electricity reduces and required
Fig. 1. Total energy, electrical energy and thermal energy required for elec-
thermal energy increases. This behavior is shown in Fig. 1.
trolysis process in solid oxide electrolyzers [19].

168
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

collector for thermal energy. They showed that under optimum condi-
tion, the system is capable of producing 0.064 kg/s hydrogen and its
rate is at its maximum at noon when solar radiation is the highest.
Another important finding of this study was that almost 93% of the
required land area for the plant is occupied by PV modules, while solar
thermal collector only occupies 3%. A similar system has been studied
by Akikur et al. [21]. They used PV module and parabolic trough col-
lector to provide energy for a reversible solid oxide fuel cell. During the
day, the cell operates as an electrolyzer to convert excess power pro-
duced by PV modules into hydrogen and during the nights, the cell
operates as a fuel cell where it uses the hydrogen to produce power.
They performed energy and exergy analyses on the system and showed
that when it is working as solar hydrogen production system its effi-
ciency is equal to 20%. As could be seen, efficiency of the system is
quite low which is due to inefficient PV modules. Solar thermal col-
lector have been used for supplying the required thermal load in the
power plants. A hybrid solid oxide fuel cell and parabolic solar dish
collector, combined cooling, heating and power (CCHP) system is
propose and analyzed [22]. A hybrid oxy-fuel cycle, high temperature
solar cycle is introduced and investigated [23].
To improve performance of the system, some researchers suggested
to use solar thermal collectors to produce the required electricity of the
electrolyzers.
Lin et al. [24] defined three different scenarios to provide the re-
quired energy for a solid oxide electrolyzer. In the first scenario, they
used a solar tower system to provide thermal energy. A part of this
energy is directly transferred to the electrolyzer as thermal energy and
the rest is converted into electricity in a power cycle. In the second
scenario, they used PV modules to provide electricity. The thermal
energy needed for the electrolysis process is provided by electric
heaters. In the third scenario, they used both solar tower and PV
module simultaneously to provide electrical and thermal energy. They
performed a techno economic analysis on these scenarios and showed
that the highest efficiency belongs to the first scenario where only solar
thermal collectors were used. AlZahrani et al. [25] integrated solar
tower collector with supercritical carbon dioxide Brayton cycle to
produce power and then the power is delivered to a solid oxide elec-
trolyzer to produce hydrogen. To enable the system to operate con-
tinuously, they also added a thermal storage system (TES). They used
solar tower collector because it could produce high temperature
thermal energy. Performing energy and exergy analyses on their system
they showed that total efficiency of the system is equal to 56.8% con-
sidering the heat supplied to TES. Balta et al. [26] used a combination
of solar tower, Brayton cycle, Rankine cycle and organic Rankine cycle
(ORC) to produce the required electricity of a high temperature solid
oxide electrolyzer. The power cycles are connected in series. They
showed that the system is capable of producing 0.057 kg/s hydrogen.
In this paper, an integrated compressed air energy storage system
with dish collector is used to produce the required electricity of an
electrolyzer. As it is known, if electricity is produced only by solar
energy, its price would be so high. Therefore to reduce the cost of
produced hydrogen, a new system is proposed in this paper. The most
important benefit of this system is that it is not entirely based on solar
energy and it uses the grid electricity during off peak hours (usually
during the nights). To provide the required thermal energy, another
dish collector is utilized. To the author’s knowledge, there is no study
on using high temperature dish collector integrated with high tem-
perature solid oxide electrolyzer. Since dish collector can achieve
higher temperature than the parabolic trough collector, it would be a
better match with solid oxide electrolyzer. Thermodynamic and

1 Intercooler 3 Aftercooler

2 7 Air
6
Comp Comp Flue gas
Compressed Air
1 2 Liquid
Storage
sodium
Oxygen
5 4 9 8
H2O+H2
Fuel 12 Organic
14 Preheater fluid
13 11 10 Water

Fuel
Turbine
Combustion Chamber
15

27
Economizer Evaporator Superheater
24
37
30 29 28
26 Turbine
18 19 20
Electric 38
Heat exchanger
Heater 1
22
21
17 32
Electric 36 39
Water Heater 2 H2+H2O 25
Tank Condenser
16
Cathode 23

SOEC
O2
Anode 31 34 33
35

Fig. 2. Schematic diagram of the proposed system.

169
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

economic analyses are performed on the system to evaluate its perfor- 3.1. Thermodynamic analysis
mance and finally sensitivity analysis on the most important parameters
of the system is done. In the following, first the proposed system is 3.1.1. Compressed air energy storage system
described in Section 2. Then in Section 3, mathematical model of the As mentioned before, compression train is consisted of two com-
system is presented. In Section 4, the results of the simulation are pressors and two heat exchangers. Outlet temperature of compressor is
shown and finally, the main findings of the study are listed in conclu- computed using Eq. (1):
sion section. γ −1
⎛ 1 ⎛ aγa ⎞
Tout = Tin ⎜1 + ⎜r p −1⎞⎟ ⎟
2. System description ηcomp ⎝ ⎠⎠ (1)
⎝
It should be noted that pressure ratio of each compressor is assumed
Fig. 2 shows schematic diagram of the proposed system. As can be
to be identical and equal to the maximum pressure of the air cavern as
seen, it is composed of 4 sections including a power cycle to produce
follows:
electrical energy, another cycle to provide thermal energy, a high
temperature electrolyzer and an ORC to recover waste energy and rp1 = rp2 = Rp (2)
produce more power.
As mentioned before, the power cycle is a compressed air energy Mass flow rate of produced hot water can be calculated by applying
storage system. It has two different sections, namely compression train energy balance for the intercooler and aftercooler:
and expansion train. These two sections operate separately. During the ṁ w (hout −hin ) = ṁ a (hout −hin ) (3)
nights when electricity price is very low, the compression train operates
to produce compressed air. Two compressors are used in the compres- Fuel mass flow rate is equal to:
sion train. To reduce the amount of power consumed by these com- ṁ a ha + ṁ f LHV = ṁ g hg + (1−ηCC ) ṁ f LHV (4)
pressors, an intercooler and an aftercooler are used in the system to
reduce temperature of the air and consequently reduce the amount of In Eq. (4), ηCC is efficiency of combustion chamber. Gas turbine
consumed power. The compressed air is stored in a cavern. During the outlet temperature is obtained using Eq. (5):
day, when there is enough solar radiation, the expansion train begins to 1 − γg
⎛ ⎛ P γg ⎞ ⎞
work. The compressed air goes through a dish collector where it re- Tout = Tin ⎜1−ηGT ⎜1−⎛ in ⎞ ⎜ ⎟
⎟⎟
ceives heat from solar energy and its temperature increases. Because ⎜ ⎜ ⎝ Pout ⎠ ⎟⎟
solar radiation changes during the day, dish outlet temperature is not ⎝ ⎝ ⎠⎠ (5)
constant. To avoid any thermal stresses in the gas turbine, a combustion The most important parameters of a CAES system is its charge and
chamber is placed after the dish collector to adjust the air temperature discharge time. These parameters are a function of mass flow rate,
by burning a small amount of fuel. It allows the gas turbine to produce a density and volume of the cavern as follows [27]:
constant power during the discharge time and therefore reduces the
(ρ7 −ρ10 ) Vstorage
power consumption from the grid. Finally a preheater is used to recover tcharge =
energy from gas turbine outlet stream. 3600ṁ 7 (6)
To provide thermal energy, another dish collector is used. The most (ρ7 −ρ10 ) Vstorage
important advantage of dish collectors is that they could reach high tdischarge =
3600ṁ 10 (7)
temperature. In this paper, liquid sodium is used as heat transfer fluid
to increase temperature of water as high as possible. Boiling point of Since compression and expansion trains of CAES system operate
sodium is equal to 873 °C which allows to have superheated water with separately, simple efficiency cannot be defined to evaluate its perfor-
high temperature. If, by any reason, dish collector couldn’t provide mance. Instead, another parameter named round trip efficiency is de-
enough energy to superheat the water, a part of water from water tank fined which shows the ratio of output energy to the input energy as
bypasses the dish collector cycle and goes through an electric heater. To follows:
further increase temperature of water, a heat exchanger is used to re- ̇ + Eheat
̇
Eelec
cover energy from SOEC outlet streams. Since water temperature RTE =
̇ ̇ ̇
Eelec + Efuel + Esolar (8)
should reach to the operating temperature of electrolyzer cell, another
electric heater is utilized to adjust water temperature. In this equation, all parameters show the amount of energy which is
Finally, an ORC is used to recover energy from electrolyzer outlet consumed or produced during a full charge/discharge cycle in kW h.
streams. After exchanging heat with water in the heat exchanger, these
two streams still have high quality energy which could be used to 3.1.2. Dish collector
produce more power and increase efficiency of the plant. Collector thermal efficiency is defined as follows:
Qu
3. Mathematical modeling ηc =
Qs (9)
Mathematical description of each section is presented here. Some In the above equation, Qu is useful energy and Qs is the energy in-
assumptions are made to simplify the analysis: cident on the concentrator aperture [28]. The energy incident on the
concentrator is a function of dish aperture area and direct normal ir-
• Heat and pressure losses inside the pipes are negligible. radiance per unit of collector area:
• Isentropic efficiency of compressors, pumps and turbines are con- Qs = Is Aape (10)
stant.
• Natural gas is used in combustion chamber. A part of this energy is absorbed by the receiver which is dependent
• Temperature gradient in the electrolyzer cell is negligible. on the optical efficiency of the collector:
• Temperature of anode and cathode outlet streams is equal to the cell ηo =
Qr
temperature.
Qs (11)
Optical efficiency of collector is determined based on optical
properties of its material and its geometry. Here it is assumed that its

170
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

value is equal to 0.85 [29]. Fig. 3 shows thermal model of the dish of the cell temperature and is calculated by Eq. (25):
collector. (25)
E 0 = 1.253−2.4516 × 10−4T
Total heat loss of the receiver is obtained by Eq. (12):
Activation overpotential shows the level of electrodes activity [33].
Ql = Qcond + Qconv + Qrad (12)
It could be calculated by the Butler-Volmer equation as follows:
Due to using insulation in the receiver, conduction loss is very
RT J ⎞
negligible compared with convection and radiation. Therefore, con- Vact = sinh−1 ⎜⎛ ⎟

duction loss is neglected in this study [30].

F ⎝ 2J0, i ⎠ (26)
Radiation heat loss is computed by: −Eact , i
J0, i = κi exp ⎛ ⎞
Qrad = Ac εeff σ (Tw4−Tamb
4
) (13) ⎝ RT ⎠ (27)

1 where Eact , i is activation energy for anode and cathode and κi Pre-ex-
εeff = ponential factor for anode and cathode exchange current density. The
1+ ( )
1
εc
−1
Ac
Aw (14) value of these parameters are shown in Table 5.
Ohmic overpotential is computed by using the Ohm’s law:
where Tw represents the temperature of the cavity wall. Based on the
A
results obtained by Kaushika et al. [31], when 0.8 < εc < 1 and Aw > 5, 10300
c Vohm = 2.99 × 10−5exp ⎛ ⎞ JL
the receiver could be considered as a black body. It should be men- ⎝ T ⎠ (28)
tioned that Ac and Aw are calculated using the following equations
where L is the electrolyte thickness. Concentration overpotential is a
where d and L are diameter and length of the cavity:
function of gas concentration difference between the electrode surface
πd 2 and the electrode–electrolyte interface [33]. Its value for cathode and
Ac =
4 (15) anode electrodes is calculated as follows:
eff 0
πd 2 RT ⎛ 1 + (jRTdc /2FD H2 O PH2 ) ⎞
Aw = πdL + Vconc, c = ln ⎜
4 (16) 2F eff 0 ⎟
⎝ 1−(jRTdc /2FD H2 O PH2 O ) ⎠ (29)
Convection heat loss is calculated by Newton’s law:
0 2
Qconv = hc Aw (Tw−Tamb) (17) RT ⎛ PO2 + JRTξda/2FBg ⎞
Vconc, a = ln ⎜ ⎟
4F PO02
In this equation, hc is heat transfer coefficient which is a function of ⎝ ⎠ (30)
Nusselt number, thermal conductivity of the bulk fluid and the en- In the above formulas, dc and da are cathode and anode thickness, Bg
semble length of the receiver as follows: is the flow permeability which can be calculated based on [35] and
Nuk D Heff2 O is the combination of molecular diffusion and Knudsen diffusion
hc = processes. Complete explanation about calculating this parameter is
Ls (18)
shown in [36]. Other constant parameters in these formulas such as
where Ls is defined as [32]: anode and cathode thickness and tortuosity are shown in Table 5.
3 Molar flow of outlet hydrogen, oxygen and water from the cell can
Ls = ∑ ai cos(ϕ + θi)bi Li be obtained by:
i=1 (19)
J
Ṅ H2, out = = Ṅ H2 O, utilized
In Eq. (19), ai , bi and θi are constant parameters. The value of these 2F (31)
parameters are shown in Table 1. Also ϕ and Li are defined as shown in
J
Fig. 4. NȮ 2, out =
4F (32)
Finally, Nusselt number is defined by the following equations [32]:
J
Nu = 0.0196RaL0.41 Pr 0.13 (20) Ṅ H2 O, out = Ṅ H2 O, in−
2F (33)
gβ (Tw−Tamb) Ls3
RaL =
υα (21)
υ
Pr =
α (22)

3.1.3. Solid oxide electrolyzer

Required voltage for water splitting reaction is equal to [33]:
VSOEC = ENernst + Vact , i + Vohm + Vconc, i (23)
In the above equation, E is the Nernst potential, Vconc, i and Vact , i are
concentration and activation overpotential in both electrodes and Vohm
is Ohmic overpotential in the electrolyte. It should be mentioned that
since the electrodes have high electrical conductivity, their Ohmic
overpotential is neglected [34]. Nernst equation is defined as Eq. (24):
0 0 1/2
RT ⎛ PH2 PO2 ⎞
ENernst = E 0 + ln ⎜
2F PH2 O ⎟
0
(24)
⎝ ⎠
In this equation, R is the universal gas constant, F is the Faraday
constant, T is operation temperature of the cell, Pi0 is partial pressure of
reactants and products and E 0 is standard potential which is a function Fig. 3. Thermal model of dish collector.

171
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

Table 1 Table 2
Constant parameters of Eq. (19). Cost equations for all components [21,37,39–44].
i ai bi θi Component Equation

1 4.08 5.41 −0.11 Compressor

2 −1.17 7.17 −0.30
Zcomp = 71.1ṁ air ( ) ln ( )/(0.9−η
Pin
Pout
Pin
Pout comp)

3 0.07 1.99 −0.08 Intercooler Zintercooler = 45.7Aintercooler

Aftercooler Zaftercooler = 45.7Aaftercooler
Heat exchanger Zpreheater = 10(4.3247 − 0.303log (AHX ) + 0.1634(log (AHX ))
2)

Dish collector Zdish = 400Aape

Combustion chamber
Zcc = ( 46.08ṁ a
0.995 − (Pout / Pin ) ) (1 + exp (0.018T−26.4))
Gas turbine P
479.34ṁ g ln( in )(1 + exp(0.036T + 273.15) − 54.4)
Pout
ZGT =
0.92 − ηGT
Pump 0.71 0.2
Zpump = 549.13Ẇ pump × 1.41 ⎜⎛1 + ⎛ ⎞ ⎞⎟
⎜ ⎟

⎝ ⎝ 1 − ηpump ⎠ ⎠
SOEC ̇
ZSOEC = 1000WSOEC
ORC turbine ZTurbine = 102.6259 + 1.4398log(W) −0.1776(log(W ))
2

ORC condenser ZSOEC = 1773ṁ CW

hydrogen. This is done using the following equation:

̇
ZCapitalCost + ZȮ & M + ZFuel
̇ + Zelectricity
̇
LCOH =
Mannual, H2 (41)
̇
In Eq. (41), ZCapitalCost is the investment cost rate which is calculated
φ as follows:
Fig. 4. Parameters ϕ and Li in Eq. (19). Ztotal CRF
̇
ZCapitalCost =
H (42)
Finally, efficiency of the cell can be computed by the following set
i (1 + i) N
of equations: CRF =
(1 + i) N −1 (43)
LHVH2 Ṅ H2
ηSOEC = where i is the interest rate, N is the plant life time and H is total annual
̇
WSOEC ̇ , SOEC + Qheat
+ Qheat ̇ , H2 O (34) working hour of the plant. Also Ztotal is total investment cost of the
̇
WSOEC = JA cell Ncell VSOEC (35) plant. Its value is calculated by computing investment cost of all com-
ponents. The required equations to do so, are shown in Table 2.
̇ , SOEC = (T ΔS−Sgen
Qheat ̇ ) Ṅ H2 O (36) To increase accuracy of economic results, investment cost of all
component should be brought to a reference year using chemical en-
̇ = 2F (Vconc, i + Vohm + Vact , i )
Sgen (37) gineering plant cost index.
The second term in Eq. (41) is ZȮ & M . This parameter shows the
̇ , H2 O = Ṅ H2 O (HH2 O, gas−HH2 O, liquid )
Qheat (38) annual cost for operating and maintenance. Its value is usually calcu-
lated based on investment cost of the components as follows:
3.1.4. ORC section ZȮ & M = αZCapitalCost
̇ (44)
Organic cycle is used to recover heat from electrolyzer outlet
streams. In a two-stage organic Rankine cycle coupled which works The value of α for dish, SOEC and other components is equal to 0.05
with solar energy is modeled and analyzed [37]. Since temperature of [39], 0.04 [21] and 0.06 [40], respectively.
these streams is high, a high temperature organic fluid should be used ̇
The third term is ZFuel which shows annual cost of fuel used in the
in the cycle. In this regard, cyclohexane is used. Since cyclohexane is a combustion chamber. To compute its value, annual fuel usage should be
dry fluid, turbine outlet stream is in superheated condition. To avoid computed and then it should be multiplied by natural gas price. In this
energy loss in condenser, a recuperator is added to the cycle to recover study, natural gas price is assumed to be equal to 0.2825 $/kg.
this energy by preheating the working fluid before going to the vapor ̇
The last term in the numerator of Eq. (41), is Zelectricity which shows
generator. This increases efficiency of the ORC section. Power produced annual cost of electricity consumed in the plant. A part of this para-
in the ORC turbine and power consumed in ORC pump are calculated meter is related to the power consumption by compressors during the
by: night. There are also two electric heaters in the plant and their elec-
tricity consumption should be considered here. Number of electrolyzer
̇ , tur = ṁ 37 (h37−h38)
WORC (39) cells is selected in a way that their power consumption is slightly lower
̇ , pump = ṁ 33 (h34−h33)
WORC than the power produced by gas turbine. The excess power produced by
(40)
gas turbine and the power produced by ORC section is delivered to
It’s worth mentioning that mass flow rate of organic cycle is cal- these electric heaters. If electricity demand by the electric heaters is
culated considering the minimum temperature difference between hot higher, then electricity from grid will be used. If not, the excess power
and cold streams. More explanation about mathematical modeling of will be sold to the grid and it would be considered as a profit. It should
ORC section can be found in [38]. be noted that the price of electricity is different during peak and off
peak hours and it is assumed to be 1.2 ₵/kW h for off peak hours and
3.2. Economic analysis 9.5₵/kW h for on peak hours [45].
The only parameter in the denominator, is annual hydrogen pro-
Economic analysis is performed to calculate cost of produced duction which can be easily calculated using formulas in Section 3.1.3.

172
A. Mohammadi, M. Mehrpooya
It should be mentioned that availability of the plant is assumed to be
80% during the year.
4. Validation
To validate the obtained results, simulation results are compared
with experimental results reported in published papers. Fig. 5 shows the
comparison for solid oxide electrolyzer cell. As can be seen, the ob-
tained results are in good agreement with experimental results at both
temperatures. This indicates that the simulation code of the SOEC is
correct.
The results for dish collector is also shown in Table 3. The com-
passion is made at cavity wall temperature equal to 823 K with the
results reported in [47]. As can be seen, with increasing inclination
angle, the deviation increases. But since at high inclination angles, the
value of convection heat transfer decreases, therefore it could be ne-
glected.
5. Results and discussion
To analyze performance of the system, a simulation program was
developed using MATLAB software. Thermodynamic properties of air,
water and organic fluid were calculated using REFPROP [48] and liquid
sodium properties were calculated using equations in Table 4. Design
condition of the system is shown in Table 5.
Using data provided above, performance of the system is evaluated.
It should be mentioned that solar radiation data for Tehran city in a
typical day of July is used in the analysis [51] which is shown in Fig. 6.
It is assumed that expansion train begins to operate at 12:00 so that
it could benefits from the highest solar radiation. Under the design
condition, thermodynamic properties of each point in the system are
calculated and shown in Table 6 and the most important parameters of
the system is shown in Table 7.
As can be seen in Table 7, charge time of the CAES system is 5.63 h
which means that the two compressors should operate for 5.63 h during
the night to fill the air cavern and for doing this, they need
1076.01 kW h electrical energy. During this period, the system produces
29.798 ton hot water which could be used for heating purposes. On the
other hand, the expansion train can operate for 4 h and it produces
1228.1 kW h energy. To do so, it consumes 0.0201 kg/day fuel. It
should be kept in mind that mass flow rate of natural gas decreased due
to utilization of dish collector which preheats the air before going into
the combustion chamber. If there was no dish collector in the system,
mass flow rate of natural gas would be equal to 0.0351 kg/day which
shows 74% increase. Generally the CAES system has a round trip effi-
ciency equal to 72.69%.
The electrolyzer voltage is equal to 1.035 V under design condition.
Between different overpotentials, activation overpotential has the
highest value. Usually in electrolyte based electrolyzers, Ohmic over-
potential dominates, but since the electrolyzer used in this study is an
anode based one, Ohmic overpotential is lower than activation over-
potential. As shown in the table, the electrolyzer can produce
41.4832 kg hydrogen per day and its efficiency is 61.70%. To produce
hydrogen, the cell consumes 299.8301 kW electricity. For a complete
discharge time (4 h), it would be equal to 1199.3 kW h. As can be seen,
this is lower than power produced by gas turbine which indicates that
the required electricity of the electrolyzer can be completely provided
by the CAES system. The excess power can be delivered to the electrical
heaters or when there is no need for electrical heaters, it could be sold
to the grid. For example, in design condition, electrical heater 1 does
not consume any power, but the second one consumes 103.6048 kW h.
On the other hand, ORC section produces 22.9041 kW h electrical en-
ergy. In general, total required electricity of the system is higher than
its power production and the system consumes 51.9118 kW h electrical
energy from the grid. Since the price of electricity during peak hours is
very expensive, this value should be as minimum as possible. Finally,
173
Energy Conversion and Management 173 (2018) 167–178
levelized cost of hydrogen is equal to 9.1203 $/kg H2 .
5.1. Sensitivity analysis
In this section, sensitivity analysis is performed to study the effect of
some parameters on the system performance. To do so, all parameters
are kept constant except the parameter under study. In the following,
detail explanation about the effect of these parameters are presented.
5.1.1. Maximum pressure of air cavern
Figs. 7 and 8 show the effect of maximum pressure of air cavern on
the system performance. Increasing the value of this parameter in-
creases outlet pressure of the compressors and therefore density of the
inlet air to the cavern changes. Consequently, both charge and dis-
charge time of the system change accordingly (Eqs. (6) and (7)). As can
be seen in Fig. 7, the higher the pressure, the higher the charge and
discharge times are which means that the system can operate for longer
hours. It should be mentioned that since the system is operating for
longer hours, more fuel will be consumed in the combustion chamber
which will increase carbon dioxide production of the system. It also
increases power consumption of compressors due to increasing their
pressure ratio. It should be mentioned that carbon dioxide production is
calculated based on the assumption of complete combustion where all
the carbon atoms convert into carbon dioxide. Having the mass flow
rate of fuel in the combustion chamber, carbon dioxide production can
be easily computed.
Generally, it could be said that higher pressure of air cavern is fa-
vorable as it leads to higher hydrogen production and round trip effi-
ciency. It also has a considerable effect on levelized cost of hydrogen.
As mentioned before, when discharge time of the cavern increases, it
means that the system should operate for longer hours. Therefore the
system operates till noon when solar radiation is not high enough.
Consequently, electricity consumption by electric heaters increases to
provide the required thermal energy for the electrolyzer. So both nu-
merator and denominator in Eq. (41) are increasing, but the effect of
hydrogen production dominates and therefore LCOH decreases. As it is
shown in Fig. 8, by increasing maximum pressure of air cavern from 13
to 17, LCOH decreases by almost 60%.
5.1.2. Minimum pressure of air cavern
The effect of minimum pressure of air cavern on the system per-
formance is shown in Figs. 9 and 10. Variation of this parameter only
affects the expansion train and it does not have any influence on the
compression train. Like the previous case, with variation of this para-
meter, both charge and discharge times of the system changes. As
shown in Fig. 9, when minimum pressure of air cavern increases, both
charge and discharge time of the system decrease. Therefore the system
operates for a shorter period of time. Consequently, lower fuel is burned
in the combustion chamber and as a result CO2 production reduces.
Present model (T=1123) Present model (T=1023)
Experiment (T=1123) Experiment (T=1023)
1.3
1.2
Cell potential (V)
1.1
1
0.9
1500 2500 3500 4500 5500 6500 7500
Current density (A/m2)
Fig. 5. Validation of solid oxide electrolyzer cell with experimental measure-
ments reported in [46].
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

Table 3 Table 5
Validation of dish collector results. Design condition of the system [20,25,33,50].
φ Tw = 823 K Parameter Value Unit

Qconv, cal (W) Qconv, exp (W) Error (%) Ambient temperature 25 °C
Ambient pressure 1 bar
0 120 123 2.44 Compressed air energy storage section
15 89 93 4.30 Maximum pressure of air cavern 15 bar
30 58 62 6.45 Minimum pressure of air cavern 10 bar
45 41 45 8.89 Volume of cavern 2000 m3
60 26 30 13.33 Compressor isentropic efficiency 85 %
75 18 20 10 Mass flow rate of air in compression train 0.55 kg/s
90 11 15 26.66 Mass flow rate of air in expansion train 0.7 kg/s
Gas turbine isentropic efficiency 90 %
Gas turbine inlet temperature 900 °C
Also since the air enters the turbine with higher pressure, power pro- Aperture diameter of dish collector 20 m
duction in the gas turbine increases. Generally, round trip efficiency of Cavity diameter 100 mm
Cavity length 200 mm
the system slightly decreases and from 73% at 8 bar reaches to 71.43%
Optic efficiency 85 %
at 12 bar.
Dish collector for liquid sodium
Since discharge of the CAES system decreased, the electrolyzer op-
Aperture diameter of dish collector 18 m
erates for a shorter period of time and therefore hydrogen production Pump isentropic efficiency 75 %
per day decreases too. This is shown in Fig. 10. As can be seen, the rate Cavity diameter 100 mm
of reduction in hydrogen production is considerable. Cavity length 200 mm
In terms of economic, having a lower pressure at the cavern outlet is Solid oxide electrolyzer cell
favorable. Like the previous case, hydrogen production rate has a Cell temperature 1100 K
higher impact on the levelized cost of hydrogen and LCOH increases Cell pressure 1 bar
Current density 4000 A/m2
considerably. It should be mentioned that electric heater 1 is used only
Activation energy for anode 1.2 × 105 J/mol
when minimum pressure of air cavern is equal to 8 bar. This is because Activation energy for cathode 1 × 105 J/mol
at this value, discharge time of the system is equal to 6 h and therefore Pre-exponential factor for anode exchange current density 2.051 × 109 A/m2
the system should operate till 18:00 when solar radiation is not high Pre-exponential factor for cathode exchange current 1.344 × 1010 A/m2
enough and solar collector cannot provide all the required thermal density
energy. Consequently, electric heater 1 should be used. On the other Anode thickness 17.5 × 10−6 m
hand, when minimum pressure of air cavern is equal to 12 bar, there is Cathode thickness 12.5 × 10−6 m

no need to use electricity from the grid and power produced by gas Electrolyte thickness 12.5 × 10−6 m
Porosity 0.48 –
turbine and ORC section is higher than the power required by the
Tortuosity 5.4 –
electrolyzer, electric heaters and the pumps. This means that the system Cell area 0.04 m2
can sell the excess power to the grid. Total number of cells 1720 –
Invertor efficiency 95 %

ORC section
5.1.3. Aperture diameter of dish collector in CAES section
Turbine inlet pressure 30 bar
Figs. 11 and 12 show the effect of aperture diameter of dish col- Condenser pressure 0.1 bar
lector in CAES system. Unlike the two previous cases, this parameter ORC turbine isentropic efficiency 90 %
does not have any effect on charge and discharge times of the system. In
fact, dish collector in CAES system was used as a preheater to increase
900
temperature of air before going to the combustion chamber. Since
800
changing dish diameter changes the amount of heat received by the
700
collector, consequently, temperature of air at dish outlet changes.
Fig. 11 shows the highest achievable temperature which is when solar 600
DNI (W/m2)

radiation is the highest (at 14:00). As can be seen, when dish diameter 500

increases from 17 to 23 m, maximum achievable temperature increases 400

by more than 100 °C. This means that air enters the combustion 300
chamber with higher temperature, therefore lower amount of fuel is 200
consumed in the combustion chamber. This also decreases the amount 100
of carbon dioxide production as it is shown in Fig. 12. On the other 0
hand, increasing aperture diameter of dish collector decreases round 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24

trip efficiency. This is because with increasing dish diameter, heat re- hour
ceived by collector increases, but due to optic inefficiency not all of the Fig. 6. Direct normal irradiance for Tehran in a typical day pf July.
received heat is delivered to the air and therefore round trip efficiency

Table 4
Thermophysical properties of liquid sodium [49].
Property Equation Unit

Density 0.5 kg/m3

(
ρ = 219 + 275.32 1−
T
2503.7 ) + 511.58 (1− T
)
2503.7
Specific heat capacity Cp = 1.6582−8.479 × 10−4T + 4.4541 × 10−7T2−2992.6T −2 kJ/kg °C
Thermal conductivity k = 124.67−0.11381T + 5.5226 × 10−5T2−1.1842 × 10−8T3 W/m °C
Dynamic viscosity ln(μ) = −6.4406−0.3958ln (T ) + 556.835/ T Pa s

174
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

Table 6 charge time discharge time W_comp CO2 production

Power consumed by compressors (kW)

Thermodynamic properties of each point of the system at 14:00. 9 335
8

CO2 production per day (kg)

Point T [°C] P [bar] m [kg/s] h [kJ/kg] 295

Discharge time (hour)

7

Charge time (hour)

1 25 1 0.550 298.45 6 255
2 191.31 3.87 0.550 466.77
5
3 40 3.87 0.550 312.96 215
4 25 1 0.675 104.92 4
5 55 1 0.675 230.33 3 175
6 216.13 15 0.550 491.68 2
7 40 15 0.550 310.70 135
1
8 25 1 0.794 104.92
9 55 1 0.794 230.33 0 95
10 40 10 0.700 311.71 12 13 14 15 16 17 18
11 405.91 10 0.700 691.46 Maximum pressure of air cavern (bar)
12 645 10 0.700 954.40
13 900 9.9 0.705 1247.61 Fig. 7. Variation of system performance with changing maximum pressure of
14 454.59 1 0.705 812.12 air cavern.
15 100 1 0.705 432.37
16 25 1 0.050 104.92
H2 production RTE LCOH
17 25.00 1.04 0.050 104.92 80 18
18 100.45 1.03 0.050 421.07

H2 production pre day (kg)

70 16

Round trip efficiency (%)

19 100.17 1.02 0.050 2675.84
20 532.10 1.01 0.050 3557.59 14
21 638.83 1 0.050 3791.65 60

LCOH ($/kg)
22 826.85 1 0.027 5450.54 12
23 826.85 1 0.023 1090.56 50
10
24 745.95 1 0.027 5151.44
25 682.10 1 0.023 932.23 40
8
26 103 1.03 0.250 581.47
30 6
27 636 1.03 0.250 1272.00
28 494.45 1.02 0.250 1095.65
20 4
29 150.45 1.01 0.250 644.69
12 13 14 15 16 17 18
30 103 1 0.250 581.46
Maximum pressure of air cavern (bar)
31 240 1 0.023 474.15
32 595.85 1 0.027 4610.53 Fig. 8. Variation of RTE, LCOH and hydrogen production rate with changing
33 19.28 0.1 0.017 −121.28 maximum pressure of air cavern.
34 20.81 30.92 0.017 −116.01
35 190 30.61 0.017 266.36
36 257.83 30.61 0.017 612.22 charge time discharge time W_GT CO2 production
37 410 30 0.017 1089.85

Power produced by gas turbine (kW)

9 360
38 288.64 0.10 0.017 787.74

CO2 production per day (kg)

39 108.64 0.1 0.017 405.37 8 320
Discharge time (hour)

7
Charge time (hour)

280
6
Table 7 5 240
Most important parameters of the system for a complete day. 4 200
Parameter Value Unit 3
160
2
Charge time 5.63 h 1 120
Power consumed by compressor 1 92.576 kW
0 80
Power consumed by compressor 2 98.294 kW
7 8 9 10 11 12 13
Electrical energy consumed by compression train 1076.01 kW h
Produced hot water in compression train 29.798 ton Minimum pressure of air cavern (bar)
Discharge time 4.43 h
Power produced by gas turbine 307 kW
Fig. 9. Variation of system performance with changing minimum pressure of
Electrical energy produced by expansion train 1228.1 kW h air cavern.
Mass flow rate of fuel in combustion chamber 0.0201 kg
Round trip efficiency 72.69 % H2 production RTE LCOH
Electrolyzer voltage 1.035 V
85 18
Activation overpotential 0.0453 V
H2 production per day (kg)

Ohmic overpotential 0.0174 V 75 16

Round trip efficiency (%)

Concentration overpotential 0.0014 V

Hydrogen production per day 41.4832 kg 65 14
LCOH ($/kg)

Efficiency of electrolyzer 61.70 % 55 12

Power consumed by electrolyzer 1199.3 kW h
Power consumed by electric heater 1 0 kW h 45 10
Power consumed by electric heater 2 103.6048 kW h
35 8
Power produced by ORC section 22.9041 kW h
LCOH 9.1203 $/kg H2 25 6

15 4
7 8 9 10 11 12 13
Minimum pressure of air cavern

Fig. 10. Variation of RTE, LCOH and hydrogen production rate with changing
minimum pressure of air cavern.

175
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

T_max m_fuel RTE maximum temperature of liquid sodium in thermal cycle with changing
Maximum dish outlet temprature (C)

750 90
dish diameter is shown in Fig. 15. Therefore when dish diameter de-
85

Fuel production per day (kg)

creases, more power should be used in electric heaters which increases

Round trip efficiency (%)

710 80
annual cost of electricity. On the other hand, lower dish diameter de-
75
670 70
creases capital cost of the system. Both of these parameters affect the
65 levelized cost of hydrogen. As can be seen in Fig. 15, there is an op-
630 60 timum value for dish diameter in which levelized cost of hydrogen is
55 minimum.
590 50
45 5.1.6. Operating temperature of electrolyzer
550 40 Operating temperature of the electrolyzer can significantly affect
16 17 18 19 20 21 22 23 24
the overpotentials. As said before, in higher temperatures, voltage of
Aperture diameter of dish collector (m)
the cell decreases and therefore the splitting reaction of water can be
Fig. 11. Variation of system performance with changing aperture diameter of done with lower power consumption. Fig. 16 shows the effect of op-
dish collector in CAES system. erating temperature on the overpotentials. As can be seen, activation
and Ohmic overpotentials reduce with increasing temperature, while
LCOH CO2 concentration overpotential slightly increases. As explained before,
9.6 280 concentration overpotential is very small compared with activation and
9.5 Ohmic losses and therefore in general, total voltage of the cell at higher

Co2 production per day (kg)

250
9.4 temperatures is higher. On the other hand, required thermal energy of
9.3 220 the cell increases. All these changes result in lower levelized cost of
LCOH ($/kg)

9.2 hydrogen, in a way that if operating temperature of the cell from

9.1
9 160
8.9
8.8
8.7
16 17 18 19 20 21 22
Aperture diameter of dish collector (m)
Fig. 12. Variation of LCOH and CO2 production per day with changing aperture
diameter of dish collector in CAES system.
of the CAES system decreases. In terms of economic, increasing dish
diameter increases capital cost of the system which results in higher
levelized cost of hydrogen as it is shown in Fig. 12.
5.1.4. Gas turbine inlet temperature
The effect of gas turbine inlet temperature on the system perfor-
mance is shown in Fig. 13. When turbine inlet temperature increases,
power produced by gas turbine increases too. On the other hand, gas
turbine outlet temperature also increases. Energy of this stream is
transferred to the air to increase its temperature in a preheater. By
increasing turbine outlet temperature, more energy is available in the
preheater to be exchanged with air. Consequently, air enters the dish
collector with higher temperature and therefore dish outlet temperature
increases. This is shown in Fig. 13. In general, although higher amount
of fuel is using in the combustion chamber, round trip efficiency in-
creases slightly and from 72% at 800 °C reaches to 73% at 950 °C. With
increasing mass flow rate of fuel, CO2 production rate also increases
which is shown in Fig. 14.
In terms of economic, higher amount of fuel is used when turbine
190
1000 °C increases to 1150 °C, levelized cost of hydrogen from 9.8 $/kg
decreases to 8.93 $/kg.
130
5.1.7. Current density
100 Figs. 17 and 18 show the effect of current density on the system
23 24 performance. As can be seen, all kinds of overpotentials increase with
increasing current density. As a result, cell voltage increases too. An-
other influence of increasing current density is increasing hydrogen
production rate. This is because based on Eq. (31), hydrogen production
rate is linearly dependent on the current density.
Fig. 18 shows the variation of levelized cost of hydrogen and elec-
trolyzer efficiency with changing current density. As it is shown, in
higher current density, electrolyzer cells have higher efficiency. In
terms of economic, when current density of electrolyzer increases its
power consumption increases too. Therefore, both capital cost (Table 2)
and annual cost of electricity increases. In general, since hydrogen
production rate increased considerably (Fig. 17), levelized cost of hy-
drogen production rate decreases, in a way that LCOH from 14 $/kg at
1500 A/m2 reached to 7.98 $/kg at 7500 A/m2.
6. Conclusion
In this paper, integration of compressed air energy storage system
and dish collector is used to provide required electrical and thermal
energy for a high temperature solid oxide electrolyzer cell. Also an
organic Rankine cycle is added to the system to recover energy from
outlet streams of the electrolyzer. The system is studied both
T_max_dish W_GT RTE
Power prodduced by gas turbine (kW)
Maximum dish outlet temprature (C)

800 74
inlet temperature increases. Therefore annual cost of fuel will increase
in Eq. (41). On the other hand, higher power produces in gas turbine
Round trip efficiecncy (%)

700
which will reduce annual cost of electricity. In general, levelized cost of 73
600
hydrogen reduces by increasing gas turbine inlet temperature as it is
shown in Fig. 14. 500 72

400
5.1.5. Aperture diameter of dish collector in thermal cycle 71
This parameter changes performance of the thermal cycle and it 300
doesn’t affect performance of the CAES system. Therefore round trip
200 70
efficiency of the CAES system and its fuel mass flow rate remain con- 750 800 850 900 950 1000
stant. Gas turbine inlet temperature (C)
Electric heaters were used in the system to increase temperature of
water whenever heat received by collector is not enough. Decreasing Fig. 13. Variation of system performance with changing gas turbine inlet
dish diameter decreases heat received by the collector. Variation of temperature.

176
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178

LCOH CO2 production LCOH eta_SOEC

9.4 250 16 80

Co2 production per day (kg)

70

Electrolyzer efficiency (%)

14
9.3
60
200 12

LCOH ($/kg)
LCOH ($/kg)

50
9.2 10
40
150 8
30
9.1
6 20

9 100 4 10
750 800 850 900 950 1000 1000 2000 3000 4000 5000 6000 7000 8000
Gas turbine inlet temperature (C) Current density (A/m2)

Fig. 14. Variation of LCOH and CO2 production per day with changing gas Fig. 18. Effect of current density on LCOH and electrolyzer efficiency.
turbine inlet temperature.
thermodynamically and economically. The main findings of this study
T_max_dish LCOH are summarized as follows:
Maximum dish outlet temperature (C)

1000 9.5

900
9.4
• In design condition, charge and discharge times of the system are
equal to 5.63 and 4.43 h. Expansion train starts to operate at 12:00
800 when there is high solar radiation. In this situation, the system is
9.3 LCOH ($/kg)
700
capable of producing 41.4832 kg/day hydrogen, while its levelized
9.2 cost of hydrogen is 9.1203 $/kg H2 .
600
• Utilization of dish collector in CAES system reduced fuel consump-
9.1 tion in combustion chamber by 74%.
•
500
Activation overpotential is higher than other overpotentials in the
400 9 cell which is due to low activity of electrodes. Also, concentration
13 14 15 16 17 18 19 20 21 overpotential is very small.
Aperture diameter of dish collector in thermal cycle (m)
• Sensitivity analysis showed that when difference between minimum
Fig. 15. Variation of LCOH and maximum outlet temperature of dish collector and maximum pressure of air cavern increases, round trip efficiency
with changing dish aperture diameter. increases and LCOH decreases.
• Aperture diameter of dish collectors have a considerable effect on
V_act V_ohm V_conc LCOH levelized cost of hydrogen. While LCOH increases by increasing
0.14 10 diameter of dish collector in CAES system, there is an optimum
0.12 value for aperture diameter of dish collector in thermal cycle in
9.5 which LCOH is minimum.
•
Overpotential (V)

0.1
Increasing gas turbine inlet temperature increases carbon dioxide
LCOH ($/kg)

0.08 production, but on the other hand, it reduces cost of hydrogen

9
0.06 production.

0.04
• Increasing current density and operating temperature of the cell
8.5 boost performance of the system as it will increase hydrogen pro-
0.02 duction rate, electrolyzer efficiency and cost of hydrogen produc-
0 8 tion.
970 1000 1030 1060 1090 1120 1150 1180 • For future works, the system could be studied exergoeconomically.
Operating temperature of electrolyzer (C) This would help to improve its performance. Also performing a
multi-objective optimization considering thermodynamic, economic
Fig. 16. Variation of system performance with changing operating temperature
and environmental objective functions could be very beneficial.
of the electrolyzer.

Acknowledgement
V_act V_ohm V_conc V_cell H2 production
1.2 25000
Authors acknowledge the financial support provided by Iran
1
20000
National science foundation (INSF).
The authors would like to acknowledge the financial support of
Overpotential (V)

0.8
15000 University of Tehran for this research under grant number 27636/01/
0.6 01.
10000
0.4
References
5000
0.2
[1] Dincer I, Acar C. Review and evaluation of hydrogen production methods for better
0 0 sustainability. Int J Hydrogen Energy 2015;40(34):11094–111.
1000 2000 3000 4000 5000 6000 7000 8000 [2] Acar C, Dincer I. Comparative assessment of hydrogen production methods from
Current density (A/m2) renewable and non-renewable sources. Int J Hydrogen Energy 2014;39(1):1–12.
[3] Mehrpooya M, et al. Modeling and optimum design of hybrid solid oxide fuel cell-
Fig. 17. Variation of voltage and hydrogen production rate with changing gas turbine power plants. Int J Hydrogen Energy 2014;39(36):21196–214.
current density. [4] Holladay JD, et al. An overview of hydrogen production technologies. Catal Today
2009;139(4):244–60.
[5] Muradov N, Veziroǧlu T. From hydrocarbon to hydrogen–carbon to hydrogen

177
A. Mohammadi, M. Mehrpooya
economy. Int J Hydrogen Energy 2005;30(3):225–37.
[6] Aghaie M, Mehrpooya M, Pourfayaz F. Introducing an integrated chemical looping
hydrogen production, inherent carbon capture and solid oxide fuel cell biomass
fueled power plant process configuration. Energy Convers Manage
2016;124:141–54.
[7] Mehrpooya M, et al. Design of an integrated process for simultaneous chemical
looping hydrogen production and electricity generation with CO2 capture. Int J
Hydrogen Energy 2017;42(12):8486–96.
[8] Dincer I. Green methods for hydrogen production. Int J Hydrogen Energy
2012;37(2):1954–71.
[9] Bhandari R, Trudewind CA, Zapp P. Life cycle assessment of hydrogen production
via electrolysis – a review. J Cleaner Prod 2014;85:151–63.
[10] Sherif SA, Barbir F, Veziroglu TN. Wind energy and the hydrogen economy—review
of the technology. Sol Energy 2005;78(5):647–60.
[11] Dincer I, Joshi AS. Solar based hydrogen production systems. Springer; 2013.
[12] Mehrpooya M, Shahsavan M, Sharifzadeh MMM. Modeling, energy and exergy
analysis of solar chimney power plant-Tehran climate data case study. Energy
2016;115:257–73.
[13] Yilmaz C, Kanoglu M. Thermodynamic evaluation of geothermal energy powered
hydrogen production by PEM water electrolysis. Energy 2014;69:592–602.
[14] Mohammadi A, Mehrpooya M. A comprehensive review on coupling different types
of electrolyzer to renewable energy sources. Energy 2018;158:632–55.
[15] Zoulias E, et al. A review on water electrolysis. Tcjst 2004;4(2):41–71.
[16] Carmo M, et al. A comprehensive review on PEM water electrolysis. Int J Hydrogen
Energy 2013;38(12):4901–34.
[17] El-Emam RS, Dincer I. Investigation and assessment of a novel solar-driven in-
tegrated energy system. Energy Convers Manage 2018;158:246–55.
[18] Khalilnejad A, Riahy GH. A hybrid wind-PV system performance investigation for
the purpose of maximum hydrogen production and storage using advanced alkaline
electrolyzer. Energy Convers Manage 2014;80:398–406.
[19] Kiaee M, et al. Demonstration of the operation and performance of a pressurised
alkaline electrolyser operating in the hydrogen fuelling station in Porsgrunn,
Norway. Energy Convers Manage 2015;94:40–50.
[20] Ngoh SK, et al. Design and simulation of hybrid solar high-temperature hydrogen
production system using both solar photovoltaic and thermal energy. Sustain
Energy Technol Assess 2014;7:279–93.
[21] Akikur R, et al. Performance analysis of a co-generation system using solar energy
and SOFC technology. Energy Convers Manage 2014;79:415–30.
[22] Moradi M, Mehrpooya M. Optimal design and economic analysis of a hybrid solid
oxide fuel cell and parabolic solar dish collector, combined cooling, heating and
power (CCHP) system used for a large commercial tower. Energy 2017;130:530–43.
[23] Mehrpooya M, Sharifzadeh MMM. A novel integration of oxy-fuel cycle, high
temperature solar cycle and LNG cold recovery–energy and exergy analysis. Appl
Therm Eng 2017;114:1090–104.
[24] Lin M, Haussener S. Techno-economic modeling and optimization of solar-driven
high-temperature electrolysis systems. Sol Energy 2017;155:1389–402.
[25] AlZahrani AA, Dincer I. Design and analysis of a solar tower based integrated
system using high temperature electrolyzer for hydrogen production. Int J
Hydrogen Energy 2016;41(19):8042–56.
[26] Balta MT, Kizilkan O, Yılmaz F. Energy and exergy analyses of integrated hydrogen
production system using high temperature steam electrolysis. Int J Hydrogen
Energy 2016;41(19):8032–41.
[27] Mohammadi A, et al. Exergy analysis of a Combined Cooling, Heating and Power
system integrated with wind turbine and compressed air energy storage system.
Energy Convers Manage 2017;131:69–78.
[28] Kalogirou SA. Solar thermal collectors and applications. Prog Energy Combust Sci
2004;30(3):231–95.
178
Energy Conversion and Management 173 (2018) 167–178
[29] Wu S-Y, et al. A parabolic dish/AMTEC solar thermal power system and its per-
formance evaluation. Appl Energy 2010;87(2):452–62.
[30] Leibfried U, Ortjohann J. Convective heat loss from upward and downward-facing
cavity solar receivers: measurements and calculations. J Sol Energy Eng
1995;117(2):75–84.
[31] Kaushika N, Reddy K. Performance of a low cost solar paraboloidal dish steam
generating system. Energy Convers Manage 2000;41(7):713–26.
[32] Paitoonsurikarn S, Lovegrove K. A new correlation for predicting the free convec-
tion loss from solar dish concentrating receivers. In: Solar. Citeseer; 2006.
[33] Ni M, Leung MK, Leung DY. Parametric study of solid oxide steam electrolyzer for
hydrogen production. Int J Hydrogen Energy 2007;32(13):2305–13.
[34] Ferguson J, Fiard J, Herbin R. Three-dimensional numerical simulation for various
geometries of solid oxide fuel cells. J Power Sources 1996;58(2):109–22.
[35] Nam JH, Jeon DH. A comprehensive micro-scale model for transport and reaction in
intermediate temperature solid oxide fuel cells. Electrochim Acta
2006;51(17):3446–60.
[36] Hernández-Pacheco E, et al. A macro-level model for determining the performance
characteristics of solid oxide fuel cells. J Power Sources 2004;138(1):174–86.
[37] Mehrpooya M, Ashouri M, Mohammadi A. Thermoeconomic analysis and optimi-
zation of a regenerative two-stage organic Rankine cycle coupled with liquefied
natural gas and solar energy. Energy 2017;126:899–914.
[38] Sadreddini A, et al. Exergy analysis and optimization of a CCHP system composed of
compressed air energy storage system and ORC cycle. Energy Convers Manage
2018;157:111–22.
[39] Abbas M, et al. Dish Stirling technology: a 100 MW solar power plant using hy-
drogen for Algeria. Int J Hydrogen Energy 2011;36(7):4305–14.
[40] Bejan A, Tsatsaronis G. Thermal design and optimization. John Wiley & Sons; 1996.
[41] Turton R, et al. Analysis, synthesis and design of chemical processes. Pearson
Education; 2008.
[42] Boyaghchi FA, Heidarnejad P. Thermoeconomic assessment and multi objective
optimization of a solar micro CCHP based on Organic Rankine Cycle for domestic
application. Energy Convers Manage 2015;97:224–34.
[43] Ansarinasab H, Mehrpooya M, Mohammadi A. Advanced exergy and ex-
ergoeconomic analyses of a hydrogen liquefaction plant equipped with mixed re-
frigerant system. J Cleaner Prod 2017;144:248–59.
[44] Ansarinasab H, Mehrpooya M. Advanced exergoeconomic analysis of a novel pro-
cess for production of LNG by using a single effect absorption refrigeration cycle.
Appl Therm Eng 2017;144:719–32.
[45] Nezamoddini N, Wang Y. Real-time electricity pricing for industrial customers:
survey and case studies in the United States. Appl Energy 2017;195:1023–37.
[46] Chacartegui R, et al. Alternative cycles based on carbon dioxide for central receiver
solar power plants. Appl Therm Eng 2011;31(5):872–9.
[47] Taumoefolau T. Experimental investigation of convection loss from a model solar
concentrator cavity receiver [Master of Philosophy Thesis]. Canberra, Australia:
Department of Engineering, Australian National University; 2004.
[48] Lemmon EW, Huber ML, McLinden MO. NIST reference fluid thermodynamic and
transport properties–REFPROP. NIST standard reference database, vol. 23; 2002.
P. v7.
[49] Boerema N, et al. Liquid sodium versus Hitec as a heat transfer fluid in solar thermal
central receiver systems. Sol Energy 2012;86(9):2293–305.
[50] Mohammadi A, Mehrpooya M. Exergy analysis and optimization of an integrated
micro gas turbine, compressed air energy storage and solar dish collector process. J
Cleaner Prod 2016;139:372–83.
[51] Mehrpooya M, Hemmatabady H, Ahmadi MH. Optimization of performance of
Combined Solar Collector-Geothermal Heat Pump Systems to supply thermal load
needed for heating greenhouses. Energy Convers Manage 2015;97:382–92.