Professional Documents
Culture Documents
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.
⁎
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
energy is the most promising option as it could be used almost every- This feature makes it easier to integrate this type of electrolyzer
where. Different types of electrolyzer based on the renewable energy with solar energy. Because solar energy could be used to produce both
sources are investigated [14]. thermal and electrical energy for the electrolyzer. Usually photovoltaic
There are three different technologies for dissociation of water in (PV) modules are used to produce electricity, while solar thermal col-
electrolysis process including alkaline, polymer membrane (PEM) and lectors are used for providing thermal energy. Ngoh et al. [20] analyzed
solid oxide [15]. The first two technologies only need electricity and a large scale hydrogen production system based on solar energy. They
their operating temperature is low. They are also quite mature and most used PV modules for producing electricity and parabolic trough
of the current electrolysis plants are based on one of these two tech-
nologies. Carmo et al. performed a review on PEM electrolyzer cells Total energy demand Electrical energy demand
[16]. El-Emam et al. [17] proposed a multi-generation system based on
Thermal energy demand
solar energy to produce power, hot water, cooling and hydrogen si-
multaneously. They used PEM electrolyzer in their system. 300
Khalilnejad et al. [18] coupled an alkaline electrolyzer with a hy-
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- heaters. In the third scenario, they used both solar tower and PV
tion, the system is capable of producing 0.064 kg/s hydrogen and its module simultaneously to provide electrical and thermal energy. They
rate is at its maximum at noon when solar radiation is the highest. performed a techno economic analysis on these scenarios and showed
Another important finding of this study was that almost 93% of the that the highest efficiency belongs to the first scenario where only solar
required land area for the plant is occupied by PV modules, while solar thermal collectors were used. AlZahrani et al. [25] integrated solar
thermal collector only occupies 3%. A similar system has been studied tower collector with supercritical carbon dioxide Brayton cycle to
by Akikur et al. [21]. They used PV module and parabolic trough col- produce power and then the power is delivered to a solid oxide elec-
lector to provide energy for a reversible solid oxide fuel cell. During the trolyzer to produce hydrogen. To enable the system to operate con-
day, the cell operates as an electrolyzer to convert excess power pro- tinuously, they also added a thermal storage system (TES). They used
duced by PV modules into hydrogen and during the nights, the cell solar tower collector because it could produce high temperature
operates as a fuel cell where it uses the hydrogen to produce power. thermal energy. Performing energy and exergy analyses on their system
They performed energy and exergy analyses on the system and showed they showed that total efficiency of the system is equal to 56.8% con-
that when it is working as solar hydrogen production system its effi- sidering the heat supplied to TES. Balta et al. [26] used a combination
ciency is equal to 20%. As could be seen, efficiency of the system is of solar tower, Brayton cycle, Rankine cycle and organic Rankine cycle
quite low which is due to inefficient PV modules. Solar thermal col- (ORC) to produce the required electricity of a high temperature solid
lector have been used for supplying the required thermal load in the oxide electrolyzer. The power cycles are connected in series. They
power plants. A hybrid solid oxide fuel cell and parabolic solar dish showed that the system is capable of producing 0.057 kg/s hydrogen.
collector, combined cooling, heating and power (CCHP) system is In this paper, an integrated compressed air energy storage system
propose and analyzed [22]. A hybrid oxy-fuel cycle, high temperature with dish collector is used to produce the required electricity of an
solar cycle is introduced and investigated [23]. electrolyzer. As it is known, if electricity is produced only by solar
To improve performance of the system, some researchers suggested energy, its price would be so high. Therefore to reduce the cost of
to use solar thermal collectors to produce the required electricity of the produced hydrogen, a new system is proposed in this paper. The most
electrolyzers. important benefit of this system is that it is not entirely based on solar
Lin et al. [24] defined three different scenarios to provide the re- energy and it uses the grid electricity during off peak hours (usually
quired energy for a solid oxide electrolyzer. In the first scenario, they during the nights). To provide the required thermal energy, another
used a solar tower system to provide thermal energy. A part of this dish collector is utilized. To the author’s knowledge, there is no study
energy is directly transferred to the electrolyzer as thermal energy and on using high temperature dish collector integrated with high tem-
the rest is converted into electricity in a power cycle. In the second perature solid oxide electrolyzer. Since dish collector can achieve
scenario, they used PV modules to provide electricity. The thermal higher temperature than the parabolic trough collector, it would be a
energy needed for the electrolysis process is provided by electric 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
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 ⎜⎛ ⎟
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)
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 − ηpump ⎠ ⎠
SOEC ̇
ZSOEC = 1000WSOEC
ORC turbine ZTurbine = 102.6259 + 1.4398log(W) −0.1776(log(W ))
2
172
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178
It should be mentioned that availability of the plant is assumed to be levelized cost of hydrogen is equal to 9.1203 $/kg H2 .
80% during the year.
5.1. Sensitivity analysis
4. Validation
In this section, sensitivity analysis is performed to study the effect of
To validate the obtained results, simulation results are compared some parameters on the system performance. To do so, all parameters
with experimental results reported in published papers. Fig. 5 shows the are kept constant except the parameter under study. In the following,
comparison for solid oxide electrolyzer cell. As can be seen, the ob- detail explanation about the effect of these parameters are presented.
tained results are in good agreement with experimental results at both
temperatures. This indicates that the simulation code of the SOEC is 5.1.1. Maximum pressure of air cavern
correct. Figs. 7 and 8 show the effect of maximum pressure of air cavern on
The results for dish collector is also shown in Table 3. The com- the system performance. Increasing the value of this parameter in-
passion is made at cavity wall temperature equal to 823 K with the creases outlet pressure of the compressors and therefore density of the
results reported in [47]. As can be seen, with increasing inclination inlet air to the cavern changes. Consequently, both charge and dis-
angle, the deviation increases. But since at high inclination angles, the charge time of the system change accordingly (Eqs. (6) and (7)). As can
value of convection heat transfer decreases, therefore it could be ne- be seen in Fig. 7, the higher the pressure, the higher the charge and
glected. discharge times are which means that the system can operate for longer
hours. It should be mentioned that since the system is operating for
5. Results and discussion longer hours, more fuel will be consumed in the combustion chamber
which will increase carbon dioxide production of the system. It also
To analyze performance of the system, a simulation program was increases power consumption of compressors due to increasing their
developed using MATLAB software. Thermodynamic properties of air, pressure ratio. It should be mentioned that carbon dioxide production is
water and organic fluid were calculated using REFPROP [48] and liquid calculated based on the assumption of complete combustion where all
sodium properties were calculated using equations in Table 4. Design the carbon atoms convert into carbon dioxide. Having the mass flow
condition of the system is shown in Table 5. rate of fuel in the combustion chamber, carbon dioxide production can
Using data provided above, performance of the system is evaluated. be easily computed.
It should be mentioned that solar radiation data for Tehran city in a Generally, it could be said that higher pressure of air cavern is fa-
typical day of July is used in the analysis [51] which is shown in Fig. 6. vorable as it leads to higher hydrogen production and round trip effi-
It is assumed that expansion train begins to operate at 12:00 so that ciency. It also has a considerable effect on levelized cost of hydrogen.
it could benefits from the highest solar radiation. Under the design As mentioned before, when discharge time of the cavern increases, it
condition, thermodynamic properties of each point in the system are means that the system should operate for longer hours. Therefore the
calculated and shown in Table 6 and the most important parameters of system operates till noon when solar radiation is not high enough.
the system is shown in Table 7. Consequently, electricity consumption by electric heaters increases to
As can be seen in Table 7, charge time of the CAES system is 5.63 h provide the required thermal energy for the electrolyzer. So both nu-
which means that the two compressors should operate for 5.63 h during merator and denominator in Eq. (41) are increasing, but the effect of
the night to fill the air cavern and for doing this, they need hydrogen production dominates and therefore LCOH decreases. As it is
1076.01 kW h electrical energy. During this period, the system produces shown in Fig. 8, by increasing maximum pressure of air cavern from 13
29.798 ton hot water which could be used for heating purposes. On the to 17, LCOH decreases by almost 60%.
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
5.1.2. Minimum pressure of air cavern
should be kept in mind that mass flow rate of natural gas decreased due
The effect of minimum pressure of air cavern on the system per-
to utilization of dish collector which preheats the air before going into
formance is shown in Figs. 9 and 10. Variation of this parameter only
the combustion chamber. If there was no dish collector in the system,
affects the expansion train and it does not have any influence on the
mass flow rate of natural gas would be equal to 0.0351 kg/day which
compression train. Like the previous case, with variation of this para-
shows 74% increase. Generally the CAES system has a round trip effi-
meter, both charge and discharge times of the system changes. As
ciency equal to 72.69%.
shown in Fig. 9, when minimum pressure of air cavern increases, both
The electrolyzer voltage is equal to 1.035 V under design condition.
charge and discharge time of the system decrease. Therefore the system
Between different overpotentials, activation overpotential has the
operates for a shorter period of time. Consequently, lower fuel is burned
highest value. Usually in electrolyte based electrolyzers, Ohmic over-
in the combustion chamber and as a result CO2 production reduces.
potential dominates, but since the electrolyzer used in this study is an
anode based one, Ohmic overpotential is lower than activation over- Present model (T=1123) Present model (T=1023)
potential. As shown in the table, the electrolyzer can produce Experiment (T=1123) Experiment (T=1023)
41.4832 kg hydrogen per day and its efficiency is 61.70%. To produce
1.3
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,
1.2
Cell potential (V)
this is lower than power produced by gas turbine which indicates that
the required electricity of the electrolyzer can be completely provided
1.1
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
1
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.
0.9
On the other hand, ORC section produces 22.9041 kW h electrical en- 1500 2500 3500 4500 5500 6500 7500
ergy. In general, total required electricity of the system is higher than Current density (A/m2)
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 Fig. 5. Validation of solid oxide electrolyzer cell with experimental measure-
very expensive, this value should be as minimum as possible. Finally, ments reported in [46].
173
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
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
174
A. Mohammadi, M. Mehrpooya Energy Conversion and Management 173 (2018) 167–178
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
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)
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
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 ($/kg)
LCOH ($/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)
0.04
• 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 Energy Conversion and Management 173 (2018) 167–178
economy. Int J Hydrogen Energy 2005;30(3):225–37. [29] Wu S-Y, et al. A parabolic dish/AMTEC solar thermal power system and its per-
[6] Aghaie M, Mehrpooya M, Pourfayaz F. Introducing an integrated chemical looping formance evaluation. Appl Energy 2010;87(2):452–62.
hydrogen production, inherent carbon capture and solid oxide fuel cell biomass [30] Leibfried U, Ortjohann J. Convective heat loss from upward and downward-facing
fueled power plant process configuration. Energy Convers Manage cavity solar receivers: measurements and calculations. J Sol Energy Eng
2016;124:141–54. 1995;117(2):75–84.
[7] Mehrpooya M, et al. Design of an integrated process for simultaneous chemical [31] Kaushika N, Reddy K. Performance of a low cost solar paraboloidal dish steam
looping hydrogen production and electricity generation with CO2 capture. Int J generating system. Energy Convers Manage 2000;41(7):713–26.
Hydrogen Energy 2017;42(12):8486–96. [32] Paitoonsurikarn S, Lovegrove K. A new correlation for predicting the free convec-
[8] Dincer I. Green methods for hydrogen production. Int J Hydrogen Energy tion loss from solar dish concentrating receivers. In: Solar. Citeseer; 2006.
2012;37(2):1954–71. [33] Ni M, Leung MK, Leung DY. Parametric study of solid oxide steam electrolyzer for
[9] Bhandari R, Trudewind CA, Zapp P. Life cycle assessment of hydrogen production hydrogen production. Int J Hydrogen Energy 2007;32(13):2305–13.
via electrolysis – a review. J Cleaner Prod 2014;85:151–63. [34] Ferguson J, Fiard J, Herbin R. Three-dimensional numerical simulation for various
[10] Sherif SA, Barbir F, Veziroglu TN. Wind energy and the hydrogen economy—review geometries of solid oxide fuel cells. J Power Sources 1996;58(2):109–22.
of the technology. Sol Energy 2005;78(5):647–60. [35] Nam JH, Jeon DH. A comprehensive micro-scale model for transport and reaction in
[11] Dincer I, Joshi AS. Solar based hydrogen production systems. Springer; 2013. intermediate temperature solid oxide fuel cells. Electrochim Acta
[12] Mehrpooya M, Shahsavan M, Sharifzadeh MMM. Modeling, energy and exergy 2006;51(17):3446–60.
analysis of solar chimney power plant-Tehran climate data case study. Energy [36] Hernández-Pacheco E, et al. A macro-level model for determining the performance
2016;115:257–73. characteristics of solid oxide fuel cells. J Power Sources 2004;138(1):174–86.
[13] Yilmaz C, Kanoglu M. Thermodynamic evaluation of geothermal energy powered [37] Mehrpooya M, Ashouri M, Mohammadi A. Thermoeconomic analysis and optimi-
hydrogen production by PEM water electrolysis. Energy 2014;69:592–602. zation of a regenerative two-stage organic Rankine cycle coupled with liquefied
[14] Mohammadi A, Mehrpooya M. A comprehensive review on coupling different types natural gas and solar energy. Energy 2017;126:899–914.
of electrolyzer to renewable energy sources. Energy 2018;158:632–55. [38] Sadreddini A, et al. Exergy analysis and optimization of a CCHP system composed of
[15] Zoulias E, et al. A review on water electrolysis. Tcjst 2004;4(2):41–71. compressed air energy storage system and ORC cycle. Energy Convers Manage
[16] Carmo M, et al. A comprehensive review on PEM water electrolysis. Int J Hydrogen 2018;157:111–22.
Energy 2013;38(12):4901–34. [39] Abbas M, et al. Dish Stirling technology: a 100 MW solar power plant using hy-
[17] El-Emam RS, Dincer I. Investigation and assessment of a novel solar-driven in- drogen for Algeria. Int J Hydrogen Energy 2011;36(7):4305–14.
tegrated energy system. Energy Convers Manage 2018;158:246–55. [40] Bejan A, Tsatsaronis G. Thermal design and optimization. John Wiley & Sons; 1996.
[18] Khalilnejad A, Riahy GH. A hybrid wind-PV system performance investigation for [41] Turton R, et al. Analysis, synthesis and design of chemical processes. Pearson
the purpose of maximum hydrogen production and storage using advanced alkaline Education; 2008.
electrolyzer. Energy Convers Manage 2014;80:398–406. [42] Boyaghchi FA, Heidarnejad P. Thermoeconomic assessment and multi objective
[19] Kiaee M, et al. Demonstration of the operation and performance of a pressurised optimization of a solar micro CCHP based on Organic Rankine Cycle for domestic
alkaline electrolyser operating in the hydrogen fuelling station in Porsgrunn, application. Energy Convers Manage 2015;97:224–34.
Norway. Energy Convers Manage 2015;94:40–50. [43] Ansarinasab H, Mehrpooya M, Mohammadi A. Advanced exergy and ex-
[20] Ngoh SK, et al. Design and simulation of hybrid solar high-temperature hydrogen ergoeconomic analyses of a hydrogen liquefaction plant equipped with mixed re-
production system using both solar photovoltaic and thermal energy. Sustain frigerant system. J Cleaner Prod 2017;144:248–59.
Energy Technol Assess 2014;7:279–93. [44] Ansarinasab H, Mehrpooya M. Advanced exergoeconomic analysis of a novel pro-
[21] Akikur R, et al. Performance analysis of a co-generation system using solar energy cess for production of LNG by using a single effect absorption refrigeration cycle.
and SOFC technology. Energy Convers Manage 2014;79:415–30. Appl Therm Eng 2017;144:719–32.
[22] Moradi M, Mehrpooya M. Optimal design and economic analysis of a hybrid solid [45] Nezamoddini N, Wang Y. Real-time electricity pricing for industrial customers:
oxide fuel cell and parabolic solar dish collector, combined cooling, heating and survey and case studies in the United States. Appl Energy 2017;195:1023–37.
power (CCHP) system used for a large commercial tower. Energy 2017;130:530–43. [46] Chacartegui R, et al. Alternative cycles based on carbon dioxide for central receiver
[23] Mehrpooya M, Sharifzadeh MMM. A novel integration of oxy-fuel cycle, high solar power plants. Appl Therm Eng 2011;31(5):872–9.
temperature solar cycle and LNG cold recovery–energy and exergy analysis. Appl [47] Taumoefolau T. Experimental investigation of convection loss from a model solar
Therm Eng 2017;114:1090–104. concentrator cavity receiver [Master of Philosophy Thesis]. Canberra, Australia:
[24] Lin M, Haussener S. Techno-economic modeling and optimization of solar-driven Department of Engineering, Australian National University; 2004.
high-temperature electrolysis systems. Sol Energy 2017;155:1389–402. [48] Lemmon EW, Huber ML, McLinden MO. NIST reference fluid thermodynamic and
[25] AlZahrani AA, Dincer I. Design and analysis of a solar tower based integrated transport properties–REFPROP. NIST standard reference database, vol. 23; 2002.
system using high temperature electrolyzer for hydrogen production. Int J P. v7.
Hydrogen Energy 2016;41(19):8042–56. [49] Boerema N, et al. Liquid sodium versus Hitec as a heat transfer fluid in solar thermal
[26] Balta MT, Kizilkan O, Yılmaz F. Energy and exergy analyses of integrated hydrogen central receiver systems. Sol Energy 2012;86(9):2293–305.
production system using high temperature steam electrolysis. Int J Hydrogen [50] Mohammadi A, Mehrpooya M. Exergy analysis and optimization of an integrated
Energy 2016;41(19):8032–41. micro gas turbine, compressed air energy storage and solar dish collector process. J
[27] Mohammadi A, et al. Exergy analysis of a Combined Cooling, Heating and Power Cleaner Prod 2016;139:372–83.
system integrated with wind turbine and compressed air energy storage system. [51] Mehrpooya M, Hemmatabady H, Ahmadi MH. Optimization of performance of
Energy Convers Manage 2017;131:69–78. Combined Solar Collector-Geothermal Heat Pump Systems to supply thermal load
[28] Kalogirou SA. Solar thermal collectors and applications. Prog Energy Combust Sci needed for heating greenhouses. Energy Convers Manage 2015;97:382–92.
2004;30(3):231–95.
178