Energy Conversion and Management 121 (2016) 186–193

Contents lists available at ScienceDirect

Energy Conversion and Management

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

An efficient hybrid system using a thermionic generator to harvest waste

heat from a reforming molten carbonate fuel cell
Chuankun Huang a,1, Yuzhuo Pan b,1, Yuan Wang a, Guozhen Su a,⇑, Jincan Chen a,⇑
a
Department of Physics, Xiamen University, Xiamen 361005, People’s Republic of China
b
College of Physics and Information Engineering, Quanzhou Normal University, 362000, People’s Republic of China

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

Article history: A new hybrid system consisting of an internal reforming molten carbonate fuel cell (IR-MCFC) and a vac-
Received 18 February 2016 uum thermionic generator (TIG) is constructed. By considering the main irreversible losses in two subsys-
Received in revised form 10 April 2016 tems and the coupling process, general expressions for the power output and efficiency of the hybrid
Accepted 9 May 2016
system are derived. The performance characteristics of the hybrid system are revealed, and the optimal
Available online 18 May 2016
regions, including the power output, efficiency, and load resistances, are determined. It is found that the
maximum power output density of the IR-MCFC/TIG system increases by 22% compared to that of a single
Keywords:
MCFC, and among the IR-MCFC- and MCFC-based hybrid systems reported in the literature, the present
Molten carbonate fuel cell
Internal reforming
hybrid system can most efficiently harvest the waste heat from the MCFC.
Vacuum thermionic generator Ó 2016 Published by Elsevier Ltd.
Hybrid system
Maximum power output density

1. Introduction
Fuel cells represent a class of efficient energy converting sys-
tems that directly turn chemical energy into electricity [1,2]. To
date, various types of fuel cells have been proposed, investigated,
and constructed [3–6]. Among them, the molten carbonate fuel cell
(MCFC), developed in the 1950s, is one of the most promising fuel
cells because of its stationary power output and high efficiency
[7,8]. Many comprehensive studies have been performed in both
theory and experiment [9–14].
Currently, much attention has been paid to reducing the costs of
MCFCs and improving the efficiency of MCFCs [15–17]. Traditional
fuel cells fueled by H2 have the difficulties of hydrogen production,
preservation, and transportation. By applying internal [18–23] or
external reforming technology [24,25], the MCFCs, working at high
temperatures of nearly 600–800 °C, can directly utilize other inex-
pensive and resourceful gases as fuel, such as natural gas, biogas,
landfill gas, and methane [26–31]. This is an attractive business
prospect. Moreover, the recovery of the waste heat produced by
MCFCs is another significant issue in the investigation of the per-
formance of fuel cells. Coupling MCFCs with different thermody-
namic systems, such as gas turbines [32–37] and heat engines
[38–42], has been intensively discussed, but the complex structure
⇑ Corresponding authors.
E-mail addresses: gzsu@xmu.edu.cn (G. Su), jcchen@xmu.edu.cn (J. Chen).
1
The two authors contribute equally to this work.
and low efficiency of these systems may limit their actual use. In
comparison, as one type of high-temperature solid state generator,
the vacuum thermionic generator uses electrons as the working
substance and has advantages that include no mechanically mov-
ing parts, a small device size, and high power density [43–46],
and it may be a better device for using the waste heat from MCFCs.
In the present paper, we establish a new hybrid model com-
posed of an IR-MCFC and a vacuum TIG. Some irreversible losses
in the hybrid system are considered. The expressions of the power
outputs and efficiencies for the two subsystems and hybrid system
are analytically derived. According to the first law of thermody-
namics, an important energy equilibrium equation between the
two subsystems is obtained. The graphs of the power output, effi-
ciency, and load resistances, varying with different parameters, are
generated. The performance characteristics of the IR-MCFC/TIG
hybrid system are revealed, and the optimal regions of some
important parameters are determined. The results obtained here
are compared with those of other IR-MCFC- or MCFC-based hybrid
systems reported in the literature.
2. The hybrid system consisting of an IR-MCFC and a TIG
Fig. 1 shows the hybrid system composed of an IR-MCFC and a
vacuum TIG, where Pf and Pt are, respectively, the power outputs of
the MCFC and the TIG; qr is the reforming energy; ql is the heat loss
from the MCFC to the environment; qc is the heat absorbed by the
cathode of the TIG; and qa is the heat released by the anode of the

http://dx.doi.org/10.1016/j.enconman.2016.05.028
0196-8904/Ó 2016 Published by Elsevier Ltd.
 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193 187

Nomenclature

A0 Richardson-Dushmann constant (A m2 K2) Ta anode temperature of the TIG (K)

Cp,k molar heat capacity of component k at constant pres- U voltage of the MCFC (V)
sure (J mol1 K1) U0 ideal equilibrium potential (V)
e unit charge (C) Uohm ohm overpotential (V)
Eact activation energy (J mol1) V voltage of the TIG (V)
E0 ideal standard potential (V) Vc optimal voltage of the TIG (V)
F Faraday constant (C mol1)
Dg molar Gibbs energy change (J mol1) Greek symbols
Dh molar enthalpy change (J mol1) vCH4 molar fraction of methane
DH_ enthalpy change rate (W) e thermal emissivity
i current density of the TIG (A m2) r Stefen-Boltzmann constant (W m2 K4)
j current density of the MCFC (A m2) dr heat loss ratio in reforming process
jco cut-off current density (A m2) j heat conductivity coefficient (W m2 K1)
jst start-work current density (A m2) g efficiency
k1 area ratio Ua anode work function of the TIG (eV)
kB Boltzmann constant (J K1)
Lm molar latent heat of H2O (J mol1) Subscripts
ne number of electrons
a anode
n_ f molar rate of fuel flowing (mol s1) c cathode
n_ H hydrogen expended rate (mol s1) f MCFC
pa partial pressure of constituent a (atm) l heat loss
P power output (W)
m maximum power output density
P⁄ power output density (W m2) s hybrid system
qa heat released by the TIG anode (W) t TIG
qc heat absorbed by the TIG cathode (W) 1, 2 two states of Ps ¼ Pfm
ql heat loss rate (W)
qr reforming energy rate (W)
Abbreviations
qw waste heat rate (W)
r load resistance (X) MCFC molten carbonate fuel cell
R universal gas constant (J mol1 K1) TIG thermionic generator
R0 reforming energy of methane per mole (J mol1) GT gas turbine
IR internal reforming
S surface area (m2)
T working temperature of the MCFC (K) SCO2 supercritical CO2
T0 environment temperature (K)

2.1. Model description of an IR-MCFC

The IR-MCFC shown in Fig. 1 has no extra supply heat for

reforming a reaction and no extra structure for reforming product
transportation. During the reforming process, pure methane and
steam are utilized as the fuel. The two main reforming reactions
are the steam/methane reforming reaction (SMR): CH4 +
H2O M CO + 3H2 and the water–gas shift reaction (WGS): CO +
H2O M CO2 + H2 [19,23,31]. The overall reforming reaction is
written as
CH4 þ 2H2 O $ CO2 þ 4H2 : ð1Þ
In reality, the fuel utilization and the chemical reactivity in the
internal reforming process are incomplete and depend on the size
of the fuel cell stack, operating conditions, catalyst activity, etc. Such
some incompleteness will reduce the performance of the IR-MCFC
system [18,19,47,48]. In the present model, the effect of the fuel uti-
lization and the incomplete chemical reactions on the system per-
formance will be not discussed quantitatively. The methane is
assumed to be completely used because we mainly focus on the
recovery of the waste heat. Thus, the extra heat provided by the
Fig. 1. The schematic diagram of the hybrid system composed of an internal MCFC during the reforming process can be calculated by
reforming molten carbonate fuel cell and a vacuum thermionic generator.   Z T   Z T 
qr ¼ n_ f vCH4 4 Dhf ðH2 Þ þ
0 0
C p;H2 ds þ Dhf ðCO2 Þ þ C p;CO2 ds
T0 T0
TIG. Identical TIGs can be set in both the top and bottom sides of  Z T   Z T 
0 0
the MCFC. For simplicity, we only draw one equivalent TIG, shown  Dhf ðCH4 Þ þ C p;CH4 ds  2 Dhf ðH2 OÞ þ Lm þ C p;H2 O ds ; ð2Þ
T0 T0
in Fig. 1.
188 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193

_
where nf is the molar rate of fuel flowing, vCH4 is the molar fraction
0
of methane in the fuel stream, Dhf ðkÞ is the molar enthalpy change
of component k at room temperature and barometric pressure, Cp,k
is the molar heat capacity of component k at constant pressure, Lm is
the molar latent heat of H2O, T is the working temperature of the
MCFC, and T0 is the environment temperature.
Within the MCFC, the anode material is typically porous Ni–Cr
or Ni–Al alloys and cathode material is porous lithiated NiO. The
electrochemical reactions at the anode and cathode of the MCFC
are, respectively, H2 þ CO2
3 ! CO2 þ H2 O þ 2e

and
1
O
2 2
þ CO2 þ 2e ! 
CO2
3 .
Faraday’s law shows that the hydrogen
expended rate in the anode is n_ H ¼ 4n_ f vCH4 ¼ jSf =ðne FÞ, where j is
the electric current density of the MCFC, Sf is the effective surface
area of the electron transport within the MCFC, ne is the number Fig. 2. The curve of the voltage output varying with the current density of the
of electrons transferred in the electrochemical reaction, and F is MCFC, where the hollow circles indicate the experimental results of the MCFC
Faraday’s constant. The power output and efficiency of a MCFC obtained in Ref. [51].
can be, respectively, expressed as
Pf ¼ USf j ð3Þ
and
and
 
Pf ne FU Ua
gf ¼ ¼ ; ð4Þ ia ¼ A0 T 2a exp  ; ð7Þ
kB T a
DH_ Dh

where DH_ ¼ n_ H Dh is the enthalpy change rate generated by the where Ta is the temperature of the anode, A0 is the Richardson-
electrochemical reaction, Dh is the molar enthalpy change, and U is Dushmann constant, Ua is the work function of the anode, V is
the voltage output of the MCFC. The voltage output of fuel cells are the voltage output, and kB is the Boltzmann constant. The power
usually affected by multi-overpotentials including the activation, output and efficiency of the vacuum TIG are given by
ohmic, concentration overpotentials. For different types of fuel cells, Pt ¼ VSt ðic  ia Þ ð8Þ
the species and magnitudes of overpotentials are different. Accord-
ing to the semi-empirical model developed by Yuh and Selman and
[11,49–52], the voltage output U is given by Pt
gt ¼ ; ð9Þ
U ¼ U 0  U a  U c  U ohm ; ð5Þ qw
h pffiffiffiffiffiffiffiffiffiffi i
where U 0 ¼ E0 þ ðRT=ne FÞ ln pH2 ;a pO2 ;c pCO2 ;c =pH2 O;a pCO2 ;a is the where St is the total surface area of the electron transport within the
0:42
vacuum TIG and qw is the waste heat rate produced by the MCFC.
ideal equilibrium potential, U a ¼ 2:27  109 jpH2 ;a p0:17 1:0
CO2 ;a pH2 O;a exp Moreover, for the vacuum TIG, the heat absorbed by the cathode
0:43
ðEact;a =RTÞ is the anode overpotential, U c ¼ 7:505  1010 jpO2 ;c and released from the anode can be written as
p0:09 4     
CO2 ;c expðEact;c =RTÞ is the cathode overpotential, Uohm = 0.5  10 j Ua þ 2kB T Ua þ 2kB T a
exp [3016  (1/T  1/923)] is the ohm overpotential, E0 =  Dg/(neF) qc ¼ St Vþ ic  V þ ia þ et rðT 4  T 4a Þ
e e
is the ideal standard potential, Dg is the molar Gibbs energy change,
ð10Þ
R is the universal gas constant, pax;a=c ða ¼ H2 ; CO2 ; H2 OÞ is the partial
pressure of constituent a in the anode or cathode, and and
Eact,a/c is the activation energy in the anode or cathode. Eq. (5) implies  
Ua þ 2kB T Ua þ 2kB T a
that in this model, the concentration overpotential is small compared qa ¼ St ic  ia þ et rðT 4  T 4a Þ ; ð11Þ
to other overpotentials and is negligible [49–51]. By using Eq. (5), the e e
curve of the voltage varying with the current density of the MCFC is where et is the thermal emissivity of the two plates and r is the
plotted in Fig. 2. It can be observed that the theoretical results derived Stefen-Boltzmann constant. The first two terms on the right side
from the model are very well consistent with experimental data of Eqs. (10) and (11) represent the energies carried by the electrons
obtained in Ref. [51]. This shows that the model of the MCFC from two electrodes, and the last term describes the thermal radia-
described above may be validly used to discuss the performance of tion loss between two plates.
the hybrid system.
2.3. Power output and efficiency of the hybrid system
2.2. Model description of a vacuum TIG
Based on Eqs. (3), (4), (8), and (9), the power output and effi-
As shown in Fig. 1, the vacuum TIG can absorb the waste heat
ciency of the hybrid system can be expressed as
from the MCFC and generate extra electricity, where two parallel
plates are set as the cathode and anode of the TIG. The temperature Ps ¼ P f þ Pt ¼ USf j þ VSt ðic  ia Þ ð12Þ
of the cathode is assumed to be identical to that of the MCFC, and
the electron transport between the two electrodes is assumed to be and
ballistic. Therefore, the current densities from the TIG cathode and Ps
gs ¼ ¼ gf þ gt ð1  gf  dr Þ; ð13Þ
anode can be, respectively, expressed as [45,46] DH_
 
Ua þ eV _ ¼ R0 =ð4DhÞ is defined as the heat loss ratio
ic ¼ A0 T 2 exp  ð6Þ where dr ¼ qr =ðDHÞ
kB T
in the reforming process of the IR-MCFC and R0 ¼ qr =ðn_ f vCH4 Þ.
 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193 189

2.4. An energy equilibrium equation Table 1

Thermodynamic parameters of component k in reforming reactions, where (g) and (l)
refer to gas and liquid phases, respectively, and Lm is the molar latent heat of H2O.
According to the first law of thermodynamics, the energy equi-
librium between the MCFC and the cathode of the TIG needs to be Component Dh0f (k) Lm Molar heat capacity
considered and can be expressed as k (J mol1) (J mol1) C p;k (J mol1 K1)

qw ¼ ql þ qc ¼ DH_  Pf  qr ;
H2(g) 0 – 29.0856  0.0008373T
ð14Þ + 0.0000020138T2
CO2(g) 393,800 – 26.0167
where the heat loss rate is
+ 0.0435259T  0.0000148422T2
h
i
CH4(g) 75,000 – 14.1555
ql ¼ Sl jðT  T 0 Þ þ ef r T 4  T 40 ; ð15Þ + 0.0755466T  0.0000180032T2
H2O(g) – – 30.00 + 0.01071T + 33,000/T2
where Sl is the heat loss area of the MCFC, j is the heat conductivity H2O(l) 285,800 40,700 75.44
coefficient, and ef is the thermal emissivity of the MCFC. Because
there are fewer lateral surfaces of the MCFC compared to the top
and bottom surfaces, the heat leakage loss area may be regarded Table 2
as Sl  2Sf  St, where Sf is the bottom or top surface area of the The parameters adopted in an MCFC [11,31,42,51,58].
MCFC and St is the total contact area between the TIG and MCFC. Parameters Values
Substituting Eqs. (2), (3), (8), (10), and (15) into Eq. (14), one can
Faraday constant, F (C mol1) 96,485
obtain
Number of electrons, ne 2
     Universal gas constant, R (J mol1 K1)
Ua þ 2kB T Ua þ 2kB T a 8.314
k1 Vþ ic  V þ ia þ et rðT 4  T 4a Þ Partial pressure of H2 in the anode, pH2 ;a (atm) 0.6
e e
" # Partial pressure of CO2 in the anode, pCO2 ;a (atm) 0.058
jðT  T 0 Þ T  T 40
4 Partial pressure of H2O in the anode, pH2 O;a (atm) 0.342
þ ef rT 30 ð2  k1 Þ þ Partial pressure of N2 in the cathode, pN2 ;c (atm) 0.59
ef rT 30 T 30 Partial pressure of O2 in the cathode, pO2 ;c (atm) 0.08
¼ j½Dh=ðne FÞ  U  R0 =ð4ne FÞ; Partial pressure of CO2 in the cathode, pCO2 ;c (atm) 0.08
Partial pressure of H2O in the cathode, pH2 O;c (atm) 0.25
ð16Þ Environment temperature, T0 (K) 298
Working temperature of the MCFC, T (K) 923
where k1 = St/Sf .
Molar enthalpy change at 923 K, Dh (J mol1) 247,430
By using the data listed in Tables 1–3 and Eq. (16), k1 can be Molar Gibbs energy change at 923 K, Dg (J mol1) 197,000
solved for using the given values of j and V. As shown in Fig. 3, Activation energy in the anode, Eact,a (J mol1) 53,500
k1 increases with an increase of j and V. At the same current den- Activation energy in the cathode, Eact,c (J mol1) 77,300
sity, the larger the value of V is, the larger the value of k1. However,
for a vacuum TIG, the large value of V impedes the electron trans-
port between the two plates, while the large value of k1 increases Table 3
the heat absorbed by the TIG. For a hybrid system, a suitable volt- The parameters adopted in a vacuum TIG [46].

age of the TIG is required. Parameters Values

Unit charge, e (C) 1.6  1019
3. Performance characteristics Stefen-Boltzmann constant, r (W m2 K4) 5.67  108
Richardson-Dushmann constant, A0 (A m2 K2) 1.2  106
Bolzmann constant, kB (J K1) 1.38  1023
Eqs. (12) and (13) show that the power output and efficiency of
Anode temperature of the TIG, Ta (K) 500
the hybrid system depend on both the current density j of the Anode work function of the TIG, Ua (eV) 1.0
MCFC and the voltage output V of the TIG, as shown in Fig. 4, where
P s ¼ P s =Sf is the power output density of the hybrid system, Psm is
the maximum power output density of the hybrid system, jsm and
Vc are the values of the current density j and voltage output V at
P sm , and Pfm is the maximum power output density of the MCFC.
Fig. 4(a) shows that the power output density has a maximum
value, Psm , and the hybrid system should be operated in the area
surrounded by the dashed line Ps ¼ Pfm . Note that the optimal volt-
age at the maximum power output density, marked as Vc in Fig. 4
(a), does not vary with j for other given parameters. Using Eqs. (12)
and (16) and the extremal conditions (o Ps/o V)j = (o Pt/o V)j = 0, one
can derive the analytical expression of Vc. Fig. 4(b) shows that the
efficiency decreases monotonously with an increase of j, and when
the current density is small, the efficiency is not affected by the
voltage output V.
Fig. 5 further illustrates the j  g and j  P⁄ curves of the two
subsystems and hybrid system for the optimal voltage V = Vc,
where P⁄ = P/Sf is the power output density; the red2 dashed, black
dash-dot, and blue solid lines correspond to the MCFC, TIG, and
hybrid system, respectively; jfm and jsm are the current densities at
the maximum power output densities P fm and P sm , respectively; jst

2
For interpretation of color in Figs. 5 and 6, the reader is referred to the web Fig. 3. The three-dimensional graph of the surface area ratio k1 varying with the
version of this article. current density j and the voltage output V.
190 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193

Fig. 4. The three-dimensional projective graphs of (a) the power output density P s
and (b) the efficiency g of the hybrid system varying with j and V. Fig. 5. The curves of (a) the power output densities and (b) the efficiencies of the
MCFC, TIG, and hybrid system varying with j, where V has been optimized, i.e.,
V = Vc = 0.41 V.
is the current density at which the TIG starts to work; and jco is the
cut-off current density at which the MCFC generates zero power out-
put and efficiency. Fig. 5(a) shows that the j  P⁄ curves of the MCFC
and hybrid system are parabolic, whereas the j  P⁄ curve of the TIG
is monotonously incremental. Fig. 5(b) shows that, when the current
density increases, the efficiencies of the MCFC and hybrid system
decrease, whereas the efficiency of the TIG increases. According to
Fig. 5(a) and (b), when the current density is small, the heat pro-
duced by the MCFC is mainly consumed by reforming reactions,
and consequently, the remaining part is too small to drive the vac-
uum TIG. When the current density j > jst, the TIG gains enough heat
and begins to work. When the current density is close to jco, a large
amount of waste heat is absorbed by the TIG, and the efficiency of
the TIG is larger than that of the hybrid system. This type of phe-
nomenon can be seen in Fig. 5(b) or directly explained by Eq. (13).
When j = jco, gf = 0, and it can be derived from Eq. (13) that
gt ¼ gs =ð1  dr Þ P gs , where the equal sign is true only for dr = 0. Fig. 6. The power output densities of the MCFC and hybrid system versus the
In Fig. 6, the power output density as a function of the efficiency efficiency for V = 0.41 V, where gsm is the efficiency of the hybrid system at the
for V = Vc is plotted, where the red dashed and blue solid lines cor- maximum power output density P sm and the shadow area is the optimally working
respond to the MCFC and hybrid system, respectively. It can be region of the hybrid system.
seen from Fig. 6 that, in the whole operating region, the power out-
when V = Vc, both (g2  gsm) and (g2  g1) attain their respective
put density Ps of the hybrid system is larger than that of the MCFC.
maximum values. In the region where V < Vc, g2 increases, whereas
The maximum power output density of the hybrid system is
both g1 and gsm decrease, as V increases. In the region where V > Vc,
2010 W/m2 and is approximately 1.22 times that of the maximum
g2 decreases, whereas both g1 and gsm increase, as V increases.
power output density of 1650 W/m2 of a single MCFC. Note that
From Figs. 6 and 7, it can be further found that the optimal
Ps ðg1 Þ ¼ P fm ¼ Ps ðg2 Þ. In the region where P s > P fm , gs satisfies the
working region of the hybrid system should be determined by
relation g1 < gs < g2, and the performance of the hybrid system is
always better than that of the MCFC at the maximum power output Pfm < Ps 6 Psm ð17Þ
density P fm . It is found that, when V = Vc, the region of P s > Pfm is
and
maximum, as shown in Fig. 6. To expound this problem, the curves
of g1, g2 and gsm varying with V are plotted in Fig. 7. It is noted that, gsm 6 gs < g2 ; ð18Þ
 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193 191

Fig. 8. The curves of the two load resistances of the MCFC and TIG varying with j,
where r f =t ¼ r f =t Sf and j2 is the current density at the efficiency g2.
Fig. 7. The curves of the efficiencies g1, g2 and gsm varying with V.

which are, respectively, indicated by the areas surrounded by the

Ps ¼ P fm curve shown in Fig. 4 and the V  gsm and V  g2 curves
shown in Fig. 7. The optimal region simultaneously determined by
Eqs. (17) and (18) is indicated by the shadow area shown in
Fig. 6. In this optimal region, the power output density decreases
with an increase of the efficiency. Thus, one should make gs as close
as possible to gsm so that the hybrid system can obtain the largest
possible power output density.
To make the hybrid system operate within the optimal region,
the matching problem of two subsystems must be solved. By using
Eqs. (5)–(7), the load resistances of the MCFC and TIG can be
expressed as

U
rf ¼
jSf
and
V Refs. [53–57], i.e., (i) IR-MCFC/GT fueled by biogas [55], (ii) IR-MCFC/GT fueled by
rt ¼ : ð20Þ
ðic  ia ÞSt
According to Eqs. (19) and (20), we can plot the curves of the
two load resistances varying with the current density j, as shown
in Fig. 8, where r f =t ¼ rf =t Sf , j2 is the current density corresponding
to g2, ðr f Þ2 and ðrf Þsm are the values of rf when j = j2 and j = jsm, and
ðrt Þ2 and ðrt Þsm are the values of rt when j = j2 and j = jsm. According
to Eq. (18), the optimal regions of r f and rt are determined by
ðrf Þ2 > r f P ðr f Þsm
and
ðrt Þ2 > r t P ðr t Þsm :
Fig. 8 shows that, in the optimal region of j2 < j 6 jsm , both r f
and r t decrease with an increase of j. In the design process, one
should, respectively, make rf and rt as close as possible to ðrf Þsm
and ðrt Þsm .
4. Discussion
The above analyses show that the present IR-MCFC/TIG hybrid
system can effectively utilize the waste heat and increase the
power output of the hybrid system at the working temperature
of T = 923 K of the MCFC. We further calculate the power output
densities of this hybrid system at different working temperatures
of the MCFC and evaluate their percentage improvements com-
pared to the maximum power output density of a single MCFC,
as shown by the black solid line in Fig. 9. The figure shows the
ð19Þ
Fig. 9. The percentage improvement of the maximum power output density of the
(IR-)MCFC/TIG hybrid system compared to that of a single MCFC, where several
symbols correspond to other (IR-)MCFC-based hybrid systems that are reported in
natural gas [53,54], (iii) MCFC/SCO2-Brayon cycle [57], (iv) MCFC/GT [56], (v) MCFC/
Stirling engine [56], and (vi) MCFC/SCO2 turbine [56].
higher the working temperature of the MCFC is, the larger the per-
centage improvement of the IR-MCFC/TIG system. When T varies
from 873 K to 923 K, the percentage improvements of the maxi-
mum power output density of the hybrid system can increase from
10.5% to 22% compared to that of a single MCFC and are larger than
ð21Þ those of the two IR-MCFC/GT systems [53–55], marked as (i) and
(ii) in Fig. 9.
Additionally, the proposed model can be conveniently used to
evaluate the performance of the MCFC/TIG hybrid system directly
ð22Þ using hydrogen as the fuel. In this condition, the internal reforming
process is unnecessary, and the IR-MCFC/TIG hybrid system
becomes an MCFC/TIG system. The MCFC/TIG hybrid system is able
to generate a higher power output density and shows a larger per-
centage improvement over the IR-MCFC/TIG hybrid system, as
shown by the red dashed line in Fig. 9, where the percentage
improvements of other MCFC-based hybrid systems reported in
Refs. [56,57] are also marked. When T varies from 873 K to
923 K, the percentage improvements of the maximum power out-
put density of the MCFC/TIG system can increase from 19% to 33%
compared to that of a single MCFC and are about 10% more than
those of the IR-MCFC/TIG system. On the other hand, system (iii)
is the most efficient configuration among other MCFC-based hybrid
systems [56]. The percentage improvement of this configuration
can theoretically be 22% at 897 K, but it is still smaller than that
of the MCFC/TIG system. It is seen from Fig. 9 that, no matter which
temperature the (IR-)MCFC is operated at, the performance of the
192 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193
(IR-)MCFC/TIG system is always superior to the performances of
the other (IR-)MCFC-based hybrid systems reported in literature.
In addition, the (IR-)MCFC/TIG system has some advantages over
other (IR-)MCFC-based hybrid systems, such as no mechanically
moving parts, a small device size, and a high power density of
the vacuum TIG. Consequently, this hybrid system is worth study-
ing further.
5. Conclusions
Based on the current models of an IR-MCFC and a vacuum TIG,
the theoretical expressions for the power output and efficiency of
the IR-MCFC/TIG hybrid system are derived. By considering the
main irreversible losses in the hybrid system and the energy equi-
librium equation between two subsystems, some influential
parameters are analyzed and graphs of the power output, effi-
ciency, and load resistances are drawn. The performance character-
istics of the hybrid system are revealed and the optimally working
regions are determined. The maximum power output density of
the hybrid system with or without internal reforming can increase
by 22% or 33% compared to that of a single MCFC. Among IR-MCFC-
and MCFC-based hybrid systems, the present hybrid system is the
most efficient hybrid configuration for harvesting the waste heat
from the MCFC.
Acknowledgements
This work has been supported by the National Natural Science
Foundation (No. 11405142), People’s Republic of China.
References
[1] Larminie J, Dicks A. Fuel cell systems explained. Chichester: Wiley; 2003.
[2] Fuel Cell Handbook. Morgantown: U.S. Department of Energy; 2004.
[3] Ormerod RM. Solid oxide fuel cells. Chem Soc Rev 2003;32(1):17–28.
[4] Mekhilef S, Saidur R, Safari A. Comparative study of different fuel cell
technologies. Renewable Sustainable Energy Rev 2012;16(1):981–9.
[5] Yang P, Zhang H. Parametric analysis of an irreversible proton exchange
membrane fuel cell/absorption refrigerator hybrid system. Energy
2015;85:458–67.
[6] Zhao M, Zhang H, Hu Z, Zhang Z, Zhang J. Performance characteristics of a
direct carbon fuel cell/thermoelectric generator hybrid system. Energy Convers
Manage 2015;89:683–9.
[7] Sundmacher K, Kienle A, Josef Pesch H, Berndt JF, Huppmann G. Molten
carbonate fuel cells: modeling, analysis, simulation,
control. Weinheim: WILEY-VCH; 2007.
[8] Antolini E. The stability of molten carbonate fuel cell electrodes: a review of
recent improvements. Appl Energy 2011;88(12):4274–93.
[9] Baranak M, Atakül H. A basic model for analysis of molten carbonate fuel cell
behavior. J Power Sources 2007;172(2):831–9.
[10] Yu L-j, Ren G-p, Jiang X-m. Experimental and analytical investigation of molten
carbonate fuel cell stack. Energy Convers Manage 2008;49(4):873–9.
[11] Zhang H, Lin G, Chen J. Performance analysis and multi-objective optimization
of a new molten carbonate fuel cell system. Int J Hydrogen Energy 2011;36
(6):4015–21.
[12] Sánchez D, Monje B, Chacartegui R, Campanari S. Potential of molten carbonate
fuel cells to enhance the performance of CHP plants in sewage treatment
facilities. Int J Hydrogen Energy 2013;38(1):394–405.
[13] Milewski J, Discepoli G, Desideri U. Modeling the performance of MCFC for
various fuel and oxidant compositions. Int J Hydrogen Energy 2014;39
(22):11713–21.
[14] Yazdanfar J, Mehrpooya M, Yousefi H, Palizdar A. Energy and exergy analysis
and optimal design of the hybrid molten carbonate fuel cell power plant and
carbon dioxide capturing process. Energy Convers Manage 2015;98:15–27.
[15] Kulkarni A, Giddey S. Materials issues and recent developments in molten
carbonate fuel cells. J Solid State Electrochem 2012;16(10):3123–46.
[16] Agll AAA, Hamad YM, Hamad TA, Thomas M, Bapat S, Martin KB, et al. Study of
a molten carbonate fuel cell combined heat, hydrogen and power system:
energy analysis. Appl Therm Eng 2013;59(1–2):634–8.
[17] Campanari S, Chiesa P, Manzolini G, Bedogni S. Economic analysis of CO2
capture from natural gas combined cycles using molten carbonate fuel cells.
Appl Energy 2014;130:562–73.
[18] Shinoki T, Matsumura M, Sasaki A. Development of an internal reforming
molten carbonate fuel cell stack. Energy Convers, IEEE Trans 1995;10
(4):722–9.
[19] Cavallaro S, Freni S. Syngas and electricity production by an integrated
autothermal reforming/molten carbonate fuel cell system. J Power Sources
1998;76(2):190–6.
[20] Fellows R. A novel configuration for direct internal reforming stacks. J Power
Sources 1998;71(1–2):281–7.
[21] Lukas MD, Lee KY, Ghezel-Ayagh H. Development of a stack simulation model
for control study on direct reforming molten carbonate fuel cell power plant.
IEEE Trans Energy Convers 1999;14(4):1651–7.
[22] Heidebrecht P, Sundmacher K. Dynamic modeling and simulation of a
countercurrent molten carbonate fuel cell (MCFC) with internal reforming.
Fuel Cells 2002;2(3–4):166–80.
[23] Heidebrecht P, Sundmacher K. Molten carbonate fuel cell (MCFC) with internal
reforming: model-based analysis of cell dynamics. Chem Eng Sci 2003;58(3–
6):1029–36.
[24] Musa A, Steeman HJ, De Paepe M. Performance of internal and external
reforming molten carbonate fuel cell systems. J Fuel Cell Sci Technol 2006;4
(1):65–71.
[25] Campanari S, Manzolini G, Chiesa P. Using MCFC for high efficiency CO2
capture from natural gas combined cycles: comparison of internal and
external reforming. Appl Energy 2013;112:772–83.
[26] Bove R, Lunghi P. Experimental comparison of MCFC performance using three
different biogas types and methane. J Power Sources 2005;145(2):588–93.
[27] Frusteri F, Freni S. Bio-ethanol, a suitable fuel to produce hydrogen for a
molten carbonate fuel cell. J Power Sources 2007;173(1):200–9.
[28] Chiodo V, Urbani F, Galvagno A, Mondello N, Freni S. Analysis of biogas
reforming process for molten carbonate fuel cells. J Power Sources
2012;206:215–21.
[29] Hamad TA, Agll AA, Hamad YM, Bapat S, Thomas M, Martin KB, et al. Study of
combined heat, hydrogen and power system based on a molten carbonate fuel
cell fed by biogas produced by anaerobic digestion. Energy Convers Manage
2014;81:184–91.
[30] Buonomano A, Calise F, Ferruzzi G, Palombo A. Molten carbonate fuel cell: an
experimental analysis of a 1 kW system fed by landfill gas. Appl Energy
2015;140:146–60.
[31] Zhang X, Liu H, Ni M, Chen J. Performance evaluation and parametric optimum
design of a syngas molten carbonate fuel cell and gas turbine hybrid system.
Renewable Energy 2015;80:407–14.
[32] Lunghi P, Ubertini S, Desideri U. Highly efficient electricity generation through
a hybrid molten carbonate fuel cell-closed loop gas turbine plant. Energy
Convers Manage 2001;42(14):1657–72.
[33] Grillo O, Magistri L, Massardo AF. Hybrid systems for distributed power
generation based on pressurisation and heat recovering of an existing 100 kW
molten carbonate fuel cell. J Power Sources 2003;115(2):252–67.
[34] Liu A, Weng Y. Performance analysis of a pressurized molten carbonate
fuel cell/micro-gas turbine hybrid system. J Power Sources 2010;195
(1):204–13.
[35] Leto L, Dispenza C, Moreno A, Calabrò A. Simulation model of a molten
carbonate fuel cell–microturbine hybrid system. Appl Therm Eng 2011;31(6–
7):1263–71.
[36] Zhang X, Guo J, Chen J. Influence of multiple irreversible losses on the
performance of a molten carbonate fuel cell-gas turbine hybrid system. Int J
Hydrogen Energy 2012;37(10):8664–71.
[37] Dimopoulos GG, Stefanatos IC, Kakalis NMP. Exergy analysis and optimisation
of a marine molten carbonate fuel cell system in simple and combined cycle
configuration. Energy Convers Manage 2016;107:10–21.
[38] Zhao Y, Chen J. Modeling and optimization of a typical fuel cell–heat engine
and
hybrid system and its parametric design criteria. J Power Sources 2009;186
(1):96–103.
[39] Chen X, Lin B, Chen J. General performance characteristics and parametric
optimum criteria of a Braysson-based fuel cell hybrid system. Energy Fuels
2009;23(12):6079–84.
[40] Sánchez D, Chacartegui R, Torres M, Sánchez T. Stirling based fuel cell hybrid
systems: an alternative for molten carbonate fuel cells. J Power Sources
2009;192(1):84–93.
[41] Zhang X, Chen J. Performance analysis and parametric optimum criteria of a
class of irreversible fuel cell/heat engine hybrid systems. Int J Hydrogen
Energy 2010;35(1):284–93.
[42] Zhang HC, Lin GX, Chen JC. Performance evaluation and parametric optimum
criteria of an irreversible molten carbonate fuel cell-heat engine hybrid
system. Int J Electrochem Sci 2011;6(10):4714–29.
[43] Xuan XC. Combined thermionic-thermoelectric refrigerator. J Appl Phys
2002;92(8):4746–50.
[44] Xuan XC, Li D. Optimization of a combined thermionic–thermoelectric
generator. J Power Sources 2003;115(1):167–70.
[45] Chen L, Ding Z, Sun F. Performance analysis of a vacuum thermionic
refrigerator with external heat transfer. J Appl Phys 2010;107(10):104507.
[46] Wang Y, Su S, Lin B, Chen J. Parametric design criteria of an irreversible
vacuum thermionic generator. J Appl Phys 2013;114(5):053502.
[47] Clarke SH, Dicks AL, Pointon K, Smith TA, Swann A. Catalytic aspects of the
steam reforming of hydrocarbons in internal reforming fuel cells. Catal Today
1997;38(4):411–23.
[48] Lukas MD, Lee KY, Ghezel-Ayagh H. Modeling and cycling control of carbonate
fuel cell power plants. Control Eng Pract 2002;10(2):197–206.
[49] Yuh CY, Selman JR. The polarization of molten carbonate fuel cell electrodes: I.
analysis of steady-state polarization data. J Electrochem Soc 1991;138
(12):3642–8.
 C. Huang et al. / Energy Conversion and Management 121 (2016) 186–193
[50] Yuh CY, Selman JR. The polarization of molten carbonate fuel cell electrodes: II.
characterization by AC impedance and response to current interruption. J
Electrochem Soc 1991;138(12):3649–56.
[51] Brouwer J, Jabbari F, Leal EM, Orr T. Analysis of a molten carbonate fuel cell:
numerical modeling and experimental validation. J Power Sources 2006;158
(1):213–24.
[52] Lee C-G, Hwang J-Y, Oh M, Kim D-H, Lim H-C. Overpotential analysis with
various anode gas compositions in a molten carbonate fuel cell. J Power
Sources 2008;179(2):467–73.
[53] Ghezel-Ayagh H, Walzak J, Patel D, Daly J, Maru H, Sanderson R, et al. State of
direct fuel cell/turbine systems development. J Power Sources
2005;152:219–25.
[54] Ghezel-Ayagh H, Walzak J, Junker ST, Patel D, Michelson F, Adriani A. DFC/TÒ
power plant: from sub-megawatt demonstration to multi-megawatt design.
ECS Trans 2008;12(1):713–7.
193
[55] Huang H, Li J, He Z, Zeng T, Kobayashi N, Kubota M. Performance analysis of a
MCFC/MGT hybrid power system bi-fueled by city gas and biogas. Energies
2015;8(6):5661.
[56] Muñoz de Escalona JM, Sánchez D, Chacartegui R, Sánchez T. Performance
analysis of hybrid systems incorporating high temperature fuel cells and
closed cycle heat engines at part-load operation. Int J Hydrogen Energy
2013;38(1):570–8.
[57] Baronci A, Messina G, McPhail SJ, Moreno A. Numerical investigation of a MCFC
(Molten Carbonate Fuel Cell) system hybridized with a supercritical CO2
Brayton cycle and compared with a bottoming organic rankine cycle. Energy
2015;93:1063–73.
[58] Koh JH, Kang BS, Lim HC. Analysis of temperature and pressure fields in molten
carbonate fuel cell stacks. AlChE J 2001;47(9):1941–56.