a Department of Chemical Engineering, Faculty of Engineering, Burapha University, Chonburi 20131, Thailandb Department of Chemical Engineering, Faculty of Engineering, Srinakharinwirot University, Nakorn Nayok 26120, Thailandc School of Chemical Engineering, Faculty of Engineering, King Mongkut's Institute of Technology Ladkrabang, Bangkok 10520, Thailandd Computational Process Engineering Research Unit, Department of Chemical Engineering, Faculty of Engineering, Chulalongkorn University, Bangkok10330, Thailand
a r t i c l e i n f o
Article history:Received 18 June 2015Received in revised form1 October 2015Accepted 31 October 2015Available online 21 November 2015
Keywords:Solid oxide fuel cellGas turbineEthanolAnodeecathode gas recirculationTurbine inlet temperature
A solid oxide fuel cell-gas turbine (SOFC-GT) hybrid system supplying liquid fuel as ethanol exhibitspromise as an auxiliary power unit. In this study, the recirculation of anode and cathode exhaust gas inthe SOFC-GT system is proposed to improve the efficiency of heat management in the SOFC-GT hybridsystem. The key operating parameters, such as fuel utilization factor and the cell and GT temperatures,are analyzed in terms of the performance of the SOFC-GT hybrid systems. The simulation results showthat the recirculation of anode and cathode exhaust gas has a direct impact on the turbine performance.To maintain the inlet temperature of the small turbine in the range of 873e1223 K, the amount of fueland air added to the combustor to control the turbine inlet temperature on the system performance isalso investigated. A SOFC-GT hybrid system with both anode and cathode exhaust gas recirculationachieves the highest system and thermal efficiency.
The world population increase drives the increase in energydemand. Moreover, pollutant emissions from conventional powerplants are a serious environmental problem. To address the aboveproblems, researchers are seeking to develop more efficient powergeneration. The solid oxide fuel cell-gas turbine (SOFC-GT) hybridsystem is an alternative power generation solution for distributedpower generation market due to its clean electrical energy pro-duction and high efficiency. In an SOFC-GT hybrid system, thewasteheat from SOFC can be used directly in the gas turbine as the bot-tom cycle. The efficiency of this hybrid system can reach up to 70%[1].
Most SOFC systems run on natural gas; however, the deficiencyof fossil fuels as a limited and non-renewable resource has beenincreasing even though it is a cost-effective feedstock. Ethanol iscurrently considered a promising fuel for SOFCs because it is arenewable fuel produced from agricultural products. It has
þ66 2 218 6877.rnwichanop).
relatively high hydrogen content and is easy to store, handle, andtransport safely [2,3]. Hence, ethanol is a suitable fuel for remoteareas where a natural gas pipeline network is unavailable, as well asa promising alternative fuel for medium-scale power generation inSOFC-GT hybrid systems.
The variation in fuel type supplied for the SOFC-GT hybrid sys-tem has a great impact on the optimal operational parameters forthe system and the system outputs [4,5]. Thus, the design andoperational condition of the SOFC-GT hybrid system should besuitable for the fuel type in the system. In the utilization of ethanolas fuel for the SOFC system, a direct ethanol supply to the SOFCleads to the degradation of anode catalysts due to carbon formation[6]. To avoid this problem, an external reformer is integrated withthe SOFC system for hydrogen production. However, the externalreformer for ethanol steam reforming requires the high heat inputbecause of its strong endothermic reaction. Additionally, theexternal reforming SOFC system requires a higher air supply forSOFC than the direct internal reforming SOFC system [7]. Theincreased air flow to the system has an adverse impact on thesystem efficiency because high heat input is needed to preheat airbefore introduction to the SOFC. The aforementioned problems
require efficient energy management in the SOFC-GT hybridsystem.
Generally, themajority of SOFC-GT hybrid system configurationscan be divided into atmospheric and pressurized SOFC-GT hybridsystems. Comparatively, the pressurized hybrid system has 7.3%higher efficiency and a 20% lower rate of destruction exergy thanthe atmospheric hybrid system [8]. In a pressurized SOFC-GT hybridsystem, the working fluid of SOFC is directly used in the turbine,resulting in less heat loss from the system; however, the operationsof the two main subsystems between SOFC and GT are related andrather complicated. Due to the different operations of the two units,the SOFC-GT hybrid system has been studied in various ways.Chinda and Brault [9] found that the SOFC-GT system performanceis limited by the cell temperature, the turbine inlet temperature,and the exhaust temperature, and the air mass flow rate for thesystem affects the variation of the heat exchanger properties.Bakalis and Stamatis [10] studied the operating parameters of theSOFC-GT hybrid system based on a SOFC scheme without anymodification, i.e., optimum compressor and turbine geometriesused to find the desired operating parameters for high system ef-ficiency. Additionally, the matched operating condition of the twomain units for the SOFC-GT hybrid system should be considered thelimit of safe operation. Song et al. [11] investigated the SOFC-GThybrid system considering the maximum allowable cell tempera-ture. Bakalis and Stamatis [12] studied an SOFC coupledwithmicro-turbines. The limit of operating conditions is considered for para-metric analysis to avoid system malfunctions. The proper designand management of the heat recovery are important to the per-formance of the hybrid system. Barelli et al. [13] studied heat re-covery from the turbine exhaust gas used to heat the air inlet ofSOFC, producing steam and heating for the steam reformer, by arecuperative heat exchanger. The thermal power loss for therecuperative heat exchanger is 10%.
The recirculation of SOFC exhaust gases affects the heat man-agement within the system. Our previous work [14] showed thatthe recirculation of anode exhaust gas can increase the system ef-ficiency for an SOFC system integrated with an ethanol reformerprocess by reducing the heat input for the preheating steam used inthe steam reformer. When considering an SOFC-GT hybrid systemwith the recirculation of cathode exhaust gas, this system can in-crease thermal efficiency and avoid a high-temperature heatexchanger, which is costly [15,16]. Thus, an SOFC-GT hybrid systemwith anode and cathode exhaust gas recycling systems is aninteresting power system that can minimize the required heatinput for the system and maximize the system efficiency. Jia et al.[17] compared the internal-reforming SOFC system with three gasrecycling types, and the results indicated that the systemwith bothanode and cathode recycling has the highest efficiency byapproximately 52%. However, there are few studies that perform adetailed analysis of the influence of recirculation of both anode andcathode exhaust gas on the performance of SOFC-GT hybrid sys-tems. The anode and cathode gas recirculation systems have adirect impact on the SOFC and gas turbine performance. Further-more, they also affect the heat recovery within the hybrid system.To fully understand the effects of the anode and cathode off-gasrecirculation on both the SOFC and gas turbine performances, aparametric analysis of the SOFC-GT hybrid system with the recir-culation of anode and cathode exhaust gas should be investigated.
In this regard, the objective of this study is to investigate theeffects of anode and cathode gas recirculation on the performanceof an SOFC-GT hybrid system combined with an external ethanolsteam reformer. The effects of matching design parameters be-tween the operation of SOFC and GT for SOFC-GT systems withanode or/and cathode recirculation were investigated. To examinethe effect of matching design parameters between two units on the
performance of different hybrid systems, the fuel utilization factorand SOFC operating temperature, which directly affect the perfor-mance of the SOFC-GT hybrid system, were studied first. Next, theadditional fuel and air supply to the combustor, which are neededto specify the turbine inlet temperature in the SOFC-GT hybridsystemwith the recirculation of cathode or/and anode exhaust gas,were also investigated.
2. Configuration of SOFC-GT systems
Fig. 1 shows the integrated systems of the ethanol reformer andSOFC with gas turbine cycles. Three SOFC system designs areconsidered: (1) SOFC-GT system with anode recirculation (AR), (2)SOFC-GT system with cathode recirculation (CR) and (3) SOFC-GTsystem with both the anode and cathode recirculation (ACR).Ethanol as a fuel is pumped, mixed with steam and reformed tosynthesis gas in an external reformer. The synthesis gas with richhydrogen is then sent to SOFC stack. Since an exhaust gas at theanode of SOFC is composed of residual fuel and steam, which can beused for the reforming of ethanol, a portion of the exhaust gas canbe recycled to the external steam reformer. This configuration isknown as the SOFC-GT system with the anode recirculation. Theremaining anode exhaust gas is mixed with an outlet oxidant gasfrom the cathode in a combustor. The combustion gas goes througha high-temperature heat exchanger before it is fed to a turbine toproduce electricity. The exiting gas from the turbine is sent to arecuperator to preheat a compressed air. For the hybrid SOFC-GTsystem with the cathode recirculation, a portion of the cathodeexhaust gas is recycled and mixed with a fresh air from the recu-perator and then sent to the SOFC. The SOFC-GT system with boththe anode and cathode recirculation is similar to that with thecathode recirculation but includes the anode recirculation (dashline in Fig 1b).
3. Model of the SOFC-GT hybrid system
A mathematical model is implemented to evaluate the systemperformance. This work focuses on the study of the configurationdesign in an SOFC-GT hybrid system. The mathematical models ofthe SOFC, gas turbine, and auxiliary units are derived from massand energy balances under steady-state operation. The main as-sumptions used for the simulation of system behavior are asfollows:
(1) All gases are considered as ideal gases.(2) Pressure drops in the SOFC are negligible.(3) Cathode and anode outlet temperatures are equal.(4) The SOFCmodel is considered as a one-dimensional variation
of parameters in the x-direction, and other components inthe system are taken as a lumped control volume.
(5) Heat losses to the environment are negligible.
3.1. SOFC
The electricity from the SOFC is produced by the electrochemicalreaction between hydrogen and oxygen. In the air channel, oxygenis reduced into oxygen ions. Then, the oxygen ions pass through theelectrolyte to the anode/electrolyte interface. Hydrogen in the fuelchannel reacts with oxygen ions, producing water and electrons atthe anode side. The electrons flow from the anode side to thecathode side, generating electricity. The electrochemical reaction isas follows:
Oxygen reduction reaction: 1/2O2 þ 2e� / O2� (1)
(a)
(b)
Mixer
SOFC
Reformer
Ethanol
Water
l
Vaporizer
Vaporizer
Combustion chamber
Air
TurbineCompressor
Blower
Cat
hode
Ano
de
Fuel
Air
Combustion chamber
Reformer
SOFC
Air
TurbineCompressor
Mixer
Water Vaporizer
MixerEthanol
Vaporizer
Cat
hode
Ano
de
Fuel
Air
Fig. 1. Schematic diagrams of the pressurized SOFC-GT hybrid system (a) SOFC-GT with anode exhaust gas recirculation and recuperative heat exchanger (b) SOFC-GT with cathodeexhaust gas recirculation and SOFC-GT with anode and cathode exhaust gas recirculation (dashed line).
In this paper, the syngas from the reformer is introduced to thefuel channel of the SOFC. The syngas consists of methane and car-bon monoxide; therefore, the steam reforming process occurringon the anode side is considered in this model as described in Eqs.(4) and (5).
The component gases in the fuel channel consist of CH4, H2O,CO, H2, and CO2. Air is continuously fed into the air channel as anoxidant; hence, the chemical species are O2 and N2. The configu-ration of the SOFC in this paper is a planar design. The mass balancein the fuel and air channels can be shown in terms of the compo-nent concentration and can be described as follows:
Fuel channel :dCi;fdx
¼ 1uf
Xk2fðiÞ;ðiiÞ;ðiiiÞg
ni;kRk1hf
(6)
Boundary condition for fuel channel : Ci;f���x¼0
¼ C0i;f (7)
D. Saebea et al. / Energy 94 (2016) 218e232 221
Air channel :dCi;adx
¼ 1ua
ni;ðiiiÞRðiiiÞ1ha
(8)
Boundary condition for air channel : Ci;a��x¼0 ¼ C0
i;a (9)
here, Ci;f and Ci;a are the concentrations of species i in the fuel andair channels, respectively; uf and ua are the inlet velocities of fueland air, respectively; and hf and ha are the heights of the fuel andair channels, respectively. The rates of all reactions within SOFC areconcluded in Table 1 [18,19].
The operating cell voltage of SOFC is expressed by Eq. (10):
V ¼ EOCV � h (10)
where EOCV is the open-circuit voltage that occurs due to the dif-ference between the thermodynamic potentials of the electrodereactions, and h is the voltage losses within the internal cell.
The open-circuit voltage can be expressed by the Nernstequation:
EOCV ¼ E0 � RT2F
ln
PH2O
PH2P0:5O2
!(11)
The dominant voltage losses are ohmic losses (hohm), concen-tration overpotentials (hconc), and activation overpotentials (hact).Activation overpotentials occur at the electrode-electrolyte in-terfaces because of the sluggishness of the electrochemical reactionand can be expressed by the ButlereVolmer equation in Eqs. (12)and (13). Ohmic losses result in resistance along the flow of ionsin the electrolyte and electrons through the electrode and currentcollectors and can be described by Ohm's law (Eq. (14)). Concen-tration overpotentials are caused by the resistance to mass trans-port due to the decrease in substance during the reaction at thesurface of the electrodes, expressed by Eqs. (15) and (16):
j ¼j0;anode
"pH2;TPB
pH2;f
exp�anFRT
hact;anode
�
� pH2O;TPB
pH2O;fexp
�� ð1� aÞnF
RThact;anode
�# (12)
j ¼j0;cathode
�exp
�anFRT
hact;cathode
�
� exp�� ð1� aÞnF
RThact;cathode
�� (13)
hohm ¼ jRohm (14)
hconc;anode ¼ RT2F
ln
pH2O;TPBpH2;f
pH2O;fpH2;TPB
!(15)
Table 1The reaction rates that occurred in the SOFC.
Steam reforming reaction R ið Þ ¼ kactpCH4 ;f exp�EaRTf
� �Water gas shift reaction
RðiiÞ ¼ kWGSRpCO;f
1� pCO2 ;f
pH2 ;f
kshiftpCO;fpH2O;f
!
Electrochemical reaction RðiiiÞ ¼ j2F
hconc;cathode ¼ RT4F
ln
pO2;a
pO2;TPB
!(16)
where Rohm is the internal electrical resistance, which is a functionof the conductivity and thickness of the individual layers; a is theapparent transfer coefficient (usually specified to be 0.5); n is thenumber of electrons exchanged in reaction step; and j0;anode andj0;cathode are the exchange current densities of the anode andcathode [20,21], as given by the following:
j0;anode ¼ RTnF
kanode exp�� Eanode
RT
�(17)
j0;cathode ¼ RTnF
kcathode exp�� Ecathode
RT
�(18)
where kanode and kcathode represent the pre-exponential factors ofthe exchange current densities of anode and cathode, and Eanodeand Ecathode represent the activation energies of the anode andcathode exchange current densities, respectively [20]. The values ofexchange current density data are presented in Table 2.
The partial pressures of H2, H2O and O2 at the three-phaseboundaries (pi;TPB) in Eqs. (15) and (16) can be determined by us-ing a gas transport model in porous media, as shown in Eqs.(19)e(21).
pH2;TPB ¼ pH2;f �RTtanode
2FDeff ;anodej (19)
pH2O;TPB ¼ pH2O;f þRTtanode
2FDeff ;anodej (20)
pO2;TPB ¼ P ��P � pO2;a
�exp
RTtcathode
4FDeff ;cathodePj
!(21)
where Deff ;anode is the effective gaseous diffusivity through theanode considering a binary gas mixture of H2 and H2O (equimolarcounter-current one dimensional diffusion of H2 and H2O andDeff ;cathode is the effective oxygen diffusivity through the cathodeconsidering a binary gas mixture of O2 and N2 (one-dimensionalself-diffusion).
In this study, the SOFC was assumed to operate under adiabaticconditions; therefore, the excess air flow to the SOFC is required tocontrol the cell operating temperature. The amount of excess inletair flow fed into the SOFC can be calculated from the energy balancearound the control volume enclosing the fuel cell, as follows:
Xinlet
_ni;an _hi;anþXinlet
_ni;ca _hi;ca
!� X
outlet
_ni;an _hi;anþXoutlet
_ni;ca _hi;ca
!
¼Wsofc
(22)
Table 2The exchange current density data for SOFC [20].
where _ni is the molar flow rate of species i, and _hi is the enthalpy ofspecies i.
The average current density can be calculated by
javg ¼ 1L
ZL0
j zð Þdz (23)
The overall fuel utilization factor is defined as follows:
Uf ¼javgLW
2F�4 _ninCH4
þ _ninH2þ _ninCO
� (24)
where _nini is the inlet molar flow rate of gas species i.The electrical power output is defined as follows:
Wsofc; dc ¼ javg � V � Ac (25)
where Ac is the cell active area.The direct current (DC) electricity produced from SOFC is
changed into alternating current in the inverter, as given by
Wsofc;ac ¼ Wsofc;dc � hinvert (26)
The excess air coefficient is the ratio of the amount of the oxygenin the inlet stream to the amount of oxygen needed for the stoi-chiometry of the electrochemical reaction, as described in Eq. (27):
lair ¼2 _nO2;inlet_nH2;inlet
(27)
The fuel utilization factor is the ratio of the amount of hydrogenused in the electrochemical reaction to the amount of hydrogen inthe inlet stream, as expressed in Eq. (28).
Uf ¼_nH2;utillized_nH2;inlet
(28)
where _nH2;utillized is the hydrogen molar flow consumed in theelectrochemical reaction, and _nH2;inlet is the inlet molar flow rate ofhydrogen fed into the SOFC. It should be noted that if the fuel isreformed internally at the anode catalyst, the hydrogen producedby the reforming reactions should be added to the term _nH2 ;inlet.
3.2. Fuel reforming
In an external reformer, ethanol is converted into syngasthrough ethanol steam reforming, which can be described by thefollowing reactions [22]:
C2H5OH þ H2O/4H2 þ 2CO (29)
CO þ H2O/H2þ CO2 (30)
CO þ 3H2/CH4þ H2O (31)
The ethanol steam reactions are considered to reach thermo-dynamic equilibrium. The reformer products are determined by astoichiometric approach. The equilibrium compositions depend onthe equilibrium constant, given by
Keq;SE ¼p4H2;rp
2CO;r
pC2H5OH;rpH2O;r(32)
Keq;WGS ¼ pH2;rpCO2;r
pCO;rpH2O;r(33)
Keq;MR ¼ pCH4;rpH2O;r
pCO;rp3H2;r
(34)
where Keq;j represents the equilibrium constants associated withthe reaction j, and pi;r represents the partial pressure of thecomponent i in the steam reformer.
The equilibrium constants of all reactions depend on the tem-perature at constant pressure and can be determined by
ln KðTÞ ¼ �DGT ; p0
RT
(35)
The energy balance of the ethanol steam reformer can bedefined as follows:
Qr ¼Xout
_ni;r _hi;r �Xin
_ni;r _hi;r (36)
The steam-to-carbon ratio at the inlet of the reformer, which isthe important parameter for the considering the anode recircula-tion ratio, can be determined as follows:
SCR ¼_ninH2O;r
_ninC2H5OH;r þ _ninCH4;r þ _ninCO;r(37)
3.3. Compressor and gas turbine
In a SOFC-GT hybrid system, the compressor is used to enhancethe air pressure before entering the SOFC, while the turbine pro-duces further electricity for the system from the high-qualityexhaust gas downstream of the SOFC component. The powerconsumed in the compressor (Wc) and the power generated in theturbine (Wt) are calculated as follows:
Wt ¼Xinlet
_ni;t _hi;t �Xoutlet
_ni;t _hi;t (38)
Wc ¼Xinlet
_ni;c _hi;c �Xoutlet
_ni;c _hi;c (39)
The compressor and turbine outlet temperatures can be deter-mined in terms of isentropic efficiency, as follows:
Toutt ¼ T int
0@1� ht
0@1�
poutt
pint
!g�1g
1A1A (40)
Toutc ¼ T inc
0@1þ 1
hc
0@�poutc
pinc
�g�1g
� 1
1A1A (41)
The total power of the gas turbine component can be expressedas follows:
Wgt ¼ ðWt �WcÞhmhg (42)
where hm and hg denote the mechanical and generator efficiencies,respectively.
Table 3Model of the auxiliary units in the SOFC-GT hybrid systems.
A SOFC-GT hybrid system consists of a pump, vaporizer, heatexchanger, mixer, and afterburner. The equations for the expressionof other units are listed in Table 3. The power work consumed bywater and ethanol pumps can be expressed by the Bernoulliequation in Eq. (43). The thermal energies for the ethanol andwatervaporizations were calculated from Eqs. (44) and (45).
A mixer was used to mix the ethanol and water. Additionally,mixers were used to mix the recycling-anode exhaust gas and freshfuel in the systems with anode gas recirculation or the recycling-cathode exhaust gas and fresh air in the systems with cathodegas recirculation. The outlet molar flow rate and temperature fromthe mixer were considered based on the mass and energy balancesin Eqs. (46) and (47).
The exhaust gases from the anode and cathode sides were fed tothe afterburner to burn the residual fuel. The combustion reactionsoccur within the afterburner according to the following reactions:
Hydrogen combustion : H2 þ 0:5O2/H2O
Carbon monoxide combustion : CO þ 0:5O2/CO2
A combustion efficiency of 98%was fixed in this work. Themolarflow rate and temperature of the exit gas from afterburner werefound using Eqs. (48) and (49).
The temperature of the air can be enhanced before feeding tothe SOFC by heat exchange with the exhaust gas from the turbineand afterburner through the recuperator and high heat exchanger.The temperatures of the cool and hot gases leaving the heatexchanger can be calculated from the heat exchanger effectivenessand the energy balance, as shown in Eqs. (50) and (51).
The desired inlet temperature of the turbine can bemanipulatedby the addition of fuel or air to the combustor. Two parameters arespecified for the consideration of the added fuel or air to adjust theturbine temperature, as follows:
FRC ¼_mf ;C_mf ;i
(52)
ARC ¼ _ma;C_ma;i
(53)
3.5. Performance criteria
The performance parameters of the system are considered fromthe SOFC electrical efficiency, given as follows:
hel;sofc ¼Wsofc; ac
_nini;anLHVi;an
(54)
where _ni is the SOFC inlet molar flow rate, and LHVi is the lowerheating value of methane, hydrogen, and carbon monoxide,respectively.
Thermal efficiency is considered based on the heat output of theexhaust gas taken out of the system ( _Qheat; sink) and the total fuelenergy input. The heat output depends on the temperature of theexhaust gas, which is set at 100 �C.
hth ¼_Qheat; sink
_nC2H5OHLHVC2H5OH þ _Quse(55)
where _nC2H5OH is the fuel molar flow rate at inlet, LHVC2H5OH is thelower heating value of ethanol and _Quse is the external thermalenergy added that is the total of thermal energy requiring forsupplying heat in the system subtracted by the thermal energy ofrecuperator downstream. The total system power is calculated fromthe sum of the SOFC and GT power which is subtracted by thepower used in the auxiliary unit of the hybrid system, as shown inEq. (56).
Fig. 2. The solution algorithm of SOFC-GT hybrid systems with anode and cathode exhaust gas recirculation.
D. Saebea et al. / Energy 94 (2016) 218e232224
Table 4Structural and material property parameters for the SOFC.
Parameter Value
Cell width, L (m) 0.4Cell width, W (m) 0.1hf (mm) 1ha (mm) 1tanode (mm) 500tcathode (mm) 50telectrolyte (mm) 20Electronic conductivity of the anode, sanode (U�1 m�1) 9:5�107
T exp��1150
T
�Electronic conductivity of cathode, scathode (U�1 m�1) 4:2�107
T exp��1200
T
�Ionic conductivity of the electrolyte, selectrolyte (U�1 m�1)
33:4� 103 exp��10300
T
�Effective gaseous diffusivity through the anode, Deff ;anode (m2 s�1) 3.66 � 10�5
Effective oxygen diffusivity through the cathode, Deff ;cathode (m2 s�1) 1.37 � 10�5
D. Saebea et al. / Energy 94 (2016) 218e232 225
Wsystem ¼�Wsofc;ac þWGT
�� WH2O;p þWC2H5OH;p þWblower
(56)
The specific work is given by:
WSp ¼ Wsystem_mair
(57)
where _mair is the inlet molar flow rate of air.The system efficiency is defined as follows:
hsys ¼Wsystem
_nC2H5OHLHVC2H5OH þ _Quse(58)
4. Solution approach
To evaluate the performance of the SOFC-GT hybrid system,the component models within the hybrid system in Section 3 are
Table 5Operating parameters used for the simulation of the SOFC/GT hybrid system undernominal conditions [20,22,26].
Parameter Value
Reformer temperature, Tr (K) 973Cell temperature, Tsofc (K) 1073Fuel temperature at system inlet (K) 298.15Air temperature at system inlet (K) 298.15Compressor pressure ratio 3Air temperature increase across cathode (K) 100Average current density (A/cm2) 0.4Cell number 3048Fuel utilization factor 0.7Steam-to-carbon ratio 1.5Air composition 21% O2, 79% N2
Fuel composition 100% C2H5OHdc-ac inverter efficiency 94Turbine isentropic efficiency (%) 82Compressor isentropic efficiency (%) 78Generator mechanical efficiency (%) 94Afterburner combustion efficiency (%) 98Recuperator effectiveness (%) 90Blower isentropic efficiency (%) 70Pump efficiency (%) 75SOFC pressure loss (%) 2Afterburner pressure loss (%) 3Recuperator pressure loss of cool stream (%) 1.5Recuperator pressure loss of hot stream (%) 2.5
simulated in Matlab 7.0. The validation of the SOFC model waspresented in our previous publication [16]. The calculationsequence for SOFC-GT hybrid systems with the recirculation ofanode exhaust gas or cathode exhaust gas is shown in Fig. 2. Inthe calculation procedure for the exhaust gas recirculation in thesystem, all mole fractions of composition i in the recycled streamare unknown variables that are initially assumed to be includewith the fresh stream as input data for the SOFC model. Thevalues from the calculation of the mole fraction of composition iin the SOFC exhaust gas are used to find the new assumed valuesof mole fractions of composition i in the recycled stream in thenext loop of the calculation. This iterative procedure provides anaccurate value of the unknown variables until the difference inthe values of the assumed and calculated variables is less than10�4.
The values of structural and material property parameters forthe SOFC are shown in Table 4 [20,21]. In the simulation of allsystems, the average current density of SOFC is given as 0.4 A/cm2.The recuperative effectiveness in all SOFC-GT hybrid systems isgiven the same value of 90%. To avoid thermal stress of the fragileceramic material within the cell due to the extreme temperaturegradient, a temperature difference across the cell of 100 K is theoperating constraint for considering the amount of air needed forthe SOFC. The carbon formation in the fuel process depends on theprocess conditions. To prevent carbon deposits in an ethanol steamreformer, excess steam is required. Thus, a steam-to-carbon ratio of1.5 is used in the investigation. The operating parameters and
40
45
50
55
60
65
70
0.55 0.6 0.65 0.7 0.75 0.8
Syst
em e
ffic
ienc
y (%
)
Fuel utilization factor (Uf)
AR systemCR systemACR system
Fig. 3. SOFC electrical efficiency of SOFC-GT hybrid systems as a function of fuel uti-lization factor.
D. Saebea et al. / Energy 94 (2016) 218e232226
design constraints for the performance analysis of the SOFC-GThybrid systems are given in Table 5 [20,22e26].
The inlet molar flow rate of fuel is adjusted to satisfy the currentdensity and fuel utilization factor. When considering SOFC-GThybrid systems with cathode-off gas recycling, the recirculationratio of the cathode exhaust gas depends on the amount of airrequired to control the cell temperature. Meanwhile, the recircu-lation ratio of the anode exhaust gas in the SOFC-GT hybrid systemwith anode-off gas recycling is adjusted to provide the given steam-to-carbon ratio for the steam reformer.
50
100
150
200
250
300
350
400
450
spec
ific
wor
k (k
J/kg
air
)
AR systemCR systemACR system
5. Results and discussion
The important operating parameters of the SOFC-GT hybridsystem that abundantly impact the matching of SOFC and gas tur-bine units are the fuel utilization factor, the cell temperature, andthe turbine inlet temperature. For insight into the behaviors ofthree configurations of the SOFC-GT hybrid system, the direct ef-fects of fuel utilization factor and temperature of SOFC on the tur-bine performance were first investigated. Consequently, the effectof adding fuel or air to the combustor to adjust the turbine inlettemperature on the system efficiency was considered.
(a)
00.55 0.6 0.65 0.7 0.75 0.8
Fuel utilization factor (Uf)
5.1. Effect of fuel utilization factor
In this part, the turbine inlet temperature directly depends onthe operation of the SOFC. Fig. 3 illustrates the electrical efficiencies
Fig. 4. (a) GT to SOFC power ratio and (b) turbine inlet temperature of SOFC-GT hybridsystems as a function of fuel utilization factor.
of the SOFC component as a function the fuel utilization factor inAR, CR, and ACR hybrid systems, respectively. As seen, the SOFCelectrical efficiencies enhance with increasing fuel utilization fac-tor. This result can be explained in that the fuel transformed toelectricity via the electrochemical reaction within the SOFC in-creases. The SOFC electrical efficiency of the CR system is thehighest, exceeding the electrical efficiency of the AR and ACR sys-tems by approximately 2.4e3%, respectively. It can be noted thatthe recirculation of the anode-off gas to the reformer dilutes the
(b)
10
15
20
25
30
35
0.55 0.6 0.65 0.7 0.75 0.8
The
rmal
eff
icie
ncy
(%)
Fuel utilization factor (Uf)
AR systemCR systemACR system
(c)
40
45
50
55
60
65
70
0.55 0.6 0.65 0.7 0.75 0.8
Syst
em e
ffic
ienc
y (%
)
Fuel utilization factor (Uf)
AR systemCR systemACR system
Fig. 5. (a) Specific work (b) thermal efficiency and (c) system electrical efficiency ofSOFC-GT hybrid systems as a function of fuel utilization factor.
D. Saebea et al. / Energy 94 (2016) 218e232 227
hydrogen and carbon monoxide concentrations in the fuel inlet ofthe SOFC, due to the accumulation of carbon dioxide within thesystem, which leads to increased anodic concentration over-potentials within the cell. When considering the AR and ACR sys-tems, the electrical efficiencies of the SOFC in both systems areinsignificant. The recirculation of cathode exhaust gas has a minorimpact on the reduction of SOFC electrical efficiency because theoperation of SOFC requires excess air to the cathode side. Thus, thereduced oxygen concentration has an insignificant effect on theSOFC electrical efficiency.
Fig. 4a represents the effect of fuel utilization factor on theturbine performance considered in terms of the GT to SOFC po-wer ratio. The results show that the influence of fuel utilizationfactor on the GT to SOFC power ratio of all SOFC-GT hybrid sys-tems has the same tendency. The GT to SOFC power ratios arenegatively correlated with the fuel utilization factor. Higher fuelutilization factor results in a decline in the amount of fuel to thecombustor, and consequently the turbine inlet temperature de-creases, as seen in Fig. 4b. Moreover, it can be observed that theGT to SOFC power ratio of the AR system is the lowest. Therecirculation of partial anode exhaust gas diminishes the fuel-to-air ratio to the combustor. The ACR system has a higher GT toSOFC power ratio than the AR system. The recirculation ofcathode-off gas back into the SOFC increases the fuel-to-air ratiofed to the combustor; thus, the combustor outlet temperatureincreases, in turn increasing the turbine inlet temperature andthe GT to SOFC power ratio.
Fig. 5 presents the influence of fuel utilization factor on thespecific work, thermal efficiency, and system efficiency of threeSOFC-GT hybrid systems. From Fig. 5a, more excess air is requiredfor cooling the cell due to the enhanced exothermic heat of theelectrochemical reaction at higher fuel utilization factor; therefore,the specific work of the AR system decreases. However, the specificwork of the CR system increases. In SOFC-GT hybrid systems withcathode exhaust gas recycling, the recirculation ratio of cathodeexhaust gas is increased to control the cell temperature when theSOFC operates at higher fuel utilization factor. The fresh air feedrequired in this system decreases. Thus, the specific work of theACR system is slightly enhanced with increasing fuel utilizationfactor.
In Fig. 5b, it can be observed that the thermal efficiency of theCR system distinctly decreases with increasing fuel utilizationfactor, which is because the reduced air flow rate to thecombustor in this system causes lower thermal efficiency of theCR system at high fuel utilization factor. However, the CR systemshows the highest thermal efficiency. For Fig. 5c, the system ef-ficiency of the CR and ACR systems would increase as the fuelutilization factor increases. The increase in SOFC power has moreimpact on the system efficiency than the decrease in turbineperformance. However, the system efficiency of the AR system
Table 6System performance of each system configuration at different cell temperatures.
Systems AR
1073 K 1173 K
SOFC efficiency (%) 41.4 44.1TIT (K) 794.2 845.5Recirculation ratio of anode exhaust gas 0.605 0.608
Recirculation ratio of cathode exhaust gas e e
GT to SOFC power ratio 0.10 0.12Specific work (kJ/kg air) 210.4 266.9Thermal efficiency (%) 23.4 21.4System efficiency (%) 50.9 53.1
exhibits inverse characteristics compared to the CR and ACRsystems.
Comparing the performance of all systems, the CR system ex-hibits the lowest system efficiency because it requires external heatto produce steam. Although the AR system reduces the externalheat, the GT performance in this hybrid system considerably de-creases. The recirculation of both anode and cathode exhaust gas inthe SOFC-GT hybrid system achieves the highest system efficiency,approximately 60% in the fuel utilization factor range studied.Additionally, the ACR system can enhance the thermal efficiencymore than the AR system by approximately 2e1%.
5.2. Effect of cell temperature
This section presents the effects of SOFC operating temperaturesof 1073 and 1173 K on the performance of each system configura-tion, as shown in Table 6. Increasing the SOFC operating tempera-ture from 1073 K to 1173 K can improve the SOFC electricalefficiency by approximately 2.7e2.9% due to the lower ohmicoverpotentials as a result of the higher conductivity of oxygen ionsthough the electrolyte. Moreover, the activation overpotentialswithin the cell are reduced when the SOFC operates at highertemperature.
Regarding the effect of the cell temperature on the downstreampart of the SOFC, the turbine inlet temperature of all systemswouldbe raised as the SOFC operating temperature increases due to thereduced excess air ratio for the system. The inlet air flow raterequired for cooling from the exothermic heat in the cell is reducedwith increased SOFC operating temperature. Among the threesystem configurations, the increase in turbine inlet temperaturewhen raising the SOFC operating temperature from 1073 to 1173 Kin the CR and ACR systems is approximately 126 K and 115 K,respectively. In the hybrid systems with cathode exhaust gasrecycling, the recirculation ratio of cathode exhaust gas is elevatedat a higher cell temperature. Due to the increased cell temperature,the air temperature at the inlet of the SOFC needs to be increased tocontrol the difference temperature across the cell of 100 K. Thus,the recirculation ratio of the cathode exhaust gas is increased from0.128 to 0.136 in order to maintain the required inlet air tempera-ture of the SOFC.
Although the turbine inlet temperature increases, the GT toSOFC power ratio in the CR and ACR systems decreases slightly atthe higher cell temperature. Meanwhile, the tendency of the GT toSOFC power ratio in the AR system is reversed from the CR and ACRsystems. It is noted that the decrease in the molar flow rate of thegas stream into the turbine due to the increased recirculation ratioof cathode gas in the CR and ACR systems results in a lower poweroutput of the turbine.
When considering the effect of the cell temperature on the ef-ficiency, upon increasing the cell temperature from 1073 to 1173 K,
Fig. 6. Effect of turbine inlet temperature on (a) additional fuel ratio (b) GT to SOFC power ratio (c) specific work (d) thermal efficiency (e) system efficiency and (f) high-temperature heat exchanger effectiveness in the AR system.
D. Saebea et al. / Energy 94 (2016) 218e232228
the thermal efficiency of the AR and ACR systems diminishes byapproximately 1.23 and 0.3%, respectively. In the systemwith anodeexhaust gas recycling, the energy of the exhaust gas from therecuperator is reduced because the anode exhaust gas is increas-ingly recycled at higher cell temperature. Additionally, the requiredexcess air flow rate is lower when the SOFC operates at high tem-perature. In contrast, the thermal efficiency of the CR system in-creases at the higher cell temperature because the influence of the
increased temperature of the gas stream at the outlet of the recu-perator in the CR system is greater than the reduction of the air flowrate.
The increase in system efficiency of the ACR system as theoperating temperature increases is highest. The system efficiency ofthe ACR system is higher than for the AR system (8.15% and 6.93%)as well as the CR system (10.18% and 10.41%) at SOFC operatingtemperatures of 1073 and 1173, respectively.
D. Saebea et al. / Energy 94 (2016) 218e232 229
5.3. Effect of the turbine inlet temperature
For the SOFC-GT based auxiliary power unit in the range of400e500 kW, a micro gas turbine generator is employed. Thecrucial problem of the micro gas turbine is the blade cooling for itssystem due to the limited ability of the materials that make up theengine to endure high temperature and stress. Therefore, themaximum turbine inlet temperature is a constraint for the
Fig. 7. Effect of turbine inlet temperature on (a) additional air ratio (b) GT to SOFC power r
operation of the SOFC-GT hybrid system. A micro gas turbineshould be operated in a temperature range between 873 and1223 K. For the SOFC-GT hybrid systems, the turbine inlet tem-perature depends directly on the SOFC operation. To operate at thenecessary turbine inlet temperature, additional fuel and air supplyratios to the combustor canmanipulate the heat downstream of theSOFC. The effects of additional fuel and air supply to the combustorto adjust the turbine inlet temperature on the performance of the
atio (c) specific work (d) thermal efficiency and (e) system efficiency in the CR system.
D. Saebea et al. / Energy 94 (2016) 218e232230
three SOFC-GT hybrid systems are considered in this part. Theadditional fuel and air supply ratios are defined by Eqs. (52) and(53), respectively.
The effects of the assigned turbine inlet temperature on thedesign parameters and performance of the AR system are presentedin Fig. 6. This system considers only the additional fuel supply ratioto the combustor. It is not necessary to add air to the combustorbecause the turbine inlet temperature of the AR systemwithout theadditional fuel and air to the combustor is low, approximately800 K. Fig. 6a shows the additional fuel supply ratio to the
Fig. 8. Effect of turbine inlet temperature on (a) additional air ratio (b) additional fuel raefficiency in the ACR system.
combustor in the AR system. For a specified cell temperature, thefuel supply ratio to the combustor increases to achieve a higherturbine inlet temperature. As the cell temperature increases, therequirement to add fuel to the combustor is lower. By comparisonwith the AR system without the additional fuel to the combustor,the addition of fuel to the combustor has a positive effect on the GTperformance, as shown in Fig. 6b. The GT to SOFC power ratio isimproved if the turbine inlet temperature is increased. Theincreased GT power enhances the specific work, as presented inFig. 6c. Although the AR system requires the additional fuel supply
tio (c) GT to SOFC power ratio (d) specific work (e) thermal efficiency and (f) system
D. Saebea et al. / Energy 94 (2016) 218e232 231
ratio to increase the turbine inlet temperature, the thermal andsystem efficiency increase as the turbine inlet temperature in-creases, as shown in Fig. 6d and e, respectively. The increased tur-bine inlet temperature of 1223 K could enhance the thermalefficiency and system efficiency from the AR system without theadditional fuel and air to the combustor by approximately 5% and3%, respectively. Additionally, the addition of fuel supply to thecombustor influences the high-temperature heat exchanger effec-tiveness, as shown in Fig. 6f. The required high-temperature heatexchanger effectiveness decreases with increasing turbine inlettemperature.
For the CR system, the turbine inlet temperature in the casewithout the additional fuel and air to the combustor is high. The airsupply to the combustor must be increased to reduce the turbineinlet temperature to the specified range. Fig. 7a shows the additionalair supply ratio to the combustor in the CR system. As the turbineinlet temperature increases, the additional air supply ratio requiredfor the combustor is lower. When the SOFC operates at a highertemperature, a greater additional air supply to the combustor isneeded. However, at a lower cell temperature of 1073 K, less fuel isadded to the combustor than for the high turbine inlet temperatureof 1223 K. The effect of the variation of the turbine inlet temperatureon the GT to SOFC power ratio is presented in Fig. 7b. The GT to SOFCpower ratio subsideswithhigher turbine inlet temperature. Becauseconsiderable air is added to the combustor at the low turbine inlettemperature, the power output of the turbine is increased due to theenhancedmolarflowrate of gas at the inlet of the turbine. Therefore,the increase in turbine power at the low turbine inlet temperature ishigher than at the higher turbine inlet temperature.
Fig. 7c, d, and e show the specific work, thermal efficiency, andsystem efficiency at various turbine inlet temperatures in the CRsystem. Although the increase in the turbine inlet temperaturedecreases the turbine performance, the specific work and systemefficiency are increased due to the lower work by the compressorfor the addition of air as the turbine inlet temperature increases.However, the addition of air to the combustor leads to higher heatenergy of the gas outlet from the system. Thus, the thermal effi-ciency is reversed from the specific work and system efficiency atthe higher turbine inlet temperature. Compared to the case of theCR system without additional fuel and air to the combustor, thesystem efficiency and specific work are lower, while the thermalefficiency is higher, as can clearly be observed at the lower turbineinlet temperature.
The additional fuel and air supply ratios to the combustor atvarious turbine inlet temperatures in the ACR system are shown inFig. 8a and b. The amount of additional air supply to the combustorin the ACR system needed to reduce the inlet turbine temperature islower than in the CR system. Meanwhile, the ACR system at theSOFC operating temperature of 1173 K requires additional fuelsupply ratios of 0.05 and 0.13 to maintain the turbine inlet tem-peratures of 1173 and 1223 K, respectively. The additions of fuel andair supply are related to the GT performance.
Fig. 8c, d, e, and f show the GT to SOFC power ratio, specificwork, thermal efficiency, and system efficiency, respectively, atvarious turbine inlet temperatures. The influences of various tur-bine inlet temperatures on these parameters are the same as for theCR system. However, at the SOFC operating temperature of 1073 K,the influences of the turbine inlet temperature in the range of1123e1223 K on the GT to SOFC power ratio and system efficiencyare different from the CR system due to the fuel added to thecombustor. The increased fuel supply ratio to the combustor in ACRincreases the GT to SOFC ratio, whereas the system efficiency be-comes lower. Comparing the performance of all systems, the ACRsystem exhibits the highest system efficiency at the studied turbineinlet temperature.
6. Conclusions
The performance analysis of the SOFC-GT hybrid system com-bined with an external steam reforming of ethanol with anode or/and cathode exhaust gas recycling was investigated in this work.The fuel utilization factor and temperature of SOFC on the systemperformance of the SOFC-GT hybrid systems without the control ofturbine inlet temperature were first studied. The results indicatethat the system efficiency of AR is higher than that of CR whereasthe thermal efficiency of AR is lowest in the all range of fuel utili-zation factor and cell temperature. When comparing three config-urations, the ACR system achieves the highest system andcogeneration efficiency for the case in the systemwithout assigningturbine inlet temperature. The SOFC operating parameters has adirect effect on turbine inlet temperature which is higher withincreasing the cell temperature and decreasing the fuel utilizationfactor. The turbine inlet temperature of AR system is lower whilethat of CR system is higher than the suitable small turbine inlettemperature. The AR system requires the additional fuel whereasthe CR system must add the high amount of air to the combustor.The ACR system requires the additional air ratio lower than CRsystem. At high turbine inlet temperature, the ACR system shouldadd less fuel when SOFC operates at low cell temperature. Althoughthe system efficiency of ACR system reduces due to the addition offuel, the ACR system is highest system efficiency and does not needthe high-temperature exchanger.
Acknowledgments
The authors would like to thank the Thailand Research Fund,Burapha University, and the Ratchadaphiseksomphot EndowmentFund of Chulalongkorn University for financial supports.
References
[1] Buonomano A, Calise F, d'Accadia MD, Palombo A, Vicidomini M. Hybrid solidoxide fuel cells-gas turbine systems for combined heat and power: a review.Appl Energy 2015;156:32e85.
[2] Hou T, Yu B, Zhang S, Xu T, Wang D, Cai W. Hydrogen production from ethanolsteam reforming over Rh/CeO2 catalyst. Catal Commun 2015;58:137e40.
[3] Nichele V, Signoretto M, Menegazzo F, Rossetti I, Cruciani G. Hydrogen pro-duction by ethanol steam reforming: effect of the synthesis parameters on theactivity of Ni/TiO2 catalysts. Int J Hydrogen Energy 2014;39:4252e8.
[4] Zabihian F, Fung AS. Performance analysis of hybrid solid oxide fuel cell andgas turbine cycle: application of alternative fuels. Energy Convers Manag2013;76:571e80.
[5] Barelli L, Bidini G, Ottaviano A. Part load operation of a SOFC/GT hybrid sys-tem: dynamic analysis. Appl Energy 2013;173e89.
[6] Laosiripojana N, Assabumrungrat S. Catalytic steam reforming of methane,methanol, and ethanol over Ni/YSZ: the possible use of these. J Power Sources2007;163:943e51.
[7] Liese EA, Gemmen RS. Performance comparison of internal reforming againstexternal refroming in a solid oxide fuel cell, gas turbine hybrid system. ASME2005;127:86e90.
[8] Gandiglio M, Lanzini A, Leone P, Santarelli M, Borchiellini R. Thermoeconomicanalysis of large solid oxide fuel cell plants: atmospheric vs. pressurizedperformance. Energy 2013;55:142e55.
[9] Chinda P, Brault P. The hybrid solid oxide fuel cell (SOFC) and gas turbine (GT)systems steady state modeling. Int J Hydrogen Energy 2012;37:9237e48.
[10] Bakalis DP, Stamatis AG. Optimization methodology of turbomachines forhybrid SOFC-GT applications. Energy 2014;70:86e94.
[11] Song TW, Sohn LS, Kim TS, Ro ST. Performance characteristics of a MW-classSOFC/GT hybrid system based on a commercially available gas turbine.J Power Sources 2006;158:361e7.
[12] Bakalis DP, Stamatis AG. Incorporating available micro gas turbines and fuelcell: matching considerations and performance evaluation. Appl Energy2013;103:607e17.
[13] Barelli L, Bidini G, Ottaviano A. Part load operation of SOFC/GT hybrid systems:stationary analysis. Int J Hydrogen Energy 2012;37:16140e50.
[14] Saebea D, Patcharavochot Y, Arpornwichanop A. Analysis of an ethanol-fuelled solid oxide fuel cell system using partial anode exhaust gas recircu-lation. J Power Sources 2012;208:120e30.
[15] Trasino F, Bozzolo M, Magistri L, Massardo AF. Modeling and performanceanalysis of the Rolls-Royce fuel cell systems limited: 1 MW plant. J Eng GasTurbines Power 2011;133(2):021701e12.
[16] Saebea D, Patcharavochot Y, Assabumrungrat S, Arpornwichanop A. Analysisof a pressurized solid oxide fuel cell-gas turbine hybrid power system withcathode gas recirculation. Int J Hydrogen Energy 2013;38:4748e59.
[17] Jia J, Li Q, Luo M, Wei L, Abudul A. Effects of gas recycle on performance ofsolid oxide fuel cell power systems. Energy 2011;36:1068e75.
[18] Achenbach E, Riensche E. Methane/steam reforming kinetics for solid oxidefuel cells. J Power Sources 1994;52:283e8.
[19] Belyaev VD, Politova TI, Marina OA, Sobyanin VA. Internal steam reforming ofmethane over Ni-based electrode in solid oxide fuel cells. Appl Catal A General1995;133:47e57.
[20] Aguiar P, Adjiman CS, Brandon NP. Anode-supported intermediate-temperature direct internal reforming solid oxide fuel cell. I: model-basedsteady-state performance. J Power Sources 2004;138:120e36.
[21] Patcharavorachot Y, Arpornwichanop A, Chuachuensuk A. Electrochemicalstudy of a planar solid oxide fuel cell: role of support structures. J PowerSources 2008;177:254e61.
[22] Freni S, Maggio G, Cavallaro S. Ethanol steam reforming in a molten carbonatefuel cell: a thermodynamic approach. J Power Sources 1996;62:67e73.
[23] Santin M, Traverso A, Magistri L, Massardo A. Liquid fuel utilization in SOFChybrid system. Appl Energy 2009;86:2204e12.
[24] Santin M, Traverso A, Magistri L, Massardo A. Thermoeconimic analysis ofSOFC-GT hybrid systems fed by liquid fuels. Energy 2010;35:1077e83.
[25] Motahar S, Alemrajabi AA. Exergy based performance analysis of a solid oxidefuel cell and steam injected gas turbine hybrid power system. Int J HydrogenEnergy 2009;34:2396e407.
[26] Akkaya AV, Sahin B, Erdem HH. An analysis of SOFC/GT CHP system based onexergetic performance criteria. Int J Hydrogen Energy 2008;33:2566e77.
