Performance assessment of concentrated solar power

i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y x x x ( 2 0 1 8 ) 1 e1 1

Available online at www.sciencedirect.com

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

Performance assessment of concentrated solar power plants based on carbon and hydrogen fuel cells Elena Dıáz a,b, Michael Epstein a, Manuel Romero a, lez-Aguilar a,* Jose Gonza a b

IMDEA Energy Institute, Avda. Ramon de La Sagra 3, 28935, Mostoles, Spain Department of Chemical and Energy Technology, ESCET, Rey Juan Carlos University, 28933 Mostoles, Spain

article info

abstract

Article history:

In spite of the recent success on the implementation of Concentrating Solar Power (CSP),

Received 22 August 2017

still this technology needs a substantial enhancement to achieve competitiveness. This

Received in revised form

paper provides thorough insight after previous analyses on an alternative concept for

22 January 2018

higher efficiency CSP systems based on the replacement of the power block by an elec-

Accepted 29 January 2018

trochemical conversion system. Concentrating solar energy is herewith used to decompose

Available online xxx

methane into hydrogen and carbon, which are used in hydrogen and carbon fuel cells for electricity generation. This approach envisages modular, efficient and flexible generation

Keywords:

plants. Dispatchability can be achieved by storing the solid carbon. Solar-to-electricity

Solar thermal electricity

efficiency was calculated assuming thermodynamic equilibrium composition and experi-

Methane decomposition

mental data available from literature, and compared with those of conventional power

Direct carbon fuel cells

generation systems and commercial CSP plants. It is concluded that this new-generation

Hydrogen fuel cells

CSP concept is potentially able to produce power more efficiently than the current stateof-the art solar thermal power plants. © 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.

Introduction Solar thermal electricity (STE) cost is currently not competitive with generation technologies based on fuels combustion or other renewables like photovoltaics or wind. Conventional strategy to accelerate cost reduction mainly lies in increasing concentrating solar power (CSP) performance through new developments on key components (such as heliostats,

receivers), heat transfer fluids and thermal storage media, and innovative thermodynamic cycles [1e3]. A long-term research approach focuses on the replacement of turbomachinery in the power block by an electrochemical system. This alternative leads to shorter response times, almost-constant partload efficiency and better grid integration. All of them are features that potentially allow increasing solar-to-electricity efficiency. Additionally water consumption is notably reduced compared to water/steam thermodynamic cycles.

Abbreviations: AFC, Alkaline fuel cell; CSP, Concentrated solar power; DCFC, Direct carbon fuel cell; FC, Fuel cell; HFC, Hydrogen fuel cell; MCFC, Molten carbonates fuel cell; PAFC, Phosphoric acid fuel cell; PEMFC, Proton exchange membrane fuel cell; PSA, Pressure swing adsorption; SOFC, Solid oxide fuel cell; SOFCIR, Solid oxide fuel cell with internal reforming; MSR, Methane steam reforming. * Corresponding author. lez-Aguilar). E-mail address: [email protected] (J. Gonza https://doi.org/10.1016/j.ijhydene.2018.01.190 0360-3199/© 2018 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved. Please cite this article in press as: Dıáz E, et al., Performance assessment of concentrated solar power plants based on carbon and hydrogen fuel cells, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/j.ijhydene.2018.01.190

2


Nomenclature Cs f :u: DHFC I LHVCH4 m_ CH4 PFC Plosses Q FC QCH4 Qpreheating Qreactor Qsolar ' Qsolar

TFC TR Xi

Concentration ratio (suns) Fuel utilization (%) Change in enthalpy in the fuel cell (kW) Normal irradiance (kW/m2) Methane low heating value (kJ/g) Methane mass flow rate (g/s) Power obtained in the fuel cell (kW) Power consumed by process losses (kW) Heat produced in the fuel cells (kW) Power supplied by the methane feedstock (kW) Heat needed for methane preheating (kW) Solar power absorbed by the reactor/receiver (kW) Concentrated solar power when preheating is not implemented (kW) Concentrated solar power when preheating is implemented (kW) Fuel cell operation temperature (ºC) Reactor/receiver operation temperature (ºC) Compound i conversion

Greek letters Reactor/receiver absorption efficiency (%) habs hCH4 Global process efficiency in terms of consumed feedstock (%) hCH4 solar Global process efficiency in terms of total energy consumed (%) Fuel cell efficiency (%) hFC s Stefan Boltzmann constant (5.67$108 W/m2K4)

Within this general concept, it has been proposed using concentrated solar energy for methane cracking into hydrogen and solid carbon, which subsequently feed a hydrogen and a direct carbon fuel cell (HFC and DCFC) [4e6]. The intrinsic features of fuel cells (FCs) mentioned above and the capability of energy dispatch by means of carbon storage allow for covering properly instantaneous hydrogen/heat/ electricity demand. Additionally energy conversion of methane-containing feedstocks are able to be used, including renewable ones such as landfill gas and biogas, after methane separation [7,8]. Finally the DCFC delivers lower and purer CO2 emissions than conventional thermal power systems based on combustion of natural gas. Potentially this new approach allows for more efficient and modular generation plants. Thermal or thermo-catalytic methane decomposition is an alternative to existing H2 production processes from CH4, like methane steam reforming (MSR). Although they seem analogous, usual MSR does not fully exploit the overall carbon content because methane is partially oxidized. Emissions are 3e5 times higher, and the required energy is higher by up to 1.7 times [9]. It has been claimed that MSR has higher environmental impact [10]. Thermodynamics shows that CH4 decomposition starts at 300 C, but temperatures in excess of 1200 C are required to obtain a reasonable decomposition rate and yield due to the strong CeH bonds and the lack of polarity

[11]. Reaction trace-products include acetylene (C2H2), ethylene (C2H4), butylene (C4H8), propylene (C3H6) and ethane (C2H6) [5]. Both, temperatures and by-products, can be significantly reduced using a catalyst, which can be carbonaceous [12], metallic [13], metal oxide based [14] or silica based [15]. Different kinds of solid carbon (carbon black, nanotubes, nanofibers, flakes, films) can be produced depending on the operation conditions, the type of reactor and the catalyst. It is possible to use solar energy for the decomposition [16]; however methane cannot be directly heated by solar irradiation because hydrocarbons poorly absorb in the solar spectrum. For that reason, solar reactors are usually based on opaque walls that absorb the solar radiation and heat up the gas (indirect heating) or a transparent window that permit direct heating of particulate material dispersed in the CH4 gas that absorbs the radiation (direct heating) [17]. This concept has been widely studied mainly by the research groups of Steinfeld at ETH-Zurich in Switzerland [18], Abanades, Rodat and Flamant at CNRS-PROMES in France [19] and Kogan, Yeheskel and Epstein of Weizmann Institute of Science in Israel [20]. In the University of Colorado, USA, by Dahl [21] and by Pinilla in CSIC, Spain, experiments have been performed as well [22]. Additionally, FCs are gaining more attention because it is based on a clean process with high efficiency and costeffective potential [23]. Use of fuel cells in stationary applications has been technically proven in a wide range of scales [24,25], although their commercialization has been hastening in the last few years based primarily on their cost [26]. Canada, Japan, South Korea, and Europe are currently developing large-scale fuel cell systems, being South Korea the country where the larger FC power plants can be found, i.e. West Incheon Power Plant (16 MWe), Noeul Green Energy Co., in Seoul (20 MW), Busan Green Energy project (30.8 MW), Gyeonggy Green Energy Park in Hwasung (59 MWe) and Pyeongteak City power plant (460 MWe) [24,25]. A variety of FCs types exists depending on the fuel and the electrolyte material, each of them with different operation temperature (TFC) and efficiency (hFC). This efficiency is defined in Eq. (1) as the electric energy produced relative to the total chemical energy change [27]: hFC ¼

PFC ; DHFC

(1)

with PFC , the power obtained, and DHFC ; the enthalpy change during the reaction. Table 1 and 2 summarize different characteristics of hydrogen and carbon FCs. An important issue regarding DCFC performance is the impurities and ash content in the carbon, which is hindering the use of biochar. However, the carbon produced by solar thermochemical decomposition of methane is pure and advantageous for use in DCFCs [5]. A FC-based power system needs the fuel and oxidant supply apart from the electrolyte management, cooling and thermal management or reaction products removal [29]. Integration of fuel cells and methane decomposition reactors in different layouts has been reported. Thus Muradov et al. consider the simplest scheme in which a hydrocarbon decomposition reactor is combined to a direct carbon and a hydrogen fuel cells [8]. In spite of the simplicity of the layout,

Please cite this article in press as: Dıáz E, et al., Performance assessment of concentrated solar power plants based on carbon and hydrogen fuel cells, International Journal of Hydrogen Energy (2018), https://doi.org/10.1016/j.ijhydene.2018.01.190

3


Table 1 e Main characteristics of hydrogen fuel cells [28,29]. Type

AFC PEMFC PAFC MCFC SOFC

KOH solution Nafion cation-exchange membranes H3PO4 þ P2O5 and SiC matrix Molten Li2CO3/K2CO3, LiAlO2 as matrix YSZ

overall chemical-to-electrical energy efficiency between 49.4 and 82.5% is estimated. Triphob et al. present a comparison between two concepts, the first one composed of a methane decomposition unit and a solid oxide fuel cell and the second one constituted by a methane steam reforming unit and a SOFC. They point out that a MD-SOFC system can be more competitive when profitability of solid carbon is taken into account [12]. A most complex concept is presented by Liu et al. They carried out an exergy analysis of a system composed of a methane catalytic decomposition reactor together with a direct carbon fuel cell and an internal reforming solid oxide fuel cell. They also consider a gas turbine to use thermal energy generated by the solid oxide fuel cell. Electrical exergy efficiency as high as 68.24% is reported [33,34]. Solar energy as heat source applied to the decarbonisation of fossil fuels by means of their thermal decomposition was proposed by Hirsch et al. [4]. Solar-to-electricity efficiency was calculated for methane solar-thermal decomposition assuming four technical routes: (1) carbon is sequestered and only H2 is used in a fuel cell; (2) carbon is used to fuel a conventional Rankine cycle and H2 is used in a fuel cell; (3) carbon is steam-gasified to syngas in a solar gasification process and the syngas further processed to H2, which, together with H2 from the CH4-decomposition reaction, is used in a fuel cell; and (4) carbon serves as a reducing agent of ZnO in a solar carbothermic process for producing Zn and CO that are further converted via water-splitting and water-shifting reactions to H2 for use in a fuel cell. Their study pointed out the viability of the third and fourth alternatives. Electricity generation by integration of direct carbon fuel cells and solar methane thermal decomposition by means of concentrating solar energy was reported by Cinti and Hemmes [6]. Numerical analysis performed using Cycle-Tempo software points out the difficulties regarding temperature matching between the solar source, due to high solar reactor

Table 2 e Main characteristics of direct carbon fuel cells [30e32]. Fuel Solid graphite rod as fuel and anode Carbon particles as fuel in the electrolyte and anode Carbon particles in a fluidized bed Carbon particles as a slurry in molten tin/salt Gasified carbon (CO or CO þ H2)

hFC (%)

TFC (ºC)

Electrolyte

Electrolyte

TFC (ºC)

Molten hydroxides (OH) Molten carbonates (CO2 3 ) Ceramic (O2)

500e600 800e850 700e900

60e90 80e90 160e220 650e700 800e1000

Theoretical

Practical

83 83 80 78 73

40e60 45e60 50e60 55e65 60e65

outlet temperature and the DCFC inlet one, and also the unreacted methane in the solar reactor. Both aspects are discussed and several options are proposed, however detailed analysis would require new developments of the software. Anyway they conclude that CSP-DCFC concept could potentially surpass conventional CSP efficiency. Vinck and Ozalp analyze the same concept using hydrogen and direct carbon fuel cells. They performed a thermodynamic analysis that incorporates heat integration and experimental data for the solar methane cracking and conclude that an overall chemical-to-electricity efficiency between 35 and 58%, depending on the fuel cells efficiency, can be achieved. However the optimization of heat integration is not described and the maximum efficiency is not clarified. This work expands the referred previous casuistry with the purpose of systematic methodology and better assessment of the potential of this concept as an alternative to conventional generation systems using renewable and/or fossil resources. To do so, firstly the analysis incorporates solar reactor outlet composition assuming reaching thermodynamic equilibrium and empirical data, which allows establishing the margin for maximum improvement in the solar reactor. Secondly, several integration layouts are proposed assuming two different hydrogen fuel cell technologies (solid oxide and proton-exchange membrane fuel cells) working with and without hydrogen purification. Finally, energy efficiency enhancement through waste heat recovery and surplus fuel recirculation has been also analyzed.

Methodology Plant description Fig. 1 illustrates the general layout of the solar thermoelectrochemical power plant. Basically, solar radiation is concentrated by a heliostat field and directed to the solar cavity reactor placed on top of a tower through a small aperture. Useful thermal energy (Qreactor) is utilized to decompose a feed stream of methane, which has been previously preheated recovering heat from the hot products. It is assumed that the solar reactor thermal losses are only by radiation, neglecting those by conduction and convection. The reactor operates at temperature TR and atmospheric pressure and contains no catalyst. Solid and gaseous products leave the reactor and are separated downstream by means of a cyclone, filter or electrostatic precipitator. Then both product streams are cooled down to the working temperature of the carbon and hydrogen fuel cells (TCFC, THFC, respectively), whereas air streams are


4


Air

Electricity Heat

H2 fuel cell CH4+H2

H2-rich gas

Air + H2O

Concentrated solar energy

CH4

HTF

Solar reactor

HTF

Air + H2O

C

HTF

HTF HTF

CO2 (pure)

CO2

C HTF

C fuel cell Air

Air

HTF

Air

Fig. 1 e General layout of the solar power plant based on methane cracking and fuel cells.

simultaneously heated up to the same FC working temperatures. This preheating takes advantage of part of the waste heat generated in both fuel cells. The molten carbonates cell shown schematically in Fig. 2, was selected as DCFC. This device electrolytically oxidizes the solid carbon particles to carbon dioxide. The pure CO2 generated in the anode is compressed to its subsequent capture, except a fraction that is recirculated to the cathode for the purpose of forming carbonate ions, which are required in the electrolyte. CO2 capture does not involve a separation stage because of the high purity of the exhaust gas. Two HFC technologies representative of low- and hightemperature operation were chosen in order to analyze the system performance, polymer electrolyte membrane fuel cells (PEMFC) and solid oxide fuel cells (SOFC) and their basic schemes are shown in Fig. 3. PEMFC was retained as benchmark of low-temperature HFCs, since it is commercially available. Hydrogen FCs usually require high-purity feed in order to preserve a proper electrochemical performance [23],

Fig. 2 e Molten carbonate DCFC layout used in the model.

which leads to the incorporation of a purification stage prior to the cell to remove hydrocarbons produced by side reactions and unreacted methane. A pressure swing adsorption (PSA) technique has been selected for the purification method, which has an impact on the energy balance due to gas compression and gas temperature control. However, analyses on electrochemical performance under the presence of methane has shown that PEMFC operation is feasible without gas purification [9]. This was explained by the unreactive nature of hydrocarbons at the operation temperature [33]. This case has been also analyzed in our model. The SOFC was selected to represent a high operation temperature HFC, avoiding the need for metal catalyst and resulting in fast kinetics [29]. The high temperature allows simplifying thermal integration with the solar reactor and heat management. Fig. 3 (c and d) shows two cases of SOFC analyzed in this study. One includes internal secondary MSR of the unreacted methane at the outlet stream of the solar reactor (SOFCIR) [34]. The second case is SOFC without internal reforming, although not enough experimental data is available about possible solid carbon formation. Thermodynamic analysis of the system sketched in Fig. 1 assumes steady-state operation. CH4 flowrate, solar concentration (Cs) and direct normal irradiance have been assumed as 1 kmol/s, 1000 and 1 kW/m2, respectively, for comparison purposes. Pressure losses and work required to transport the particles or gases were neglected except for the gas compression in the PSA and the capture of released carbon dioxide. It was assumed that the compressor mechanical efficiency, PSA final pressure and the CO2 compression pressure


5


Fig. 3 e HFC layouts used in the model: PEMFC with (a) and without (b) prior purification step (PSA þ PEMFC and PEMFC cases); SOFC with (c) and without (d) internal reforming (SOFCIR and SOFC cases).

were 80%, 11 bar [35] and 110 bar [36], respectively. Heat transfer effectiveness was assumed as 100% except for the solar reactor, which has certain absorption efficiency (habs). Ambient parameters were fixed at 25 C and 1 bar with molar fractions of nitrogen and oxygen of 0.79 and 0.21, respectively. Operation parameters applied in the model for fuel cells are based on literature data [28,29,37] and are shown in Table 3. With regard to the solar reactor, two output chemical compositions have been considered; first one assumes thermodynamic equilibrium and second one is from empirical data. The comparison allows detecting the parameters to be optimized in the reactor in future developments. Previous studies claim that thermodynamic equilibrium is not fully reached in practice [38,39], therefore chemical composition obtained from this approach were used to determine the ideal system performance. They were implemented from 800 to 1600 C to analyze the effect of temperature on both the reactor and the power plant. Experimental data were collected from Ref. [40] and it was assumed that they could be applied to a mean temperature of 1200 C. Table 4 shows the calculated chemical conversions for reactions in Eqs. (2)e(5), which represent the syntheses of C1eC2 hydrocarbons [39]: 2CH4 ¼ C2 H6 þ H2

(2)

C2 H6 ¼ C2 H4 þ H2

(3)

C2 H4 ¼ C2 H2 þ H2

(4)

C2 H2 ¼ 2C þ H2

(5)

Aspen Plus® was used to simulate the different plant flowsheets and to obtain the required data for calculating the efficiencies (c.f. section 2.2). Main process components are the solar reactor and the carbon and hydrogen fuel cells. In addition heat exchangers, separation units and compressors have been taken into account. The solar reactor was simulated by two unit operation blocks in series, where the first one models the chemical transformation and the second one is a separation unit that creates two streams for the gaseous and solid products. RGibbs block was used to obtain the products composition at thermodynamic equilibrium by means of the Gibbs free energy minimization method, whilst an RStoic block was employed to implement experimental data by reactions (Eqs. (2)e(5)), stoichiometry and conversions (Table 4). Fuel cells were also modeled as an assembly composed of an RStoic block (specified stoichiometry and conversions) followed by a separator, which expresses the anode and cathode currents as

Table 4 e Conversion factors of C1eC2 chemical reactions (Eqs. (2)e(5)) calculated from off-gas compositions in Ref. [40]. Run

Table 3 e Operation parameters of fuel cells applied in the model [28,29,37]. Parameter

TFC (ºC) Fuel Utilization (f.u.) (%) hFC (%)

DCFC

800 95 80

HFC PEMFC

SOFC

85 85 45

850 85 60

1 2 3 4 5 5 7 8 9

XCH4 (Eq. (2))

XC2H6 (Eq. (3))

XC2H4 (Eq. (4))

XC2H2 (Eq. (5))

0.700 0.950 1.000 1.000 1.000 0.740 0.970 0.910 0.960

1.000 1.000 1.000 1.000 1.000 0.990 1.000 1.000 1.000

0.968 0.967 0.973 0.974 0.980 0.965 0.974 0.977 0.981

0.566 0.499 0.492 0.430 0.617 0.509 0.459 0.399 0.419


6


separated streams. In the PEMFC case with prior hydrogen purification, a PSA is modeled as a compressor followed by an ideal separator due to the high purification the PSA can reach. Finally, in the case of SOFCIR (SOFC with internal reforming) it was necessary to implement the fuel processor step. This was done by adding a new reaction block (RGibbs) at the same temperature. Aspen Plus® calculates the mass and energy balance solutions, namely the flow rates and conditions of every stream together with the heat or work generated or consumed in each unit.

Efficiency analysis Power plant performance is evaluated by two energy conversion efficiencies (Eqs. (6) and (7)) according to the input and useful energy. Thus Eq. (6) refers to the ratio between the useful power and the input energy from methane, while Eq. (7) refers to the ratio between the useful power and the total input power (chemical from methane and solar). hCH4 ¼

PHFC þ PCFC Plosses QCH4

hCH4 solar ¼

PHFC þ PCFC Plosses QCH4 þ Qsolar

(6)

PFC ¼ DHFC $hFC

(12)

Q FC ¼ DHFC $ð1 hFC Þ

(13)

Two approaches were applied to further enhance the system efficiency. First concept is waste heat recovery, beside system heat integration. This recovery refers to the heat retrieved thanks to the streams that must or could be cooled down together with the residual heat produced in the fuel cells. Specifically, it involves (1) surplus of heat produced in the HFC, at 85 or 850 C depending on the type of the cell, and (2) heat recovered when cooling down the cathodes currents, at 800 C for DCFC and 85 or 850 C for hydrogen cell, (3) heat recovered from cooling the carbon dioxide stream before the compression step, from 800 to 25 C, and (4) heat released by carbon and hydrogen streams when cooled from the reactor TR (800e1600 C) to the PSA operating temperature (25 C) or THFC (85/850 C) and TCFC (800 C). Second approach is recirculation of non-converted fuel in a single pass in the system. This recirculation is implemented in the simulations and involves carbon and hydrogen not oxidized in the cells (returned to each cell) as well as unconverted methane stream (returned to the reactor).

(7)

Where PHFC and PCFC are the power generated in hydrogen and carbon fuel cells, respectively, Plosses is the power consumed and QCH4 and Qsolar are the power supplied by the methane feedstock and solar, respectively. Solar energy supply has been calculated depending on whether a preheater to heat the feedstock to 500 C is used or not according to Eqs. (8) and (9). When feedstock preheating is not applied, the additional energy contribution is supplied by solar energy increasing the input share as shown in Eq. (9). These energy supplies relate to the concentrated solar radiation reaching the solar reactor, disregarding the optical losses of the solar field. The solar reactor efficiency is obtained assuming the thermal equilibrium and is mainly controlled by radiation heat transfer (Eq. (10)) [41]. Finally, the energy delivered by the methane flow rate is given by Eq. (11). 0 ¼ Qsolar

Qreactor habs

(8)

Qsolar ¼

Qreactor þ Qpreheating habs

(9)

habs ¼ 1

sT4R ICs

QCH4 ¼ LHVCH4 $m_ CH4

(10) (11)

where I is the normal irradiance, Cs is the concentration ratio, LHVCH4 is the methane low heating value (50.2 kJ/g) and m_ CH4 is the methane mass flow rate. Regarding the fuel cells, their efficiency given by Eq. (1) allows to describe both electrical (PFC) and heat energy (QFC) produced in the fuel cell in terms of enthalpy change and efficiency (Eq. (12) and (13)). Enthalpy values are provided by Aspen Plus® simulations while efficiencies are taken from literature (Table 3)

Fig. 4 e a) hCH4 and XCH4 vs. reactor temperature assuming thermodynamic equilibrium; b) hCH4 vs. XCH4 for experimental data and thermodynamic equilibrium calculations (red bordered diamonds) at 1200 C. (For interpretation of the references to color in this figure legend, the reader is referred to the Web version of this article.)



7

Results and discussion Results Fig. 4 (a) shows the chemical-to-electrical conversion efficiency (as defined by Eq. (6)) and methane conversion for each proposed scheme as a function of the reactor temperature, TR. It can be observed that efficiency and conversion trends are very similar, which means efficiency is directly affected by the methane conversion. Efficiency depends on the temperature because more hydrogen and carbon are produced at higher reactor temperatures, increasing the fuel feeding rate to both fuel cells and thus increasing the electricity generation. However, methane conversion, and thus efficiency, are stabilized once the reactor temperature is higher than 1100 C and approach an asymptotic value. The lowest performance is obtained for the system with hydrogen purification (PSA þ PEMFC) due to the penalties associated with the compression losses. These accounted for 4e5% of the power generated by the fuel cells, so that the PEMFC efficiency is higher than the PSA þ PEMFC by about 2.5%. Higher efficiencies are achieved by the systems containing a SOFC owing to the higher cell efficiency (0.6 vs. 0.45). Differences between SOFCIR and SOFC come from the additional electricity resulting from the internal reforming of the non-converted methane or by-products. For that reason, this distinction is more significant with smaller conversion values, which lead to a higher hydrocarbons concentration in the reactor output. Here the additional energy due to the reforming as small a 0.03e3% due to the high conversions achieved in thermodynamic equilibrium although it is more noticeable at lower temperatures, entailing slightly lower methane conversions. Fig. 4 (b) compares the efficiency vs. methane conversion using experimental and thermodynamic equilibrium data at 1200 C. As it was observed with equilibrium inputs, experimental hCH4 also increases with methane conversion. However, experimental values for the efficiency are smaller than those from thermodynamic equilibrium due to the presence of gaseous byproducts other than hydrogen such as acetylene (Table 4). SOFC and SOFCIR-based systems reach similar efficiencies when equilibrium methane conversion is achieved and the deviation increases at low XCH4 due to the augmentation of the hydrocarbons fraction. Fig. 5 (a) presents system efficiency (hCH4 solar ) as defined by Eq. (7), which takes into account the solar energy input, as a function of the solar reactor temperature. Solar reactor efficiency decreases when TR increases because radiation losses are proportional to the fourth power of the reactor temperature (Eq. (10)) while the concentration is fixed. Consequently, the global efficiencies (solid lines) follow a curved trend with a rising section on account of the conversion increase with temperature and a decreasing zone due to the radiation losses effect together with the conversion stabilization. This descent effect might be partly moderated if some process heat is applied to the methane preheating (dash-dot lines). This implies saving an important fraction of the required solar energy, with a more significant effect at higher TR. Efficiency can be further improved when waste heat is exploited (dashed line). Here it was assumed an ideal 100%

Fig. 5 e a) hCH4 solar vs. solar reactor temperature assuming thermodynamic equilibrium for various FCs schemes. b) hCH4 solar vs. acetylene conversion for experimental data and thermodynamic equilibrium results at 1200 C (red bordered diamonds) for PEMFC case. (For interpretation of the references to color/colour in this figure legend, the reader is referred to the Web version of this article.)

heat recovery, which provides the maximum achievable potential. In a real situation the heat recovery depends on the process design. Assuming heat recovery, the efficiencies from all the different cases converges. This is a result of the additional amount of thermal energy that is recovered in the solar plant layouts with low temperature HFCs, which makes them approaching the SOFC case. Nevertheless, since SOFC temperature is higher, thermodynamically, there is a potential to obtain even higher energy quality. Efficiencies considering experimental and thermodynamic equilibrium compositions are compared in Fig. 5 (b) for the PEMFC case. The increase in efficiency owing to methane preheating occurs but there is a bigger potential for improvement with heat recovery. This figure illustrates efficiency versus acetylene conversion in order to verify the effect of byproducts formation on the system performance. It can be seen that efficiency is higher when acetylene is dissociated to a larger extent. Either way, efficiency is also affected by the methane conversion as shown in the non-linear trend of the graph. Besides feedstock preheating and waste heat recovery, recirculation of non-converted fuel is other option to improve the process efficiency. Naturally, it entails an improvement in


8


Discussion

Fig. 6 e a) hCH4 solar vs. reactor temperature; b) hCH4 solar vs. experimental methane conversions assuming 1200 C for SOFC case.

efficiency values, as can be seen in Fig. 6. In both cases, experimental and equilibrium data, the effect of the methane conversion becomes almost negligible after recirculation due to the complete fuel utilization. The difference with the base case increases at lower methane conversion because the effect of the recirculation is more significant. Fig. 6 (b) also includes the case implementing recirculation and heat recovery which provides the maximum potential gain obtained from the process, around 70%.

The proposed power plant based on solar methane cracking and fuel cells is auto-thermal; revealing solar shares of 14e23% depending on the reactor temperature and methane conversion. Technological competitiveness of the concept is based on exploiting both the carbon and hydrogen products avoiding additional or more energy-intensive transformations, such as those based on MSR. The DCFC share of the total produced electricity ranges from 30 to 55% experimentally and can reach 55e65% at equilibrium conditions. Around half of the energy produced in the system comes from the obtained carbon, which means that its complete valorization is essential for efficient electricity production from methane and thus one of the most important contribution of this work compared to actual processes. Energy efficiency values (hCH4 solar , Eq. (7)) surpasses commercial CSP plants based on molten salts, of which efficiency is currently established between 32 and 35% excluding the solar field optical efficiency [1], as can be seen in Fig. 7. It was found that efficiencies are lower when the experimental compositions are applied in the reactor analysis compared to the thermodynamic calculations. This emphasizes that the reactor design is a key component in order to achieve high methane conversion and low byproduct formation. Even so, schemes based on high-temperature fuel cells and PEMFC with heat recovery and/or fuel recirculation show also better efficiencies. It has to be underlined that mentioned CSP efficiency refers nominal conditions. Unlike turbines commonly used in power cycles (for instance, steam turbines in Rankine cycles) for which the efficiency usually decreases under partload operation, efficiency of fuel cells is almost constant [42]. Consequently it is expected that the overall performance will be superior using fuel cells compared to turbines if transient is considered. In terms of costs, the target for PEMFC is 1000 $/kWe by 2020 [46]. For SOFC with large-scale manufacturing the target cost is expected to be reduced to 402e532 $/kWe [26]. This can be compared with the capital cost of the current power blocks for a tower CSP with storage of 950e1300 $/kWe [47]while operating at lower efficiencies as expected with the FC [48].

Fig. 7 e Average efficiency (Eq. (7)) results compared with a range of values for current CSP state of the art (32e35%).



9

separation phases, which usually represent between 45 and 85% of the overall carbon capture and storage losses [43,45]. Considering the modeled carbon dioxide compression losses and 2 kJ/mol CO2 assumed for pumping the compressed CO2 into its storage site [45], less than 7% of the electrical power generated by the fuel cells is consumed. In common power generation, this fraction typically represents 30% with a range between 11 and 40% [45]. In terms of the overall global warming effect, the proposed process does not produce particle emissions and its calculated released CO2 accounts for less than half of the current U.S. electric mix and also surpasses the values for the less harmful power generation processes such as combined cycles, as illustrated in Fig. 9.

Fig. 8 e Energy comparison of fuel cells CSP plant (SOFC case) with common energy generating systems.

Comparison with current thermal power generation plants based on turbines [43,44] shows that the above proposed power plant exceeds the typical commercial conversion efficiencies. A comparison can be seen in Fig. 8 where the different process steps are represented in terms of energy efficiency for both systems. First of all, common plants spend part of the fuel energy in the combustion while solar upgrades methane energy by 20% in the solar reactor. Furthermore, fuel cells efficiency overcomes those from turbomachinery cycles (60% vs. 40%). Consequently, the analyzed power plant produces more electricity starting with a lower fuel energy content. Furthermore, the emission handling entails other big advantage of the fuel cells-based system due to the exhaust CO2 purity, which largely decreases carbon sequestration efforts. SOFCIR is the only exception since the internal methane reforming produces it, although its content is less than 1% vol. The pure CO2 prevents consumption of energy associated with capture and

Fig. 9 e CO2 emissions compared with common generation systems (Sources of comparison data [43]).

Conclusions This work systematically analyzes the efficiency of an alternative concept of concentrating solar power plant in which the solar receiver and conventional turbine are replaced by a solar reactor and fuel cells. The methane-containing feedstock can be extracted from renewable sources like landfills or water treatment processes. It has been selected as an excellent hydrogen source, however the carbon contribution is essential for the process economic viability since it entails around half of the produced energy. Energy storage is easily applicable to the system by means of retaining the reaction products, hydrogen and carbon, to be used in the fuel cells when necessary. The present analysis allows identifying the most important parameters influencing the overall performance: fuel cell types, methane conversion, byproducts formation, reactor temperature and heat recovery. The advisable option among the HFCs is SOFC due to the possible utilization of internal reforming, higher cell efficiency, simpler heat integration with the solar reactor, less auxiliary subsystems for gas purification and no need for recirculation. Even so, the solar reactor performance is the key factor that has to be further optimized to maximize methane conversion and reduce byproducts production together with solving the carbon formation handling. Solar-to-electricity efficiency has been calculated, assuming both thermodynamic equilibrium and experimental composition, and compared with commercial thermal power plants (including CSP). The proposed concept is viable and competitive because it avoids significant energy losses related to the thermal-to-mechanical and mechanical-to-electrical energy conversions. Average efficiencies if thermodynamic equilibrium is reached are 47, 49 and 56% for PSA þ PEMC, PEMFC and SOFCIR cases respectively. Although experimental composition implies lower methane conversions some recirculation and heat recovery can be applied, thus achieving 46, 49 and 59% of efficiency for each case. Furthermore, it implies a more flexible operation that enhances the electrical grid integration. This process also improves the current methane-to-electricity conversion carried out in common thermal power plants due to the solar energy upgrading, the high fuel cells efficiency and easier emissions handling.


10


Acknowledgements [14]

The authors would like to thank the Spanish Ministry of Economy and Competiveness and the FEDER (European Funds for Regional Development) funds for their support to the ARROPAR-CEX project (ENE2015-71254-C3-1-R) and the “Comunidad de Madrid” and European Structural Funds for their financial support to ALCCONES project (S2013/MAE2985). E. Dıáz is grateful to Spanish Ministry of Education, Culture and Sport by funding through internship FPU (FPU16/ 00217).

[15]

[16]

[17]

references

lez-Aguilar J. High-flux/high-temperature [1] Romero M, Gonza solar thermal conversion: technology development and advanced applications. Renew Energy Environ Sustain 2016;1:26. https://doi.org/10.1051/rees/2016011. [2] IRENA. Concentrating solar power. Renew Energy Technol Cost Anal Ser 2012;1. https://doi.org/10.1016/B978-0-08087872-0.00319-X. [3] Teske S, Leung J. Solar thermal electricity - global outlook. 2016. p. 2016. [4] Hirsch D, Epstein M, Steinfeld A. The solar thermal decarbonization of natural gas. Int J Hydrogen Energy 2001;26:1023e33. https://doi.org/10.1016/S0360-3199(01) 00040-4. [5] Ozalp N, Vinck I. Energy conversion analysis of a novel solar thermochemical system coupled with fuel cells. In: Proc ASME 2015 Int Mech Eng Congr Expo, Houston-TX; 2015. https://doi.org/10.1115/IMECE2015-51498. [6] Cinti G, Hemmes K. Integration of direct carbon fuel cells with concentrated solar power. Int J Hydrogen Energy 2011;36:10198e208. https://doi.org/10.1016/j.ijhydene. 2010.11.019. [7] Muradov N, Veziroǧlu TN. From hydrocarbon to hydrogencarbon to hydrogen economy. Int J Hydrogen Energy 2005;30:225e37. https://doi.org/10.1016/ j.ijhydene.2004.03.033. nault M, Roux C, Bultel Y, Georges S. Direct [8] Klein JM, He methane solid oxide fuel cell working by gradual internal steam reforming: analysis of operation. J Power Sources 2009;193:331e7. https://doi.org/10.1016/ j.jpowsour.2008.11.122. [9] Muradov N, Choi P, Smith F, Bokerman G. Integration of direct carbon and hydrogen fuel cells for highly efficient power generation from hydrocarbon fuels. J Power Sources 2010;195:1112e21. https://doi.org/10.1016/ j.jpowsour.2009.09.010. lvez JL, Moreno J, Garcıá C. Life cycle [10] Dufour J, Serrano DP, Ga assessment of processes for hydrogen production. Environmental feasibility and reduction of greenhouse gases emissions. Int J Hydrog 2009;34:1370e6. https://doi.org/ 10.1016/j.ijhydene.2008.11.053. [11] Abbas HF, Wan Daud WMA. Hydrogen production by methane decomposition: a review. Int J Hydrogen Energy 2010;35:1160e90. https://doi.org/10.1016/j.ijhydene.2009.11.036. pez R, Pizarro P, [12] Serrano DP, Botas JA, Fierro JLG, Guil-Lo mez G. Hydrogen production by methane decomposition: Go origin of the catalytic activity of carbon materials. Fuel 2010;89:1241e8. https://doi.org/10.1016/j.fuel.2009.11.030. [13] Guil-Lopez R, Botas JA, Fierro JLG, Serrano DP. Comparison of metal and carbon catalysts for hydrogen production by

[18]

[19]

[20]

[21]

[22]

[23]

[24] [25] [26]

[27]

[28]

[29] [30]

[31]

methane decomposition. Appl Catal A Gen 2011;396:40e51. https://doi.org/10.1016/j.apcata.2011.01.036. Carrillo AJ, Sastre D, Zazo L, Serrano DP, Coronado JM, Pizarro P. Hydrogen production by methane decomposition over MnOx/YSZ catalysts. Int J Hydrogen Energy 2016:4e11. https://doi.org/10.1016/j.ijhydene.2016.04.138. Serrano DP, Botas JA, Pizarro P, Moreno I, Gomez G. Hydrogen production through catalytic methane decomposition promoted by pure silica materials. Int J Hydrogen Energy 2014;0:1e7. https://doi.org/10.1016/j.ijhydene.2015.01.056. Ozalp N, Kogan A, Epstein M. Solar decomposition of fossil fuels as an option for sustainability. Int J Hydrogen Energy 2009;34:710e20. https://doi.org/10.1016/ j.ijhydene.2008.11.019. Abanades S, Flamant G. Production of hydrogen by thermal methane splitting in a nozzle-type laboratory-scale solar reactor. Int J Hydrogen Energy 2005;30:843e53. https:// doi.org/10.1016/j.ijhydene.2004.09.006. Maag G, Lipinski W, Steinfeld A. Particle-gas reacting flow under concentrated solar irradiation. Int J Heat Mass Tran 2009;52:4997e5004. https://doi.org/10.1016/ j.ijheatmasstransfer.2009.02.049. Rodat S, Abanades S, Flamant G. Co-production of hydrogen and carbon black from solar thermal methane splitting in a tubular reactor prototype. Sol Energy 2011;85:645e52. https:// doi.org/10.1016/j.solener.2010.02.016. Kogan A, Kogan M, Barak S. Production of hydrogen and carbon by solar thermal methane splitting. III. Fluidization, entrainment and seeding powder particles into a volumetric solar receiver. Int J Hydrogen Energy 2005;30:35e43. https:// doi.org/10.1016/j.ijhydene.2004.03.028. Dahl JK, Buechler KJ, Weimer AW, Lewandowski A, Bingham C. Solar-thermal dissociation of methane in a fluid-wall aerosol flow reactor. Int J Hydrogen Energy 2004;29:725e36. https:// doi.org/10.1016/j.ijhydene.2003.08.009. zaro MJ, Suelves I, Moliner R, Pinilla JL, Torres D, La ~ adas I, et al. Metallic and carbonaceous -based Can catalysts performance in the solar catalytic decomposition of methane for hydrogen and carbon production. Int J Hydrogen Energy 2012;37:9645e55. https://doi.org/10.1016/ j.ijhydene.2012.03.075. Kirubakaran A, Jain S, Nema RK. A review on fuel cell technologies and power electronic interface. Renew Sustain Energy Rev 2009;13:2430e40. https://doi.org/10.1016/ j.rser.2009.04.004. Jackson C. Fuel cells for stationary applications. CPD Fund 2016:1e8. E4Tech. The fuel cell industry review. 2016. https://doi.org/ 10.1595/147106712X657535. Lewis J. Stationary fuel cells - insights into commercialisation. Int J Hydrogen Energy 2014;39:21896e901. https://doi.org/10.1016/j.ijhydene.2014.05.177. Pilatowsky I, Romero RJ, Isaza CA, Gamboa SA, Sebastian PJ, Rivera W. Thermodynamics of fuel cells. Cogener Fuel CellsSorption Air Cond Syst 2011:25e37. https://doi.org/10.1007/ 978-1-4020-6995-6_2. Wendt H, Kreysa G. Electrochemical engineering. Science and technology in chemical and other industries. SpringerVerlag; 2010. Li X. Fuel cells. Energy effic renew energy handb. 2016. p. 1703e53. Giddey S, Badwal SPS, Kulkarni A, Munnings C. A comprehensive review of direct carbon fuel cell technology. Prog Energy Combust Sci 2012;38:360e99. https://doi.org/ 10.1016/j.pecs.2012.01.003. Gu¨r TM. Critical review of carbon conversion in “Carbon Fuel Cells”. Chem Rev 2013;113:6179e206. https://doi.org/10.1021/ cr400072b.



[32] Desclaux P, Rzepka M, Stimming U, Hempelmann R. Actual state of technology in direct carbon fuel cells. Z Phys Chem 2013;227:627e49. https://doi.org/10.1524/zpch.2012.0253. [33] Appleby AJ. Fuel cells. Encycl Electrochem Power Sources 2009;2:277e96. [34] Piroonlerkgul P, Assabumrungrat S, Laosiripojana N, Adesina AA. Selection of appropriate fuel processor for biogas-fuelled SOFC system. Chem Eng J 2008;140:341e51. https://doi.org/10.1016/j.cej.2007.10.007. [35] Yand J, Lee C-H, Chang J-W. Separation of hydrogen mixtures by a two bed pressure swing adsorption process using zeolite 5A. Ind Eng Chem Res 1997;36:2789e98. [36] Damen K, Troost M, Van Faaij A, Turkenburg W. A comparison of electricity and hydrogen production systems with CO 2 capture and storage. Part A: review and selection of promising conversion and capture technologies. Prog Energy Combust Sci 2006;32:215e46. https://doi.org/10.1016/ j.pecs.2005.11.005. [37] Dodds PE, Staffell I, Hawkes AD, Li F, Gru¨newald P, McDowall W, et al. Hydrogen and fuel cell technologies for heating: a review. Int J Hydrogen Energy 2015;40:2065e83. https://doi.org/10.1016/j.ijhydene.2014.11.059. [38] Keipi T, Tolvanen KES, Tolvanen H, Konttinen J. Thermocatalytic decomposition of methane: the effect of reaction parameters on process design and the utilization possibilities of the produced carbon. Energy Convers Manag 2016;126:923e34. https://doi.org/10.1016/ j.enconman.2016.08.060. nades A. Low carbon production of hydrogen by [39] Aba methane decarbonization. Prod Hydrogen Renew Resour 2015;5:149e77. https://doi.org/10.1007/978-94-017-73300_6.

11

[40] Rodat S, Abanades S, Sans JL, Flamant G. A pilot-scale solar reactor for the production of hydrogen and carbon black from methane splitting. Int J Hydrogen Energy 2010;35:7748e58. https://doi.org/10.1016/j.ijhydene.2010. 05.057. lez-Aguilar J, Zarza E. Concentrating solar [41] Romero M, Gonza thermal power. Energy effic renew energy handb. 2015. p. 1237e345. [42] Kartha S, Grimes P. Fuel cells: energy conversion for the next century. Phys Today 1994;47:54e61. https://doi.org/10.1063/ 1.881426. [43] Willnow K. Energy efficiency technologies. World Energy Counc Sustain Energy 2013:1e30. [44] Poullikkas A. An overview of current and future sustainable gas turbine technologies. Renew Sustain Energy Rev 2005;9:409e43. [45] House KZ, Harvey CF, Aziz MJ, Schrag DP. The energy penalty of post-combustion CO2 capture & storage and its implications for retrofitting the U.S. installed base. Energy Environ Sci 2009;25. https://doi.org/10.1039/b811608c. [46] U.S. Department of Energy. Fuel cell technologies office. Multi-Year Research, Development, and Demonstration Plan. 2012. [47] Black & Vetch Holding Company. Cost and performance data for power generation technologies. Natl Renew Energy Lab; 2012. p. 1e106. [48] Hirsch T, Khenissi A. A systematic comparison on power block efficiencies for CSP plants with direct steam generation. Energy Procedia 2013;49:1165e76. https://doi.org/ 10.1016/j.egypro.2014.03.126.
