Review of experimental investigation on directly irradiated particles solar reactors

Review of experimental investigation on directly irradiated particles solar reactors Elisa Alonso a, n , Manuel Romero 1,b a Universidad de Antofagasta. Centro de Desarrollo Energético Antofagasta, Avda. Angamos, 601, Antofagasta, Chile b IMDEA, Energy Avda Ramón de la Sagra 3, Móstoles 28935, Spain article info Article history: Received 14 May 2014 Accepted 12 August 2014 Available online 6 September 2014 Keywords: Solar reactors Particle receivers Entrained reactors Fluidized reactors Fixed reactors Mobile reactors Rotary kilns abstract Solar thermochemistry is a technology that has been demonstrated to contain a high potential development capability. In order to carry out efficient solar chemical reactions, optimized reactors adapted to each chemical process are necessary. In last 30 years many solar reactors of different configurations, performances and sizes have been designed and fabricated by the main solar chemistry research groups. Among them, directly irradiated particles solar reactors operate in a high temperature range that usually correspond to gas–solid thermochemical reactions. This work compiles more than 20 directly irradiated particles reactors designed, constructed and experimentally investigated in the last 30 years. Their description, schemes and main parameters of their performance are given. Detected problems associated are also mentioned. Reactors are classified from the point of view of chemical engineering in entrained, fluidized and stacked beds. Finally, a summary of the main characteristics of reviewed reactors is provided. &2014 Elsevier Ltd. All rights reserved. Contents 1. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53 1.1. Indirect reactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 1.2. Direct reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 2. Early studies on particles absorbance. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54 3. Particles reactors classification..........................................................................................54 3.1. Entrained reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.2. Fluidized reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 3.3. Stacked reactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.3.1. Fixed reactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59 3.3.2. Mobile reactors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 3.3.3. Rotary reactors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 4. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 Acknowledgment . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66 References..............................................................................................................66 1. Introduction The objective of solar chemistry is to use solar radiation for producing fuels and chemicals [1]. It transforms solar energy in long-term storable and transportable energy carriers, what represents a significant contribution to the requirements for improving the current energetic system. In order to increase the efficiency, these processes should be performed at an upper temperature as high as possible because the rate of chemical reaction increases Contents lists available at ScienceDirect journal homepage: www.elsevier.com/locate/rser Renewable and Sustainable Energy Reviews http://dx.doi.org/10.1016/j.rser.2014.08.027 1364-0321/&2014 Elsevier Ltd. All rights reserved. Abbreviations: SNL, Sandia National Laboratories; SPCR, solid particle central receiver; NREL, National Renewable Energy Laboratory CNRS-ENSIC; NG, natural gas; CPC, compound parabolic concentrator; CSIRO, Commonwealth Scientific and Industrial Research Organization; PSI, Paul Scherrer Institut; ETH, Eidgenössische Technische Hochschule; DLR, Deutschen Zentrums für Luftund Raumfahrt n Corresponding author. Tel.: þ56 552513530. E-mail address: [email protected] (E. Alonso). 1 Tel.: þ34 917371120. Renewable and Sustainable Energy Reviews 41 (2015) 53–67 exponentially with temperature. To achieve that, apart from an adequate concentration system, optimized solar reactors capable to withstand high temperatures, minimizing heat loss and favouring mass and heat transfer between chemicals involved are required. Solar reactors are particular cases of solar receivers where the absorber heat is employed to carry out endothermic chemical reactions. Thus, a preliminary classification of solar reactor could correspond to that generally used to categorize solar receivers. According to the heat integration mode into the reaction chamber, solar reactors are indirectly or directly irradiated [2]. 1.1. Indirect reactors The external opaque walls of the reaction chamber are heated by solar energy. Endothermic reactions are promoted by the heat flux transferred by conduction from the walls to the reactants. Most of indirect irradiated reactors found in literature are catalytic tubular reformers [3–7] where the catalyst is fixed inside a tube and the gas is forced to flow across. The solar flux is distributed along the external walls of the tubes. Additional concepts of indirect reactors have been also developed for thermochemical applications, such as double-cavity reactors with a reaction chamber physically separated from the one that receives the radiation [8–10]. 1.2. Direct reactors Reactants are directly irradiated and heated by incoming solar concentrated radiation. Reactors are opened to air or closed by a transparent window through which radiations enters into the reaction chamber. Due to the absorption of radiation occurs on the reactants surface, higher temperatures are expected by working with directly irradiated reactors. Apart from volumetric reactors, which physical processes have been widely studied from the point of view of volumetric absorbers [11,12], particles solar reactors represents the largest group of studied solar reactors, particularly in a laboratory scale. Thus, a compilation and examination of designed, constructed and experimentally investigated hitherto particles solar reactors are the aims of the present work. 2. Early studies on particles absorbance Sandia National Laboratories (SNL) was pioneer in proposing solid particles as heat absorber medium for concentrated solar systems. The initial study was made by Martin and Vitko [13]. Pebbles and sand were tested as heat solid carriers with successful results. Detected advantages of such as scheme included direct absorption of the incident radiation, direct heat transfer, use of the working fluid as a storage medium, and ease of hybridizing with a fossil-fired system. Moreover, higher temperatures were theoretically possible. Their study led to the first concept of Solid Particle Central Receiver (SPCR) consisting in a falling cloud of 100– 1000 μm solid particles [14]. Solar energy was directed to the particles through the aperture of a cavity receiver [14,15].By means of a theoretical parametric study they analyzed how the material kind, the particles size or the infrared scattering albedo affected to the particles temperature, the convective loss and the optical thickness. Further studies were done on this topic, most of them collected on internal reports of SNL [16–20]. Such as research line concluded to a prototype of SPR that was tested on top of Sandia's 61 m tall central receiver located at the National Solar Thermal Test Facility in Albuquerque, NM. The heliostat field of this facility is able to provide 5 MW th [21].The SPR consisted of a 6 m tall cavity through where a 1 m wide curtain of spherical ceramic particles was dropped and directly heated with concentrated solar energy. Particles were collected in an internally insulated lower hopper. Fig. 1 shows the system layout. First experimental research demonstrated a cavity temperature higher than 1000 1C, receiver efficiency in excess of 55% and particle exit temperature in excess of 200 1C, depending on the operating conditions. Siegel and Kolb [21] reported an expected improvement of the results after an optimization of the receiver design. In fact, SNL are still in the process of developing SPCR. In 2012 DOE's SunShot Initiative awarded three research projects in this area that currently have been developed by SNL, NREL and San Diego State University. Some other studies on particles receivers have been found in literature. For example, Bertocchi et al. [22] reported experimental evaluation of a solar particle receiver designed for a power input of 10 kW. It consisted of a conical cavity of 40 mm high and 78 mm of inner diameter close to the ambient by a quartz window. Particles were injected in a gas/particle suspension at the focal plane through a duct placed close to the window. The exit takes place at the rear of the receiver. Achieved exit gas temperatures exceeded 2100 K. 3. Particles reactors classification The knowledge developed on particles receivers helped to the conception of the first receiver-reactors where particles perform both heat absorption and chemical transformation. Particles may be arranged with different configuration depending on their required residence time inside the reactor and the existence of a Fig. 1. Lay out of the SPCR developed in SNL [21]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–6754 carrier fluid and its contact mode with the particles. These parameters affect directly to the heat and mass transfer occurring inside the reaction zone. Thus, particles solar reactors can be distinguished, according to the classification proposed by Villermaux [23] in entrained, fluidized and stacked reactors (see Table 1). Falling cloud of particles solar reactors such as the receiver described before–in case it had chemical proposalshould be included among the entrained type reactors. Selecting one type of particles solar reactor may fundamentally depend on the type of application. Villermaux [23] recommended the use of fix beds, included in the stacked reactors group, for solar catalytic reactions. For those reactions that require good thermal transfer properties, he suggested the employment of fluidized beds. The close contact between gas and fluidized particles favours a homogeneous heating at the reaction medium. Moreover, higher temperatures are expected to be achieved. His recommendation about entrained reactors varies depending on the sub-type what they belong to. For example, cyclones are interesting if the further separation between solid and gas is desired. Villermaux also mentioned the interest of rotary kilns due to the wide knowledge and experience existing already on solid treatment in this type of reactors. Some other classification criteria could have been employed to distinguish types of solar reactor. Application is one of the possibilities selected by other authors. For example, Puig-Arnavat et al. [24] published in 2012 a state of the art on solar reactors applied to the gasification of carbonaceous feedstock or Steinfeld [25] relates different solar reactors to different routes for hydrogen production through thermochemical routes. Since the initial investigations on the particles solar reactors, many authors have designed different prototypes with their particular characteristics. Villermaux's selecting criteria have not been always followed; nevertheless the compilation that is presented in this work distinguishes between entrained, fluidized and stacked reactors. 3.1. Entrained reactors While SNL were starting to develop the first concepts regarding SPCR, chemical engineers from CNRS-ENSIC, France [26], thought out a cyclone as an innovative concept of particle solar reactor. It was purposed to study the continuous flash pyrolysis of wood sawdust at 1143 K. Results concluded that such a reaction could be carried out on a large scale in such a new type of reactor by use of concentrated solar energy. In 1991, Imhof et al. [27] proposed a new solar cyclonic reactor conceived to develop gas–solid thermochemical processes. With this concept, an easy way to continuously feed reactants and remove products was sought in combination to some of the advantages of volumetric receivers such as high absorbance of radiation [27]. The cyclonic reactor, shown in Fig. 2, consisted of a truncated conical cavity 30 cm-height that was opened to the atmosphere. Inner walls were covered by a ceramic insulation layer. The remainder of the cavity was formed by two concentric cones that formed a conical gap for the gas exhaustion. The reactor was designed to be used at the Paul Scherrer Institute (PSI) 17 kW solar furnace. The authors studied the thermal decomposition of calcium carbonate as an example of gas–solid reaction. Employed experimental set-up, shown in Fig. 2, mainly included a gas and particle injection system, the solar reactor, a cyclone separator for collecting solid products, and a heat exchanger to cool down the outlet gas stream. They reported to have achieved reactants temperatures about 1300 K, high degree of calcination and a global efficiency of 43%. This efficiency is based on the ratio of the overall energy absorbed, sensible and process heat, to the energy incident on the aperture. A further prototype, of same configuration but larger dimensions was fabricated and tested at the 55 kW McDonnel Douglas dish [28]. Fig. 2. Scheme of the cyclonic solar reactor and experimental set-up reported by Steinfeld et al. [28]. Table 1 Gas–solid reactors classification proposed by Villermaux in 1980 [23]. Falling particles reactors were not included in the original classification. However it has been added to the Villermaux's table taking into account the previous section of this work. Stacked beds Fixed Mobile Conveyor belt Blast-furnace Brewed oven Multistage Rotary kiln Vibrated or pulsated Fluidized or suspended beds Fluidized bed Circulating fluidized bed Blown bed Drooping bed Entrained beds Pneumatic transport Cyclone Falling particles E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–67 55 A novel solar chemical reactor was designed in 1998 by Steinfeld et al. [29] also based on a vortex flow confined to a solar cavity-receiver. A small prototype of 5 kW, called SynMet was firstly fabricated and tested at the PSI solar furnace. It was conceived for the co-production of metallic Zn and syngas starting from ZnO and natural gas (NG). It consisted of a cylindrical cavity provided with a quartz window. Concentrated radiation entered through the window and heated the reactants and the cavity walls. Particles of ZnO, conveyed in a flow of NG, were continuously injected into the cavity via a tangential inlet. A vortex flow of reactants progressed from the back to the front of the reactor. Achieved temperatures exceeded 1600 K and chemical conversion from ZnO to Zn reached 90%. An innovative incorporation on this reactor design was the actively cooled window. An auxiliary flow of gas was injected tangentially and radially at the window in order to maintain it cooled and clear of particles. This reactor was further employed to produce CaO and syngas by the combined CaCO 3 -decomposionand CH 4 -reforming processes [30]. SYNPET project (2003–2009) resulted in a solar reactor similar to SynMet but applied to the steam-gasification of petcoke. Main modifications were associated to the reactant feeding system that, after several attempts, was finally a slurry generator [31,32]. The reactor tested in a high-flux solar furnace in the range 1300–1800 K yielded up to 87% petcoke conversion. The solar-to-chemical energy conversion efficiency attained 9% without accounting for the products sensible heat, and 20% when the sensible heat was recovered for steam generation and pre-heating. The vortex reactor was then scaling up from 5 kW to 300 kW [33,34]. Scaling up improved the solar-to-chemical efficiency to 24% due to the advantageous volume-to-surface ratio. It was also found a big dependence of the conversion to the particle size and residence time inside the reactor. Fig. 3 shows both schemes of 5 kW reactors, the one proposed in 1998 and the subsequent in 2006. Researchers at Weizmann Institute worked on a new solar cyclone reactor to perform the thermal splitting of methane. Such a configuration was chosen in order to allow an effective reactor window screening and to keep it free of generated carbon particles. Thus, first efforts were destined to analyze the advantages of a tornado flow in an unseeded reactor provided with a transparent window [35].Temperatures up to 1320 K were achieved and the maximum extent of reaction reported was 28.1%. The main problem found was a high degree of carbon deposition on the reactor walls and also on the exit port, what often triggered the experimental test termination. In order to favor the generation of carbon particles by methane splitting in the central region on the chamber instead of the walls, it was proposed to seed the reaction chamber with radiation absorbing particles [36]. This way, temperature was increased at the middle of the cavity. However, it was necessary to prevent destruction of the reaction window by contact with incandescent solid particles. Different geometries were examined in order to optimize the reactor design. Fig. 4 shows the last proposed geometry that was considered a satisfactory solution. The main reactor body was composed of three segments. Main gas streams were introduced through a distributor assembled at the top of the reactor to generate a tornado flow pattern inside the chamber. Two extra gas enters were located between the segments and seeding gas can be injected optionally through either one of the lateral ducts d 1 ,d 2 or d 3 . On the basis of a cloud of particles absorbing solar direct radiation Ganz et al. [37,38] proposed in 1994 a novel reactor concept. It consisted of a cylindrical cavity of heat-resistant steel that contains a windowless aperture. Reactants in powder impinged on a cone and were conveyed in a swirling air stream. They were directly exposed to high solar flux. The radial air jet helped to prevent the particles from leaving the reactor. Products exited via a water-cooled axial tube where reactants were quenched. A schematic of the solar reactor is represented in Fig. 5. From previously mentioned experiences, it can be inferred that vortex-flow based reactors are normally feasible choice for heating Fig. 3. Schemes of the two 5 kW vortex flow reactors. Left: Prototype designed for ZnO reduction with NG [29]. Right: Prototype designed to the steam-gasification of petcoke [32]. Fig. 4. Satisfactory geometry selected by Kogan et al. to create a proper tornado flow preventing the window destruction [36]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–6756 solid particles in the range of 1300–1800 K, that are common for many thermochemical reactions (for solar fuels production, thermochemical storage, etc.). Geometry of the cavity and injection of gas and reactants are the main aspects to take into account to get a properly designed prototype. 3.2. Fluidized reactors In comparison to entrained reactors, fluid beds improve gas– solid contact and increase particles residence time. Such a characteristic may represent an advantage particularly for those chemical reactions associated to slower kinetic mechanism. An early solar fluidized bed was proposed by Flamant in 1980 [39].It consisted of a transparent silica tube (34 mm in diameter) between two metallic brackets. The tube was directly irradiated by concentrated solar radiation (Fig. 6). This device was tested in a 2 kW solar furnace for heating refractory materials at 600–1300 1C and decarbonation of calcite at 850 1C. Graphite particles were attached with calcite in order to increase the chemical rate. Thermochemical conversion was also improved with graphite upper to 14%. The scaling-up to 50 kW of this reactor was then realized by Flamant et al. [40]. Other concepts improving this prototype were later proposed [41] such as an opaque fluidized bed provided with a transparent window on the top or an annular fluidized reactor with opaque external walls that is also irradiated from the top. Preliminary solar and non-solar thermochemical tests are often carried out in a laboratory scale fluidized bed due to their easy construction and operation. Fig.7 shows a small solar fluidized reactor designed by Steinfeld et al. [42] to perform the ZnO reduction and CH 4 reforming. It consisted of a 2 cm in diameter quartz tube. A compound parabolic concentrator (CPC) and an involute provided uniform irradiation on the tubular reactor. With this arrangement the design offered high thermal efficiency, low thermal capacitance and good thermal shock resistance. ZnO particles were fluidized in CH 4 and reaction was activated by the concentrated radiation given by the PSI solar furnace. The fluidizedbed, operated under vigorous bubbling conditions, was likely to be at uniform temperature. The reactor-receiver operated at 1373 K and under uniform solar flux of 57 W/cm 2 system exhibited very low thermal inertia, good termal shock resistance and proved to be well adapted for direct absorption processes. A maximum of 43% of the CH 4 in the reducing gas was converted. A similar design, showed in Fig. 8 was employed in 2009 to perform consecutively CaO-carbonation and CaCO 3 -calcination to Fig. 6. Scheme of the lab-scale fluidized bed proposed by Flamant et al. [40]. Fig. 7. Scheme of the solar fluidized bed proposed by Steinfeld et al. [42]. Fig. 8. Fluidized solar reactor tested in the PSI high flux solar simulator and set up of the process [43]. Fig. 5. Solar reactor proposed by Meier et al. [38]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–67 57 capture CO 2 from atmosphere [43]. Solar reactor also consisted of a quartz tube, 25 mm-outer diameter, 25 cm-height, containing a fluidized bed of reacting particles. A high flux solar simulator was employed to irradiate the tube at its top with a power up to 75 kW. Simulator focal plan was located at the middle of the fluidized bed. Successful results were obtained with a mass balance of CO 2 after five consecutive cycles close to 99%. Maximum temperatures of 1150 K were achieved for the calcination step. Gokon et al. [44] proposed a novel solar fluidized bed. It was based on the concept that concentrated solar radiation passed through a transparent window on the top of the reactor and directly heated an internally circulating bed of reactant particles. They specifically used NiFe 2 O 4 /m-ZrO 2 , which two-step thermochemical cycle was investigated. On the basis of this concept two laboratory reactors were developed (Fig. 9) and tested using a 6 kW Xe-arc solar simulator. The employed input of power was 2.4–2.6 kW. First one was constructed using quartz tubes, with an outer diameter of 45 mm. Carrier gas was introduced through a conical cap and a draft tube in order to create an internally circulating bed. The reactor was preheated at 900 1C. A second fluidized reactor was constructed of stainless steel with approximately the same dimensions. A quartz window was located at the top of the reactor to let the concentrated radiation enter in the reactor. Chemical conversion was found to increase by including a draft tube in the reactor design that is, generating an internal circulation. Moreover, better results were achieved with the stainless steel prototype, presumable due to an increasing on the thermal efficiency caused by the opaque walls. The stainless steel reactor was later employed to perform both successive reactions of thermal-reduction and water-decomposition steps, also starting with NiFe 2 O 4 /m-ZrO 2 particles [45].The gas feed was switched from an inert gas in thermal decomposition to an inert gas and steam mixture in water decomposition. This second reaction was performed under 1.6–1.7 kW of incoming power. Successful hydrogen production was achieved with a reported chemical conversion of 45% and maximum measured temperature of 1200 1C. Several authors have studied the solar gasification process in a directly irradiated fluidized bed reactor [24]. This is the case of the beam down irradiated fluidized bed concept proposed by Gokon et al., what was applied for the CO 2 gasification of coal coke with successful results. It was reported a maximum energy conversion efficiency of 14%. Some other examples of fluidized bed employed Fig. 9. Internally circulating fluidized solar reactors proposed by Gokon et al. Left: quartz prototype. Right: stainless steel prototype [44]. Fig. 10. Side view of TREMPER [47]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–6758 for solar gasification of carbonaceous material are compiled in [24] carried out by centers such as ETH and University of Minnesota. 3.3. Stacked reactors According to the classification reported by Villermaux [23] it is appropriate to distinguish between fixed, mobile and rotary reactors or kilns. In general terms, it has been observed that rotary reactors favor mass and heat transfer phenomena. However, design, construction and operation on fixed reactors are usually easier and more economic. Particularly, those prototypes conceived as small and preliminary laboratory tools to start studying chemical processes are often simple fixed bed. 3.3.1. Fixed reactors TREMPER is a fixed reactor developed by Frey et al. [46] to study chemical kinetics of reactions on a second scale at temperatures up to 2100 K. TREMPER was composed of a set of elements all of them enclosed in a quartz tube (Fig. 10). The sample was placed in a shallow hole on a water-cooled copper support. This arrangement prevented the sample from reacting with the support. A 451mirror guided the concentrated solar radiation from a horizontal source to the sample. Carrier gas swept the quartz tube and was collected with a silica capillary. A gas flow was directed towards the mirror to avoid deposition of evaporated material. A quenching unit was included at the outlet. A flow patterns analysis determined the whole apparatus behaved as a continuous ideally stirred tank [47] and hence, gas release rates could conveniently be obtained from on-line gas analysis at the outlet. Reduction of iron and manganese oxides were performed at TREMPER and maximal chemical conversion reached was 25% for iron oxide and 85% for manganese oxide. A pattern flow analysis was also done by Alonso et al. to study the behavior of a 1 kW solar reactor conceived to perform thermal reductions of non-volatile metallic oxides [48]. The solar reactor is schematically represented in Fig. 11. It consisted of a vertical stainless steel and internally insulated vessel 100 mm high and 80 mm in diameter that was closed to the ambient by a quartz window. Concentrated radiation provided by a 7 kW e solar simulator is guided by a 451-sloped water-cooled mirror towards the cavity of the reactor. Samples were introduced inside a crucible forming a packed-bed of particles. Carrier gas was introduced into the reaction chamber to sweep it and collect gas products. Flow into the reactor was found to behave close to an ideal plug flow. Manganese oxides, Mn 2 O 3 yMn 3 O 4 , were reduced at the top layer of the packed-bed with a maximum conversion of 60% for Mn 2 O 3 . A maximum temperature of 1400 1C was measured at the top of sample. However, high thermal gradients were produced between the irradiated surface and the bottom of the sample. Similar flow pattern characterization as well as thermal tests was done on a novel 2 kW prototype designed by Alonso et al. [49] (Fig. 12). The solar reactor mainly consisted of a central alumina cavity 110 mm length and 60 mm in diameter closed at the front by a cone, also made of alumina and by an alumina cover at the back. There was placed the gas outlet. The cone included a watercooled quart window. The housing, made of stainless steel was separated from the cavity by an insulation layer. Gas inlet took place by four perforations located in cross at the frontal cone. After crossing the cavity it left through a duct placed at the back of the reactor. Reactants were packed inside a 20 mm inner diameter, 26 mm outer diameter and 50 mm in length cylindrical ceramic sample-holder that rested in the middle of the cavity. Radiation entered through the quartz window and an aperture. It was found that the fluid flow behavior was closed to an ideal Plug Flow Reactor, particularly at high temperature and high carrier gas flow. An effective thermal efficiency of 47% was achieved at 530 W of incoming power. High thermal gradients were found at sample with a maximum measured temperature of 1450 1C. Reductions of Mn 2 O 3 ,Mn 3 O 4 and CeO 2 were performed achieving complete conversion for the first one and variable for the rest, depending on the reaction conditions. Fig. 11. Schematic of the 1 kW vertical solar reactor developed by Alonso et al. [48]. Fig. 12. Schematic section of the reactor developed by Alonso et al. [49]. Fig. 13. Fixed reactor developed by Moller and Palumbo [50] to dissociate ZnO. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–67 59 Moller et al. [50,51] also employed a fixed reactor to dissociate pre-sintered ZnO. A schematic of the reactor used is shown in Fig. 13. The reactor chamber was filled with stabilised ZrO 2 insulation. In the center of the chamber, at the focal point of the solar furnace, a ZrO 2 receptacle accepted a cylindrical pellet of ZnO. Concentrated solar energy passed through the window and illuminated the ZnO front surface. Several inert gas streams prevented products vapor from condensing on the window. The gaseous reaction products and the inert gas flowed continuously through the reactor to a chimney and out to a heat exchanger where the products were quenched. Temperatures up to 2100 K were achieved at the sample. To study the reduction step of the CeO 2 based water splitting thermochemical cycle, Abanades et al. [52] developed a solar device consisting in a spherical vessel of Pyrex ™which permitted a pellet of reactant heating by concentrated solar energy and the control of atmosphere composition and pressure. It is shown in Fig. 14. This kind of solar reactor was closed to air and it could operate under controlled atmospheres: vacuum, inert or reactive gases. Results showed that total thermal reduction of Ce (IV) to Ce (III) could be obtained working with proper parameters such as initial sample mass, reaction time, pressure and sample temperature. Successful results also depended on the gas flow hydrodynamics surrounding the sample. Temperatures of 2000 1Cwere detected at the sample because it was observed CeO 2 melting. A post-treatment characterization of the sample revealed high thermal gradients in the sample. Chueh et al. [53] studied the solar thermochemical H 2 O–CO 2 splitting cycle using also cerium oxides. A novel fixed solar reactor was employed to perform the two stages that are involved in such a cycle, first one at higher temperature and second at lower temperature. A schematic of the solar device was shown in Fig. 15. It consisted on a thermally insulated cavity receiver containing a porous monolithic ceria cylinder. Concentrated solar radiation entered through a windowed aperture and impinged on ceria inner walls. Reacting gases flowed radially across the porous ceria toward the cavity inside, whereas product gases exited the cavity through an axial outlet port at the bottom. The feasibility of the complete cycle was demonstrated and stable and rapid generation of CO and H 2 was achieved by dissociating CO 2 and H 2 O. Material stability was also demonstrated over 500 thermochemical cycles. However, solar-to-fuel efficiencies of 0.7–0.8% were only achieved. Authors suggest thermal transfer controls the high temperature stage of the process. Moreover, low efficiency was also attributed Fig. 14. Solar reactor used by Abanades and Flamant [52] to reduce CeO 2 under controlled atmosphere. Fig. 15. Schematic of the solar reactor proposed to study the solar thermochemical H 2 O–CO 2 splitting cycle [53]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–6760 to the system scale and design, due to it is limited by thermal losses. Thus, the need of a reactor optimization was suggested. As a part of the two-step thermochemical water splitting cycle the high-temperature thermal reaction of ZnO and SnO 2 was also investigated by Chambon et al. [54] in a 1 kW solar fixed reactor. It was composed of a cylindrical water-cooled shell (depth: 76 mm, diameter: 88 mm) made of stainless steel. It was close to the ambient by a convex window made of Pyrex ™. The cavity was internally coated with a ceramic insulation. Pellets of reactant were stacked in a ceramic tube exceeding of 7 mm the bottom of the refractory cavity. They formed a rod that was pushed upward via a screw piston manually rotated for achieving a continuous reactant injection during an experimental run. Temperatures of 1900 K were measured at the cavity. Kinetic parameters for ZnO and SnO 2 dissociations were obtained. Although the yield of product particles recovery was not very high (around 2%), authors expected to achieve better results by optimizing a quenching device at the reactor output. Fig. 16 shows and scheme of this reactor. A different concept of solar fixed reactor led to a novel design developed by the Tokyo Institute of Technology, (Japan). It was conceived for solar hydrogen production with two step water splitting process. The basis of the reactor were two fixed cells located onto a rotor that switch two different type reaction rooms, one for discharging oxygen and another for water splitting reaction. The reactor rotated in order to alternate the reaction cell that is exposed to concentrated solar radiation. Fig. 17. Shows a schematic outline of the rotary-type solar furnace for two-step water splitting cycles. Experiment performing two thermochemical cycles, CeO 2 and Ni,Mn-ferrite, were developed in a laboratory scale reactor and using an infrared lamp as heat source [55]. Repetition of the two stages of both process were achieved, and successive evolution of H 2 and O 2 in respective water splitting and reduction cell were observed. Optimum reaction temperatures found were 1623 K and 1273 K in case of CeO 2 cycle and 1473 K and 1173 K for ferrite cycle. The solar reactor was scheduled to be scaled-up and operated at the solar concentrating system of CSIRO, using an input solar power of 10–30 kW. This work was developed in the frame of the Asia-Pacific Partnership on Clean Development and Climate project [56]. The idea of rotating reaction cells also led to develop another concept in SNL. The main objective was to perform ferrites thermochemical cycles [57]. It was called Counter-Rotating-Ring Receiver/Reactor/Recuperator (CR5) and used a stack of counterrotating rings or disks with fins along the perimeter. The fins contained ferrite reactant, presumably on a support. Each ring rotated in the opposite direction to its neighbor at a rotational speed on the order of one RPM or less. Solar flux illuminated the fins on the stack of rings on edge along nominally 1/4 of the perimeter. On the opposite side of the stack, the water oxidation reaction took place. The remaining half of the stack (two 1/4 sections between) was adiabatic and is utilized for counter current recuperation occurs, primarily by thermal radiation. Equal pressures are maintained in the two reactors to minimize flow through the recuperator sections. Fig. 18 shows a schematic of the CR5, with a detail of the set of counter-rotating rings. Preliminary thermal efficiency reported was 29.9% although it depended on operating temperature and the irreversibility of their internal Fig. 16. Fixed solar reactor developed by Chambon et al. [54] to investigate the high-temperature thermal reaction of ZnO and SnO 2 . Fig. 17. Novel solar reactor developed by the Tokyo Institute of Technology [55]. E. Alonso, M. Romero / Renewable and Sustainable Energy Reviews 41 (2015) 53–67 61