Large-scale high-temperature solar energy storage using natural minerals

1 Large-scale high-temperature solar energy storage using natural natural minerals Monica Benitez-Guerrero 1, 2, Beatriz Sarrion 2, Antonio Perejon 2, 3, Pedro E. Sanchez- Jimenez 2, Luis A. Perez-Maqueda 2, Jose Manuel Valverde 1* 1 Facultad de Física, Universidad de Sevilla, Avenida Reina Mercedes s/n, 41012 Sevilla, Spain. 2 Instituto de Ciencia de Materiales de Sevilla, C.S.I.C.-Universidad de Sevilla, C. Américo Vespucio nº49, 41092 Sevilla, Spain. 3 Departamento de Química Inorgánica, Facultad de Química, Universidad de Sevilla, Sevilla 41071, Spain. *Prof. Dr. J.M. Valverde Facultad de Física Universidad de Sevilla Avenida Reina Mercedes s/n, 41012 Sevilla (Spain) Tel +34 954550960 Fax +34 954239434 E-mail: jm[email protected] 2 Large-scale high-temperature solar energy storage using natural minerals Abstract The present work is focused on thermochemical energy storage (TCES) in Concentrated Solar Power (CSP) plants by means of the Calcium-Looping (CaL) process using cheap, abundant and non-toxic natural carbonate minerals. CaL conditions for CSP storage involve calcination of CaCO⁠3 in the solar receiver at relatively low temperature whereas carbonation of CaO is carried out at high temperature and high CO2 concentration to use the heat of reaction for power production by means of a CO2 closed power cycle. Under these conditions, large CaO particles derived from limestone to be used in industrial processes are rapidly deactivated due to pore plugging, which limits the extent of the reaction. This is favored by the relatively small pores of the CaO skeleton generated by low temperature calcination, the large thickness of the CaCO3 layer built upon the CaO surface and the very fast carbonation kinetics. On the other hand, pore plugging at CaL conditions for CSP storage does not limit carbonation of CaO derived from dolomite (dolime). Dolime is shown to exhibit a high multicycle conversion regardless of particle size, which is explained by the presence of inert MgO grains that allow the reacting gas to percolate inside the porous particles Keywords: Concentrated Solar Power; Natural carbonates; Multicycle conversion; CaL-CSP storage; CaL-CO2 capture; particle size. 3 1. Introduction It is widely accepted that the massive deployment of power generation from renewable energy sources is one of the essential measures urgently needed to mitigate global warming [1]. Among the different renewable energies, concentrated solar power (CSP) offers the possibility of large scale electricity generation and relatively low cost energy storage in the form of heat for base load power generation. Nowadays, commercial CSP plants are capable of generating power overnight by means of sensible energy storage in molten salts [2-5]. In order to reach a large scale deployment stage, CSP with energy storage needs from cheaper massive energy storage technologies to compete against fossil fuel power plants [6-9]. Thermochemical energy storage (TCES) is being currently investigated as a possible alternative to CSP storage in molten salts. TCES basically consists of using the high temperatures achievable in CSP (up to ~1000ºC in CSP with tower technology) to carry out an endothermic chemical reaction. The byproducts of the reaction are separately stored and brought together to carry out the reverse exothermic reaction on demand. Main general advantages of TCES systems are the high energy density potentially attainable and the possibility of storing energy in the long term [10-13]. Among diverse alternatives currently under research, the Calcium-Looping (CaL) process based on the calcination/carbonation of CaCO3 offers a great potential as it relies on the use of extremely cheap, widely abundant and non-toxic materials such as natural carbonates [14]. The Calcium-Looping (CaL) process has been widely investigated in the last few years for CO2 capture in coal fired power plants [15]. CO2 capture is performed in this process by the exothermic carbonation reaction of CaO (Eq.1). CaO(s)+ CO2(g) ⇄ CaCO3(s) ; ∆Hr0 = -178 kJ/mol (1) The reaction takes place inside a high temperature fluidized bed reactor (T~650ºC) where the post-combustion flue gas at atmospheric pressure carries a CO2 concentration near 15% volume. Once CaO particles are carbonated, they are transported into a second reactor (calciner) where CaO is regenerated by calcination at high CO2 partial pressure and high 4 temperatures (T~950ºC). This way CO2 at high concentration can be extracted from the calciner to be compressed and stored. The CaO particles regenerated from calcination are recirculated into the carbonator to be used in a new cycle. The use of the CaL process in CSP plants for TCES was early proposed in the late 1970s [16-18] and has gained renewed interest in the last years [11,14,19,20]. Figure 1 shows an example of integration scheme recently proposed with potentially high global efficiencies [14, 21]. The system essentially consists of a power unit, a solar calciner, a carbonator, a CO2 compression-storage system and two reservoirs for both CaO and CaCO3 storage. The process starts with calcination of CaCO3 particles in a fluidized bed reactor (calciner) using concentrated solar energy as the source of heat. Once calcination occurs, sensible heat of the reaction products (CaO and CO2) is recovered by heat exchangers before storing them separately. Conditions and time of storage can be adapted to energy demand. When needed, CaO and CO2 are circulated to the carbonator where the heat used for calcination is recovered through the carbonation reaction enthalpy. This heat is carried by the excess CO2 not participating in the carbonation reaction to a gas turbine where electricity is generated in a closed cycle. According to process simulations [14], maximum efficiency CaL conditions for the integration into CSP plants according to this scheme are obtained when carbonation of CaO takes place at high temperature (above 850ºC) and under high CO2 partial pressure. It must be remarked that these CaL carbonation conditions are radically different from the optimum conditions corresponding to the CaL process for CO2 capture. A potential improvement of the CaL-CSP integration consists of using He in the calciner for reducing the calcination temperature by the enhancement of thermal conductivity of the gas mixture and the great increase of CO2 diffusivity in He [22,23]. The CO2/He gas mixture exiting the calciner could be separated by selective membranes currently available at commercial scale [24], allowing the storage of CO2 in compression tanks and the recirculation 5 of He to be used in the calciner environment. The use of He in the calciner allows the calcination reaction to be fully achieved in short residence times at temperatures of just around 700ºC [21], which would make possible the use of commercial and relatively cheap solar receivers based on metal alloys in the calciner. Despite the high price and limited availability of He, it must be taken into account that this gas would be used in a closed loop. Further technoeconomic analysis should be carried out in future works to assess the effect of possible He losses and membrane separation inefficiency. Natural limestone (almost pure CaCO3) is a widely available, low-cost (< 10$/ton) and non-toxic CaO precursor that would help further reducing the cost of the CaL-CSP integration. However, a main potential drawback for the use of natural limestone at industrial scale is the rapid deactivation widely observed in previous works [25,26] under CO2 capture conditions mainly due to agglomeration and sintering of the regenerated CaO grains, which drastically reduces the surface area available for carbonation in short residence times. The majority of strategies that have been developed to use CaO based sorbents for CO2 capture involve doping and other expensive treatments, but the compromise between improving the sorbent performance and the final cost of the technology necessarily requires the use of cheap natural carbonates [27]. On the other hand, a recent work has shown that carbonation/calcination conditions in the CaL process for CSP storage process yield a much higher residual conversion for natural limestone and dolomite derived CaO as compared to CO2 capture conditions [21], which would lead to a high efficiency of the CaL-CSP integration. In the present work, we explore the influence of a further critical parameter on the process performance, which is particle size of these natural carbonates. Generally, particle size plays a relevant role in industrial processes based on circulating fluidized beds (CFBs) relying on the efficiency of open cyclone exchangers for separation of the particles from the gas streams to transport solids between different reactors [14]. The collection efficiency of commercial cyclones drops dramatically for particles under ~10 m in size [28]. Furthermore, the particles trajectories and residence time in the cyclones depend on their size. 6 Thus, the typical particle size lower limit to ensure an acceptable efficiency of cyclones is around 50 m [29]. Most of the investigations focused on the effect of particle size on CaO conversion during carbonation have been carried out under CO2 capture conditions [30-32]. Under these conditions, the main limitation to carbonation regarding particle size is the diffusion of CO2 molecules through the pores of the CaO particles. Intraparticle pore diffusion hinders carbonation for particles larger than about 300 m [33,34]. Taking into account the limitations imposed by elutriation in commercial cyclones and pore diffusion the optimum particle size used in CaL pilot scale plants to capture CO2 is in the range 100-300 m [35-38]. Another important phenomenon that might limit gas-solid reactions concerning particle size is pore-plugging if the pore size is not sufficiently large and carbonation conditions lead to a very fast buildup of a thick CaCO3 product layer on the surface of the CaO particles. At the harsh calcination conditions used in the CaL cycle to capture CO2 for sorbent regeneration (high temperatures and high CO2 concentration), the CaO structure becomes quite sintered with large pores (usually above ~100 nm) [39,40] as compared with the thickness of the carbonate layer built up on the surface (~40 – 60 nm) [41] during the chemically controlled fast reaction phase. Thus, pore-plugging rarely poses a limitation for carbonation. In the present work, we investigate whether pore plugging might limit carbonation of natural carbonates in the CaL process at conditions for CSP storage. Under these conditions, calcination is carried out at relatively low temperatures, which would lead to a high porosity CaO skeleton, whereas carbonation occurs at high temperature under high CO2 concentration leading to the formation of a thick CaCO3 product layer upon the CaO surface in a chemically controlled and quite fast reaction regime [42]. Thus, it might be foreseen that these conditions favor pore plugging, which would pose a limit to carbonation for particles large enough to be used in CFB reactors.This potentially serious drawback for the practical integration of the CaL process in CSP plants is addressed in this manuscript. 7 2. Experimental 2.1. Materials The natural carbonate minerals analyzed in our work were: limestone (99.8 wt% CaCO3) from Matagallar quarry in Pedrera (Sevilla, Spain) supplied by Segura S.L., calcitic marble (99.4 wt% CaCO3) from Purchena (Almería, Spain) supplied by Omya Clariana and dolomite (94.4 wt% CaMg(CO3)2 and 5 wt% CaCO3) from Bueres (Asturias, Spain) provided by Dolomitas del Norte (Spain). All the samples were originally supplied in powder form and were sieved (mesh size 45 m) in order to analyze the multicycle CaO carbonation behavior for particles of different size range (<45 and > 45m) at CaL conditions for CSP storage. This mesh size has been chosen as a typical low limit to ensure efficient gas-solid separation in commercial cyclones. Particle size upper limit was below 300 µm allowing us to dismiss carbonation hindrance by resistance to intraparticle pore diffusion [43]. 2.2. Characterization methods Multicycle carbonation/calcination tests were carried out using a Q5000IR thermogravimetric analyzer (TA Instruments) provided with a high sensitivity balance (< 0.1 µg) and a furnace heated by IR halogen lamps, which allows for high heating/cooling rates up to 300ºC/min and stable isotherms. Under CaL-CSP storage conditions (CaL-CSP test) the experiments started with a precalcination stage from room temperature to the calcination temperature 725ºC, at 300ºC/min under helium atmosphere. Then, the temperature was augmented at 300ºC/min to 850ºC to carry out the carbonation stage under a pure CO2 atmosphere during 5 min followed by a quick decrease of temperature at 300ºC/min to 725ºC, which was maintained for 5 min for CaO regeneration by changing the gas to pure He. After calcination the temperature was decreased down at 100ºC/min to 300ºC and kept for 2 min in a He atmosphere. This cooling stage aimed at mimicking the extraction of sensible heat from the solids after calcination and before being stored. At this point, pure CO2 atmosphere was 8 introduced to repeat the cycle through a carbonation stage. A total of 20 carbonation/calcination cycles were run for each sample. For comparison, multicycle CO2 capture tests were also carried out by employing CaL-CO2 capture conditions involving carbonation at 650ºC under a 15% CO2/85% air v/v atmosphere and calcination at 950ºC under high CO2 concentration (70% CO2/30% air v/v) at atmospheric pressure, high cooling/heating rates (300ºC/min). Short residence times (5 min) for both calcination and carbonation stages were used as expected in practice. In order to avoid undesired effects due to CO2 diffusion resistance across the sample, small masses (10 mg) was tested in all cases. Prior to thermogravimetric analyses, both fractions of sieved dolomite were precalcined under N2 at 800ºC for 30 min in order to decompose ex-situ CaMg(CO3)2 to avoid crepitation phenomena in the TGA apparatus, which is a well-known phenomenon observed especially for relatively large dolomite particles during their decomposition [44]. The ex-situ precalcined samples initially with particles of size > 45 m were checked not to pass across the 45 m mesh size sieve. Scanning electron micrographs were acquired in a Hitachi S4800 SEM-FEG microscope. Pore size distributions were analyzed by mercury intrusion porosimetry, performed using an Autopore IV instrument (Micromeritics). 3. Results and Discussion Figure 2 shows the thermograms recorded during the first cycles at both CaL- CSP storage and CaL-CO2 capture conditions for marble (similar results are observed for limestone). As can be seen in Fig. 2a, first calcination of marble under He is hindered for the larger particles. This may be explained by the high degree of crystallinity of marble, which hampers CO2 to diffuse out of the solid during decarbonation as observed in previous works [45,46]. In contrast, we observe that calcination is not hindered for dolomite as inert MgO grains would favor CO2 diffusion through the solid structure as reported in previous studies 9 [21,47]. In this regard. In this regard, it would be advantageous to use dolomite instead of natural CaCO3, which may exhibit hindered calcination in the case of high crystallinity materials such as marble. The use of dolomite would favor calcination at relatively lower temperatures and/or shorter residence times. Concerning the carbonation stage, Fig. 2 demonstrates a critical role of the carbonation conditions on the kinetics of this stage. As well know from previous studies [16,34], carbonation takes place through two well differentiated phases. A first fast reaction controlled phase occurs on the free surface of CaO particles, which is followed by a relatively slower diffusion controlled phase characterized by the counter-current diffusion of CO32- and O2- anions across the CaCO3 product layer built up on the CaO surface [31, 48]. As seen in Fig. 2a, the fast reaction controlled phase is the main contribution for CO2 uptake under carbonation conditions for CSP storage, which can be explained by the high CO2 concentration and carbonation temperature that enhance the reaction kinetics. In contrast, the fast reaction phase is relatively hindered under carbonation conditions for CO2 capture, which could be explained by the lower CO2 partial pressure (15% v/v) and lower carbonation temperature but mainly by the harsh calcination conditions for CaO regeneration, which drastically reduce the CaO surface area available for carbonation in the fast reaction controlled stage. On the other hand, the slower diffusion phase is promoted under these conditions (Fig. 2b). As reported from process simulations [49] on the integration of the CaL process for CO2 capture, the CO2 capture efficiency and energy penalty could be improved by prolonging the solids residence time in the carbonator beyond a few minutes due to the relevant role of carbonation in the slower diffusion controlled phase as observed in Fig. 2b. Regarding the integration of the CaL process for CSP storage, our results (Fig. 2a) suggest otherwise. As may be observed in Fig. 2a, carbonation in the slow diffusion controlled phase under CaL-CSP storage conditions is negligible. In order to quantify the multicycle CaO conversion performance of the materials employed in our work we will use the effective conversion Xef defined as the ratio of the mass of 16 [20] R. Chacartegui, et al. Sistema de almacenamiento de energía termoquímica a partir fuente térmica a media temperatura basado en la integración de ciclo calcinación-carbonatación (Calcium Looping) y ciclo cerrado de potencia de CO2. Patent P201500493. [21] B. Sarrion, J.M. Valverde, A. Perejon, L. Perez-Maqueda, P.E. Sanchez-Jimenez, On the multicycle activity of natural limestone/dolomite for thermochemical energy storage of concentrated solar power, Energy Technology, 4 (2016) 1013-1019. [22] E.E. Berger, Effect of steam on the decomposition of limestone 1,1, Industrial & Engineering Chemistry, 19 (1927) 594-596. [23] E.L. Cussler, Diffusion: Mass Transfer in Fluid Systems, Cambridge University Press, 1997. [24] E. Taketomo, M. Fujiura, Porous materials for concentration and separation of hydrogen or helium, and process therewith for the separation of the gas, in, Google Patents, 1984. [25] B.R. Stanmore, P. Gilot, Review—calcination and carbonation of limestone during thermal cycling for CO2 sequestration, Fuel Processing Technology, 86 (2005) 1707-1743. [26] P. Sun, J.R. Grace, C.J. Lim, E.J. Anthony, The effect of CaO sintering on cyclic CO2 capture in energy systems, AIChE Journal, 53 (2007) 2432-2442. [27] M. Erans, V. Manovic, E.J. Anthony, Calcium looping sorbents for CO2 capture, Applied Energy, 180 (2016) 722-742. [28] B. Zhao, Y. Su, J. Zhang, Simulation of gas flow pattern and separation efficiency in cyclone with conventional single and spiral double inlet configuration, Chemical Engineering Research and Design, 84 (2006) 1158-1165. [29] R. Utikar, N. Darmawan, M. Tade, Q. Li, G. Evans, M. Glenny, V. Pareek, Hydrodynamic Simulation of Cyclone Separators, 2010. [30] J.C. Abanades, D. Alvarez, Conversion limits in the reaction of CO2 with lime, Energy and Fuels, 17 (2003) 308-315. [31] S.K. Bhatia, D.D. Perlmutter, Effect of the product layer on the kinetics of the CO2-lime reaction, AIChE Journal, 29 (1983) 79-86. [32] Y. Li, C. Zhao, H. Chen, Y. Liu, Enhancement of Ca-based sorbent multicyclic behavior in Ca looping process for CO2 separation, Chemical Engineering & Technology, 32 (2009) 548- 555. [33] G.S. Grasa, J.C. Abanades, M. Alonso, B. González, Reactivity of highly cycled particles of CaO in a carbonation/calcination loop, Chemical Engineering Journal, 137 (2008) 561-567. [34] G. Grasa, R. Murillo, M. Alonso, J.C. Abanades, Application of the random pore model to the carbonation cyclic reaction, AIChE Journal, 55 (2009) 1246-1255. [35] H. Dieter, C. Hawthorne, M. Zieba, G. Scheffknecht, Progress in Calcium Looping post combustion CO2 capture: Successful pilot scale demonstration, Energy Procedia, 37 (2013) 48- 56. [36] B. Arias, M.E. Diego, J.C. Abanades, M. Lorenzo, L. Diaz, D. Martínez, J. Alvarez, A. Sánchez-Biezma, Demonstration of steady state CO2 capture in a 1.7 MWth calcium looping pilot, International Journal of Greenhouse Gas Control, 18 (2013) 237-245. [37] J. Ströhle, M. Junk, J. Kremer, A. Galloy, B. Epple, Carbonate looping experiments in a 1 MWth pilot plant and model validation, Fuel, 127 (2014) 13-22. [38] D.P. Hanak, E.J. Anthony, V. Manovic, A review of developments in pilot-plant testing and modelling of calcium looping process for CO2 capture from power generation systems, Energy & Environmental Science, 8 (2015) 2199-2249. [39] D. Alvarez, J.C. Abanades, Pore-Size and Shape Effects on the Recarbonation Performance of Calcium Oxide Submitted to Repeated Calcination/Recarbonation Cycles, Energy & Fuels, 19 (2005) 270-278. [40] J.M. Valverde, P.E. Sanchez-Jimenez, L.A. Perez-Maqueda, Limestone calcination nearby equilibrium: Kinetics, CaO crystal structure, sintering and reactivity, The Journal of Physical Chemistry C, 119 (2015) 1623-1641. [41] D. Alvarez, J. Carlos Abanades, Determination of the critical product layer thickness in the reaction of CaO with CO2, Industrial and Engineering Chemistry Research, 44 (2005) 5608- 5615. 17 [42] Z.S. Li, F. Fang, X.Y. Tang, N.S. Cai, Effect of temperature on the carbonation reaction of CaO with CO2, Energy & Fuels, 26 (2012) 2473-2482. [43] F. Garcıa-Labiano, A. Abad, L.F. de Diego, P. Gayán, J. Adánez, Calcination of calciumbased sorbents at pressure in a broad range of CO2 concentrations, Chemical Engineering Science, 57 (2002) 2381-2393. [44] J.M. Valverde, P.E. Sanchez-Jimenez, L.A. Perez-Maqueda, Ca-looping for postcombustion CO2 capture: A comparative analysis on the performances of dolomite and limestone, Applied Energy, 138 (2015) 202-215. [45] A. de la Calle Martos, J.M. Valverde, P.E. Sanchez-Jimenez, A. Perejon, C. Garcia- Garrido, L.A. Perez-Maqueda, Effect of dolomite decomposition under CO2 on its multicycle CO2 capture behaviour under calcium looping conditions, Physical Chemistry Chemical Physics, 18 (2016) 16325-16336. [46] P.E. Sanchez-Jimenez, J.M. Valverde, A. Perejón, A. de la Calle, S. Medina, L.A. Pérez- Maqueda, Influence of ball milling on CaO crystal growth during limestone and dolomite calcination: Effect on CO2 capture at Calcium Looping conditions, Crystal Growth & Design, (2016). In Press: DOI: 10.1021/acs.cgd.6b01228 [47] J.M. Valverde, A. Perejon, S. Medina, L.A. Perez-Maqueda, Thermal decomposition of dolomite under CO2: Insights from TGA and in situ XRD analysis, Physical Chemistry Chemical Physics, 17 (2015) 30162-30176. [48] Z. Sun, S. Luo, P. Qi, L.-S. Fan, Ionic diffusion through Calcite (CaCO3) layer during the reaction of CaO and CO2, Chemical Engineering Science, 81 (2012) 164-168. [49] C. Ortiz, R. Chacartegui, J.M. Valverde, J.A. Becerra, L.A. Perez-Maqueda, A new model of the carbonator reactor in the calcium looping technology for post-combustion CO2 capture, Fuel, 160 (2015) 328-338. [50] V. Manovic, J.-P. Charland, J. Blamey, P.S. Fennell, D.Y. Lu, E.J. Anthony, Influence of calcination conditions on carrying capacity of CaO-based sorbent in CO2 looping cycles, Fuel, 88 (2009) 1893-1900. [51] A. Perejón, J. Miranda-Pizarro, L.A. Pérez-Maqueda, J.M. Valverde, On the relevant role of solids residence time on their CO2 capture performance in the Calcium Looping technology, Energy, 113 (2016) 160-171. 18 Figures Fig. 1. Calcium-Looping thermochemical energy storage system for concentrated solar power plants. The integration scheme is developed in Ref. [14], where the interested reader may find additional details. 19 Fig. 2. Time evolution of temperature and sample weight for the first carbonation cycle of CaCO3 using different particles sizes, under CaL-CSP storage (a) and CaL-CO2 capture conditions (b), as indicated. 20 Fig. 3. Multicycle effective conversion (Xef) of sieved limestone and marble samples subjected to CaL cycles under CaL-CSP storage (a) and CaL-CO2 capture conditions (b), and sieved dolomite compared to limestone under CaL-CSP (c) and CaL-CO2 capture conditions (d). Fig. 4. Pore size distribution measured by Hg intrusion porosimetry of CaO obtained after calcination under pure N2 at 750 ºC and pure CO2 at 950 ºC corresponding to CSP storage and CO2 capture conditions, respectively. The insets show the growing process of a CaCO3 layer on a porous CaO particle under CaL-CSP storage and CaL-CO2 capture conditions leading to pore-plugging for relatively large CaO particles under CSP conditions. 21 Fig. 5. SEM micrographs of CaO sorbents, derived for natural CaCO3 (<45 μm) subjected to CaL cycles after calcination at the 20th-cycle under CaL-CSP storage (a) and CaL-CO2 capture (b) conditions, and the corresponding pore size distributions obtained from image analysis (c, d). 22 Fig. 6. SEM micrographs obtained for natural CaCO3 particles subjected to CaL cycles after calcination at the 20th-cycle under CaL-CSP storage conditions: (a) >45 μm; (b) <45 μm. 23 Fig. 7. SEM micrographs obtained for dolomite (<45 μm) subjected to CaL cycles after calcination at the 20th-cycle under CaL-CSP storage (a) and CaL-CO2 capture (b) conditions.