Zhangke Ma ,Yingjie Li, *,Boyu Li ,Zeyan Wang ,Tao Wang ,Wentao Lei

1 School of Energy and Power Engineering,Shandong University,Jinan 250061,China

2 State Key Laboratory of Crystal Materials,Shandong University,Jinan 250100,China

3 Shandong Naxin Electric Power Technology Co.,Ltd.,Jinan 250101,China

Keywords:Calcium looping heat storage Fluidization CaO-based pellets Mechanical property

ABSTRACT The CaO-based pellets were fabricated using extrusion-spheronization method for calcium looping thermochemical heat storage under the fluidization.The effects of adhesive,biomass-based pore-forming agent,binder and particle size on the heat storage performance and mechanical property of the CaObased pellets were investigated in a bubbling fluidized bed reactor.The addition of 2% (mass)polyvinylpyrrolidone as an adhesive not only helps granulate,but also improves the heat storage capacity of the pellets.All biomass-templated CaO-based pellets display higher heat storage capacity than biomass-free pellets,indicating that the biomass-based pore-forming agent is beneficial for heat storage under the fluidization.Especially,bagasse-templated pellets show the highest heat storage conversion of 0.61 after 10 cycles.Moreover,Al2O3 as a binder for the pellets helps obtain high mechanical strength.The CaO-based pellets doped with 10%(mass)bagasse and 5%(mass)Al2O3 reach the highest heat storage density of 1621 kJ.kg-1 after 30 cycles and the highest crushing strength of 4.98 N.The microstructure of the bagasse-templated pellets appears more porous than that of biomass-free pellets.The bagassetemplated CaO-based pellets doped with Al2O3 seem promising for thermochemical heat storage under the fluidization,owing to the enhanced heat storage capacity,excellent mechanical strength,and simplicity of the synthesis procedure.

1.Introduction

The solar energy for electricity production in concentrated solar power(CSP)plants has been in commercial operation,owing to its advantages of renewability and cleanness [1].According to the International Energy Agency,total installed capacity in CSP will rise to 982 GW in 2050 and 11%of electricity generation will come from CSP[2].Under a high solar energy scenario,the development in energy storage technologies becomes essential for CSP due to the natural shortcomings of solar irradiation such as intrinsic variability and dispatchability [3,4].In the recent decade,thermal energy storage (TES) for CSP has been developed significantly [5].The 1st and 2nd generation CSPs use heat-conduction oil and molten salt as sensible energy storage(SES)materials,respectively[6].However,SES systems suffer from several drawbacks,such as operating at the low temperature (below 560 °C),limited power cycle efficiency and salt corrosiveness [7,8].To this end,it is urgent to develop the 3rd generation CSP aimed at increasing the operating temperature and electric efficiency to over 700°C and 40%,respectively [9].

Thermochemical energy storage (TCES) is a suitable alternative to molten salts-based systems for high-temperature application[10,11].The heat storage density of TCES system is usually higher than 1000 kJ.L-1,which is about 15 times as high as that of SES.TCES is based on reversible chemical reactions by using the solar energy to carry out the endothermic reaction [12].When the energy needs to be released,the stored chemical energy is released through the exothermic reaction.A diverse set of materials have been proposed for TCES,such as hydroxides [13],carbonates [14],metal oxides [15],sulfates [16],hydrides [17] and ammonia [18].Calcium carbonate (CaCO3) stands out from the numerous TCES materials due to high reaction temperature and cheap raw materials.Calcium looping(CaL)heat storage using CaCO3is based on the reversible carbonation/calcination reaction (CaCO3?CaO+CO2).The volumetric thermal energy density and reaction enthalpy of CaL heat storage are 3.26 GJ.m-3and 178 kJ.mol-1,respectively[19].The precursors of CaO (such as limestone and dolomite) are cheap,non-toxic and widely available [20].

The CaL process is Type Ⅱchemical looping technology which has been studied widely for post-combustion CO2capture from fossil fuel-fired power plants [21–24].However,the application of CaL for CSP thermochemical heat storage is entirely different from that of CaL for CO2capture.At present,the most acceptable CSP-CaL integration scheme has been proposed by Chacartegui et al.[25] in which the heat produced in the carbonator is used for power generation through a CO2closed Brayton cycle.Fig.1 shows the flow diagram of this configuration and the efficiency of 43%–48% is achieved.This system begins with decomposition of CaCO3(calcination) in the calciner using concentrated solar energy,and then the generated CaO and CO2are separately stored.The calcination can be carried out under helium at below 750 °C[26].In the heat released mode,the CaO and CO2are introduced to the carbonator for the carbonation to generate heat.The exothermic carbonation reaction is operated at high temperature(above 850 °C) to obtain the maximum overall efficiency [26].

One of the main drawbacks of the CaL heat storage process is the progressive deactivation of the regenerated CaO,as the number of heat storage cycles increases.The severe loss in activity of the CaO-based material is attributed to the sintering and poreplugging [26,27].A number of additives such as Al2O3[28],MgO[29,30],SiO2[31],TiO2[32] and ZrO2[33] have been used to enhance the multicycle CaO activity in CaL cycles.Wang et al.[34] reported a stable porous MgO-stabilized nano-CaO powder which achieved high and stable cyclic effective conversion for CaL heat storage.Benbenz-guerrero et al.[35] synthesized a Ca/Al composite for CaL heat storage and found that the Ca4Al6O13helped stabilize the CaO microstructure and mitigate poreplugging.Valverde et al.[26] synthesized nanosilica/CaO composite by mechanical blending for CaL heat storage and found that the addition of nanosilica led to a significant increase in the surface area due to its nanosized porous structure.

As mentioned above,the cyclic heat storage activities of the CaO-based materials have been measured in the fixed bed reactor[27]and thermogravimetric analyzer(TGA)[36].However,chemical reaction efficiencies show better values in the case of a fluidized bed reactor for solid–gas chemical reactions (e.g.carbonation of CaO).Although Gokon et al.[37]and Meier et al.[38]proposed different kinds of solar fluidized bed designed for highly efficient solar thermal utilization,the CaL thermochemical heat storage performance of CaO-based materials under the fluidization has been rarely reported.In our previous work[39],the heat storage density and cyclic stability of the limestone carbonated at higher fluidization velocity are higher than that carbonated at static (solid-like)state.It was proven that higher fluidization velocity significantly mitigates the pore-plugging and sintering of the limestone.However,it should be noted that the limestone carbonated under the fluidization suffered fragmentation and attrition during 20 heat storage cycles.The attrition rate of the limestone remains 0.25 μm per cycle after 20 cycles.To be repetitively used in fluidized bed reactor,CaO-based materials should possess good mechanical property and attrition resistance.Otherwise,the broken fragments resulting from the collision and thermal stress are elutriated out of the reactor and more fresh materials are required.

Granulation is one of the effective ways to improve the mechanical properties of CaO-based materials.The extensive studies of granulation methods have been previously reported,including extrusion granulation [40],rotary drum granulation [41] and extrusion-spheronization granulation [42].The CaO-based pellets prepared by extrusion-spheronization method have been extensively used in the CaL for CO2capture [43].However,extrusionspheronization results in a loss of specific surface area and inferior CO2sorption performance.Consequently,biomass-based poreforming materials are used to improve the pore structure of pellets.Sun et al.[41]prepared carbide slag pellets for CaL CO2capture via extrusion-spheronization and two types of biomass-based poreforming materials (microcrystalline cellulose and rice husk) were used.They found that the addition of 20% (mass) microcrystalline cellulose and pre-washed rice husk displayed high carbonation conversion of 53% and 51% in the 25th cycle,respectively.Xu et al.[44] prepared porous spherical calcium-based sorbents by templating with bamboo for CO2capture and found that the pellets with 2% (mass) bamboo showed the highest capture capacity of 0.26 g.g-1after 25 cycles as well as favorable mechanical strength.Ridha et al.[45] synthesized composite pellets with 10% (mass)leaves which exhibited a high pore volume of 0.11 cm3.g-1,compared to that of 0.07 cm3.g-1for pellets without biomass addition.Li et al.[46] found that the addition of biomass in the CaO pellets mitigated the decrease rate in the heat storage capacity during the repetitive cycles.Pelletization still allows the possibility of doping different binders to improve the mechanical strength,such as bentonites [47],kaolin [48] and calcium aluminate cement[49,50].Sakellariou et al.[51] obtained the promising results for nearly spherical structured formulations using kaolinite as a binder.The combination of natural limestone with 25%(mass)kaolinite rendered mechanically strong materials with a hydration capacity of up to 50%compared to the maximum hydration capacity of pure CaO.Wu et al.[52] found the cement-bound pellets showed stable CO2capture capacity during repeated calcination/-carbonation cycles.It should be noted that the above-mentioned CO2capture performance and heat storage capacity of CaO-based pellets were examined in the fixed bed reactor and TGA.The heat storage and mechanical property performance of CaO-based pellets under the fluidization is still unclear.

Fig.1.Schematic diagram of CaL heat storage system [25].

In this work,we aim to fabricate CaO-based pellets with a high heat storage capacity and good mechanical property for application in the fluidized bed reactor.Although biomass-templated pellets are used for CaL CO2capture,they have seldom been studied for CaL thermochemical heat storage under the fluidization.Thus,the CaL heat storage performance of the biomass-templated CaObased pellets prepared by the extrusion-spheronization method was investigated under the fluidization in this work.The different additives (adhesive,biomass-based pore-forming agent,and binder) were used in the simple and cost-effective procedure to prepare CaO-based pellets.The effects of the adhesive,biomass,binder and particle size on the heat storage performance and mechanical property of CaO-based pellets were investigated in a bubbling fluidized bed reactor.Furthermore,we determined the optimal doping content of additives and analyzed the microstructural characteristics of the CaO-based pellets during CaL heat storage cycles.

2.Experimental

2.1.Materials

The CaO-based pellets as the heat storage material were fabricated by the extrusion-spheronization method.Ca(OH)2(>95%,Tianjin Kermel Chemical Reagent Co.,Ltd.,China) was employed as the source of the CaO.Polyvinylpyrrolidone (PVP,Shandong Yousuo Chemical Technology Co.,Ltd.,China) was chosen as an adhesive for palletization process.Four biomass materials were used as pore-forming agents:bagasse,rice husk,pine and entermorpha.The biomass materials were ground using a crusher (Gilson LC-34) and sieved into a particle size of <0.075 mm by a sieve shaker(Retsch AS400 Control).The proximate analyses of the biomass materials are shown in Table 1.Three powder binders such as Al2O3(>93%,Sinopharm Chemical Reagent Co.,Ltd.,China),SiO2and MgO (>99%,Tianjin Kermel Chemical Reagent Co.,Ltd.,China)were used to act as the inert supports for the pellets.Additionally,a natural limestone with a CaCO3content of 95% was obtainedfrom Henan province,China.Table 2 shows the chemical components of the calcined limestone measured by an X-ray fluorescence(XRF,ZSX Primus II) analyzer.The limestone was ground and sieved into a particle size range of 0.5–0.9 mm.

Table1 Proximate analyses of biomass on an air-dried basis/% (mass)

Table2 Chemical components of calcined limestone/% (mass)

2.2.Preparation of CaO-based pellets

The preparation procedure of CaO-based pellets was presented in Fig.2.First,30 g Ca(OH)2and some powder additives(PVP,biomass and binder) were mixed with the desired proportions in a beaker.Then,7–15 ml deionized water was stirred constantly with well-blended mixture and the amount of deionized water was depended on the used additives.Second,the mixture was extruded through a 0.8 mm sieve to produce columnar shape with uniform diameters in an extruder (Xinyite E25/Mini-S).Then,the samples were sent to a spheronizator (Xinyite E25/Mini-S) to form the spherical CaO-based pellets.The rotational speed of the spheronizer was set at 2000 r.min-1for 1–2 min,and then the speed was reduced to 1000 r.min-1and maintained for 3–5 min.Thus,the CaO-based pellets were obtained and then sieved into a particle size of 0.5–0.9 mm.Lastly,the pellets were dried for 12 h at 100 °C and then pre-calcined at 850 °C under air for 10 min in BFBR.The CaO-based pellets were prepared with different PVP contents (0–2% (mass)),different biomass types (bagasse,rice husk,pine and entermorpha) and quantities (0–15% (mass)) as well as different binder types (Al2O3,SiO2and MgO) and quantities(0–15% (mass)).The obtained pellets were denoted as Ca-Pa-Xb-Yc,where P represented PVP;X was the first letter of the biomass used;Y was the element other than oxygen in the binder;a,b and c were the contents of PVP,biomass and binder,respectively.For instance,the CaO-based pellets doping with 2% PVP,10% bagasse and 5% Al2O3were denoted as Ca-P2-B10-Al5.Moreover,biomass-templated pellets without any binder and biomass-free pellets without any binder were designated as Ca-Pa-Xb and Ca-Pa,respectively.To investigate the effect of the particle size on heat storage performance,the pellets with the larger particle sizes in the range of 0.9–1.5 mm were also prepared according to the above-mentioned procedure.

2.3.Cyclic heat storage test

The cyclic CaL heat storage performance of CaO-based pellets was investigated in a bubbling fluidized bed reactor (BFBR) operated at 850 °C,as shown in Fig.3.The BFBR system basically includes an electric heating furnace,a flue gas analyzer and a data collector.The furnace was a quartz tube with 12 mm inner diameter and 900 mm height.The isothermal region was about 200 mm from the perforated distributor plate in the furnace.Two gas cylinders were used to provide CO2and N2.The fluidization gases in the calcination and carbonation stage were pure N2and CO2,respectively,which were controlled by the mass flowmeters (Flowmethod FL-802).The minimum fluidization velocities of the CaO-based pellets in the size ranges of 0.5–0.9 and 0.9–1.5 mm were 0.3 and 0.4 m.s-1,respectively,which were determined by the descending velocity method in BFBR.The fluidization velocity during heat storage cycles was controlled at 0.6 m.s-1and all the pellets were under the bubbling fluidization.

Fig.2.Preparation procedure of CaO-based pellets by extrusion-spheronization method.

Fig.3.Schematic diagram of BFBR heat storage system.

During each test,the furnace was firstly heated in pure N2by the electrical heating with a rate of 20°C.min.When the temperature of the furnace reached 850°C as the carbonation temperature for CaL heat storage,about 8 g CaO-based pellets were rapidly added into the furnace.Meanwhile,the atmosphere was switched to the reaction atmosphere of pure CO2and the pellets were carbonated for 10 min.After the carbonation,the gas was changed into pure N2.In this stage,the concentration of CO2in the exhaust gas was detected and recorded by a flue gas analyzer (MRU Vario Plus,with a resolution of 0.01% CO2).The sample was calcined completely until CO2concentration was reduced from 100 to 0.Subsequently,the gas was changed into pure CO2again.The above-mentioned procedure was repeated for the cyclic heat storage cycles.The empty bed tests were also carried out according to the above-mentioned procedure,except for the addition of the pellets.In addition,the cyclic CaL heat storage performance of the calcined limestone was performed according to the above-mentioned procedure as a comparison with the CaO pellets.The effective conversion and heat density were calculated by Eqs.(1) and (2),respectively,as follows:

where,N represents number of CaL heat storage cycles;Xef,Nrepresents the effective conversion of the CaO-based material during the Nth cycle;t is calcination time,min;t0is complete calcination time,min;Q is gas flow,L.min-1;φCO2,0(t) is CO2concentration in the exhaust gas from BFBR in the absence of sample at t min,% (vol);φCO2(N,t) represents CO2concentration in gas out of BFBR in the presence of sample at t min during the Nth cycle,% (vol);MCaOdenotes molar mass of CaO,g.mol-1;m0is the gross mass of the CaO-based material before heat storage cycles,g;Hef,Ndenotes heat density of the CaO-based material during the Nth cycle,kJ/kg;ΔH0is the standard reaction heat,kJ.mol-1.

2.4.Mechanical property test

The mechanical property of CaO-based pellets was evaluated from two aspects:crushing strength and attrition resistance.The influence of static mechanical stress on pellets was expressed in the form of crushing strength which was measured by a manual precision pressure tester with a resolution of 0.01 N (LYYS-1000 N,Jinan Lingyue Precision Instrument Co.,measuring range:0–20 N),as shown in Fig.4(a).The pellets with a particle size of 0.5–0.9 mm were placed on the lower disc.The distance between the upper disc and the lower disc was 2 mm.Applied a force to the upper disc to make it move slowly downward.When contacting the pellet,gradually increased the force until the pellet was broken.Recorded the peak value of the force applied when the pellet was broken.Lots of CaO-based pellets after the calcination were tested for every sample,and the average value was seemed as their crushing strength.The method was adopted by Qin et al.[40,53]and Kawatra et al.[54].

The attrition resistance test of biomass-templated pellets was performed on a Friability Tester (Tianjin Jingtuo Instrument Technology Co.,CS-2A).As plotted in Fig.4(b),the operating principle of the Friability Tester was similar to the ones adopted by Qin et al.[40]and Sun et al.[42].The samples were added into the two symmetrical drums (rotation speed:25 r.min-1).After undergoing 3000/5000 rotations in the drum,the pellets were collected by screening the broken pieces using a standard sieve of 0.9 mm and weighed via an electronic balance.The mass loss compared to their initial mass was measured to identify the anti-attrition performance.

It has been proven that the fragmentation and attrition occur because of thermal stress and internal pressure during the CaL heat storage cycles[39].The attrition performance of the limestone and CaO-based pellets were tested after heat storage cycles.The fine particles escaped from BFBR were all caught by the cyclone and the filters,as shown in Fig.3.After finishing the calcination stage of CaL energy storage cycle,the particles in the tube were cooled by introducing pure N2to the room temperature.Then,the particle sizes of the resident particles in the tube and the fine particles escaped from BFBR were measured by a Retsch AS400 Control sieve shaker.The particles with different sizes were weighted by Mettler Toledo-XS105DU electronic balance (with a resolution of 0.1 mg).Thus,the particle size distributions of the limestone and CaObased pellets during heat storage cycles were obtained.The Sauter mean particle diameter and attrition rate were used to describe attrition performance of the samples [39],as follows:

Fig.4.Principle of mechanical property test:(a) crushing strength;(b) attrition resistance test.

where DNdenotes the Sauter average diameters of sample during the Nth cycle,mm;i is number of sieve size;direpresents average diameter of the ith sieve,mm;and xiis mass fraction of particles on the ith sieve,% (mass);RNis attrition rate of sample after N CaL energy storage cycles,μm.

2.5.Characterization

The surface morphology of the CaO-based pellets was obtained using Scanning Electron Microscopy (SEM) with a thermal fieldemission SUPRATM55 instrument.The SEM analysis were performed with an accelerating voltage of 5.0 kV in high vacuum conditions(10-4–10-5Pa).Before test,the samples were sprayed with gold for 60 s in an Emitech K550 Telstar sputter-coating to ensure conductivity.

3.Results and Discussion

3.1.Effect of adhesive in the pellets on CaL heat storage performance

Ca(OH)2is powder and has poor adhesion,adding the appropriate amount of PVP as the adhesive can help granulate and indirectly improve the mechanical property of pellets [49].The effect of PVP content in CaO-based pellets on CaL heat storage performance is shown in Fig.5.In the first 3 cycles,Xef,Nof the CaObased pellets is slightly lower than that of the limestone.However,Xef,Nof the limestone shows a rapid decrease with the number of cycles.After 10 heat storage cycles,Xef,10of the limestone is 9%lower than that of Ca-P0.When 1%(mass)PVP is added,it is found that the CaO-based pellets are easy to granulate,but the effective conversion of pellets increases slightly.With increasing PVP content from 1% to 2% (mass),Xef,1and Xef,10increase by 7% and 10%,respectively.The cyclic stability of Ca-P2 is higher than that of limestone in CaL heat storage cycles.With increasing the number of cycles from 1 to 10,Hef,Nof Ca-P2 and limestone decrease by 22% and 42%,respectively.However,as PVP content increases further,the viscosity of PVP becomes too high,which leads to the difficulty of granulation as well as the increase of cost.Therefore,the feasible PVP content in the CaO-based pellets is 2%.

Fig.5.Effect of PVP content in the pellets on CaL heat storage performance(original particle size,0.5–0.9 mm).

3.2.Effect of biomass in the pellets on CaL heat storage performance

It is known that the combustion of the biomass-based material as the pore-forming agent in the CaO-based pellets releases gases that alters the porosity of the sorbent for CO2capture [49].However,it is unclear whether the combustion of biomass also affects the CaL heat storage performance of the pellets.Fig.6 presents the effect of biomass in the pellets on CaL heat storage performance under the fluidization.It is found that biomass-templated pellets show higher heat storage capacity than biomass-free pellets.For example,Xef,1and Xef,10of Ca-P2-B5 are about 15% and 12%higher than that of Ca-P2,respectively.The biomass in the pellets burns quickly during the pre-calcination stage,so the porosity of the pellets is improved by the released combustion gases.This is beneficial for the carbonation of the pellets.During 10 heat storage cycles,Ca-P2-B5 shows the highest heat storage capacity.Hef,1of Ca-P2-B5 is about 2575 kJ.kg-1,which is 10% higher than that of Ca-P2-E5.After 10 CaL heat storage cycles,Ca-P2-B5 exhibits Xef,10of 0.61,which is 24%lower than Xef,1.Although the heat storage capacity decreases as the number of cycles increases,the decay differs in four kinds of biomass-templated pellets.Xef,10of Ca-P2-P5,Ca-P2-R5 and Ca-P2-E5 achieve 0.59,0.57 and 0.56,which are 24%,26% and 23% lower than their Xef,1,respectively.Table 1 shows that in comparison to rice husk,pine and entermorpha,more volatile and less ash are generated after the combustion of bagasse,which may lead to larger porosity of Ca-P2-B5.

Fig.6.Effect of biomass in the pellets on CaL heat storage performance (original particle size,0.5–0.9 mm).

Fig.7.Effect of bagasse content in the pellets on CaL heat storage performance(original particle size,0.5–0.9 mm).

Fig.7 illustrates the effect of bagasse content in the CaO-based pellets on heat storage performance.With increasing bagasse content from 5% to 10% (mass),Xef,1and Xef,10of bagasse-templated pellets increase by 3% and 7%,respectively.A higher content of bagasse in the pellets leads to more gases released in the precalcination stage,which is beneficial for CaL heat storage of the pellets.However,with increasing bagasse content above 10% further,the pellets show an inferior heat storage capacity.After 10 cycles,the heat storage density of Ca-P2-B15 drops to 1844 kJ.kg-1,which is 5% and 9% lower than that of Ca-P2-B5 and Ca-P2-B10,respectively.This is in accordance with the previous results [45],where low dopant concentration of biomass improved the performance of CaO sorbent,but increasing the dopant concentration resulted in a decline due to more severe sintering.A higher content of bagasse in CaO-based pellets leads to a higher concentration of ash,which results in a negative effect on CaL heat storage.Nonetheless,even with less CaO and more ash,the heat storage capacity of Ca-P2-B15 is still higher than that of Ca-P2.It indicates that bagasse is beneficial for the CaL heat storage and the optimal bagasse content in Ca-P2 is 10%.

3.3.Effect of binder in the pellets on CaL heat storage performance

The purpose of doping the binder in the CaO-based pellets is to obtain high mechanical strength.Al2O3,SiO2and MgO were widely used to act as an inert support of CaO for carbonation [28–31].However,the effects of inert support in the CaO-based pellets on CaL heat storage performance under fluidization have not been reported.The effect of binder type on the CaL heat storage performance of Ca-P2-B10 is shown in Fig.8.It is observed that the doping of MgO and SiO2results in lower heat storage capacity,indicating that MgO and SiO2have the negative effect on the carbonation of CaO.However,the solid phase reaction between CaO and Al2O3forms mayenite (Ca12Al14O33) [55] which results in the stabilization of the structure and mitigation of pore-plugging.After 10 cycles,Hef,10of Ca-P2-B10-Al5 is 2107 kJ.kg-1,which is 2%higher than that of Ca-P2-B10.It is concluded that Al2O3is a good candidate as the binder for the pellets.

The effect of Al2O3content in the pellets on CaL heat storage performance during 30 heat storage cycles is presented in Fig.9.The pellets doped with Al2O3show higher heat storage capacity than Ca-P2-B10 after 30 cycles.For example,Xef,30of Ca-P2-B10-Al5 is 0.51,which is 25% higher than that of Ca-P2-B10.Ca-P2-B10-Al5 shows higher cyclic stability than Ca-P2-B10 during 30 cycles.With increasing cycle number from 1 to 30,Hef,Nof Ca-P2-B10-Al5 and Ca-P2-B10 decreases by 38%and 51%,respectively.Moreover,as Al2O3content rises from 5% to 15% (mass),the heat storage capacity of the pellets decreases by 9%after 30 cycles.This is because more CaO is consumed by the reaction between CaO and Al2O3.Therefore,Al2O3content of 5% in the CaO-based pellets is optimal for heat storage.

Fig.8.Effect of binder in the pellets on CaL heat storage performance (original particle size,0.5–0.9 mm).

Fig.9.Effect of Al2O3 content in the pellets on CaL heat storage performance(original particle size,0.5–0.9 mm).

3.4.Effect of pellet particle size on CaL heat storage performance

The CaL heat storage capacities of Ca-P2-B10 with two particle size ranges of 0.5–0.9 mm and 0.9–1.5 mm are illustrated in Fig.10.Xef,1and Xef,10of Ca-P2-B10 with a particle size range of 0.5–0.9 mm are 18% and 28% higher than those with the particle size range of 0.9–1.5 mm,respectively.This indicates that smaller CaO-based pellets exhibit much higher heat storage capacity.The Ca-P2-B10 with a smaller particle size possesses better heat and mass transfer properties under the same heat storage conditions,which probably leads to a higher CaO reactivity.

3.5.Mechanical property

Fig.10.Effect of particle size on heat storage performance of Ca-P2-B10 during 10 cycles.

Fig.11 displays the attrition behaviors of the CaO-based pellets and calcined limestone under the fluidization in CaL heat storage cycles.As shown in Fig.11(a),after 10 heat storage cycles,the small particles in the new size ranges of 0.2–0.5 and <0.2 mm in limestone,Ca-P2 and Ca-P2-B10 are generated due to the fragmentation and attrition.The CaO-based pellets exhibit higher attrition resistance than the limestone.It is found that after 10 cycles the Ca-P2 and Ca-P2-B10 particles in the size range of 0.5–0.9 mm only decrease by 0.4% (mass) and 0.9% (mass),respectively.However,the limestone particles show an obvious decrease (about 5.8%) in the size range of 0.5–0.9 mm.This indicates that the extrusion-spheronization method is beneficial for improving the mechanical property of CaO-based pellets.In addition,Ca-P2-B10 particles display a slighter increase in the size ranges of 0.2–0.5 and <0.2 mm than Ca-P2 particles after 10 cycles.As illustrated in Fig.11(b),D10of limestone is 10% and 9% lower than that of Ca-P2 and Ca-P2-B10,respectively.It indicates that the breakage of the limestone particles during the cycles is aggravated.Moreover,the attrition rate of Ca-P2-B10 is 2.5 times as high as that of Ca-P2 after 10 cycles.It is explained that the release of the gas during combustion of bagasse contributes to more fluffy and porous structure,but leads to the lower mechanical property.

Fig.11.Attrition behaviors of the CaO-based pellets and calcined limestone after 10 heat storage cycles (original particle size,0.5–0.9 mm):(a) particle size distributions,(b) D10 and R10.

Fig.12.Weight loss of the CaO-based pellets during the attrition test (original particle size,0.5–0.9 mm).

Fig.12 illustrates the attrition testing results of biomasstemplated CaO-based pellets on the Friability Tester.The effect of kinetic stress on the attrition of the pellets is evaluated in the form of weight loss after the certain rotations.It is found that all CaO-based pellets exhibit good attrition resistance and the weight losses are less than 0.3%.Ca-P2-B5 exhibits the highest attrition resistance after the same rotations.The weight loss of Ca-P2-B5 after 3000 rotations is only 0.07%,which is 58%,30%and 34%lower than that of Ca-P2-E5,Ca-P2-P5 and Ca-P2-R5,respectively.It is noted that the weight loss of CaO-based pellets grows as the rotation number increases.For instance,the weight loss of Ca-P2-B5 after 5000 rotations is about 0.15%,which is twice as high as that after 3000 rotations.The weight loss of the pellets during the test indicates that the bagasse-templated pellets have good resistance to kinetic attrition.

Fig.13 shows the effect of binder type on the crushing strength of the CaO pellets during 10 CaL heat storage cycles under the fluidization.The Ca-P2-B10 presents a crushing strength value of 3.11 N,while the value increases by 26% after 10 heat storage cycles.This indicates that the mechanical strength of the CaO-based pellets is improved by sintering and agglomeration of CaO with the number of CaL cycles.A similar phenomenon is also found in the other pellets.It is noted that the Ca-P2-B10-Al5 exhibits the highest crushing strength.For example,the crushing strength of the original Ca-P2-B10-Al5 reaches 4.21 N,which is 13% and 11% higher than that of Ca-P2-B10-Si5 and Ca-P2-B10-Mg5,respectively.The high crushing strength should be related to the good stabilization of Al2O3in the pellets.

The comparison (mass fraction) in the crushing strength of the CaO-based pellets obtained in this work and other results published in the references [34,40,49,53] is shown in Table 3.The crushing strength of CaO pellets in this work is obviously higher than that reported in the reference.For example,the crushing strength of the Ca-P2-B10-Al5 after 10 cycles reaches 4.98 N,which is 59% and 290% higher than that of the CaO pellets reported by Wang et al.[34] and Qin et al.[40],respectively.

Fig.13.Effect of binder on crushing strength of CaO-based pellets(original particle size,0.5–0.9 mm).

Table3 The crushing strength of CaO pellets reported in the references

In a word,the optimum content of PVP as the adhesive in the CaO-based pellets for CaL energy storage under the fluidization is 2%(mass).Considering the heat storage capacity and the mechanical property,the most favorable biomass-based pore-forming agent and binder of CaO-based pellets are 10% bagasse and 5%Al2O3,respectively.The comparison in the heat storage capacity of the CaO-based pellets carbonated in the fluidized bed reactor and the fixed bed reactor or TGA reported in the references[34,46] during 10 cycles is shown in Fig.14.The Ca-P2-B10-Al5 carbonated in the fluidized bed reactor exhibits a higher cyclic heat storage capacity than that carbonated in the fixed bed [46]or TGA [34].For example,Hef,10of Ca-P2-B10-Al5 in this work is 32% and 86% higher than that reported by Wang et al.[34]and Li et al.[46],respectively.This suggests that the fluidized bed reactor is a good choice for CaL heat storage using the CaObased pellets.

3.6.SEM analysis

Fig.15 displays the SEM images of original Ca-P2 and Ca-P2-B10.As illustrated in Fig.15(a)and(c),the surface of Ca-P2 appears more compact than that of Ca-P2-B10,which indicates that the original Ca-P2 has better fragmentation resistance.It is observed that the surface of Ca-P2-B10 has a slight crack due to the combustion of bagasse.At higher magnification,Ca-P2-B10 seems more porous than Ca-P2,as shown in Fig.15(b) and (d).This indicates that the combustion of bagasse improves the porosity of the CaO-based pellets as a result of the gas release.The SEM images of Ca-P2 and Ca-P2-B10 after 10 CaL heat storage cycles are depicted in Fig.16.As shown in Fig.16(a) and (c),Ca-P2 and Ca-P2-B10 after 10 CaL heat storage cycles appear less fluffy,compared with the original pellets.The sintering and agglomeration of CaO grains during cycles lead to the high mechanical property.As exhibited in Fig.16(b),few pores can be observed on the surface of Ca-P2 after 10 CaL cycles.However,the surface of the Ca-P2-B10 after 10 CaL cycles shows more pores than Ca-P2.This indicates that bagasse as the pore-forming agent improves the porosity structure of CaO-based pellets which contribute to the higher heat storage capacity.

Fig.14.Comparison in heat storage performance of CaO-based pellets.

Fig.15.SEM images of original CaO-based pellets:(a) Ca-P2,80×;(b) Ca-P2,20000×;(c) Ca-P2-B10,80×;(d) Ca-P2-B10,20000× (original particle size,0.5–0.9 mm).

Fig.16.SEM images of CaO-based pellets after 10 CaL heat storage cycles:(a) Ca-P2,80×;(b) Ca-P2,20000×;(c) Ca-P2-B10,80×;(d)Ca-P2-B10,20000×(original particle size,0.5–0.9 mm).

4.Conclusions

This work investigated the CaL heat storage performance and mechanical property of CaO-based pellets under the fluidization.The pellets were synthesized with Ca(OH)2,adhesive,biomass,and binder by the extrusion-spheronization method.PVP is beneficial to granulation and the feasible content in CaO-based pellets is 2% (mass).In contrast to biomass-free pellets,biomass-templated pellets show higher heat storage capacity.Xef,1and Xef,10of Ca-P2-B5 are 15% and 12% higher than that of Ca-P2,respectively.The attrition resistance of Ca-P2-B5 is the highest among CaObased pellets.With increasing the bagasse content from 5% to 10% (mass),Xef,10of Ca-P2-B10 achieves 0.65,which is 7% and 12% higher than that of Ca-P2-B5 and Ca-P2-B15,respectively.The addition of the binders such as Al2O3,SiO2and MgO enhances the heat storage of the CaO-based pellets during 30 cycles.Ca-P2-B10-Al5 shows the highest Hef,30of 1621 kJ.kg-1,compared to 1303 kJ.kg-1for Ca-P2-B10.However,an increase of Al2O3content in the pellets does not further enhance the heat storage performance of the pellets.Moreover,Xef,1and Xef,10of the CaO-based pellets with a particle size range of 0.5–0.9 mm are 18% and 28%higher than those with a particle size range of 0.9–1.5 mm,respectively.The CaO-based pellets exhibit higher fragmentation and attrition resistance than the limestone.The dopant of Al2O3in the pellets contributes to higher crushing strength.In addition,the addition of bagasse improves the porosity of the CaO-based pellets which is beneficial for heat storage.It seems promising to use CaO-based pellets for CaL thermochemical heat storage in the fluidized bed reactor.The further investigation for CaL heat storage in a large-scale fluidized bed reactor is necessary for the industrial application.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Acknowledgements

This research was supported by the National Natural Science Foundation of China(51876105),the Fundamental Research Funds of Shandong University(2018JC039)and Major Scientific and Technological Innovation Projects of Key Research&Development Program of Shandong Province (2019JZZY020118).