Thermal Driving Demonstration of Li₄SiO₄/CO₂/Zeolite Thermochemical Energy Storage System for Efficient High-Temperature Heat Utilizations

Abstract

Thermochemical energy storage (TcES) system using lithium orthosilicate/carbon dioxide (Li₄SiO₄/CO₂) reaction was developed for recovery and utilization of high temperature thermal energy generated from high temperature industrial process. Li₄SiO₄/CO₂ TcES packed bed reactor (LPR) and zeolite packed bed reactor (ZPR) were developed as thermal energy storage and CO₂ reservoir. Both reactors, LPR and ZPR, were connected by flexible tube and thermal driving operation of TcES system was demonstrated. For lithium orthosilicate packed bed reactor, tablet forms of Li₄SiO₄ named K-tablet was developed and used in this study.

Li₄SiO₄ carbonation (thermal energy output process) and lithium carbonate (Li₂CO₃) decarbonation (thermal energy storage process) were conducted sequentially with specific condition. All experimental results showed similar tendency; a middle temperature in the Li₄SiO₄ packed bed reactor rapidly increased and decreased at the initial time of carbonation and decarbonation respectively. From kinetic analysis, it was confirmed that the developed K-tablet reacted around 80% and a thermal energy output density of LPR was estimated 331–395 kJ/L-packed bed, 759–904 kJ/L-material. The thermal driving demonstration of Li₄SiO₄/CO₂/Zeolite TcES system shows high possibility to utilize surplus heats in low-carbon ironmaking system efficiently.

1. Introduction

Waste heat recovery from high-temperature industrial processes, such as ironmaking system and concentrated solar power system, is expected to be useful for energy saving and efficiency enhancement for low-carbon energy systems.¹⁾

Especially, Thermal energy storage (TES) systems are expected to increase the thermal energy system efficiency and to be bridge the gap between demand and supply of thermal energy. This kind of TES system can be carried out by three different methods: as sensible heat, as latent heat, or as thermochemical energy storage.^2,3) The most common types of TES systems are sensible heat and latent heat, however, sensible heat and latent heat TES systems incur thermal energy losses while storing; the large thermal energy loss can impede storage ability for long term use.^4,5)

Among the thermal energy storage methods, thermochemical energy storage (TcES), that stores the absorbed heat chemically, is considered as an alternative to conventional sensible and latent TES systems. The energy losses during the storing period often become a challenge in conventional TES systems as mentioned above and TcES system is expected to overcome this weakness through storing the thermal energy in chemical bonds by the sorption property of the working pair.^6,7,8,9)

Various kinds of gas-solid reaction working pairs are under investigation but candidate reaction for high temperature, around 700°C, has been limited: dehydrogenation of metal hydride (80–400°C),^10,11) dehydration of metal hydroxide (250–800°C),^12,13) decarboxylation of metal carbonates (100–950°C),^14,15) and thermal deoxygenation of metal oxides (600–1000°C).^16,17) Lithium orthosilicate/carbon dioxide (Li₄SiO₄/CO₂) reaction, Eq. (1), that has −127 kJ mol⁻¹ of reaction enthalpy at 700°C was proposed and investigated by author’s group for using thermal energy around 700°C.¹⁸⁾ A TcES system for high temperature is expected to enhance the utilization of surplus heat from high temperature systems including ironmaking industry, concentrated solar power, high-temperature gas cooled reactor, and so on.   

$$\begin{array}{l}{\text{LI}}_{\text{4}}{\text{SiO}}_{\text{4}}{\text{(s) + CO}}_{\text{2}}\text{(g)}\rightleftarrows {\text{Li}}_{\text{2}}{\text{CO}}_{\text{3}}{\text{(s)+Li}}_{\text{2}}{\text{SiO}}_{\text{3}}\text{(s),}\\ \mathrm{\Delta }{H}_r^{700°\text{C}}\text{=}-\text{127}\text{.0 kJ}\mathrm{\hspace{0.17em} }{\text{mol}}^{-1}\end{array}$$(1)

In this study, zeolite was proposed for CO₂ storage during decarbonation of lithium carbonate (Li₂CO₃) and CO₂ supply for carbonation of lithium orthorsilicate (Li₄SiO₄). Figure 1 shows the schematic diagram of lithium orthosilicate/carbon dioxide/zeolite (Li₄SiO₄/CO₂/Zeoilite) TcES system. The decarbonation is endothermic reaction corresponding to thermal energy storage process, Fig. 1(a), and carbonation is exothermic reaction corresponding to thermal energy output process, Fig. 1(b). This TcES system store the surplus heat and release it when needed by using low grade of thermal energy. The lithium orthosilicate packed bed reactor (LPR) plays a major role in storing and releasing thermal energy, and zeolite packed bed reactor (ZPR) function as a CO₂ reservoir.

Fig. 1.

Schematic diagram of Li₄SiO₄/CO₂/Zeolite TcES system. (a) Decarbonation/thermal energy storage process, (b) Carbonation/Thermal energy output process.

There were many studies for TcES’s working pair in aspect of material and theoretical operation, however there have been few studies on large scale demonstration. In this study, we investigated two different reactors in large scale and demonstrated thermal driving feasibility of high temperature TcES system without any mechanical power. In addition, performance of developed material and investigated system were estimated through kinetic analysis.

2. Experimental Preparation

2.1. Materials

The advanced Li₄SiO₄ tablet (K-tablet) was prepared for lithium orthosilicate packed bed reactor by the solid state method as follows. Lithium carbonate (Li₂CO₃, 99.0%, Wako pure Chemical Industries, Ltd.), silicon dioxide (SiO₂, 99.0%, Wako pure Chemical Industries, Ltd.), and potassium carbonate (K₂CO₃, 99.0%, Wako pure Chemical Industries, Ltd.) were prepared in 2.0:1.0:0.1 molar ratio as a precursor; the SiO₂ and K₂CO₃ were added for enhancing the synthesizing and CO₂ sorption property of material.¹⁹⁾ The prepared precursors with chemical additive were compressed into tablets (Φ6 mm × 2 mm, ρ = 0.9 g/cm³) under 10 MPa, then precast tablet was sintered at 785°C for 12 hrs under Ar atmosphere, Fig. 2(a). As a preliminary study, performance evaluation of developed K-tablet was conducted by using thermogravimetric analysis (TGD-9600, Advance RIKO Inc.). On single isothermal experiment, Fig. 2(b), K-tablet showed higher performance than pure Li₄SiO₄ powder at all temperature. Stability of K-tablet was also investigated through 10 cyclic experiment at 700°C, Fig. 2(c). Developed K-tablet shows good durability and stable reacted conversion, Δx, during cyclic experiment. From preliminary study, we could confirm the superior performance and durability of K-tablet and 329 g of K-tablet was used for LPR in this study.

Fig. 2.

Li₄SiO₄ materials’ performances. (a) Exterior image of K-tablet, (b) Single cycle operations under isothermal condition for the powders and K-tablet, and (c) 10 cyclic experimental result for K-tablet.

The zeolite packed bed reactor was developed as a CO₂ reservoir and 3300 g of commercial zeolite (F-9, 1.40–2.36 mm, Wako pure chemical industries, Ltd.) was filled into reactor designed for specific operating condition.

2.2. Reactors and Experimental Apparatus

The two reactors, LPR and ZPR (SUS 304), were designed in consideration of actual operating condition. LPR works under high temperature, around 700°C, in vacuum state and ZPR implement rapid temperature change in specific time. Additionally, thermocouples were set in each reactor for measuring the temperature change during experiment.

Figure 3 shows the cross section of vacuum chamber and top view of LPR. The three alumina Tammann tubes (SSA-H-T8, OD × ID × L = 50.0 × 40.0 × 180 mm, Volume = 225 mL) with sheath heater were used for electric furnace. The electric furnaces were fixed in vacuum chamber and vacuum chamber was encased by ribbon and jacket heaters. A ZPR was designed for effective heating and cooling of zeolite bed as a CO₂ reservoir. The ZPR consisted of two layers of cooling tube and sheath heater on inner and outer sides of vacuum chamber, Fig. 4. The volume of vacuum chamber is 4.9 liter and metal mesh tube was positioned in the center of reactor for enhancing CO₂ distribution.

Fig. 3.

Image of (a) cross section of vacuum chamber and (b) top view of LPR.

Fig. 4.

Image of (a) cross section of vacuum chamber and (b) side view of ZPR.

The schematic diagram of thermal driving demonstration system for Li₄SiO₄/CO₂/Zeolite thermochemical energy storage is shown on Fig. 5. Two reactors were connected by flexible and rigid pipe with solenoid valve, and mounted on an electric balance to measure the weight change during experiment. The solenoid valve controlled supplying CO₂ pressure according to setting by receiving the current value from pressure gauge. All temperature, pressure, and mass data is collected through data acquisition system.

Fig. 5.

Schematic diagram of thermal driving demonstration system.

2.3. Experimental Procedure

The experimental procedure can be divided into two process for each reactor; carbonation of LPR/desorption of ZPR (CD experiment) corresponding to thermal energy output process and decarbonation of LPR/adsorption of ZPR (DA experiment) equivalent to thermal energy storage process. Before start each experiment, Li₄SiO₄ tablet was fully decarbonated by heating in vacuum state, around 10 Pa. Table 1 shows experimental conditions and procedure is as follow. First, the LPR heaters maintain a crucible temperature at 715°C during all experiment. In CD experiment, ZPR was heated up to 120°C for CO₂ desorption, then desorbed CO₂ was introduced to LPR with set pressure for 4 hours, in this study 150 kPa. After finishing the carbonation, we closed the main valve and cool down ZPR for DA experiment, around 2 hours. When middle temperature of ZPR had been reached 15°C, DA experiment was carried out for 6 hours. In this study, 2 times of thermal driving demonstration were conducted according to above procedure; the one is single operation and the other is cyclic operation.

Table 1. Thermal driving experiment condition.

LPR			ZPR			Note
Carbonation	Tcar [°C]	715	Desorption	Tdesorp [°C]	120	- Time: 4 hrs
	Pcar [kPa]	150.0		Pdesorp [kPa]	150.0	- Sol. valve: Pressure control
Decarbonation	Tdecar [°C]	715	Adsorption	Tsorp [°C]	15	- Time: 6 hrs
	Pdecar [kPa]	13.0		Psorp [kPa]	13.0	- Sol. valve: Open

3. Experimental Results and Discussion

3.1. Experimental Results for Thermal Driving

The results of thermal driving demonstration are shown in Fig. 6 and Table 2. Each experiment showed similar temperature and mass change trend as follows.

Fig. 6.

Thermal driving demonstration results. (a) Single operation, (b) Cyclic operation.

Table 2. Results for thermal driving experiments.

Experiment	Highest temp. T [°C]	Carbonation Δmcarb [g]	Decarbonation Δmdecb [g]	Lowest Temp. T [°C]
Single Operation	792	93.7	84.1	683
1st Cycle of CO	786	91.4	81.5	690
2nd Cycle of CO	748	80.3	78.2	693

- Middle temperature was sharply increased at the initial time of CD experiment (thermal energy output process), on the other hand, the temperature was suddenly decreased in the beginning of DA experiment (thermal energy storage process).

- Bottom temperature did not show particular fluctuation during the whole experiment. Not enough CO₂ diffusion in electric crucible is thought to be the cause of small temperature change and also it will influence on insufficient reacted conversion of LPR.

- Upper temperature measured above K-tablet packed bed was lower than other temperatures and it represents atmosphere of electric crucible. On CD experiment, temperature was increased by influence of exothermic reaction and it was also increased on DA experiment by high temperature desorbed CO₂ gas from K-tablet.

The overall weight change of LPR and ZPR were corresponded well each other and these results were used for calculation of reacted mole fraction and thermal energy capacity.

The single operation (SO), Fig. 6(a), and first cycle of cyclic operation (1st CO), Fig. 6(b), showed almost similar results. The highest temperatures were 792°C and 786°C, and mass conversion in carbonation, Δm_carb, were 93.7 g and 91.4 g respectively. However, on second cycle of cyclic operation (2nd CO), the highest temperature and mass conversion in carbonation were decreased to 748°C and 80.3 g. On the other hand, mass conversion in decarbonation, Δm_decb, and lowest temperature were similar in all experiment, Table 2. The highest temperature and mass conversion in carbonation were decreased on 2nd CO due to not enough decarbonation on 1st CO; carbonation on SO and 1st CO was initiated with fully decarbonated state.

3.2. Kinetic Analysis of Results

The reacted mole fraction, x [−], was defined as follow Eq. (2) and calculated from mass change of ZPR that can be measured mass change more precisely than LPR, since lighter weight and generous operating condition.   

$$x=\frac{\mathrm{\Delta }m/M_{{\text{CO}}_{\text{2}}}}{m_{\text{o}}/M_{{\text{Li}}_{\text{4}}{\text{SiO}}_{\text{4}}}}$$(2)

Δm: Mass change of ZPR [g]

$M_{{\text{CO}}_2}$: Molecular weight of CO₂ [g mol⁻¹]

$M_{{\text{Li}}_4{\text{SiO}}_4}$: Molecular weight of Li₄SiO₄ [g mol⁻¹]

m_o: Initial mass of Li₄SiO₄ in LPR [g]

The all calculated reacted conversion, Δx, on Fig. 7(a), shows similar tendency with mass conversion, Δm, on Table 2. Even though, 1st CO shows higher reaction rate during first 50 min by enhanced CO₂ diffusivity due to repeated experiment, total reacted conversion of SO and 1st CO are similar, Δx = 0.79 and 0.76. In case of 2nd CO, reacted conversion and reaction rate for carbonation were lower than 1st CO and this result reflected similar results of cyclic experiment on preliminary experiment, Fig. 2(c). Meanwhile, all reacted mole fraction for decarbonation shows similar reacted conversion, Δx = 0.62–0.68.

Fig. 7.

Kinetic analysis of thermal driving results. (a) Reacted mole fraction, (b) Thermal energy output capacity and output rate.

The thermal energy output capacity, Q_LPR [kJ/L] that is thermal energy released per unit volume of LPR was estimated from calculated reacted mole fraction and defined by following equation, Eq. (3). The evaluating of thermal energy output rate, w_LPR [kW/L], was also evaluated through Eq. (4).   

$$Q_{\text{LPR}}=-\frac{m_0}{M_{{\text{Li}}_{\text{4}}{\text{SiO}}_{\text{4}}}}\cdot \mathrm{\Delta }x\cdot \mathrm{\Delta }{H}_{\text{r}}$$(3)

  

$$w_{\text{LPR}}=\frac{\text{d}{Q}_{\text{LPR}}}{\text{d}{t}_{carb}}$$(4)

where ΔH_r [kJ mol⁻¹] is standard enthalpy change of reaction and t_carb [sec] is the time interval during the carbonation.

In Fig. 7(b), all experiment shows similar highest thermal energy output rate, w_LPR, around 5 kW/L-bed, however, that of 2nd CO was decreased rapidly and lasted shorter than other experiments. As a result, 2nd CO shows lower thermal energy output rate and capacity. Total thermal energy output capacity, Q_LPR, of SO and 1st CO are estimated 395 kJ/L-bed and 381 kJ/L-bed respectively and that was decreased to 331 kJ/L-bed on 2nd CO, Table 3. The overall results of thermal energy output capacities show similar tendency with reacted mole fraction results on Fig. 7(a).

Table 3. Thermal energy output and storage capacities of LPR and Li₄SiO₄ tablet.

Experiment	Li4SiO4 packed bed reactor		Li4SiO4 tablet material
	Thermal energy output QLPR [kJ/L-bed]	Thermal energy storage QLPR [kJ/L-bed]	Thermal energy output QLPR [kJ/L-material]	Thermal energy storage QLPR [kJ/L-material]
Single Operation	395	342	904	786
1st Cycle of CO	381	332	874	765
2nd Cycle of CO	331	316	759	727

Figure 8 shows the thermal energy output and storage capacities according to experimental time. Both graphs on Fig. 8, thermal energy output and thermal energy storage, do not show particular increasing after 150 min. In thermal energy output process, Fig. 8(a), 1st CO shows the highest thermal energy output capacity at first 50 min, but after 150 min, SO shows higher thermal energy output capacity than others. On the contrary, there is no certain change in the order on thermal storage process, Fig. 8(b). About thermal energy recovery ratio, the ratio of thermal energy storage capacity to thermal energy output capacity (Q_storage/Q_ouput) SO and 1st CO showed same value, 87%, and that value is increased to 96% on 2nd CO. From the above result it is expected that if increasing the number of cycles, thermal energy output and storage capacities would be stable. The total heat capacities of developed Li₄SiO₄ tablet were also estimated from a material perspective on Table 3. The developed tablet shows over 700.0 kJ/L-material on all thermal energy output and storage process. These values are meaningful because TcES method has over 20 times higher energy density than sensible and latent TES method theoretically.²⁰⁾ The high energy density of TcES system could be expected to downsize existing sensible and latent TES system for high temperature.

Fig. 8.

Thermal energy output and storage capacity changes. (a) Thermal energy output process, (b) Thermal energy storage process.

4. Conclusion

Li₄SiO₄/CO₂/Zeolite TcES system was investigated and thermal driving feasibility of the system was demonstrated. The advanced Li₄SiO₄ tablet and special reactors, LPR and ZPR designed to specific operating condition, were developed for practical thermal driving demonstration. All experiments were carried out in vacuum state with follow conditions. The carbonation (thermal energy output process) was conducted under 715°C with 150 kPa of CO₂ pressure for 4 hours, and decarbonation (thermal energy storage process) was implemented by using pressure difference between LPR and ZPR for 6 hours. Thermal energy output density of LPR was estimated at 331–395 kJ/L-packed bed, and 759–904 kJ/L-material. In the cyclic operation, while first cycle shows similar results with single operation, overall performance was decreased on second cycle. However, the heat capacity is expected to be stable as increasing the number of cycles.

The developed Li₄SiO₄/CO₂/Zeolite TcES system shows high possibility to enhance efficiency of high temperature industrial process and to substitute conventional TES system for low-carbon ironmaking system. More investigation, such as temperature and CO₂ pressure dependency or durability of performance, is required as a further study.

Acknowledgement

This work was supported by the SIP Program (Cross-Ministerial Strategic Innovation Promotion Program), Energy Carrier Research, operated by the Japan Science and Technology Agency.

References

• 1)   T.  Akiyama and  J.  Yagi: Tetsu-to-Hagané, 82 (1996), 177 (in Japanese).
• 2)   C.  Prieto,  R.  Osuna,  A. l.  Fernandez and  L. F.  Cabeza: Renew. Energy, 99 (2016), 852.
• 3)   S. M.  Hasnain: Energy Convers. Manag., 39 (1998), 127.
• 4)   M.  Al-Zareer,  I.  Dincer and  M. A.  Rosen: J. Heat Transf., 140 (2018), 022802.
• 5)   K.  Pielichowska and  K.  Pielichowski: Prog. Mater. Sci., 65 (2014), 67.
• 6)   N.  Yu,  R. Z.  Wang and  L. W.  Wang: Prog. Energy Combust. Sci., 39 (2014), 489.
• 7)   S.  Narayanan,  X.  Li,  S.  Yang,  I.  McKay,  H.  Kim and  E. N.  Wang: Proc. ASME 2013 Heat Transfer Summer Conf., ASME, New York, (2013), 17472.
• 8)   A.  Hauer: Sorption Theory for Thermal Energy Storage, Springer, Dordrecht, (2007), 393.
• 9)   J.  Janchen,  D.  Ackermann,  H.  Stach and  W.  Brosike: Sol. Energy, 76 (2004), 339.
• 10)   C.  Corgnale,  B.  Hardy,  T.  Motyka and  R.  Zidan: Int. J. Hydrog. Energy, 41 (2016), 20217.
• 11)   M.  Paskevicius,  D. A.  Sheppard,  K.  Williamson and  C. E.  Buckley: Energy, 88 (2015), 469.
• 12)   C.  Agrafiotis,  M.  Roeb,  M.  Schmucker and  C.  Sattler: Sol. Energy, 102 (2014), 189.
• 13)   J.  Yan and  C. Y.  Zhao: Appl. Energy, 175 (2016), 277.
• 14)   L.  Andre and  S.  Abanades: J. Energy Storage, 13 (2017), 193.
• 15)   Y.  Kato,  Y.  Watanabe and  Y.  Yoshizawa: Proc. 31st Intersociety Energy Conversion Engineering Conf. (IECEC), IEEE, Piscataway, NJ, (1996), 763.
• 16)   J. E.  Miller,  M. D.  Allendorf,  R. B.  Diver,  L. R.  Evans,  N. P.  Siegel and  J. N.  Stuecker: J. Mater. Sci., 43 (2008), 4714.
• 17)   A. J.  Carrillo,  D.  Sastre,  D. P.  Serrano,  P.  Pizarro and  J. M.  Coronado: Phys. Chem. Chem. Phys., 18 (2016), 8039.
• 18)   H.  Takasu,  J.  Ryu and  Y.  Kato: Appl. Energy, 193 (2017), 74.
• 19)   M.  Seggiani,  M.  Puccini and  S.  Vitolo: Int. J. Greenh. Gas Control, 17 (2013), 25.
• 20)   K. E.  N’Tsoukpoe,  H.  Liu,  N.  Le Pierres and  L.  Luo: Renew. Sustain. Energy Rev., 13 (2009), 2385.