Progress in Sorption Thermal Energy Storage

Abstract

There are various ways for thermal energy storage, such as sensible, latent, sorption, and chemical reaction. Sensible thermal energy storage and latent thermal energy storage are already in use. However, the drawbacks of bulk size (small energy storage density) and the strict requirement for thermal insulation have hindered their wide applications. Sorption and thermochemical reactions used for thermal energy storage have been considered as a future great potential product for thermal energy storage of solar energy, waste heat. or even electric heating, etc. The market thus needs such a “thermal battery,” which should be with a variety of kWhs capacities. Several key challenges remain in the way of the development of an efficient sorption thermal battery: sorption materials with high storage density and low cost, sorption bed with good heat and mass transfer to ensure charging power and discharging power, being stable after repeated cycles, minimum heat capacity ratio between the inert materials to the sorption thermal energy; control of the output temperature and power to meet the use demand. In this chapter, recent progress in sorption thermal energy storage, including materials, systems, and demonstrations, were described. The detailed future researches and developing maps were also discussed.

1 Introduction

A major concern for the full utilization of renewable energies like solar energy is overcoming their intermittence nature by decreasing the mismatch between the supply side and the demand side. Hence, TES (Thermal Energy Storage) plays an important role in the process of converting renewable energy by which the excess heat can be stored during peak periods and later released for heating during those rainy, cloudy days, and nights.

According to the type of heat used in the storage process, methods for TES can be mainly divided into three groups: sensible method by the temperature difference of the storage medium, latent method by the phase change process of a material at a constant temperature, and thermochemical method by the chemical bond between sorbent and sorbate [1, 2]. Thermochemical method could be further divided into chemical reactions and sorption processes, based on the binding forces and past conventions.

Sensible and latent thermal storage are the most studied technologies in recent decades. Most currently available TES systems used for space heating or cooling of buildings rely on sensible energy storage using tanks of chilled water for cooling or ceramic bricks or concrete for storing heat, or latent energy storage using ice or paraffin wax phase change materials. The traditional heat storage technologies suffer major disadvantages, including the inability to be downsized economically, low energy densities, and for consideration in seasonal storage, unacceptable energy losses. This means that current seasonal TES systems require too large a unit size and are ultimately too expensive to be accommodated within a single family residence.

Compared with the sensible and latent TES methods, the technology of using sorption processes in TES systems is relatively “new,” which started in the late 1970s and became a hot research topic since the beginning of this century. Sorption processes reversibly store thermal energy in molecular bonds in the form of chemical potential. Some distinct advantages can be provided with the reversibly sorption processes [3].

Substantially high energy storage density, heat of sorption, which is a function of sorption quantity of sorbate, is higher than heat of condensation/evaporation, leading to a high energy density in theory. Comparison of energy storage densities of different TES technologies shows that thermochemical processes, including sorption and chemical reaction, offer the highest storage densities, 3–30 times that obtained by sensible method [1].

Being able to preserve energy for long periods with limited heat losses, a sorption process will not occur until the sorbent contacts with the sorbate, meaning that it stores energy as chemical potential. The potential does not require thermal insulation and would not degrade with time, yielding a great opportunity for using sorption processes in long-term TES application.

Combining cold storage and heat storage functions in one system, the chemical potential stored in the sorption systems could be transformed to heating effect via the sorption reaction in the reactor or to cooling effect via the evaporation process in the evaporator, exhibiting more flexibility than traditional sensible and latent TES systems, which only serve for single storage purpose.

In this review, starting from the fundamental issues, various aspects related to sorption TES technology were presented, including some basic storage concepts, material selection for diverse sorbents in varied temperature ranges, and key issues concerning about the design of reactors in solid adsorption systems. Emphasis was put on past and present demonstration projects for both short-term and long-term applications. These projects serve as important references for our future endeavors. The detailed future researches and development hot spots were also discussed.

2 Sorption TES Concept

2.1 Storage Principle

The basic storage principle of which could be illustrated by the following equation:

$$ A \cdot (\text{m} + \text{n}){\text{B}} + \text{Heat}\underset {} \longleftrightarrow \text{A} \cdot \text{mB} + \text{nB} $$

(1)

where A and B represent the sorbent and sorbate respectively. A·(m + n)B and A·mB denote compound materials when (m + n) mole B and m mole are adsorbed or absorbed by A. The enrichment or dilution process of B on/in A is accompanied with a large amount of heat. During the desorption (charging) process, when heat is added to A·(m + n)B, the binding force between A and B is broken and a part of B is released from A; during the exothermal sorption (discharging) process, A·mB contacts with B to form A·(m + n)B again and the chemical potential is transferred into thermal energy.

Akin to an electric battery, a TES system, which is referred to as “thermal battery” [4], include two operation procedures: charging stage to take energy in and discharging stage to release energy out. For a sorption TES system, the charging stage consists of a desorption reaction in a reactor and a condensation process in a condenser, as shown in Fig. 1a. The paths of a Sorption TES cycle in the Clausius–Clapeyron diagram are given in Fig. 1b. The charging stage is initiated from point 1 at the ambient temperature Tamb with a maximum sorbate uptake xmax. When heat is added to the reactor, the temperature of the reactor begins to rise, following an isosteric heating process 1–2 until the pressure of the reactor equals to the pressure of the condenser (Pc). Then the value connecting the reactor and the condenser are opened and the vapor (sorbate) which clings to the sorbent starts to escape from the binding force and condenses into liquid in a condenser at a lower temperature level. The heat of condensation is released to the environment—air source, ground source, or water source. Ground and water sources may be the preferable options for its state temperature conditions, which is lower than air temperature in summer and higher than air temperature in winter. After an isobaric desorption process 2–3, the minimum sorbate uptake xmin is achieved at the highest desorption temperature of Tdes.

Fig. 1

Principle of sorption TES system: a operation procedures; b storage cycle in the Clausius–Clapeyron diagram

The discharging stage works in a reverse direction: it includes a sorption reaction in the reactor and an evaporation process in the evaporator. Between the charging stage and discharging stage, there is an inactive storage stage, period of which depends on the application of the TES systems. For the application in short-term TES, the whole cycle is completed in several hours, so the system temperature at the beginning of the discharging state may still remain high. However, for the application in long-term TES, without perfect thermal insulation, the system temperature will fall back to the ambient temperature (point 4 in Fig. 1b), meaning that extra heat is needed to increase the reactor temperature to the required temperature and pressure levels.

The above discussed is a typical sorption thermal storage principle, in which a sorbent-sorbent working pair is used and a condenser/evaporator is needed, however, there is another type of sorption thermal storage– resorption thermal storage, in which condenser/evaporator is replaced by another sorption bed. Cooling effect is produced by the low-temperature sorption bed when a resorption process occurs between the high-temperature sorption bed and low-temperature sorption bed. For detailed information about this resorption thermal storage process, please refer to the work of Li et al. [5].

2.2 Classification

Based on the storage mechanisms and established customs, the discussed scope related to sorption processes is suggested to be divided into the following three groups, namely liquid/gas absorption, solid/gas adsorption, and solid/gas sorption reaction, as presented in Fig. 2.

Fig. 2

Classification of sorption processes [3]

Liquid/gas absorption. The most familiar liquid absorption working pairs encompass LiBr/H₂O and H₂O/Ammonia, which have been successfully applied in commercial absorption chillers and heat pumps. Conventionally, liquid absorption is referred to as two-phase absorption. Another innovative kind of absorption thermal storage process, using lithium chloride crystals to increase energy density, is called “three-phase absorption” [6, 7].

Solid/gas adsorption. Silica gel and zeolite are widely studied and applied as common adsorbents using water as working fluid. Some new classes of materials, including AlPOs (aluminophosphates), SAPOs (silico-aluminophosphates), and MOFs (metal organic frameworks), have recently emerged as promising porous materials for heat storage.

Solid/gas sorption reactions or chemical adsorption. This type of sorption processes is often referred to as “solid/gas thermochemical sorption reaction/process” [5, 8–10] or “Chemical adsorption” [11, 12]. Here it is simply called sorption reactions. Sorption reactions mainly consist of two kinds of reactions, namely reactions of ammoniate with ammonia and salt hydrate with water. Strictly speaking, both the two reactions can be regarded as coordination reactions—molecules of ammonia or water vapor are attracted by metal ions to form coordinate bonds. Unlike absorption and adsorption, solid/gas sorption reaction is monovariant and induces volume modification of the solid. They have been widely researched for TES systems due to the high storage densities and multiple choices of salts with varied temperature requirement.

In addition to the above three groups, another group of sorption materials, composite sorbents, called CSPM (composite “salt porous matrix”) [13–16] has been widely adopted for sorption cooling, heat pump, and energy storage applications. Composite sorbents consists of two or more elements: one element is a porous matrix, such as silica gel, zeolite, expanded vermiculite, aerogel, etc., which are used for improving heat or mass transfer properties; others are inorganic salt, such as LiCl, CaCl₂, MgCl₂, MgSO_4, etc., which are responsible for most of the sorbate uptake quantities. The sorption mechanism of composite materials is just combination of the three basic processes.

2.3 Comparison of Sorption Refrigeration and TES

Sorption processes have been extensively researched for use in refrigeration systems since the end of 1970s. A large number of small-scale experimental prototypes have been built and tested in labs around the world. Right now, various new sorption chillers with small-and medium-scale refrigeration capacities have passed over from the prototype stage to serial production. As the market demand for solar cooling is increasing, a rising amount of commercial products can be expected. Compared with sorption refrigeration systems, the development of sorption TES systems are still in their infancy.

A detailed comparison between sorption refrigeration and sorption TES systems for a residential house are given in Table 1. Essentially, the basic principle for sorption refrigeration/heat pumps and sorption TES systems are the same, i.e., using reversible sorption process between the sorbent and sorbate to produce cooling or heating effect. However, different application areas give rise to different requirements for system configurations.

Table 1 Comparison between sorption refrigeration and sorption TES systems for a residential house

COP and SCP (Specific Cooling Power) are two important parameters for evaluating the performance of sorption refrigeration systems. COP can be improved by means of using heat and mass recovery processes and advanced cycles, while SCP is closely related with the design of the sorption reactor. Advanced sorption materials, expanded heat transfer area and short heat and mass transfer distance are common measures to increase the output power. With a good design of the reactor, it is possible to realize a higher power output with less mass of sorbent. Recent trends applied in some adsorption machines are the technology of coating the heat exchanger surface with adsorbent to obtain an high SCP. In this case, the thermal capacity ratio between the adsorbent and the auxiliary metal components is high, leading to a high SCP but a low COP. This principle of design may be suitable for adsorption refrigeration systems but it is inapplicable for adsorption TES systems, as the prime evaluation factors for sorption TES systems are HSD in Wh/kg or kWh/m³ and heat storage efficiency, not the cooling power. Storage materials with high HSDs are prerequisite to realize compact and efficient sorption thermal storage applications. With a given material, the mass of sorbent required for a specific storage capacity is defined, which cannot be reduced even with better design of the reactor.

Materials with low desorption temperatures and low costs are always attractive for sorption systems. However, different application objectives result in different requirements for sorption materials in some aspects. As cooling effect is created by the evaporation process of the sorbate, higher heat of evaporation and lower heat of sorption lead to higher COP for sorption refrigeration systems. On the contrary, higher heat of sorption will facilitate the HSD on the ground that heat is supplied by the desorption process. As mentioned before, the sorption TES system could be considered for heat storage and cold storage at the same time, and then a contradictory requirement for heat of sorption is confronted. As a rule, cold storage capacity is inferior to heat storage capacity under the same working conditions. The requirement for materials should be chosen according to the alternative emphasis of a specific system, i.e., either heat storage or cold storage.

3 Material Selection Criteria

It is a common knowledge that energy storage performance of a sorption TES system, to a great extent, is determined by storage materials. As mentioned before, basic principles for selecting appropriate sorption working pairs for TES are similar to those for sorption refrigeration, which are listed in detail as follows [1, 3, 17, 18]:

• High energy storage density (Wh/kg or kWh/m³).

• Low charging temperature.

• Appropriate heat and mass transfer properties—to ensure designed output power.

• Easy to handle—nonpoisonous and no corrosion.

• Low cost—low price per kWh heat energy stored.

• Thermal stability—no deterioration after many charging-discharging cycles.

• Mechanical stability.

High energy storage density is a prerequisite requirement for a suitable sorption working pair for TES, which implies more particular claims including high sorbate uptake change per cycle and high heat of sorption (for heating). According to the theory given by Aristov [16], solid sorbents with S-shape isotherms or isobars (like AlPOs and SAPOs), which feature a sharp rise in sorption quantity between the charging and discharging conditions, are optimal materials for heat transformation applications.

Selections of sorbates are confined in a small range, among which water and ammonia are the most commonly used choices. Water offers the advantages of environmentally friendly properties and high evaporation heat. For open systems, water is the only choice of sorbate as open systems are connected with ambient environment. As water systems operate in vacuum pressures, sorption systems with water often face a critical mass transfer problem. Another challenge for water systems is that they cannot work when the environment temperature is lower than 0 °C. In that case, methanol and ammonia should be considered. In contrast to water, ammonia systems hold positive pressures, leading to a better mass transfer situation.

Unlike other sensible TES materials and PCMs, heat storage densities of sorption working pairs are influenced by many factors, including charging temperature and condensation temperature, discharging temperature and evaporation temperature. This fact brings the difficulty to compare the heat storage performance of all kinds of sorption couples. However, for a given working pair, its TES potential is largely determined by the charging temperature, which means that the relations between HSD and charging temperature is the most significant information for reference.

4 TES Based on Absorption

4.1 Working Pairs

Frequently used absorption working pairs include LiBr/H₂O, LiCl/H₂O, CaCl₂/H₂O, H₂O/NH_3, and NaOH/H₂O. The former three could be widely used in both open and closed systems and the latter two could only be used in open systems. Except for H₂O/NH₃, which uses ammonia as the sorbate, other working pairs are combinations of solutions of hygroscopic salt with water. At the same temperature and pressure, LiCl solution has a lower mass concentration of salt than LiBr and CaCl₂ solutions, meaning that more water uptake per unit mass of salt is available for LiCl solution. At a given pressure, LiBr/H₂O has the highest crystallization temperature. The crystallization points at a condensation pressure of 5.63 kPa are 86.1 °C for LiCl/H₂O, 72.2 °C for CaCl₂/H₂O and 103.4 °C for LiBr/H₂O, respectively. Essentially, absorption technology owes its storage ability to the concentration difference of the sorbate attracted to the sorbent-sorbate solution between absorption and regeneration processes. Thus, larger concentration differences promise greater heat storage densities. To avoid crystallization problem at high concentrations, especially for a single effect LiBr/H₂O liquid absorption process, the concentration difference should be maintained below 10 %, which is too narrow to yield competitive storage density for TES systems.

4.2 Systems

The development of absorption refrigeration dates back to the early 1900s, when ammonia/water refrigeration was an applicable technology [19]. After several decades of research effort, nowadays absorption refrigeration has become a proven and mature thermally activated cooling technology. The widespread uses of absorption machines is due to its low electric input requirement, smaller physical dimensions than adsorption machines for the same capacity, great flexibility caused by the fluidity of solutions. However, when researchers try to employ absorption for the application in TES, many new challenges have been encountered in some projects.

In spite of its high corrosive property, the merits of low cost and high selectivity for water make NaOH an attractive option for energy storage. In EMPA, Weber and Dorer [18] developed a single-stage closed absorption prototype in 2008 for long-term heat storage using NaOH/H₂O. This experimental setup mainly consists of a reactor and three storage tanks for water, strong solution, and weak solution respectively. Two heat exchangers—one for water and one for NaOH solution—are both arranged in the reactor, with a radiation protection to decrease direct heat losses between the heat and cold heat exchangers. The single-stage NaOH prototype requires a charging temperature of 150 °C if taking all kinds of heat losses into account. To reach higher concentrations at lower temperature levels in the desorber, a double stage cycle was suggested. The second stage was constructed and directly connected to the original first stage to obtain a double stage system [20]. The drawbacks of the double stage cycle included a low discharging rate, crystallization of NaOH solution and overheating of the ground heat exchanger.

Liu et al. [21] proposed a long-term absorption energy storage cycle for house heating, as shown in Fig. 3. This cycle allows the solution to reach the crystallization point in the storage tank. The evaluation study illustrated that the appearance of salt crystal in the storage tank significantly increases the HSD and LiCl/H₂O couple possesses excellent storage capacity and efficiency. Whereas, a note must be made concerning this cycle that the charging temperature is still restrained as crystallization is only permitted in the storage tank, rather than in the reactor/desorber during the desorption process. Based on the proposed long-term absorption storage cycle, a demonstrative LiBr/H₂O prototype with a heat storage capacity of 8 kWh and a discharging rate of 1 kW has recently been set up in LOCIE-CNRS [22]. The obtained charging performance seemed to be fine but the discharging process was verified a failure due to an inadequate design of the absorber.

Fig. 3

Long-term LiBr/H₂Oabsorption storage cycle [22]

Recently, Li et al. [23] established a LiBr/H₂O absorption energy storage device with a highest power output of 30 kW, which was designed for improving the performance of a CCHP (Combined Cooling, Heating, and Power generation) system by storing excess heat in LiBr solution. To make the device compact and reliable, the generator/absorber and solution storage tanks are integrated into one vessel, meaning that the discharged solution is just collected under the heat exchanger. Heat transfer performance in the generator/absorber and evaporator/condenser is guaranteed by spray-type heat exchangers. Energy storage performance of the device were carried out under three different conditions, namely 7 °C chilled water output for cooling in summer, 60 °C hot water output for heating in winter and 45 °C hot water output for domestic hot water throughout the year. Energy storage densities (based on the total volume of LiBr solution and liquid water) of 31, 42, 22 kWh/m³, and storage efficiencies of 0.5, 0.6, 0.4 were obtained under the three working conditions.

In conventional absorption systems, the process of crystallization is perceived as a major barrier to further expand the concentration difference between the strong solution and the weak solution. Lourdudoss and Stymne [24] presented an idea of using crystals of hydrophilic salts to enhance storage density. To distinguish the cycle from common two-phase absorption cycles, it is named “three-phase absorption cycle.” A major difference between this three-phase cycle and the long-term storage cycle is that the former allows existence of crystals in the charging process but the latter only permits crystals in the storage tank. The principle of the three-phase absorption thermal storage system is shown in Fig. 4. During the charging process, the weak solution stored at the bottom of the reactor is pumped and sprayed over the heat exchanger. The solution is heated until it achieves the saturation point at the charging temperature. As the water vapor is released from the solution, solid crystal is formed continuously, flushed to the bottom of the vessel and stopped by a sieve. Three phases, including vapor, liquid solution, and solid crystal coexist simultaneously, so a monovariant process is ensured.

Fig. 4

Working principle of so-called “three-phase” absorption cycle [7]

Based on the three-phase absorption cycle, a Swedish company named Climate-Well [7, 25–28] has developed several generations of prototypes which target the Mediterranean heating and cooling market. Test of the original prototype named Thermo-Chemical Accumulator was accomplished in the framework of IEA-SHC Task 32 [7, 29]. The claimed energy density for LiCl salt was equivalent to 253 kWh/m³, 3.6 times that of water (25–85 °C).

To reduce the volume of separate storage tanks for water, strong and weak solutions, Quinnell et al. [30] presented a novel concept of using a single storage vessel for storing liquid calcium chloride in a closed liquid absorption system. The system schematic is presented in Fig. 5. During the charging process in summer, diluted CaCl₂ solution is circulated to the solar collector where it boils at 117–138 °C. The concentrated CaCl₂ solution and condensed water are stored until the winter. When the collector temperature is not high enough—below the boiling point, the stored solution can be heated to provide sensible energy storage; when the required temperature is satisfied, the solution can be discharged directly. To prevent mixing of solutions in single storage vessel, a heat exchanger, and a stratification manifold are immersed in the storage tank to thermally stratify the solution and minimize mixing between regions of different CaCl₂ mass fractions. First, the feasibility of the proposed concept was demonstrated via a computational fluid dynamic study of heat and mass transfer in the system [30–32]. A 1500-liter storage tank was built to elucidate mixing of solutions [31, 33]. Results suggest that this concept of absorption storage can be preserved for 160–902 days, which are adequate for maintaining the stratification of mass fraction for seasonal storage.

Fig. 5

Closed-cycle CaCl₂/H₂O absorption heating system [33]

5 TES Based on Adsorption

5.1 Working Pairs

Development in the porous material science offers a diverse variety of adsorption materials on the market, and some of them have been evaluated in TES processes. Silica gel is the most widely used desiccant because of its low cost and low desorption temperature. Hot water lower than 100 C provided by common solar collectors is able to charge silica gel sufficiently in the desorption process. Silica gels with narrow and large pores are the most used types. Commercially available zeolites mainly include types 4A, 5A, 10X, 13X, and Y with varied pore sizes. FAM-Z02 (Functional Adsorption Material Type Z02) is a zeolite-based molecular sieve (CHA-type silico-aluminophosphate) recently developed, which is also reported as SAPO-34 [34, 35]. Water adsorption isotherms (40 °C) for above-mentioned adsorption materials have been presented in the scope of IEA-SHC Task 32 [36]. As shown in Fig. 6, water sorption on zeolites could be almost finished at a low relative pressure below 0.1, indicating that desorption of zeolites requires more efforts including high desorption temperatures or extremely low relative pressures. The final water uptake amounts of silica gels are not low but the water adsorption mainly occurs at too high relative pressures, about 0.5 for silica gel with narrow pore and 0.9 for silica gel with wide pore. This fact leads to a low water exchange within a typical cycle, especially for closed systems. FAM-Z02 undergoes the S-shape transition between hydrophobic and hydrophilic behavior which causes exchange of large amount of water (up to 0.2 g/g) within narrow relative pressure ranges.

Fig. 6

Water adsorption isotherms (40 °C) for a selection of adsorption materials [36]. zeolite Na5A: zeolite 5A; zeolite SC Y 1/16: zeolite Y; silica gel 127B: silica gel with narrow pore; silica gel LE-32: silica gel with wide pore; SWS-1L: mesoporous silica gel impregnated with CaCl₂; FAM-Z02: a type of zeolite-based molecular sieve

5.2 Systems

Similar to the evolution trend in adsorption heat pumps, research interest in adsorption systems for TES started from silica gel/H₂O, but then steered to zeolite/H₂O due to its high storage density. Though novel materials such as AlPOs, SAPOs, and MOFs show much greater storage potential that conventional silica gel and zeolite (13X, 4A, 5A, Y), their application in practical TES prototypes could not be found right now. That may be due to the higher costs in large-scale TES systems.

Within the framework of the EU-project MODESTORE [37, 38], a prototype storage module was developed using silica gel/H₂O. The silica gel used in this prototype was microporous silica gel Grace 127B. As shown in Fig. 7, a spiral heat exchanger containing approximately 200 kg of silica gel is connected to an evaporator/condenser at the bottom of the vessel. Operating between a water content interval 3 and 13 %, a storage capacity of the lab scale unit was only 13 kWh—significantly lower than initially expected. Thus, the author concluded that silica gel was not suitable for long-term sorption thermal storage and the task of future projects was to find other advanced materials to fulfill the requirements of sorption systems for storage purpose.

Fig. 7

Schematic of Silica gel/H₂O TES prototype in the MODESTORE project

In Shanghai Jiao Tong University, Lu et al. [39] presented a closed adsorption cold storage system with a zeolite 13X/water working pair, which is designed for locomotive air conditioning. This system was driven by exhaust gases from a locomotive. This prototype compasses only one adsorber and a cold storage tank. The averaged cooling power was about 4.1 kW. The total experimental capacity of the cold storage was 5.5 kWh when the temperature of adsorption bed reached its maximum value of 125 °C. After being tested and optimized in lab, the system was actually installed in a locomotive, which ran in the Zhejiang Province, to refrigerate the driver’s cabin. The operating performance proved to be stable and reliable.

A large-scale open adsorption thermal storage system employing zeolite 13X/H₂O, was installed in Munich by ZAE Bayern [40, 41] to heat a school building in winter and to cool a jazz club in summer. The school (with a heating load of 130 kW) and the club (with a cooling load of 50 kW) are connected to the local district heating system of Munich. Figure 8a, b shows heat flux during charging and discharging modes. At night, zeolite is charged by the hot steam from the district heating system and heated to about 130–180 °C. During the day, the zeolite bed is discharged in times of peak power demand. An additional cold recovery device, consisting of an exhaust air humidifier with an integrated heat exchanger and the supply air heat exchanger, are required in the cooling process for the jazz club. The obtained storage densities were 124 kWh/m³ for heating and 100 kWh/m³ for cooling with COP of 0.9 and 0.86, respectively.

Fig. 8

Open adsorption thermal storage system on district heating net in Munich: a charge mode for heating; b discharge mode for heating [41]

The Monosorp project [29], which was involved in the scope of IEA-SHC Task 32 from 2003 to 2007, was about developing a Zeolite 4A/water sorption process integrated in a mechanical ventilation system for solar heating for a single family. In this concept, regeneration of zeolite proceeds in summer when excess solar heat is available for desorption. During winter when solar heat is insufficient, room exhaust air is directed into the sorption bed and the sorption heat released by the sorption between zeolite and water results in a temperature rise of the air flow. Zeolite honeycomb monoliths were developed to get a good kinetic performance and low pressure drop of the air flow. A small-scale pilot plant was built up and measured energy density is about 130 kWh/m³. Based on the results of the Monosorp project, a new project named Solspace was initiated in ITW in 2012 [42, 43], which aimed at developing an open sorption TES in several segments which can be desorbed or adsorbed separately. The research goal is to cover over 90 % of heat demand for space heating and 80 % of heat demand for hot water with this concept.

Recently, adsorption thermal storage has been researched in some new application areas such as electric vehicles and dishwashers. Climate-control systems in today’s electric vehicles consume as much as 30 % of the power provided by the electric batteries. To lessen the power drain of heating and cooling systems, Narayanan et al. [4] from Massachusetts Institute of Technology (MIT) presented an idea of using a thermal battery based on adsorption to deliver both heating and cooling by storing thermal energy. The design of the thermal battery is shown in Fig. 9. The thermal battery was designed to be compact (<30 l) and could provide heating and cooling capacities over 2.5 kWh. If successful, the technology could potentially extend electrical vehicles driving range by 30–40 %. Hauer and Fischer [44] from ZAE Bayern has tried to integrate an open adsorption process into a dishwasher to reduce its energy consumption. The adsorbent is zeolite 13X. The water heating phase of the main washing cycle has been used to generate a packed bed of zeolites. The conventional water heating phase before drying of dishes is replaced by an adsorption phase in which the dished are dried by hot air. The experimental setup is presented in Fig. 10. Their experimental results showed that the energy consumption compared to a conventional dishwasher was reduced from 1.06 down to 0.80 kWh per cycle, leading to energy savings of 24 %.

Fig. 9

Thermal battery for electric vehicles [4]

Fig. 10

Energy-saving dishwasher with open adsorption thermal storage [75]

6 TES Based on Sorption Reaction and Composite Materials

6.1 Working Pairs

Due to the strong chemical bond between the sorbent and the sorbate, thermochemical sorption reactions are capable of providing much greater storage densities at lower temperatures than zeolite/water. Among the large body of inorganic salts, Na₂S/H₂O [45–47], SrBr₂/H₂O [9, 10, 48, 49], MgCl₂/H₂O [50, 51], MgSO₄/H₂O [52] are attractive choices for researchers. A severe problem for these hygroscopic salts is forming of saturated solution at high relative humidity. This process is called deliquescence, which is an important solid–water interaction phenomenon. Deliquescence is defined as a first-order phase transformation of the solid to a saturated solution when the RH (Relative Humidity) reaches a certain threshold value, namely, the DRH (Deliquescence Relative Humidity) [53]. The value of DRH depends on the properties of the salt and the temperatures. The DRH values for LiCl and LiBr are only 11.3 and 6.2 % at 30 °C [54], implying that it is more likely for these solids to form solutions directly. Although the DRH of MgCl₂ (32.4 % at 30 °C) is higher than LiBr and LiCl, deliquescence is still possible at some situations, spastically at open cycles. Therefore, treating of MgCl₂ in sorption storage systems needs special care. With a DRH higher than 90 % at 30 °C, MgSO₄ is considered to be a suitable salt which is inclined to have hydration reactions with water vapor.

To overcome the deliquescence problem, composite materials are usually proposed by impregnating salt in the pore space of a porous matrix. The porous structure is conductive to water-salt sorption kinetics with the high specific surface area and is able to contain some amount of liquid solution with its internal pore volume. Various combinations of salts with porous matrices have been raised up, such as silica gel–CaCl₂ [55, 56], silica gel–LiCl [57, 58], zeolite 13X–CaCl₂ [59], zeolite 13X–MgSO₄ [60], zeolite 13X–MgCl₂ [61]. In terms of the porous matrix, adding of hygroscopic salts greatly increases the water uptake quantities; in terms of the salts, dispersing of the salt particles in porous matrix could improve the sorption kinetics.

The water sorption and desorption behaviors for three commonly used hygroscopic salts, namely LiCl, CaCl_2, and LiBr, have been analyzed by the authors [62] from Shanghai Jiao Tong University, and results showed that LiCl possessed a great potential for sorption TES. To find a proper porous matrix for LiCl, we have investigated and compared the sorption properties of silica gel–LiCl composite [58] and activated carbon–LiCl composite [63] in the previous research and activated carbon was regarded as a better choice. Then, a new type of consolidated composite sorbent was developed using activated carbon (AC) as a porous host matrix to carry LiCl, mixing with expanded natural graphite treated with sulfuric acid (ENG-TSA) to increase heat transfer and adding silica solution (SS) as a binder to enhance mechanical strength [63]. Consolidated sorbents with ENG-TSA not only help to enhance the thermal conductivity but also to increase the total macro pore volume to contain more liquid LiCl solution.

Sorption reactions between ammoniate and ammonia offer such wide choices of sorption working pairs at varied reaction temperatures and pressure levels, so it is possible to make use of thermal energy efficiently by selecting appropriate sorption working pairs according to the external heat sources at different working temperatures. Table 2 shows the reaction parameters and energy densities of some typical solid-gas thermochemical sorption working pairs at different operating temperatures. The working equilibrium temperature is calculated on the assumption that the heat sink temperature is 30 °C. For most reactive salt for the integrated energy storage and energy upgrade based on solid-gas thermochemical sorption process, the energy density is higher than 300 Wh/kg, which is about 2–5 times that obtained with latent heat storage method using PCMs. Therefore, the higher energy density is beneficial to reduce the volume size of energy storage device.

Table 2 Energy storage densities of some typical solid-gas thermochemical sorption working pairs at different operating temperatures [5]

6.2 Systems

In the SWEAT (Salt-water Energy Accumulation and Transformation) project, Boer et al. [46] developed a prototype of a modular sorption cooling system using Na₂S/H₂O. The scheme of the prototype is illustrated in Fig. 11. The shell and tube design of the module consists of sorbent-filled copper-wire-fin tubular heat exchangers, a condenser and an evaporator coil. The vessel of the accumulator/reactor is connected through a vapor channel with another vessel of the evaporator/condenser. There are six Sipro-tubes in hexagonal packing and one diffuser tube located in the accumulator. Test results showed that a cold storage capacity of 2.1 kWh and a cooling COP of 0.56 were achieved by inputting a heat capacity of 3.7 kWh with 3 kg of Na₂S in one module.

Fig. 11

The SWEAT Na₂S/H₂O storage module [46]

Mauran et al. [10] in PROMES-CNRS set up a storage prototype using the reversible reaction between SrBr₂·6H₂O and SrBr₂·H₂O. The solid/gas reactor, which integrates an evaporator/condenser, has a volume of 1 m³. It was designed at a scale of a heat storage capacity of 60 kWh at 35 °C and a cold storage capacity of 40 kWh at 18 °C. The prototype has a modular structure in which thin layers of reactive composite (12 mm) are laid out in alternation with plates of a heat exchanger and steam diffuser that are also thin (5–6 mm). Unexpected discouraging results were obtained, owing to the low heat transfer at the interface between the reactive layer and the heat exchanger wall. Later, they switched their research direction from the closed TES systems to open TES systems, still based on the solid/gas sorption reaction between SrBr₂ and water. A small-scale experimental unit was built in PROMES-CNRS to characterize the mass transfer and heat storage potential of open porous reaction bed for long-term storage of solar energy using SrBr₂/H₂O [64]. Kinetic experimental results showed that energy storage densities about 430–460 kWh/m³ are available with SrBr₂/H₂O in an open system. But the obtained specific output powers are quite low, only in the range of 1.93–2.88 W/kg.

A solar air conditioning pilot plant was installed in PROMES laboratory in 2006 [8]. This pilot, with a daily cooling capacity of 20 kWh, consists of four subsystems: a solar heating loop, a thermochemical unit using the working pair BaCl₂/NH₃, a ground cooling loop and a chilled water production loop, as shown in Fig. 12. The thermochemical process is powered at 60–70 °C by 20 m² of flat plate solar collectors. The reactive sorbent, which consists of a compressed mixture of 140 kg of anhydrous BaCl₂ and 35 kg of expanded natural graphite, is loaded into a set of 19 tubes in the thermochemical sorption unit. The hot PCM storage enables the storage of excess solar heat if the desorption reaction in the reactor is completed and the cold PCM storage makes a supply of cooling effect possible when the sorption reaction is not available. Two years of experimental operation illustrated an average yearly efficiency of solar collectors and a COP ranging from 0.4–0.5 to 0.3–0.4 respectively. A daily cold storage capacity is about 0.8–1.2 kWh per m² plate solar collector at 4 °C cooling output.

Fig. 12

View of the solar sorption pilot plant for air conditioning, flat plate solar collectors, and the thermochemical reactor design [8]

In many industrial processes, there exists a large amount of low-grade heat resources which is directly released to the ambient environment because the low-temperature levels cannot meet the high-temperature requirements. Meanwhile, sometimes the conventional heat pumps cannot provide enough high-temperature lift for space heating when the outside temperature is low in winter. These situations introduce a practical requirement for energy upgrade by developing advanced heat transformer technologies. Recently Li et al. [5] proposed a new concept named target-oriented sorption heat transformer for integrated energy storage and energy upgrade. The working temperature of low-grade thermal energy can be upgraded from 87 to 171 °C using a group of sorption working pair MnCl₂-CaCl₂-NH₃, and it would be upgraded from 130 to 282 °C when a group of sorption working pair NiCl₂-SrBr₂-NH₃ is employed. Li et al. [65] also proposed another concept called a dual-mode heat transformer for seasonal storage of solar energy with little heat losses. The proposed system is able to operate in two modes according to practical working conditions. Except the conventional sorption TES mode for heating, another energy upgrade mode is enabled to upgrade the discharging temperature level when the ambient temperature is low, by introducing a heat recovery process between the low-temperature reactor and high-temperature evaporator.

ITW has recently proposed a new design, namely CWT-NT concept [42, 66, 67], for a long-term thermochemical energy storage integrated into a solar thermal combisystem for the composite material of zeolite and salt. A detailed description of the concept is presented in Fig. 13. The thermochemical storage operates as a low power heating system and is connected to the combistore of the solar thermal system. It is designed as an open system using ambient or exhaust room air for charging and discharging process. The separate reactor concept is realized, meaning that the sorption material is put into another vessel during storage process and is pumped to the reactor when the desorption/sorption process takes place. Thermal heat capacities and heat losses could be significantly reduced though this separate design. A recent progress of this project [67] was referred to a proposal of a new highly efficient regeneration process, reducing the regeneration temperature from 180 to 130 °C due to a decrease of humidity of air flow. Additional desiccants and dehumidification devices are needed to dry the air flow before it enters the reactor.

Fig. 13

Schematic of CWS-NT storage concept [42]

The sorption TES activity at ECN focuses on developing open sorption system with MgCl₂/H₂O [51]. Four potentially interesting hydrates, namely MgSO₄·7H₂O, Al₂(SO₄)·18H₂O, MgCl₂·6H₂O and CaCl₂·6H₂O, were identified by Zondag et al. [51] experimentally and MgCl₂·6H₂O was recognized by them as the best option to be used in open reactors. An open reactor with a packed reaction bed with 17 L MgCl₂ was established and it could achieve a HSD of about 140 kWh/m³, with a discharging power of only 50 W.

Recently various kinds of composite sorption materials have been developed as an important means to improve the heat and mass transfer in an open system from the viewpoint of materials’ novel structure. Hongois et al. [60] prepared a zeolite/MgSO₄ composite for low-temperature heat storage applications. The optimum mass percentage of MgSO₄ in the composite was founded to be 15 %. An energy density of 166 kWh/m³ was obtained with an airflow rate of 8 L/min and inlet air RH of 50 % in an open cycle. The CaCl₂-silica gel composite has been extensively studied by Wu and Zhu [55, 68, 69] in an open setup with 40 kg sorbent pellets for TES. A stable energy storage density of 264 Wh/kg could be obtained from the sorption thermal storage experiments over multiple repeated sorption/desorption cycles [55]. Liu et al. [70] prepared a ceramic honeycomb filter with 0.28 mm-thickness cell wall for an open TES system, using a natural mesoporous Wakkanai siliceous shale (WSS) as the matrix to confine CaCl₂. The honeycomb structure is able to ensure a large contact area and low pressure loss, resulting in a good kinetic performance. For the WSS impregnated with 22.4 wt%, the volumetric heat rate was maintained at an average value of 10.5 kW/m³ when the outlet temperature was greater than 40 °C and the obtained heat storage densities were 708 Wh/kg in terms of mass and 328 kWh/m³ in terms of volume.

7 Conclusions and Outlooks

7.1 Summary of Current Sorption TES Technologies

In recent years, sorption technology, including liquid/gas absorption, solid/gas adsorption, and sorption reactions, with potentially high energy density and without significant heat losses over long periods, has been regarded as one of the promising approaches for heat storage of solar energy. Research activities devoted for the area of sorption TES keep increasing, with many novel concepts, materials, and demonstrations being presented. As illustrated by the aforementioned developments, the concept of sorption thermal storage is a realistic and sustainable technology for the future.

Various sorption working pairs have been proposed and tested in material level, and heat storage densities above 1000 Wh/kg and 600 kWh/m³ can be expected. Figure 14 summarizes some typical literature data for different sorption working pairs, to reveal the maxim potential of various working pairs. Energy densities of PCM and water are also provided for comparison. According to a theoretical study done by Liu et al. [21], required charging temperatures for most liquid absorption couples, including LiCl/H₂O, LiBr/H₂O, KOH/H₂O, NaOH/H₂O, CaCl₂/H₂O, are lower than 100 °C, among which LiCl/H₂O presents the best energy storage density (1219 Wh/kg)—even higher than the densities obtained from chemical reactions—with a charging temperature of only 66 °C.

Fig. 14

Heat storage densities of different sorption working pairs. The default sorbate is water and it is omitted, Details of the data about sorption working pairs can be found in Table 2 in Yu et al. [3]’s work. Details of the data about PCMs are sited from Farid et al. [76]. The reference temperature of sensible heat storage of water (the blue line) is 10 °C

For silica gels, though the required charging temperatures are lower than 100 °C, their storage densities are even lower than sensible heat method of water, making application prospect of silica gel/water in sorption storage obscure. Except for silica gel/water, heat storage densities of other sorption working pairs are superior to those of conventional sensible and latent materials. Zeolites (type 4A, 13X, Y) show varied and moderate storage densities (71–128 Wh/kg) with charging temperatures higher than 150 °C. Novel porous solids like AlPOs, SAPOs and MOFs have been proposed for TES systems recently. Out of the classes of AlPOs, SAPOs and MOFs, AlPO-18 (243 Wh/kg) [71], SAPO-34 (203 Wh/kg) [71] and MIL-101 (690 Wh/kg) [72] are regarded as the most attractive choices respectively. All kinds of thermochemical sorption reactions provide superior storage potentials within a wide range of temperatures. Their storage densities range from 408 Wh/kg (BaCl₂/NH₃) to 1067 Wh/kg (Na₂S/H₂O). The composite materials, such as the well-known SWS-1L, exhibit intermediate densities between the salt and the matrix (silica gel or zeolite), as expected.

Though many sorption materials show excellent energy storage performance in small-scale sample level, some critical problems have been met in practical setups. The current development is concluded as follows:

The advantages of absorption TES are good heat transfer performance, which could ensure sufficient thermal power output, and high calculated storage densities at low charging temperatures (<100 °C). Although absorption has been regarded as a proven and mature thermally activated cooling technology, applying absorption for TES systems have met great challenges. Number of demonstration projects is still limited, leading to a deficient understanding of absorption TES processes. In the aforementioned completed projects, compared with charging process, discharging process seems to be more difficult to live up to the designed expectations. Spray-type heat exchangers, which have been widely used in commercial absorption machines, are still the most reliable and efficient option for absorption systems. Some novel designs about trying to apply other types of [22] exchangers in the reactor seem fail to meet the designers’ expectations due to lack of design experience or manufacturing difficulties. The highest energy density was obtained from the three-phase Climate-Well absorption heat pump, amounting to 253 kWh/m³ [29], which is 3.6 times that of water (25–85 °C). Other prototypes could not achieve heat storage densities higher than 50 kWh/m³ in the tests.

Adsorption TES systems offer the benefits of no need for solution pumps or rectification devices, no corrosion to metal components and no worry about crystallization of salt solutions. Silica gel/H₂O and zeolite/H₂O are the most used working pairs in TES systems, but it seems that Silica gel/H₂O has disadvantages for the storage energy density, the possibility to improve it is the use of composite materials (Silica gel/LiCl for example) which boost the concentration changes on water for 2–3 times. A highest HSD of 130 kWh/m³ [29] was obtained with zeolite 4A/H₂O at a charging temperature above 170 °C in practical systems. For novel porous adsorption materials like AlPOs, SAPOs and MOFs, which show much greater storage densities ranging from 200 to 600 Wh/kg, their application in TES should be further investigated in practical systems. The costs of these materials may be of great concern for large-scale applications.

Solid/gas thermochemical sorption reaction has become the research hotspot for long-term seasonal TES owing to their high storage potential. Among all kinds of inorganic salts, Na₂S/H₂O, SrBr₂/H₂O, MgCl₂/H₂O, MgSO₄/H₂O, ammoniate/NH3 are the most attractive choices. Currently, the highest energy storage density up to 460 kWh/m³ [42] at a charging temperature of 80 °C has been observed for SrBr₂/H₂O in an open reactor. The main drawback for sorption reaction TES systems is low specific power (kW/kg), due to the poor mass transfer in sorption bed. Composite sorption materials have been adopted as an efficient means to improve the mass transfer yet it may decrease the value of volumetric energy storage density. Reactions between ammoniate and ammonia offer wide choices of sorption working pairs at varied reaction temperatures and pressure levels, so it is possible to make use of thermal energy efficiently by selecting appropriate reactive salts according to the external heat sources at different working temperatures. As Li et al. [65] proposed, sorption, resorption, and heat recovery processes between different reactive salts provides flexible operation modes, such as energy update or energy storage.

7.2 Future Developing Roadmaps

Based on the past research activities, some important concerns are outlined as the future developing roadmaps for highly efficient and compact sorption thermal storage technologies. General concerns for both liquid/gas absorption and solid/gas sorption TES systems are listed as follows:

Open or closed systems. For ammonia systems, closed type is the only option. While for water systems, both open and closed systems could be considered. Design of closed systems is more intricate than open ones with the need of considering both heat and mass transfer issues. Open systems hold more possibility to achieve the high storage density as it can operates in a wider relative pressure range (from nearly 0–0.95 or even 1) than closed systems (from 0.1 to 0.4, relative pressure in closed systems = vapor pressure/saturated vapor pressure at the sorption temperature). But both temperature and relative pressure conditions of ambient air will influence this possibility. So additional humidity control devices such as humidifiers are required for open systems. The benefits of closed systems are their impregnability of outside relative pressure and short cycle periods. It seems that many researchers are inclined to use open systems for long-term TES given the large system capacity.

Minimizing the heat capacity of inert parts. Heat capacity of inert materials strongly influences the practical performance of sorption systems. Large thermal masses of inert materials such as metal components bring unwanted heat losses during charging and discharging stages as extra heat are required to heat them up. Calculation and minimization of the heat capacity ratio between inert parts to sorption materials is a vital task in the design period.

Maintaining constant output temperature and controllable thermal power during discharging process is a practical requirement for TES systems. Owing to the variations of discharging temperature and sorbate concentration in/on sorbents, it is a tough task to keep a relative stable output temperature and power during discharging process. Using variable frequency pump to change the flow rate of heat transfer fluid may be a solution but it needs an in-depth study on control strategy. Adding a buffer vessel for heat transfer fluid is a frequently used and easy solution yet it could only partially solve the problem. Solid-gas thermochemical sorption TES usually has the advantages of high energy density and stable working temperature in comparison with liquid-gas absorption and solid-gas adsorption TES. Moreover, the working temperature can be easily controlled by changing the operating pressure due to the monovariant characteristic of chemical reaction in solid-gas thermochemical sorption TES system. Thus, it is beneficial to realize the close match between the energy supply and the energy demand because of stable temperature.

System costs should always be emphasized in the way of producing commercially competitive prototypes. Large-scale systems like long-term systems only need low-cost sorption materials, yet some small-scale systems can stand materials with a little higher price.

Sorption materials should be stable after many repeated TES cycles. The sorption bed should have no obvious performance deterioration after long-term uses. Cycle stability of the sorption material should be checked.

For liquid/gas absorption TES systems, the following additional issues should be addressed:

Usually, apart from one reactor and one evaporator/condenser, three separate storage tanks for strong and weak solutions and liquid water are needed in an absorption TES system, which consequently gives rise to a significant decrease of energy storage density considering the volume of the whole system. The fact also adds complexity in system design and control strategy. Integrated storage tank concept is proposed [31] but the challenge regarding mass stratification and mixing effect should be further investigated experimentally.

Crystallization remains a pivotal bottleneck for normal two-phase absorption systems, by imposing restrictions on charging temperature and needing peculiar measure to get rid of crystals in charging. Long-term systems are more likely to face crystallization as it operates in a large temperature and period span. The operational concentration differences among common absorption cycles are not sufficient to yield competitive high storage densities. Thus, some advanced cycles are proposed to allow the appearance of salt crystals to improve the storage performance, for example, the long-term storage cycles and the three-phase cycles. Additional advantage can be expected from the transition from liquid solution to solid crystals or even anhydrous salts. Still crystals should be handled with great discretion as pumps are always in danger of being blocked by them.

Corrosion of metal components in liquid solution is another concern for long-term stability of absorption systems, as all the common used couples are corrosive salts, especially for NaOH/H₂O. Additional corrosion protection methods should be considered for long-term systems as the solution is kept in closed vessels for such a long period.

For solid/gas TES systems (adsorption, sorption reaction, and composite materials), which have similar system configurations, the following additional issues can be addressed:

Realization of the storage potential for promising working pairs greatly relies on the efficient and compact design of sorption bed to ensure good heat and mass transfer. Varied kinds of composite sorption materials have been developed by impregnating salt into pores of matrix. For closed solid/gas sorption systems, consolidation of sorbent pellets with expanded graphite is often used for improving heat conductivity. Recently, compressed activated carbon and ENG-TSA (Expanded Graphite Treated with Sulfuric Acid) was reported by Wang et al. [73] to have a maximum thermal conductivity of 337 W/(m·K) at a density of 831 kg/m³. Wang et al. [74] also prepared consolidated composite sorbents with activated carbon and ENG-TSA and obtained a Highest thermal conductivity of 34.2 W(m·K), which was 150 times higher than ordinary granular activated carbon. It should be noted that relations between heat transfer and mass transfer should be optimized for these consolidated sorbents. For open solid/gas sorption systems, honeycomb structure proved to be a fine design for kinetic performance with its large contact area with air and low pressure loss.

Composite water sorbents offer new opportunities for designing advanced sorption thermal storage systems. One crucial problem needed to be paid attention is the carryover of deliquesced salt solution droplets, which has been ignored by some previous studies. To prevent the carryover from happening, salt content in the matrix should be controlled according to the designed conditions. Tolerated salt content in closed systems can be larger than that in open systems, since the relative pressure for discharging in closed systems are lower than 0.4 while that in open systems can be greater than 0.95. From another aspect, itis easier to avoid the carryover of deliquesced salt in closed systems than in open systems by simply controlling the filling water amount in the evaporator.

The challenges mentioned here show not only that solar energy storage through sorption technology is far from a significant state of efficiency, but they also point the directions toward future investigation. Right now research activities devoted to sorption TES are extensively increasing. To bring this sorption storage solution into the market, more intensive studies in fields of evaluation of advanced materials and development of efficient, compact prototypes are still required.