.png)
Heat storage is a key lever for decarbonizing energy systems. Among the three families of thermal storage—sensible, latent, and thermochemical (TCES)—the latter stands out with theoretical energy densities of 140 to 560 kWh/m³, compared to 15–50 kWh/m³ for sensible storage and 55–110 kWh/m³ for latent storage. Furthermore, it allows for loss-free storage at ambient temperature, paving the way for seasonal storage, though it remains largely in the R&D phase.
Figure 1 – Performance comparison of the three thermal storage families (Kwasi-Effah et al., 2025)
During charging, a heat input dissociates a material into two products stored separately at ambient temperature, without any thermal loss, much like a sponge that is wrung out and can be stored indefinitely while dry. The discharge is triggered by the recombination of the two products, releasing heat through an exothermic reaction. A distinction is made between sorption, where water vapor physically attaches to a porous solid, and chemical reactions proper, which involve the transformation of molecular bonds.
Figure 2 – TCES principle (Gbenou et al., 2021)
The sorption systems are suitable for low temperatures (<200 °C) (TRL 5-7). They rely on a simple principle: certain mineral salts spontaneously absorb water vapor while releasing heat, and can be regenerated by heating them moderately. Magnesium sulfate, calcium chloride, and potassium carbonate are the most studied. Their recharging would take place at temperatures accessible via industrial waste heat (80–150 °C), and their discharge would release 42 to 110 kWh/m³. These materials are inexpensive and could theoretically store summer heat for use in winter. Zeolites and new porous materials based on metal-organic frameworks (MOFs)(TRL 3-5), which can be custom-designed, offer superior kinetic performance but at costs that remain high.
The hydroxides and carbonates could allow for the storage of medium temperatures (200–650 °C) (TRL 4-5). The dehydration of calcium hydroxide into calcium oxide at around 450–550 °C can release approximately 130 kWh/m³ in a fully reversible manner. The calcium carbonate/calcium oxide pair (~700–900 °C) offers even higher densities, but suffers from degradation due to sintering during repeated cycles.
Finally, redox metal oxides could store high temperatures (750–1,050 °C) (TRL 3-4). The cobalt oxide, manganese oxide, and copper oxide pairs use air as both a reactant and a heat transfer fluid, simplifying integration into solar towers.Cobalt oxide reaches a theoretical ~220 kWh/m³. Agglomeration and low thermal conductivity necessitate fluidized bed architectures or nanoparticle doping.
No commercial deployment exists yet, but application niches are becoming clearer. The EU has mobilized over €95 million via Horizon Europe since 2021, and 1–10 MW pilots are expected before 2028. TCES primarily targets high-temperature industrial heat for sectors such as cement and steel. It also represents a solution for recovering waste heat from data centers, whose global electricity consumption is expected to double by 2030: the European project THUNDER explores the precise thermochemical seasonal storage of waste heat from these infrastructures to supply urban heating networks. With an energy density up to ten times higher than sensible storage, it represents a structural option for the transition, provided the pilot demonstration phase is completed before 2030.