(506c) Performance Analysis of Fluidized Bed Reactors for Thermochemical Energy Storage Based on the CaO/CaCO3 and MnAl2O4/MnAl2O4-? Systems

Thermochemical energy storage is an interesting tool to decouple the intermittent production of electricity from renewable sources from its use. The process is based on gas-solid reactions characterized by high heats of reaction: the endothermic reaction is carried out when excess energy is available (charge phase), while the exothermic reaction is allowed to take place when the energy demand is higher than the production (discharge phase).

The efficiency of these storage systems and the possibility of coupling them with downhill energy-demanding processes depend on both the characteristics of the material employed, such as heat of reaction, kinetics, and cyclability, and the reactor design. More specifically, the system should be capable of guaranteeing a constant outlet temperature of the gas during the discharge phase for long periods of time. Both the CaO/CaCO₃ and the MnAl₂O₄/MnAl₂O_4-δ materials have been found to be promising for application in thermochemical storage reactions. To improve the thermal management of the process, the use of fluidized bed reactors has been proposed, because of the high heat transfer rate between solid and gas and their capability of maintaining almost isothermal conditions; however, these reactors are less flexible in terms of operating conditions because of the need to maintain adequate fluidization conditions. These are, in turn, strongly affected by changes in density of the solid particles that can take place during gas-solid reactions. Consequently, it is important to tailor the operating conditions and reactor geometries to the specific system in use and to ensure that the fluidization regime is maintained throughout the reaction.

In this work, a model for a fluidized bed reactor operating with either the CaO/CaCO₃ or the MnAl₂O₄/MnAl₂O_4-δ was developed. The effect of the degree of advancement of the reaction on the fluidization characteristics was analyzed and simplified models were developed to evaluate key performance indicators at a low computational cost.