CFD Modelling and Multi-Objective Optimization of a Segmented Open Sorption Heat Storage

Abstract

Sorption heat storage offers high energy density and low thermal losses, making it attractive for short-term and seasonal solar heat storage in buildings. However, implementation depends on reactor design, particularly in terms of scalability, flow distribution, and pressure losses. Modular open-reactor designs improve scalability, while internal segmentation can reduce pressure drop by shortening the flow path. However, segmentation may also cause non-uniform heat and mass distribution, leading to outlet-temperature non-uniformity and less reliable heat delivery. This paper combines a computational fluid dynamics (CFD) model of a segmented reactor with response surface methodology (RSM)-based multi-response optimization of key design variables. Coupled heat and mass transfer in a zeolite packed bed is simulated using adsorption equilibrium and kinetic relations. A central composite design (CCD) samples the geometry space, and quadratic response-surface models are developed for energy storage density, useful recovered energy, pressure drop, average outlet temperature, outlet-temperature non-uniformity, the 95% threshold temperature, and the corresponding duration above this threshold. The surrogate models are evaluated using analysis of variance (ANOVA) and diagnostic checks. A desirability-function approach combines the seven responses into a single objective and identifies high-desirability geometries under an approximately constrained module envelope. The optimized CFD-confirmed design increased useful recovered energy by 18.28%, energy storage density by 5.21%, reduced outlet-temperature non-uniformity by 70.25%, and extended the duration above the 95% temperature threshold by 34.65%, with only a 10% increase in pressure drop. These results demonstrate that the proposed CFD–RSM framework is an efficient, reliable approach for multi-objective optimization of sorption reactors.