Charging/discharging performance examination in a finned-tube heat storage tank: Based on artificial neural network, pareto optimization, and numerical simulation

The efficient use of solar energy is essential for sustainable energy development, and latent heat storage systems using phase change materials (PCMs) are promising candidates for enhancing energy storage density and reliability. This study aims to optimize the thermal performance of a novel heat storage tank inspired by shell-and-tube heat exchangers, using RT50 as the PCM. The system’s geometry was systematically modified by adjusting finned tubes’ horizontal and vertical displacements and altering fin angles to direct natural convection. A numerical model based on the enthalpy-porosity technique simulated the melting and solidification processes across 27 configurations. Due to the high computational cost, artificial neural networks (ANNs) were trained on the numerical data to provide accurate predictive models for melting behavior. These models were then integrated with a genetic algorithm and Pareto optimization to identify optimal configurations for rapid melting and high energy storage capacity. At first, the time needed to melt 50% and 94% of RT50 for the different storage tank configurations was determined. The originality of this study lies in the combined use of ANN and genetic algorithm for multi-objective optimization, the inclusion of fin angle effects on convection, and the presentation of ready-to-use predictive models to accelerate future research and industrial design. Using prediction models and genetic algorithms, two optimized configurations were obtained. Optimal configuration 1 minimized the time to reach 50% melting. Optimal configuration 2 minimized the time for 94% melting, with differences of up to 60 minutes observed between the best and worst-performing cases. These differences are critical in real-world solar energy systems, where charging periods are typically limited to 4–6 hours. In a 5-hour charging test, optimal configuration 1 and optimal configuration 2 achieved 25.12% and 28.06% higher melting, respectively, compared to the fin-free case. Ultimately, the solidification process was examined for the optimized configurations over the remaining 19 hours, and superior energy retention was confirmed in these designs. • A latent heat storage tank with finned tubes and PCM enclosure was analyzed. • The charging and discharging processes of PCM were investigated. • A sensitivity analysis was conducted on the configuration of tubes and fins. • ANNs were used to achieve a more detailed modeling and in-depth investigation. • Single/multi-objective optimization alongside a cost-effective analysis was done.

Перевод пока недоступен