Assessment and Prediction of the Thermal Performance of a Centralized Latent Heat Thermal Energy Storage Utilizing Artificial Neural Network

Abstract

A simulation tool is developed to analyze the thermal performance of a centralized latent heat thermal energy storage system (LHTES) using computational fluid dynamics (CFD). The LHTES system is integrated with a mechanical ventilation system. Paraffin RT20 was used as a phase change material (PCM) and fins are used to enhance its performance. Due to the fact that the numerical calculations take a longer time, the simulations are performed on the first day of each week through summer months and then the database is used to train an artificial neural network (ANN) for predicting of the performance. Then, the LHTES's outlet air-temperature function is integrated into the TRNSYS building thermal response model. The trained ANN is able to improve the prediction of the LHTES's outlet air-temperature for a wide range of inlet conditions (i.e., air-temperature and flow rate). We found that the indoor air-temperature is reduced by about 1.5-2.5.