Heat and mass transfer prediction in fluidized beds of cooling and desalination systems by AI approach

Abstract

Since greenhouse gas emissions and freshwater scarcity are the top global risks, looking for new methods to reduce CO2 emissions and increase drinking water production is becoming a significant civilization challenge. One of the promising approaches to addressing these dares has proven to be adsorption cooling and desalination systems powered with low-grade thermal energy, including waste heat of the near ambient temperature. Due to poor heat and mass transfer and the low performance of the existing adsorption chiller with conventional packed beds, the innovative concept of fluidized beds application was elaborated on in the paper. Furthermore, the article introduces a novel approach based on artificial intelligence methods for predicting heat and mass transfer within the adsorption bed of cooling and desalination systems. Silica gel, as the parent adsorption material, and two additives, aluminium and carbon nanotubes, with different shares, are applied in tests. The water vapour uptake and the convective heat transfer coefficient, measured during experiments and predicted by the developed models, are investigated and compared. The data evaluated by models are in good agreement with experimental results. The developed models allow the study of input parameters' effect on the outputs and optimize the operating strategy of the bed. The highest water vapour uptake and the convective heat transfer coefficient, which can be obtained for the considered range of input parameters, are equal to 1.65 g/g and 1212.62 W/m2 K, respectively, and can be achieved only due to the fluidization of the adsorption bed.