الفهرس | Only 14 pages are availabe for public view |
Abstract Desiccant dehumidification is a simple and cheap mean of obtaining dry air as required in many industrial applications. Researchers have proposed and investigated different types of desiccant materials and bed configurations with objective of achieving high performance for the desiccant dehumidification system. The present study presents a few ideas like using composite desiccant particles and bed of varying particle diameter in line to improve bed performance in terms of total mass adsorbed and savings in pressure DROP for a vertical silica gel packed bed. The general two dimension mathematical formulations for moisture diffusion and heat conduction in a spherical silica gel particle subjected to a flow of moist air have been presented by applying energy and moisture conservation laws. Following this the PGC, SSR, SSR-AC and MSSR models have been deduced to study in detail the heat and moisture interactions in packed bed considering axial heat conduction in the bed, composite particle, varying particle diameter bed and cyclic operation. The experimental results extracted during the present study for single blow adsorption, single blow desorption and cyclic operation tests for silica gel packed bed have been used for validating the different mathematical models which have been used for predicting the transient variation of various operating parameters like exit air humidity ratio, exit air temperature, bed water content and bed temperature. For the first time the PGC model for desiccant packed bed has been solved using semi analytical method. The results from the semi analytical solution as well as numerical solution have been compared with the experimental data. A range of relative errors for exit air temperature and humidity ratio are 1.08% to 13.49% and 3.77% to 27.17%, respectively. The significance of the axial heat conduction in the bed as a function of the flow velocity, particle diameter, bed length and number of units of mass transfer ( ) has been investigated. Through a comparison for the numerical results of SSR and SSR–AC models it has been found that the difference in the predicted bed performance is negligible when the number of mass transfer units is less than unity. |