An experimentally optimized model for heat and mass transfer in direct contact membrane distillation

Abstract

Membrane distillation (MD), a thermally driven process involving hydrophobic micro-porous membranes has gained widespread interest in academic research and is set to become an alternative solution to other membrane separation processes such as reverse osmosis (RO). Although extensive experimental studies have been carried out since the 1980s [1,2], clear understanding of the heat and mass transport phenomena has yet to be established. This manuscript presents experimental results of direct contact membrane distillation (DCMD) with de-ionized water and aqueous salt solutions of NaCl with concentration levels of up to 15ppt as feed together with an experimentally optimized and validated model for the prediction of the permeate flux in DCMD for GE Aspire Membrane QL 833 (GE Energy). Different heat transfer prediction methods in combination with the three different forms of the Dusty Gas model for mass transport were used in the comparison of our experimental data in the laminar and turbulent flow regimes under steady-state conditions. The comparison between experimental and predicted results confirmed our expectation that the Knudsen-molecular diffusion transition model yielded the best prediction. We have also identified, based on the comparison of the data, the most accurate heat transfer correlation for the laminar and turbulent flow regimes, taking into account the experimental and permeate prediction uncertainties to optimally address the heat and mass transport equations used in DCMD studies. Hence, it is highly recommended that these heat transfer correlations and the Knudsen-molecular mass transport equation be used in the prediction of heat and mass transfer for flat sheet DCMD experiments.