3-D numerical simulations of conjugate heat transport in vacuum membrane distillation systems with applied membrane heating

Abstract

Membrane distillation has gained attention recently for its capabilities to treat hyper-saline brine and its compatibility with renewable heat. But the effects of temperature and concentration polarization are major inhibitors to its permeate production and ultimately its commercial viability. To alleviate these effects, we investigate an improvement to vacuum membrane distillation (VMD) such that a thin, thermally conductive, porous metal mesh is placed beneath the membrane. This mesh is heated laterally with low-grade heat to actively heat the membrane and feed, thereby countering effects of temperature polarization. We develop a three-dimensional CFD code to simulate heat and vapor transport conjugately in the feed, membrane, and mesh for this system, which included deriving new equations governing heat and mass in the membrane and mesh. We discretize these governing equations with second-order accurate finite volume methods. These methods are verified using manufactured solutions. We then perform a comprehensive parametric study of fully developed duct flow over a heated plate. We identify the optimal combination of plate properties, duct dimensions, and operating conditions to maximize uniform heating of the duct-plate interface. With this, we identified that decreasing channel width, decreasing inlet flow rate, and increasing plate thickness provided the best results of uniform heating. We then validate our solver against experimental measurements of vapor flux, and determined the best fit for membrane vapor permeability Am. For best fit Am, we were able to reproduce experimental results to within 9% mean error. Following that, we performed a second comprehensive parametric study of the VMD system to investigate the effect of operating conditions, mesh properties, and system geometry on temperature polarization and vapor flux measurements. We observe that polarization effects could be reversed for systems with a high input heat, faster flow rate, slim channel width, thicker mesh, and high vacuum pressure.