(653e) Computational Fluid Dynamics Modeling of Internal Concentration Polarization in Forward Osmosis: Effects of Pore Geometry, Hydrodynamics, and Solute Diffusivity | AIChE

(653e) Computational Fluid Dynamics Modeling of Internal Concentration Polarization in Forward Osmosis: Effects of Pore Geometry, Hydrodynamics, and Solute Diffusivity

Authors 

R.Esfahani, M. - Presenter, Tennessee Technological Universsity
Languri, E. - Presenter, Tennessee Technological University
Arias Chavez, L. H. - Presenter, Tennessee Technological University

Internal concentration polarization remains a primary limiting factor for water flux in forward osmosis.  Modeling transport through the membrane support layer can provide new insights to guide design of membranes and membrane systems.  Experimental results have definitively shown that membrane support layer structure and module hydrodynamics significantly influence the severity of concentration polarization.  Previous numerical modeling efforts have confirmed the importance of the structural parameter in determining transport and have also characterized hydrodynamic conditions within support layer pores.  We investigate here the interplay between pore geometry and cross-flow velocity in determining the extent of internal concentration polarization and its impact on water flux.  In addition, we examine the potential for error to be introduced in transport models by the assumption of a constant solute diffusivity.  A simplified pore geometry was adopted, with a single cylindrical pore of fixed diameter and length being considered.  Diffusion of water and salt through the pore was modeled together with convection effects induced inside the pore by cross-flow velocity applied at the support layer – bulk solution interface.  The bulk feed and draw solution concentrations were fixed at experimentally relevant values while cross-flow velocity, pore length, and pore diameter were varied.  The draw solution concentration at the support layer – active layer interface and the resulting water flux were then determined via computational fluid dynamics simulations.   Results obtained with concentration-dependent and concentration-independent solute diffusivities were compared.  Implications of this work for future modeling efforts, for membrane support layer design, and for operating conditions will be discussed.