Abstract
Reverse Osmosis (RO) is a process whereby solutes are removed from a solution by means of a semipermeable membrane. Providing access to clean water is one of our generation's grand engineering challenges, and RO processes are taking center stage in the global implementation of water purification technologies. In this work, computational fluid dynamics simulations are performed to elucidate the steady state phenomena associated with the mass transport of solution through cylindrical hollow fiber membranes in hopes of optimizing RO technologies. The Navier-Stokes and mass transport equations are solved numerically to determine the flow field and solute concentration distribution in the hollow fiber membrane bank, which is a portion of the three-dimensional feed channel containing a small collection of fibers. The k-co Shear Stress Transport turbulence model is employed to characterize the flow field. Special attention is given to the prediction of water passage through hollow fiber membranes by the use of the solution-diffusion model, which couples the salt gradient, water flux, and local pressure at the membrane surface. This work probes hollow fiber membrane arrangement in the feed channel by considering inline and staggered alignments. Feed flow rates for Reynolds number values ranging between 400 and 1000 are considered. Increased momentum mixing within the feed channel solution can substantially enhance the system efficiency, and hollow fiber membrane arrangements and feed flow rates dictate the momentum mixing intensity. Velocity and vorticity isosurfaces of the flow domain are presented in order to assess the momentum mixing achieved with various hollow fiber membrane arrangements and flow rates. The total water permeation rate per hour is calculated to compare system efficiencies, and the coefficient of performance is calculated to compare membrane performance relative to the necessary power input, both for the various hollow fiber membrane arrangements and feed flow rates.