Computational modeling of discrete-object feed spacers attached directly onto reverse osmosis membranes for enhanced module packing capacity and improved hydrodynamics

Abstract

In the recent decade, various printing technologies have been under development for producing spacers in reverse osmosis (RO) and nanofiltration (NF) modules. One manufacturing technology directly attaches discrete objects or features to the feed-side membrane surface. The novelty of this method is that tangential flow over the membrane is much less obstructed than with conventional feed spacers, and thus thinner spacers can be built. Within the same cylindrical module volume, more layers can be packed and more membrane surface area can be added to achieve higher permeate flow rates. The research goal of this paper is to design efficient models that help us discover the best parameters for discrete-object spacers. Spacer heights and spacer patterns are directly related to longitudinal pressure drop and concentration polarization (CP). Computational fluid dynamics (CFD) is useful in helping expedite the process of discovering the best designs with both low pressure drop and low CP. Simulations investigated spacer shapes including regular cylinders, elliptical cylinders, and airfoils. Ellipses with a length:width ratio of 2.4 while maintaining a between-feature distance of 6 mm were optimal for minimizing pressure drop. Spacer heights ranging from 200 µm to 500 µm were simulated to discover the height to achieve the same pressure drop as a 30 mil (0.762 mm) conventional spacer. Simulation results indicate that discrete-object spacers with elliptical design can greatly increase water productivity in a spiral-wound module by increasing packing capacity. These designs also reduce CP by 7% by improving the hydrodynamics, compared to empty channels.