Loading...
Application of membrane mass transfer models to direct potable reuse projects employing reverse osmosis and nanofiltration
Ohm, Kevin K.
Ohm, Kevin K.
Citations
Altmetric:
Advisor
Editor
Date
Date Issued
2023
Date Submitted
Keywords
Collections
Research Projects
Organizational Units
Journal Issue
Embargo Expires
Abstract
There is increased interest in alternative sources of drinking water such as potable reuse of wastewater due to growing water demand combined with depletion of traditional water sources. Because wastewater contains higher concentrations of contaminants and pathogens compared to conventional drinking water sources, potable reuse requires additional treatment to meet drinking water regulations and protect human health. Commonly used treatment processes for potable reuse applications include reverse osmosis (RO) and nanofiltration (NF) membranes, which have demonstrated high contaminant removal, often greater than 99\%, and above 5-log pathogen removal. However, there is a need to predict the rejection of contaminants over a range of membrane system operating conditions and develop integrity testing approaches to verify pathogen log-removal credits during potable reuse applications.
Membrane models have been developed and employed to evaluate mass transfer of solutes across high-pressure membranes and incorporate process phenomena including osmotic pressure, surface adsorption, concentration polarization, and electrokinetic effects as well as system conditions such as cross-flow velocity, permeate flux, and feed concentration. Two membrane mass transfer models, the phenomenological and solution diffusion model, were employed to evaluate their effectiveness for describing the rejection of several perfluoroalkyl acids (PFAAs) by four RO and NF membranes (SWRO, CR100, NF90, and NF270) in a closed-circuit desalination system and identify integrity defects in an RO (ESPA2-LD-4040) membrane system by describing conductivity rejection during baseline operation and operation with a defect.
The solution diffusion model was found to describe PFAA rejection by high-pressure membranes more accurately than the phenomenological model. Both models were found to describe PFAA rejection by the loose NF membrane (NF270) less accurately than the tight NF membrane (NF90) and RO membranes (SWRO and CR100).
The phenomenological model was found to identify membrane defects more reliably than the solution diffusion model. During baseline operation, error in conductivity rejection predicted by the phenomenological model remained constant, while the error in rejection predicted by the solution diffusion model increased over time, which could result in defects being incorrectly identified. However, both models failed to identify three of the six defects, suggesting conductivity is not a reliable marker for this method.
Associated Publications
Rights
Copyright of the original work is retained by the author.