A Nanofiltration Decision Tool for Potable Reuse: A New Rejection Model for Recalcitrant CECs

Effluent samples from three water resource recovery facilities (WRRFs) were individually characterized for presence and concentration of 94 different chemicals of emerging concern (CECs), using analytical methods with reporting limits in the low-parts-per-trillion. Following CEC analysis, each sample was subjected to dead-end, pressurized membrane filtration with either a nanofiltration (NF) or reverse osmosis (RO) membrane. The majority of the measured CECs were rejected by both membranes by 1-log removal (90%) or greater. However, nine of the 94 CECs had average rejection rates by the NF membrane less than 90%. A multilevel, multivariable model was developed to predict the probable rejection coefficients of CECs with the studied NF membrane. The resulting Quantitative Molecular Properties Model (QMPM) predicted the NF rejection of CECs based on size, ionic charge, and hydrophobicity. The model parameters that successfully predicted NF rejection in bench testing were log (Kow/Kaw) and the polar surface area of the CEC molecule.