Importance of membrane pore size distribution for virus removal in greywater membrane bioreactors

Abstract

Safe use of treated greywater requires effective control of enteric viruses, yet there is limited full-scale evidence on virus removal in membrane-based greywater treatment systems. We performed monthly challenge tests with MS2 bacteriophage, a commonly used conservative surrogate for enteric viruses, over one year in a full-scale membrane bioreactor (MBR) treating light greywater and equipped with ultrafiltration modules (nominal pore size 0.04 μm). Observed log10 removal values (LRVs) ranged from 1.0 to 3.4, substantially lower than ∼4.0 LRVs previously reported for a comparable MBR treating municipal wastewater. LRVs were lowest (∼1.0–1.5) in the weeks following chemical cleaning and increased only modestly during operation, consistent with limited fouling development under low flux conditions (2–8 L m−2 h−1). Scanning electron microscopy (SEM)-based pore size analysis of the pristine ultrafiltration membrane indicated a broad distribution with an upper tail extending to ∼85 nm. Using this measured pore size distribution as input, a mechanistic sieving model predicted a size-exclusion-only LRV of ∼0.5 for MS2-sized particles, providing a mechanistic baseline consistent with measured LRVs observed at full-scale post-cleaning. Overall, these results underscore the need for full-scale validation of greywater MBRs and show that membrane specifications on pore size distribution are needed to support more accurate predictions of virus LRVs in membrane-based greywater treatment applications.