Evaluation of Potential Indicators of Virus Removal During Advanced Physical Treatment of Wastewater

Abstract

The increasing threat of water scarcity throughout the world has led to the implementation of potable wastewater reuse. During potable wastewater reuse, final treated wastewater effluent is purified to drinking water quality. This is performed to augment current drinking water sources. However, treated wastewater effluent is of degraded quality when compared to conventional drinking water sources, therefore extensive treatment and quality monitoring is required for the protection of public health. Removal of pathogens, particularly viruses, is a major concern during wastewater reuse, as sewage is a significant source of enteric pathogens. Viruses are the smallest of enteric pathogens and have physical attributes which allow for persistence in the environment for extended periods of time while maintaining viability. Therefore, strict requirements pertaining to virus removal during wastewater reuse are enforced. Treatment facilities are required to demonstrate the capability of different treatment technologies implemented for virus removal. This is largely done by the spiking of laboratory propagated MS2 coliphage, which is recommended by the National Water Research Institution Framework for Direct Potable Reuse (2015). Additionally, if membrane processes are implemented, they must undergo continual monitoring ensuring membrane integrity, which is performed by spike evaluations as well as the use of non-viral surrogates such as conductivity and total organic carbon. However, spiking coliphages can quickly become cumbersome at full-scale facilities. Advances in virus detection technology has facilitated the potential use of naturally occurring indicator viruses as a means of evaluating system performance. Wastewater is a rich reservoir of viruses, with non-human viruses greatly outnumbering viral pathogens. These abundant non-human viruses can be utilized as indicators of viral pathogen fate, allowing for continual monitoring of virus removal performance by a variety of different reuse technologies. This dissertation summarizes three studies which evaluate the fate of different potential indicator viruses found abundantly in wastewater, and their potential use to model virus removal during wastewater reuse. Evaluated indicators include plant viruses Pepper Mild Mottle Virus (PMMoV) and Cucumber Green Mottle Mosaic Virus (CGMMV), Bacteroides infecting crAssphage, human adenoviruses (HAdVs), as well as a recently discovered eukaryotic virus, wastewater circular rep-encoding single stranded (CRESS) DNA virus 2 (WCDV-2). Two physical treatment processes were evaluated: soil aquifer treatment (SAT) and integrated ultrafiltration (UF) and reverse osmosis (RO) membrane processes. Removal, persistence, and abundance of the different indicators across these two treatment processes was assessed. Additionally, metagenomic techniques were utilized across an integrated membrane process to identify groups of viruses which may exhibit an increased propensity to pass through damaged membranes, and thus become a valuable tool in monitoring membrane integrity. The first study evaluated removal by SAT using PMMoV, crAssphage, and HAdVs. During SAT, water is infiltrated through the vadose zone and blended with natural groundwater where it can later be extracted. The physicochemical properties of soil allow for additional treatment of water, including pathogen removal. This study examined two SAT sites receiving treated wastewater effluent from the same wastewater treatment plant, yet implementing different operational parameters, including wetting drying cycles and infiltration rates. Both PMMoV and crAssphage were found most prevalently (100%, n = 6) in the feed water, however greater concentrations of PMMoV were exhibited (6.6 ± 1.0 log10 gc per L). PMMoV also exhibited the highest prevalence in the groundwater (22%, n = 18), where it was frequently detected from an SAT site which utilizes faster infiltration rates as well as shorter wetting drying cycles. Results from this study indicate that PMMoV is a valuable indicator of virus removal, revealing that parameters such as wetting-drying cycles might be influencing subsurface virus transport more than initially understood. The second study evaluated PMMoV, CGMMV, WCDV-2, and HAdVs as potential indicators of membrane treatment performance. Two pilot-scale and one full-scale integrated UF and RO membrane processes at three different treatment facilities were evaluated. PMMoV and CGMMV were found prevalently before and after treatment at all three facilities, and differences in their fate were largely indiscernible. WCDV-2 showed potential as a conservative indicator of membrane integrity, however, it exhibited a pattern of regional specificity, reducing the confidence that this novel CRESS virus could be utilized universally. The two pilot facilities exhibited greater log10¬ reduction values (LRVs) of potential indicators than full-scale, by as much as 2 LRVs. This discrepancy between pilot and full-scale highlights the need for integrity monitoring, as full-scale facilities utilize 100 times as many membranes as a pilot scale facility, increasing the chances of a non-integral membrane to go undetected and allow for the passage of pathogens. The final study utilized next generation sequencing across a pilot-scale integrated UF and RO membrane process, evaluating the presence of RNA and DNA viruses in the feed water, RO concentrate, and the final permeate. Over 6,000 individual operational taxonomic units (OTUs) were determined in this study, with the majority being prokaryotic viruses (bacteriophages). CRESS viruses, which are among the smallest of eukaryotic viruses, were found to increase in relative abundance after treatment by UF and RO, providing further evidence that these small viruses have an increased propensity to pass through membrane damage and thus potentially provide a sensitive estimator of membrane integrity. Interestingly, no members of Virgaviridae were present in RO permeate, likely due to lowered sensitivity of sequencing methods. However, the presence of OTUs belonging to Tombusviridae, a group of icosahedral, singe stranded RNA plant viruses, in permeate samples suggests that other plant viruses might be more resistant to physical removal than Virgaviridae members. Overall, the conclusions of these studies have provided valuable information regarding the potential use of virus indicators that are found abundantly in wastewater. PMMoV has showed promising capabilities as an indicator of virus removal during both SAT and membrane processes. It allows for the evaluation of different parameters of SAT, as well as exposes the weaknesses of membrane processes at full-scale. The closely related CGMMV provides similar information, and metagenomic analysis performed found many abundant members of the genus Tobamovirus in treated wastewater effluent, opening the possibility of exploring removal of Tobamoviruses as a whole. However, neither of these Virgaviridae members were detected during metagenomic analysis of RO permeate. CRESS viruses have potential to play an important role regarding membrane integrity evaluation, as exhibited by their prevalence and diversity in membrane permeate streams, however more research determining CRESS viruses common to different water sources is required. Overall, naturally occurring viruses seem to provide an alternative method of evaluating LRVs and membrane integrity during wastewater reuse, without the need to spike coliphages.