Demonstrating the reduction of enteric viruses by drinking water treatment during snowmelt episodes in urban areas

Abstract

This study investigates short-term fluctuations in virus concentrations in source water and their removal by full-scale drinking water treatment processes under different source water conditions. Transient peaks in raw water faecal contamination were identified using in situ online β-d-glucuronidase activity monitoring at two urban drinking water treatment plants. During these peaks, sequential grab samples were collected at the source and throughout the treatment train to evaluate concentrations of rotavirus, adenovirus, norovirus, enterovirus, JC virus, reovirus, astrovirus and sapovirus by reverse transcription and real-time quantitative PCR. Virus infectivity was assessed through viral culture by measurement of cytopathic effect and integrated cell culture qPCR. Virus concentrations increased by approximately 0.5-log during two snowmelt/rainfall episodes and approximately 1.0-log following a planned wastewater discharge upstream of the drinking water intake and during a β-d-glucuronidase activity peak in dry weather conditions. Increases in the removal of adenovirus and rotavirus by coagulation/flocculation processes were observed during peak virus concentrations in source water, suggesting that these processes do not operate under steady-state conditions but dynamic conditions in response to source water conditions. Rotavirus and enterovirus detected in raw and treated water samples were predominantly negative in viral culture. At one site, infectious adenoviruses were detected in raw water and water treated by a combination of ballasted clarification, ozonation, GAC filtration, and UV disinfection operated at a dose of 40 mJ cm^-2. The proposed sampling strategy can inform the understanding of the dynamics associated with virus concentrations at drinking water treatment plants susceptible to de facto wastewater reuse.

Keywords: Drinking water; Enteric viruses; Risk assessment; β-d-glucuronidase.

Graphical abstract

Fig. 1

Unit processes involved in the treatment train of drinking water treatment plants A and B and sampling points (red). (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 2

Time series of daily rainfall, β-d-glucuronidase (GLUC) activity, snow cover, raw water turbidity, raw water ammonia concentration and river flow rate during snowmelt periods in 2017 and 2018 at drinking water treatment plants (DWTPs) A and B. Yellow rectangles indicate targeted events. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 3

Time series of β-d-glucuronidase (GLUC) activity and rotavirus, adenovirus, norovirus GII, and JC virus concentrations during snowmelt episodes at drinking water treatment plants (DWTPs) A and B. Yellow rectangles indicate targeted events. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 4

Histograms for rotavirus, adenovirus, norovirus, and JC virus concentrations in raw water, settled water, and filtered water during events A1 and A2 at drinking water treatment plant (DWTP) A. Error bars represent the 95% credibility interval on the virus concentration assuming that purified virus nucleic acids in each PCR are Poisson-distributed. Columns with no colour represent the limit of detection. Orange glowing bars represent samples positive by ICC-qPCR. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 5

Histograms for rotavirus, adenovirus, norovirus GII, JC virus, and enterovirus concentrations in raw water, settled water, filtered water and UV disinfected water under baseline and event conditions (Event B2) at drinking water treatment plant (DWTP) B. Error bars represent the 95% credibility interval on the virus concentration assuming that purified virus nucleic acids in each PCR are Poisson-distributed. Columns with no colour represent the limit of detection. Orange glowing bars represent sample positives by ICC-qPCR. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)

Fig. 6

Change in virus removal performances of coagulation/flocculation in response to enteric virus peak concentrations in raw water during snowmelt episodes at drinking water treatment plants (DWTPs) A and B. White circles and squares represent minimum removal performance values due to the inability to quantify the virus in settled water (below the limit of detection).