Variability analysis of pathogen and indicator loads from urban wastewater systems along a river for drinking water risk estimations

BACKGROUND AND OBJECTIVESPathogens in source waters represent a health risk for drinking water consumers that may result in waterborne diseases, if not sufficiently removed within water treatment plants. According to the water safety plans in the WHO Drinking Water Guidelines (3rd ed.), identification and comparison of drinking water hazards within the catchment is of significant importance to accomplish effective risk management. In comparison to baseline concentrations significantly higher levels have been documented after heavy rainfalls (Kistemann et al. 2002; Signor et al. 2005) and a large number of waterborne outbreaks documented have been preceded by periods of heavy rainfalls (Rose et al. 2000; Curriero et al. 2001). Heavy rainfalls with wastewater discharges, in combination with treatment failures, resulted in the largest waterborne outbreak in Sweden in 1988. In this paper we have assessed the spatial and temporal distribution of wastewater discharges from urban areas along the river Göta älv used for drinking water supply to about 700 000 people including the city of Göteborg. The objectives were to assess the pathogen loads and variability by simulations of wastewater discharges from the urban sewage systems. The simulations have been compared with measured concentrations of pathogens and faecal indicator bacteria in the river.METHODSMicrobial point sources in municipal sewage systems along the river Göta älv, upstream the raw water intake to the city of Göteborg, were identified and quantified in relation to time and magnitude. Data on discharges of treated wastewater from wastewater treatment plants (WWTPs), as well as untreated wastewater from combined sewer overflows (CSOs) and sanitary sewer overflows (SSOs), were collected, including onsite monitoring and urban wastewater modelling results. Literature data on the concentrations as well as the reduction of indicator bacteria (E. coli and spores of clostridia), Campylobacter, noroviruses, Giardia and Cryptosporidium within sewage systems were used. Through Monte Carlo simulations the daily microbial loads to the river were calculated using pathogen concentrations and discharge information from the wastewater systems for each single discharge point. Microbial loads were calculated for baseline conditions and for rain event conditions, based on a half year data period. Baseline conditions were defined as dry weather periods with discharges predominantly of treated wastewater. Rain event conditions were separately described for two wet weather periods, including discharges from CSOs and SSOs. Assuming a total mixing of wastewater pathogens discharged into the river, the downstream microbial concentration was approximated and compared to sampling data on indicator and pathogens from the river Göta älv for the same data period.RESULTSThe discharge volumes of treated wastewater varied highly between WWTPs along the river as a function of persons connected and the portion of stormwater within the system. CSOs as well as SSOs activated during wet weather periods increased the discharge volumes from the urban sewer systems. Daily loads simulated for the indicator bacteria as well as the pathogens were presented in figures, showing the contributions for the separate municipalities along the river. In general, the pathogen loads during baseline conditions were defined by the flow and the number of persons connected to each system, while the pathogen loads during rain conditions were additionally set by the degree of combined and sanitary sewer overflows. Downstream positive detections of norovirus were made during a wet weather period with multiple discharges of combined and sanitary sewer overflows registered along the river, confirming the simulated results. Overall, the river concentrations approximated by these simulations, were in the same range for the faecal indicator bacteria as compared to sampling data for the same period, but higher for the Campylobacter and slightly lower for norovirus, Giardia and Cryptosporidium. CONCLUSIONS AND SIGNIFICANCE OF THE WORK In this study, local flow and volume registrations were used to theoretically assess the pathogen load along a river used as a source for drinking water production. The range of pathogen loads from the urban areas, as well as the differences between dry and wet weather conditions, illustrates the potential variability of source water pathogen concentrations. This information may be used for assumptions on pathogen concentration, such as within quantitative microbial risk assessments. Present simulation results also suggest a first prioritization for reduction of microbial loads within urban wastewater systems along this river, such as the need for local additional wastewater treatment steps in order to bring down the spread of pathogens. This investigation is also a starting point to further investigate presently unquantified microbial sources, such as unregistered CSOs, SSOs and animal contaminations, as well as the microbial effects from sedimentation, die-off and the downstream microbial transport mechanisms in the river.