Assessing microbial risk through event-based pathogen loading and hydrodynamic modelling.

The aim of this study was to assess the variability of microbial risk associated with drinking water under various contaminant loading conditions in a drinking water source. For this purpose, a probabilistic-deterministic approach was applied to estimate the loadings of Cryptosporidium, Giardia, and Escherichia coli (E. coli) from fecal contamination sources during both dry and wet weather conditions. The relative importance of loads originating from various fecal contamination sources was also determined by a probabilistic approach. This study demonstrates that water resource recovery facilities were the dominant source of Giardia, yet rivers were more important with regards to Cryptosporidium. Estimated loadings were used as input to a three-dimensional hydrodynamic model of Lake Ontario; the fate and transport of microbial organisms were simulated at the influent of a drinking water intake. Discharge-based hydrodynamic modelling results were compared to observed concentrations. Simulated probability distributions of concentrations at the intake were used as an input to a quantitative microbial risk assessment (QMRA) model such that the variability of microbial risk in the context of drinking water could be examined. Depending on wind and currents, higher levels of fecal contamination reached the intake during wet weather loading scenarios. Probability distribution functions of Cryptosporidium, Giardia and E. coli concentrations at the intake were significantly higher during wet weather conditions when compared to dry conditions (p < 0.05). For all contaminants studied, the QMRA model showed a higher risk during wet weather (over 1 order of magnitude) compared to dry weather conditions. When considering sewage by-pass scenarios, risks remained below 2.7 × 10 person day for Giardia and E. coli O157:H7. Limited data were available for Cryptosporidium in by-pass effluents and the risk is unknown; hence it is critical to obtain reliable loading data for the riskiest scenarios, such as those associated with water resource recovery facility by-passes.