Low pressure transients can be mechanisms for intrusion ofmicrobial contaminants into the distribution system. Prior researchhas demonstrated episodes of negative pressure (-16 psi) in drinkingwater systems following planned or unplanned pump shutdowns.This poster paper models the risk of infection fromthese negative pressure events and evaluates mitigation options. Aquantitative microbial risk assessment (QMRA) model developedestimates of norovirus and rotavirus concentrations in untreatedsewage based on a Monte Carlo analysis of censored published values.The volume of sewage that could intrude into the system was basedon a logarithmic distribution of the duration of negative pressuretransients and an estimate of the distribution leakage. The dilutionof the intrusion volume was calculated based on the distribution offlow in the system and resulted in an estimation of the final virusconcentration. Dose response data for norovirus and rotavirus alongwith known estimates of water consumption were used to estimatethe risk of infection based on the distribution of customers (demand)within the system. Sensitivity analysis showed that the coincidence ofconsumption (the probability of consuming the slug of contaminatedwater) has the strongest influence on risk and was related to theduration of the negative pressure event. Estimates of the disinfectantdemand due to the intruded diluted sewage, and virus inactivation,showed that maintenance of free chlorine residuals nearly eliminatedthe risk, but viruses persisted for more than 24 hours inchloraminated systems. A single event in an undisinfected systemcould pose significant risk (>1/10,000 annual risk of infection), butdata on electrical power interruptions showed that pump shutdownscould happen as frequently as once per month. Overall the QMRAmodel was useful for organizing known data, estimating data gaps,and testing hypotheses for future research.
