Assessing Inhalation Exposures Associated with Contamination Events in Water Distribution Systems

Abstract

When a water distribution system (WDS) is contaminated, short-term inhalation exposures to airborne contaminants could occur as the result of domestic water use. The most important domestic sources of such exposures are likely to be showering and the use of aerosol-producing humidifiers, i.e., ultrasonic and impeller (cool-mist) units. A framework is presented for assessing the potential effects of short-term, system-wide inhalation exposures that could result from such activities during a contamination event. This framework utilizes available statistical models for showering frequency and duration, available exposure models for showering and humidifier use, and experimental results on both aerosol generation and the volatilization of chemicals during showering. New models for the times when showering occurs are developed using time-use data for the United States. Given a lack of similar models for how humidifiers are used, or the information needed to develop them, an analysis of the sensitivity of results to assumptions concerning humidifier use is presented. The framework is applied using network models for three actual WDSs. Simple models are developed for estimating upper bounds on the potential effects of system-wide inhalation exposures associated with showering and humidifier use. From a system-wide, population perspective, showering could result in significant inhalation doses of volatile chemical contaminants, and humidifier use could result in significant inhalation doses of microbial contaminants during a contamination event. From a system-wide perspective, showering is unlikely to be associated with significant doses of microbial contaminants. Given the potential importance of humidifiers as a source of airborne contaminants during a contamination event, an improved understanding of the nature of humidifier use is warranted.

An overlooked route of inhalation exposure to tap water constituents for children and adults: Aerosolized aqueous minerals from ultrasonic humidifiers

“…Data are not available for the types of water that consumers actually use to fill their ultrasonic humidifiers. As it requires approximately 2.5 L of water per 8-h cycle, it is likely that consumers may use local tap water for convenience ( Davis et al., 2016 ) even though about three decades ago the United States Environmental Protection Agency ( USEPA, 1991 ) recommended using distilled or low mineral water. Select ultrasonic humidifiers manufacturers make recommendations of water type to use in their products.…”

Response to comments by P. Moore on the paper “Mass imbalances in EPANET water-quality simulations”

“…Except as noted, all simulations used a hydraulic time step of 3600 s. Various water-quality time steps were used for time-driven simulations to determine the influence of the time step on the MBR and the constituent concentrations. The default water-quality time step in EPANET is 300 s, as noted above; however, some studies, e.g., Diao et al (2016), Helbling and VanBriesen (2009), and Wang and Harrison (2014), use substantially longer time steps. Therefore, in our analysis we include water-quality time steps longer than the default value.…”

Predicting Contaminated Water Removal From Residential Water Heaters Under Various Flushing Scenarios

“…Researchers have shown that storage‐type water heaters can be modeled as a plug‐flow reactor (PFR) while turned on because of tank temperature stratification and because mixing does not usually occur during the first 80% of water volume used under low flow conditions (<2 gpm) (Brazeau & Edwards 2013a). However, best practices for water heater flushing should include turning the water heater off to avoid volatilization of potential harmful contaminants (Davis et al 2016, Omur‐Ozbek et al 2016, Casteloes et al 2015, Brazeau & Edwards 2013a); therefore, if the water heater is turned off and there is no temperature stratification, it is likely that mixing will occur during flushing, and a PFR model is not appropriate. In this case, a continuously stirred tank reactor (CSTR) model was evaluated, and researchers predicted that some flushing protocols would not have resulted in “safe” heaters (Casteloes et al 2015).…”