Human health risk modeling for various exposure routes of trihalomethanes (THMs) in potable water supply

Abstract

Deterministic and probabilistic approaches are used here to calculate the human health risk associated with trihalomethanes (THMs) in the water supply of three communities namely: St. John's, Clarenville and Shoal Harbour located in Canada's Newfoundland and Labrador province. Every effort is made to utilize the latest information on chloroform toxicology to quantify cancer risks through different exposures. Showering and drinking activities are identified as major sources of exposure. Chloroform is considered as the most significant THM compound during risk assessment because of its high concentrations in the chlorinated water supply and its carcinogenic characteristics. During shower, the major exposure pathways of chloroform are inhalation and dermal absorption. Inhalation pathway is due to high volatility of chloroform, which causes high concentration in confined space (shower stall) while showering with hot water. Two different deterministic approaches are used to determine the shower air concentration. The first approach is based on statistical model developed by Kar [2000, Environmental and Health Risk Assessment of Trihalomethanes in Drinking Water – A Case Study. M.Eng Dissertation. Faculty of Engineering and Applied Science, Memorial University of Newfoundland, St. John's, NL]. The second deterministic approach is based on a transient mass balance of the air in the shower stall. For dermal exposure mode, three different approaches (i.e. traditional steady state approach, membrane approach and statistical model by Kar) are used. Ingestion is considered the major pathway to account for the drinking activity. Among the three exposure routes only the inhalation route due to shower activity has predicted risk values several times higher than the generally acceptable risk of 1 per million. Probabilistic risk assessment is also conducted to account for uncertainty and variability in the analysis. The maximum likelihood of worst consequence is also identified by probabilistic methods.