Estimating Exposure to Chemical Contaminants in Drinking Water

• Published in: Environmental science & technology Vol. 38; no. 6; pp. 1799 - 1806
• Main Authors: Kim, Eunyoung, Little, John C, Chiu, Nancy
• Format: Journal Article
• Language: English
• Published: Washington, DC American Chemical Society 15.03.2004
• Subjects: Adsorption; Biological and medical sciences; Chemical contaminants; Drinking water; Environmental Exposure; Estimating techniques; Food toxicology; Forecasting; Housing; Humans; Inhalation Exposure; Medical sciences; Models, Theoretical; Monte Carlo Method; Toxicology; Ventilation; Volatilization; Water Pollutants - analysis; Water Supply; United States; North America; America; Human; Drinking water; Insecticide; Pesticides; Oral administration; Tap water; Housing; Modeling; Inhalation; Heavy metal; Percutaneous route; Organochlorine compounds; Lead; Public health; Topical administration; Chemical contamination
• ISSN: 0013-936X; 1520-5851
• DOI: 10.1021/es026300t

A model is developed that predicts exposure and absorbed dose for chemical contaminants in household drinking water via three pathways:  inhalation, direct and indirect ingestion, and dermal penetration. Extensive probability distributions for building characteristics, activity and water use patterns, operating conditions of water devices, and physiological characteristics of the general population are developed. The impacts of different operating conditions on mass transfer coefficients for the shower, bath, washing machine, dishwasher, and faucet are established. Dichlorobromomethane, inorganic lead, and endosulfan, three compounds associated with adverse birth outcomes that have significantly different chemical properties, are selected for analysis. The primary exposure pathways for dichlorobromomethane are inhalation (62%) and ingestion (27%). Seventy percent of total exposure to endosulfan comes from ingestion, and 18% from dermal sorption with the remaining 12% due to inhalation. Virtually all (99.9%) of the exposure to lead occurs via ingestion. A nested Monte Carlo analysis shows that natural variability contributes significantly more (a factor of 10) toward total uncertainty than knowledge uncertainty (a factor of 1.5). Better identification of certain critical input variables (ventilation rate in the shower and bathroom, ingestion rate, the boiling water mass transfer coefficient, and skin permeability) is required.