Statistical Models for the Concentrations of Chemicals in Source and Treated Drinking Water Supplies

EPA Grant Number: R826890
Title: Statistical Models for the Concentrations of Chemicals in Source and Treated Drinking Water Supplies
Investigators: Schervish, Mark J. , Small, Mitchell J.
Institution: Carnegie Mellon University
EPA Project Officer: Louie, Nica
Project Period: September 1, 1998 through August 31, 2001 (Extended to September 30, 2002)
Project Amount: $250,000
RFA: Environmental Statistics (1998) RFA Text |  Recipients Lists
Research Category: Health , Ecosystems , Environmental Statistics


Improvements in Risk Assessment: The distributions of contaminant concentrations and the distributions of treatment efficiencies should help the EPA to determine the sources and degrees of uncertainty associated with some of the information and assumptions that they use in making decisions for drinking water regulation.

Objectives: We propose to develop statistical models for the concentrations of multiple contaminants in both raw and finished drinking water sources. These models will be used to quantify the variability across the United States of the concentrations of these contaminants as well as the uncertainty associated with estimates of concentration. The key components will include models for concentrations in raw water, models for that choice of treatment to be used and models for the efficiency of the treatments. These models can be combined to produce a model for concentrations in finished water.


We will model the way that various characteristics such as size of system, geographical location and water source affect concentrations, choice of treatment and treatment efficiency. We will create Bayesian hierarchical statistical models for each of the components and use several available data sets to form posterior and predictive distributions for quantities of interest.

Expected Results:

We expect to produce joint distributions for concentrations of contaminants that can be used by the EPA for assessing the potential impacts of proposed maximum contaminant levels mandated by the Safe Drinking Water Act of 1996. We also expect to develop new statistical methodology for dealing with multiple data sets and sources of expert opinion that each impact on different aspects of the overall research problem.

Publications and Presentations:

Publications have been submitted on this project: View all 6 publications for this project

Journal Articles:

Journal Articles have been submitted on this project: View all 1 journal articles for this project

Supplemental Keywords:

Co-occurrence,, RFA, Scientific Discipline, Economic, Social, & Behavioral Science Research Program, Toxics, Water, POLLUTANTS/TOXICS, Environmental Chemistry, Chemistry, Contaminant Candidate List, Arsenic, Environmental Statistics, Water Pollutants, Drinking Water, Environmental Engineering, monitoring, public water systems, data synthesis, CCL, Bayesian space-time model, co-pollutant effects, exposure and effects, Bayesian hierarchical statistical models, environmental risks, predictive distributors, exposure, other - risk assessment, community water system, drinking water supplies, treatment, statistical models, maximum contaminant levels, data analysis, drinking water contaminants, water treatment, data models, hierarchical statistical analysis, innovative statistical models, chemical concentrations, drinking water treatment, drinking water system, multiple contaminants

Relevant Websites: Exit EPA icon

Progress and Final Reports:

  • 1999 Progress Report
  • 2000 Progress Report
  • 2001
  • 2002
  • Final