ABSTRACT

As part of the National Water-Quality Assessment Program of the U.S. Geological Survey (USGS), water samples were collected during 1991–2004 from domestic wells (private wells used for household drinking water) for analysis of drinking-water contaminants, where contaminants are considered, as defined by the Safe Drinking Water Act, to be all substances in water. Physical properties and the concentrations of major ions, trace elements, nutrients, radon, and organic compounds (pesticides and volatile organic compounds) were measured in as many as 2,167 wells; fecal indicator bacteria and radionuclides also were measured in some wells. The wells were located within major hydrogeologic settings of 30 regionally extensive aquifers used for water supply in the United States. One sample was collected from each well prior to any in-home treatment. Concentrations were compared to water-quality benchmarks for human health, either U.S. Environmental Protection Agency (USEPA) Maximum Contaminant Levels (MCLs) for public water supplies or USGS Health-Based Screening Levels (HBSLs).

No individual contaminant was present in concentrations greater than available health benchmarks in more than 8 percent of the sampled wells. Collectively, however, about 23 percent of wells had at least 1 contaminant present at concentrations greater than an MCL or HBSL, based on analysis of samples from 1,389 wells in which most contaminants were measured. Radon, nitrate, several trace elements, fluoride, gross alpha- and beta-particle radioactivity, and fecal indicator bacteria were found most frequently (in one or more percent of wells) at concentrations greater than benchmarks and, thus, are of potential concern for human health. Radon concentrations were greater than the lower of two proposed MCLs (300 picocuries per liter or pCi/L) in about 65 percent of the wells and greater than the higher proposed MCL (4,000 pCi/L) in about 4 percent of wells. Nitrate, arsenic, manganese, strontium, and gross alpha-particle radioactivity (uncorrected) each were present at levels greater than MCLs or HBSLs in samples from about 5 to 7 percent of the wells; boron, fluoride, uranium, and gross beta-particle radioactivity were present at levels greater than MCLs or HBSLs in about 1 to 2 percent of the wells. Total coliform and Escherichia coli bacteria were detected in about 34 and 8 percent, respectively, of sampled wells. Thus, with the exception of nitrate and fecal indicator bacteria, the contaminants that were present in the sampled wells most frequently at concentrations greater than human-health benchmarks were naturally occurring.

Anthropogenic organic compounds were frequently detected at low concentrations, using typical analytical detection limits of 0.001 to 0.1 micrograms per liter, but were seldom present at concentrations greater than MCLs or HBSLs. The most frequently detected compounds included the pesticide atrazine, its degradate deethylatrazine, and the volatile organic compounds chloroform, methyl tert-butyl ether, perchloroethene, and dichlorofluoromethane. Only 7 of 168 organic compounds were present in samples at concentrations greater than MCLs or HBSLs, each in less than 1 percent of wells. These were diazinon, dibromochloroprane, dinoseb, dieldrin, ethylene dibromide, perchloroethene, and trichloroethene. Overall, concentrations of any organic compound greater than MCLs or HBSLs were present in 0.8 percent of wells, and concentrations of any organic compound greater than one-tenth of MCLs or HBSLs were present in about 3 percent of wells.

Several other properties and contaminants were measured at values or concentrations outside of recommended ranges for drinking water for aesthetic quality (for example, taste or odor) or other non-health reasons. About 16 percent of the sampled wells had pH values less than (14.4 percent) or greater than (1.9 percent) the USEPA recommended range of 6.5 to 8.5. Total dissolved solids were greater than the USEPA Secondary Maximum Contaminant Level (SMCL) of 500 milligrams per liter in about 15 percent of wells. Iron and manganese concentrations were greater than SMCLs in about 19 and 21 percent of wells, respectively. Concentrations of fluoride, which can be harmful at high levels but prevents tooth decay at lower levels, were less than those recommended by the U.S. Centers for Disease Control and Prevention in about 90 percent of the sampled wells.

Geographic patterns of occurrence among principal aquifers showed that several contaminants and properties may be of greater potential concern in certain locations or regions than nationally. For example, radon concentrations were greater than the proposed MCLs in 30 percent (higher proposed MCL) and 90 percent (lower proposed MCL) of wells in crystalline-rock aquifers located in the Northeast, the central and southern Appalachians, and Colorado. Nitrate was present at concentrations greater than the MCL more frequently in agricultural areas than in other land-use settings. Contaminant concentrations also were related to geochemical conditions. For example, uranium concentrations were correlated with concentrations of dissolved oxygen in addition to showing regional patterns of occurrence; relatively high iron and manganese concentrations occurred everywhere, but were inversely correlated with dissolved oxygen concentrations.

Mixtures of two or more contaminants at concentrations greater than human-health benchmarks occurred in only about 4 percent of wells, but mixtures of two or more contaminants with concentrations greater than one-tenth of their benchmarks occurred in about 73 percent of wells. The more complex mixtures, with the largest numbers of contaminants, were most common in several aquifers in the western and south-central United States in ground water with high concentrations of dissolved solids overall. Two-thirds of the unique mixtures of contaminants with concentrations greater than one-tenth of their benchmarks that occurred in 5 percent or more of wells were composed of two or more of six contaminants—nitrate, arsenic, radon, and uranium, and to a lesser extent, molybdenum and manganese. Organic compounds were rarely components of unique mixtures of contaminants at concentrations greater than human-health benchmarks or greater than one-tenth of their benchmarks. However, mixtures of naturally occurring contaminants at concentrations greater than one-tenth of their benchmarks and organic compounds detected at any concentration were common, occurring in about 90 percent of wells. Several combinations of organic compounds in mixtures with possible health effects were identified—specifically, atrazine and deethylatrazine, atrazine or simazine with nitrate, and perchloroethene and three other solvents—but combined concentrations either were less than the health benchmarks or no benchmarks were available for the mixtures. These co-occurrences may be a potential concern for human health, but the long-term cumulative effects of low concentrations of multiple contaminants on human health currently are unknown.

---

Suggested citation:

DeSimone, L.A., 2009, Quality of water from domestic wells in principal aquifers of the United States, 1991–2004: U.S. Geological Survey Scientific Investigations Report 2008–5227, 139 p., available online at http://pubs.usgs.gov/sir/2008/5227.

---

Contents

Foreword

Abstract

Introduction

What is a Contaminant?

Methods

Well Selection

Sample Collection and Analysis

Quality Assurance and Quality Control

Data Sources and Treatment

Statistical Methods and Data Presentation

Water-Quality Benchmarks for Human Health

Domestic Wells Sampled for the NAWQA Program, 1991–2004

Quality of Water from Domestic Wells

Physical Properties and Dissolved Solids

Major Ions

Trace Elements

Nutrients and Dissolved Organic Carbon

Radon and Other Radionuclides

Organic Compounds

Fecal Indicator Bacteria

Summary Comparison of Contaminant Concentrations to Human-Health Benchmarks

Co-occurrence of Contaminants as Mixtures

Approach for Assessment of Mixtures

Occurrence and Distribution of Mixtures

Composition of Mixtures

Assessing the Potential Significance of Mixtures to Human Health

Comparison with Previous Studies of Domestic Well Water Quality

Summary and Conclusions

Acknowledgments

References

APPENDIXES IN BACK OF REPORT:

Appendix 1. Summary of selected studies of domestic well water quality in the United States.

Appendix 2. Characteristics, principal aquifers, study units, and assessment studies of domestic wells sampled for the NAWQA Program, 1991–2004, by aquifer type.

Appendix 3. Sampling and analytical methods and summary of field quality-control data for chemical contaminants measured in domestic wells sampled for the NAWQA Program, 1991–2004.

Appendix 4. Physical properties and contaminants analyzed in samples collected from domestic wells for the NAWQA Program, 2001–2004, and associated human-health benchmarks for drinking water.

APPENDIXES ON CD-ROM IN BACK OF REPORT:

Appendix 5. Summary statistics for physical properties and concentrations of major ions, trace elements, nutrients, radionuclides, and fecal indicator bacteria in samples collected from domestic wells for the NAWQA Program, 1991–2004.

Appendix 6. Percentages of wells in principal aquifers with values of physical properties or concentrations of major ions, trace elements, nutrients, radionuclides, or organic compounds greater than human-health benchmarks or guidelines for drinking water in samples collected from domestic wells for the NAWQA Program, 1991–2004.

Appendix 7. Detection frequencies of organic compounds at any concentration and at several common reporting levels in samples collected from domestic wells for the NAWQA Program, 1991–2004.

Revised November 2009

First posted March 20, 2009

Report PDF (12.1 MB)