Volatile Organic Compounds in the Nation's Ground Water and Drinking-Water Supply Wells: Supporting Information

Home News FAQ About VOCs Aquifer Studies Occurrence Maps Download Data


Chapter 2

  1. What are VOCs?
  2. How are VOCs Used?
  3. Assessing the Quality of Ground Water
  4. Assessing the Quality of Ground Water Captured by Drinking-Water Supply Wells
  5. What are Assessment Levels, and Why are They Used?
  6. How Do Ground-Water Concentrations from VOC Sources Differ?

Chapter 3

  1. Occurrence Information Helps in Managing Ground-Water Resources
  2. Urban Land Use Contributes More VOCs to Ground Water than Do Other Land Uses
  3. Hydrogeologic Conditions Can Partly Control the Occurrence of VOCs in an Aquifer
  4. Analysis and Reporting at the Principal-Aquifer Scale Help Link National and Local-Scale Findings
  5. Production Rates Do Not Fully Explain VOC Group Detection Frequencies
  6. Local and National Detection Frequencies of VOC Groups Can Differ
  7. Specific VOC Mixtures Occurred Infrequently in Aquifer Samples
  8. Some VOC Detections Could be the Result of the Degradation of a Parent Compound
  9. How were Associations Developed for the Occurrence of VOCs in Aquifers?
  10. Dissolved Oxygen Varies Along a Ground-Water Flowpath

Chapter 4

  1. MCLs Serve as Drinking-Water-Quality Benchmarks for PWSs
  2. HBSLs Can be Applied to VOCs with no MCLs
  3. Most Government Agencies Do Not Require Routine Monitoring of Water Quality for Domestic Wells
  4. Ground Water is Used by Many PWSs
  5. Unregulated Contaminants, Including VOCs, are Monitored in PWSs
  6. VOCs Occur More Frequently in the Water Supply of Very Large CWSs than Other System Sizes
  7. NAWQA's VOC Assessment Can Be Placed in the Context of Other Studies for Public Wells
  8. Specific VOC Mixtures Occurred Infrequently, But Were More Common in Public Well Sa

Chapter 5

  1. Chloroform has a Long History of Use, Research, and Regulation
  2. THMs Result from Water Chlorination
  3. Some THMs Occur Naturally
  4. THMs are Associated with Human-Health Problems
  5. THMs Persist in an Aquifer in the Western United States
  6. Production and Use of Chlorinated Solvents have Declined
  7. Some Solvents are Associated with Human-Health Problems
  8. Solvents Occur Frequently in Drinking Water Supplied by CWSs
  9. What are the Properties of Solvents?
  10. What is MTBE and Why Does it Persist in Ground W
  11. MTBE has been Completely or Partially Banned in Some States
  12. Artificial Recharge Areas are a Source of MTBE to Some California Aquifers
  13. MTBE has been Detected in Drinking Water of Some CWSs
  14. Why are Gasoline Hydrocarbons Detected Infrequently in Aquifers?
  15. Gasoline Hydrocarbons Occur More Frequently in Drinking Water from CWSs than in Domestic Wells

Chapter 2 (Top of page)

1. What are VOCs?
VOCs are a subset of organic compounds with inherent physical and chemical properties that allow these compounds to move between water and air. This behavior is the fundamental basis for the USGS's laboratory analysis of VOCs in water samples, in which compounds that are sufficiently volatile are purged from a water sample by an inert gas and then identified and quantified by gas chromatography/ mass spectrometry (GC/MS). In general, VOCs have high vapor pressures, low-to-medium water solubilities, and low molecular weights. Some VOCs may occur naturally in the environment, other compounds occur only as a result of manmade activities, and some compounds have both origins.

(Top of page)

2. How are VOCs Used?
VOCs have been used extensively in the United States since the 1940s. VOCs are common components or additives in many commercial and household products including gasoline, diesel fuel, other petroleum-based products, carpets, paints, varnishes, glues, spot removers, and cleaners. Example industrial applications include the manufacturing of automobiles, electronics, computers, wood products, adhesives, dyes, rubber products, and plastics, as well as in the synthesis of other organic compounds. VOCs also are used in the dry cleaning of clothing, in refrigeration units, and in the degreasing of equipment and home septic systems. VOCs are present in some personal care products such as perfumes, deodorants, insect repellents, skin lotions, and pharmaceuticals. Some VOCs also have been applied as fumigants in agriculture and in households to control insects, worms, and other pests.

(Top of page)

3. Assessing the Quality of Ground Water
Ground water is an important supply of drinking water in the United States, and the study of aquifers is a large component of NAWQA's ground-water assessments. Aquifer studies have been completed in nearly every NAWQA Study Unit and have provided a comprehensive picture of the chemical quality of water in locally and regionally important aquifers. More information on specific aquifer studies is available on the Circular's Web site. Many pesticides, VOCs, nutrients, and naturally occurring chemicals are monitored in aquifer studies. Typically the aquifer (or portion thereof) selected for study is locally one of the most intensively used aquifers for drinking water. Aquifer studies are designed to provide an overall picture of the aquifer's water-quality condition and, as such, are considered resource assessments. To achieve this spatially large aquifer characterization, wells selected for sampling are randomly located but distributed approximately equally across the study area. A variety of well types with different water uses are included in the assessment of aquifer studies. None of the sampled wells were selected because of prior knowledge of nearby contamination.

(Top of page)

4. Assessing the Quality of Ground Water Captured by Drinking-Water Supply Wells
NAWQA's studies of drinking-water supply wells focus on the quality of ground water captured by domestic and public wells, in contrast to the quality of tap water (that is, drinking water). USGS field personnel collect samples of ground water from domestic and public wells at the wellhead and before any treatment or blending. As such, NAWQA's studies complement drinking-water-compliance- monitoring programs required by other agencies; these programs usually specify monitoring after treatment or blending. Comparisons of concentrations for domestic and public well samples to primary drinking-water standards and Health-Based Screening Levels (HBSLs) in this report are made only in the context of the quality of untreated and unblended ground water. Human exposure from tap water and other pathways is not quantified. During NAWQA's first decade of assessments, many domestic wells and some public wells were sampled. During its second decade, additional emphasis has been placed on understanding the quality of drinking-water supplies including the monitoring of river intakes and production wells of large CWSs, as well as the continued sampling of domestic wells. In addition, major factors that influence the transport of chemicals to public wells are being studied. Studies of drinking-water supplies are important because these studies (1) identify the presence and concentrations of those chemicals that may reach domestic and public wells (or surface-water intakes); and (2) provide information on the need for enhanced source control. Through these studies, the USGS will continue to collaborate with other agencies, organizations, and water utilities involved with the supply of the Nation's drinking water.

(Top of page)

5. What are Assessment Levels, and Why are They Used?
The detection frequency of VOCs in ground water is an important indicator of water quality in occurrence assessments. In order to compare detection frequencies for individual VOCs, groups of VOCs, or VOC data from different agencies with different reporting levels, an "assessment level" must be established. An assessment level is a fixed concentration that is the basis for computing detection frequencies. An assessment level is necessary because the detection frequency computed for a specific VOC depends on the laboratory reporting level for that compound.(21) Laboratory reporting levels for VOCs may vary from compound to compound and from one laboratory to another due to differences in laboratory equipment, equipment sensitivity, experience and skill of equipment operators, or laboratory conditions. In addition, data sets collected for different monitoring objectives or analyzed by different laboratory methods also can have different reporting levels. Thus, different detection frequencies for VOC data sets with different reporting levels may not represent true differences in water quality, but rather they may only reflect the above noted factors.

(Top of page)

6. How Do Ground-Water Concentrations from VOC Sources Differ?
VOC contamination can originate from the release of liquids, such as petroleum hydrocarbons or solvents, at one location. The release of VOCs from a LUST is an example of such contamination and commonly results in concentrations of VOCs in ground water near the source at the milligram or gram per liter level. These large concentrations are one reason why this type of contamination can spread over a large area. Contamination also can originate over large areas from sources such as leaking water and sewer lines, stormwater runoff, and atmospheric deposition. Typically, these sources result in small concentrations (microgram per liter or smaller) in water.

(Top of page)

Chapter 3 (Top of page)


7. Occurrence Information Helps in Managing Ground-Water Resources
The occurrence of VOCs in aquifers provides important information to those responsible for managing ground-water resources. Contamination of aquifers by one or more VOCs also is a national issue of potential concern because of the widespread and long-term use of many of these compounds. Detecting one or more VOCs in aquifer samples provides evidence that conditions favor VOCs reaching the sampled wells. Contaminant occurrence depends on aquifer properties, the associated sources of water to the aquifer, and stresses on the aquifer such as pumping. Contamination also depends on the locations and types of VOC sources, the relative locations of wells, and the transport and fate of VOCs.(27) Knowledge that VOC contamination is present in an aquifer provides the rationale for assessment of the human-health significance of the contamination, as well as the possible need for more in-depth studies to determine the source(s) of contamination and remedial action if concentrations are of potential concern. The occurrence of lowlevel contamination of one or more VOCs in an aquifer also can provide managers with an early indication of the presence of VOCs that eventually might adversely affect the quality of water from domestic and public wells.

(Top of page)

8. Urban Land Use Contributes More VOCs to Ground Water than Do Other Land Uses
Detection frequencies of 1 or more of the 55 VOCs differ in shallow ground water partly depending on the overlying land use-38 percent in residential/commercial urban settings and 11 percent in agricultural settings at an assessment level of 0.2 μg/L. The residential/ commercial findings may be attributable to one or more of several factors related to VOC sources in the urban environment compared to other settings. For example, the urban setting may have more sources and releases of VOCs than other settings. Also, recharge of VOCs to ground water may be enhanced in urban areas by structures such as recharge basins and dry wells. In addition, differences in detection frequencies could be attributable to distance traveled by VOCs and to the transport and fate properties of the VOCs associated with the land-use setting. The finding that urban settings contribute more VOCs to underlying ground water indicates that these waters generally are more vulnerable to VOC contamination than ground water underlying other settings. However, this is not always the case locally. In Oahu, Hawaii, for example, the largest VOC contamination occurs in the agricultural areas of central Oahu, where fumigants have been intensively applied but the aquifers are unconfined, as compared to the minimal contamination underlying urban Honolulu, where the aquifers are somewhat protected by a confining unit.(28)

(Top of page)

9. Hydrogeologic Conditions Can Partly Control the Occurrence of VOCs in an Aquifer
The Edwards aquifer is a sole-source carbonate aquifer used for drinking-water supply in south-central Texas. This aquifer demonstrates the control that hydrogeologic conditions can have on VOC occurrence in an aquifer.(29) VOC detection frequencies in the Edwards aquifer for the unconfined recharge area (61 percent) differed from the confined area (38 percent). The aquifer's recharge area is a faulted and fractured limestone that allows unrestricted downward movement of water and contaminants into the aquifer. The confined part of the Edwards aquifer, however, is overlain by a unit composed of several hundred feet of low-permeability rocks (the Navarro-Del Rio confining unit). This unit restricts the downward movement of water and contaminants to the underlying confined part of the Edwards aquifer, resulting in a smaller VOC occurrence in the confined area than in the unconfined recharge area. Cross section of the Edwards aquifer near San Antonio, Texas, showing the recharge zone and the confined zone of the Edwards aquifer.(29) Figure showing cross-section of Edwards Aquifer

(Top of page)

10. Analysis and Reporting at the Principal-Aquifer Scale Help Link National and Local-Scale Findings
Analysis and reporting of NAWQA's first decade of sampling have focused on national and Study-Unit (local-scale) assessments. Future NAWQA efforts will expand this focus to include analysis and reporting at the principal-aquifer scale. National assessments provide summaries of the national occurrence and distribution of water-quality conditions. However, the large variability in hydrogeologic and other conditions across the Nation often confound the scientist's ability to sort out factors that affect water quality. Study-Unit assessments describe water-quality conditions locally, and often the scientist is able to determine the factors that affect water quality. Extrapolating those findings to beyond the study area often is problematic. Analysis and reporting at an intermediate regional scale, such as by principal aquifer, is intended to help link the findings between the national and local scales. The principal aquifers used as the framework for this intermediate scale of analysis and reporting are located throughout the United States. Sixty-two principal aquifers have been identified as regionally extensive aquifers or aquifer systems that could potentially be used as a source of potable water.(2) NAWQA sampled parts of 33 of these 62 principal aquifers during its first decade of assessments. The principal aquifers vary widely in size, thickness, hydrogeologic properties, yield, and use as drinking-water supplies. Basic descriptions of these principal aquifers and many of their characteristics are available at http://www.nationalatlas.gov.

(Top of page)

11. Production Rates Do Not Fully Explain VOC Group Detection Frequencies
If estimated production rates of VOCs (shown below) alone were the primary governing factor explaining detection frequencies, the gasoline hydrocarbons would be detected in aquifers much more frequently than the other six VOC groups. Likewise, the fumigants, refrigerants, and THMs would have smaller detection frequencies than the other four VOC groups. Comparison of production rates with detection frequencies of VOCs by group (fig. 5) shows that this generally is not the case. Figure showing VOC production rates by group There are many possible reasons for this lack of correspondence between VOC production rates and detection frequency in aquifers. For example, production data (see Circular's Web site) are not available for all VOCs in each group, so actual production could be considerably more than the estimates shown above. In addition, even if the production data were complete, production is not necessarily an exact measure of a VOC source that is contributing a VOC to ground water. Although preferable in this analysis, national data sets of releases for all VOC groups are not available. Additionally, factors such as the hydrogeologic setting, geochemistry of the ground water, and transport and fate properties of VOCs can control the occurrence of VOCs in aquifers (p. 14 and 15).

(Top of page)

12. Local and National Detection Frequencies of VOC Groups Can Differ
Fumigant detections in aquifers on the Island of Oahu, Hawaii, and in the Central Valley of California provide examples of a VOC group with much higher local detection frequencies than the national detection frequency of about 2 percent (assessment level of 0.2 μg/L). The fumigant detections in Oahu are the result of fumigant application to pineapple fields. In 1970, for example, about 1.8 million pounds of fumigants were applied to combat rootworms. (30) Fumigant formulations containing 1,2-dichloropropane, 1,2,3-trichloropropane, and EDB were banned in the late 1970s-early 1980s after about 20 to 30 years of use. (31, 32) In spite of the discontinuation of their use more than 20 years ago, fumigants were detected in more than 30 percent of wells sampled in NAWQA's aquifer study. Fumigants occur in ground water in Oahu because of a combination of factors, including extensive use in recharge areas of the unconfined aquifer in central Oahu, high rainfall that promotes infiltration from the surface, and slow rates of biodegradation. (28) Fumigant detections in the Central Valley of California also are associated with the historical application of a fumigant-in this case, DBCP-on vineyards and almond orchards. DBCP was detected in shallow ground water beneath the vineyards and orchards as well as in the regional aquifer. Detection frequencies of DBCP were as large as 60 percent in shallow ground water beneath the vineyards and orchards and about 10 percent in the regional aquifer.(33)

(Top of page)

13. Specific VOC Mixtures Occurred Infrequently in Aquifer Samples
Specific mixtures of VOCs in the 3,498 aquifer samples occurred relatively infrequently at an assessment level of 0.2 μg/L. Of the 10 most common mixtures, the two most frequently detected VOC mixtures, PCE-TCE and PCE-chloroform, occurred in 1.5 percent of samples (table below). Only one other mixture, TCE-chloroform, occurred in more than 1 percent of the samples. Of the 55 VOCs measured, only 7 compounds (5 solvents and 2 THMs) were found in the 10 most frequently occurring mixtures. Although specific VOC mixtures are a relatively infrequent occurrence at an assessment level of 0.2 μg/L, mixtures do occur more frequently when lower VOC concentrations are considered.(12) table showing detection percentages for different VOCs

(Top of page)

14. Some VOC Detections Could be the Result of the Degradation of a Parent Compound
Some VOCs can degrade through abiotic or biotic processes to another VOC or other compound under oxic and/or anoxic conditions. Several possible degradation by-products are among the 15 most frequently detected VOCs in aquifers. Four of these are (1) methylene chloride from chloroform; (2) chloromethane from methylene chloride; (3) TCE from PCE; and (4) 1,1-dichloroethane (1,1-DCA) from TCA. There is a high degree of co-occurrence of these four by-product/parent VOCs in aquifer samples where the parent compound was detected. Some VOCs can originate both as degradation by-products and from industrial production for use in industrial, commercial, or domestic applications. For VOCs with these dual origins, information on the sources of both the parent and potential by-product to ground water would be helpful to determine whether a detected VOC was a degradation by-product or a result of anthropogenic use.

(Top of page)

15. How were Associations Developed for the Occurrence of VOCs in Aquifers?
Many natural and anthropogenic factors were tested in individual logistic regression models, hereafter termed "statistical models," for 10 frequently detected VOCs. Numerous models were tested for each VOC, but one was selected as the final model on the basis of several statistical measures. Details of these measures and the general procedures for the modeling are described elsewhere. (19) The detections of individual VOCs were significantly associated with particular natural or anthropogenic factors (table 4). In these models, numerous factors are considered together in a single model allowing one to account for differences in one factor (for example, sources) while testing for significance of other factors (for example, dissolved oxygen).

(Top of page)

16. Dissolved Oxygen Varies Along a Ground-Water Flowpath
Dissolved-oxygen concentrations in ground water can vary by location in an aquifer and with the age of the ground water. Young ground water usually has a larger dissolvedoxygen concentration compared to the old ground water. This is because dissolved oxygen can become depleted along a flowpath through various abiotic and biotic processes. Samples collected by NAWQA with age-date information indicate that water recharged after 1955 (referred to here as young ground water) had higher dissolved-oxygen concentrations compared to ground water recharged prior to 1955 (referred to here as old ground water). Because ground water in recharge areas of aquifers is younger than ground water farther along a flowpath, a comparison of detection frequencies of VOCs between young, oxic ground water and old, anoxic ground water should be similar to a comparison of ground water at points along a hypothetical flowpath (fig. 11).

(Top of page)

Chapter 4 (Top of page)


17. MCLs Serve as Drinking-Water-Quality Benchmarks for PWSs
Under the authority of the SDWA, the USEPA establishes drinking-water standards, such as MCLs, to limit the level of contaminants in the Nation's drinking water. An MCL is a legally enforceable standard that sets the maximum permissible level of a contaminant in water that is delivered to any user of a PWS. (40) When setting an MCL, the USEPA also establishes a non-enforceable health goal or Maximum Contaminant Level Goal (MCLG). The MCLG is the maximum level of a contaminant in drinking water at which no known or anticipated adverse effect on human health would occur, and which allows an adequate margin of safety. (40) The MCL is set as close to the MCLG as feasible, taking into account the best available technology, treatment techniques, and cost considerations, as well as expert judgment and public comments. The USEPA reviews drinking-water standards every 6 years to determine if revisions are needed. Established MCLs apply to 29 VOCs included in this NAWQA assessment. However, because MCLs apply to drinking water supplied to the public by PWSs, comparisons of VOC concentrations for samples collected at the well head in this assessment to MCLs are used only to indicate concentrations of potential humanhealth concern. Actual human exposure from drinking water is not described (sidebar 4).

(Top of page)

18. HBSLs Can be Applied to VOCs with no MCLs
HBSLs are estimates of benchmark concentrations of contaminants in water that may be of potential human-health concern. HBSLs are based on health effects alone and have been calculated for unregulated contaminants (those with no MCLs) analyzed by the NAWQA Program. HBSLs were developed by the USGS in collaboration with others (p. 13) using (1) standard USEPA Office of Water methodologies; and (2) the most current, USEPA peer-reviewed, publically available human-health toxicity information. HBSLs are regularly reviewed and, as needed, revised to incorporate the most recent toxicity information and research findings. HBSLs are not regulatory standards and are not legally enforceable. HBSLs were calculated for 15 of the 26 unregulated VOCs in this assessment, but were not calculated for the remaining 11 VOCs due to a lack of toxicity information. Measured contaminant concentrations may be compared to HBSLs to evaluate water-quality data in a humanhealth context. Such comparisons can provide an early indication of when contaminant concentrations in water resources may merit additional study or monitoring. Since 1998, the USGS, in collaboration with others, has made substantial progress in providing additional information about the potential human-health implications of its water-quality findings. USGS will continue its research to develop and refine approaches to expand its ability to evaluate contaminant concentrations in a human-health context at the State and national scales. Additional information about HBSLs and ongoing research is available in other publications (20, 41, 42) and at http://water.usgs.gov/nawqa/HBSL.

(Top of page)

19. Most Government Agencies Do Not Require Routine Monitoring of Water Quality for Domestic Wells
Although regulations vary by State, and also within States, the quality of water from privately owned domestic wells generally is the homeowner's responsibility. Routine monitoring is not required; however, most States and some local agencies provide guidance to domestic well owners through Web sites and printed materials.(45) Raising awareness about the importance of regularly testing private wells is an important step towards ensuring a safe drinking-water supply for the population relying on domestic wells. As such, private well owners are advised by State and local agencies to test water annually to identify possible contaminants such as coliform bacteria, nitrate and nitrite, pesticides, radionuclides, heavy metals, and VOCs, and to compare test results to USEPA and State standards. No States currently require homeowners to take action to improve water quality if contaminants are detected in domestic well water. However, some States have introduced measures to assess water quality to aid in protection of human health. For example, in 2002 New Jersey passed a law that required "raw" or untreated water to be tested in wells included in real estate transactions. (46) Additionally, landlords must test water every 5 years and provide the test results to new tenants. VOCs, including benzene and TCE, were among the required compounds to be tested in the well samples. Because water from domestic wells usually is not treated prior to use, the VOC occurrence data provided by this NAWQA assessment may reflect the quality of tap water used by many rural households. Prior to NAWQA's assessment, no major national studies had been conducted for a large number of VOCs in domestic well samples.

(Top of page)

20. Ground Water is Used by Many PWSs
Nearly 145,000 PWSs provide ground water for human consumption to about 112 million people in the United States. (38) PWS categories established by the USEPA include CWSs and non-community water systems (NCWSs). CWSs serve a residential population such as a municipality, mobile home park, or nursing home. NCWSs are divided into non-transient, non-community water systems (NTNCWSs), such as schools, hospitals, and factories; and transient non-community water systems (TNCWSs), such as campgrounds, motels, and gasoline stations. Nearly 60 percent of PWSs are TNCWSs, but more than 80 percent of the U.S. population is served by CWSs.(38) Graph showing percent of systems and population served by community, non-community and transient water systems Ownership and size of population served by PWSs may vary from very small, privately owned systems whose primary business is something other than water supply (such as mobile home parks) to large, publicly owned water utilities that serve millions of people.(47)

(Top of page)

21. Unregulated Contaminants, Including VOCs, are Monitored in PWSs
The 1996 SDWA Amendments require the USEPA to identify and publish a list of unregulated contaminants (referred to as the CCL) that are known or anticipated to occur in PWSs and that may require regulation with a national primary drinking-water standard.(49) In making regulatory determinations for compounds on the CCL, USEPA must determine whether (1) the contaminant may have an adverse effect on human health; (2) the contaminant is known to occur or there is substantial likelihood that the contaminant will occur in PWSs with a frequency and at levels of public health concern; and (3) in the sole judgment of the Administrator, regulation of such contaminant presents a meaningful opportunity for health risk reduction for people served by PWSs. (50) SDWA requires that the USEPA publish the CCL every 5 years and make regulatory determinations for at least five contaminants (also every 5 years). The Unregulated Contaminant Monitoring (UCM) Program is the mechanism used to collect data for unregulated contaminants suspected to occur in drinking water. (51) The UCM list is revised every 5 years by the USEPA and is based primarily on the CCL. Five VOCs monitored by NAWQA are currently listed on the second CCL published by the USEPA (52) and are prioritized for research and data collection efforts by the USEPA. These VOCs include bromomethane (methyl bromide), 1,1-DCA, 1,3-dichloropropene, MTBE, and 1,2,4-trimethylbenzene. (52) In public well samples collected for this national assessment, the isomers cis- and trans-1,3-dichloropropene were not detected, and bromomethane and 1,2,4-trimethylbenzene were detected in less than 1 percent of the samples. MTBE and 1,1-DCA had the largest detection frequencies of these five VOCs, 5.4 percent and 2.0 percent, respectively (Appendix 10).

(Top of page)

22. VOCs Occur More Frequently in the Water Supply of Very Large CWSs than Other System Sizes
The USGS assisted in a nationwide survey (1999-2002) to characterize the occurrence of VOCs in ground water that served as a drinking-water supply for CWSs. (13) The survey used a statistically stratified design for sampling 575 public wells from randomly selected CWSs representative of the five size categories as follows:
CWS Size Population Served
Very smallless than 500
Small 501 to 3,300
Medium 3,301 to 10,000
Large 10,001 to 50,000
Very large more than 50,000
In general, VOCs were detected most frequently in the very large CWSs. As of 1998, very large CWSs using ground water collectively provided drinking water to about 26 million people. VOC detections were significantly related to urban land use and population density associated with large systems. (13) In particular, detections of gasoline hydrocarbons, solvents, and refrigerants were detected more frequently in ground-water supplies from very large CWSs than from smaller sized systems. figure showing VOC detections by system size: very small, 26 percent; small, 20 percent; medium, 20 percent; large, 24 percent; very large, 42 percent

(Top of page)

23. NAWQA's VOC Assessment Can Be Placed in the Context of Other Studies for Public Wells
This NAWQA study provides one of the few existing national assessments of VOCs in public well samples. The few previous studies of VOC occurrence generally were based on samples of drinking water. One such study was completed by the USEPA during 1981-1982 and focused on 29 VOCs in treated water from CWSs. (8) In this study, VOC concentrations greater than 5 μg/L were found in 2.9 percent of samples from CWSs serving less than 10,000 people and in 6.5 percent of the samples from CWSs serving more than 10,000 people. Although untreated water was sampled for this NAWQA assessment, VOC occurrence in these size categories was similar, 2.4 and 7.8 percent, respectively. A recent study completed in 2002 by the USEPA, (55) based on compliance monitoring data, also provided information on the quality of drinking water from PWSs. In the USEPA study, nine VOCs-benzene, carbon tetrachloride, 1,2-DCA, 1,1-DCE, methylene chloride, 1,2-dichloropropane, PCE, 1,1,2-trichloroethane, and TCE-each had concentrations greater than their MCL in less than 0.7 percent of the PWSs sampled. In public wells sampled by the NAWQA Program, only four of these VOCs-1,1-DCE, methylene chloride, PCE, and TCE-each had concentrations greater than their MCL in less than 0.9 percent of the samples.

(Top of page)

24. Specific VOC Mixtures Occurred Infrequently, But Were More Common in Public Well Samples
All specific mixtures of VOCs occurred in less than 1 percent of domestic well samples, and in less than 4 percent of public well samples at an assessment level of 0.2 μg/L. The table below ranks the 10 most common VOC mixtures by their detection frequency in domestic well samples and in public well samples, and shows the more common occurrence of mixtures in public well samples.

[MTBE, methyl tert-butyl ether; PCE, perchloroethene; TCA, 1,1,1-trichloroethane; TCE, trichloroethene]

Rank VOC mixture Detection
frequency,
in percent
Domestic well samples
1PCE-TCA 0.62
2Chloroform-PCE .50
2Chloroform-MTBE .50
2PCE-TCE .50
5Dibromochloromethane-
chloroform
.42
6Chloroform-TCA .37
6TCA-TCE .37
8PCE-MTBE .33
9Bromoform-
dibromochloromethane
.29
9Bromoform-chloroform .29
Public well samples
1Bromodichloromethane-
dibromochloromethane
3.4
2Bromodichloromethane-
chloroform
3.2
3Bromoform-
dibromochloromethane
2.9
4Dibromochloromethane-
chloroform
2.5
5Bromodichloromethane-
dibromochloromethane-
chloroform
2.4
6PCE-TCE 2.3
7Bromodichloromethane-
bromoform
2.0
7Bromodichloromethane-
bromoform-dibromochloromethane
2.0
9Bromoform-chloroform 1.7
1 Bromoform-
dibromochloromethane-
chloroform
1.6

(Top of page)


Chapter 5 (Top of page)

25. Chloroform has a Long History of Use, Research, and Regulation
figure showing time-line of history of Chloroform

(Top of page)

26. THMs Result from Water Chlorination
THMs are a group of VOCs classified as disinfection by-products. THMs in drinking water were first identified by Rook (77) and are formed as a result of the haloform reaction when dissolved chlorine combines with dissolved organic matter, such as humic and fulvic acids. The disinfection of drinking water in the United States by chlorination commonly uses chlorine gas, sodium hypochlorite, and calcium hypochlorite. The chemistry associated with the chlorination of water and the formation of chloroform and the brominated THMs are described elsewhere. (71) The primary purpose of chlorination of drinking water is to prevent the spread of waterborne diseases, which can include such fatal diseases as cholera and typhoid. Another advantage of chlorination is that a chlorine residual is retained, which provides ongoing disinfection within the distribution system. (78) Although the chlorination of drinking water provides many advantages, THMs remain a human-health concern (sidebar 28). Because of this concern, the USEPA has established an MCL of 80 μg/L for the combined concentrations of four THMs (chloroform, bromodichloromethane, dibromochloromethane, and bromoform), also known as TTHMs.

(Top of page)

27. Some THMs Occur Naturally
Chloroform was originally considered solely of anthropogenic origins; however, several natural sources of chloroform recently have been identified. These include volcanic gases, (79) soil fungi, (64) and marine algae. (80, 81) Although natural sources contribute approximately 90 percent of the total global chloroform flux, (64) evidence in the NAWQA-collected data is inconclusive about whether natural sources contribute to chloroform in ground water. Marine algae have been identified as a natural source for bromodichloromethane, dibromochloromethane, and bromoform. (80)

(Top of page)

28. THMs are Associated with Human-Health Problems
THMs are associated with both acute and chronic human-health problems, including nausea, vomiting, dry mouth, dizziness, headaches, and damage to the liver and kidneys with prolonged exposure. (62) USEPA Office of Water indicates that three THMS-chloroform, bromodichloromethane, and bromoform- are likely to be carcinogenic to humans. Chloroform, however, is likely to be carcinogenic only at high doses. Chloroform is the only potentially carcinogenic THM for which the more stringent, but not regulatory, MCLG has been raised. Historically, the USEPA has established an MCLG of zero for all carcinogens, including chloroform, based on the presumption that any exposure to carcinogens represents a non-zero health risk. Because chloroform is presumed to be non-carcinogenic to humans at concentrations less than those that cause cell regeneration, the USEPA removed the zero MCLG in May 2000 (82) and more recently revised the MCLG to 70 μg/L. (83) Recent studies have shown a weak association between the ingestion of THMs and adverse birth outcomes (such as low birth weight, stillbirths, spontaneous abortions, and neural tube defects); however, many of these studies were considered inconclusive. More systematic epidemiological data on reproductive anomalies, including health effects of mixtures of THMs, and better exposure characterization would improve such studies (84) and aid regulatory agencies in making informed decisions to further protect human health.

(Top of page)

29. THMs Persist in an Aquifer in the Western United States
Chlorinated surface water was injected into an oxic, unconsolidated sand and gravel aquifer in Antelope Valley, California, as part of a program to assess the long-term feasibility of using injection, storage, and recovery as a water-supply method and to reduce waterlevel declines and land subsidence in Antelope Valley. (88) In the sand and gravel aquifer, THM formation continued after the treated water was injected. Once all the residual chlorine reacted, the concentrations of THMs were controlled primarily by mixing and dilution with native ground water. Potential for natural THM attenuation in the aquifer by biodegradation and sorption was low because the aquifer has oxic conditions and low organic matter content. A model used to forecast the effects of repeated cycles of injection, storage, and recovery indicated that the cycles increased concentrations of THMs in the aquifer. These repeated cycles could yield aquifer THM concentrations approaching 100 percent of the injection-water THM concentration within 10 annual cycles, provided mixing within the aquifer does not lower concentrations markedly.

(Top of page)

30. Production and Use of Chlorinated Solvents have Declined
Chlorinated solvents are used in the aerospace and electronics industries, dry cleaning, manufacture of flexible urethane foam, paint removal/stripping, manufacture of pharmaceuticals, metal cleaning and degreasing, and wood manufacturing. (91) Solvents also are present in a variety of household consumer products including drain and pipe cleaners, oven cleaners, shoe polish, household degreasers, deodorizers, leather dyes, photographic supplies, tar remover, waxes, and pesticides. (92) Some solvents, such as carbon tetrachloride, have been used as fumigants in grain storage bins. Production of solvents began in the United States in the early 20th century, and usage of solvents increased markedly after World War II. Production has since declined. Nonetheless, large quantities of solvents continue to be used in industrial, commercial, and domestic applications. For example, methylene chloride is an active ingredient in many formulations of paint removers, and PCE is used by more than 80 percent of commercial dry cleaners. (91)
31. Some Solvents are Associated with Human-Health Problems
Chlorinated solvents have been associated with both cancer and non-cancer humanhealth problems. USEPA Office of Water indicates that both methylene chloride and TCE are probable human carcinogens, although the cancer classification of TCE is under review. TCA is not classifiable for carcinogenicity by the USEPA because of the lack of reported human data and the inadequacy of the available animal studies. The USEPA has not issued any qualitative judgment on the carcinogenicity of PCE. (83) The USEPA currently is reassessing the health effects of all four solvents. The final drafts of the reassessments are expected during 2006-2008. (93) MCLs for drinking water (Appendix 9) have been set for all the solvents considered here. (82) The potential of solvents to affect drinking water is large because the water solubilities of the solvents are much greater than their MCLs. This means that even small spills can result in ground-water concentrations of potential human-health concern.

(Top of page)

32. Solvents Occur Frequently in Drinking Water Supplied by CWSs
Drinking water supplied by CWSs in 12 New England and Mid-Atlantic States had a larger occurrence of solvents compared to samples from domestic wells in the same 12-State area. One or more solvents were detected in about 10 percent of samples from CWSs that supply drinking water from ground water and in about 6 percent of domestic well samples. (70) Individual solvents also were detected more frequently in CWS samples than in domestic well samples with the exception of methylene chloride. Individual solvents were detected much more frequently in large CWSs than in small CWSs. The differences in detection frequencies by CWS size probably are related to larger pumping rates and to more urban land use and higher population density in areas surrounding the supply wells of large CWSs compared to small ones. In drinking water from CWSs, concentrations of three solvents were greater than MCLs more often than other VOCs. Concentrations of PCE, TCE, and methylene chloride were greater than their MCLs in 1.5 percent, 1.2 percent, and 0.2 percent, respectively, of CWS samples.

(Top of page)

33. What are the Properties of Solvents?
In general, chlorinated solvents, as pure liquids, have relatively high densities, vapor pressures, and solubilities. In ground water, they also have relatively long half-lives. All solvents considered here have densities greater than 1. This means that they are denser than water and that releases of pure solvents can penetrate the water table (fig. 27). The relatively high vapor pressures of the four solvents means that these compounds can volatilize when spilled onto a surface or exposed to the atmosphere. The aqueous solubilities of the four solvents also generally are high. (96) Consequently, some mass of solvents exposed to land surfaces can move in solution to the water table. Finally, the biotic half-lives of solvents in ground water are longer than those of other commonly used VOCs, like the gasoline hydrocarbons. (97) This means that solvents biodegrade slowly and therefore can persist for long periods of time in certain aquifers. For more highly chlorinated molecules, the biodegradation of one chlorinated solvent can result in a by-product that also is a chlorinated solvent. For example, PCE can degrade to TCE (fig. 26), especially under anoxic conditions. If this occurs, the detection frequencies and concentrations of PCE should be larger in oxic ground water compared to anoxic ground water. PCE was detected in about 13 percent of samples from oxic ground water, but in only about 6 percent of samples from anoxic ground water. In addition, the concentration ratios of PCE to TCE were three times larger in oxic ground water compared to anoxic ground water. These lines of evidence indicate that some PCE is being degraded to TCE in aquifers with anoxic conditions.

(Top of page)

34. What is MTBE and Why Does it Persist in Ground Water?
MTBE was first introduced in gasoline in 1979 as an octane enhancer resulting from the phase-out of leaded gasoline. (99) Since the mid-1990s, MTBE also has been used in large quantities as an oxygenate in reformulated gasoline. MTBE is an organic compound with ether bonds. The ether bonds and tertiary carbon atom of MTBE make it slow to biodegrade in ground water. graphic showing structure of MTBE molecule The physical properties of MTBE include high water solubility compared to gasoline hydrocarbons, low sorption to organic matter in soil and aquifer material, and a tendency to partition from air into water. MTBE also can undergo significant vapor phase transport in the unsaturated zone. (100, 101, 102) Collectively, these properties can allow MTBE to reach ground water and to travel faster and farther than other common gasoline components. In ground water, MTBE is slow to biodegrade. (102, 103) MTBE is much less biodegradable than the hydrocarbons benzene, toluene, ethylbenzene, and xylenes (BTEX) in ground water; thus, dissolved MTBE can persist longer in aquifer systems relative to BTEX hydrocarbons. (104)

(Top of page)

35. MTBE has been Completely or Partially Banned in Some States
As of June 2004, 19 States had enacted legislation to completely or partially ban MTBE use in gasoline. (111) Also, several Federal bills and resolutions have been introduced in Congress that would ban or limit the use of MTBE nationwide. As of 2004, however, none of the Federal legislation to restrict MTBE use had been enacted. After 1995, most of the gasoline in California contained 2 percent oxygen by weight (11 percent by volume) in order to comply with Federal regulations. (112) Because of its favorable blending and transfer characteristics in gasoline, MTBE was used as the oxygenate in gasoline in California to comply with air-quality standards. However, because of concerns about the occurrence of MTBE in water supplies, an Executive Order was issued by the Governor of California in 1999 to ban the use of MTBE in gasoline by the end of 2002. In 2002, implementation of the ban was extended to the end of 2003. (112) In 2005, Congress passed the Energy Policy Act that eliminated the oxygen requirement in gasoline as established by the Clean Air Act Amendments (CAAA) of 1990. (98) Application of the elimination of the oxygen requirement in gasoline can begin as of the date of the Act (Aug. 8, 2005) for States that received a waiver under the CAAA. Application of the elimination of the oxygen requirement for any other State can begin 270 days after the date of the Act. (98) The elimination of an oxygen requirement is expected to result in less use of oxygenates, such as MTBE, in reformulated gasoline.

(Top of page)

36. Artificial Recharge Areas are a Source of MTBE to Some California Aquifers
In a study of public wells in the Los Angeles Basin (aquifer) of southern California, factors such as population density and the density of LUSTs were not associated with the occurrence of VOCs like MTBE. (114) This information indicates that localized sources like LUSTs and vertical migration of contaminants are not the dominant factors affecting the occurrence of MTBE in this area. Instead, movement of ground water along flowpaths from artificial recharge areas appeared to be the dominant factor in the occurrence of MTBE in active public wells. (114) In general, ground water near the focused recharge areas contained more VOCs like MTBE compared to water deeper in the basins. The occurrence of MTBE close to the areas of recharge is consistent with the relatively recent use of MTBE as a fuel additive and the lateral downgradient flow of recharge water in this area. Further investigation of ground water from public wells in three aquifers in southern California confirmed that MTBE was being introduced into areas of artificial recharge in the aquifers and that concentrations of MTBE decreased along flowpaths in the aquifers. (115)

(Top of page)

37. MTBE has been Detected in Drinking Water of Some CWSs
MTBE was the fifth most frequently detected VOC in CWS samples from 10 New England and Mid-Atlantic States. It was detected in about 9 percent of 985 CWS samples. However, only 0.9 percent of CWSs reported MTBE concentrations greater than 20 μg/L, the lower limit of the USEPA's Drinking-Water Advisory. (70) Most MTBE concentrations in drinking water were less than 5.0 μg/L. The USEPA required monitoring of MTBE in drinking water provided by public water systems under the UCM, which was designed to support drinking-water regulations. As of January 2005, 1,859 public water systems with ground-water sources have been sampled for MTBE under the UCM. Only four systems reported MTBE concentrations greater than 20 μg/L. (116)

(Top of page)

38. Why are Gasoline Hydrocarbons Detected Infrequently in Aquifers?
A variety of physical and chemical properties can limit the movement of gasoline hydrocarbons to and transport by ground water. In ground water, a compound's solubility is one of the most important chemical properties. The median solubility of gasoline hydrocarbons is the lowest of any VOC group. (118) Thus, gasoline hydrocarbons have less chance to reach ground water if released because less mass can be dissolved in recharge water. At a gasoline release site in Beaufort, South Carolina, aerobic biodegradation and volatilization were found to be important in limiting the transport and occurrence of gasoline hydrocarbons in the unsaturated zone. (119) This could help to explain, in part, the lower detection frequency of gasoline hydrocarbons relative to other VOC groups. Another important property of organic chemicals is their ability to adsorb to organic carbon in soil and aquifer material. A good measure of the sorption potential is the organic carbon partitioning coefficient. Gasoline hydrocarbons have the highest median organic carbon partitioning coefficient of any other VOC group. (118) Because of this, once in the subsurface the movement of gasoline hydrocarbons will be retarded with respect to the velocity of ground water. Also, less mass is left in ground water as sorption occurs. Finally, and perhaps most importantly, biodegradation can reduce the mass of gasoline hydrocarbons in the unsaturated zone and in ground water. For aerobic biodegradation, gasoline hydrocarbons have some of the shortest half-lives of VOCs (see Circular's Web site). This means that gasoline hydrocarbons biodegrade quickly in oxic ground water relative to other VOCs.

(Top of page)

39. Gasoline Hydrocarbons Occur More Frequently in Drinking Water from CWSs than in Domestic Wells
Drinking water from CWSs in 12 New England and Mid-Atlantic States (70) had a more frequent occurrence of one or more gasoline hydrocarbons compared to domestic well samples in the same 12-State area. One or more gasoline hydrocarbons were detected in about 8 percent of CWSs deriving drinking water from ground water compared to about 2 percent of domestic well samples. Individual gasoline hydrocarbons also were detected more frequently in CWSs than in domestic well samples, with the exception of isopropylbenzene. The larger detection frequency of gasoline hydrocarbons in CWSs than in domestic well samples is likely due to several factors as discussed previously (p. 40 and 41). Few concentrations of gasoline hydrocarbons in drinking water from CWSs or in domestic well samples were greater than MCLs. One CWS had a concentration of one gasoline hydrocarbon-benzene-that was greater than its MCL. In domestic well samples from the same 12-State area, no concentrations of gasoline hydrocarbons were greater than MCLs.

(Top of page)