Occurrence Summary and Use
     Support Document /
  for the Six-Year Review of
National Primary Drinking Water
        Regulations

-------
Office of Water (4606)
EPA-815-D-02-006
www.epa.gov
March 2002
                  ^^ Printed on Recycled Paper

-------
                               Disclaimer

  This report is issued in support of the preliminary revise/not revise
  decisions for the Six-Year Review Notice of Intent. It is intended for
  public comment and does not represent final agency policy.  EPA expects
  to issue a final version of this report with the publication of the
  final notice in 2002, reflecting corrections due to public comment on
  the preliminary notice and the supporting documents.

  Mention of trade names or commercial products does not constitute
  endorsement or recommendation for use.
Occurrence Summary and Use Support Document                                        March 2002

-------
                                  ACKNOWLEDGMENTS


The compilation and analysis of the data summarized in this report were undertaken by EPA's Office of
Ground Water and Drinking Water (OGWDW) in support of the Six-Year Review of National Primary
Drinking Water Regulations.  This effort was directed by Mr. Peter Lassovszky, with contributions and
reviews by Ms. Wynne Miller, of OGWDW.

We would like to thank the many States and State staff who contributed data sets and valuable advice.
Thank you also to the many public water systems and their staff who conducted the monitoring that
provided the extensive amount of contaminant occurrence data used in this report.

EPA also appreciates the technical support provided by The Cadmus Group, Inc., the prime contractor for
this project.  Dr. Jonathan Koplos and Dr. George Hallberg served as Cadmus Project Managers in the
provision of support for background research, contaminant occurrence analyses, and report development.
The major contributions of Erin Hartigan, Alison Kotros, and Jane Parkin are gratefully acknowledged.
Occurrence Summary and Use Support Document          111                                    March 2002

-------
                                TABLE OF CONTENTS

1.0 Introduction	  1

       1.1  Purpose and Scope	  1
       1.2  Review Process  	  1
       1.3  Data Sources 	  2
       1.4  Occurrence Analysis  	  9
       1.5  References 	  11

2.0 Inorganic Contaminants (lOCs) 	  13

       2.1  Beryllium	  15
       2.2  Chromium 	  30
       2.3  Fluoride 	  56
       2.4  Mercury 	  81
       2.5  Thallium	  98

3.0 Synthetic Organic Contaminants (SOCs)	  Ill

       3.1  Alachlor	  112
       3.2  Bis(2-ethylhexyl)phthalate (DEHP)  	  130
       3.3  Carbofuran	  142
       3.4  Chlordane	151
       3.5  l,2-Dibromo-3-chloropropane (DBCP)	  170
       3.6  Diquat	  182
       3.7  Glyphosate	  192
       3.8  Heptachlor & Heptachlor Epoxide  	 200
       3.9  Hexachlorobenzene  	 219
       3.10 Hexachlorocyclopentadiene	 230
       3.11 Oxamyl	 240
       3.12 Picloram  	 251
       3.13 Simazine	 262
       3.14 Toxaphene 	 280

4.0 Volatile Organic Contaminants (VOCs)	 294

       4.1  Benzene 	 295
       4.2  Carbon Tetrachloride	 312
       4.3  1,4-Dichlorobenzene  	 331
       4.4  1,2-Dichloroethane  	 346
       4.5  1,1-Dichloroethylene	 366
       4.6  Dichloromethane  	 382
       4.7  1,2-Dichloropropane  	 393
       4.8  Tetrachloroethylene	 406
       4.9  1,1,2-Trichloroethane	 425
       4.10 Trichloroethylene	 438

Abbreviations and Acronyms   	 453
Occurrence Summary and Use Support Document          iv                                     March 2002

-------
1.0    Introduction



Table of Contents

1.1  Purpose and Scope  	  1

1.2  Review Process	  1

1.3  Data Sources	  2
       1.3.1  Agency for Toxic Substances and Disease Registry	  2
       1.3.2 Toxic Release Inventory	  2
       1.3.3  National Water Quality Assessment	  3
       1.3.4 The National Highway Runoff Data and Methodology Synthesis	  5
       1.3.5  16-State Cross-Section Data	  5
               1.3.5.1  Data Management	  6
       1.3.6 Additional Data	  8

1.4  Occurrence Analysis	  9

1.5  References	  11


List of Tables & Figures

Figure 1.3-1.  Map of the 16 Cross-Section States 	  6
Table 1.3-1. Contaminant Occurrence Data From the 16-State Cross-Section 	  7
Occurrence Summary and Use Support Document          v                                     March 2002

-------
1.1 Purpose and Scope

This document summarizes contaminant occurrence findings for 30 regulated contaminants in support of
the Environmental Protection Agency's (EPA's) Six-Year Review of National Primary Drinking Water
Regulations ("Six-Year Review").  Included is detailed information regarding each contaminant's
occurrence in drinking water and related information relevant to initial exposure assessments. Brief
reviews regarding each contaminant's production, uses and occurrence in ambient water are also
included.  To provide some regulatory context, the Six-Year Review process is briefly described.

This document is divided into four sections. This section (Section 1) contains general information
- applicable to all contaminants- which includes a description of all data sources used for this analysis
and the statistical methodology used to estimate national occurrence from the data that are national in
scope. The following three sections are organized by contaminant group. Section 2 provides information
for all inorganic contaminants (lOCs), Section 3 provides information for all synthetic organic
contaminants (SOCs), and  Section 4 provides  information for all volatile organic contaminants (VOCs).
(See Page ii for a list of the contaminants included in each section.)  Within each contaminant group
section there is a detailed Table of Contents specific to each contaminant. For a general overview or
summary of a particular contaminant's occurrence findings, please refer to the conclusion section for that
specific contaminant.

1.2 Review Process

The Environmental Protection Agency's Office of Ground Water and Drinking Water (OGWDW) is
responsible for implementing the provisions of the Safe Drinking Water Act (SDWA). The 1996 SDWA
amendments require the U.S. EPA to review existing National Primary Drinking Water Regulations
(NPDWRs) no less often than every six years  and, if appropriate, revise them. As long as an NPDWR
revision maintains or provides for greater protection of public health, the SDWA 1996 amendments give
the Administrator discretion to determine if revision is appropriate.  EPA believes the revision must
continue to meet the basic statutory requirements of the SDWA (e.g., generally setting the maximum
contaminant level as close to the maximum contaminant level goal as is feasible), to determine if a
revision is appropriate.  EPA also believes the revision must present significant opportunities to improve
the level of public health protection and/or to achieve cost savings while maintaining, or improving, the
level of public health protection. The  Six-Year Review workgroup is reviewing  the factors relevant to
this formal re-assessment of each national primary drinking water regulation.

EPA developed a protocol document — EPA Protocol for the Review of Existing National Primary
Drinking Water Regulations (USEPA, 2001) — to describe the process and strategy for Review that EPA
will use to meet its statutory requirement.  To most efficiently utilize limited resources, EPA plans to
perform a series of analyses at the beginning of each Review cycle, intended to target those NPDWRs
that are the most appropriate candidates for revision. The Agency plans to use available,  scientifically-
sound data to make decisions regarding whether or not to revise a regulation. EPA will review the
following key information to make decisions regarding regulatory changes:  current health risk
assessments, technology assessments (including reviews of laboratory analytical methods and treatment
techniques), and occurrence and exposure assessments.  This current document presents information
specific to occurrence and exposure in support of the Six-Year Review.

EPA will consider regulatory revisions based on the various components of each primary  drinking water
regulation, including possible changes to Maximum Contaminant  Levels (MCLs), Maximum
Contaminant Level Goals (MCLGs), treatment techniques,  analytical method capabilities, and treatment
capabilities. In some cases, EPA may also consider revisions to monitoring or system reporting
requirements as part of the  Six-Year Review; however, in most cases, these types of revisions will be
considered through other vehicles. In rare instances, EPA may consider dropping a contaminant from
regulation. For any NPDWR that is a potential candidate for revision based on its review, EPA will also
take economic considerations into account before making its "revise/not revise"  decision. Moreover,
EPA will apply basic risk management principles to determine whether these candidate regulations


Occurrence Summary and Use Support Document          1                                      March 2002

-------
warrant regulatory revision to ensure that any revision will present a significant opportunity to improve
the level of public health protection and/or present a significant opportunity for cost savings while
maintaining, or improving, the level of public health protection.

1.3 Data Sources

Numerous data sources were used in this report to provide information on contaminant use, production,
occurrence, and potential exposure. The primary data sources used for background use, production, and
release reviews include the Agency for Toxic Substances and Disease Registry (ATSDR) and Toxics
Release Inventory (TRI).  Information on contaminant occurrence in ambient water primarily derives
from the National Water Quality Assessment Program (NAWQA).  Drinking water contaminant
occurrence analyses are generated from State drinking water compliance monitoring data from a national
cross-section comprised of occurrence data from 16 States.  This  16-State national cross-section is the
largest compliance monitoring data set compiled to date by EPA.  These primary data sources are
described in detail in the subsequent sections of this Introduction. In addition to the primary data
sources, supplemental contaminant occurrence information for drinking water and ambient water is also
reviewed whenever available.  These supplemental data sources and findings are included in the
contaminant specific sections of this report.

1.3.1  Agency for Toxic Substances and Disease Registry

In 1980, Congress created the Agency for Toxic Substances and Disease Registry (ATSDR) to
implement the health-related sections of laws that protect the public from hazardous wastes and
environmental  spills of hazardous substances (ATSDR, 2001).  The Comprehensive Environmental
Response, Compensation, and Liability Act of 1980 (CERCLA),  commonly known as the "Superfund"
Act, provided the Congressional mandate to remove or clean up abandoned and inactive hazardous waste
sites and to provide federal assistance in toxic emergencies. As the lead Agency within the Public Health
Service for implementing the health-related provisions  of CERCLA, ATSDR is charged under the
Superfund Act to assess the presence and nature of health hazards at specific Superfund sites, to help
prevent or reduce further exposure and the illnesses that result from such exposures, and to expand the
knowledge base about health effects from exposure to hazardous  substances (ATSDR, 2001).

In the 1984 amendments to the Resource Conservation and Recovery Act of 1976  (RCRA), which
provides for the management of legitimate hazardous waste storage or destruction facilities, ATSDR was
authorized to conduct public health assessments at these sites, when requested by EPA, States, or
individuals.  ATSDR was also authorized to assist EPA in determining which substances should be
regulated and the levels at which substances may pose a threat to  human health (ATSDR, 2001).

With the passage of the Superfund Amendments and Reauthorization Act of 1986  (SARA), ATSDR
received additional responsibilities in environmental public health.  This act broadened ATSDR's
responsibilities in the areas of public health assessments, establishment and maintenance of toxicologic
databases, information dissemination, and medical education (ATSDR, 2001).

ATSDR issues  Toxicological Profiles for over 250 substances, including 23 of the  30 contaminants
discussed in this report. These profiles contain exhaustive reports on the substances' health effects,
chemical and physical properties, use and production, potential for human exposure, and analytical
methods. Whenever available, ATSDR was used as a primary source in this report for contaminant use
and production information.

1.3.2  Toxic  Release Inventory

The Toxics Release Inventory  (TRI) is a publicly available EPA database that contains information on
specific toxic chemical releases and other waste management activities reported annually by certain
covered industry groups as well as federal facilities. Facilities are required to use their best readily
available data to calculate their releases and waste management estimates. If facilities do not have actual


Occurrence Summary and Use Support Document          2                                      March 2002

-------
monitoring data, submitted values are derived from various estimation techniques. Reporting Year (RY)
1999 is the most recent TRI data available.

In 1986, the Emergency Planning and Community Right-to-Know Act (EPCRA) established the TRI of
hazardous chemicals. Through EPCRA, Congress mandated that larger facilities publicly report when
TRI chemicals are released into the environment. TRI provides citizens with accurate information about
potentially hazardous chemicals and their use so that communities have more power to hold companies
accountable and make informed decisions about how toxic chemicals are to be managed.  This public
reporting is required for facilities with more than 10 full-time employees that annually manufacture or
produce more than 25,000 pounds, or use more than 10,000 pounds, of TRI chemical (USEPA, 1996;
USEPA, 2000c). Under these conditions, facilities are required to report the pounds per year of the
contaminant released into the environment both on- and off-site.  The on-site quantity is subdivided into
air emissions, surface water discharges, underground injections, and releases to land.

Although the TRI data can be useful in giving a general idea of release trends, it is far from exhaustive
and has significant limitations.  For example, only industries which meet TRI criteria (at least 10 full-
time employees and manufacture and processing of quantities exceeding 25,000 Ibs/yr, or use of more
than 10,000 Ibs/yr) are required to report releases. These reporting criteria, therefore, do not account for
releases from smaller industries. Threshold manufacture and processing quantities also changed from
1988 to  1990 (dropping from 75,000 Ibs/yr in 1988 to 50,000 Ibs/yr in 1989 to its current 25,000 Ibs/yr in
1990) creating possibly misleading data trends. Also, reported releases are annual estimates. The
amounts reported could have been released evenly over the course of the year or, possibly, in a single
large burst. Finally, the TRI data is meant to reflect releases and should not be used to estimate general
exposure to a chemical (USEPA, 2000a; USEPA, 2000b).

Twenty-six of the  30 contaminants discussed in this report are listed as Toxic Release Inventory (TRI)
chemicals. Whenever available, TRI data were used to describe environmental releases of each
contaminant; the distribution of these releases (with respect to the method of release and overall
geographic distribution); and the geographic distribution of the releases relative to the 16 cross-section
States.

1.3.3  National Water Quality Assessment

Although the Six-Year Review assesses contaminant occurrence in finished drinking water (assessing
occurrence in water sampled from within public water systems), an evaluation of contaminant occurrence
in ambient (raw, unfinished) water provides background information regarding the presence of a
contaminant in the environment. In a drinking water context, ambient water is source water existing in
surface waters and aquifers before treatment.  The most comprehensive and nationally consistent data
describing ambient water quality in the United States are being collected and managed through the efforts
of the United States Geological Survey's (USGS) and its National Water Quality Assessment (NAWQA)
program.  (NAWQA, however, is a relatively young program and complete national data are not yet
available from their entire array of sites across the nation. Also, data is not yet available on all the
contaminants in this report.  As the NAWQA program continues, important addition spatial and temporal
coverage of this ambient occurrence data will become available for occurrence analyzes.)

The USGS instituted the NAWQA program in 1991 to examine water quality status and trends in the
United States.  NAWQA is designed and implemented in such a manner to allow consistency and
comparison between representative study basins located around the country, facilitating interpretation of
natural and anthropogenic factors affecting water quality (Leahy and Thompson, 1994).

The NAWQA program consists of 59 significant watersheds and aquifers referred to as "study units."
The study units represent approximately two thirds of the overall water usage in the United States and a
similar proportion of the population served by public water systems. Approximately one half of the
nation's land area is represented (Leahy and Thompson, 1994).
Occurrence Summary and Use Support Document          3                                      March 2002

-------
To facilitate management and make the program cost-effective, approximately one third of the study
units at a time engage in intensive assessment for a period of 3 to 5 years. This is followed by a period of
less intensive research and monitoring that lasts between 5 and 7 years. This way all 59 study units
rotate through intensive assessment over a ten-year period (Leahy and Thompson,  1994). The first round
of intensive monitoring (1991-96) targeted 20 watersheds and the second round monitored 16 basins
beginning in 1994.

Whenever available, ambient data for the inorganic contaminants were downloaded directly from the
USGS website.  These data follow the parameters and guidelines discussed above. No particular
limitations apply specifically to the  downloaded IOC data.

For pesticides (classified as SOCs), NAWQA includes analyses of 76 pesticides and 7 selected pesticide
degradation products in about 8,200 samples of ground water and surface water in 20 of the nation's
major hydrologic basins (USGS, 1998).  The 76 herbicides, insecticides, and fungicides targeted in the
study account for approximately 75 percent of the total amount (by weight) of pesticides used for
agriculture in the U.S., and also a substantial portion of urban and suburban use. The minimum reporting
limit (MRL) for the SOCs varied by contaminant.

For this Six-Year Review report, data for pesticide occurrence  in ambient water was used from Pesticides
in Surface and Ground Water of the United States: Summary of Results of the National Water Quality
Assessment Program (USGS,1998).  This published report includes the USGS analyzed NAWQA sample
data from 1992 to 1996.  (More recent NAWQA sampling round results have recently become available
and include sample results data from 1996 to 1998. However,  the data so far published are not as
comprehensive as the 1992-1996 data, and the more comprehensive data, while available on the USGS
website, has not been completely quality reviewed or analyzed by USGS. Therefore, this current Six-
Year Review occurrence assessment uses the data from the published 1998 report.  In the subsequent
draft of this Six-Year Review report, the 1996 to 1998 NAWQA data will be  reviewed and potentially
incorporated.)

For volatile organic chemicals (VOCs), the national synthesis will compile data from the first and second
rounds of intensive assessments.  Study units assessed in the second round represent conditions in more
urbanized basins, but initial results are not yet available.  However, VOCs were analyzed in the first
round of intensive monitoring and data are available for these study units (Squillace et al., 1999).  The
minimum reporting limit (MRL) for all VOCs listed in this report was 0.2 (ig/L (Squillace et al., 1999).
Additional information on  analytical methods used in the NAWQA study units, including method
detection limits, are described by Gilliom and others (1998).

Furthermore, the NAWQA program has  compiled, by study unit, data collected from local, State, and
other Federal agencies to augment its own data. The data set provides an assessment of VOCs in
untreated ambient ground water of the conterminous United States for the period 1985-1995 (Squillace et
al.,  1999).  Data were included in the compilation if they met certain criteria for collection, analysis, well
network design, and well construction (Lapham et al.,  1997).  They represent both rural and urban areas,
but should be viewed as  a progress report as NAWQA data continue to be collected that may influence
conclusions regarding occurrence and distribution of VOCs (Squillace et al.,  1999).

In addition to the NAWQA studies/data described above, information is also  presented regarding
pesticide concentrations  in reservoirs and finished drinking water (Blomquist, et al., 2001) . In 1999, a
pilot monitoring program was initiated by USGS and EPA to provide information on pesticide
concentrations in drinking  water. Prior to implementation of this pilot program, there were few available
datasets that contained information on pesticide concentrations in finished drinking water on a national
scale, as most available data sets generally cover only selected compounds or local areas. This pilot
program was implemented to begin to fill this important data gap, and to provide more information about
appropriate methods for  a national monitoring of pesticides in drinking water. The NAWQA data sets
provide the only nationally consistent pesticide concentration data for a large suite of compounds. OPP
currently uses the NAWQA data in their drinking-water and aquatic-exposure assessments; however,


Occurrence Summary and Use Support Document          4                                      March 2002

-------
these data are not collected from public water supplies and, therefore, may not directly reflect human
exposure to pesticides via drinking water.

Drinking water utilities that withdrew from reservoirs were sampled because reservoirs are vulnerable to
pesticide contamination, integrate pesticide loading from their watersheds, and show smaller temporal
variations than streams.  Sampling frequencies were designed to measure long-term mean and short-term
peak concentrations of pesticides in drinking water.  Samples were collected quarterly throughout the
year and at weekly or bi-weekly intervals following the primary pesticide application periods. Water
samples were collected from the raw-water intake and from the finished drinking-water tap prior to
entering the distribution system. At some sites, samples were also collected at the reservoir outflow.
Three different laboratory methods were selected for use during the pilot monitoring program— a USGS
approved method and two developmental methods. Only results from the USGS approved method
(known as "method 2001" or "schedule 2001") are presented in this report.

Twelve water-supply reservoirs were sampled. The sites were in California, Indiana, Ohio, Oklahoma,
Louisiana, Missouri, South Carolina, South Dakota, New York, North Carolina, Pennsylvania, and
Texas. In 1999, drought conditions affected parts of the Eastern United States and California; therefore
sampling was extended through 2000 at nine sites. This report presents results for alachlor, carbofuran,
and simazine in both raw and finished water.

1.3.4 The National Highway Runoff Data and Methodology Synthesis

The National Highway Runoff Data and Methodology Synthesis has reviewed 44 highway and urban
runoff studies implemented  since 1970 (Lopes and Dionne, 1998). Two national studies were included in
this review: the National Urban Runoff Program (NURP) and studies associated with the USEPA
National Pollution Discharge Elimination System (NPDES) municipal stormwater permits. NURP,
conducted in the 1970s and  early 1980s, had the most extensive geographic distribution.  The NPDES
studies took place in the early to mid- 1990s (Lopes and Dionne, 1998).

1.3.5 16-State Cross-Section Data

The State  public water system (PWS) compliance monitoring occurrence data used in this analysis were
submitted by States for EPA review and study of the occurrence of regulated contaminants in PWSs (see
USEPA, 1999). In  the USEPA (1999) review, all 50 States were evaluated through a methodology that
included ranking of States' pollution potential, dividing States into quartiles based on these rankings, and
then selecting States that equally represent the four pollution potential quartiles. Another factor
considered (when selecting  states equally across the four pollution potential quartiles) was selection of
states distributed geographically to include  a broad representation of climatic and hydrologic variability
across the United States. In this way, a subset, or cross-section, of States could be selected to reflect a
national representation of pollution potentials and climatic/hydrologic difference.

An initial  national cross-section of 8 States was selected because it provided a balanced national cross-
section of State occurrence data, was based only on high quality and adequately complete State
occurrence data sets, and that, in aggregate, is indicative of national contaminant occurrence. This
national cross-section development and use methodology was the subject of critical internal review and
external peer review in the USEPA (1999) study.  An additional group of 8 States were added to build
upon and expand the coverage of the 8-State cross-section of occurrence data (USEPA, 2002). The
additional 8 States were selected using the same criteria for selection of the initial  8 cross-section States,
and therefore were  added in a manner that maintained the pollution potential and geographic national
balance of the cross-section. The analyses presented in this report are based on data from the final group
of 16 States, including Alabama, California, Florida, Illinois, Indiana, Kentucky, Michigan, Montana,
Nebraska, New Jersey, New Mexico, Oregon, South Carolina, South Dakota, Texas, and Vermont (see
Figure 1.3-1).
Occurrence Summary and Use Support Document          5                                      March 2002

-------
Figure 1.3-1.  Map of the 16 Cross-Section States
                                                          |    | Initial 8-State Cross-Section
                                                          |    | Additional 8-State Cross-Section
                                                          |    | States not included in Cross-Section
The 8 initial plus 8 additional States that are shaded in the figure comprise the 16-State cross-section used for analysis.
1.3.5.1  Data Management

A brief discussion of data management is included below; for further detailed discussion of the extensive
data management of the State data sets used in the analyses included in this report, see USEPA (1999),
Cadmus (2000), Cadmus (2001), and USEPA (2002).  The data used in the cross-section analyses were
limited to only those data with confirmed drinking water source and population served information.  Only
standard SDWA compliance samples were used; samples identified as "special," "duplicate," or
"investigation," or samples of unknown type were not used in the analyses. (For example, investigation
samples can signify an investigation of identified contamination at a particular PWS. Inclusion of the
investigation sample results would bias the calculation of mean concentration for the system, and
therefore the investigation samples are excluded.)

Raw data from the States were received in a very wide variety of formats, structures, and content. Each
data set was reviewed to ensure it contained the basic data elements (data fields) necessary to conduct a
consistent analysis for this study.  These elements were reviewed with State data management staff both
Occurrence Summary and Use Support Document
March 2002

-------
before and after data were received and reviewed to ensure consistent and appropriate data set
interpretations. Every State data set reviewed for this study contained unique data elements or unique
treatment of common elements. Even after initial screening and conversion to roughly similar formats
and data set structure, unique factors were always uncovered during data analysis. Many of the
confounding factors were resolved only through direct consultation with the States. As a general rule,
when errors or ambiguities in various data elements could not be resolved, those particular data elements
were not included in the analyses to avoid problematic results or results based on data of questionable
quality.  This data quality measure eliminated relatively very few observations (compared to the
thousands of analytical results included in the data sets).

Decisions also had to be made on how to quantitatively include "less-than" or "non-detection" data.
Some states record the minimum reporting level (or limit) (MRL) in the analytical result column and also
include a "<" in a corresponding column to identify an analytical result record for which no contaminant
was detected.  (Such samples are often referred to as "no-detects" or "non-detections." More  precisely,
these  samples can be referred to as "censored data," meaning that there apparently is no quantitative
information for those records below the MRL, or censoring, concentration.) Other states simply include
a zero in the analytical result column to signify a non-detection.  Although non-detection data do not play
a role in the Stage 1 analysis, the non-detection data did pose a problem within Stage 2 analysis
(described below in Section 1.4).

Summary record counts were generated to determine what (and how many) MRLs were present in the
data for each contaminant. (There are sometimes multiple approved laboratory analytical methods that
can be used to analyze drinking water samples, and different analytical methods  can have different
minimum reporting levels.) For the parametric statistical analysis conducted for the Six-Year Review,
each record must have a quantitative, non-zero value. Therefore, all non-detection data (often reported as
"zero" or simply "<") were set equal to the non-zero modal MRL for the particular state reporting the
record in question. For the states that set all non-detections equal to zero, the non-detections were set
equal to the overall non-zero modal MRL for the entire 16 state cross-section. Table 1.3-1 describes the
number and type of public water systems, the population  served by those systems, and the overall non-
zero modal detection limit for each contaminant from the 16-State cross-section data set used  for Stage 2
analysis.
Table 1.3-1. Contaminant Occurrence Data From the 16-State Cross-Section
Contaminant
Total
Number of
Systems
Number of
Ground
Water
Systems
Number of
Surface
Water
Systems
Total
Population
Served by
Systems
Non-Zero
Modal MRL
(mg/L)
lOCs
Beryllium
Chromium
Fluoride
Mercury
Thallium
18,933
19,695
20,803
18,995
17,972
17,509
18,169
19,210
17,445
16,504
1,424
1,526
1,593
1,550
1,468
104,573,700
105,380,000
107,075,700
105,096,700
104,291,600
0.001
0.01
0.1
0.001
0.001
SOCs
Alachlor
Bis(2-ethylhexyl)phthalate
Carbofuran
14,330
9,418
13,925
12,917
8,591
12,531
1,413
827
1,394
95,678,600
78,293,000
94,338,000
0.0002
0.0006
0.0009
Occurrence Summary and Use Support Document
March 2002

-------
Contaminant
Chlordane
1 ,2-Dibromo-3-chloropropane
Diquat
Glyphosate
Heptachlor
Heptachlor Epoxide
Hexachlorobenzene
Hexachlorocyclopentadiene
Oxamyl
Picloram
Simazine
Toxaphene
Total
Number of
Systems
13,184
14,042
9,159
7,862
14,245
14,133
14,011
13,922
13,157
12,907
14,533
13,805
Number of
Ground
Water
Systems
11,854
13,008
8,337
7,069
12,835
12,729
12,625
12,536
11,798
11,555
13,136
12,408
Number of
Surface
Water
Systems
1,330
1,034
822
793
1,410
1,404
1,386
1,386
1,359
1,352
1,397
1,397
Total
Population
Served by
Systems
97,459,900
87,727,200
73,602,900
70,081,900
96,563,400
96,222,900
94,035,300
93,429,200
92,345,800
93,235,500
98,178,100
95,108,100
Non-Zero
Modal MRL
(mg/L)
0.0002
0.00002
0.0004
0.006
0.00004
0.00002
0.0001
0.005
0.002
0.0001
0.001
0.001
voc.
Benzene
Carbon Tetrachloride
1 ,4-Dichlorobenzene
1 ,2-Dichloroethane
1 , 1 -Dichloroethy lene
Dichloromethane
1 ,2-Dichloropropane
Tetrachloroethylene
1 ,1 ,2-Trichloroethane
Trichloroethylene
23,266
23,028
18,961
23,038
19,101
21,530
21,988
22,362
22,284
23,035
21,670
21,454
17,615
21,463
17,576
20,019
20,410
20,795
20,758
21,461
1,596
1,574
1,346
1,575
1,525
1,511
1,578
1,567
1,526
1,574
110,866,600
110,605,500
72,994,500
110,794,100
106,607,600
110,146,100
110,450,100
110,557,800
110,366,500
110,612,900
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
0.0005
The reduced number of systems sampling for SOC data, as compared to IOC and VOC, may relate to state waivers for pesticides and herbicides.
For example, New Jersey has an extensive SOC waiver protocol.
Note: All population values are rounded to the nearest hundred.


1.3.6 Additional Data

In addition to the primary data sources described above, previously compiled contaminant occurrence
information is also presented whenever available. This information generally includes reviews of
previous occurrence surveys and studies of contaminant occurrence in both ground water and surface
water drinking and non-drinking water sources.  Furthermore, some of the information is national and
some regional in nature, providing additional, and sometimes historical, contaminant information that
supplements the other contaminant occurrence, use, and production information presented in this report.
However, the detailed national estimates of contaminant occurrence, described in the following Section
1.4, were conducted only using the 16-state national cross-section data.
Occurrence Summary and Use Support Document
March 2002

-------
 1.4 Occurrence Analysis

 A two stage occurrence estimation approach has been developed to assess the 16-State cross-section data
 of regulated contaminant occurrence. The initial stage, "Stage 1 Analysis," provides a straight-forward,
 clear, and conservative assessment of occurrence.  Using the occurrence data aggregated from the 16-
 State cross-section, calculations were made  of the percent of PWSs with at least one analytical result
 that exceeds a specified health threshold. These initial occurrence measures were developed for each
 contaminant according to water system source water type (ground water, surface water) for the
 aggregated 16-State cross-section data set. These "Stage 1 Analyses" are descriptive, non-parametric
 statistics that provide a general characterization of contaminant occurrence.

 Similar estimates based on the population served by water system provide a "Stage 1" preliminary
 characteristic of exposure potential.  Stage 1 results are most closely representative of short term
 exposure since the Stage  1 analyses are based on the single maximum analytical value recorded at each
 water system. Assessed relative to MCLs (which reflect public health considerations for long-term
 exposure to contaminants in drinking water), these Stage 1 analyses are conservative (cautious regarding
 public health concerns) in the sense that they are descriptive statistics based on peak, rather than long-
 term mean, concentrations. For a complete  description of the Stage 1 methodology and a comprehensive
 presentation of the detailed Stage  1 findings, please refer to First Stage Occurrence and Exposure Report
for Six-Year Review (Cadmus, 2000).

 The second "Stage 2 analysis," a rigorous parametric statistical modeling approach, was developed to
 enable more specific and detailed occurrence estimations. While Stage 1 analysis estimates, for example,
 the percent of PWSs with at least one analytical result exceeding a specified health threshold, Stage 2
 analysis estimates the percent of PWSs with an estimated system mean concentration exceeding a
 specified health threshold.  The Stage 2 analytical results are stratified by source water type and system
 size.  (For example, estimates can be made for the number of systems, with a mean concentration of a
 contaminant greater than some specified concentration, that use ground water and serve fewer than 500
 people, that use surface water and serve fewer than 500 people, that use ground water and serve 501 to
 3300 people, etc.).  These estimates are expressed as the probability of threshold exceedance. These
 Stage 2 occurrence estimates also include quantified estimation errors for the values estimated.  (These
 quantified estimation errors are important because they provide a measure of the level of confidence in
 an statistical estimation.)

 Since the Stage 2 occurrence estimations represent mean concentration values, they are more
 representative of long-term exposure to contaminants in drinking water.  These Stage 2 estimates
 contribute to the initial assessments of health risks from exposure to contaminants through consumption
 of drinking water.  The Stage 2 estimates also can be used in assessments of the costs and benefits
 associated with the establishment of a revised regulatory limit for any particular drinking water
 contaminant. A general description of the Stage 2 analysis is presented below. For a complete
 description of the Stage 2 methodology and a comprehensive presentation of the stratified Stage 2
 findings, please refer to Occurrence Estimation Methodology and Occurrence Findings for Six-Year
 Review of National Primary Drinking Water Regulations (USEPA, 2002). The entire two stage
 occurrence estimation approach (including the development of the national cross-section of state data, as
 well as the Stage 1 and Stage 2 analyses) was peer-reviewed and received generally favorable support.

 The statistical model developed for the Stage 2 analyses for the Six-Year Review is a Bayesian-based
 hierarchical model. One advantage of this modeling approach is that it is able to fully use the occurrence
 information contained in sample "non-detections" (all the analytical results with values less than the
 MRL) in estimating system mean concentration. (This is important because the sample non-detections
 typically comprise the majority of occurrence data for most drinking water contaminants.)  The Bayesian-
 based model first estimates mean concentration values for each PWS in the  16-State cross-section.  In
 the process of generating the system means, a standard deviation associated with the mean is also
 generated. (The standard deviation defines, in a general sense, the variability of the estimated system
Occurrence Summary and Use Support Document          9                                       March 2002

-------
mean concentrations.) Importantly, the Bayesian-based modeling approach also generates estimations of
error which enable a quantiative assessment of the uncertainty associated with all model estimates.

The estimated system means and standard deviations are then used as input to a Markov chain Monte
Carlo method. The Monte Carlo method is, in general terms, any technique using a large number of
randomly selected values to determine a single estimated total.  This method works well for estimating
the total number of public water systems with an (estimated) mean concentration greater than some
specified threshold concentration, because the underlying probability distributions -defined by the
individual estimated system means and standard deviations- are known, but the final results (the actual
total number of systems with actual contaminant means greater than some specified threshold) is more
difficult to determine. Approximately 500 Monte Carlo  simulations (using the stratum-level mean and
standard deviation as model input) are used to estimate the number of systems for each source water type
and system-size category that are expected to exceed each specified concentration threshold.  The
estimated number of systems that exceed each threshold for a given stratum is then divided by the total
number of systems in that stratum, resulting in the percent of systems estimated to exceed a specified
threshold for a specific stratum (the estimated mean "probability of threshold exceedance").

Once the probability of threshold exceedance has been estimated through the Stage 2 analysis, the
number of systems (and population served by systems) with potential threshold exceedances can be
estimated. The total number of systems in the 16-State cross-section with mean contaminant
concentrations that are expected to exceed specified threshold concentrations is calculated by multiplying
the percentage of systems with an estimated mean concentration threshold exceedance (estimated by the
statistical model) by the total number of systems in the 16-State cross-section with data for that particular
contaminant.  The total population served by systems in the 16-State cross-section potentially exposed to
contaminant concentrations greater than the health threshold is estimated by multiplying the population
served by systems in the 16-State cross-section with data for that particular contaminant by the model-
generated percentage of population served by systems with estimated mean concentration threshold
exceedance. (See Table  1.3-1 for the total number of systems and population served by systems in the
16-State cross-section for each of the  30 contaminants.)

Because the 16-State cross-section data set  was developed to contain data on contaminant occurrence that
is indicative of national occurrence, national contaminant occurrence can be also estimated. The total
national number of systems (or population served by systems) estimated to exceed a specified threshold
is extrapolated by multiplying the representative cross-section probability of exceedance by the national
numbers for systems (and population served by systems) documented in the Water Industry Baseline
Handbook, Second Edition (USEPA, 2000d).  The total number of ground and surface water community
water systems (CWSs) plus non-transient, non-community water systems (NTNCWSs) in the Baseline
Handbook is 65,030, and the total population served by ground and surface water CWSs plus NTNCWSs
is 213,008,182 persons.  (The handbook presents the system and population served numbers stratified by
source water type and population served size categories as well.)

To derive the  national occurrence estimate for a specific threshold/source water type/population served
size category, the national number of PWSs (or population served by PWSs) from the handbook is simply
multiplied by the probability of exceedance (a percentage) estimated by the statistical model.  In the
tables presented in this report, only the total number of systems (and population served by systems)
estimated to exceed each threshold are presented.  For stratified contaminant occurrence findings, please
refer to Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of National
Primary Drinking Water Regulations (USEPA, 2002).

Contaminant occurrence findings based on the Stage 1 and Stage 2 analyses (using the 16-State cross-
section data) are summarized in tables presented in the specific contaminant chapters that follow this
introductory chapter. All Stage 1 and Stage 2 occurrence findings (presented as percentages)  are
displayed to 3 significant figures and all population values are rounded to the nearest hundred.
Importantly, note that the Stage 2 findings are based on model estimations conducted separately for
ground water systems, surface water systems, and combined ground and surface water systems. Since


Occurrence Summary and Use Support Document          10                                     March 2002

-------
these groups of systems were modeled separately, the sum of occurrence estimates for ground water
systems and surface water systems will not equal the total occurrence estimate for combined ground and
surface water systems. For more information of the details of the Stage 2 Bayesian-based modeling,
please refer to USEPA (2002).


1.5 References

Agency for Toxic Substances and Disease Registry (ATSDR).  2001. About the Agency for Toxic
       Substances and Disease Registry. Available on the Internet at:
       http://www.atsdr.cdc.gov/about.html (Last updated June 28, 2001).

Blomquist, J.D., J.M. Denis, J.L. Cowles, J.A. Hetrick, R.D. Jones, and N.B. Birchfield.  2001.
       Pesticides in Selected Water-Supply Reservoirs and Finished Drinking Water, 1999-2000:
       Summary of Results from a Pilot Monitoring Survey. U.S. Geological Survey Open-File Report
       01-456. 65pp.

Cadmus. 2000. First Stage Occurrence and Exposure Report for Six-Year Review. Draft report
       submitted to EPA for review May 12, 2000.

Cadmus. 2001. Data Quality Overview of Contaminant Occurrence Data - DRAFT.  Draft report
       submitted to EPA for review January 31,2001.

Gilliom, R.J., O.K. Mueller, and L.H. Nowell.  1998.  Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589 (Last updated October 09, 1998).

Lapham, W.W., K.M. Neitzert, M.J. Moran, and J.S. Zogorski. 1997. USGS compiles data set for
       national assessment of VOCs in ground water. G. WaterMon. Rem.  17(4): 147-157.

Leahy, P.P., and T.H. Thompson.  1994.  The National Water-Quality Assessment Program. U.S.
       Geological Survey Open-File Report 94-70. 4 pp.  Available on the Internet at:
       http://water.usgs.gov/nawqa/NAWQA.OFR94-70.html  (Last updated May 29, 2001).

Lopes, T.J. and S.G. Dionne.  1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban  Stormwater. U.S. Geological Survey Open-File Report 98-409. 67
       pp.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, R.M. Clawges, and J.S. Zogorski.  1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Em. Sci. and Tech..  33(23):4176-4187.

USEPA.  1999. A Review of Contaminant Occurrence in Public Water Systems.  EPA Report/816-R-
       99/006, Office of Water, 78 pp.

USEPA.  2000a. TRIExplorer: Are Year-to-Year Changes Comparable? Available on the Internet at:
       www.epa.gov/triexplorer/yearsum.htm (Last modified May 5, 2000).

USEPA.  2000b.  The Toxic Release Inventory (TRI) and Factors to Consider when Using TRI Data.
       Available on the Internet at: http://www.epa.gov/tri/tri98/98over.pdf (Last modified August 11,
       2000).

USEPA.  2000c.  What is the Toxic Release Inventory. Available on the Internet at:
http://www.epa.gov/tri/general.htm (Last modified February 28, 2000).


Occurrence Summary and Use Support Document          11                                     March 2002

-------
USEPA. 2000d. Water Industry Baseline Handbook, Second Edition (Draft). March 17.

USEPA. 2001. EPA Protocol for the Review of Existing National Primary Drinking Water Regulations -
       DRAFT. U.S. Environmental Protection Agency, Office of Drinking Water. May 2001.

USEPA. 2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 1998. Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA). PROVISIONAL DATA - SUBJECT TO
       REVISION. Available on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/,  last updated
       October 9, 1998.
Occurrence Summary and Use Support Document         12                                    March 2002

-------
2.0 INORGANIC CONTAMINANTS
Occurrence Summary and Use Support Document        13                             March 2002

-------
2.1    Beryllium
Table of Contents
   . 1  Introduction, Use and Production  	  15
   .2  Environmental Release  	  17
   .3  Ambient Occurrence 	  18
   .4  Drinking Water Occurrence Based on the 16-State Cross-Section	  21
   .5  Additional Drinking Water Occurrence Data  	  26
   .6  Conclusion	  28
   .7  References  	  28
Tables and Figures

Table 2.1-1: Imports of Beryllium to the United States (thousand metric tons, gross weight)  	  16

Table 2.1-2: Beryllium and Beryllium Compound Manufacturers and Processors by State 	  16

Table 2.1-3: Environmental Releases (in pounds) for Beryllium in the United States,
       1988-1999	  17

Table 2.1-4: Environmental Releases (in pounds) for Beryllium Compounds in the
       United States,  1988-1999 	  18

Table 2.1-5: Beryllium Detections and Concentrations in Surface Water and Ground Water	  18

Table 2.1-6: Beryllium Occurrence in Raw Water - National Surveys 	  19

Table 2.1-7: Beryllium Occurrence in Raw Water - Regional Surveys	  20

Table 2.1-8: Stage 1 Beryllium Occurrence Based on  16-State Cross-Section - Systems	  22

Table 2.1-9: Stage 1 Beryllium Occurrence Based on  16-State Cross-Section - Population	  22

Table 2.1-10: Stage 2 Estimated Beryllium Occurrence Based on 16-State Cross-Section -
       Systems	  24

Table 2.1-11: Stage 2 Estimated Beryllium Occurrence Based on 16-State Cross-Section -
       Population	  25

Table 2.1-12: Estimated National Beryllium Occurrence - Systems and Population Served	  26

Table 2.1-13: Estimated Beryllium Exceedance as Reported in the NIRS - Systems 	  27

Table 2.1-14: Estimated Beryllium Exceedance as Reported in the NIRS - Population  	  27

Table 2.1-15: Beryllium Occurrence in 100 Largest U.S. Cities  	  28

Occurrence Summary and Use Support Document          14                                     March 2002

-------
2.1.1  Introduction, Use and Production

Beryllium (Be) is the lightest structural metal known. Pure beryllium is hard and grayish. Beryllium is
never found free in the environment but can be found in compounds in mineral rocks, coal, soil, and
volcanic dust (ATSDR, 1993).  Beryllium compounds have no particular smell. Beryllium exists in the
environment in approximately 40 mineralized forms. The most common forms for beryllium are beryl, a
beryllium aluminum silicate (3BeO.Al2O3.6SiO2) and bertrandite, ahydrated disilicate
(4BeO.2SiO2.H2O). The most important commercial forms of beryllium are the metal itself, beryllium-
copper alloys, and beryllium oxide (Ciccone & Associates, 1984).

Most beryllium ore that is mined is  converted into alloys. Most of these alloys are used in making
electrical and electronic parts or as construction materials for machinery and molds for plastics. Pure
beryllium metal is used in nuclear weapons and reactors, aircraft and space vehicle structures,
instruments, x-ray machines, and mirrors.  Beryllium oxide is also made from beryllium ores and is used
to make specialty ceramics for electrical and high-technology applications (ATSDR, 2000).

Beryllium metal  is used in aircraft disc brakes, x-ray transmission windows, space vehicle optics and
instruments, aircraft/satellite structures, missile guidance systems, nuclear reactor neutron reflectors,
nuclear warhead triggering devices, fuel containers, precision instruments, rocket propellants,
navigational systems, heat shields, mirrors, high speed computers, and audio components, with other
assorted miscellaneous uses (Cunningham, 1998, as cited in ATSDR, 2000). In 1998, the use of
beryllium (as an  alloy, metal, or oxide) in the electronic and electrical components, aerospace, and
applications accounted for more than 80% of its consumption (Cunningham, 1999, as cited in ATSDR,
2000).

Beryllium oxide  is used in high technology ceramics, electronic heat sinks, electrical insulators,
microwave oven components, gyroscopes, military vehicle armor, rocket nozzle crucibles, thermocouple
tubing, laser structural components, and substrates.  Beryllium is also used for high-density electrical
circuits, automotive ignition  systems, and radar electronic countermeasure systems (Cunningham, 1998,
as cited in ATSDR, 2000).

Beryllium-copper alloys are useful in a wide variety of applications because of their electrical and
thermal conductivity, high strength  and hardness, good  corrosion and fatigue resistance, and non
magnetic properties.  Beryllium-copper alloys are manufactured into springs, electrical connectors and
relays, precision instruments, brushings and bearings in aircraft and heavy machinery, non sparking tools,
submarine cable  housing and pivots, wheels and pinions, switches in automobiles, molds for injection
molded plastics,  radar, telecommunications, factory automation, computers, home appliances,
instrumentation and control systems, tubing in oil and drilling equipment, connectors for fiber optics, and
integrated circuits, as well as many  other uses (Cunningham, 1998, as cited in ATSDR, 2000).

Both anthropogenic and natural processes result in emissions of beryllium to the atmosphere. In addition
to ore processing, beryllium is also released into the atmosphere during the production and use of
beryllium alloys  and chemicals.  Beryllium is released into the atmosphere from anthropogenic sources
including the combustion of coal and fuel oil, the incineration of municipal and solid waste, the
production, use, and recycling of beryllium alloys and chemicals, and, to a minor extent, the burning of
solid rocket fuel. Emissions  from coal and fuel oil combustion account for a majority of the United
States beryllium  emissions from natural and anthropogenic sources. Natural emission sources include
windblown dusts and volcanic particles. However, the  amount of beryllium released to the atmosphere
from these sources is small compared with anthropogenic sources.
Occurrence Summary and Use Support Document          15                                      March 2002

-------
Anthropogenic sources of beryllium release to water include industrial waste water effluents. Beryllium
concentrations are highest in waste waters from electric utility industries.  Deposition of atmospheric
beryllium is also a source in surface waters. Beryllium also enters water from the weathering of rocks
and soils.  Since coal contains beryllium, it is also likely that beryllium will enter surface water via
leaching of coal piles.

Beryllium is naturally present in soils and sediments. The majority of beryllium and beryllium
compound releases to land are by facilities that manufacture or process beryllium. Coal fly ash and
municipal solid waste containing beryllium are disposed of in landfills and used in building materials.
This contributes to beryllium concentrations in soil.  About 100 million tons of coal fly ash containing
various levels of beryllium are generated each year.  Land application of sewage sludge containing higher
than background concentrations of beryllium can be a source of beryllium contamination of soil.
Deposition of atmospheric aerosols on terrestrial surfaces is another source of beryllium in soil.
Table 2.1-1:  Imports of Beryllium to the United States (thousand metric tons, gross weight)
Imports For Consumption
Beryllium Ore and Metal
1995
32
1996
20
1997
20
1998
50
1999
20
 Source: USGS, 2000
Table 2.1-2:  Beryllium and Beryllium Compound Manufacturers and Processors by State
State"
AL
AZ
FL
GA
IL
IN
KY
MI
MO
MT
NC
NM
OH
OK
PA
SC
TN
TX
UT
WI
wv
WY
Number of facilities
6
4
2
5
1
5
4
3
1
1
4
6
8
2
5
1
3
2
4
8
1
9
Range of maximum amounts on
site in thousands of pounds'1
1,000-99,999
10,000-9,999,999
1,000-9,999
10,000-99,999
1,000-9,999
1,000-999,999
0-99,999
0-99,999
1,000-9,999
10,000-99,999
10,000-99,999
0-99,999
1,000-99,999
100-99,999
100-999,999
0-99
10,000-99,999
10,000-99,999
100-49,999,999
1,000-99,999
1,000-99,999
1.000-99.999
Activities and uses0
1,3,4,5,6,9,10,13
1,3,4,5,6,8,9,10
1,4,5,6,10,13
1,3,4,5,6,10,13
1,5,10
1,5,9,10
1,5,6,10,13
1,5,9,10
8,10
1,3,4,5,6,10,13
1,3,4,5,6,9,10,13
1,3,4,5,6,10,13
1,3,4,5,6,10,13
1,5,9
1,2,3,4,5,6,7,9,10
13
1,5,8,9
1,3,4,5,6,10,13
1,4,7,8,10,13
8,9
1,3,4,5,6,10,13
1.5.6.10.13
Tost office State abbreviations used
b Range represents maximum amounts on site reported by facilities in each State
cActivities/Uses
                        8. As a formulation component
                        9. As an article component
                        10. For repackaging only
                        11. As a chemical processing aid
                        12. As a manufacturing aid
                        13. Ancillary or other uses
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
cNumber of facilities reporting "no data" regarding maximum amount of the substance on site
Source: AT SDR, 2000 compilation of 1998 TRI data
Occurrence Summary and Use Support Document
                                                   16
March 2002

-------
2.1.2  Environmental Release

Beryllium and beryllium compounds are listed as Toxics Release Inventory (TRI) chemicals (see Tables
2.1-3 and 2.1-4). For both beryllium and beryllium compounds, releases to land constitute most of the
on-site releases.  Beryllium compounds have a gradual increase in releases to land until 1992, then
decrease from 1993 through 1999, while releases of beryllium decrease until  1993 and then increase in
1994. It is unclear whether these sharp beryllium decreases are real or a function  of changes in TRI
reporting requirements in the late 1980s and early 1990s (see discussion in Introduction). Air emissions
are also an important mode of on-site release. Though the first several years of record for air emissions
of beryllium are markedly higher for certain years, no trend is apparent for the remainder.  Also, air
emissions of beryllium compounds have fluctuated modestly with no trend evident.  Surface water
discharges of beryllium and beryllium  compounds are less significant on-site releases, with low levels
continuing until the present and no other trends apparent. There are no releases by underground injection
of either beryllium or beryllium compounds until 1999, where a high percentage of on-site releases of
beryllium compounds are attributed to  underground injection.

 Increases in releases to land have contributed to increases in total on- and off-site releases in recent
years.  Off-site releases of beryllium and beryllium compounds are considerable.  Off-site releases of
beryllium are a large component of total releases. Though there is a large  drop in beryllium releases in
1992 when compared to the previous year, the late  1990s show a steady increase in pounds released.
Off-site releases of beryllium compounds fluctuate throughout the years, with highest releases in 1988
and 1998, although no other trends are apparent.  These TRI data for beryllium were reported from 21
States, with only 3 States  reporting data for all years (USEPA, 2000). Eight of these 21 States that
reported TRI data for beryllium are contained in  the 16-State cross-section (used for analyses of
beryllium occurrence in drinking water; see  Section 2.1.4).  The TRI data for beryllium compounds were
reported from  17 States, with only two States reporting data for all years (USEPA, 2000).  Six of the 17
States that reported TRI data for beryllium compounds are contained in the 16-State cross-section.  (For a
map of the 16-State cross-section, see Figure 1.3-1.)
Table 2.1-3:  Environmental Releases (in pounds) for Beryllium in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
769
799
816
1,114
1,087
899
903
1,868
1,378
1,375
1,895
2,763
Surface Water
Discharges
57
26
27
36
31
36
24
39
101
42
122
74
Underground
Injection
-
-
-
-
-
-
-
-
-
-
-
-
Releases
to Land
53,271
57,818
56,123
31,245
21,255
22,860
14,594
21,358
29,023
6,517
31,522
37,000
Off-Site Releases
20,081
20,404
5,741
4,852
7,595
9,632
5,142
14,774
117,582
1,371
1,199
3,160
Total On- &
Off-site
Releases
74,178
79,047
62,707
37,247
29,968
33,427
20,663
38,039
148,084
9,305
34,738
42,997
 Source: USEPA, 2000
Occurrence Summary and Use Support Document
17
March 2002

-------
Table 2.1-4: Environmental Releases (in pounds) for Beryllium Compounds in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
473
383
365
395
360
610
363
511
242
212
962
862
Surface Water
Discharges
27
6
1
1
2
2
4
5
9
88
25
17
Underground
Injection
4,100
-
-
-
-
-
-
-
-
-
-
-
Releases
to Land
19
-
96
16,188
23,000
17,000
8,087
48,000
30,000
40,000
36,000
12,000
Off-Site Releases
5,028
6,754
4,602
2,333
2,391
2,901
3,055
4,618
2,180
972
5,085
8,261
Total On- &
Off-site
Releases
9,647
7,143
5,064
18,917
25,753
20,513
11,509
53,134
32,431
41,272
42,072
21,140
 Source: USEPA, 2000
2.1.3 Ambient Occurrence

Beryllium is an analyte for both surface and ground water NAWQA studies, with a method detection
limit (MDL) of 0.001 mg/L. Additional information on analytical methods used in the NAWQA study
units, including method detection limits, are described by Gilliom and others (1998).

Beryllium was never detected in groundwater. A possible explanation for the higher detection
frequencies in surface water is its greater sensitivity to anthropogenic releases.  Both the median and 99th
percentile concentrations of beryllium are below the  MCL (0.001 mg/L and 0.002 mg/L, respectively).
Beryllium detection frequencies in surface water greater than the MCL (0.004 mg/L) are approximately
10 times less than the percentage of all beryllium detections in surface water.
Table 2.1-5: Beryllium Detections and Concentrations in Surface Water and Ground Water
                          Detection frequency
                               >MDL*
           Detection frequency      Concentration percentiles
                 > MCL*             (all samples; mg/L)
                         % samples     % sites     % samples      % sites
                                   median
                                      99*
surface -water
                          0.64%
3.83%
0.06%
0.55%
0.001
0.002
    ground -water
                                    N/A
                                      N/A
* The Method Detection Limit (MDL) for beryllium in water is 0.001 mg/L and the Maximum Contaminant Level (MCL) is 0.004 mg/L.
Occurrence Summary and Use Support Document
           18
                                        March 2002

-------
2.1.3.2 Additional Ambient Occurrence Data

A summary document entitled "Beryllium in Water: An Assessment of Occurrence and Exposure"
(Ciccone & Associates, 1984), was previously prepared for past USEPA assessments of beryllium.  This
document included various studies and information are presented regarding levels of beryllium in
ambient water on both the regional and national level.  The following information is taken directly from
"Beryllium in Water: An Assessment of Occurrence and Exposure" (Ciccone & Associates, 1984).

2.1.3.2.1 National Surveys

Since 1967, no new information on beryllium concentrations in naturally occurring waters in the U.S. has
been cited in the literature. However, no drastic changes would be expected. A study by Kopp and
Kroner (1969, as cited in Ciccone & Associates, 1984) suggests a minimum beryllium concentration to
be 0 |ig/L in raw surface water with a maximum positive value of 1.22 |ig/L and a mean positive value of
0.19 |ig/L (Table 2.1-6). It should be noted, however, that the authors used only positive values for
beryllium in calculating the average level of beryllium occurrence.  Further, the frequency of occurrence
of beryllium in this study was only 5.4%. Thus, the average concentration of beryllium in raw water
suggested by Kopp and Kroner may be artificially inflated.

In a 1970 USGS survey of 143  surface water samples, beryllium levels were found to be mostly less than
1.0 |ig/L (Table 2.1-6).  The highest level detected was less than 40 |ig/L in  California Gulch of Malton,
Colorado. No detection limits were reported for this study. Therefore, the meaning of less than 1 or less
than 40 |ig/L is unclear. Of the few actual numerical values reported by the  USGS for beryllium, all
were in New Jersey.  These values ranged from 0.1- 2.4 |ig/L (mean 0.65 M-g/L), the highest value being
detected in the Delaware River of Trenton, NJ.

Durum and Haffty (1961, as cited in Ciccone & Associates, 1984) monitored 15 rivers in the U.S. and
Canada for the presence of beryllium (Table 2.1-6). At 13 sites, beryllium was below detection limits in
all samples examined. Beryllium was detected in 1 of 4 samples from the Apalachicola River, FL
(<0.058 |ig/L) and in 2 of 4 samples from the Atchafalaya River, LA (for one sample no numerical value
reported; for the other, <0.22 M-g/L). Detection limits for beryllium in this survey were not given. The
national surveys described above all indicate that beryllium occurrence is very low in naturally occurring
raw surface waters.
Table 2.1-6:  Beryllium Occurrence in Raw Water - National Surveys
Survey
Trace Metals in
Waters of the
United States.
Oct. 1, 1962-
Sept. 30, 1967
Sampling # Samples/ Site
Site
1 30 R, S Weekly samples
composited for 3
months
(occasionally 1
month).
Composite
samples taken
twice yearly.
Mean
Concentration
(Range) jig/L
0.19
(0-1.22)
System Size Survey
Area
U;U.S. Nationwide
divided into
15 major
river basins.
Notes
1,577 samples taken. 5.4%
frequency of occurrence
Occurrence Summary and Use Support Document
19
March 2002

-------
Survey

Quality of
Surface Waters
in the United
States. USGS,
1970
Occurrence of
Minor Elements
in Water.
Durum and
Haffty, 1961

Public Water
Supplies of the
100 Largest
Cities in the
United States,
1962. Durfor
and Becker,
1962
Sampling # Samples/ Site
Site

181R, S Samples
collected daily
and monthly at
143 locations.

15R, S Variable, 2 to 7
samples/site.

100 G, S Variable





Mean System Size
Concentration
(Range) jig/L
0.65 U;U.S.
(0.1-2.4) divided into
16 major
river basins.

0.22-0.058 U, U.S. and
Canada 1 5
sampling
sites

ND - 0.75 U





Survey Notes
Area

Nationwide Highest concentrations of Be
detected in the Delaware River,
Trenton, NJ. No limits of
detection given.

Nationwide Apalachicola River, FL: 1 of 4
and Canada samples positive at < 0.058
Hg/L. Atchafalaya River, LA: 2
of 4 samples positive at O.22
Hg/L. No detection limits
given.
Nationwide





2.1.3.2.2 State and Regional Surveys

Regional studies of the occurrence of beryllium in natural waters show a spectrum of results (Table 2.1-
7).  Silvey (1967, as cited in Ciccone & Associates, 1984) found no detectable beryllium in water
samples from the State of California in a study of groundwater, surface water, ocean water and oil wells.
On the other hand,  Page (1981, as cited in Ciccone & Associates,  1984) found beryllium in all 1,064
groundwater samples and all 590 surface water samples obtained from New Jersey. The highest
concentration of beryllium found by Page was 84 |ig/L in groundwater. This same study showed that, in
surface water, the highest concentration of beryllium was 1 |ig/L.  For both surface and ground water, the
median level of beryllium was 1 |ig/L. No detection limits were specified in this study.
Table 2.1-7:  Beryllium Occurrence in Raw Water - Regional Surveys
Survey
Occurrence of Selected
Minor Elements in the
Waters of California.
Silvey, 1967.
Sampling
Site
72 Springs
63 Water
Wells
19 Oil Wells
# Samples/
Site
One (3. 51)
Mean System
Concentration Size
(Range) fig/L
ND U
Survey
Area
State of
California
Notes
Be limit of detection was 0
l-ig/L. No Be found at any
sampling site.

3
Occurrence Summary and Use Support Document
20
March 2002

-------
Survey Sampling
Site
Comparison of Ground 1064 U, G
Water and Surface 590 U, S
Water for Patterns and
Levels of
Contamination by
Toxic Substances,
1977-1979. Page,
1981.
# Samples/ Mean System Survey
Site Concentration Size Area
(Range) fig/L
Usually one 1-84 U State of
grab New Jersey
sample,
sometimes
more.
Notes
Median data reported. No limits
of detection reported. 84 ng/L
highest concentration found in
groundwater (median 1 ng/L). 1
Hg/L highest concentration
found in surface water (median 1
l-ig/L). Be detected in all
samples.
2.1.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of beryllium occurrence presented in the following section is based on State compliance
monitoring data from the  16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest: 0.004 mg/L; 0.01 mg/L; and 0.001 mg/L.

All sixteen cross-section State data sets contained occurrence data for beryllium.  These data represent
approximately 48,000 analytical results from more than 19,000 PWSs during the period from  1983 to
1998 (with most analytical results from 1992 to 1997). The number of sample results and PWSs vary by
State, although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness. The overall modal detection limit for beryllium in the 16 cross-section States is equal to
0.001 mg/L.  (For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this  report.)

2.1.4.1 Stage 1 Analysis Occurrence Findings

Table 2.1-8 illustrates the Stage 1 analysis of beryllium occurrence in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds: 0.01 mg/L, 0.004 mg/L (the current
MCL), and 0.001 mg/L (the modal MRL). A total of 18 ground water and surface water PWSs
(approximately 0.0951%) had at least  one analytical result exceeding 0.01 mg/L; 0.217% of systems (41
systems) had at least one analytical result exceeding the MCL (0.004 mg/L); and 1.28% of systems (243
systems) had at least one analytical result exceeding 0.001 mg/L.

Approximately 0.0971% of ground water systems (17 systems) had at least one analytical result greater
than 0.01 mg/L. About 0.223% of ground water systems (39 systems) had at least one analytical result
above the  MCL (0.004 mg/L).  The percentage of ground water systems with at least one result greater
than 0.001 mg/L was equal to 1.27% (223 systems).

Only 1 (0.0702% of) surface water systems had at least one analytical result greater than 0.01 mg/L. A
total of 2 (0.140% of) surface water systems had at least one analytical result greater than the  MCL
(0.004 mg/L). Twenty surface water systems (1.40%) had at least one analytical result exceeding  0.001
mg/L.
Occurrence Summary and Use Support Document
21
March 2002

-------
Table 2.1-8:  Stage 1 Beryllium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.01
0.004
0.001
Percent of Systems
Exceeding Threshold
0.0971%
0.223%
1.27%
Number of Systems
Exceeding Threshold
17
39
223

Surface Water
0.01
0.004
0.001
0.0702%
0.140%
1.40%
1
2
20

Combined Ground &
Surface Water
0.01
0.004
0.001
0.0951%
0.217%
1.28%
18
41
243
Reviewing beryllium occurrence in the 16 cross-section States by PWS population served (Table 2.1-9)
shows that approximately 0.0683% of the population (over 71,000 people) was served by PWSs with at
least one analytical result of beryllium greater than 0.01 mg/L. A total of 649,000 (0.621% of) people
were served by systems with an exceedance of the MCL. Approximately 3.2 million people (3.09%)
were served by systems in the 16-State cross-section with at least one analytical result greater than 0.001
mg/L.

The percentage of population served by ground water systems with analytical results greater than 0.01
mg/L was equal to 0.150% (almost 68,800 people).  When evaluated relative to 0.004 mg/L and 0.001
mg/L, the percent of population exposed was equal to 1.38% (635,000 people) and 4.25% (almost 2
million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.01 mg/L was equal
to 0.00441% (2,600 people).  Approximately 0.0240% (about 14,100 people) of the population served by
surface water systems in the 16-State cross-section was exposed to beryllium concentrations greater than
0.004 mg/L. When evaluated relative to 0.001 mg/L, the percent of population exposed was equal to
2.18% (over 1.2 million people).

Table 2.1-9:  Stage 1 Beryllium Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.01
0.004
0.001
Percent of Population
Served by Systems
Exceeding Threshold
0.150%
1.38%
4.25%
Total Population Served
by Systems Exceeding
Threshold
68,800
635,000
1,954,500
Occurrence Summary and Use Support Document
22
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population
Served by Systems
Exceeding Threshold
Total Population Served
by Systems Exceeding
Threshold

Surface Water
0.01
0.004
0.001
0.00441%
0.0240%
2.18%
2,600
14,100
1,277,600

Combined Ground &
Surface Water
0.01
0.004
0.001
0.0683%
0.621%
3.09%
71,400
649,000
3,232,100
2.1.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 2.1-10 and 2.1-
11.  The statistically generated best estimate values, as well as the ranges around the best estimate value,
are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.
For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology
and Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations -
DRAFT (USEPA, 2002)).

A total of 2 (0.00809%) ground water and surface water PWSs in the 16 States had an estimated mean
concentration of beryllium exceeding 0.01 mg/L. Approximately  15 (0.0781% of) PWSs in the 16 States
had an estimated mean concentration exceeding 0.004 mg/L, and 203 (1.07%) had an estimated mean
concentration exceeding 0.001 mg/L.

An estimated 2 ground water PWSs in the 16 cross-section States (0.00870%) had a mean concentration
greater than 0.01 mg/L, 15 (0.0833%) had a mean concentration greater than 0.004 mg/L,  and 192
(1.10%) had a mean concentration greater than 0.001 mg/L.  Approximately  1 (0.000562%), 1 (0.014%),
and 10 (0.731%) surface water PWSs in the 16 States had estimated mean concentrations exceeding 0.01
mg/L, 0.004 mg/L, and 0.001 mg/L, respectively.
Occurrence Summary and Use Support Document
23
March 2002

-------
Table 2.1-10:  Stage 2 Estimated Beryllium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.01
0.004
0.001
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.00870%
0.0833%
1.10%
Range
0.000% - 0.0229%
0.0400% -0.1 37%
0.908% -1.29%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
2
15
192
Range
0-4
7-24
159-225

Surface Water
0.01
0.004
0.001
0.000562%
0.0140%
0.731%
0.000% - 0.000%
0.000% - 0.0702%
0.351% -1.1 9%
1
1
10
0-0
0-1
5-17

Combined Ground &
Surface Water
0.01
0.004
0.001
0.00809%
0.0781%
1.07%
0.000% -0.0211%
0.0370% -0.127%
0.882% -1.25%
2
15
203
0-4
7-24
167-237
Reviewing beryllium occurrence by PWS population served (Table 2.1-11) shows that an estimate of
approximately 2,000 people (approximately 0.0019%) of population served by all PWSs in the 16 cross-
section States were potentially exposed to beryllium levels above 0.01 mg/L.  The percentage of
population served by PWSs in the 16 States with levels of beryllium above 0.004 mg/L and 0.001 mg/L
was 0.0208% (an estimated 21,800 people) and 0.699% (over 731,000 people), respectively.

When the percent of population served by ground water systems was evaluated relative to a threshold of
0.01 mg/L, 0.004 mg/L, and 0.001 mg/L, the percentage of population exposed in the 16 cross-section
States was equal to 0.00403% (an estimated 1,900 people), 0.0430% (an estimated 19,800 people) and
1.17% (an estimated 538,500 people), respectively.

The percentage of population served by surface water systems in the  16 States with levels above 0.01
mg/L was  equal to 0.000227% (an estimated 100 people), while the percentage of population served with
levels above 0.004 mg/L was 0.00341% (an estimated 2,000 people). The percentage of the population
served by surface water systems in the 16 States with levels above 0.001 mg/L was 0.330% (almost
193,000 people).
Occurrence Summary and Use Support Document
24
March 2002

-------
Table 2.1-11:  Stage 2 Estimated Beryllium Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.01
0.004
0.001
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.00403%
0.0430%
1.17%
Range
0.000% - 0.0255%
0.00586% -0.174%
0.595% -1.96%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
1,900
19,800
538,500
Range
0-11,700
2,700 - 79,900
273,800 - 903,400

Surface Water
0.01
0.004
0.001
0.000227%
0.00341%
0.330%
0.000% - 0.000%
0.000% - 0.0427%
0.0558% -0.911%
100
2,000
192,900
0-0
0 - 25,000
32,600 - 533,200

Combined Ground &
Surface Water
0.01
0.004
0.001
0.00190%
0.0208%
0.699%
0.000% -0.0128%
0.00278% -0.0781%
0.356% -1.1 8%
2,000
21,800
731,300
0-13,400
2,900-81,700
372,400-1,237,100
2.1.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 5 PWSs nationally serving approximately 4,100 people could be exposed to
beryllium concentrations above 0.01 mg/L.  About 51 systems serving 44,400 people had estimated mean
concentrations greater than 0.004 mg/L. Approximately 696 systems serving about 1.5 million people
nationally were estimated to have beryllium concentrations greater than 0.001 mg/L. (See Section 1.4 for
a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 5 PWSs serving about 3,500 people nationally had mean
concentrations greater than 0.01 mg/L.  Approximately 50 systems serving about 36,800 people
nationally had estimated mean concentration values that exceeded 0.004 mg/L. About 653 ground water
systems serving just over 1 million people had estimated mean concentrations greater than 0.001 mg/L.

Approximately 1 surface water system serving 300 people was estimated to have a mean concentration of
beryllium above 0.01 mg/L. An estimated 1 surface water system serving 4,300 people had an estimated
mean concentration greater than 0.004 mg/L. An estimated 41 surface water systems serving
approximately 419,500 people had mean concentrations greater than 0.001 mg/L.
Occurrence Summary and Use Support Document
25
March 2002

-------
Table 2.1-12:  Estimated National Beryllium Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.01
0.004
0.001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
5
50
653
Range
0-14
24-81
540 - 764
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
3,500
36,800
1,002,500
Range
0-21,900
5,000 - 148,700
509,700-1,681,900

Surface Water
0.01
0.004
0.001
1
1
41
0-0
0-4
20-67
300
4,300
419,500
0-0
0 - 54,400
71,000-1,159,600

Combined Ground &
Surface Water
0.01
0.004
0.001
5
51
696
0-14
24-82
574 - 814
4,100
44,400
1,489,600
0 - 27,200
5,900 - 166,300
758,500-2,519,900
2.1.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of beryllium in drinking water are also
reviewed. Beryllium data contained in the National Inorganics and Radionuclides Survey (NIRS)
database are summarized below. Previously compiled occurrence information, from an OGWDW
summary document entitled "Beryllium in Water: An Assessment of Occurrence and Exposure"
(Ciccone & Associates, 1984), is presented in Section 2.1.5.2. In that review, a literature search was
conducted and knowledgeable sources within the Office of Water were contacted. Only one study was
found that addressed the occurrence of beryllium in finished water. (All information in Section 2.1.5.2 is
taken directly from "Beryllium in Water:  An Assessment of Occurrence and Exposure" (Ciccone &
Associates, 1984).) Note that the study presented in the following  section does not provide the
quantitative analytical results or comprehensive coverage that would enable direct comparison to the
occurrence findings estimated with the cross-section occurrence data presented in Section 2.1.4. These
additional studies, however, do enable a broader assessment of the Stage 2 occurrence estimates
presented for this Six-Year Review.

2.1.5.1 National Inorganics and Radionuclides Survey (NIRS)

In 1981, the USEPA's Office of Drinking Water (ODW) initiated the National Inorganics and
Radionuclides Survey (NIRS) to characterize the occurrence of various contaminants in community
drinking water supplies. The survey focused on the presence of 36 inorganics, including beryllium, and
four radionuclides in ground water supplies from throughout the  United States. Implementation of the
survey and sampling were accomplished by ODW's Technical Support Division (TSD) between July
1984 and October 1986.

The NIRS sampling program was designed to reflect the national distribution of community ground water
supplies by size of population served as inventoried by the Federal Reporting Data System (FRDS).  The
Occurrence Summary and Use Support Document
26
March 2002

-------
FRDS data was stratified into the following four population-size categories: very small (serving 25-500),
small (serving 501-3,300), medium (serving 3,301-10,000), and large/very large (serving >10,000). A
total of 1,000 sites were selected randomly from the FRDS data in proportion to the four size categories.
Approximately 2.1% of the supplies in each size category were chosen for sampling.  Of the  1,000
targeted sites, 990 were actually sampled in the NIRS.

Sample collection and location within each  supply were designed to reflect the quality of water actually
received by the consumer.  Samples were collected after three minutes of flushing in order to represent
the finished water in the distribution system. To the extent possible, the sampling location was chosen at
a point of maximum use in the distribution system. The method used to analyze for beryllium was not
reported. The minimum reporting limit (MRL) for beryllium was 0.001 mg/L.

Beryllium results are available for all of 989 sites sampled in the NIRS. The mean of the analytical
detections was 0.04 mg/L.  The maximum value detected was 0.21 mg/L.  Table 2.1-13 presents the
estimated percentages of beryllium exceedances and the total number of systems to exceed the given
threshold level.  Table 2.1-14 shows the estimated percentages of beryllium exceedances and the total
number of population to exceed the given threshold level.
Table 2.1-13:  Estimated Beryllium Exceedance as Reported in the NIRS - Systems
Threshold
0.01 mg/L
0.004 mg/L
0.001 mg/L
Percent of Systems That
Exceed Threshold
0.101%
0.101%
0.506%
Number of Systems
Estimated to Exceed
Threshold
60
60
301
Table 2.1-14:  Estimated Beryllium Exceedance as Reported in the NIRS - Population
Threshold
0.01 mg/L
0.004 mg/L
0.001 mg/L
Percent of Population Served
by Systems That Exceed
Threshold
0.038%
0.038%
0.106%
Total Population Served by
Systems Estimated to Exceed
Threshold
33,000
33,000
91,000
2.1.5.2 100 Largest Cities - Finished Water Survey

Only one study was identified which examined the concentration of beryllium in finished water from
treatment plants (Table 2.1-15).  Durfor and Becker (1962, as cited in Ciccone & Associates, 1984)
studied the occurrence of beryllium in raw and finished water from the 100 largest U.S. cities (1960
census). No detectable beryllium (detection limits 0.0001% of dissolved solids) was found in finished
water in any of the cities. However, beryllium was detected in raw water from the Lock Raven
Reservoir, Baltimore, MD at a concentration of 0.75 |ig/L.  The treatment process yielded finished water
that contained no detectable beryllium.
Occurrence Summary and Use Support Document
27
March 2002

-------
Table 2.1-15: Beryllium Occurrence in 100 Largest U.S. Cities
Survey
Public Water
Supplies of the
100 Largest
Cities in the
U.S., 1962.
Durfur and
Becker, 1962.
Sampling # Samples/ Mean System Survey Area Notes
Site Site Concentration Size
(Range) fig/L
100 G, S Variable ND U Nationwide 1 sample of raw water from Lock Raven
Reservoir, Baltimore, MD contained 0.75
l-ig/L. No Be detected (detection limits:
0.0001% of dissolved solids) following
treatment, plain sedimentation,
prechlorination, coagulation with alum,
sedimentation, rapid sand filtration and
adjustment of pH to 7.8 with lime.
2.1.6  Conclusion

Beryllium and many of its compounds are naturally occurring and found at low levels in soil, water, and
air. Beryllium metal has a wide range of uses in nuclear weapons and reactors. Beryllium alloys are
used in making electrical and electronic parts.  Industrial releases of beryllium and beryllium compounds
have been reported to TRI since 1988 from 21  States and 28 States, respectively. Releases to land, such
as spills or leaks within the boundaries of the reporting facility, constitute the greatest proportion of the
total on- and off-site releases of beryllium and beryllium compounds. Beryllium is also a national
NAWQA analyte.  Only 4% of all surface water sites had analytical detections of beryllium, compared to
0% of ground water sites. The percentage of surface water sites with analytical detections of beryllium
greater than the MCL (0.004 mg/L) was equal to 0.55%. The Stage 2 analysis, based on the 16-State
cross-section, estimated that approximately 0.0781% of combined ground water and surface water
systems serving 0.0208% of the population had estimated mean concentrations of beryllium greater than
the MCL of 0.004 mg/L.  Based on this estimate, approximately 51 PWSs nationally serving about
44,400 people are expected to have estimated mean concentrations of beryllium greater than 0.004 mg/L.

Beryllium is a naturally occurring element. Therefore, the balanced geographic distribution of the 16-
State cross-section should adequately cover the range of natural occurrence of beryllium from low to
high.  The 16-State cross-section also contains a substantial proportion of the States with reported TRI
releases. Based on this use and release evaluation, the 16-State cross-section appears to  adequately
represent beryllium occurrence nationally.

2.1.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1993. ToxFAQs for Beryllium. U.S.
        Department of Health and Human Services, Public Health Service.  Available on the Internet at:
        http ://atsdr. cdc .gov/tfacts4 .html

Agency for Toxic Substances and Disease Registry (ATSDR).  2000. Toxicological Profile for
        Beryllium. U.S. Department of Health and Human Services, Public Health Service. 239 pp. +
        Appendices. Available on the Internet at: http://www.atsdr.cdc.gov/toxprofiles/tp4.pdf

Ciccone, V.J & Associates.  1984. Beryllium in Water: An Assessment of Occurrence and Exposure.
        Prepared for and submitted to EPA on August 8, 1984.
Occurrence Summary and Use Support Document
28
March 2002

-------
Cunningham, L.D.  1998.  Beryllium, Mineral Yearbook. U.S. Geological Survey, pp. K1-K4.

Cunningham, L.D.  1999.  Beryllium, Mineral Commodity Summary.  U.S. Geological Survey, pp. 34-35.

Gilliom, R.J., D.K. Mueller, and L.H. Nowell. 1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589/, last updated October 9, 1998.

USEPA.  2000.  TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 2000. Mineral Commodity Summaries, February, 2000 - Beryllium.  Available on the Internet
       at: http://minerals.usgs.gov/minerals/pubs/commodity.

Need references for: Kopp and Kroner; Duram and Haffty; Silvey; Page; Durfor and Decker
Occurrence Summary and Use Support Document          29                                    March 2002

-------
2.2    Chromium
Table of Contents

2.2.1 Introduction, Use and Production  	  32
2.2.2 Environmental Release  	  34
2.2.3 Ambient Occurrence  	  35
2.2.4 Drinking Water Occurrence Based on the 16-State Cross-Section	  36
2.2.5 Additional Drinking Water Occurrence Data  	  42
2.2.6 Conclusion	  53
2.2.7 References  	  53
Tables and Figures

Table 2.2-1:  Imports of Chromium to the United States (thousand metric tons, gross weight)	  33

Table 2.2-2:  Chromium and Chromium Compound Manufacturers and Processors by State 	  33

Table 2.2-3:  Environmental Releases (in pounds) for Chromium in the United States,
       1988-1999	  34

Table 2.2-4:  Environmental Releases (in pounds) for Chromium Compounds in the
       United States, 1988-1999  	  35

Table 2.2-5:  Chromium Detections and Concentrations in Surface Water and Ground Water	  36

Table 2.2-6:  Stage 1 Chromium Occurrence Based on 16-State Cross-Section - Systems  	  37

Table 2.2-7:  Stage 1 Chromium Occurrence Based on 16-State Cross-Section - Population 	  38

Table 2.2-8:  Stage 2 Estimated Chromium Occurrence Based on 16-State Cross-Section -
       Systems	  39

Table 2.2-9:  Stage 2 Estimated Chromium Occurrence Based on 16-State Cross-Section -
       Population	  40

Table 2.2-10: Estimated National Chromium Occurrence - Systems and Population Served	  42

Table 2.2-11: Cumulative Occurrence of Chromium as Reported inNIRS	  43

Table 2.2-12: Estimated Chromium Exceedance as Reported in the NIRS - Systems	  44

Table 2.2-13: Estimated Chromium Exceedance as Reported in the NIRS - Population	  44

Table 2.2-14: National Summary of Chromium Violations in Ground Water Systems	  47
Occurrence Summary and Use Support Document         30                                    March 2002

-------
Table 2.2-15:  Estimated Cumulative National Occurrence of Chromium in Community
       Ground Water Supplies Based on the Delta Lognormal Distribution Model	  49

Table 2.2-16:  Estimated Cumulative Occurrence of Chromium in Surface Water Supplies
       Based on the Delta Lognormal Distribution Model 	  50

Table 2.2-17:  Population Exposed to Chromium at Levels Exceeding the Current MCL in
       Ground Water Supplies as Reported in FRDS 	  51

Table 2.2-18:  Estimated Cumulative Population (in thousands) Exposed to Chromium in
       Drinking Water Exceeding the Indicated Concentrations from Community
       Ground Water Supplies	  52

Table 2.2-19:  Estimated Cumulative Population (in thousands) Exposed to Chromium in
       Drinking Water Exceeding the Indicated Concentrations from Community
       Surface Water Supplies	  52
Occurrence Summary and Use Support Document         31                                    March 2002

-------
2.2.1  Introduction, Use and Production

Chromium (Cr) is a member of Group VIA of the periodic table. It is a naturally occurring element
found in rocks, animals, plants, soil, and in volcanic dust and gases.  Elemental chromium is a hard blue-
white metal that is generally obtained through reduction of the chromite ore (FeO-Cr2O3).  Chromium is
present in the environment in several different forms. The primary valence states of chromium are +3
and +6, which are generally the only ones found in natural waters (Battelle, 1984; USEPA, 1979, as cited
in Wade Miller, 1990). Chromium(III) occurs naturally in the environment and is an essential nutrient.
Chromium(VI) and chromium(O) are generally produced by industrial processes (ATSDR, 2001). No
taste or odor is associated with chromium compounds.

The major source of chromium in ground water is through leaching from rocks, mineral deposits, and
mining and mill wastes. The major sources of chromium in surface waters are leaching from rocks,
mineral deposits, soil runoff, industrial effluents, and deposition/precipitation of airborne particulates
(Wade Miller, 1990).

Chromium has a wide range of uses in metals, chemicals, and refractories. Chromium is one of the
United States' most important strategic and critical materials.  Chromium use in iron,  steel, and
nonferrous alloys enhances hardenability and resistance to corrosion and oxidation. The use  of chromium
to produce stainless steel and nonferrous alloys are two of its more important applications. Chromium is
also used in bricks in furnaces, dyes and pigments, and for chrome plating, leather tanning, and wood
preserving. Other applications are  in making pigments, leather processing, catalysts, surface treatments,
and refractories (USGS, 2001).

In the United States, most chromite ore is smelted to produce stainless steel and non-ferrous  alloy
(ATSDR, 2000).  The latter is used primarily in the production of steel to improve stiffness, hardness,
and strength.  Almost all of the chromite ore used in steel production in the United States is imported (see
Table 2.2-1; USGS, 2001). Table 2.2-2 provides further information by State of the widespread
manufacture and  processing of chromium (ATSDR, 2000).

Drinking water generally contains the same chromium levels as the surface and ground waters, which
serve as its source.  Although some piping materials contain significant levels of chromium (corrosion
resistant steel, 8-14%; cement, 5-120 ppm chromium), little is leached into the water.  However, it should
be noted that chromium (III) may be oxidized to chromium (VI) during the chlorination process (NLM,
2001).

The two largest sources of chromium emission in the atmosphere are from the chemical manufacturing
industry and combustion of natural gas, oil, and coal. Other sources include wind transport from road
dust, cement producing plants (since cement contains chromium), the wearing down of asbestos brake
linings from automobiles or similar sources of wind carried asbestos  (since asbestos contains chromium),
incineration of municipal refuse and sewage sludge, exhaust emission from automotive catalytic
converters, emissions from cooling towers that use chromium compounds as rust inhibitors, waste waters
from electroplating, leather tanning, and textile  industries when discharged into surface waters, and solid
wastes from chemical manufacture  of chromium compounds or from municipal incineration  when
disposed  of improperly in landfill sites (NLM, 2001).
Occurrence Summary and Use Support Document         32                                     March 2002

-------
Table 2.2-1: Imports of Chromium to the United States (thousand metric tons, gross weight)
Import for Consumption
chromite ore
1995
416
1996
362
1997
350
1998
381
1999
476
    Source: USGS, 2001
Table 2.2-2: Chromium and Chromium Compound Manufacturers and Processors by State
State"
AK
AL
AR
AZ
CA
CO
CT
DE
FL
GA
HI
IA
ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
PR
RI
SC
SD
TN
TX
UT
VA
VT
WA
Number of facilities
1
87
49
31
147
33
66
6
45
90
9
55
6
212
199
38
72
34
79
32
18
162
49
74
49
4
91
5
22
20
48
7
6
94
333
67
37
316
4
14
72
8
81
177
28
51
4
39
Range of maximum amounts on
site in thousands of pounds'"
0-99,999
1,000-999,999
100-999,999
100-999,999
0-999,999,999
100-999,999
100-9,999,999
1,000-999,999
100-9,999,999
0-49,999,999
0-99,999
100-9,999,999
100,000-49,999,999
0-9,999,999
0-99,999,999
0-999,999
0-9,999,999
100-9,999,999
100-999,999
100-49,999,999
1,000-999,999,999
0-9,999,999
100-9,999,999
100-99,99,999
1,000-999,999
10,000-99,999
0-999,999,999
1,000-999,999
1,000-999,999
1,000-99,999
100-9,999,999
1,000-999,999
10,000-99,999
0-9,999,999
0-999,999,999
100-49,999,999
1,000-999,999
0-999,999,999
0-99,999
1,000-999,999
0-999,999,999
1,000-99,999
0-9,999,999
0-99,999,999
0-999,999
0-9,999,999
0-999,999
1 00-9 999 999
Activities and uses0
11
1,2,3,5,6,7,8,9,10,11,12,13
1,2,3,4,5,8,9,11,12,13
1,2,3,4,5,6,8,9,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,7,8,9,11,12
1,2,3,4,5,7,8,9,10,11,12,13
1,2,3,4,5,7,8,9
5,7,8,9,10,11,13
1,2,3,4,5,6,7,8,9,10,11,12,13
8
1,2,3,5,7,8,9,10,12,13
1,5,7,9
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
2,3,7,8,9,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,7,8,9,10,11,12
1,2,3,4,5,7,8,9,10,11,12,13
1,2,3,4,5,6,7,8,9,10,11
1,3,5,7,8,9,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,3,5,7,8,9,10,12,13
1,2,5,7,8,9,10,11,12,13
1,5,7,8,9,11,12,13
7,8,13
1,2,3,4,5,6,7,8,9,10,12,13
2,3,9
1,2,3,5,8,9,11,12,13
8,9,11,12
1,2,3,4,5,7,8,9,10,11,12,13
1,2,3,5,6,8,9,12
4,8,9,10
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,6,8,9,10,11,12,13
1,2,3,5,8,9,10,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,5,9,11
1,2,3,8,9,10,13
1,2,3,4,5,7,8,9,10,11,12,13
2,3,8,9
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,7,8,9,11,13
1,2,4,5,7,8,9,10,12,13
3,4,9
1 7. 3 6 7 8 9 1 7. 1 3
Occurrence Summary and Use Support Document
33
March 2002

-------
 State"
               Number of facilities
                       Range of maximum amounts on
                       site in thousands of pounds'1
                                                                Activities and uses0
 WI
 WV
 WY
197
27
3	
0-9,999,999
100-9,999,999
0-99.999
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,5,6,7,8,9,10,12,13
1.6.9	
aPost office State abbreviations used
b Range represents maximum amounts on site reported by facilities in each State
cActivities/Uses
1. Produce                         8. As a formulation component
2. Import                         9. As an article component
3. For on-site use/processing             10. For repackaging only
4. For sale/distribution                 11. As a chemical processing aid
5. As a byproduct                    12. As a manufacturing aid
6. As an impurity                     13. Ancillary or other uses
7. As a reactant

Source: AT SDR, 2000 compilation of 1997 TRI data
2.2.2 Environmental Release

Chromium and chromium compounds are listed as Toxics Release Inventory (TRI) chemicals.  Table 2.2-
3 illustrates the environmental releases of chromium from 1988 to 1999.  (There are only chromium data
for these years.)  Releases to land, constituting most of the on-site releases of chromium, gradually
decreased from 1988 to  1995 and then dropped dramatically in  1996.  It is unclear whether these sharp
chromium decreases are real or a function of changes in TRI reporting requirements in the late  1980s and
early 1990s (see discussion in the Introduction). Since 1995, land releases have fluctuated modestly for
chromium with no apparent trend. Air emissions are also an important mode of on-site release.  Though
the first several years are markedly higher for air emissions of chromium, no trend is apparent for the
remaining years.  Surface water discharges and underground injection of chromium are less significant
on-site releases.  Off-site releases of chromium, a large component of total releases, are considerable.
Though there was a large drop in total chromium releases in 1991 when compared to previous years, the
late  1990s show a steady increase in pounds released.  The TRI data for chromium were reported from 49
States with the exception of Hawaii and included Puerto Rico (USEPA,  2000).  All 16 of the cross-
section States (used for analyses of chromium occurrence in drinking water; see Section 2.2 A) reported
releases of chromium. (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 2.2-3:  Environmental Releases (in pounds) for Chromium in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
On-Site Releases
Air Emissions
300,814
478,733
305,113
434,107
418,408
622,337
599,014
492,851
427,901
434,330
Surface Water
Discharges
11,236
13,112
11,914
574,456
17,266
20,566
24,246
23,341
17,857
41,265
Underground
Injection
56
9
1
7
33
48
269
333
531
95
Releases
to Land
715,100
685,766
514,449
535,711
1,109,958
1,176,881
1,063,935
1,138,756
1,209,587
2,738,681
Off-Site Releases
15,773,121
12,938,825
5,552,216
6,022,527
5,771,254
5,877,781
7,901,753
6,047,944
6,576,924
10,585,379
Total On- &
Off-site
Releases
16,800,327
14^16,445
6,383,693
7,566,808
7,316,919
7,697,613
9,589,217
7,703,225
8,232,800
13,799,750
Occurrence Summary and Use Support Document
                                   34
                                                     March 2002

-------
Year
1989
1988
On-Site Releases
Air Emissions
945,625
566,498
Surface Water
Discharges
69,681
75,442
Underground
Injection
693
2,249
Releases
to Land
3,369,485
9,282,766
Off-Site Releases
9,865,483
11,710,612
Total On- &
Off-site
Releases
14,250,967
21,637,567
 Source: USEPA, 2000
Table 2.2-4 illustrates the environmental releases for chromium compounds.  Releases to land, ranging
from about 20 to 30 million pounds, constitute most of the on-site releases of chromium compounds from
1988 to 1999. Though the first several years of recorded air emissions of chromium compounds are
slightly higher, no real trend is apparent. Surface water discharges and underground injection of
chromium compounds are less significant on-site releases.  Surface water discharges of chromium
compounds have gradually decreased since 1988.  Underground injections of chromium compounds
showed no discernable trend besides a sharp increase from 1995 to 1997. Off-site releases of chromium
compounds fluctuate throughout the years, with highest releases in 1995 and 1997, although no other
trends are apparent. The TRI data for chromium compounds were reported from 49 States with the
exception of Vermont and included Puerto Rico (USEPA, 2000).  All 16 of the cross-section States (used
for analyses of chromium occurrence in drinking water; see Section 2.2.4) reported releases of chromium
compounds. (For a map of the 16-State cross-section, see Figure  1.3-1.)
Table 2.2-4: Environmental Releases (in pounds) for Chromium Compounds in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
467,071
347,629
382,843
420,051
650,311
547,907
405,083
515,820
606,278
721,150
1,398,526
764,851
Surface Water
Discharges
97,379
112,520
100,082
138,591
138,551
159,655
230,548
276,401
343,298
409,916
475,850
326,027
Underground
Injection
816,717
874,795
1,131,559
1,193,808
1,084,747
38,061
42,493
32,137
34,603
83,147
59,110
52,653
Releases
to Land
29,591,378
30,241,083
29,271,102
26,463,701
22,090,165
22,185,322
24,634,864
25,085,551
26,267,556
23,445,165
31,081,444
30,938,106
Off-Site Releases
14,203,028
16,206,382
21,003,330
14,685,755
20,389,031
16,029,364
12,975,431
12,749,479
13,735,176
15,898,149
18,526,178
14,898,699
Total On- &
Off-site
Releases
45,175,573
47,782,409
51,888,916
42,901,906
44,352,805
38,960,309
38,288,419
38,659,388
40,986,911
40,557,527
51,541,108
46,980,336
 Source: USEPA, 2000
2.2.3  Ambient Occurrence

Chromium is an analyte for both surface and ground water NAWQA studies, with a method detection
limit (MDL) of 0.01 mg/L. Additional information on analytical methods used in the NAWQA study
units, including method detection limits, are described by Gilliom and others (1998).
Occurrence Summary and Use Support Document
35
March 2002

-------
Typical of many inorganic contaminants, chromium occurrence in ambient surface waters is high (Table
2.2-5). This is not surprising, considering that chromium and its compounds are used in many products.
Contrary to occurrence of chromium in surface waters, no chromium was detected in ground waters.
This is not surprising, however, considering the bonding tendencies of chromium with organic matter in
soil particles and the low tendency for chromium to be found in leachate because of its low mobility in
soil. Detection frequencies are also greater for surface water than for ground water, possibly because
surface waters are more sensitive to anthropogenic releases. For all surface water samples, the median
concentration is 0.001 mg/L and the 99th percentile concentration is 0.006 mg/L.  However, chromium
detection frequencies exceeding the MCL (0.1 mg/L) do not occur for surface water samples. This
could be expected because surface waters subject to large anthropogenic inputs of chromium are more
easily diluted by waters integrated from other parts of the watershed where chromium concentrations
may be lower.
Table 2.2-5:  Chromium Detections and Concentrations in Surface Water and Ground Water

                         Detection frequency        Detection frequency      Concentration percentiles
                              > MDL*                 > MCL*             (all samples; mg/L)

                        % samples     % sites     % samples     % sites       median         99th

    surface water           27.4%      66.0%       0.0%       0.0%       0.001         0.006


    ground water            0.0%       0.0%        0.0%       0.0%        N/A          N/A

* The Minimum Reporting Level (MRL) for chromium in water is 0.01 mg/L and the Maximum Contaminant Level (MCL) is 0.1 mg/L.


2.2.3.2 Additional Ambient Occurrence Data

A summary document entitled "Occurrence and Exposure Assessment for Chromium in Public Drinking
Water Supplies" (Wade Miller, 1990), was previously prepared for past USEPA assessments of
chromium. However, no information on the ambient occurrence of chromium was included in that
document. (The document did include information regarding chromium occurrence in drinking water,
which is discussed in Section 2.2.5 of this report.)

2.2.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of chromium occurrence presented in the following section is based on State  compliance
monitoring data from the  16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest: 0.1 mg/L; 0.07 mg/L; 0.05 mg/L; 0.02 mg/L; and
0.01 mg/L.

All sixteen cross-section State data sets contained occurrence data for chromium.  These data represent
over 65,000 analytical results from approximately 20,000  PWSs during the period from 1983 to 1998
(with most analytical results from 1992 to 1997).  The number of sample results and PWSs vary by State,
although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness. The overall modal detection limit for chromium in the 16 cross-section States is equal to
0.01 mg/L. (For details regarding the 16-State cross-section, please refer to Section 1.3.5  of this report.)

Occurrence Summary and Use Support Document          36                                     March 2002

-------
2.2.4.1 Stage 1 Analysis Occurrence Findings

Table 2.2-6 illustrates the Stage 1 analysis of chromium occurrence in drinking water for the public water
systems in the 16-State cross-section relative to five thresholds: 0.1 mg/L (the current MCL), 0.07 mg/L,
0.05 mg/L, 0.02 mg/L, and 0.01 mg/L (the modal MRL). A total of 25 ground water and surface water
PWSs (approximately 0.127%) had at least one analytical result exceeding the MCL (0.1 mg/L); 0.513%
(101 systems) of PWSs had at least one analytical result exceeding 0.05 mg/L; and 2.83% (557 systems)
of PWSs had at least one analytical result exceeding 0.01 mg/L.

Approximately 0.110% (20 systems) of ground water PWSs had at least one analytical result greater than
the MCL (0.1 mg/L). About 0.484% (88 systems) of ground water PWSs had at least one analytical
result above 0.05 mg/L. The percentage of ground water systems with at least one result greater than
0.01 mg/L was equal to 2.81 % (511 systems).

Only 5 (0.328% of) surface water systems had at least one analytical result greater than the MCL (0.1
mg/L). A total of 13 (0.852% of) surface  water systems had at least one analytical result greater than
0.05 mg/L. Forty-six surface water systems (3.01%) had at least one analytical result exceeding 0.01
mg/L.
Table 2.2-6:  Stage 1 Chromium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.1
0.07
0.05
0.02
0.01
Percent of Systems
Exceeding Threshold
0.110%
0.352%
0.484%
1.33%
2.81%
Number of Systems
Exceeding Threshold
20
64
88
242
511

Surface Water
0.1
0.07
0.05
0.02
0.01
0.328%
0.524%
0.852%
1.57%
3.01%
5
8
13
24
46

Combined Ground &
Surface Water
0.1
0.07
0.05
0.02
0.01
0.127%
0.366%
0.513%
1.35%
2.83%
25
72
101
266
557
Occurrence Summary and Use Support Document
37
March 2002

-------
Reviewing chromium occurrence in the 16 cross-section States by PWS population served (Table 2.2-7)
shows that approximately 0.666% (over 702,000 people) of the population was served by PWSs with at
least one analytical result of chromium greater than the MCL. Almost 10.5 million people (9.95%) were
served by systems with an exceedance of 0.05 mg/L. Approximately 14 million people (13.3%) were
served by systems with at least one analytical result greater than 0.01 mg/L.

The percentage of population served by ground water systems with analytical results greater than 0.1
mg/L was equal to 0.543% (244,300 people). When evaluated relative to 0.05 mg/L or 0.01 mg/L, the
percent of population exposed was equal to 1.76% (almost 789,000 people) and 8.83% (almost 4 million
people), respectively.

The percentage of population served by surface water systems with exceedances of 0.1 mg/L was equal
to 0.758% (about 458,000 people). The percentage of population served by surface water systems
dramatically increased to 16.1% (approximately 9.7 million people) when evaluated relative to 0.05
mg/L. The percentage of population exposed to concentrations greater than 0.01 mg/L was equal to
16.6% (over 10 million people).
Table 2.2-7:  Stage 1 Chromium Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.1
0.07
0.05
0.02
0.01
Percent of Population Served
by Systems
Exceeding Threshold
0.543%
1.03%
1.76%
4.49%
8.83%
Total Population Served by
Systems Exceeding
Threshold
244,300
462,500
789,000
2,020,400
3,967,200

Surface Water
0.1
0.07
0.05
0.02
0.01
0.758%
13.0%
16.1%
16.2%
16.6%
457,800
7,876,700
9,699,800
9,804,900
10,055,100

Combined Ground &
Surface Water
0.1
0.07
0.05
0.02
0.01
0.666%
7.91%
9.95%
11.2%
13.3%
702,100
8,339,100
10,488,800
11,825,300
14,022,300
Occurrence Summary and Use Support Document
38
March 2002

-------
2.2.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 2.2-8 and 2.2-9.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

One (0.00424%) PWS in the 16 States was estimated to have a mean concentration of chromium above
0.1 mg/L (the current MCL).  Approximately 7 (0.0366% of) PWSs in the 16-State cross-section are
estimated to have mean concentrations greater than 0.05 mg/L.  The percentage of PWSs in the 16 States
with estimated mean concentrations exceeding 0.01 mg/L (the modal detection limit) was about 1.56%
(307) PWSs nationally.

A significantly greater proportion of ground water systems, as compared to surface water systems, were
estimated to exceed each threshold. Approximately 1 (0.00459% of) ground water system in the 16
States had an estimated mean concentration of chromium above 0.1 mg/L, compared to zero surface
water systems. About 0.0393% (an estimated 7 PWSs) of the 16 States' ground water systems had
estimated mean concentrations greater than 0.05 mg/L. This compares with about 0.00406% (about 1
system) of the surface water systems with estimated mean concentrations greater than 0.05 mg/L.  The
estimated mean concentration values for approximately 299 (1.65% of) ground water PWSs in the 16-
State cross-section exceed 0.01 mg/L, compared to only 7 (0.474% of) surface water systems with
estimated mean concentrations exceeding the modal detection limit.
Table 2.2-8:  Stage 2 Estimated Chromium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.1
0.07
0.05
0.02
0.01
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.00459%
0.0144%
0.0393%
0.396%
1.65%
Range
0.000% -0.0165%
0.000% - 0.0385%
0.011% -0.0716%
0.292% -0.501%
1.4% -1.89%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
1
3
7
72
299
Range
0-3
0-7
2-13
53-91
255 - 343

Surface Water
0.1
0.07
0.05
0.02
0.01
0.000%
0.000786%
0.00406%
0.0696%
0.474%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0655%
0.000% - 0.262%
0.1 31% -0.786%
0
1
1
1
7
0-0
0-0
0-1
0-4
2-12
Occurrence Summary and Use Support Document
39
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
Range
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.1
0.07
0.05
0.02
0.01
0.00424%
0.0133%
0.0366%
0.371%
1.56%
0.000% -0.0152%
0.000% - 0.0355%
0.0152% -0.0660%
0.274% - 0.467%
1.33% -1.78%
1
3
7
73
307
0-3
0-7
3-13
54-92
262 - 350
Reviewing chromium occurrence by PWS population served (Table 2.2-9) shows that approximately
0.00139% (an estimate of approximately 1,500 people) of the population served by PWSs in the 16 States
were potentially exposed to chromium levels above 0.1 mg/L.  When evaluated relative to a threshold of
0.05 mg/L, the percent of the population exposed estimated about 0.0108% (over 11,000 people served
by systems in the 16 cross-section States). The percentage of population served by PWSs with estimated
mean concentrations greater than 0.01 mg/L was approximately 0.431% (approximately 453,700 people).

About 1,500 people (0.00325% of the population served by ground water systems in the 16 States) were
served by systems with estimated mean concentrations of chromium above 0.1 mg/L. An estimated
0.0248% (over 11,000 people) of the 16-State population were served by ground water systems whose
mean concentration value exceeded 0.05 mg/L.  The percentage of population served by ground water
PWSs with estimated mean concentration values exceeding 0.01 mg/L was approximately 0.893% (about
401,300 people).

Zero surface water PWSs (therefore, none of the population served by surface water systems) had
estimated mean concentrations of chromium above 0.1 mg/L. About 0.000336% (approximately 200
people in the 16 States) of the population served by surface water PWSs had mean concentrations greater
than 0.05 mg/L. An estimated 52,300 (0.0866% of) people served by surface water in the 16-State cross-
section were served by systems with estimated mean concentrations of chromium above 0.01 mg/L.  The
percentage (and estimated number) of population served by surface water systems that exceeded each
threshold was always less than the percentage of ground water systems that exceeded the threshold.
Table 2.2-9: Stage 2 Estimated Chromium Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.1
0.07
0.05
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.00325%
0.00987%
0.0248%
Range
0.000% -0.0186%
0.000% -0.1 13%
0.00124% -0.130%
Total Population Served by Systems in th<
16 States Estimated to Exceed Threshold
Best Estimate
1,500
4,400
11,100
Range
0 - 8,400
0 - 50,700
600 - 58,600
Occurrence Summary and Use Support Document
40
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.02
0.01
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.228%
0.893%
Range
0.0959% -0.357%
0.635% -1.32%
Total Population Served by Systems in th<
16 States Estimated to Exceed Threshold
Best Estimate
102,600
401,300
Range
43,100-160,500
285,500 - 592,500

Surface Water
0.1
0.07
0.05
0.02
0.01
0.000%
0.000104%
0.000336%
0.00667%
0.0866%
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.00182%
0.000% - 0.0532%
0.00604% - 0.358%
0
100
200
4,000
52,300
0-0
0-0
0- 1,100
0 - 32,100
3,700-216,100

Combined Ground &
Surface Water
0.1
0.07
0.05
0.02
0.01
0.00139%
0.00427%
0.0108%
0.101%
0.431%
0.000% - 0.00793%
0.000% -0.0481%
0.000580% - 0.0559%
0.0447% -0.1 59%
0.298% - 0.649%
1,500
4,500
11,300
106,600
453,700
0 - 8,400
0 - 50,600
600 - 58,900
47,100 - 167,700
313,700-683,900
2.2.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold (see Table 2.2-10), an estimated 3 PWSs serving approximately 3,000 people nationally could
be exposed to chromium concentrations above 0.1 mg/L.  About 24 systems serving almost 23,000
people had estimated mean concentrations greater than 0.05 mg/L. Approximately 1,013 systems serving
about 917,000 people nationally were estimated to have chromium concentrations greater than 0.01
mg/L. (See Section  1.4 for a description of how Stage 2 16-State estimates are extrapolated to national
values.)

For ground water systems, an estimated 3 PWSs serving about 2,800 people nationally had mean
concentrations greater than 0.1 mg/L. Approximately 23 ground water systems serving about 21,200
people nationally had estimated mean concentration values that exceeded 0.05 mg/L.  About 980 ground
water systems serving almost 765,000 people had estimated mean concentrations greater than 0.01 mg/L.

Zero surface water systems were estimated to have mean concentrations of chromium above 0.1 mg/L.
An estimated 1 surface water system serving 400 people had an estimated mean concentration greater
than 0.05 mg/L.  About 26 surface water systems serving approximately 110,200 people had estimated
mean concentrations greater than 0.01 mg/L.
Occurrence Summary and Use Support Document
41
March 2002

-------
Table 2.2-10:  Estimated National Chromium Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.1
0.07
0.05
0.02
0.01
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
3
9
23
236
980
Range
0-10
0-23
7-43
173-298
834 - 1,122
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
2,800
8,500
21,200
195,500
764,900
Range
0-15,900
0 - 96,600
1,100-111,600
82,200 - 306,000
544,200 - 1,129,300

Surface Water
0.1
0.07
0.05
0.02
0.01
0
1
1
4
26
0-0
0-0
0-4
0-15
7-44
0
100
400
8,500
110,200
0-0
0-0
0 - 2,300
0 - 67,700
7,700 - 455,400

Combined Ground &
Surface Water
0.1
0.07
0.05
0.02
0.01
3
9
24
241
1,013
0-10
0-23
10-43
178 - 304
865-1,156
3,000
9,100
22,900
215,600
917,000
0 - 16,900
0 - 102,400
1,200-119,000
95,200 - 338,900
634,100 - 1,382,400
2.2.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of chromium in drinking water are also
reviewed. Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Exposure Assessment for Chromium in Public Drinking Water Supplies" (Wade Miller,
1990), is presented in this section. This variety of studies and information are presented regarding levels
of chromium in drinking water, with the scope of the reviewed studies ranging from national to regional.
Note that none of the studies presented in the following section provide the quantitative analytical results
or comprehensive coverage that would enable direct comparison to the occurrence findings estimated
with the cross-section occurrence data presented in Section 2.2.4.  These additional studies, however, do
enable a broader assessment of the Stage 2 occurrence estimates presented for this Six-Year Review. All
the following information in Section 2.2.5 is taken directly from "Occurrence and Exposure Assessment
for Chromium in Public Drinking Water Supplies" (Wade Miller,  1990).

2.2.5.1 National Inorganics and Radionuclides Survey (NIRS)

In 1981, the USEPA's Office of Drinking Water (ODW) initiated the National Inorganics and
Radionuclides Survey (NIRS) to characterize the occurrence of various contaminants in community
drinking  water supplies. The survey focused on the presence of 36 inorganics, including chromium,  and
Occurrence Summary and Use Support Document
42
March 2002

-------
four radionuclides in ground water supplies from throughout the United States.  Implementation of the
survey and sampling were accomplished by ODW's Technical Support Division (TSD) between July
1984 and October 1986.

The NIRS sampling program was designed to reflect the national distribution of community ground water
supplies by size of population served as inventoried by the Federal Reporting Data System (FRDS).  The
FRDS data was stratified into the following four population-size categories: very small (serving 25-500),
small (serving 501-3,300), medium (serving 3,301-10,000), and large/very large (serving >10,000).  A
total of 1,000 sites were selected randomly from the FRDS data in proportion to the four size categories.
Approximately 2.1% of the supplies in each size category were chosen for sampling.  Of the 1,000
targeted sites, 990 were actually sampled in the NIRS.

Sample collection and location within each supply were designed to reflect the quality of water actually
received by the consumer. Samples were collected after three minutes of flushing in order to represent
the finished water in the distribution system. To the extent possible, the sampling location was chosen at
a point of maximum use in the distribution system.  The method used to analyze for chromium was not
reported. The minimum reporting limit (MRL) for chromium was 0.002 mg/L.

Table 2.2-11 presents the NIRS results in terms of the cumulative number of systems exceeding the
indicated concentrations for each of the population size strata. Chromium results are  available for all of
990 sites sampled in the NIRS. Approximately 94 percent of the samples had chromium concentrations
below the MRL of 0.002 mg/L.  Six percent (63) of the samples contained chromium in concentrations
equal to or greater than 0.002 mg/L. The mean of the positives was 0.007 mg/L. The maximum value
detected was 0.041 mg/L which is below the current MCL of 0.05 mg/L and the proposed MCL of 0.1
mg/L.  Table 2.2-12 presents the estimated percentages of chromium exceedances and the total number of
systems to exceed the given threshold level. Table 2.2-13 shows the estimated percentages of chromium
exceedances and the total number of population to exceed the given threshold level.
Table 2.2-11:  Cumulative Occurrence of Chromium as Reported in NIRS
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
Number of
Systems Sampled
338
337

113
120

54

22
3
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002
27
21

2
4

8

1
0
>0.05
0
0

0
0

0

0
0
>0.1
0
0

0
0

0

0
0
>0.2
0
0

0
0

0

0
0
Occurrence Summary and Use Support Document
43
March 2002

-------
System size
(population served)
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Number of
Systems Sampled
2
0
1
990
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002
0
0
0
63
>0.05
0
0
0
0
>0.1
0
0
0
0
>0.2
0
0
0
0
Table 2.2-12: Estimated Chromium Exceedance as Reported in the NIRS - Systems
Threshold
0.10 mg/L
0.07 mg/L
0.05 mg/L
0.02 mg/L
0.01 mg/L
Percent of Systems That
Exceed Threshold
0.00%
0.00%
0.00%
0.51%
0.81%
Number of Systems Estimated
to Exceed Threshold
0
0
0
301
481
Table 2.2-13: Estimated Chromium Exceedance as Reported in the NIRS - Population
Threshold
0.10 mg/L
0.07 mg/L
0.05 mg/L
0.02 mg/L
0.01 mg/L
Percent of Population Served
by Systems That Exceed
Threshold
0.00%
0.00%
0.00%
0.16%
0.48%
Total Population Served by
Systems Estimated to Exceed
Threshold
0
0
0
137,876
411,201
2.2.5.2 Federal Survey Data

Several national-scale surveys, in addition to the NIRS Survey, have been conducted that provide data on
chromium in public drinking water supplies. These include the 100 Largest Cities Study, the 1969
Community Water Supply Study, the 1978 Community Water Supply Survey, the Rural Water Survey,
and the National Organic Monitoring Survey. The following sections describe those surveys and present
the data on chromium levels in ground water and surface water supplies.  It should be noted that none of
the surveys differentiate between chromium in the (III) and (VI) valence states.
Occurrence Summary and Use Support Document
44
March 2002

-------
2.2.5.2.1 100 Largest Cities Study

Durfor and Becker (1964, as cited in Wade Miller, 1990) reported on the water quality, use, and
treatment at public water supplies serving the 100 largest U.S. cities at the time of the 1960 census.
Included were 66 cities that used only surface water, 20 that used only ground water, and 14 that used
mixed surface and ground water sources.  Both raw and finished water samples were taken for most of
the locations.

Chromium was measured using a flame emission spectrographic technique. There was not sufficient
information provided to discern the minimum detectable or minimum quantifiable concentration. Only
very generalized results are available for chromium. Chromium concentrations ranged from undetectable
to 0.035 mg/L. The majority of finished water samples with chromium present had concentrations below
0.005 mg/L.

2.2.5.2.2 1969 Community Water Supply Study (1969 CWSS)

The U.S. Public Health Service (USPHS) conducted the Community Water Supply Study (CWSS) in
1969 to assess the Nation's water supply facilities and drinking water quality (McCabe et al., 1970;
USPHS, 1970; both as cited in Wade Miller, 1990). Finished water from 969 community supplies
located in  9 geographically distributed areas was studied. These areas were Vermont; New York, New
York; Charleston, West Virginia; Charleston, South Carolina; Cincinnati, Ohio; Kansas City, Missouri-
Kansas; New Orleans, Louisiana; Pueblo, Colorado; and San Bernardino-Riverside-Ontario, California.
Except for Vermont (in which all supplies in the State were sampled), the study locations are standard
metropolitan statistical areas (SMSAs). Water samples were taken at randomly selected sites in the
distribution system after flushing for several seconds.

Of the 969 systems studied, 678 were groundwater supplies; 109 were surface water supplies; and the
remaining 182 were mixed sources, purchased water, or of unspecified source. Analytical results for
chromium were provided for 676 ground water supplies and 109 surface water supplies.

Results were published in several volumes addressing each of the study areas and the national findings.
The published report did not provide complete data on the water source, population served, or chromium
level measured for each system sampled.  However, a computer file with the requisite data was prepared
by Science Applications International Corporation (SAIC,  1987, as  cited in Wade Miller, 1990) using the
published  data, and additional information was provided by Dr. Rolf Deininger of the University of
Michigan, and EPA staff.

The information does not specify a detection limit nor a minimum quantifiable concentration for
chromium. The "negative" findings for individual supplies were reported as having a value of 0 mg/L.

Overall, chromium was observed in 107 of the 676 ground water supplies (16 percent) for which data are
available.  Chromium was also present in 35 of the 109 surface water supplies sampled (32 percent) at
concentrations of 0.0005-0.21 mg/L.

2.2.5.2.3 1978 Community Water Supply Survey (1978 CWSS)

The  1978 Community Water Supply Survey (CWSS) was conducted by EPA to determine the occurrence
of organic and inorganic compounds in public water supplies throughout the U.S. Drinking water
samples were provided by approximately 500 supplies; however, due to  analytical problems, reliable data
Occurrence Summary and Use Support Document         45                                     March 2002

-------
for chromium were available for only 47 ground water and 12 surface water supplies (USEPA, 1983, as
cited in Wade Miller, 1990).

Details on the analytical method used for chromium were not available; from the information provided by
Glick (1984, as cited in Wade Miller, 1990) the minimum quantifiable concentration appeared to be
0.005 mg/L. Supplies provided one to five samples of raw, finished, and/or distribution water. However,
Brass (1983, as cited in Wade Miller, 1990) indicated that reporting inconsistencies made it impossible to
distinguish between finished and distribution samples. Therefore, distribution and finished sample
results were averaged; raw water data were not used.

Chromium was not found above the (apparent) minimum quantifiable concentration of 0.005 mg/L in any
of the 47 ground water or 12 surface water supplies for which analytical data are available.

2.2.5.2.4 Rural Water Survey (RWS)

The Rural Water Survey (RWS) conducted between 1978 and 1980 evaluated the status of drinking water
in rural America as required by Section 3 of the Safe Drinking Water Act. More than 2,000 households
served by 648 public water supplies (494 ground water,  154 surface water) were surveyed.  Many of
these households used private wells or very  small systems serving fewer than 25 people, and only a
subsample of the supplies evaluated included analyses for chromium (71 ground water and 21 surface
water supplies). Results of the inorganic analyses were provided to Science Applications International
Corporation (SAIC, 1987, as cited in Wade Miller, 1990) as a computer file by Brower (1983, as cited in
Wade Miller, 1990).

A problem with the RWS was that the number of service connections associated with water systems was
reported in lieu of the actual populations served by the systems.  Dr. Bruce Brower of Cornell University,
who collaborated in the National Statistical Assessment of Rural Water Conditions (based on the RWS
data), provided a factor to convert the data from  service connections to the number of people served,
based on the average number of persons per household observed in the RWS. It must be noted, however,
that these population values are only approximations. Details were not available on the sample collection
nor the analytical methodology used. However,  the minimum quantifiable concentration for chromium
appeared to be 0.005 mg/L.  Of the 71 ground water supplies studied, 4 (6 percent) contained chromium,
all at levels of 0.005 mg/L.

For surface water supplies, 2 of the 21 supplies sampled had  chromium present, also at levels of 0.005
mg/L.

2.2.5.2.5 National Organic Monitoring Survey (NOMS)

EPA conducted the National Organic Monitoring Survey (NOMS) in  1976 and 1977, primarily to provide
data for establishing MCLs for organic compounds in drinking water. A substudy of NOMS analyzed
samples for 27 trace elements in the 113 supplies sampled (91 surface water, 19 ground water, 3 mixed
sources) (USEPA, 1980, as cited in Wade Miller, 1990). Chromium data are available for 90 surface
water and 18 ground water supplies. Samples were taken of treated, finished water leaving the treatment
plant; occasionally, samples were taken of distribution system water in close proximity to the treatment
plant. The results of the inorganic analyses  for NOMS were  provided to SAIC by the Technical Support
Division of EPA's Office of Drinking Water (USEPA, 1985,  as cited in Wade Miller, 1990). No
information on the analytical procedures used was available; the minimum quantifiable concentration
appeared to be approximately 0.003 mg/L.
Occurrence Summary and Use Support Document         46                                     March 2002

-------
Chromium was found in 3 of the 18 ground water supplies sampled (17 percent) at concentrations
ranging from 0.0045 to 0.0895 mg/L. Only one supply had a value exceeding the current MCL of 0.05
mg/L.  Chromium was not detected in concentrations greater than the proposed MCL of 0.1 mg/L.

For surface water supplies, chromium was observed in 24 of the 90 supplies sampled (27 percent) at
concentrations ranging from 0.0032 mg/L to 0.0522 mg/L.  Only one supply had a value exceeding the
current MCL of 0.05 mg/L and none of the concentrations were greater than the proposed MCL of 0.1
mg/L.

2.2.5.2.6  Compliance Monitoring Data

The Federal Reporting Data System (FRDS) provides information on public water supplies in violation
of current MCLs as determined through compliance monitoring of all supplies performed by the States
under the requirements of the National Interim Primary Drinking Water Regulations. Only violations of
current MCLs (i.e., 0.05 mg/L for chromium) and approved variances and exemptions from the standards
are recorded in FRDS. Monitoring is required annually for surface water supplies and every three years
for ground water supplies.

The FRDS database (FRDS, 1990, as cited in Wade Miller, 1990) was searched for ground water
violations reported during the three year period from 1987 through 1989. The data indicate that there are
5 public water supplies providing drinking water having chromium levels above the current MCL of 0.05
mg/L.  As summarized in Table 2.2-14, all of those supplies have a ground water source and are in the
very small size categories. Four of the supplies in violation had chromium levels  in the range of 0.05 to
0.075 mg/L and the fifth supply reported a value of 0.15 mg/L. No violations were reported for supplies
having a surface water source. Also, no variances or exemptions from the current chromium  MCL of
0.05 mg/L have been granted.
Table 2.2-14:  National Summary of Chromium Violations in Ground Water Systems
System size TT .„ ,
, , ,. Unspecified
(population _ '' ..
.. Concentration
served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000

0
0

0
0

0

0
0
0
Number of Supplies in Violation
>0.05-
0.075

1
3

0
0

0

0
0
0
> 0.075 -
0.1

0
0

0
0

0

0
0
0
>0.1-
0.125

0
0

0
0

0

0
0
0
by Concentration Range (mg/L):
> 0.125 -
0.15

0
1

0
0

0

0
0
0
>0.15

0
0

0
0

0

0
0
0
Totals

1
4

0
0

0

0
0
0
Occurrence Summary and Use Support Document
47
March 2002

-------
Number of Supplies
System size
(population
served)
75,001 - 100,000
>100,000
Totals
Unspecified
Concentration
0
0
0

>0.05-
0.075
0
0
4

> 0.075 -
0.1
0
0
0
in Violation

0.125
0
0
0
by Concentration Range (mg/L):

> 0.125 -
0.15
0
0
1

>0.15
0
0
0

Totals
0
0
5
Source: FRDS 1990
2.2.5.3 Results of Survey Data

2.2.5.3.1  Estimated National Occurrence in Public Water Supplies

The preceding sections presented the results of several studies that provide information on the occurrence
of chromium in public drinking water supplies, as reported by Wade Miller (1990). This section presents
national estimates of chromium occurrence in public drinking water supplies based on those study
results. A document entitled Overview of methodology used to estimate national occurrence and
exposure to inorganic drinking water contaminants, describes the methodology used to estimate
chromium occurrence in drinking water (USEPA, 1986, as cited in Wade Miller, 1990).

In general, national survey data (presented in section 2.2.5.2) are used as the basis  for the national
estimates. The most representative data sets are selected and combined, stratified appropriately by water
source and system size sampled, and applied to a "delta-lognormal" distribution model.  Using that
model, the probability of contaminant occurrence above any given concentration is determined for each
source/size category of supplies and applied to the total number of water supplies in those groups. The
resulting  national occurrence estimates are presented in tables that show the number of public water
supplies within various water source and size categories expected to have contaminant levels falling
within certain concentration ranges.  Tables are also provided that show the cumulative number of
supplies,  by  source and size category, expected to exceed certain concentrations. The cumulative
estimates also include upper and lower bounds based on the 95 percent confidence intervals.

The chromium occurrence estimates are based on the results of the NIRS survey for ground water
supplies and the combined results of the 1969 CWSS, the 1978 CWSS, the RWS, and the NOMS for
surface water supplies.  The NIRS  data were the most current ground water data available at the time of
the study, and the analytical results are considered to be highly reliable because of the extensive quality
assurance program employed. None of the surveys that analyzed  surface water supplies could be shown
to be any more or less representative of the universe of water supplies than the other surveys, and it was
therefore determined to be most appropriate to combine the results from all of the surveys to form the
basis of the national occurrence estimates. (There were insufficient details for the  100 Largest Cities
Survey to permit inclusion in the national estimates.)

Table 2.2-15 shows the estimated cumulative number of public ground water supplies, by system size,
having chromium levels exceeding various concentrations. It is estimated that more than 93 percent of
the water supplies have chromium  levels below 0.002 mg/L.  Because 0.002 mg/L was the minimum
quantifiable concentration for chromium, it is  not known whether those supplies are free of chromium or
have chromium present at very low levels.
Occurrence Summary and Use Support Document          48                                      March 2002

-------
Table 2.2-15 indicates that two supplies, both serving small populations, are expected to have chromium
present at levels exceeding both the proposed MCL and MCLG of 0.1 mg/L. The FRDS compliance
data, given in Section 2.2.5.2.6, reported only one ground water supply as having chromium levels above
0.1 mg/L. The supply was listed as serving 200 people with a concentration of 0.15 mg/L chromium.
Table 2.2-15:  Estimated Cumulative National Occurrence of Chromium in Community Ground
Water Supplies Based on the Delta Lognormal Distribution Model
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Number of
Systems Sampled
17,079
15,354

5,038
5,185

2,308

823
278
77
17
43
46,202
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002
1,216
1,093

130
134

253

90
31
8
2
4
2,961
>0.05
2
2

5
5

0

0
0
0
0
0
14
>0.1
0
0

1
1

0

0
0
0
0
0
2
>0.2
0
0

0
0

0

0
0
0
0
0
0
* Number of systems in the United States includes those that purchase water.
Table 2.2-16 provides the national cumulative occurrence estimates for chromium from surface water
supplies. These estimates, based on the combined results of the 1969 CWSS, 1975 CWSS, NOMS and
RWS indicate that zero surface water supplies have chromium levels above the proposed MCL of 0.1
mg/L.  This estimate is consistent with the FRDS data which also shows no supplies exceeding the 0.1
mg/L MCL.

It should be noted that the values shown in Tables 2.2-15 and 2.2-16 are rounded to the nearest whole
number. Because they are based on probabilities of occurrence at various concentrations, some partial
probability of occurrence may exist even when the value in the table indicates "0" systems in a particular
size category at or above a given concentration.
Occurrence Summary and Use Support Document
49
March 2002

-------
Table 2.2-16:  Estimated Cumulative Occurrence of Chromium in Surface Water Supplies Based
on the Delta Lognormal Distribution Model
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Number of
Systems Sampled
833
779

754
1,040

1,156

569
328
157
108
233
5,957
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002
208
195

189
260

289

142
82
39
27
59
1,490
>0.05
2
2

2
2

3

1
1
0
0
0
13
>0.1
0
0

0
0

0

0
0
0
0
0
0
>0.2
0
0

0
0

0

0
0
0
0
0
0
* Number of systems in the United States includes those that purchase water.
2.2.5.3.2 Estimated National Exposure from Public Water Supplies

The previous section noted that a separate methodology document has been prepared to describe the
approach used to estimate national occurrence of inorganic contaminants in public water supplies.  That
document also addresses the approach to estimating the population exposed to contaminant levels in
public drinking water supplies.  In summary, the occurrence probability density functions obtained from
the national survey data on supplies contaminated at various levels are applied to the number of people
using water supplies in the various size categories. The results of the population exposed estimates are
presented in tables similar to those in section 2.2.5.3.1 presenting the national occurrence estimates.

Table 2.2-17 presents the population exposed to chromium from ground water supplies exceeding the
current MCL of 0.05 mg/L as reported in FRDS (1990, as cited in Wade Miller, 1990). The data show
that 1,060 people are exposed to chromium at levels exceeding the current MCL of 0.05 mg/L.  Of the
1,060 people, 806 are exposed to levels ranging from 0.05 to 0.075 mg/L and an additional 200 people
are exposed to chromium in ground water at levels above 0.1 mg/L.  Table  2.2-18 gives the estimated
cumulative population exposed to chromium from ground water systems exceeding various
concentrations.  This table indicates that 3,000 people are expected to be exposed to chromium at levels
above the proposed MCL of 0.1 mg/L.
Occurrence Summary and Use Support Document
50
March 2002

-------
National cumulative population exposure estimates based on the national survey data for chromium from
surface water supplies are shown in Table 2.2-19. An estimated 20,000 people are expected to be
exposed to chromium levels above the proposed MCL of 0.1 mg/L.

The apparent inconsistency between the estimate in Section 2.2.5.3.1 that there are no surface water
supplies exceeding the proposed MCL of 0.1 mg/L and the estimate here that 20,000 people using surface
water supplies are exposed to levels  above this is a result of the application of the probability of
occurrence and exposure derived from the delta-lognormal distribution and applied to the universe of
water supplies and populations using those supplies.  In the case of estimating supplies, the numbers are
rounded to the nearest whole number, whereas the population estimates are rounded to the nearest
thousand. Where very small probabilities of occurrence exist, the number of supplies may be rounded to
0, whereas the number of people exposed is rounded to some number greater than 0.

Using a drinking water consumption rate of 2 liters/day, the intake of chromium for an adult at the
proposed MCL is 0.200 mg/day. As indicated above  from the national estimates, 23,000 people are
estimated to have an intake of that amount or more from drinking water. Typically, however, exposure to
chromium in drinking water is at levels below 0.002 mg/L, which would contribute less than 0.004
mg/day to chromium intake.
Table 2.2-17:  Population Exposed to Chromium at Levels Exceeding the Current MCL in Ground
Water Supplies as Reported in FRDS
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Unspecified
Concentration
0
0

0
0

0

0
0
0
0
0
0
Number of Supplies in Violation by Concentration Range (mg/L):
> 0.05 -
0.075
40
820

0
0

0

0
0
0
0
0
860
> 0.075 -
0.1
0
0

0
0

0

0
0
0
0
0
0
>0.1-
0.125
0
0

0
0

0

0
0
0
0
0
0
> 0.125 -
0.15
0
200

0
0

0

0
0
0
0
0
200
>0.15
0
0

0
0

0

0
0
0
0
0
0
Totals
40
1,020

0
0

0

0
0
0
0
0
1,060
Source: FRDS 1990
Occurrence Summary and Use Support Document
51
March 2002

-------
Table 2.2-18: Estimated Cumulative Population (in thousands) Exposed to Chromium in Drinking
Water Exceeding the Indicated Concentrations from Community Ground Water Supplies
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Number of
Systems
Sampled

950
3,850

3,910
10,000

13,310

13,110
9,540
4,770
1,360
10,360
71,160
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002

68
274

101
258

1,461

1,439
1,047
524
149
1,137
6,458
>0.05

0
0

4
10

0

0
0
0
0
0
14
>0.1

0
0

1
2

0

0
0
0
0
0
3
>0.2

0
0

0
0

0

0
0
0
0
0
0
* Number of systems in the United States includes those that purchase water.
Table 2.2-19: Estimated Cumulative Population (in thousands) Exposed to Chromium in Drinking
Water Exceeding the Indicated Concentrations from Community Surface Water Supplies
System size
(population served)
Very Small
25 - 100
101-500
Small
501 - 1,000
1,001-3,300
Medium
3,301 - 10,000
Number of —
Systems Sampled

90
570

1,280
4,330
10,200
Cumulative Number of Supplies with Concentrations (mg/L) of:
> 0.002

23
143

320
1,083
2,550
>0.05 >0.1

0 0
1 0

3 0
10 1
24 1
>0.2

0
0

0
0
0
Occurrence Summary and Use Support Document
52
March 2002

-------
System size
(population served)
Large/Very Large
10,001 -25,000
25,001 - 50,000
50,001 - 75,000
75,001 - 100,000
>100,000
Totals
Number of —
Systems Sampled

12,640
15,910
10,310
10,090
89,240
154,660
Cumulative Number of Supplies
> 0.002

3,160
3,978
2,578
2,523
22,311
38,669
>0.05

30
37
24
24
210
363
with Concentrations (mg/L) of:
>0.1

2
2
1
1
12
20
>0.2

0
0
0
0
0
0
* Number of systems in the United States includes those that purchase water.
2.2.6  Conclusion

Chromium and many of its compounds are naturally occurring and found at low levels in soil, water, and
air. Furthermore, chromium compounds are produced in the United States from chromite ore and are in
widespread use.  Chromium has a wide range of uses in metals, chemicals, and refractories.  Chromium
use in iron, steel, and nonferrous alloys enhances hardenability and resistance to corrosion and oxidation.
Recent statistics regarding import for consumption indicate production and use are robust. Industrial
releases of chromium and chromium compounds have been reported to TRI since 1988 from 49 of the 50
States. Off-site releases constitute a considerable amount of total releases, with releases to land the most
significant on-site releases.  Chromium is also a national NAWQA analyte. Approximately 66% of all
surface water sites had analytical detections of chromium, compared to 0% of ground water sites. No
surface or ground water had any analytical detections of chromium greater than the MCL (0.1 mg/L).
The Stage 2 analysis, based on the 16-State cross-section, estimated that approximately 0.00424% of
combined ground water and surface water systems serving 0.00139% of the population had estimated
mean concentrations of chromium greater than the MCL of 0.1 mg/L.  Based on this estimate,
approximately 3 PWSs  nationally serving about 3,000 people are expected to have estimated mean
concentrations of chromium greater than 0.1 mg/L.

Chromium is a naturally occurring element. Therefore, the balanced geographic distribution of the 16-
State cross-section should adequately cover the range of natural occurrence of chromium from low to
high.  Fifteen of the 16 cross-section States, excluding Vermont, have  reported TRI releases. Based on
this use and release evaluation, the 16-State cross-section appears to adequately represent chromium
occurrence nationally.

2.2.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 2000.  Toxicological Profile for
       Chromium. U.S. Department of Health and Human Services, Public Health Service. 419 pp. +
       Appendices.  Available on the Internet at: http://www.atsdr.cdc.gov/toxprofiles/tp7.pdf

Agency for Toxic Substances and Disease Registry (ATSDR). 2001.  ToxFAQs for Chromium. U.S.
       Department of Health and Human Services, Public Health Service.  Available on the Internet at:
       http://atsdr.cdc .gov/tfacts7 .html
Occurrence Summary and Use Support Document          53                                      March 2002

-------
Battelle.  1984.  Battelle, Pacific Northwest Laboratories. Chemical attenuation rates, coefficients, and
       constants in leachate migration. Volume 1:  A critical review. Prepared for the Electric Power
       Research Institute. Research Project 2198-1.

Brass, H.  1983. U.S. Environmental Protection Agency, Office of Drinking Water, Technical Support
       Division, Cincinnati, OH. Personal communication to author of SAIC, 1987.

Brower, B. 1983. Computer data file of analytical results for public water supplies sampled in the Rural
       Water Survey. Cornell University, Department of Rural Sociology.

Durfor, C.N., and E. Becker.  1964.  Public water supplies of the 100 largest cities in the United States,
       1962. U.S. Geological Survey Water Supply Paper 1812. Washington, DC: U.S. Government
       Printing Office.

Federal Reporting Data System (FRDS). 1990. Computer file extracts of compliance monitoring data
       for the Safe Drinking Water Act.

Gilliom, R.J., O.K. Mueller, and L.H. Nowell.  1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95.  U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589/, last updated October 9, 1998.

Glick, E.  1984. U.S. Environmental Protection Agency, Office of Drinking Water, Technical Support
       Division, Cincinnati, OH. Personal communication to author of SAIC, 1987.

McCabe, L.J., J.M. Symons, R.D. Lee, and G.G. Robeck. 1970. Survey of community water supply
       systems. Journal of the American Water Works Association,  v.  62, no. 9, pp. 670-687.

Morel, P.M.   1983. Principles of aquatic chemistry.  John Wiley and Sons, Inc.

National Library of Medicine (NLM). 2001.  NLM Hazardous Substances Data Base Toxicology Data
       Network.  Available on the Internet at: http://toxnet.nlm.nih.gov/, last updated May 16, 2001.

Science Applications International Corporation (SAIC).  1987.  Estimated National Occurrence and
       Exposure to Chromium in Public Drinking Water Supplies (Revised Draft).

USEPA.  1979.  Water-related environmental fate of 129 priority pollutants. Volume  1.  Office of Water
       Planning and  Standards, USEPA.  EPA-440/4-79-029a.

USEPA.  1980.  Trace Elements in 113 U.S. Drinking Waters. A substudy of the National Organics
       Monitoring Survey. Cincinnati, OH: Technical Support Division, Office  of Drinking Water,
       USEPA. January, 1980.

USEPA.  1983.  Data on inorganics from the Community Water Supply Survey. Provided by E.M. Glick,
       Drinking Water Quality Assessment Branch, Technical Support Division, Office of Drinking
       Water, USEPA, Cincinnati, OH.
Occurrence Summary and Use Support Document          54                                     March 2002

-------
USEPA.  1985.  Computerized listing of inorganics data for the National Organics Monitoring Survey
       (NOMS).  Provided to Science Applications International Corporation by E.B. Dotson, Water
       Supply Technology Branch, Technical Support Division, Office of Drinking Water, USEPA,
       Cincinnati, OH. November 12, 1985.

USEPA.  1986.  Overview of methodology used to estimate national occurrence and exposure to
       inorganic drinking water contaminants (Draft).  Prepared by Science Applications International
       Corporation for Science and Technology Branch, Office of Drinking Water, USEPA.  October
       30, 1986.

USEPA.  2000.  TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 2001. Mineral Commodity Summaries, January, 2001 - Chromium. Available on the Internet at:
       http ://minerals .usgs .gov/minerals/pubs/commodity/chromium/1803 00 .pdf.

USPHS.  1970.  Community water supply study. Analysis of national survey findings. Bureau of Water
       Hygiene, Environmental Health Service, Department of Health, Education, and Welfare.

Wade Miller Associates, Inc. 1990. Occurrence and Exposure Assessment for Chromium in Public
       Drinking Water Supplies. Prepared for and submitted to EPA on July 24, 1990.
Occurrence Summary and Use Support Document         55                                     March 2002

-------
2.3    Fluoride
Table of Contents

2.3.1  Introduction, Use and Production  	  57
2.3.2  Environmental Release  	  59
2.3.3  Ambient Occurrence  	  60
2.3.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  61
2.3.5  Additional Drinking Water Occurrence Data  	  68
2.3.6  Conclusion	  78
2.3.7  References 	  79
Tables and Figures

Table 2.3-1: Imports of Fluorspar to the United States (thousand metric tons, gross weight)	  57

Table 2.3-2: Hydrogen Fluoride Manufacturers and Processors by State  	  58

Table 2.3-3: Environmental Releases (in pounds) for Fluorine in the United States, 1995-1999  	  59

Table 2.3-4: Environmental Releases (in pounds) for Hydrogen Fluoride in the
       United States,  1988-1999 	  60

Table 2.3-5: Fluoride Detections and Concentrations in Surface Water and Ground Water	  61

Table 2.3-6: Stage 1 Fluoride Occurrence Based on 16-State Cross-Section - Systems  	  62

Table 2.3-7: Stage 1 Fluoride Occurrence Based on 16-State Cross-Section - Population 	  63

Table 2.3-8: Stage 2 Estimated Fluoride Occurrence Based on 16-State Cross-Section - Systems ....  65

Table 2.3-9: Stage 2 Estimated Fluoride Occurrence Based on 16-State Cross-Section - Population . .  66

Table 2.3-10: Estimated National Fluoride Occurrence - Systems and Population Served	  67
Occurrence Summary and Use Support Document         56                                     March 2002

-------
2.3.1  Introduction, Use and Production

Fluorine (F), the element, is a pale, yellow-green, irritating gas that has a strong, sharp odor. Because it
is so chemically reactive, it is almost always found naturally combined with metals as a salt. The term
"fluorides" is used to refer to the common salts of the element fluorine, with the most common salts
being sodium fluoride and calcium fluoride. Because toxic effects are due to the fluoride ion, the term
fluoride is most commonly used in discussing health effects and regulations.

Fluorides occur naturally in a variety of geologic settings (for example, in sedimentary and volcanic
rocks) in coal, clay, and certain minerals, and also are found in sea water. Ground water levels of
fluoride are typically higher than surface water levels because their content is influenced more by the
mineral and rock through which they flow (WHO, 1984, as cited in ATSDR, 1993). The biggest natural
source of fluorides released to the air is volcanic eruptions.  Although the greatest total volume of
fluorides in the environment is released from natural sources (e.g., volcanoes and oceans), the highest
concentrations are found near anthropogenic point sources related to steel, aluminum, and glass
production.

Fluoride compounds are used in a wide range of industrial and other applications.  The steel industry is
the largest consumer of fluorides, followed by the chemical industry and the glass and ceramics industry
(ATSDR, 1993).

In the United States, more than 80% of fluorspar, the primary source of fluorine and its compounds, is
imported (ATSDR, 1993). Fluorspar (a term sometimes used to refer to calcium fluoride)  is used directly
or indirectly to manufacture products such as aluminum, gasoline, insulating foams, refrigerants, steel,
and uranium fuel.  Byproduct fluorosilicic acid production from some phosphoric acid producers
supplements fluorspar as a domestic source of fluorine, but is not included in fluorspar production or
consumption calculations. The imports originate primarily from China, the Republic of South Africa,
and Mexico (USGS, 2000).
Table 2.3-1:  Imports of Fluorspar to the United States (thousand metric tons, gross weight)
Imports For Consumption
Acid grade
Metallurgical grade
Fluorspar equivalent from
hydrofluoric acid plus cryolite
1994
434
59
108
1995
470
88
114
1996
474
39
131
1997
485
51
175
1998
462
41
204
1999
419
59
192
Source: USGS, 2000
Usage of fluoride compounds are varied, implying potentially widespread environmental release.
Molecular fluorine has been used as an oxidizer in rocket fuels, in the production of metallics and other
fluorides, and in glass, enamel, and brick production.  Currently, molecular fluorine is used by most
manufacturers for the production of various inorganic fluorides.  The main use of another fluoride
compound, sodium fluoride, is as a drinking water additive for prevention of dental caries (tooth decay).
It is also used as a disinfectant for fermentation apparatus in breweries and distilleries, in wood
Occurrence Summary and Use Support Document
57
March 2002

-------
preservation, and in rimmed steel manufacturing. Calcium fluoride, the primary compound used by the
fluoride chemical industry, has been used as a flux in steel manufacturing. Other uses are frosting glass
and enamels, coating welding rods, fluoridation of drinking water, paint pigment, and as a catalyst in
wood preservation. Prior to 1930, hydrogen fluoride was used mainly for glass etching and polishing,
foundry scale removal, and minor production of metal fluorides. Other uses include uranium processing,
petroleum alkylation, and stainless steel pickling. A sharp decrease in demand for hydrogen fluoride
occurred in 1978 when the United States prohibited the use  of chlorofluorocarbon gases in pressure
packaging (ATSDR, 1993).

Table 2.3-2 shows the number of facilities in each State that manufacture and process hydrogen fluoride,
the intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1993). Because only certain types of facilities are required to report,
this is not an exhaustive list. Neither sodium fluoride nor any other fluoride salts are listed on TRI.
Table 2.3-2:  Hydrogen Fluoride Manufacturers and Processors by State
State"
AL
AR
AZ
CA
CO
CT
DE
FL
GA
HI
IA
ID
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
ND
NJ
NM
NV
NY
OH
OK
OR
PA
Number of facilities
5
1
8(l)d
60 (6) d
6(l)d
9
1
11
13(l)d
2
3
2(l)d
13
14 (2) d
8
8(l)d
11
15
6(l)d
1
12
1
12(l)d
1
5
13
1
17 (2) d
4(l)d
1
21 (2) d
38 (6) d
8(l)d
8
48 (2) d
Range of maximum amounts on
site in thousands of pounds'"
1-999
100-999
0.1-999
0-49,999
1-99
0.1-99
10-99
0-999
0.1-999
1-9
0-99
0.1-0.9
0-999
0-999
1-9,999
0-9,999
10-9,999
0.1-99
1-999
10-99
0.1-9,999
100-999
0-99
1-9
0-999
0-99
1-9
0-999
0.1-99
1-99
0-9,999
0-999
10-999
0.1-999
0.1-9.999
Activities and uses0
3,7,12
7
7,11,12,13
2,3,4,7,8,9,10,11,12,
11,12,13
1,5,8,11,12,13
7,8
1,3,5,6,7,12,13
8,9,11,12,13
12
3,7,9,11,13
1,5,12,13
2,3,4,7,8,9,11,12
5,7,8,11,12,13
7,10,11
1,2,3,4,5,7,10,11,12
1,2,3,4,5,7,11,12,13
8,11,12,13
5,11,12
12
2,7,8,11,12,13
7,13
5,7,8,11,12,13
13
1,3,5,11
1,5,7,11,12,13
11




13























1,2,3,4,5,7,8,10,11,12,13
11,12,13
2,7


1,2,5,6,7,10,11,12,13
1,2,3,5,7,8,11,12,13
1,2,3,7,11,12
1,5,11,12,13
2,5.7.8.10.11.12.13




Occurrence Summary and Use Support Document
58
March 2002

-------
State"
PR
RI
SC
TN
TX
UT
VA
VT
WA
WI
wv
WY
Number of facilities
7(l)d
3(l)d
6
8(l)d
42
7
3
1
16(l)d
11
3
4md
Range of maximum amounts on
site in thousands of pounds'1
1-99
0.1-99
0-999
0-9,999
0-49,999
0-999
0-99
100-999
0-999
0.1-99
0-9
10-99
Activities and uses0
5,11,12,13
11,12
5,7,11,12,13
1,3,4,5,6,7,11,12
1,2,3,4,5,7,11,12,
2,3,5,8,11,12,13
5,11,12
12
1,5,7,11,12,13
9,10,11,12,13
2,5,12
5.11




13







Tost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
                         8. As a formulation component
                         9. As an article component
                         10. For repackaging only
                         11. As a chemical processing aid
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
                        12. As a manufacturing aid
                        13. Ancillary or other uses
dNumber of facilities reporting "no data" regarding maximum amount of the substance on site
Source: AT SDR, 1993 compilation of 1989 TRI data


2.3.2 Environmental Release

Fluorine and hydrogen fluoride are both listed as Toxics Release Inventory (TRI) chemicals. Table 2.3-3
illustrates the environmental releases for fluorine from 1995 - 1999.  (There are only fluorine data for
these years.) Air emissions constitute most of the on-site releases, with a steady increase over the years.
The increase in air emissions, as well as in surface water discharges, have solely contributed to increases
in fluorine total on- and off-site releases in recent years. No underground injection, releases to land
(such as spills or leaks within the boundaries of the reporting facility), or off-site releases (including
metals or metal compounds transferred  off-site) were reported for fluorine. These TRI release data for
fluorine were reported from 14 States and Puerto Rico (USEPA, 2000).  Of the  14 States that reported
TRI data, 4 are contained within the  16-State cross-section (used for analyses of fluoride occurrence in
drinking water; see Section 2.3.4). (For a map of the 16-State cross-section,  see Figure 1.3-1.)


Table 2.3-3: Environmental Releases (in pounds) for Fluorine in the United States, 1995-1999
Year
1999
1998
1997
1996
1995
On-Site Releases
Air Emissions
86,302
81,938
30,091
25,460
18,319
Surface Water
Discharges
54,153
49,857
54,200
48,300
15,000
Underground
Injection
-
—
—
—
-
Releases
to Land
-
—
—
—
-
Off-Site Releases
—
—
—
—
—
Total On- &
Off-site
Releases
140,455
131,795
84,291
73,760
33,319
 Source: USEPA, 2000
Occurrence Summary and Use Support Document
                                                    59
March 2002

-------
Table 2.3-4 illustrates the environmental releases for hydrogen fluoride between 1988 and 1999. Air
emissions constitute most of the on-site releases, with a moderate fluctuation over the years. No real
trend is suggested in the data.  Surface water discharges have also fluctuated since 1988, but have stayed
considerably lower than the high levels observed in  1988. Underground injection and releases to land are
less significant on-site releases, with underground injections sharply decreasing to zero in 1998 after its
5-year long peak between 1993 and 1997.  Off-site releases of hydrogen fluoride are considerable. From
1988 to 1998 there has been a general decrease in off-site releases, but in 1999 there was a large increase
in pounds released. The TRI data for hydrogen fluoride data were reported from 49 States and do not
include Alaska or Puerto Rico (USEPA, 2000).  All  16 of the cross-section States (used for analyses of
fluoride occurrence in drinking water; see  Section 2.3.4) reported releases of hydrogen fluoride.  (For a
map of the 16-State cross-section, see Figure 1.3-1.)
Table 2.3-4:  Environmental Releases (in pounds) for Hydrogen Fluoride in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
14,434,432
15,612,197
13,235,977
13,696,720
11,616,575
8,901,538
9,388,840
11,577,121
11,345,059
11,577,926
13,066,004
14,732,294
Surface Water
Discharges
16,983
23,194
31,680
10,691
8,702
14,989
10,340
3,400
5,469
13,868
35,918
189,928
Underground
Injection
0
0
2,879
2,620
3,845
2,174
3,520
1
1
25
0
250
Releases
to Land
5,353
12,740
15,047
36,604
24,078
33,443
33,260
27,886
25,259
8,329
10,943
13,002
Off-Site Releases
522,159
72,358
110,848
167,240
1,020,034
736,747
856,706
1,269,429
1,080,205
1,658,769
1,398,278
3,467,471
Total On- &
Off-site
Releases
14,978,927
15,720,489
13,396,431
13,913,875
12,673,234
9,688,891
10,292,666
12,877,837
12,455,993
13,258,917
14,511,143
18,402,945
 Source: USEPA, 2000
2.3.3  Ambient Occurrence

Fluoride is an analyte for both surface and ground water NAWQA studies, with a method detection limit
(MDL) of 0.1 mg/L. Additional information on analytical methods used in the NAWQA study units,
including method detection limits, is described by Gilliom and others (1998).

Typical of many inorganic contaminants, fluoride occurrence in ambient surface and ground waters is
common (Table 2.3-5). This is not surprising, considering that the element and its compounds are used
in many products.  Significantly, fluoride compounds are used in drinking water treatment.

Detection  frequencies are consistently greater for surface water than for ground water, possibly because
surface waters are more sensitive to anthropogenic releases. Median concentrations are also generally
higher for surface water (median concentration is 0.2 mg/L in surface water and 0.1 mg/L in ground
water).  However, fluoride detection frequencies greater than the MCL (4 mg/L) are higher in ground
water, and 99th percentile ground water concentrations are two times larger than corresponding 99th
percentile  surface water concentrations. Locally high concentrations in ground water, higher than any
Occurrence Summary and Use Support Document
60
March 2002

-------
seen in surface water, are not surprising given the possibility of long contact times between ground water
and rocks enriched in fluoride at a given location.  Contact times between surface waters and naturally
occurring fluoride are orders of magnitude shorter, hence concentrations are lower.  Furthermore, surface
waters subject to large anthropogenic inputs of fluoride are more easily diluted by waters integrated from
other parts of the watershed, where fluoride concentrations may be lower.
Table 2.3-5:  Fluoride Detections and Concentrations in Surface Water and Ground Water

                         Detection frequency       Detection frequency      Concentration percentiles
                             > MDL*                > MCL*             (all samples; mg/L)

                        % samples     % sites     % samples     % sites      median        99th

    surface water           65.7%      69.0%       0.02%      0.05%        0.2           1.3


    ground water           60.3%      67.0%       0.4%        0.5%        0.1           2.6

* The Method Detection Limit (MDL) for fluoride in water is 0.1 mg/L and the Maximum Contaminant Level (MCL) is 4.0 mg/L.


2.3.3.1 Additional Ambient Occurrence Data

A summary document entitled "Occurrence of Fluoride in Drinking Water, Food, and Air" (JRB
Associates, 1984), was previously prepared for past USEPA assessments of fluoride. However, no
information on the ambient occurrence of fluoride was included in that document.  (The document did
include information regarding fluoride occurrence in drinking water, which is discussed in Section 2.3.5
of this report.)

2.3.4 Drinking Water Occurrence Based on  the 16-State Cross-Section

Fluoride is unique as a drinking water constituent in that there are both a MCL to limit levels in drinking
water to protect the public from the adverse effects of fluoride, and a recommended Optimum Level for
protection against dental caries. A number of drinking water systems achieve the Optimum Level by the
addition of fluoride during the drinking water treatment process. The recommended optimum level
ranges from 0.7 mg/L for warmer climates to 1.2 mg/L for cooler climates. The current MCL of 4 mg/L
was promulgated by EPA in 1986.  At the same time, a National Secondary Drinking Water Regulation
(NSDWR) was promulgated, establishing a Secondary Maximum Contaminant Level (SMCL) of 2.0
mg/L. (The SMCL was based upon the finding that a 2.0 mg/L concentration of fluoride in drinking
water would be protective against dental decay while preventing the majority of cases of water-related
dental fluorosis which is cosmetically objectionable.)

The analysis of fluoride occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. Fluoride occurrence in
drinking water was assessed relative to several  concentration thresholds of interest: 4 mg/L; 3 mg/L; 2
mg/L; 1.5 mg/L;  1.2 mg/L; 0.7 mg/L; 0.5 mg/L; and 0.1  mg/L.
Occurrence Summary and Use Support Document          61                                      March 2002

-------
All sixteen cross-section State data sets contained occurrence data for fluoride. These data represent
more than 93,000 analytical results from approximately 21,000 PWSs during the period from 1983 to
1998 (with most analytical results from 1992 to 1997).  The number of sample results and PWSs vary by
State, although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness. The overall modal detection limit for fluoride in the 16 cross-section States is equal to 0.1
mg/L.  (For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this report.)

2.3.4.1 Stage 1 Analysis Occurrence Findings

The percentage of systems indicates the proportion of systems (or population served by systems) with
any analytical results exceeding the specified threshold of concern.  Table 7 illustrates the Stage 1
analysis of fluoride occurrence in drinking water for the public water systems in the 16-State cross-
section. The percentage of total ground and surface water PWSs with at least one analytical result
exceeding the MCL (4 mg/L) was equal to 1.28% (266 systems).  Approximately 4.45% of total ground
and surface water systems (925 systems) had any analytical results greater than the 1A MCL (2 mg/L).
Over 74.1 % of PWSs (15,414 systems) had at least one analytical result greater than 0.1 mg/L (the modal
detection limit for fluoride).
Table 2.3-6:  Stage 1 Fluoride Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
4
3
2
1.5
1.2
0.7
0.5
0.1
Percent of Systems
Exceeding Threshold
1.26%
2.26%
4.51%
7.41%
10.7%
23.3%
31.1%
73.7%
Number of Systems
Exceeding Threshold
243
434
867
1,424
2,062
4,475
5,980
14,164

Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
1.44%
2.00%
3.64%
7.03%
15.6%
44.8%
50.4%
78.5%
23
32
58
112
249
713
803
1,250

Combined Ground &
Surface Water
4
3
1.28%
2.24%
266
466
Occurrence Summary and Use Support Document
62
March 2002

-------
Source Water Type

Threshold
(mg/L)
2
1.5
1.2
0.7
0.5
0.1
Percent of Systems
Exceeding Threshold
4.45%
7.38%
11.1%
24.9%
32.6%
74.1%
Number of Systems
Exceeding Threshold
925
1,536
2,311
5,188
6,783
15,414
Reviewing fluoride occurrence in the 16 cross-section States by PWS population served (Table 2.3-7)
shows over 4.56% of the total population (over 4.8 million people) was served by PWSs with at least one
analytical result greater than the MCL. Approximately 6.76% of the population served by all systems
(about 7.2 million people) had any analytical results greater than the 1A MCL. More than 93.5% of the
population (over 100 million people) was served by PWSs with analytical detections of fluoride greater
than 0.1 mg/L.
Table 2.3-7:  Stage 1 Fluoride Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
4
3
2
1.5
1.2
0.7
0.5
0.1
Percent of Population
Served by Systems
Exceeding Threshold
3.84%
5.29%
8.06%
12.2%
21.3%
52.5%
62.4%
93.0%
Total Population
Served by Systems
Exceeding Threshold
1,685,400
2,323,400
3,539,100
5,362,200
9,363,800
23,069,200
27,432,700
40,839,200

Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
5.06%
5.09%
5.86%
9.34%
15.7%
62.4%
66.6%
93.9%
3,195,800
3,213,800
3,700,100
5,894,600
9,921,200
39,421,200
42,022,900
59,262,900
Occurrence Summary and Use Support Document
63
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population
Served by Systems
Exceeding Threshold
Total Population
Served by Systems
Exceeding Threshold

Combined Ground &
Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
4.56%
5.17%
6.76%
10.5%
18%
58.4%
64.9%
93.5%
4,881,200
5,537,200
7,239,200
11,256,800
19,285,000
62,490,400
69,455,600
100,102,100
2.3.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 2.3-8 and 2.3-9.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Approximately  106 (ground water and surface water) PWSs in the 16-State cross-section (0.511%) are
estimated to have mean concentrations of fluoride above 4 mg/L (the current MCL). Approximately 603
(2.90% of) PWSs in the 16 States are estimated to have mean concentrations greater than 2 mg/L.  The
percentage of PWSs in the  16 States with estimated mean concentrations exceeding 0.1 mg/L (the  modal
detection limit) was about 82.2% (17,098 PWSs).

A significantly greater proportion of ground water systems, as compared to surface water systems, was
estimated to exceed each threshold.  Approximately 106 ground water systems in the 16-State cross-
section (0.550%) had estimated mean concentrations of fluoride above 4 mg/L, compared to
approximately 1 surface water system (0.0491%). About 3.05% of ground water systems (an estimated
586 systems in the 16 States) had estimated mean concentrations greater than 2 mg/L. This compares
with about 1.11% of the surface water systems (about  18 systems) with estimated mean concentrations
greater than 2 mg/L. The estimated mean concentration values for approximately 15,846 ground water
PWSs in the 16  States (82.5%) exceed 0.1 mg/L. Approximately 1,251 surface water systems in the 16
States (78.6%) had estimated mean concentrations exceeding the modal detection limit.
Occurrence Summary and Use Support Document
64
March 2002

-------
Table 2.3-8:  Stage 2 Estimated Fluoride Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
4
3
2
1.5
1.2
0.7
0.5
0.1
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.550%
1.20%
3.05%
5.55%
8.54%
19.8%
29.3%
82.5%
Range
0.469% - 0.635%
1.07% -1.33%
2. 85% -3.23%
5. 29% -5. 78%
8.24% - 8.80%
19.4% -20.2%
28.8% - 29.7%
82.1% -82.9%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
106
230
586
1,066
1,640
3,804
5,619
15,846
Range
90 - 122
205-255
547-621
1,017-1,110
1,583-1,691
3,729 - 3,875
5,540 - 5,702
15,769-15,925

Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
0.0491%
0.193%
1.11%
3.91%
9.37%
30.2%
39.6%
78.6%
0.000% -0.1 88%
0.000% - 0.377%
0.691% -1.51%
3. 08% -4.77%
8.22% -10.5%
29.0% -31. 3%
38.6% - 40.6%
77.4% -79. 9%
1
3
18
62
149
481
631
1,251
0-3
0-6
11-24
49-76
131-167
462 - 499
615-646
1,233-1,272

Combined Ground
& Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
0.511%
1.12%
2.90%
5.42%
8.60%
20.6%
30.0%
82.2%
0.437% -0.591%
1.01% -1.23%
2. 72% -3.08%
5. 19% -5. 63%
8. 32% -8. 89%
20.2% -21.0%
29.7% -30.5%
81. 8% -82.6%
106
233
603
1,128
1,789
4,283
6,249
17,098
91 - 123
209 - 256
566 - 640
1,079-1,172
1,730-1,849
4,208 - 4,358
6,170-6,337
17,017-17,177
Reviewing fluoride occurrence by PWS population served (Table 2.3-9) shows that approximately
96,000 (0.0897% of) the PWS population in the 16-State cross-section were served by systems with mean
fluoride concentrations above 4 mg/L. When evaluated relative to a threshold of 2 mg/L, the percent of
population exposed increased significantly to about 0.929% (about 994,600 people served in the 16
States). The percentage of population served by PWSs with estimated mean concentrations greater than
0.1 mg/L was approximately 89.4% (almost 96 million people).
Occurrence Summary and Use Support Document
65
March 2002

-------
For ground water systems, about 91,300 people in the 16 States (0.208% of the population served by
ground water systems) were served by systems with estimated mean concentrations of fluoride above 4
mg/L. An estimated 1.68% of the 16-State population (approximately 736,800 people) were served by
ground water systems whose mean concentration value exceeded 2 mg/L. The percentage of population
served by ground water PWSs with estimated mean concentration values exceeding 0.1 mg/L was
approximately 90.4% (almost 40 million people).

Approximately 0.00754% (about 4,800 people) of the population served by surface water PWSs in the
16-State cross-section were served by systems with an estimated mean concentration of fluoride above 4
mg/L. About 0.408% (about 257,800 people) of the population served by surface water PWSs were
served by systems with mean concentrations greater than 2 mg/L.  Over 56 million people in the 16
States (88.7%) were served by PWSs with mean fluoride concentrations above 0.1 mg/L.  The percentage
of population served by surface  water systems that exceeded each threshold was generally always less
than the percentage of ground water systems that exceeded the threshold (except greater than 0.7 mg/L).
However, the total population served by systems estimated to exceed each threshold was much greater
for surface water systems for thresholds less than 1.5 mg/L.
Table 2.3-9:  Stage 2 Estimated Fluoride Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
4
3
2
1.5
1.2
0.7
0.5
0.1
Percent of Population Served by Systems
in the 16 States Estimated to Exceed
Threshold
Best Estimate
0.208%
0.556%
1.68%
3.52%
6.51%
24.6%
40.1%
90.4%
Range
0.131% -0.314%
0.393% -0.771%
1.35% -2.04%
2.87% -4.18%
5. 59% -7.57%
20.8% - 27.4%
38.3% - 42.0%
89.6% -91. 3%
Total Population served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
91,300
244,100
736,800
1,545,600
2,858,000
10,825,500
17,613,500
39,730,300
Range
57,400 - 137,900
172,400 - 338,800
594,400 - 896,300
1,261,800-1,836,500
2,455,500 - 3,325,000
9,151,600-12,042,500
16,844,600-18,461,400
39,361,200-40,108,100

Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
0.00754%
0.0400%
0.408%
1.92%
5.59%
28.3%
37.8%
88.7%
0.000% - 0.0687%
0.000% -0.175%
0.1 30% -1.44%
0.961% -3.41%
3. 55% -8.41%
25.6% -30.9%
35.4% -41.1%
85. 5% -92. 5%
4,800
25,300
257,800
1,212,900
3,527,700
17,837,300
23,892,500
56,012,300
0 - 43,300
0-110,300
81,800 - 908,600
606,900-2,153,100
2,244,000 - 5,309,500
16,189,300-19,491,600
22,364,500 - 25,938,300
53,953,900 - 58,392,700
Occurrence Summary and Use Support Document
66
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
in the 16 States Estimated to Exceed
Threshold
Best Estimate
Range
Total Population served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground &
Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
0.0897%
0.252%
0.929%
2.58%
5.96%
26.8%
38.8%
89.4%
0.0555% -0.141%
0.169% -0.36%
0.707% -1.55%
1.96% -3. 46%
4.61% -7.64%
24.3% - 28.8%
37.1% -40.9%
87.4% -9 1.6%
96,000
269,400
994,600
2,758,300
6,384,900
28,664,200
41,502,600
95,736,400
59,400-151,300
181,000-387,300
756,600-1,655,400
2,093,300 - 3,707,000
4,937,300 - 8,179,500
26,040,800 - 30,837,800
39,671,600-43,794,000
93,573,500 - 98,102,800
2.3.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold (Table 2.3-10), an estimated 332 PWSs serving approximately 191,000 people nationally could
be exposed to fluoride concentrations above 4 mg/L. About 1,885 systems serving almost 2 million
people had estimated mean concentrations greater than 2 mg/L. Approximately 53,448 systems serving
over 190 million people nationally were estimated to have fluoride concentrations greater than 0.1 mg/L.
(See Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 327 PWSs serving about 178,000 people nationally had mean
concentrations greater than 4 mg/L. Approximately 1,812 ground water systems serving about 1.4
million people nationally had estimated mean concentration values that exceeded 2 mg/L.  About 49,032
ground water systems serving over 77 million people had estimated mean concentrations greater than 0.1
mg/L.

Three surface water systems serving 9,600 people were estimated to have mean concentrations of
fluoride above 4 mg/L.  An estimated 62 surface water systems serving 519,900 people had estimated
mean concentrations greater than 2 mg/L. An estimated 4,392 surface water systems serving almost 113
million people had mean concentrations greater than 0.1 mg/L.
Table 2.3-10:  Estimated National Fluoride Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
4
3
2
1.5
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
327
711
1,812
3,297
Range
278 - 378
634 - 789
1,692-1,922
3,147-3,434
Total Population Served by Systems Nationallj
Estimated to Exceed Threshold
Best Estimate
178,000
476,000
1,436,900
3,014,300
Range
112,000-268,900
336,300 - 660,800
1,159,300-1,747,900
2,460,800-3,581,500
Occurrence Summary and Use Support Document
67
March 2002

-------
Source Water Type

Threshold
(mg/L)
1.2
0.7
0.5
0.1
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
5,076
11,769
17,386
49,032
Range
4,898 - 5,233
11,537-11,989
17,142 - 17,642
48,794 - 49,276
Total Population Served by Systems Nationallj
Estimated to Exceed Threshold
Best Estimate
5,573,600
21,112,000
34,349,800
77,482,000
Range
4,788,700 - 6,484,400
17,847,500-23,485,400
32,850,400 - 36,003,400
76,762,200-78,218,800

Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
3
11
62
219
524
1,688
2,213
4,392
0-11
0-21
39-84
172-267
460 - 586
1,621 - 1,751
2,158-2,267
4,327 - 4,464
9,600
51,000
519,900
2,445,900
7,113,700
35,969,700
48,180,300
112,951,300
0 - 87,400
0 - 222,400
164,900-1,832,200
1,223,900-4,341,800
4,525,200 - 10,706,900
32,646,500 - 39,305,700
45,099,000 - 52,305,700
108,800,500-117,751,500

Combined Ground &
Surface Water
4
3
2
1.5
1.2
0.7
0.5
0.1
332
728
1,885
3,526
5,594
13,390
19,535
53,448
284 - 385
654 - 801
1,769-2,000
3,373 - 3,664
5,408 - 5,780
13,156-13,624
19,288-19,808
53,195-53,695
191,000
535,900
1,978,600
5,487,100
12,701,700
57,022,300
82,562,000
190,450,600
118,200-301,000
360,000 - 770,500
1,505,100-3,293,100
4,164,300 - 7,374,300
9,821,800-16,271,700
51,803,600-61,346,400
78,919,500-87,120,300
186,147,900-195,158,100
2.3.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of fluoride in drinking water are also reviewed.
A preliminary comparison of the Stage 2 model findings to MCL violation records in SDWIS/FED was
conducted. The Stage 2 occurrence estimates for fluoride were also compared to fluoride occurrence
findings reported in the CDC Fluoridation Census 1992. Previously compiled occurrence information,
from an OGWDW summary document entitled "Occurrence of Fluoride in Drinking Water, Food, and
Air" (JRB Associates, 1984), is presented in Sections 2.3.5.3 - 2.3.5.8.  This variety of studies and
information are presented regarding levels of fluoride in drinking water, with the scope of the reviewed
studies ranging from national to regional. (All the information in Sections 2.3.5.3 - 2.3.5.8 is taken
directly from "Occurrence of Fluoride in Drinking Water, Food, and Air" (JRB Associates, 1984).) Note
that none of the studies presented in the following section provide the quantitative analytical results or
comprehensive coverage that would enable direct comparison to the occurrence findings estimated with
the cross-section occurrence data presented in Section 2.3.4. These additional studies, however, do
enable a broader assessment of the Stage 2 occurrence estimates presented for this  Six-Year Review.
Occurrence Summary and Use Support Document
March 2002

-------
2.3.5.1 SDWIS/FED Comparison

A preliminary comparison of Stage 2 model findings to MCL violation records in the Safe Drinking
Water Information System (SDWIS) was conducted. Due to many qualifying factors, this must be
regarded as a very general, indirect comparison.  A primary reason inhibiting a direct comparison is the
somewhat incomplete State reporting to the SDWIS database over the time frame of interest (roughly
1993-1999). Also, the method for calculating a contaminant's concentration in a system is somewhat
different for the Stage 2 analysis as compared to  MCL violation determinations.  A brief description of
some key topics related to MCL compliance information is presented below to provide background on
SDWIS MCL violation data.

For systems that monitor more frequently than annually, compliance with the MCL is determined by a
running annual average of results from all samples taken at each sampling point. If this contaminant
mean concentration exceeds the MCL, then the system is out of compliance.  For systems that monitor
annually or less frequently, if the level of a contaminant at any sampling point exceeds the MCL, the
system is out of compliance with the MCL.

More systems have MCL exceedances than actual MCL violations. A system with an MCL violation
always has an MCL exceedance. However, a system with an MCL exceedance may not always incur an
MCL violation. For example, if a system on quarterly monitoring has one quarter in which the
concentration is above the MCL, but the running annual average of this high sample result and the three
preceding quarters are below the MCL, the system would have an MCL exceedance but not an MCL
violation.  Also, if the State requires a confirmation sample, compliance with the MCL is calculated
using the average of the routine and confirmation sample. If the average is below the MCL, the system
would have an exceedance but not an MCL violation.

Many States experienced delays in implementation of the Phase II/V rule. In some cases, approval of
State primacy took many years. Laboratory capacity, resource and staffing levels, and waivers were
recurring issues.  States are required to report MCL violations to SDWIS. However, delays in
determining MCL violations and reporting them to SDWIS were commonplace.  States had to create new
databases or modify existing databases for Phase II/V compliance tracking and reporting. As a result,
many States had delays in reporting chemical violation data to SDWIS. Underreporting of violations for
chemicals may still be an issue for some States.  SDWIS has the capability of storing data on MCL
exceedances; however, reporting of these is optional.

When comparing the modeled national occurrence estimates to the SDWIS MCL violation data, one must
also consider the different time frames at hand. The SDWIS MCL violation data are roughly from 1993-
1999. The Stage 2 estimate is based on compliance monitoring analytical results predominantly between
1992 and 1997. (Potentially more than five years of monitoring results are used to estimate a single
system mean concentration). The Stage 2 estimated number of systems with a mean concentration
greater than the MCL (based on data for many years) is an approximation of, though not directly
comparable to, the number of SDWIS MCL violations for a single year.

Approximately 247 PWSs reported MCL violations to SDWIS/FED between January 1,  1993 and
December 31, 1999. The Stage 2 analysis estimated that approximately 332 PWSs nationally had mean
concentrations of fluoride greater than the MCL (See Table 2.3-10). (The Stage 2 estimate is based on
compliance monitoring analytical results predominantly between 1992 and 1997.  Potentially more than
five years of monitoring results are used to estimate  a single system mean concentration). Based on the
qualifying factors described above, it is  not surprising that the Stage 2 estimates are higher than the
reported SDWIS MCL violations. Even given the inherent differences between the SDWIS MCL

Occurrence Summary and Use Support Document          69                                    March 2002

-------
violation records and the Stage 2 analytical findings measured relative to the MCL, the comparison
between the two assessments of occurrence are quite comparable, providing additional measures that
suggest a validation of the Stage 2 modeling approach.

2.3.5.2  Centers for Disease Control and Prevention Fluoridation Census

This section provides a general comparative assessment between the Six-Year Review's Stage 2 national
fluoride occurrence estimates (based on the 16-State national cross-section) primarily with public water
system fluoridation findings reported in the Centers for Disease Control and Prevention (CDC)
Fluoridation Census 1992 (CDC, 1992) as well as with findings provided by CDC staff from the
unpublished Fluoridation Census 2000 (CDC, 2002). A rigorous and direct comparison cannot be made
between the CDC Fluoridation Census 1992 (or 2000) findings and the EPA-OGWDW's Six-Year
Review statistical model estimations. The CDC census findings report the voluntary provision of
qualitative/semi-quantitative information from public drinking water systems that identify if a particular
system is operating as a fluoridating system (with either natural or "adjusted" concentrations of fluoride
within the optimum range of fluoride). The OGWDW Six-Year Review Stage 2 findings are
quantitative, parametric statistical estimations of fluoride occurrence based on compliance monitoring
analytical results of fluoride concentrations in public drinking water systems from the 16-State cross-
section (with the results from the 16-State cross-section then extrapolated to national occurrence
estimates).

Despite the significant differences in the underlying sources of fluoride occurrence information, the
comparison between the  CDC and OGWDW Stage 2 findings is informative.  The comparison suggests
that the Stage 2 modeled national estimates are valid, broadly reflecting and correlating with the general
fluoride concentrations implied by the fluoridation census findings when considering details of the
differences between the census and statistical estimation approaches. Details of the comparative
assessment are included below.

The CDC periodically conducts a national fluoridation census which records the total national  number of
public water systems, and population served by those systems, that operate with natural or adjusted levels
of fluoride in drinking water at optimum levels. The "optimum range" of fluoride in drinking water
(regarding prevention of dental cavities) is from 0.7 mg/L to 1.2 mg/L (although a system can be
considered to be operating as a fluoridating system if it operates within the broader "control range" of
fluoride concentrations from 0.6 mg/L to 1.7 mg/L)1. To complete the fluoridation census, States
voluntarily report: the name, location, and public water system identification number of each fluoridated
water system; the population served by each system;  whether the system operates with adjusted, or
natural, levels of fluoride; the chemical used for fluoridation, if adjusted; and whether or not the system
purchased water. However, no quantitative analytical results are presented, no information is provided
for systems with fluoride occurrence less than 0.7 mg/L (the low end of the optimum range), and the
source water type is not specified.  Therefore, the implication is that all systems that reported as
fluoridating (i.e., allsystemslisted in the CDC census) are considered for comparison purposes  to have a
minimum (average) fluoride concentration of 0.7 mg/L, the low end of the optimum fluoridation range.
  The optimum amount of fluoride in drinking water at PWSs is the range of fluoride that assists in the prevention of dental cavities. The
specific optimum level of fluoride for a given PWS is generally inversely proportional to temperature. It is assumed that individuals drink more
water in warmer climates and higher temperatures. The ingestion of fluoride via drinking water is directly related to the volume of water
consumed. Therefore, since high temperatures result in consumption of higher volumes of water, the amount of fluoride considered optimal is at
the low end of the optimal range (0.7 mg/L) in warmer regions (or warmer seasons) and at the high end of the optimum range (1.2 mg/L) in
cooler regions (or cooler seasons). For example, the optimum level of fluoride for PWSs in southern Florida is 0.7 mg/L and for PWSs in Maine
is 1.2 mg/L. The "control range" recommended by CDC for the optimum concentration is 0.1 mg/L below to 0.5 mg/L above the optimum.

Occurrence Summary and Use Support Document          70                                       March 2002

-------
Table VI.J. 1 .a shows a comparison between the quantitative results based on the  16-State cross-section
data and the qualitative reported findings of the CDC Fluoridation Census 1992 and the unpublished
CDC Fluoridation Census 2000.  The table specifically presents the Stage 2 modeled estimates based on
the Six-Year Review 16-State cross-section compared to the number of fluoridating systems (and
population served by those systems) as reported to the CDC.
Table VI. J.I.a.  Comparison of the National Extrapolations of the Stage 2 Modeled Estimates with
the CDC Fluoridation Census Findings
Stage 2 Modeled Estimates based on
Compliance Monitoring Results of the
16-State Cross-Section
Fluoride
Threshold
(mg/L)
4
2
0.7
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
332
1,885
13,390
Range
284-385
1,769 - 2,000
13,156- 13,624
Total Population Served by Systems
Nationally Estimated to
Exceed Threshold
Best Estimate
191,000
1,978,600
57,022,300 2
Range
118,200-301,000
1,505,100-3,293,100
51,803,600-61,346,400 2
CDC Fluoridation Census
1992 and unpublished
2000 1 Findings
Total
Number of
Systems
Fluoridating
145 '
746 '
14,496 3
Total Population
Served by
Systems
Fluoridating
152,527 '
849,591 '
141, 107,164 3
For a summary of how the Stage 2 modeled estimations are derived, please refer to Section 1.4 of this report. For a detailed description of the
Stage 2 method estimations, please refer to Occurrence Estimation Methodology and Occurrence Findings Report For Six-Year Review of
National Primary Drinking Water Regulations, EPA-815-D-02-005,  March 2002.

1.  The number of systems and population-served by systems with reported fluoride concentrations greater than 2 mg/L and 4 mg/L were
provided by the CDC from the unpublished Fluoridation Census 2000. (These measures of fluoride occurrence relative to these specific fluoride
thresholds are not included in the 1992 or earlier census publications.)
2.  There were no compliance monitoring records for fluoride for the Chicago Water System in the State of Illinois' compliance monitoring data
set. Therefore, since Illinois is one of the states in the 16-state cross-section and the Chicago Water System is known to fluoridate, the Stage 2
modeled estimates presented here do not reflect the population served by the fluoridated water provided by the Chicago Water System (and its
consecutive systems). (The Chicago Water System fluoride were requested, but were received after the Stage 2 modeled estimates were
generated for this draft report.)
3.  This estimate includes public water systems that operate within the "control range" of optimum fluoride concentrations. Therefore, this
estimate includes systems that maintain fluoride concentrations as low as 0.6 mg/L while the Stage 2 model estimates are based on the 0.7 mg/L
fluoride concentration values which is the low end of the optimum (rather than control) range.
The CDC Fluoridation Census 1992 indicates that a total of 14,496 public water systems, serving
141,107,164 people, report that they operate as a fluoridated system2. In comparison, the Stage 2
national occurrence estimates based on the 16-State cross-section indicate a total of 13,390 systems,
serving 57,022,300 people, with estimated mean concentrations of fluoride  greater than 0.7 mg/L.  (For a
detailed presentation of Stage 2 modeled estimations of fluoride occurrence, please refer to Appendix C,
Tables C.16.f and C.16.n). A system with a mean concentration of fluoride greater than 0.7 mg/L is
approximately equivalent to  a "fluoridated" system (in the  CDC census) that reports operation at
optimum fluoride levels, though systems are included in the CDC census as fluoridating if systems
operate within the broader control range and above 0.6 mg/L fluoride.
  The system and population totals listed here are equal to the exact system and population summations of the 50 States. These total sums do
not equal the sums presented in the CDC Fluoridation Census 1992, in part due to the inclusion of the District of Columbia.
Occurrence Summary and Use Support Document
71
March 2002

-------
The differences between the national estimates for fluoridating systems (CDC-14,496 PWSs and national
cross-section-13,390 PWSs) and population served by fluoridating systems (CDC-141,107,164 people
and national cross-section-57,022,300 people) appear to relate to several factors.  First, the lack of
fluoride occurrence data from the Chicago Water System certainly results in an underestimate for the
model estimated number of systems, and population served by those systems.  (Fluoride data for Chicago
were not included in the original compliance monitoring data sets obtained directly from the state of
Illinois. Fluoride compliance monitoring data are reported to the state's public health agency.  These
fluoride data were subsequently requested, but were received after the Stage 2 estimations prepared for
this draft report.)

The size distribution of fluoridating systems (based on population served size categories) is a second
factor influencing the difference in population served between the State data and the CDC census data.
Table VI.J. 1 .b illustrates the distribution of fluoridating systems based on six population served size
categories. Although the number of systems serving 1,000 people or less is similar in the State data and
CDC data (2,824 and 2,540 respectively), this equals a much larger percentage of the smallest systems in
the cross-section data set (66%) than are represented in the CDC census (only 42%).  Therefore, the
cross-section data set (comprised of compliance monitoring data records acquired directly from the
States) has a proportionately larger amount of the smallest systems than does the CDC census (based on
voluntary reporting of which systems  fluoridate). This differing system size profile can result in a
smaller population served by a  similar number of systems (as is the case for the national extrapolations
based on the cross-section data).
Table Vl.J.l.b. Distribution of Fluoridating Systems in the 16 Cross-Section States and the CDC
Fluoridation Census 1992
Population
Served
< 1,000
1,001-5,000
5,001 - 10,000
10,001 - 50,000
50,001 - 100,000
> 100,000
Total
Number of Systems
that Fluoridate -
State Data 1
2,824
750
244
347
54
42
4,261
Percent of State
Total Number of
Fluoridating
Systems
66%
18%
6%
8%
1%
1%
100%
Number of Systems
that Fluoridate -
CDC Data 2
2,540
1,983
578
716
103
79
5,999
Percent of CDC
Total Number of
Fluoridating
Systems
42%
33%
10%
12%
2%
1%
100%
1. The State data results of "number of systems that fluoridate" are derived by using the cross-section data, calculating a simple arithmetic mean
fluoride concentration for each system in a particular population served system size category, and then counting all systems with a mean
concentration greater than 0.7 mg/L (which represents the low end of the range of optimum fluoride concentration for fluoridation).
2. CDC "number of systems that fluoridate" are derived from qualitative reported findings of the CDC 1992 Fluoridation Census.
A third factor influencing the system number and population differences between the CDC and national
cross-section estimates relates to systems that report as "fluoridated systems."  In the CDC Fluoridation
Census 1992, systems that operate within the control range (of optimum fluoride concentration) of 0.6
mg/L and 1.7 mg/L are considered to fluoridate. The Stage 2 statistical estimations and related national
extrapolations were based on a fluoride concentration threshold of 0.7 mg/L (the low end of the
Occurrence Summary and Use Support Document
72
March 2002

-------
optimum, not the control, range of fluoride levels). Therefore, the Six-Year Review Stage 2 estimates of
systems with estimated mean concentrations of fluoride greater than 0.7 mg/L will likely be lower the
CDC census number of systems reporting to operate within the control range (between 0.6 mg/L and  1.7
mg/L fluoride).

Other factors relate to the type of systems included in EPA's Six-Year Review and the CDC fluoridation
census.  The Six-Year Review, based on compliance monitoring, does not include analytical results for
consecutive systems. Because consecutive systems compliance monitoring requirements are at the
discretion of the individual States, the consecutive system compliance monitoring record is not uniform
from state to state.  Therefore, systems identified as consecutive (or, synonymously, as "purchased")
were removed in the raw data sets prior to Stage 2 estimations. The CDC fluoridation census does
include consecutive systems. (CDC estimates that there may be approximately 1,696 consecutive
systems serving a population of 12,850,000 in the 16 states that comprise the 16-State cross-section used
in this Six-Year Review analysis.) Also, while the Six-Year Review 16-State cross-section (compliance
monitoring) data does include monitoring results from non-transient non-community water systems
(NTNCWSs), the CDC does not include NTNCWs in the  fluoridation census.

In summary, given the differences between the CDC census numbers and the EPA model estimates,
rigorous and systematic comparisons cannot be directly made between the two assessments of fluoride
occurrence in public water systems.  However, general comparisons suggest that the EPA Six-Year Stage
2 modeling approach is valid in that it tracks relatively closely to CDC voluntarily reported qualitative
census findings. These comparisons are closer still when the underlying differences between the CDC
census findings and the Six-Year Review model estimates are considered.

The comparison of the number of systems estimated by the Six-Year model to have an estimated mean
fluoride concentration greater than 0.7 mg/L compared to the number of systems reporting to the CDC
census as fluoridating exhibits the largest difference between the two fluoride occurrence assessments.
However, the two differing estimates are likely much closer when considering that: the CDC fluoridating
systems include systems operating with the "control range" (as low as 0.6 mg/L); the Six-Year Review
estimates did not have data for the Chicago Water System, and; the Six-Year Review estimates generally
do not include systems considered to be "consecutive" (or purchasing water).

The comparison between the Six-Year Review model estimates and the unpublished 2000 Census
numbers for systems with  fluoride concentrations above 2 mg/L and 4 mg/L are generally good. For
example, relative to 4 mg/L, the Six-Year Review model estimates are not dramatically higher than the
CDC estimates, especially when comparing the total population served. The EPA system numbers are
modeled estimates, and therefore, it is appropriate to consider the estimated range of results rather than
the single "best estimate." The modeled estimates indicate that from 118,200 people to 301,000 people
could be served by systems with a mean fluoride concentration greater than 4 mg/L. The CDC estimate
(of approximately 153,000 people) falls into this range. The other comparisons with the unpublished
2000 CDC census findings, while not within the statistical model estimate ranges, are relatively similar
(not orders of magnitude apart) given the differences between the two (CDC and EPA) sources of
information/data.  The differences in hundreds of systems  or hundreds of thousands of population served
by systems can be interpreted as somewhat close when considered  relative to the total US population or
the total US population served by fluoridated water.
Occurrence Summary and Use Support Document          73                                     March 2002

-------
2.3.5.3 1969 Community Water Supply Study

The U.S. Public Health Service (USPHS) conducted a Community Water Supply Study (CWSS) in 1969
to assess the status of the nation's water supply facilities and drinking water quality (USPHS 1970b, as
cited in JRB Associates, 1984). Finished water from a total of 969 community supplies in nine
geographically distributed areas were studied.  Except for the State of Vermont (in which all supplies
were sampled), the study locations were standard metropolitan statistical areas (SMSAs). Water samples
were reported to have been taken at various places in the distribution system of each supply studied. Of
the 969 systems, 678 were groundwater supplies, 109 were surface water supplies, and the remaining  182
were mixed sources, purchased water, or of unspecified source. A shortcoming in the 1969 CWSS data
file is that no distinction can be made between those systems for which fluoride was found not to be
present at the detection limit of 0.1 mg/L and those having an actual measured value of 0.1 mg/L.
Therefore, all systems in the 1969 CWSS reporting a value of 0.1 mg/L were treated as though fluoride
were present at that level.

Of the  678 groundwater supplies sampled, 628 (92.6%) had fluoride levels reported to be at or below  1.0
mg/L.  Of the 50 supplies with higher levels, 34 (5.0 %) were found to have fluoride between 1.0 and  2.0
mg/L, 7 (1.0%) between 2.0 and 3.0 mg/L, and 9 (1.3%) between 3.0 and 4.0 mg/L.  No groundwater
systems were observed in the  1969 CWSS to have fluoride levels exceeding  4.0 mg/L.  Of the 16
groundwater systems exceeding 2.0 mg/L, 13 were small systems serving fewer than 500 people. The
mean value of fluoride levels in groundwater systems sampled in the 1969 CWSS was 0.39 mg/L;  the
median was 0.17 mg/L.

In surface water, 102 (93.6%) of the 109 systems sampled had  fluoride levels at or below 1.0 mg/L. Of
the remaining 7 systems, 6 (5.5%) were between 1.0 and 2.0 mg/L, and 1 (0.9%) was between 2.0  and 3.0
mg/L.  The mean value of fluoride levels in surface water systems sampled in the 1969 CWSS was 0.30
mg/L; the median value was 0.18 mg/L.

These 1969 survey findings are generally consistent with the Stage 2 occurrence findings estimated for
this report.  However, the different scopes of the two studies do not enable a direct comparison.

2.3.5.4 1978 Community Water Supply Survey

The 1978 CWSS, conducted by the USEPA, provided data on fluoride levels in a total of 157 surface
water and 345 groundwater supplies dispersed throughout the United States (USEPA, 1978, as cited in
JRB Associates, 1984).  The survey examined systems ranging  in size from 25 people served to more than
100,000 people served.  One to five samples were taken from each system. Water samples classified as
raw, finished (i.e., treated water sampled at the supply), and distribution (i.e., water sampled at a user's
tap) were all included.  However, for the purpose of this analysis, distribution sample data were used
when available. When data for distribution samples were not available, data for finished water samples
were used.  Data on raw water were not included in the analysis.

Of the  345 groundwater supplies sampled, 310 (89.9%) had fluoride levels reported to be at or below  1.0
mg/L.  Of the remaining 35 systems, 23 (6.7%) were found to have fluoride levels between 1.0 and 2.0
mg/L, 5 (1.4%) between 2.0 and 3.0 mg/L, 3 (0.9%) between 3.0 and 4.0 mg/L, 2 (0.6%) between 4.0  and
5.0 mg/L, and 2 (0.6%) between 5.0 and 6.0 mg/L.  Of the 12 groundwater systems exceeding 2.0 mg/L,
seven were small systems serving fewer than 500 people; the remaining  five systems served between 500
and 2,500 people. The mean fluoride concentration in groundwater systems  sampled in the 1978 CWSS
was 0.58 mg/L; the median value was 0.33 mg/L.
Occurrence Summary and Use Support Document          74                                     March 2002

-------
In surface water, 134 (85.4%) of the 157 systems sampled had fluoride levels at or below 1.0 mg/L.  Of
the remaining 23 systems, 22 (14.0%) were between 1.0 and 2.0 mg/L, and 1 (0.6%) was reported to be at
5.0 mg/L. This latter supply was a small system serving between 101-500 people. The mean fluoride
concentration in surface water systems in the 1978 CWSS was 0.69 mg/L; the median value was 0.79
mg/L.

These 1978 survey findings are generally consistent with the Stage 2 occurrence findings estimated for
this report.  However, the different scope and coverage of the two studies prohibits  a more direct
comparison.

2.3.5.5 Rural Water Survey

The Rural Water Survey (RWS) (USEPA, 1982, as cited in JRB Associates, 1984) was conducted
between 1978 and 1980 to evaluate the status of drinking water in rural America as required by Section 3
of the Safe Drinking Water Act. Although the RWS examined drinking water from over 2,000
households in rural areas for a variety of water quality parameters, samples from only 91 public water
supplies were examined for fluoride levels (the sources of water for the remaining households were
private wells or very small systems serving fewer than 25 people). A second limitation regarding the
RWS was that the number of service connections associated with water systems was reported in lieu of
the actual population served by the systems. Dr. Bruce Brower of Cornell University, who collaborated
in the National Statistical Assessment of Rural Water Conditions (based on the RWS data) provided a
factor to convert the data from service connections to the number of people served, based on the average
number of persons per household observed in the RWS.  It must be noted, however, that these population
values are only approximations.

Of the 70 groundwater supplies sampled, 62 (88.6%) had fluoride levels at or below  1 mg/L. Of the
remaining 8 systems, 6 (8.6%) had fluoride levels between 1.0 and 2.0 mg/L and 2 (2.9%) had fluoride
levels between 2.0 and 3.0 mg/L.  The two systems with levels between 2.0 and 3.0 mg/L served between
501 and 2,500 people.  The mean value of the fluoride levels in systems sampled in the RWS was 0.45
mg/L.

Of the 21 surface water systems sampled in the RWS, 19 (90.5%) had fluoride levels at or below 1.0
mg/L.  The remaining 2 (9.5%) systems were reported to have fluoride between 1.0 and 2.0 mg/L. The
mean value of fluoride in surface water systems in the RWS was 0.67 mg/L.

2.3.5.6 Federal Reporting Data System

The Federal Reporting Data System (FRDS) was developed to provide information on public water
supplies with MCL violations as determined through monitoring of all supplies performed under the
requirements of the National Interim Primary Drinking Water Regulations. The FRDS was the
forerunner of the Safe Drinking Water Information System (SDWIS) database.  The Wentworth (1983, as
cited in JRB Associates, 1984) study on MCL violations recorded in the FRDS was conducted prior to the
1986 Fluoride Rule that set the MCL at 4 mg/L. Therefore, the Wentworth (1983, as cited in JRB
Associates, 1984) study assessed violations in a different regulatory context, but the results are
summarized here for historical perspective.

An estimated 558 groundwater supplies and 29 surface water supplies were delivering  drinking water
with fluoride levels exceeding the MCL in place prior to the 1986 fluoride rule (Wentworth, 1983, as
cited in JRB Associates, 1984). Because of the several different temperature-dependent standards
(ranging from 1.4-2.4  mg/L) for fluoride that were in place throughout the U.S. prior to 1986,

Occurrence Summary and  Use Support Document          75                                    March 2002

-------
Wentworth's results should be interpreted with caution. For example, in the fluoride concentration
ranges of 1.0-2.0 mg/L and 2.0-3.0 mg/L, the number of systems counted as exceeding the MCL is an
estimate of the number of systems that by virtue of having fluoride levels in those ranges are in violation
of the locally applicable fluoride standard.  Other systems having fluoride concentrations in those ranges,
but not in violation of the local standard, are not included in the FRDS data. For the concentration
ranges greater than 3.0 mg/L, however, the  FRDS data were considered to be a reasonably accurate
reflection of the total number of systems nationally, since fluoride present at such levels would always be
in violation of the MCL regardless of the local air temperature as based on the standards in place in 1983.

2.3.5.7 U. S. Public Health Service Reports on Optimum Levels

Two reports produced by the USPHS discuss fluoride in drinking water supplies, both of which deal with
systems operating at optimal fluoride levels. These two studies are earlier versions of the fluoridation
census.  (The CDC Fluoridation Census 1992, for example, was described in Section 2.3.5.2)  Note that
these studies are based on "communities" and "places" (see below) rather than on public water systems
and the direct population served by those water systems. These differences should be considered when
comparing the results of these two USPHS  fluoridation studies to the results of the 1992 census and to
the result of the Stage 2 16-State cross-section fluoride analyses. Note also that the USPHS (1970a, as
cited in JRB Associates, 1984) study is for communities served by natural fluoride levels at or above the
minimum optimal level of 0.7 mg/L.

The study titled National Fluoride Content of Community Water Supplies - 1969 (USPHS, 1970a, as
cited in JRB Associates, 1984) was conducted to identify the number and location of community water
supplies in the United States that have natural fluoride levels at or above the minimum optimal level of
0.7 mg/L.  This report provided the names of the communities, their populations, and the reported
concentration in the community water supplies.  However, no information was given on the source of the
water (i.e., ground or surface) in those communities, nor was any information provided for communities
having natural levels less than 0.7 mg/L.

The results of the survey summarized indicated that 2,630 communities with a combined population of
8,106,435 (based on 1960 U.S. Census data) had natural fluoride levels above 0.7 mg/L.  Of these, 1,517
communities (57.7%) with a combined population of 5,752,628 had drinking water supplies with natural
levels in the optimum range of 0.7-1.2 mg/L. There were 1,017 communities (38.7%) with a combined
population of 2,172,706 reported to have natural fluoride levels at or above 1.4 mg/L, the lowest
temperature based MCL value.  Of these, 596 communities with 1,070,222 people were reported to have
levels above 2.0 mg/L. Only 138 communities were reported to have fluoride levels of 4 mg/L or more
and seven communities exceeded 8 mg/L.

The average community size in this survey  was 3,082, ranging from less than 25 to one community of
313,900. Only eight communities had reported populations greater than 100,000; over 75% had fewer
than 2,500. There is some trend toward smaller average community  size for those communities with
levels that exceed the current MCL.  Communities reporting very high levels (e.g., 8 mg/L or more) all
had populations of about 500 persons or less.

The USPHS (1970a, as cited in JRB Associates, 1984) indicated that most (62%) of the communities
having natural fluoride levels of 0.7 mg/L or more were located in the States of Arizona, Colorado,
Illinois, Iowa, New Mexico, Ohio, Oklahoma, South Dakota, and Texas. These same States account for
about 66% of the 1,017 communities exceeding the lowest MCL of 1.4 mg/L. Six States - Delaware,
Hawaii, Massachusetts, Pennsylvania, Tennessee, and Vermont - as well as the District of Columbia
were reported to have no communities with natural levels above 0.7 mg/L.  Although this information

Occurrence Summary and Use Support Document          76                                     March 2002

-------
indicates that there is a regional aspect to high natural fluoride levels in drinking water supplies, it is
important to note that in all but one State (other than those listed above as having no natural fluoride of
0.7 mg/L or more), at least one community had 1.4 mg/L or more.  (The one exception to this, New
Hampshire, had one community at 1.3 mg/L.)

The second report by the U.S. Public Health Service that addressed optimum fluoride levels in drinking
water supplies is the 7975 Fluoridation Census (USPHS, 1977, as cited in JRB Associates, 1984).  This
report is similar to the USPHS report discussed above, although it includes locations and population data
for places having both natural and adjusted fluoride levels of 0.7 mg/L or more. However, this report
presents no data on the concentration of fluoride in these systems.  Bock (1982, as cited in JRB
Associates, 1984) indicates that, like the  1969 report, the 1975 census included all systems with levels
above 0.7 mg/L, not just those in the optimum fluoride range. Again, like the 1969 study, no information
was given on water source type nor for actual water sample analytical results.

The 7975 Fluoridation Census reported that 9,425 places3 with a total population of 105,338,343 (using
1970 U.S. Census data) have drinking water with natural or adjusted fluoride levels of 0.7 mg/L or more.
Of these, 2,630 places having a population of 10,711,049 have natural fluoride levels at 0.7 mg/L or
more. (The appearance of "2,630" in both the USPHS surveys is coincidental. The terms "community"
used in the  1969 study and "place" used in the 1975 Census do not appear to be synonymous. Some
States show significantly more "places"  in the 1975 report than "communities" in the 1969 survey, while
others show fewer; States showing the same number in both studies often list different locations.)

Although of limited value for this discussion of fluoride occurrence in drinking water, the 7975
Fluoridation Census does indicate a higher population receiving natural fluoride levels of 0.7 mg/L or
more  as compared to the 1969 study. Also, the Fluoridation Census supports the 1969 study conclusion
that the nine States listed earlier are responsible for most (60% in the 1975 fluoridation census) of the
places having natural levels  of 0.7 mg/L or more.

2.3.5.8 Estimated National Occurrence of Fluoride in Public Drinking Water Supplies Based on
Survey Data

The JRB Associates (1984)  study developed estimates of the number of drinking water supplies
nationally within each of the source/size categories expected to have fluoride present within various
concentration ranges.  The study developed the national estimates using both Federal survey data and
compliance monitoring data. The Public Health Service data (briefly reviewed above in Section 2.3.5.7)
were reportedly not useful in the JRB Associates (1984) development  of the national estimates for
several reasons. Those data were for "places" or "communities" rather than supplies, and there was no
indication as to the source of the water for those locations.  The 7975 Fluoridation Census provided no
quantitative data on fluoride levels in the drinking water, indicating only whether the water had natural or
adjusted levels of 0.7 mg/L or more; the 1969 USPHS  study addressed only communities with natural
levels at or above 0.7 mg/L.

The Federal survey data together with the compliance monitoring data provided information on water
source and system size for extrapolating to all public water supplies. The compliance monitoring data is
believed to provide a reasonably accurate picture of those supplies in the U.S. with high fluoride levels.
However, because it addressed only supplies in violation of the current MCL, the compliance monitoring
data provided no information on supplies with lower levels of fluoride. The Federal survey data, which
3  "Place" refers to a geographic entity listed in the Worldwide Geographic Location Codes prepared by the General Services Administration.

Occurrence Summary and Use Support Document          77                                     March 2002

-------
provides some information on supplies at all concentrations, is of limited value for estimating the number
of systems nationally having high fluoride levels because of the small number of supplies sampled. That
is, because there are relatively few supplies in the U.S. having high levels, the chances of observing them
in the surveys is small.  On the other hand, when a supply with a high fluoride level was observed in the
surveys, extrapolating to the national level results in a questionably high national estimate. It was,
therefore, decided that the compliance monitoring data would be used to describe national fluoride
occurrence at the higher concentrations and the survey data would be used to estimate national fluoride
occurrence at lower concentrations.

The estimates of supplies having > 3.0 mg/L are taken directly from the compliance monitoring data,
since systems having levels > 3.0 would always be reported through FRDS as MCL violations (at the
MCL in place prior to the 1986 fluoride rule). For lower levels (i.e., < 3.0 mg/L), national estimates
were calculated as follows. First, the results of the three surveys were combined for groundwater and
surface water, respectively, to determine the total number of systems sampled with values < 3.0 mg/1, as
well as the number falling within each concentration range of interest.  National estimates of supplies
within each concentration range were then calculated in proportion to that observed  in the combined
Federal survey data.

For the > 2.0-3.0 mg/L range for groundwater, the estimates calculated from the Federal survey data were
compared to the compliance monitoring data for each source/size category and the larger value was
chosen for the national estimates. This choice was based on the recognition that the compliance
monitoring data may underestimate the actual number of supplies in the > 2.0-3.0 mg/L range since not
all supplies having fluoride in that range are necessarily in violation of the MCL and, therefore, would
not be reported through FRDS. Higher estimates, computed from the survey data, were considered to be
more conservative. On the other hand, in those instances where the compliance data showed more
supplies in that range than the estimates from the survey data, the compliance data were considered more
representative of actual occurrence, since the lower estimates from the survey data probably resulted
from the small sample size.  Specifically, for ground water in the > 2.0-3.0 mg/L ranges, the national
estimates for supplies serving 5,000 or fewer people are based on the Federal survey data and on the
compliance data for supplies serving more than 5,000 persons. For surface water, all values in the > 2.0-
3.0 mg/L range are from the compliance monitoring data. All estimates of ground water and surface
water supplies having < 2.0 mg/L of fluoride are based on the Federal survey data.

The estimates indicate that 82.8% of all public water supplies in the U.S. in the mid-1980s using ground
water had fluoride present in the range of 0.1 to  1.0 mg/L. About 8.6% of ground water supplies  (4,214
total) were estimated to have fluoride levels > 1.0 mg/L, though most of these fall in the 1.0-2.0 mg/L
range.  Most of the estimated 1,324 groundwater supplies with fluoride levels exceeding 2.0 mg/L are
expected to be small or medium sized supplies, while all of the supplies having very high levels (> 7.0
mg/L) are expected to be small systems serving fewer than  1,000 people.

For surface water, the majority of systems (78.5%) are also expected to have fluoride present in the 0.1-
1.0 mg/L range. Of the 726 surface water supplies estimated to have levels above 1.0 mg/L,  almost all
(704) are estimated to be in the 1.0-2.0 mg/L range. As in the case of groundwater,  most of the surface
water supplies expected to have levels exceeding 2.0 mg/L are small and medium sized systems.

2.3.6  Conclusion

The record of use, production and release of fluoride and its compounds suggest it is found at low levels
in soil, water, and air. Most fluorides are used in steel production, while other fluoride compounds are
used in a variety of applications including drinking water additives.  Recent statistics regarding import

Occurrence Summary and Use Support Document          78                                      March 2002

-------
for consumption of fluorspar (the primary source of fluorine and its compounds) indicate production and
use are robust. Industrial releases of fluorine have occurred since 1995 in 11 States and Puerto Rico,
while releases of hydrogen fluoride have occurred since 1988 in 49 States.  Air emissions constitute the
greatest proportion of the total on- and off-site releases of fluorine and hydrogen fluoride. Fluoride is
also a national NAWQA analyte. Fluoride occurrence in ambient surface and ground waters is high.
Approximately 69.0% of all surface water sites had analytical detections of fluoride, compared to 67.0%
of ground water sites. The percentage of sites with fluoride detections greater than the MCL (4 mg/L)
drastically decreases, however, to 0.05% for surface water sites and 0.5% for ground water sites. The
Stage 2 analysis, based on the  16-State cross-section, estimated that approximately 0.511% of combined
ground water and surface water systems serving 0.0897% of the population had estimated mean
concentrations of fluoride greater than the MCL of 4 mg/L. Based on this estimate, approximately 332
PWSs nationally serving about 191,000 people are expected to have estimated mean concentrations of
fluoride greater than 4 mg/L.

Since fluoride and its compounds occur naturally occurring, the balanced geographic distribution of the
16-State cross-section should adequately cover the range of natural occurrence of fluoride from low to
high.  The 16-State cross-section also contains a substantial proportion of the States with reported TRI
releases of fluorine and hydrogen fluoride. Based on this use and release evaluation, the 16-State cross-
section appears to adequately represent fluoride occurrence nationally.

2.3.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 1993. Toxicological Profile for
       Fluorides, Hydrogen Fluoride, and Fluorine (F). U.S. Department of Health and Human
       Services, Public Health Service. 244 pp. + Appendices. Available on the Internet at:
       http://www.atsdr.cdc.gov/toxprofiles/tp 11 .pdf

Bock, W.  1982.  Census Program, Center for Disease Control, U.S. Public Health Service,  Athens, GA.
       Private Communication with author of JRB Associates, 19 84.

CDC.  1992.  Fluoridation Census 1992. Atlanta, GA: Centers for Disease Control and Prevention,
       Public Health Service, U.S. Department of Health and Human Services.

Gilliom, R.J., O.K. Mueller, and L.H. Nowell.  1998.  Methods for comparing water-quality conditions
       among National  Water-Quality Assessment Study Units, 1992-95. U.S. Geological  Survey
       Open-File Report 97-589.  Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589/, last updated October 9, 1998.

JRB Associates.  1984.  Occurrence of Fluoride in Drinking Water, Food, and Air. Prepared for and
       submitted to EPA on February 9, 1984.

USEPA.  1978.  Community Water Supply Survey. Computer data files on inorganic contaminants.

USEPA.  1982.  Rural Water Survey. Computer data provided by Department of Sociology, Cornell
       University.

USEPA.  2000.  TRI Explorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.
Occurrence Summary and Use Support Document         79                                      March 2002

-------
USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 2000. Mineral Commodity Summaries, February, 2000 - Fluoride. Available on the Internet at:
       http://minerals.usgs.gov/minerals/pubs/commodity.

USPHS.  1970a. Natural Fluoride Content of Community Water Supplies; 1969.  Bethesda, MD:
       National Institutes of Health, Public Health Service, Department of Health, Education, and
       Welfare.

USPHS.  1970b. Community Water Supply Survey. Washington, DC: Bureau of Water Hygiene, Public
       Health Service, Department of Health, Education, and Welfare.

USPHS.  1977. Fluoridation Census 1975.  Atlanta, GA: Center for Disease Control, Public Health
       Service, Department of Health, Education, and Welfare.

Wentworth, N. 1983. Summary of fluoride drinking water violations data derived from: (a) FRDS24 of
       FY82 (3/14/83); (b) FRDS Chemical Violation List of FY78-FY81; (c) Summary of Variances
       and Exemptions Issued FY82; (d) Report on "Defluoridation Study for Suffolk Area of VA,"
       (10/82); and (e) Region VIII report on "Solving Public Health Problems"  (2/8/83). U.S.
       Environmental Protection Agency, Office of Drinking Water. Personal communication to author
       of JRB Associates, 1984 on March 24,  1983.

WHO. 1984. Fluorine and Fluorides.  Geneve, Switzerland:  World Health Organization, Distribution
       and Sales Service, EHC Number 36.
Occurrence Summary and Use Support Document         80                                    March 2002

-------
2.4    Mercury
Table of Contents

2.4.1  Introduction, Use and Production  	  82
2.4.2  Environmental Release  	  84
2.4.3  Ambient Occurrence 	  85
2.4.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  85
2.4.5  Additional Drinking Water Occurrence Data  	  90
2.4.6  Conclusion	  95
2.4.7  References  	  96
Tables and Figures

Table 2.4-1: Facilities that Manufacture or Process Mercury  	  83

Table 2.4-2: Environmental Releases (in pounds) for Mercury in the United States, 1988-1999	  84

Table 2.4-3: Environmental Releases (in pounds) for Mercury Compounds in the
       United States, 1988-1999 	  85

Table 2.4-4: Stage 1 Mercury Occurrence Based on 16-State Cross-Section - Systems  	  86

Table 2.4-5: Stage 1 Mercury Occurrence Based on 16-State Cross-Section - Population	  87

Table 2.4-6: Stage 2 Estimated Mercury Occurrence Based on 16-State Cross-Section - Systems ....  88

Table 2.4-7: Stage 2 Estimated Mercury Occurrence Based on 16-State Cross-Section - Population . .  89

Table 2.4-8: Estimated National Mercury Occurrence - Systems and Population Served	  90
Occurrence Summary and Use Support Document          81                                     March 2002

-------
2.4.1  Introduction, Use and Production

Mercury (Hg) is a member of Group IIB of the periodic table. Mercury is a naturally occurring metal
which has several forms. The principal valence states of mercury are +1 (mercurous) and +2 (mercuric),
both of which are found in natural waters (USEPA, 1979, as cited in Wade Miller, 1990). The metallic
mercury is a shiny, silver-white, odorless liquid. If heated, it is a colorless, odorless gas. Mercury
combines with other elements, such as chlorine, sulfur, or oxygen, to form inorganic mercury compounds
or "salts," which are usually white powders or crystals. Mercury also combines with carbon to make
organic mercury compounds (ATSDR, 1999b).

The occurrence of mercury in surface and ground water is influenced by natural and anthropogenic
sources. Although mercury occurs in the environment as a result of human activities, the major source of
mercury in the environment results from geothermal activity, volatilization from mineral deposits, or
volcanic activity (Wade Miller, 1990). Mercury in surface water may be the result of human activities or
naturally occurring sources. Because mercury does not readily leach through soils, contamination of
ground water is believed to be unlikely.  Therefore, most mercury in ground water is considered to be
natural in origin (Battelle, 1977; Perwak et al., 1980, all as cited in Wade Miller, 1990).

Like all elements, the same amount of mercury has existed on the planet since the Earth was formed.
However,  the amount of mercury mobilized and released into the environment has increased since the
beginning of the industrial age.  The human activities that are most responsible for causing mercury to
enter the environment are the burning of materials (such as batteries), use of fuels (such as coal) that
contain mercury, and certain industrial processes (USEPA, 2001).

Mercury has many applications in industry due to its unique properties, such as its fluidity, its uniform
volume expansion over the entire liquid temperature range, its high surface tension, and its ability to
alloy with other metals (ATSDR, 1999a). Mercury is used in its pure form in thermometers, barometers,
and other consumer products (NSC, 2001). However, domestic consumption of mercury has
demonstrated a downward trend since the early 1970s. In 1995, consumption was 102,000 pounds, down
10% from 1994.  The largest commercial use of mercury in the United States was for electrolytic
production of chlorine and caustic soda in mercury cells, accounting for 35% of domestic consumption.
Manufacture of wiring devices and switches accounted for 19%, measuring and control instruments for
9%, dental equipment and supplies used 7%,  electric lighting used 7%, and other uses amounted to 21%
(USGS, 1997). Due to the high toxicity of mercury in most of its forms, many applications have been
canceled as a result of attempts to limit the amount of exposure to mercury waste (ATSDR, 1999a).

The canceled uses of mercury include use of phenylmercuric acetate as a fungicide in interior latex paints
and exterior paints. The former use was banned in 1990 and the latter in 1991.  This occurred because
the paint released mercury vapors as it degraded.  Most agricultural applications of mercury compounds
in bactericides and fungicides have been canceled  due to the toxicity of mercury. Minor uses of mercury
in the production of felt hats and as a wood preservative have also been terminated (ATSDR, 1999a).

In electrical applications, mercury is a critical element in alkaline batteries. Due to its toxicity, the
amount of mercury in batteries is being reduced from 0.1% to 0.025%. Mercury vapors are used in
discharge tubes of some electrical lamps because this makes the lamps efficient, long-lasting, and more
energy efficient.  In 1985, 64% of the mercury used in the U.S. was for electrical applications, and this
amount had fallen to 29% in 1992 (ATSDR, 1999a).

Probably one of mercury's most familiar applications is in medicine in dental restorations. It is used in
dentistry because of its ability to alloy with other metals.  Estimates of annual mercury usage by U.S.

Occurrence Summary and Use Support Document          82                                      March 2002

-------
dentists range from approximately 220,500 pounds in the 1970s to 154,300 pounds in 1995. Until 30
years ago, mercury was also used in pharmaceuticals.  Mercury salts were used in antiseptics, diuretics,
skin lightening creams, and laxatives. Organic mercury compounds were used in antisyphilitic drugs and
some laxatives.  Phenylmercury acetate was used in contraceptive gels and foams and as a disinfectant.
Almost all of these uses have now been replaced by others that are more effective and less toxic
(ATSDR,  1999a).

In the chemical and mining industries, mercury is used as a catalyst in reactions to form polymers and in
the preparation of chlorine and caustic soda from brines. In mining, mercury is used in gold mining to
extract gold from ores through amalgamation (ATSDR, 1999a).

 Organic mercury (especially methyl mercury and dimethyl mercury) can imperil human welfare because
of its tendency to bioaccumulate in the flesh of certain fish. Low levels of mercury contamination in
oceans and lakes can lead to toxic contamination offish, making it hazardous for human consumption,
especially for children and pregnant women (NSC, 2001).

The primary method of obtaining mercury is from mining.  Ten percent of mercury mining is open pit
mining and 90% is from underground mining techniques. There are currently 34 facilities that produce
or process mercury in the U.S. (ATSDR, 1999a). U.S. production of mercury in 1985 was 1,254,000
pounds, and world production in 1986 was  13,376,000 pounds (NSC, 2001).  As of 1995, eight mines in
California, Nevada, and Utah produced  mercury  as a by-product from gold mining operations.
Approximately 127,900 pounds of mercury were produced from eight mines in 1991 and 141,100 pounds
were produced as a by-product from nine mines in  1992. Since then, production volumes have been
withheld to avoid disclosing Bureau of Mines company proprietary data (ATSDR, 1999a).

Table 2.4-1 lists the facilities in each State that manufacture or process mercury, the intended use, and
the range of maximum amounts of mercury that are stored on site derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1999a).
Table 2.4-1: Facilities that Manufacture or Process Mercury
Facility
Occidental Chemical
Tuscaloosa Steel Corp.
Occidental Chemical
Olin Chlor-Alkali Prods.
Alexander Mfg. Co.
Micro Switch
Valspar Corp.
Durakool Inc.
Hermaseal Co.
U.S. Steel
United Techs.
Koch Sulfur Prods. Co.
BF Goodrich Co.
DuPont
Borden Chemicals &
Dow Chemical Co.
Pioneer Chlor Alkali Co.
PPG Ind. Inc.
Holtrachem Mfg.
Location"
Muscle Shoals, AL
Tuscaloosa, AL
New Castle, DE
Augusta, GA
Mason City, IA
Freeport, IL
Rockford, IL
Elkhart, IN
Elkhart, IN
Gary, IN
Edinburgh, IN
De Soto, KS
Calvert City, KY
Louisville, KY
Geismar, LA
Plaquemine, LA
Saint Gabriel, LA
Lake Charles, LA
Orrington, ME
Range of maximum
amounts on site in
pounds
100,000-999,999
0-99
100,000-999,999
100,000-999,999
0-99
10,000-99,999
10,000-99,999
10,000-99,999
10,000-99,999
10,000-99,999
10,000-99,999
0-99
100,000-999,999
not available
100,000-999,999
1,000-9,999
100,000-999,999
100,000-999,999
100,000-999,999
Activities and uses
Chemical processing aid
Article component
Chemical processing aid
Chemical processing aid
Import, On-site use/processing, Article component
Article component
Formulation component
Article component
Article component
Produce, Byproduct
Article component
Ancillary/other use
Chemical processing aid
not available
Import, On-site use/processing, Chemical processing aid
Produce, Byproduct
Chemical processing aid
Chemical processing aid
Chemical processing aid
Occurrence Summary and Use Support Document
83
March 2002

-------
Facility
Elm Plating Co.
Kerr Corp.
Holtrachem Mfg. Co.
Mercury Refining Co.
Ashta Chemicals Inc.
Component Repair
Sinclair Oil Corp.
Advanced Environmental
Bethlehem Apparatus
Zinc Corp. of America
Olin Corp.
Occidental Chemical
Georgia-Pacific West
Vulcan Materials Co.
PPG Ind. Inc.
Location"
Jackson, MI
Romulus, MI
Riegelwood, NC
Albany, NY
Ashtabula, OH
Mentor, OH
Tulsa, OK
Allentown, PA
Hellertown, PA
Monaca, PA
Charleston, TN
Deer Park, TX
Bellingham, WA
Port Edwards, WI
New Martinsville.
Range of maximum
amounts on site in
pounds
0-99
1,000-9,999
100,000-999,999
10,000-99,999
10,000-99,999
not available
100-999
10,000-99,999
100,000-999,999
10,000-99,999
100,000-999,999
100,000-999,999
100,000-999,999
100,000-999,999
100.000-999.999
Activities and uses
Article component
Repackaging
Chemical processing aid
Produce, Sale/Distribution, Repackaging, Ancillary/other use
Chemical processing aid
not available
Produce, Byproduct
Produce, Sale/distribution
Produce, Import, On-site use/processing, Sale/distribution,
Produce, Impurity
Chemical processing aid
Chemical processing aid
Chemical processing aid
Chemical processing aid
Chemical processing aid
Tost office State abbreviations used
Source: ATSDR, 1999a compilation O/TR196 1998 data
2.4.2 Environmental Release

Mercury and mercury compounds are both listed as Toxics Release Inventory (TRI) chemicals.  Table
2.4-2 illustrates the environmental releases for mercury from 1988 - 1999. (There are only mercury data
for these years.) Air emissions constitute most of the on-site releases, with the amount released
decreasing in earlier years and generally remaining constant in more recent years. The decrease in air
emissions, as well as decreases in surface water discharges and off-site releases (including metals or
metal compounds transferred off-site), have contributed to decreases in mercury total on- and off-site
releases in previous years.  Releases to land decreased until 1996, at which point they began to rise.  No
underground injection releases were recorded or reported for mercury. These TRI data for mercury were
reported from 31 States and Puerto Rico with 12 States reporting all 12 years (USEPA, 2000).  Of the 31
States, 10 are included in the  16 State cross-section (used for analyses of mercury occurrence in drinking
water; see Section 2.4.4).  (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 2.4-2: Environmental Releases (in pounds) for Mercury in the United States, 1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
11,275
12,591
12,163
14,286
13,262
11,277
11,642
14,222
18,208
22,391
25,095
22,905
Surface Water
Discharges
133
134
391
468
192
175
267
273
629
751
1,555
1,397
Underground
Injection
-
-
-
-
-
-
-
-
0
0
0
0
Releases
to Land
2,419
3,069
1,016
537
1,016
1,351
1,801
3,122
5,292
4,184
4,942
13,279
Off-Site Releases
7,070
15,721
26,346
13,012
14,228
14,097
18,355
43,854
112,969
177,015
126,087
258,718
Total On- &
Off-site
Releases
20,897
31,515
39,916
28,303
28,698
26,900
32,065
61,471
137,098
204,341
157,679
296,299
 Source: USEPA, 2000
Occurrence Summary and Use Support Document
84
March 2002

-------
Table 2.4-3 illustrates the environmental releases for mercury compounds between 1988 and 1999. Air
emissions constitute most of the on-site releases, with some fluctuation over the years.  Both surface
water discharges and releases to land have also fluctuated, with releases to land dramatically increasing
in 1998-1999.  Underground injections contribute relatively little to total releases, and have not been
registered since 1998. Off-site releases of mercury compounds are considerable.  The levels of total on-
and off-site releases have generally followed the highs and lows of the levels of off-site releases, with no
discernible trend. These TRI data for mercury compounds were reported from 32 States with one State,
Pennsylvania, reporting every year (USEPA, 2000). Of the 32 States, 12 are included in the 16 State
cross-section (used for analyses of mercury occurrence in drinking water; see Section 2.4.4). (For a map
of the 16-State cross-section, see Figure 1.3-1.)
Table 2.4-3:  Environmental Releases (in pounds) for Mercury Compounds in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
2,110
2,372
2,356
2,916
3,156
2,716
3,421
3,249
3,080
1,158
4,009
2,376
Surface Water
Discharges
36
34
34
73
136
151
184
307
52
58
13
9
Underground
Injection
-
-
41
9
6
7
15
9
9
21
36
27
Releases
to Land
5,700
2,550
0
0
0
0
11
17
2
15
260
0
Off-Site Releases
53,051
19,858
25,448
29,437
207,097
26,166
56,009
191,945
36,875
36,041
56,113
17,916
Total On- &
Off-site
Releases
60,897
24,814
27,879
32,435
210,395
29,040
59,640
195,527
40,018
37,293
60,431
20,328
 Source: USEPA, 2000
2.4.3  Ambient Occurrence

The most comprehensive and nationally consistent data describing ambient water quality in the United
States are being produced through the United States Geological Survey's (USGS) National Water Quality
Assessment (NAWQA) program. However, no national NAWQA data are available for mercury.

Also, there are no other ambient data available for mercury. A summary document entitled "Occurrence
and Exposure Assessment for Mercury in Public Drinking Water Supplies" (Wade Miller, 1990), was
previously prepared for past USEPA assessments of mercury. However, no information on the ambient
occurrence of mercury was included in that document. (The document did include information regarding
mercury occurrence in drinking water, which is discussed in Section 2.4.5 of this report.)

2.4.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of mercury occurrence presented in the following section is based on State compliance
monitoring data from the  16 cross-section States. The 16-State cross-section is the largest and most
Occurrence Summary and Use Support Document
85
March 2002

-------
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest:  0.002 mg/L; 0.001 mg/L; and 0.0005 mg/L.

All sixteen cross-section State data sets contained occurrence data for mercury. These data represent
more than 64,000 analytical results from approximately 19,000 PWSs during the period from 1984 to
1998 (with most analytical results from 1992 to 1997). The number of sample results and PWSs vary by
State, although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness.  The overall modal detection limit for mercury in the  16 cross-section States is equal to
0.001 mg/L. (For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this report.)

2.4.4.1 Stage 1 Analysis Occurrence Findings

Table 2.4-4 illustrates the low occurrence of mercury in drinking water for the public water systems in
the 16-State cross-section relative to three thresholds: 0.002 mg/L (the current MCL), 0.001  mg/L (the
modal MRL), and 0.0005 mg/L. A total of 50 ground water and surface water PWSs (approximately
0.263%) had at least one analytical result exceeding the MCL; 0.900% (171 systems) of PWSs had at
least one analytical result exceeding 0.001 mg/L; and 2.90% (551 systems) of PWSs had at least one
analytical  result exceeding 0.0005 mg/L.

Approximately 0.252% (44 systems) of ground water PWSs had at least one analytical result greater than
the MCL.  About 0.848% (148 systems) of ground water PWSs  had at least one analytical result above
0.001 mg/L. The percentage of ground water systems with at least one result greater than 0.0005 mg/L
was equal to 2.68% (467 systems).

Only 6 (0.387% of) surface water systems had at least one analytical result greater than the MCL.  A total
of 23 (1.48% of) surface water systems had at least one analytical result greater than 0.001 mg/L.
Eighty-four surface water systems (5.42%) had at least one analytical result exceeding 0.0005 mg/L.
Table 2.4-4:  Stage 1 Mercury Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0005
Percent of Systems
Exceeding Threshold
0.252%
0.848%
2.68%
Number of Systems Exceeding
Threshold
44
148
467

Surface Water
0.002
0.001
0.0005
0.387%
1.48%
5.42%
6
23
84

Combined Ground &
Surface Water
0.002
0.001
0.0005
0.263%
0.900%
2.90%
50
171
551
Occurrence Summary and Use Support Document
86
March 2002

-------
Reviewing mercury occurrence in the 16 cross-section States by PWS population served (Table 2.4-5)
shows that approximately 0.276% of the population (almost 300,000 people) was served by PWSs with at
least one analytical result of mercury greater than the MCL (0.002 mg/L). Approximately 4.5 million
(4.33% of) people were served by systems with an exceedance of 0.001 mg/L. A total of 9,268,500
people (8.82%) were served by systems with at least one analytical result greater than 0.0005 mg/L.

The percentage of population served by ground water systems with analytical results greater than the
MCL was equal to 0.538% (almost 232,000 people). When evaluated relative to 0.001 mg/L or 0.0005
mg/L, the percent of population exposed was equal to 4.18% (1,803,400 people) and 10.2%
(approximately 4.4 million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.002 mg/L was
equal to 0.0937% (58,000 people). Approximately 4.43% (about 2.7 million people) of the population
served by surface water systems were served by systems with estimated mean concentrations of mercury
greater than 0.001 mg/L. When evaluated relative to 0.0005 mg/L, the percent of population exposed
was equal to 7.84% (almost 4.9 million people).
Table 2.4-5:  Stage 1 Mercury Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0005
Percent of Population Served
by Systems Exceeding
Threshold
0.538%
4.18%
10.2%
Total Population Served by
Systems Exceeding Threshold
231,900
1,803,400
4,409,500

Surface Water
0.002
0.001
0.0005
0.0937%
4.43%
7.84%
58,000
2,746,300
4,859,000

Combined Ground &
Surface Water
0.002
0.001
0.0005
0.276%
4.33%
8.82%
290,000
4,549,700
9,268,500
2.4.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 2.4-6 and 2.4-7.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).
Occurrence Summary and Use Support Document
87
March 2002

-------
A total of 13 (0.0672%) ground water and surface water PWSs in the 16 States had an estimated mean
concentration of mercury exceeding 0.002 mg/L. Approximately 73 (0.385% of) PWSs in the 16 States
had an estimated mean concentration exceeding 0.001 mg/L, and 431 (2.27%) had an estimated mean
concentration exceeding 0.0005 mg/L.

An estimated 13 ground water PWSs in the 16 cross-section States (0.0727%) had a mean concentration
greater than 0.002 mg/L, 72 (0.411%) had a mean concentration greater than 0.001 mg/L, and 416
(2.38%) had a mean concentration greater than 0.0005 mg/L. Approximately 1 (0.00529%), 1
(0.0863%), and 16 (1.03%) surface water PWSs in the 16 States had estimated mean concentrations
exceeding 0.002 mg/L, 0.001 mg/L, and 0.0005 mg/L, respectively.
Table 2.4-6:  Stage 2 Estimated Mercury Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0727%
0.411%
2.38%
Range
0.0459% -0.103%
0.321% -0.504%
2. 13% -2.62%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
13
72
416
Range
8-18
56-88
372 - 457

Surface Water
0.002
0.001
0.0005
0.00529%
0.0863%
1.03%
0.000% - 0.0645%
0.000% - 0.258%
0.581% -1.55%
1
1
16
0-1
0-4
9-24

Combined Ground
& Surface Water
0.002
0.001
0.0005
0.0672%
0.385%
2.27%
0.0421% -0.0948%
0.300% - 0.474%
2.05% - 2.49%
13
73
431
8-18
57-90
390 - 473
Reviewing mercury occurrence by PWS population served (Table 2.4-7) shows that approximately
0.00627% (an estimate of approximately 6,600 people) of population served by all PWSs in the 16 cross-
section States were potentially exposed to mercury levels above 0.002 mg/L. The percentage of
population served by PWSs in the 16 States with levels of mercury above 0.001 mg/L and 0.0005 mg/L
were 0.0987% (an estimated 103,700 people) and 1.56% (about 1.6 million people), respectively.

When the percent of population served by ground water systems was evaluated relative to a threshold of
0.002 mg/L, 0.001 mg/L, and 0.0005 mg/L, the percentage of population exposed in the 16 cross-section
States was equal to 0.0147% (an estimated 6,300 people), 0.151% (an estimated 65,100 people) and
1.36% (an estimated 585,500 people), respectively.

The percentage of population served by surface water systems with levels above 0.002 mg/L was equal to
0.000429% (an estimated 300 people served by systems in the 16 States), and the percentage of
population served with levels above 0.001 mg/L was 0.0624% (an estimated 38,700 people in the  16-
State cross-section). The percentage of the population served by surface water systems with levels above
0.0005 mg/L was 1.70% (just over 1 million people).
Occurrence Summary and Use Support Document
March 2002

-------
Table 2.4-7:  Stage 2 Estimated Mercury Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.0147%
0.151%
1.36%
Range
0.00224% - 0.0535%
0.0661% -0.288%
0.923% -2.26%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
6,300
65,100
585,500
Range
1,000-23,100
28,500 - 124,100
397,800 - 975,300

Surface Water
0.002
0.001
0.0005
0.000429%
0.0624%
1.70%
0.000% -0.00921%
0.000% -1.36%
0.473% - 2.27%
300
38,700
1,056,100
0 - 5,700
0 - 840,400
293,000-1,408,800

Combined Ground
& Surface Water
0.002
0.001
0.0005
0.00627%
0.0987%
1.56%
0.000924% - 0.0223%
0.0308% - 0.853%
0.946% - 2.06%
6,600
103,700
1,641,600
1,000-23,400
32,300 - 896,900
994,500-2,166,000
2.4.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 44 PWSs serving approximately 13,400 people nationally could be exposed to
mercury concentrations above 0.002 mg/L. About 250 systems serving 210,200 people had estimated
mean concentrations greater than 0.001 mg/L. Approximately 1,477 systems serving about 3.3 million
people nationally were estimated to have mean mercury concentrations greater than 0.0005 mg/L.  (See
Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 43 PWSs serving about 12,600 people nationally had mean
concentrations greater than 0.002 mg/L.  Approximately 244 systems serving about 129,300 people
nationally had estimated mean concentration values that exceeded 0.001 mg/L. About 1,416 ground
water systems serving over 1.1 million people had estimated mean concentrations greater than 0.0005
mg/L.

Approximately 1 surface water system serving 500 people was estimated to have a mean concentration of
mercury above 0.002 mg/L. About 5 surface water systems serving 79,400 people had estimated mean
concentrations greater than 0.001 mg/L.  An estimated 58 surface water systems serving approximately
2,169,600 people had mean concentrations greater than 0.0005 mg/L.
Occurrence Summary and Use Support Document
89
March 2002

-------
Table 2.4-8:  Estimated National Mercury Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
43
244
1,416
Range
27-61
191-300
1,267-1,557
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
12,600
129,300
1,163,600
Range
1,900-45,900
56,700 - 246,600
790,600-1,938,100

Surface Water
0.002
0.001
0.0005
1
5
58
0-4
0-14
32-87
500
79,400
2,169,600
0-11,700
0-1,726,500
601,900-2,894,100

Combined Ground
& Surface Water
0.002
0.001
0.0005
44
250
1,477
27-62
195-308
1,335-1,619
13,400
210,200
3,327,200
2,000 - 47,400
65,500-1,817,800
2,015,700-4,390,100
2.4.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of mercury in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Exposure Assessment for Mercury in Public Drinking Water Supplies" (Wade Miller,
1990), is presented in this section.  This variety of studies and information are presented regarding levels
of mercury in drinking water, with the scope of the reviewed studies ranging from national to  regional.
Note that none of the studies presented in the following section provide the quantitative analytical results
or comprehensive coverage that would enable direct comparison to the occurrence findings estimated
with the cross-section occurrence data presented in Section 2.4.4.  These additional studies, however, do
enable a broader assessment of the Stage 2 occurrence estimates presented for this Six-Year Review. All
the following information in Section 2.4.5 is taken directly from "Occurrence and Exposure Assessment
for Mercury in Public Drinking Water Supplies" (Wade Miller, 1990).

2.4.5.1 Reported Occurrence in Drinking Water

This section presents information on measured mercury levels in drinking water from public water
supplies. Section 2.4.5.1.1 provides the results from several national-scale surveys in which mercury
levels were measured in drinking water.  Section 2.4.5.1.2 provides the most current information from the
U.S. Environmental Protection Agency's Federal Reporting Data System (FRDS) on the status of
drinking water supplies reported to be in violation of the current standard of 2 |ig/L for mercury.

2.4.5.1.1 Federal Survey Data

Several national-scale surveys have been conducted that provide data on mercury in public drinking
water supplies.  These include the 1978 Community Water Supply Survey, the Rural Water Survey, the
National Organic Monitoring Survey, and the National Inorganics and Radionuclides Survey.  The
Occurrence Summary and Use Support Document
90
March 2002

-------
following sections describe these surveys and present the data on mercury levels in ground water and
surface water supplies.

1978 Community Water Supply Survey (1978 CWSS)

The 1978 Community Water Supply Survey (CWSS) was conducted by the USEPA to determine the
occurrence of organic and inorganic compounds in public water supplies.  Drinking water samples were
provided by approximately 500 supplies; however, due to analytical problems, reliable data for mercury
were available for only 35 ground water and  10 surface water supplies (USEPA, 1983, as cited in Wade
Miller, 1990).

Details on the analytical method used for mercury were not available; from the information provided by
Glick (1984, as cited in Wade Miller, 1990) the minimum quantifiable concentration appeared to be 0.5
l-ig/L. Supplies provided one to five samples of raw, finished, and/or distribution water.  However, Brass
(1983, as cited in Wade Miller, 1990) indicated that reporting inconsistencies made it impossible to
distinguish between finished and distribution samples.  Therefore, distribution and finished sample
results were averaged; raw water data were not used.

Of the 35 ground water supplies sampled, nine (26 percent) contained mercury at concentrations ranging
from 0.5 to 175 |ig/L. Except for the one high value of 175 |ig/L, none of the samples exceeded 2 |ig/L.
None of the 10 surface water supplies sampled contained mercury above the minimum quantifiable
concentration of 0.5 |ig/L.

Rural Water Survey (RWS)

The Rural Water Survey (RWS) conducted between 1978 and 1980 evaluated the status of drinking water
in rural America as required by Section 3 of the Safe Drinking Water Act. More than 2,000 households
served by 648 public water supplies (494 ground water, 154 surface water) were surveyed. Many of
these households used private wells or very small systems serving fewer than 25 people, and only a
subsample of the supplies evaluated included analyses for mercury (71 ground water and 21 surface
water supplies).  Results of the inorganic analyses were provided to Science  Applications International
Corporation (SAIC) as a computer file by Brower (1983, as cited in Wade Miller, 1990).

A problem with the RWS was that the number of service connections associated with water systems was
reported in lieu of the actual populations served by the systems. Dr. Bruce Brower of Cornell University,
who collaborated in the National Statistical Assessment of Rural Water Conditions (based on the RWS
data) provided a factor to convert the data from service connections to the number of people served,
based on the average number of persons per household observed in the RWS. It must be noted, however,
that these population values  are only approximations.

Details were not  available on the sample collection or analytical methodology used. However, the
minimum quantifiable concentration for mercury appeared to range from 0.2 to 0.5  |ig/L. Of the 71
ground water supplies studied, 24 (34 percent) contained mercury at levels ranging from 0.3 to 12.3 |ig/L.
For surface water supplies, 11 of the 21 supplies sampled (52 percent) had mercury present at
concentrations ranging from 0.3 to 16.0 |ig/L.

National Organic Monitoring Survey (NOMS)

The National Organic Monitoring Survey (NOMS), which was conducted by the USEPA in 1976 and
1977, was intended primarily to provide data for establishing MCLs for organic compounds in drinking

Occurrence Summary and Use Support Document          91                                     March 2002

-------
water.  A substudy of NOMS analyzed samples for 27 trace elements in the 113 supplies sampled (91
surface water, 19 ground water, 3 mixed sources) (USEPA, 1980, as cited in Wade Miller, 1990).
Mercury data are available for 86 surface water and 15 ground water supplies.  Samples were taken of
treated, finished water leaving the treatment plant. Occasionally, samples were taken of distribution
system water in close proximity to the treatment plant. The results of the inorganic analyses for NOMS
were provided to SAIC by the Technical Support Division of EPA's Office of Drinking Water (USEPA,
1985, as cited in Wade Miller, 1990).  No information on the analytical procedures used was available;
the minimum quantifiable concentration appeared to be approximately 0.5 |ig/L.

Mercury was not found in any of the 15 ground water supplies sampled. For surface water supplies,
mercury was observed in two of the 86 supplies sampled (2 percent) at concentrations of 0.5 |ig/L and 1.5
p-g/L.

National Inorganics and Radionuclides Survey (NIRS)

In 1981 the USEPA's Office of Drinking Water (ODW) initiated the National Inorganics and
Radionuclides Survey (NIRS) to characterize the occurrence of various contaminants in community
drinking water supplies.  The survey focused on the presence of 36 inorganics, including mercury, and
four radionuclides in ground water supplies from throughout the United States. Implementation of the
survey and sampling were accomplished by the ODW's Technical Support Division between July 1984
and October 1986.

The NIRS sampling program was designed to reflect the national distribution of community ground water
supplies by size of population served as inventoried by the Federal Reporting Data System (FRDS).  The
FRDS data was stratified into the following four population-size categories: very small (serving 25-500),
small (serving 501-3,300) medium  (serving 3,301-10,000), and large/very large (serving >10,000).  A
total of 1,000 sites were selected randomly from the FRDS data in proportion to the four size categories.
Approximately 2.1% of the supplies in each size category were chosen for sampling. Of the 1,000
targeted sites, 990 were actually sampled in the NIRS.

Sample collection and location within  each supply were designed to reflect the quality of water actually
received by the consumer. Samples were collected after three minutes of flushing in order to represent
the finished water in the distribution system.  To the extent possible, the sampling location was chosen at
a point of maximum use in the distribution system.  The method used to analyze for mercury was not
reported. The minimum reporting limit (MRL) was 0.2 |ig/L.

Mercury results are available for all of the 990 sites sampled in the NIRS. Approximately 98% of the
samples had mercury concentrations less than 0.5 |ig/L. Of the 19 supplies (two percent) with  higher
mercury concentrations, only one exceeded the current MCL of 2 |ig/L.  This maximum value was 2.1
|ig/L for a small supply, serving between 501 and 3,300 people. The overall mean of the samples was
approximately 0.2 |ig/L, which was the MRL.

2.4.5.1.2 Compliance Monitoring Data

The Federal Reporting Data System (FRDS) provides information on public  water supplies in violation
of current MCLs as determined through compliance monitoring of all supplies performed by the States
under the requirements of the National Interim Primary Drinking Water Regulations. Only violations of
current MCLs (i.e., 2 |ig/L for mercury) and approved variances and exemptions from the  standards are
recorded in FRDS. Monitoring is required annually for surface water supplies and every three  years for
ground water supplies.

Occurrence Summary and Use Support Document          92                                     March 2002

-------
The FRDS database (FRDS, 1990, as cited in Wade Miller, 1990) was searched for violations reported
during the three year period from 1987 through 1989. The data indicated that there were three public
ground water and three public surface water supplies providing drinking water which had mercury levels
above the current MCL of 2 |ig/L. The ground water supplies were all in the very small and small size
categories. A maximum value of 4.3 |ig/L was reported for a supply serving 70 people.

The three surface water violations were distributed between the very small, medium, and large/very large
size categories. A maximum value of 7 |ig/L was reported for a system serving 8,000 people.
Additionally, no variances or exemptions to the current mercury MCL of 2 |ig/L were reported to have
been granted for surface  water supplies.

2.4.5.2 Estimated National Occurrence in Public Water Supplies

Section 2.4.5.1 presented the results of several studies that provide information on the occurrence of
mercury in public drinking water supplies. This section presents national estimates of mercury
occurrence in public drinking water supplies based on those study results.

2.4.5.2.1  Methodology

National  estimates of the occurrence of contaminants in drinking water supplies are needed to serve as
input to the analysis of regulatory costs and benefits. A draft document has been prepared that describes
the methodology used to estimate national occurrence of the inorganic drinking water contaminants
addressed under Phase II of the national primary drinking water regulations development process
(USEPA, 1986, as cited in Wade Miller, 1990). In general, national survey data are used as the basis for
the national estimates. The most representative data sets are selected and combined, stratified
appropriately by water source and system size sampled, and applied to a "delta-log normal" distribution
model. Using that model, the probability of contaminant occurrence above any given concentration is
determined for each source/size category of supplies and applied to the total number of water supplies in
those groups. The resulting national occurrence estimates are presented in tables that show the
cumulative number of public water supplies, within various water source  and size categories, expected to
have certain concentrations.

The mercury occurrence  estimates are based on the results of the NIRS survey for ground water supplies
and the combined results of the 1978 CWSS, the RWS, and the NOMS for surface water supplies. The
NIRS data are the most current ground water data available, and the analytical results are considered to
be highly reliable because of the extensive quality assurance program being employed. None of the
surveys that analyzed surface water supplies could be shown to be any more or less representative of the
universe of water supplies than the other surveys, and it was therefore determined to be most appropriate
to combine the results  from all of the surveys to form the basis of the national occurrence estimates.

2.4.5.2.2  Results

Section 2.4.5.2.1 noted that the national estimates of mercury in finished water are based on the results
from NIRS for ground water supplies, and on the combined results from the 1978 CWSS, RWS and
NOMS for the surface water supplies.

For ground water, the model estimates that 45 public ground water supplies have mercury levels greater
than the proposed MCLG and MCL of 2 |ig/L. Six of those supplies are estimated to have mercury levels
exceeding 3  |ig/L. The majority of the supplies expected to have mercury concentrations above 2 |ig/L
Occurrence Summary and Use Support Document          93                                     March 2002

-------
are in the small and very small size categories. FRDS reports only three systems in violation of the 2
|ig/L MCL.

For surface water, the model predicts that approximately 7 percent or 395  surface water supplies contain
mercury in excess of 2 |ig/L. An estimated 3 percent (189 supplies) are expected to contain greater than
6 |ig/L mercury.  The systems expected to exceed the proposed MCLG and MCL of 2 |ig/L are
distributed throughout all size categories. FRDS reports only three systems in violation of the current
MCL of 2 |ig/L, one of which contains mercury at levels greater than 6 |ig/L.

The occurrence estimates formulated from the delta lognormal distribution model predict a larger number
of systems contain mercury in excess of 2 |ig/L than reported by FRDS for both ground and surface water
supplies. It cannot be determined whether this is due to an over-estimate by the model or an under-
reporting of actual violations in FRDS. For the surface water estimates the higher number of supplies
predicted to exceed 2 |ig/L are attributed to the use of the Rural Water Survey results. Six of the 21
samples from the RWS exceeded 2 |ig/L and five of the six were above 6 |ig/L. None of the surface
water samples from the  1978 CWSS or the NOMS contained mercury above 2 |ig/L.  The RWS data may
have significantly skewed the estimated occurrence numbers.

2.4.5.3 Estimated National Exposure from Public Water Supplies

2.4.5.3.1 Methodology

Section 2.4.5.2.1 noted that a separate methodology document has been prepared to describe the
approach used to estimate national occurrence of inorganic contaminants in public water supplies. That
document also addresses the approach to estimating the population exposed to contaminant levels in
public drinking water supplies.  In summary, the occurrence  probability density functions obtained from
the national survey data on supplies contaminated at various levels are applied to the  number of people
using water supplies in the various size categories.

2.4.5.3.2 Results

Based on the national occurrence estimates derived from the national survey data, 72,000 people are
expected to be exposed to mercury from ground water supplied drinking water systems above the
proposed MCLG and MCL of 2 |ig/L. The model also estimates that 6,000 of those people will be
exposed to mercury at levels greater than 3 |ig/L. Zero people are expected to be exposed to mercury in
drinking water at levels greater than 4 |ig/L.  The FRDS data indicate that only 424 people are exposed to
mercury at levels exceeding the current MCL of 2 |ig/L. Additionally, 70 people receive drinking water
with a mercury concentration between 4 and  5  |ig/L.

Based upon the cumulative national population exposure estimates from the national  survey data for
mercury in surface water supplies, the model predicts that approximately 10 million people are exposed
to mercury levels above 2 |ig/L.  Information regarding the population exposed to mercury at levels
exceeding the current MCL of 2 |ig/L as reported in FRDS indicates that 18,766 people are exposed to
mercury levels above 2 |ig/L.

The FRDS data and the modeled exposure estimates differ by several orders of magnitude. Once again,
this may be due to an over estimation by the mathematical model or an underreporting of FRDS
violations. It should also be noted that the higher national occurrence and exposure estimates for surface
water supplies are due to the fact that they are based on the (apparently) anomalous high results of the
RWS.

Occurrence Summary and Use Support Document          94                                      March 2002

-------
Overall, the model predicts that, of the estimated 219 million people using public drinking water
supplies, about 5 percent or 10.3 million are exposed to mercury at levels above 2 |ig/L. The remaining
95 percent of the population (208.7 million people) are either not exposed to mercury or are receiving
levels less than 2 |ig/L. Of the 10.3 million people exposed, approximately 5 million are expected to
receive drinking water with mercury levels in excess of 6 |ig/L.

2.4.6  Conclusion

Mercury and many of its compounds are naturally occurring and found at low levels in soil, water, and
air. Furthermore, mercury is mined in the United States for widespread use. Most mercury is used in
production of mercury cells, while other uses for mercury include manufacture of wiring devices and
switches, dental equipment, and electric lighting. Recent statistics regarding production and use of
mercury  indicate they are steady. Industrial releases of mercury and mercury compounds have been
reported to TRI since 1988 from 31 States and 32 States, respectively.  Off-site releases constitute a
considerable amount of total releases, with releases to land the most significant on-site releases. No
national NAWQA data were available for mercury.  The Stage 2 analysis, based on the 16-State cross-
section, estimated that approximately 0.0672% of combined ground water and surface water systems
serving 0.00627% of the population had estimated mean concentrations of mercury greater than the MCL
of 0.002 mg/L. Based on this estimate, approximately 44 PWSs nationally serving about 13,400 people
are expected to have estimated mean concentrations of mercury greater than 0.002 mg/L.

Mercury is a naturally occurring element.  Therefore, the balanced geographic distribution of the 16-State
cross-section should adequately cover the  range of natural occurrence of mercury from low to high.  The
16-State  cross-section also contains a substantial proportion of the States with reported TRI releases of
mercury  and mercury compounds. Based  on this use and release evaluation, the 16-State cross-section
appears to adequately represent mercury occurrence nationally.

2.4.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1999a. Toxicological Profile for
       Mercury. U.S. Department of Health and Human Services, Public Health Service.  617 pp. +
       Appendices. Available on the Internet at: http://www.atsdr.cdc.gov/toxprofiles/tp46.pdf

Agency for Toxic Substances and Disease Registry (ATSDR).  1999b. ToxFAQs for Mercury.  U.S.
       Department of Health and Human Services, Public Health Service.  Available on the Internet at:
       http://atsdr.cdc .gov/tfacts46 .html

Battelle.  1977'. Multimedia levels: Mercury. Washington, DC: Office of Toxic Substances, USEPA.
       EPA-560/6-77-031.

Brass, H. 1983.  U.S. Environmental Protection Agency, Office of Drinking Water, Technical Support
       Division, Cincinnati, OH. Personal communication to Frank Letkiewicz, SAIC.

Brower, B. 1983. Computer data file of analytical results for public water supplies sampled in the Rural
       Water Survey. Cornell University, Department of Rural Sociology.

Federal Reporting Data System (FRDS).  1990.  Computer file extracts of compliance monitoring data
       for the Safe Drinking Water Act.  1987-1989.
Occurrence Summary and Use Support Document          95                                     March 2002

-------
Glick, E. 1984. U.S. Environmental Protection Agency, Office of Drinking Water, Technical Support
       Division, Cincinnati, OH. Personal communication to Frank Letkiewicz, SAIC.

National Safety Council (NSC). 2001. Mercury Chemical Backgrounder.  Itasca, IL: National Safety
       Council.  Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?id=117&chempath=chemicals, accessed
       July 23, 2001.

Perwak, J., M. Goyer, L. Nelken, K. Scow, M. Wald, and D. Wallace.  1980. An exposure and risk
       assessment mercury. Final draft report prepared by A.D. Little submitted to Office of Water and
       Waste Management, USEPA, Washington, DC. EPA Contracts 68-01-3857, 68-01-5949.

USGS.  1997. Mercury. United States Geological Survey.

USEPA.  1979. Water-related  environmental fate of 129 priority pollutants. Volume 1. Office of Water
       Planning and Standards, USEPA.  EPA-440/4-79-029a.

USEPA.  1980. Trace elements in 113 U.S. drinking waters. A substudy of the National Organics
       Monitoring Survey.  Cincinnati, OH: Technical Support Division, Office of Drinking Water,
       USEPA.

USEPA.  1983. Data on inorganics from the Community Water Supply Survey. Provided by E.M. Glick,
       Drinking Water Quality Assessment Branch, Technical Support Division, Office of Drinking
       Water, Cincinnati, Ohio.

USEPA.  1985. U.S. Environmental Protection Agency. Computerized listing of inorganics data for the
       National Organics Monitoring Survey (NOMS).  Provided to Science Applications International
       Corporation by E.B. Dotson, Water Supply Technology Branch, Technical Support Division,
       Office of Drinking Water, Cincinnati, Ohio.  November 12, 1985.

USEPA.  1986. Overview of methodology used to estimate national occurrence and exposure to
       inorganic drinking water contaminants (draft).  Prepared by Science Applications International
       Corporation for U.S. Environmental Protection Agency, Office of Drinking Water, Science and
       Technology Branch.

USEPA.  2000. TR1Explorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on: Mercury.
       Office of Ground Water and Drinking Water, USEPA. Available on the Internet at:
       http://www.epa.gov/safewater/dwh/c-ioc/mercury.html, last updated March 9, 2001.

Wade Miller Associates, Inc.  1990. Occurrence and Exposure Assessment for Mercury in Public
       Drinking Water Supplies.  Prepared for and submitted to EPA on July 26, 1990.
Occurrence Summary and Use Support Document          96                                     March 2002

-------
2.5    Thallium
Table of Contents

2.5.1  Introduction, Use and Production  	  98
2.5.2  Environmental Release  	  99
2.5.3  Ambient Occurrence 	  100
2.5.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  101
2.5.5  Additional Drinking Water Occurrence Data  	  105
2.5.6  Conclusion	  106
2.5.7  References  	  107
Tables and Figures

Table 2.5-1: Facilities that Manufacture or Process Thallium and Compounds	  98

Table 2.5-2: Environmental Releases (in pounds) for Thallium in the United States, 1989-1999	  99

Table 2.5-3: Environmental Releases (in pounds) for Thallium Compounds in the
       United States, 1988-1999 	  100

Table 2.5-4: Stage 1 Thallium Occurrence Based on 16-State Cross-Section - Systems	  101

Table 2.5-5: Stage 1 Thallium Occurrence Based on 16-State Cross-Section - Population	  102

Table 2.5-6: Stage 2 Estimated Thallium Occurrence Based on 16-State Cross-Section -
       Systems	  103

Table 2.5-7: Stage 2 Estimated Thallium Occurrence Based on 16-State Cross-Section -
       Population	  103

Table 2.5-8: Estimated National Thallium Occurrence - Systems and Population Served  	  104
Occurrence Summary and Use Support Document          97                                     March 2002

-------
2.5.1  Introduction, Use and Production

Pure thallium (Tl) is a bluish-white metal that is found in trace amounts in the earth's crust. Thallium
metal forms alloys with other metals and readily amalgamates with mercury (Wade Miller, 1989).
Thallium forms compounds in which it occurs in either the +1 (thallous) or +3 (thallic) valence states.
The +1 salts are slightly to moderately soluble in water; however, trivalent thallium is reportedly the
predominant ionic species found in seawater and freshwater in equilibrium with atmospheric oxygen
(Wade Miller, 1989).

Thallium is found mainly in several rare minerals, deposits of which are so small that they have not been
of commercial importance. The main commercial sources of thallium have been as by-products from the
production or processing of other materials, notably flue dusts from pyrites (FeS2), lead, zinc and
cadmium (Wade Miller, 1989).

In its pure form, thallium is odorless and tasteless. It can also be found combined with other substances
such as bromine, chlorine, fluorine, and iodine.  When it's combined, it appears colorless-to-white or
yellow (ATSDR, 1995).

Thallium is a metal found in natural deposits in ores containing other elements.  The greatest use of
thallium is in specialized electronic research equipment (USEPA, 2001).  The domestic production of
thallium ceased in 1981. Prior to that, thallium  had been recovered as a byproduct from the flue dust and
residuals that resulted from the smelting of zinc, copper, and lead ores. Based upon the estimated
thallium content of zinc ores,  U.S. mine production of thallium was 992 pounds in 1986 and  1987,
compared to 31,000 pounds in the rest of the world (ATSDR, 1992).

Currently thallium can only be obtained by importation.  In 1987, about 4,500 pounds of thallium were
imported (USEPA, 2001) and in  1990, approximately 1,540 pounds were imported (HSDB, 2000).
About 60-70% of thallium use is concentrated in the semiconductor industry in the production of
switches and closures (ATSDR, 1992). Thallium is also used in the pharmaceutical industry for
myocardial imaging, in the manufacture of thallaphide cells, atomic beam clocks, photoelectric cells,
lamps, thermometers, alloys, scintillation counters (HSDB, 2000), and in highly refractive optical glass
(ATSDR,  1992). Thallium was formerly used as a rodenticide and ant killer, but it has not been available
for that function in the U.S. since 1975 and its use as such is restricted in many other parts of the world
(HSDB, 2000).

Table 2.5-1 shows the six facilities that either import thallium, use thallium and  its compounds in
manufacturing processes, or produce them as byproducts. The table also gives the uses of the product
and the range of maximum amounts on site. All information was derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1992).
Table 2.5-1:  Facilities that Manufacture or Process Thallium and Compounds3
Facility
Philips Industries, Inc., Dexter Axle
Tenneco Oil Company
Koch Refining Company
River Cement Company
Location
Albion, IN
Chalmette, LA
Saint Paul, MN
Festus, MO
Maximum amount on site in
pounds
10,000-99,999
0-99
1,000-9,999
100,000,000-499,999,999
Use
Import; as a manufacturing aid
As a processing aid
As an impurity
As a reactant
Occurrence Summary and Use Support Document          98                                     March 2002

-------
 Facility
                              Location
Maximum amount on site in
pounds	
                                                                        Use
 Sohio Oil Company Toledo Refinery    Oregon, OH
 Dana Corporation                 Reading, PA
100-999
0-99
As an impurity
As an impurity
"Derived from TRI 1989
Source: AT SDR, 1992 compilation of 1989 TRI data


2.5.2  Environmental Release

Thallium and thallium compounds are both listed as Toxics Release Inventory (TRI) chemicals. Table
2.5-2 illustrates the environmental releases for thallium from 1989 - 1999. No thallium data were
reported in 1992 or 1996.  Air emissions and releases to land are the only factors that have consistently
contributed to on-site releases. Air emissions decreased (from a high of 1,000 pounds in 1989 to a low of
15 pounds in 1998) until 1999, when they spiked to over 2,000 pounds.  Total on- and off-site releases of
thallium remained relatively stable for many years until just recently.  Off-site releases have been on an
upward trend since 1997, and releases to land went from 755 pounds to 4,355 pounds from 1995-1999.
No underground injection releases and very few surface water discharges were reported for thallium.
These TRI data for thallium were  reported from 10 States (USEPA, 2000).  Of the 10 States, four are
included in the  16 State cross-section (used for analyses of thallium occurrence in drinking water; see
Section 2.5.4).  (For a map of the  16-State cross-section, see Figure 1.3-1.)
Table 2.5-2: Environmental Releases (in pounds) for Thallium in the United States, 1989-1999
Year
1999
1998
1997
1995
1994
1993
1991
1990
1989
On-Site Releases
Air Emissions
2,137
15
256
255
255
255
30
750
1,000
Surface Water
Discharges
-
-
-
-
-
-
1
5
-
Underground
Injection
-
-
-
-
-
-
-
-
-
Releases
to Land
4,355
3,400
1,000
755
755
755
-
-
500
Off-Site Releases
4,828
3,665
1,500
195
255
5
953
916
250
Total On- &
Off-site
Releases
11,320
7,080
2,756
1,205
1,265
1,015
984
1,671
1,750
 Source: USEPA, 2000
Table 2.5-3 illustrates the environmental releases for thallium compounds between 1988 and 1999.
There were no data, however, reported in 1993 or 1995. Releases to land have generally constituted most
of the on-site releases, with dramatic increases in 1994 and 1998. No real trend is suggested in the data.
Air emissions were relatively constant from 1988-1991, but since then have fluctuated from zero to over
1,000 pounds. Releases to land have fluctuated dramatically; from 1988-1991 they were constant, then
they increased to about 3,700 pounds in 1994 and over 400,000 pounds in 1998.  Off-site releases of
thallium compounds vary from almost zero to under 2,000 pounds. No underground injection releases
and very few surface water discharges were reported for thallium compounds. Overall, total on- and off-
site releases have ranged from 5 to over 400,000 pounds. The fluctuations are relatively modest except
Occurrence Summary and Use Support Document
99
                March 2002

-------
for increases in 1998, mostly as a result of the increased releases to land. The TRI data for thallium
compounds were reported from 20 States (USEPA, 2000). Of the 20 States, eight are included in the 16
State cross-section (used for analyses of thallium occurrence in drinking water; see  Section 2.5.4). (For a
map of the 16-State cross-section, see Figure 1.3-1.)
Table 2.5-3:  Environmental Releases (in pounds) for Thallium Compounds in the United States,
1988-1999
Year
1999
1998
1997
1996
1994
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
654
1,060
0
0
171
755
255
255
256
253
Surface Water
Discharges
750
250
--
--
--
--
--
--
--
--
Underground
Injection
--
--
--
--
--
--
--
--
--
-
Releases
to Land
252,800
409,000
--
--
3,695
505
255
255
250
250
Off-Site Releases
1,838
759
180
5
5
255
5
5
504
1,256
Total On- &
Off-site
Releases
256,042
411,069
180
5
3,871
1,515
515
515
1,010
1,759
 Source: USEPA, 2000
2.5.3  Ambient Occurrence

The most comprehensive and nationally consistent data describing ambient water quality in the United
States are being produced through the United States Geological Survey's (USGS) National Water Quality
Assessment (NAWQA) program.  However, no national NAWQA data are available for thallium.

2.5.3.1 Additional Ambient Occurrence Data

Information is very limited with regard to concentrations of thallium in ambient waters. A summary
document entitled "Occurrence and Exposure Assessment of Thallium in Public Drinking Water
Supplies" (Wade Miller, 1989), was previously prepared for past USEPA assessments of beryllium. No
additional national or regional studies on the occurrence of thallium in ground water were available.
However, one additional study was presented regarding levels of thallium in ambient surface waters.  The
following information is taken directly from that document.

V.J. Ciccone and Associates (USEPA, 1984, as cited in Wade Miller,  1989) provided information on a
study by Durum and Haffty (1961, as cited in Wade Miller, 1989) in which analyses of ambient water for
the occurrence of minor elements were reported. The study involved analysis of samples collected at 15
surface water sites throughout the United  States and Canada. The number of samples collected varied
from two to seven per site.  Samples were analyzed for thallium using the emission spectroscopy
technique.  The results of this study revealed that thallium was below the analytical detection limit in all
samples collected; however, the detection limit was not provided.
Occurrence Summary and Use Support Document
100
March 2002

-------
2.5.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of thallium occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date.  These data were evaluated
relative to two concentration thresholds of interest:  0.002 mg/L; and 0.001 mg/L.

All sixteen cross-section State data sets contained occurrence data for thallium.  These data represent
more than 47,000 analytical results from approximately 18,000 PWSs during the period from 1985 to
1998 (with most analytical results from 1992 to 1997).  The number of sample results and PWSs vary by
State, although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness.  The overall modal detection limit for thallium in the 16 cross-section States is equal to
0.001 mg/L. (For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this report.)

2.5.4.1 Stage 1 Analysis Occurrence Findings

Table 2.5-4 illustrates the Stage 1 analysis of thallium occurrence in drinking water for the public water
systems in the 16-State cross-section relative to two thresholds of interest: 0.002 mg/L (the current
MCL) and 0.001 mg/L (the modal MRL). Approximately 0.679% (122  systems) of all ground water and
surface water PWSs had any analytical results of thallium exceeding the MCL.  Approximately 1.89%
(340 systems) of PWSs had any exceedances of the modal MRL.

A greater proportion of surface water systems, as compared to ground water systems, exceeded each
threshold. However, the actual number of ground water systems with threshold exceedances  was much
greater than the actual number of surface water systems with threshold exceedances.  Over 100
(approximately 0.660% of) ground water PWSs had any analytical results exceeding the MCL, compared
to 13 (about 0.886% of) surface water PWSs. About 1.81% (299 systems) of ground water PWSs had
any analytical results exceeding 0.001 mg/L. This compares to about 2.79% (41 systems) of surface
water PWSs with any analytical results greater than the modal MRL.
Table 2.5-4:  Stage 1 Thallium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
Percent of Systems
Exceeding Threshold
0.660%
1.81%
Number of Systems
Exceeding Threshold
109
299

Surface Water
0.002
0.001
0.886%
2.79%
13
41

Combined Ground &
Surface Water
0.002
0.001
0.679%
1.89%
122
340
Occurrence Summary and Use Support Document
101
March 2002

-------
Reviewing thallium occurrence in the 16 cross-section States by PWS population served (Table 2.5-5)
shows approximately 1.85% (a total of 1,927,200 people) of the population was served by PWSs with
any analytical results greater than the MCL. The percentage of population served by PWSs with any
exceedance of the modal MRL was 4.32% (4,500,600 people).

The number of people served by ground water systems with any analytical results greater than 0.002
mg/L was equal to 858,700 (about 2.01%). A total of 1,068,500 people (approximately 1.73%) were
served by surface water systems with any analytical results greater than 0.002 mg/L. About 4.46% of the
population served by ground water PWSs (1,903,500 people) had any exceedances of 0.001 mg/L,
compared to approximately 4.21% (2,597,100 people) of the population served by surface water systems.
Table 2.5-5:  Stage 1 Thallium Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
Percent of Population
Served by Systems
Exceeding Threshold
2.01%
4.46%
Total Population Served bj
Systems Exceeding
Threshold
858,700
1,903,500

Surface Water
0.002
0.001
1.73%
4.21%
1,068,500
2,597,100

Combined Ground &
Surface Water
0.002
0.001
1.85%
4.32%
1,927,200
4,500,600
2.5.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 2.5-6 and 2.5-7.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Approximately 0.283% (51 systems) of all ground water and surface water PWSs in the 16-States are
estimated to have mean concentrations of thallium above 0.002 mg/L (the current MCL). The percentage
of PWSs in the 16 cross-section States with estimated mean concentrations exceeding 0.001 mg/L (the
modal detection limit) was about 1.20% (215 PWSs).

A greater proportion of ground water systems, as compared to surface water systems, were  estimated to
exceed each threshold.  Approximately 48 (0.292% of) ground water systems in the 16 cross-section
States had estimated mean concentrations of thallium above 0.002 mg/L, compared to approximately 3
(0.180% of) surface water systems. About 1.21% (an estimated 199 systems) of ground water systems in
the 16-State cross-section had estimated mean concentrations greater than  0.001 mg/L. This compares
Occurrence Summary and Use Support Document
102
March 2002

-------
with about 1.07% (about 16 systems) of the surface water systems with estimated mean concentrations
greater than 0.001 mg/L.
Table 2.5-6: Stage 2 Estimated Thallium Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.292%
1.21%
Range
0.206% -0.382%
1.01% -1.41%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
48
199
Range
34-63
167-232

Surface Water
0.002
0.001
0.180%
1.07%
0.000% - 0.409%
0.545% -1.64%
3
16
0-6
8-24

Combined Ground
& Surface Water
0.002
0.001
0.283%
1.20%
0.195% -0.367%
1.01% -1.39%
51
215
35-66
181-250
Reviewing thallium occurrence by PWS population served (Table 2.5-7) shows that approximately
77,500 (0.0743% of) the PWS population in the 16 States were served by systems with mean thallium
concentrations above 0.002 mg/L. When evaluated relative to a threshold of 0.001 mg/L, the percent of
population exposed increased significantly to about 0.527% (approximately 550,000 people served in the
16 States).

About 63,800 (0.150% of) people served by ground water systems in the 16 States were served by
systems with estimated mean concentrations of thallium above 0.002 mg/L. An estimated 0.887% of the
population (about 378,100 people) were served by ground water systems whose mean concentration
value exceeded 0.001 mg/L.

Approximately 0.0222% of the population served by surface water PWSs (an estimated 13,700 people
nationally) had estimated mean concentrations of thallium above 0.002 mg/L.  About 0.279% of the
population served by surface water PWSs had mean concentrations greater than 0.001 mg/L (almost
172,000 people).
Table 2.5-7: Stage 2 Estimated Thallium Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.150%
0.887%
Range
0.0319% -0.573%
0.383% -1.41%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
63,800
378,100
Range
13,600-244,300
163,300 - 602,200
Occurrence Summary and Use Support Document
103
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Surface Water
0.002
0.001
0.0222%
0.279%
0.000% - 0.0866%
0.0664% -1.41%
13,700
171,900
0 - 53,400
40,900 - 869,800

Combined Ground
& Surface Water
0.002
0.001
0.0743%
0.527%
0.0156% -0.246%
0.244% -1.24%
77,500
550,000
16,300-256,900
254,600-1,296,300
2.5.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 184 PWSs serving approximately 158,300 people nationally could be exposed to
thallium concentrations above 0.002 mg/L. Approximately 778 systems serving about 1.1 million people
served nationally were estimated to have thallium concentrations greater than 0.001 mg/L. (See Section
1.4 for a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 174 PWSs serving about  128,200 people nationally had mean
concentrations greater than 0.002 mg/L.  Approximately 717 systems serving about 759,7000 people
nationally had estimated mean concentration values that exceeded 0.001 mg/L.

Approximately 10 surface water systems serving 28,300 people were estimated to have mean
concentrations of thallium above 0.002 mg/L.  About 60 surface water systems serving 355,100 people
had estimated mean concentrations greater than 0.001  mg/L.
Table 2.5-8:  Estimated National Thallium Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
174
717
Range
122 - 227
602 - 836
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
128,200
759,700
Range
27,300 - 490,900
328,100-1,209,800

Surface Water
0.002
0.001
10
60
0-23
30-91
28,300
355,100
0-110,300
84,500 - 1,796,600

Combined Ground
& Surface Water
0.002
0.001
184
778
127-239
655 - 905
158,300
1,123,400
33,300 - 524,600
520,000 - 2,647,700
Occurrence Summary and Use Support Document
104
March 2002

-------
2.5.5  Additional Drinking Water Occurrence Data

Additional data sources regarding the occurrence of thallium in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Exposure Assessment of Thallium in Public Drinking Water Supplies" (Wade Miller,
1989), is presented in this section.  Only one national scale study - the National Inorganics and
Radionuclides Survey - provided data on the occurrence of thallium in drinking water sources. That
survey, however, is limited to ground water sources. No regional ground water or surface water studies
or national surface  water studies were found. Note  that none of the studies presented in the following
section provide the quantitative analytical results or comprehensive coverage that would enable direct
comparison to the occurrence findings estimated with the cross-section occurrence data presented in
Section 2.5.4.  These additional studies, however, do enable a broader assessment of the Stage 2
occurrence estimates presented for this Six-Year Review.  All the following information in Section 2.5.5
is taken directly from "Occurrence and Exposure Assessment of Thallium in Public Drinking Water
Supplies"  (Wade Miller, 1989).

2.5.5.1 Ground Water Sources - National Studies

The National Inorganics and Radionuclides Survey  (NIRS) was a two-year national sampling and
analytical  study designed to gather information on the occurrence of selected inorganic compounds and
radionuclides in drinking water.  From July 1984 through May 1986 approximately 1,000 samples of
finished drinking water from ground water supplies throughout the United States were collected.  Of 989
utilities for which thallium data were reported, 28 systems served populations greater than 10,000 (large
and very large, combined), 54 served populations of 3,301 to 10,000 (medium), 233 served populations
of 501 to 3,300 (small), and 674 served populations of 25 to 500 (very small).

The analytical results of the NIRS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water. Additional information regarding the location and sources of
supplies, population served, methods, and detection limits was available in an EPA Status Report, also
provided by the Technical Support Division (USEPA, 1985, as cited in Wade Miller, 1989).

Samples were collected in the field by representatives of local utilities. Samplers were provided with
complete kits including detailed instructions, and a rigorous quality control procedure was developed to
ensure the representativeness of the samples. The samples collected for metals analyses were preserved
in the field with nitric acid and shipped to one of several EPA laboratories for testing.

Of the 989 utilities  for which thallium data were reported, only one returned a sample that revealed a
thallium concentration above the minimum reporting level of 8 |ig/L. That positive sample was collected
at a very small utility and was reported to contain 10 |ig/L thallium.

2.5.5.2 Ground Water and Surface Water Sources - STORET

The EPA computerized water quality data base, known as STORET, was devised to help Federal and
State institutions meet the objectives of Public Law 92-500 to maintain and enhance the physical,
chemical,  and biological quality of the nation's ambient waterways by providing for the collection and
dissemination of basic water quality data (Staples et al., 1985, as cited in Wade Miller, 1989).  Data are
collected by States, EPA regional offices, and other government agencies and maintained in the STORET
system.  STORET contains approximately 80 million pieces of data, including data for drinking water
from ground water  and surface water sources.
Occurrence Summary and Use Support Document         105                                     March 2002

-------
Before presenting a summary of the drinking water data in STORET, it is important to note that there are
significant limitations in using the data base to estimate representative concentrations of a contaminant
such as thallium.  Data entered into STORET are gathered from an array of studies conducted for various
purposes.  Analyses are conducted in a number of different laboratories employing different
methodologies with a range of detection limits. In many cases, detection limits are not reported, making
the reliability of the data highly questionable.  Where detection limits have been reported, STORET
assigns the detection limit value to those observations reported as not detected. This can lead to errors in
interpretation and overestimation of concentrations in cases in which there is a preponderance of
nondetectable values. Additionally, a few high values can inflate mean values and result in large
standard deviations relative to the means (Staples et al., 1985, as cited in Wade Miller, 1989). Very high
values may not be correct, as they may reflect sample contamination or analytical error, and can
significantly distort assessment of average concentrations. Staples et. al. (1985,  as cited in Wade Miller,
1989) also notes that the use of data collected prior to the 1980s is not recommended, since such data was
obtained using less sensitive laboratory techniques and quality assurance procedures were not yet
mandated  for the data entered into the system.

With these limitations in mind, a  summary of STORET's most recently obtained data for drinking water
from ground water and surface water sources is presented here (USEPA, 1988, as cited in Wade Miller,
1989). For ground water, there were two positive observations for total thallium  from February 1978 to
April 1979, with both values reported at 0.8 |ig/L. There were 221 samples reported as undetected; the
apparent detection limit for these  samples was 50 |ig/L. Including the undetected samples and detections
known to be less than the reported value, there were 227 observations for total thallium in ground water
from February 1978 to November 1987. With detection limit value assigned to the undetected samples,
STORET reported an overall mean value of 48.7 |ig/L and a range of 0.2 to 50 |ig/L.  The standard
deviation for all observations was 7.9 |ig/L.  Detection limits and other sampling information were not
reported.

For surface water, there were 11 positive observations for total thallium from April 1978 to September
1981, with an overall mean value of 3.4 |ig/L and a range of 0.2 to 9.0 |ig/L. The standard deviation for
these observations was 2.9 |ig/L.  Seven samples were reported as undetected, with a mean value of 30.3
l-ig/L, apparently reflecting a detection limit value range of 20 to 50 |ig/L.  Including the undetected
samples and the detections known to be less than the reported value, there were 54 observations for total
thallium in surface water from February 1978 to July 1987, with an overall mean value of 5.4 |ig/L and a
range of 0.2 to 50 |ig/L.  The standard deviation for all observations was 12.9 |ig/L. Detection limits and
other sampling information were  not reported.

2.5.6 Conclusion

Thallium and some of its compounds are naturally occurring and found at low levels as natural  deposits.
Thallium is no longer produced domestically, and all thallium used in the United States is imported.
Most thallium is used in production of specialized electronic research equipment. Some statistics
regarding  import for consumption of thallium indicate that imports may be declining. Industrial releases
of thallium have occurred since 1989 in 10 States, while releases of thallium compounds have occurred
since 1988 in 20 States.  Air emissions and off-site releases constitute the greatest proportion of the total
on- and off-site releases of thallium and thallium compounds. No national NAWQA data were available
for thallium. The Stage 2 analysis, based on the 16-State cross-section, estimated that approximately
0.283% of combined ground water and surface water systems serving 0.07439%  of the population had
estimated mean concentrations of thallium greater than the MCL of 0.002 mg/L.  Based on this estimate,
approximately 184 PWSs nationally serving about 158,300 people are expected to have estimated mean
concentrations of thallium greater than 0.002 mg/L.

Occurrence Summary and Use Support Document          106                                      March 2002

-------
Since thallium and its compounds occur naturally occurring, the balanced geographic distribution of the
16-State cross-section should adequately cover the range of natural occurrence of thallium from low to
high. The 16-State cross-section also contains a substantial proportion of the States with reported TRI
releases of thallium and thallium compounds.  Based on this use and release evaluation, the 16-State
cross-section appears to adequately represent thallium occurrence nationally.

2.5.7 References

Agency for Toxic Substances and Disease Registry (ATSDR).  1992.  Toxicological Profile for Thallium.
       U.S. Department of Health and Human Services, Public Health Service.  90 pp. + Appendices.
       Available on the Internet at: http://atsdrl.atsdr.cdc.gov/toxprofiles/tp54.pdf

Agency for Toxic Substances and Disease Registry (ATSDR).  1995.  ToxFAQs for Thallium.  U.S.
       Department of Health and Human Services, Public Health Service.  Available on the Internet at:
       http://atsdr.cdc.gov/tfacts54.html

Center for Disease Control and Prevention (CDC). 1992. Fluoridation Census 1992. Atlanta, GA:
       Centers for Disease Control and Prevention, Public Health Service, U.S.  Department of Health
       and Human Services.

Center for Disease Control and Prevention (CDC). 2002. Written comments (including information
       from the unpublished Fluoridation Census 2000) from CDC staff to OGWDW staff, January  8,
       2002.

Durum, W.H., and J. Haffty.  1961.  Occurrence of minor elements in water. U.S. Geological Survey
       Circular 445, U.S. Geological Survey, U.S. Department of the Interior, as cited in USEPA. 1984.

Hazardous Substances Data Bank (HSDB).  2000. Search for Thallium. Available on the Internet
       through TOXNET, sponsored by the National Institute of Health's National Library of Medicine.
       Available on the Internet at: http://toxnet.nlm.nih.gov/cgi-bin/sis/htmlgen7HSDB,

Staples, C.A., A.F. Werner, and T.J. Hoogheem.  1985. Assessment of priority pollutant concentrations
       in the United States using STORET database. In: Environmental Toxicology and Chemistry,  v.
       4, pp. 131-142. New York, New York: Pergamon Press, Ltd.

USEPA.  1984.  Thallium in water: An assessment of occurrence and exposure. First Draft. Prepared
       by: V.J. Ciccone & Associates, Inc., Woodbridge, VA.  Prepared for: Science and Technology
       Branch, Criteria and Standards Division, Office of Drinking Water, USEPA, Washington, DC.
       September 28, 1984.

USEPA.  1985.  Status report: National Inorganics and Radionuclides Survey.  By Jon P. Longtin.
       Cincinnati, OH: Technical Support Division, Office of Drinking Water, USEPA. August 30,
       1985.

USEPA.  1988.  Computer printout of STORET water quality data base retrieval  conducted March 23,
       1988 by Science Applications International Corporation. Data available through the Office of
       Water Regulations and Standards, USEPA.

USEPA.  2000.  TRI Explorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

Occurrence Summary and Use Support Document          107                                    March 2002

-------
USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on: Thallium.
       Office of Ground Water and Drinking Water, USEPA.  Available on the Internet at:
       http://www.epa.gov/OGWDW/dwh/c-ioc/thallium.html, last updated March 9, 2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Wade Miller Associates, Inc.  1989. Occurrence and Exposure Assessment of Thallium in Public
       Drinking Water Supplies. Prepared for and submitted to EPA on February 24, 1989.
Occurrence Summary and Use Support Document         108                                    March 2002

-------
3.0 SYNTHETIC ORGANIC CONTAMINANTS
Occurrence Summary and Use Support Document       109                         March 2002

-------
3.1    Alachlor
Table of Contents

3.1.1  Introduction, Use and Production  	  Ill
3.1.2  Environmental Release  	  112
3.1.3  Ambient Occurrence 	  113
3.1.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  116
3.1.5  Additional Drinking Water Occurrence Data  	  119
3.1.6  Conclusion	  124
3.1.7  References  	  125
Tables and Figures

Figure 3.1-1:  Estimated Annual Agricultural Use for Alachlor (1992) 	  112

Table 3.1-1: Environmental Releases (in pounds) for Alachlor in the United States, 1995-1999 ....  113

Table 3.1-2: Alachlor Detections and Concentrations in Surface Water and Ground Water	  113

Table 3.1-3: Stage 1 Alachlor Occurrence Based on 16-State Cross-Section - Systems	  116

Table 3.1-4: Stage 1 Alachlor Occurrence Based on 16-State Cross-Section - Population Served  ...  117

Table 3.1-5: Stage 2 Estimated Alachlor Occurrence Based on  16-State Cross-Section - Systems  . .  118

Table 3.1-6: Stage 2 Estimated Alachlor Occurrence Based on  16-State Cross-Section -
       Population	  118

Table 3.1-7: Estimated National Alachlor Occurrence - Systems and Population Served	  119

Table 3.1-8. Summary Alachlor Monitoring in Ground Water Conducted by the ARP  	  121

Table 3.1-9. Summary of 1995 and 1996 Alachlor Monitoring in Surface Water Conducted
       by the ARP 	  121

Table 3.1-10. Summary of Wells with Detections of Alachlor from Various Studies Presented
       in the Alachlor RED	  122

Table 3.1-11. Summary of Major Surface Water Sources with Detections of Alachlor from
       Various Studies Presented in the Alachlor RED  	  122
Occurrence Summary and Use Support Document          110                                     March 2002

-------
3.1.1  Introduction, Use and Production

Alachlor [2-chloro-2', 6'-diethyl-N-(methoxymethyl) acetanilide] is a pre-emergent amide herbicide. It is
an odorless, white crystalline solid. Alachlor is found in a variety of commercial herbicides including
Lasso, Lariat, and Crop Star. It also mixes well with other herbicides such as Bullet, Freedom, and Rasta
and in mixed formulations with atrazine, glyphosate, trifluralin and imazaquin (EXTOXNET, 2001).

The greatest use of alachlor is as a herbicide for control of annual grasses and broadleaf weeds in crops,
primarily on corn, sorghum and soybeans. Alachlor is the second most widely used herbicide in the U.S.,
with particularly heavy use on corn and soybeans in Illinois, Indiana, Iowa, Minnesota, Nebraska, Ohio,
and Wisconsin (USEPA, 2001).  It is a selective systemic herbicide, absorbed by germinating shoots and
by roots. The chemical works by interfering with a plant's ability to produce protein and by interfering
with root elongation (EXTOXNET, 2001).

Recent national estimates of agricultural use for alachlor are available. The United States Geological
Survey (USGS, 1998a) estimates approximately 25.65 million pounds of alachlor active ingredient were
used in 1992. These estimates were derived using State-level data sets on pesticide use rates available
from the National Center for Food and Agricultural Policy (NCFAP) combined with county-level data on
harvested crop acreage from the  Census of Agriculture (Thelin and Gianessi, 2000).  Corn accounts for
the majority of usage (13.9 million pounds alachlor ai), while moderate use can be found on several other
crops  as well (e.g., soybeans, sorghum, beans). According to EPA estimates, from 1993-1995 annual
alachlor use ranged between 29.3 and 44.6 million pounds, with 20.5 to 28.0 million acres treated
(number of acres times number of times treated) (USEPA, 1998). Although no production data is
available, Monsanto in St. Louis, MO is the basic manufacturer of alachlor (EXTOXNET, 2001).

Figure 3.1-1 shows the USGS (1998a) derived geographic distribution of estimated average annual
alachlor use in the United States for 1992. A breakdown of use by crop is also included.  The greatest
amount of alachlor is used in corn production, with the largest concentration of alachlor use seen in the
Midwest, where corn and soybeans are plentiful. While non-agricultural uses are not reflected here and
any sharp spatial differences in use within a county are not well represented (USGS,  1998c), existing
data suggest that non-agricultural use  of alachlor is minimal to non-existent (USEPA, 1998). A
comparison of this use map with the map of the 16 cross-section States (Figure  1.3-1) shows that States
across the range of high of low alachlor use are well represented in the cross-section.
Occurrence Summary and Use Support Document         111                                     March 2002

-------
Figure 3.1-1: Estimated Annual Agricultural Use for Alachlor (1992)
                                         ALACHLOR
                                 ESTIMATED ANNUAL AGRICULTURAL USE
                    Average use of
                   Active Ingredient
                  Pounds per square mile
                   of county per year
                   D No Estimated Use
                   D  < 0.415
                   D 0.415-2.456
                   D Z.457-B.BBS
                   D 0.989 - 29.195
                   •  >= 29.198
Crops
com
soybeans
sorghum
sweet own
dry beans
peanuts
oottan
sunflower
Total
Pounds Applied
13,902,747
8,852,899
1,829,517
507,232
336,353
122,871
81,870
23,734
Percent
National Use
54.21
34.56
7.13
1.98
1.31
0. IB
0.24
0.09
Source: USGS 1998a
3.1.2 Environmental Release

Alachlor is listed as a Toxics Release Inventory (TRI) chemical. Table 3.1-1 illustrates the
environmental releases for alachlor from 1995 to  1999.  (There are only alachlor data for these years.)
Air emissions constitute most of the on-site releases, with a decrease since 1997.  Total on- and off-site
releases first increased from 1995 to 1998 and then dropped off dramatically in 1999, as both air
emissions and off-site releases (including metals or metal compounds transferred off-site) decreased in
1999.  Surface water discharges have remained relatively stable over the five-year period. No
underground injection or releases to land (such as spills or leaks within the boundaries  of the reporting
facility) were reported for alachlor. These  TRI data for alachlor were reported from five States: Illinois,
Iowa, Nebraska, Ohio, and Texas, with only Iowa reporting every year (USEPA,  2000).  Illinois,
Nebraska, and Texas are included in the 16-State cross-section (used for analyses of alachlor occurrence
in drinking water; see Section 3.1.4).  (For a map of the 16-State cross-section, see Figure 1.3-1.)
Occurrence Summary and Use Support Document
112
March 2002

-------
Table 3.1-1: Environmental Releases (in pounds) for Alachlor in the United States, 1995-1999
Year
1999
1998
1997
1996
1995
On-Site Releases
Air Emissions
755
1,510
2,700
2,340
756
Surface Water
Discharges
390
220
290
330
280
Underground
Injection
—
-
-
-
—
Releases
to Land
—
—
—
—
—
Off-Site Releases
1,270
9,100
3,600
4,100
2,940
Total On- &
Off-site
Releases
2,415
10,830
6,590
6,770
3,976
 Source: USEPA, 2000
3.1.3 Ambient Occurrence

Alachlor is an analyte for both surface and ground water NAWQA studies, with a method detection limit
(MDL) of 0.002 |ig/L.  Additional information on analytical methods used in the NAWQA study units,
including method detection limits, are described by Gilliom and others (1998).

Alachlor concentrations at all of the ground and surface water sites exceed the detection limit, with the
exception of urban ground water sites.  High percentages of surface water samples exceed detection
frequencies of 0.01 and 0.05 |ig/L. Within surface water sites, urban sites have consistently lower
percentage exceeding the detection frequencies then agricultural or integrator sites. Within ground water
sites, agricultural sites have a slightly higher percentage exceeding the detection frequencies than urban
or integrator sites. The maximum concentration exceeds the detection limit in all ground and  surface
water sites. The 95th percentile values exceed the detection limit for all surface water sites, while none of
the 95th percentile values exceed the detection limit for any ground water sites. None of the median
values exceed the detection limit in ground or surface water sites.
Table 3.1-2: Alachlor Detections and Concentrations in Surface Water and Ground Water
                               Detection frequency
                                 (% of samples)
                    Concentration percentiles
                      (all samples; Jlg/L)
                      all samples    >0.01 llg/L    > 0.05 llg/L
                       median
95"
surface water
agricultural
urban
integrator
all sites

36.36%
13.46%
39.02%
30.58%

27.37%
8.87%
26.02%
22.62%

11.49%
3.67%
10.16%
9.91%


-------
                              Detection frequency                    Concentration percentiles
                                 (% of samples)                         (all samples; Jlg/L)
ground -water
agricultural
urban
major aquifers
all sites

3.14%
0.33%
1.61%
3.04%

1.49%
ND
0.86%
1.95%

0.54%
ND
0.54%
1.03%


-------
wells in areas where alachlor was extensively used.  Of the 246 wells sampled, 10 wells had detectable
levels of alachlor. The highest level reported was 5.8 |ig/L. The remaining nine wells had levels below 1
l-ig/L. The detection limit was 0.1 |ig/L.

3.1.3.2.2 Surface Water Sources

Baker et al. (1981, as cited in USEPA, 1989) presented data from a study that examined concentrations of
alachlor in stream water in Ohio. A total of 292 samples were collected from 12 different streams during
the spring and summer of 1981 and analyzed for alachlor. The analysis identified 235 (80%) positive
samples. The maximum concentration observed was 104.6 |ig/L. No detection limit for the analysis was
given.

Baker (1983, as cited in USEPA, 1989) reported on levels of alachlor in water samples collected from
two Ohio rivers. Between May 28, 1983, and July 27, 1983, a total of 46 samples were collected (23
samples from each river). The average alachlor concentrations for samples collected and analyzed from
each of the rivers were 1.24 |ig/L and 3.11 |ig/L, respectively. The number of positive values and the
detection limit for alachlor were not reported.

Another study examining surface waters in Ohio was reviewed for alachlor occurrence data. Datta (no
date, as cited in USEPA,  1989) reported analyses of a creek in southwest Ohio during 1981, and again in
1982. Mean concentrations of alachlor for the two years were 13.9 |ig/L and 7.6 |ig/L, respectively.  No
other information was reported. The same source (Datta, no date, as cited in USEPA, 1989) also  reported
that for 5 northwest Ohio rivers, 233 samples were found to have mean peak concentrations of 23.2 |ig/L
(maximum concentration = 69.6 M-g/L). The number of positives and the detection limit were not
reported.

River samples from the Little Sioux River in northwest Iowa and Big Spring Basin, Iowa, were analyzed
for alachlor by the Iowa Department of Water, Air, and Waste Management and as part of the Review of
Hydrogeology, Water Quality, and Land Management in the Big Spring Basin, respectively (Kelley and
Wnuk, 1986; Datta, no date, all as cited in USEPA, 1989).  During the overall study period from  1981 to
1985, 18 samples from 5 locations were taken, and 10 proved positive for alachlor. These positive
samples represent a mean of 2.59 |ig/l (range = 0.06-20.0 M-g/L). No detection limit was given. A
reservoir on the Des Moines River was sampled for alachlor during 1977-1978 by Leung et al. (1982, as
cited in USEPA, 1989). Three sites were sampled (upstream, within, and downstream of the reservoir).
A mean concentration of 0.089 |ig/L (range = 0 to 0.82 |ig/L) was reported for positive samples.  The
number of positives, number of samples, and detection limit were not reported.

Twenty-five samples from River Raisin, Michigan, were taken at U.S. Geological Survey stations during
1982 (Datta, no date, as cited in USEPA, 1989). A maximum concentration of 8.16 |ig/L was reported
for alachlor. No other information was reported.

Dudley and Karr (1980, as cited in USEPA, 1989) analyzed levels of atrazine and alachlor in 45 samples
of water, sediment, and fish. The samples were collected in mid-July from a stream draining an
agricultural watershed (Black Creek) in Allen County, Indiana. No samples of water showed levels of
alachlor in excess of the detection limit of 100 |ig/l. The exact number of water samples was not
reported.

Schepers et al. (1980, as cited in USEPA, 1989) assessed concentrations of alachlor in 30 samples of
water collected from a watershed in Nebraska.  Concentrations of alachlor in the 30 samples ranged
Occurrence Summary and Use Support Document          115                                     March 2002

-------
between "non-detectable" and 1.41 |ig/L. However, the period of collection, detection limit, and number
of positive samples for alachlor were not reported.

Lake Erie water samples were collected during a study by Konasevich et al. (1978, as cited in USEPA,
1989). Three samples were positive for alachlor at a mean concentration of 3.05 |ig/L (range = 0.07 to
9.0 |ig/L). The number of samples and detection limit were not reported.

3.1.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of alachlor occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States. The  16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to two concentration thresholds of interest: 0.002 mg/L;  and 0.0002 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
alachlor.  These data represent more than 58,000 analytical results from approximately 14,000 PWSs
during the period from 1984 to 1998 (with most analytical results from 1992 to 1997). The number of
sample results and PWSs vary by State, although the State data sets have been reviewed and checked to
ensure adequacy of coverage and completeness. The overall modal detection limit for alachlor in the 16
cross-section States is equal to 0.0002 mg/L. (For details regarding the  16-State cross-section, please
refer to Section 1.3.5 of this report.)

3.1.4.1 Stage 1 Analysis Occurrence Findings

Table 3.1-3 illustrates the Stage 1 analysis of alachlor occurrence in drinking water for the public water
systems in the 16-State cross-section relative to two thresholds: 0.002 mg/L (the current MCL), and
0.0002 mg/L (the modal MRL).  Approximately 0.0419% of (a total of 6 systems) all ground water and
surface water PWSs had any analytical results of alachlor exceeding the MCL. Approximately 0.488%
(70 systems) of PWSs had any exceedances of the modal MRL.

A greater proportion of surface water systems, as compared to ground water systems, exceeded each
threshold. Only 1 (approximately 0.00774% of) ground water PWSs had any analytical results exceeding
the MCL, compared to 5 (about 0.354% of) surface water systems.  About 0.132% (17 systems) of
ground water PWSs had any analytical results exceeding  0.0002 mg/L.  This compares to about 3.75%
(53 systems) of surface water systems with any analytical results greater than the modal MRL.
Table 3.1-3:  Stage 1 Alachlor Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.0002
Percent of Systems
Exceeding Threshold
0.00774%
0.132%
Number of Systems
Exceeding Threshold
1
17

Surface Water
0.002
0.0002
0.354%
3.75%
5
53

Occurrence Summary and Use Support Document
116
March 2002

-------
Source Water Type
Combined Ground &
Surface Water
Threshold
(mg/L)
0.002
0.0002
Percent of Systems
Exceeding Threshold
0.0419%
0.488%
Number of Systems
Exceeding Threshold
6
70
Reviewing alachlor occurrence in the 16 cross-section States by PWS population served (Table 3.1-4)
shows approximately 0.259% of the population (a total of 247,300 people) was served by PWSs with any
analytical results greater than the MCL.  The percentage of population served by PWSs with any
exceedance of the modal MRL was 1.19% (over 1.1 million people).

The number of people served by ground water systems with any analytical results greater than 0.002
mg/L was equal to 700 (about 0.00171%). A total of 246,700 (approximately 0.445% of) people were
served by surface water systems with any analytical results greater than 0.002 mg/L.  About 0.249% of
the population served by ground water PWSs (100,200 people) had any exceedances of 0.0002 mg/L,
compared to approximately 1.87% of the population served by surface water systems (1,036,700 people).
Table 3.1-4:  Stage 1 Alachlor Occurrence Based on 16-State Cross-Section - Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.0002
Percent of Population
Served by Systems
Exceeding Threshold
0.00171%
0.249%
Total Population Served bj
Systems Exceeding
Threshold
700
100,200

Surface Water
0.002
0.0002
0.445%
1.87%
246,700
1,036,700

Combined Ground &
Surface Water
0.002
0.0002
0.259%
1.19%
247,300
1,136,900
3.1.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.1-5 and 3.1-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

No ground water or surface water PWSs had an estimated mean concentration of alachlor exceeding
0.002 mg/L (the current MCL). Approximately 8 (0.0559% of) ground water and surface water PWSs in
Occurrence Summary and Use Support Document
111
March 2002

-------
the 16-States were estimated to have mean concentrations of alachlor above 0.0002 mg/L (the modal
detection limit).

A greater proportion of surface water systems, as compared to ground water systems, were estimated to
exceed 0.0002 mg/L. Approximately 3 (0.0266% of) ground water systems in the 16 States had
estimated mean concentrations of alachlor above 0.0002 mg/L, compared to approximately 5 (0.323% of)
surface water systems. As stated above, no ground water or surface water PWSs had an estimated mean
concentration of alachlor exceeding 0.002 mg/L.
Table 3.1-5:  Stage 2 Estimated Alachlor Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.0002
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.0266%
Range
0.000% - 0.000%
0.0077% - 0.0542%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
3
Range
0-0
1-7

Surface Water
0.002
0.0002
0.000%
0.323%
0.000% - 0.000%
0.0708% -0.566%
0
5
0-0
1 -8

Combined Ground
& Surface Water
0.002
0.0002
0.000%
0.0559%
0.000% - 0.000%
0.0279% - 0.0907%
0
8
0-0
4-13
Reviewing alachlor occurrence by PWS population served (Table 3.1-6) shows that approximately
68,500 (0.0716% of) the PWS population in the 16 States were served by systems with mean alachlor
concentrations above 0.0002 mg/L. When evaluated relative to a threshold of 0.002 mg/L, the percent of
population exposed was equal to 0%.

For ground water systems, about 15,500 people nationally (0.0385% of the population served by ground
water systems in the 16 cross-section States) were exposed to alachlor levels above 0.0002 mg/L.
Approximately 0.0956% of the population served by surface water PWSs (about 53,000 people in the 16-
State cross-section) had estimated mean concentrations of alachlor above 0.0002 mg/L.

Table 3.1-6: Stage 2 Estimated Alachlor Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.0002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.0385%
Range
0.000% - 0.000%
0.00383% -0.1 12%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
15,500
Range
0-0
1,500-45,200

Occurrence Summary and Use Support Document
118
March 2002

-------
Source Water Type
Surface Water
Threshold
(mg/L)
0.002
0.0002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.0956%
Range
0.000% - 0.000%
0.00250% - 0.408%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
53,000
Range
0-0
1,400-226,200

Combined Ground
& Surface Water
0.002
0.0002
0.000%
0.0716%
0.000% - 0.000%
0.0104% -0.2576%
0
68,500
0-0
9,900 - 246,500
3.1.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, no PWSs (therefore, no population served by systems) had estimated mean concentrations of
alachlor above 0.002 mg/L.  Approximately 36 systems serving about 152,500 people served nationally
were estimated to have alachlor concentrations greater than 0.0002 mg/L.  An estimated 16 ground water
systems serving about 33,000 people nationally had mean concentration values that exceeded 0.0002
mg/L. Approximately 18 surface water systems serving 121,700 people had estimated mean
concentrations greater than 0.0002 mg/L. (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)
Table 3.1-7:  Estimated National Alachlor Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.0002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
16
Range
0-0
5-32
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
33,000
Range
0-0
3,300 - 96,200

Surface Water
0.002
0.0002
0
18
0-0
4-32
0
121,700
0-0
3,200-519,500

Combined Ground
& Surface Water
0.002
0.0002
0
36
0-0
18-59
0
152,500
0-0
22,100 - 548,700
3.1.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of alachlor in drinking water are also reviewed.
Previously compiled alachlor occurrence information is presented in following section.  Information from
the EPA's Office of Pesticide Program's (OPP's) "Reregistration Eligibility Decision (RED):  Alachlor"
Occurrence Summary and Use Support Document
119
March 2002

-------
(USEPA, 1988) is presented in Section 3.1.5.1, and information from an OGWDW summary document
entitled "Occurrence and Human Exposure to Pesticides in Drinking Water, Food, and Air in the United
States of America" (USEPA, 1989) is presented in Sections 3.1.5.2 - 3.1.5.5.  The following variety of
studies and information is presented regarding levels of alachlor in drinking water, with the scope of the
reviewed studies ranging from national to regional. Note that none of the studies presented in the
following section provide the quantitative analytical results or comprehensive coverage that would
enable direct comparison to the occurrence findings estimated with the cross-section occurrence data
presented in Section 3.1.4. These additional studies, however, do enable a broader assessment of the
Stage 2 occurrence estimates presented for this Six-Year Review.  The information in Section 3.1.5.1 is
taken directly from "Reregistration Eligibility Decision (RED): Alachlor" (USEPA, 1988). All the
information in Sections 3.1.5.2 - 3.1.5.5 is taken directly from "Occurrence and Human Exposure to
Pesticides in Drinking Water, Food, and Air in the United States of America" (USEPA, 1989).

3.1.5.1  Summary of Alachlor Monitoring Information from OPP's Alachlor Reregistration

The Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA)  states that "the Administrator shall
determine whether pesticides containing such active ingredient are eligible for reregistration," before
either reregistering products or taking "other appropriate regulatory action." Thus, reregistration
involves a thorough review of the scientific data base underlying a pesticide's registration. The purpose
of the Agency's review is to reassess the potential hazards arising from the currently registered uses of
the pesticide; to determine the need for additional data on health and environmental effects; and to
determine whether the pesticide meets the "no unreasonable adverse effects" criterion of FIFRA.

In 1988, OPP published  a "Reregistration Eligibility Decision" (RED) for alachlor (USEPA,  1988).
Several sources of information on monitoring/detections of alachlor in ground water and surface water
were reviewed in the RED, and are summarized below.  However, because most of these studies were
conducted in predominantly high atrazine use areas, the occurrence findings cannot be directly compared
to the Stage 2 national occurrence estimates  for alachlor.

The most extensive review of alachlor occurrence data included in the RED was collected through the
Acetochlor Registration Partnership (ARP) Monitoring Program. This study included monitoring for
alachlor as part of Monsanto's conditional registration monitoring program for Acetochlor. Since 1995,
wells and surface waters in the major corn growing regions of the U.S. have been monitored.

The ARP Ground Water Monitoring Program involved the creation of a unique network of 175 ground
water monitoring wells.  These wells are of great relevance to U.S. agriculture because they represent the
range of soil and agriculture in seven key Midwest states (Illinois, Indiana, Iowa, Kansas, Minnesota,
Nebraska, and Wisconsin) and are positioned directly next to treated corn fields. As a requirement for
the registration of acetochlor, the two acetochlor registrants are conducting  a ground water monitoring
program in seven major use states.  Analytes include alachlor, acetochlor, and atrazine, dimethenamid,
and metolachlor. Ground water samples are collected monthly from 175 wells located in corn producing
areas.  The  limits of detection and quantification for all analytes are 0.03 |ig/L and 0.05 |ig/L,
respectively.

Results for alachlor are summarized in Table 3.1-8.  Fourteen of the wells had alachlor detects greater
than limit of quantification (LOQ - 0.05 |ig/L). Twenty-seven wells had alachlor detections above the
limit of detection (LOD) of 0.03 |ig/L. Approximately  36% of the alachlor detections exceeded the MCL
(2.0 |ig/L) and 54% exceeded a one-in-million cancer risk level of 0.4 |ig/L. Approximately  75% of the
detections exceeded 0.11 |ig/L.
Occurrence Summary and Use Support Document          120                                      March 2002

-------
Table 3.1-8. Summary Alachlor Monitoring in Ground Water Conducted by the ARP
Statistic
Number of Samples with Detects >0.05 |_ig/L (% of samples)
Number of Wells (%of!73)
Number of Samples1
Mean
Minimum
Median
Maximum
Detection/Concentration
Qlg/L)
30(1.7%)
14(8.1%)
1,720
3.38
0.05
0.73
15.89
The Acetochlor Registration Partnership also provides the most extensive data on alachlor concentrations
in finish surface drinking water available to OPP. Samples were collected at 179 different sites (drinking
water utilities) in the following 12 states: DE, IL, IN, IA, KS, MD, MN, MO, NE, OH, PA, and WI.
Samples were collected approximately once every two weeks from April through early September. Two
to three additional samples were collected at most sites, one to two in the fall and one in the winter.
Unfiltered samples were analyzed for total alachlor.

Additional detail is provided in a review and analysis of the data by R. David Jones (EFED memo dated
May 24, 1996 to B. Montague) and in an electronic spreadsheet (summary statistics) by S. Abel. These
analyses provide yearly summary statistics for peak, 96-hour time weighted mean concentrations
(TWMC), annual TWMC for 1995 (R. David Jones), and peak  and annual TWMC for 1996 (S. Abel).
Additionally, running peak and annual TWMC are provided for both years data. Based on the re-analysis
of data from 1995 through 1996, Table 3.1-9 provides maximum and 90th percentile (upper 10th
percentile) concentrations (peak and annual TWMC).
Table 3.1-9. Summary of 1995 and 1996 Alachlor Monitoring in Surface Water Conducted by the
ARP
Statistic
Peak
Concentrations
Annual Time
Weighted Mean
Calculation for a given site
Highest Value Observed 1995-1996,
for any Site
Weighted Mean for 1995-1996,
for any Site (weight by time)
Summary Across Sites
Maximum Value for any Site (
Jig/1)
4
0.36
Value Equaled/Exceeded
onlO%ofSites(u.g/l)
0.63
0.1
In addition to the ARP data, several other studies of alachlor monitoring data were summarized in the
Alachlor RED.  Table 3.1-10 presents a summary of additional ground water occurrence findings and
Table 3.1-11 summarizes the major surface water studies contained in that report.
Occurrence Summary and Use Support Document
121
March 2002

-------
Table 3.1-10. Summary of Wells with Detections of Alachlor from Various Studies Presented in
the Alachlor RED
Study1
USGS (1991 -1994)
ARP-GWMP
PGWDB
NPS
NAWWS
Well type
drinking
monitoring
mixed, most drinking
drinking
drinking
Number of Wells (%)
Sampled
303
173
25,933
1300
1430
Alachlor
Detected
10 (3.3%)
27(15.6%)
467(1.8%)
1(<0.1%)
28 (2.0%)
Concentration

-------
3.1.5.3 Ground Water Sources - Regional Studies

Drinking water wells from eight counties in Maryland were examined by the State of Maryland's Office
of Environmental Programs, Department of Health and Mental Hygiene, in the fall of 1983 (State of
Maryland, 1983, as cited in USEPA, 1989). Thirteen samples were collected from 1 1 locations and
analyzed for alachlor.  Two samples were positive at concentrations of 0.1 and 0.8 |ig/L.  The detection
limit was 0.1 |ig/L.

The Suffolk County Department of Health  Services analyzed drinking water wells in Long Island, New
York, during 1984 (Holden, 1986, as cited  in USEPA, 1989). Alachlor was not detected in any of 24
samples collected (detection limit not reported).

Drinking water wells throughout the State of Wisconsin were analyzed by Union Carbide during 1983-
1984 as part of a Wisconsin Department of Natural Resources program (Holden, 1986, as cited in
USEPA,  1989). Possible contamination may have occurred from both point and nonpoint sources. Of
the 377 samples analyzed, 47 were found positive for alachlor, with a maximum concentration of 88
l-ig/L. The mean concentration, range of values, and detection limit were not reported.

Two studies of drinking water wells in Iowa were available for information  on the occurrence of
alachlor:  one conducted by Iowa State University and the other by the Iowa Department of Water, Air,
and Waste Management (Baker and Austin, 1983: Kelley and Wnuk, 1986,  all as cited in USEPA, 1989).
Combined, the studies  included six Iowa counties sampled between 1981 and 1985. Fifty-nine samples
were analyzed from 25 sites, with all but one positive sample coming from Humboldt County. The
overall range was 0 to  2.7 |ig/L (the detection limit was 0.01 |ig/L). The one positive sample from
outside Humboldt County was 0.18 |ig/L.  The mean for Humboldt County samples, all taken at one  site,
was 0.08 |ig/L (the detection limit was 0.01 |ig/L).  The total number of positives and other detection
limits were not reported.

3.1.5.4 Surface Water Sources - National Study

The National Screening Program for Organics in Drinking Water (NSP) (Boland, 1981, as cited  in
USEPA,  1989) also contained information  on alachlor contamination in drinking water from surface
water sources.  Finished drinking water samples, collected from  104 surface water systems of varying
size throughout the United States, were analyzed for alachlor. Drinking water samples from four very
large systems (serving  greater than 100,000 individuals) were found to contain levels of alachlor in
excess of the quantification limit of 0.1 |ig/L, ranging between 0.1 and 0.9 |ig/L, with an average value of
0.38
3.1.5.5 Surface Water Sources - Regional Studies

Alachlor was monitored in known high-use areas of Illinois, Indiana, Iowa, Michigan, Missouri, North
Carolina, and Ohio (Monsanto, 1986, as cited in USEPA, 1989). Sampling occurred at 24 community
water treatment plants in these seven States, which primarily or exclusively utilize surface water
supplies. Populations of the 24 communities ranged from 356 to 388,000 and, overall, the study is
representative of over 1.3 million people using public drinking water. Both raw and finished water
samples were collected daily from each location for 1 year, and were turned into weekly composites.
Sampling took place between April 1985 and April 1986. The lower limit of method validation (LLMV)
was 0.02 |ig/L.
Occurrence Summary and Use Support Document          123                                     March 2002

-------
Alachlor was not detected in any weekly sample of finished water from 10 of the 24 plants. Of the other
14 plants, the annualized mean concentration (AMC, i.e., a time-weighted average concentration) at 12
plants was less than 0.50 |ig/L. The other two plants had AMCs of 0.69 |ig/L and  1.4 |ig/L. Rarely was
alachlor found above 2.0 |ig/L in a weekly composite: only 2.6 percent of the weekly composites
exceeded 2.0 |ig/L during the sampling year (Monsanto, 1986, as cited in USEPA, 1989).  The weekly
composite maximum concentrations ranged from <0.20 to 10.7 |ig/L for raw water samples and from
<0.20 to 10.9 |ig/L for finished water samples.  The AMCs ranged  from 0 to 1.5 |ig/L for raw water
samples, and from 0 to 1.4  |ig/L  for finished water samples. [Note: For the AMC range, the lower value
assumes that all non-detected = 0 |ig/L and the higher value assumes that all non-detected = 0.20 |ig/L.]
Individual sample results were not presented in the report.

Baker (1983, as cited in USEPA, 1989) provided the results of analysis of finished drinking water
samples collected from a water supply in Ohio from 1981-1982.  The supply obtained water from a river
draining an agricultural area. Between June 1981 and July 1982, 15 finished drinking water samples
were collected from this plant. Although no detection limits were given, 14 of the samples had positive
concentrations of alachlor that ranged between 0.03 and 14.3  |ig/L, with an average of 4.5 |ig/L.

In the study by Baker (1983, as cited in USEPA,  1989), 49 samples were collected from 3 water supplies
in Ohio and analyzed for alachlor. The supplies obtained their raw water from two rivers that drain
agricultural areas. Average concentrations for samples collected at each of the supplies between May 28
and July 27, 1983 were 1.08 (18  samples), 0.22 (15 samples), and 1.87 |ig/L (16 samples). Peak
concentrations observed in 1983 were 2.73, 0.47, and 5.91 |ig/L, respectively. Datta (no date, as cited in
USEPA, 1989) reported an overall mean concentration of 1.07 |ig/L for these supplies. The detection
limit and number of positive samples were not reported.

Finished drinking water samples from New Orleans, Louisiana were analyzed by Keith et al. (1976, as
cited in USEPA, 1989). The range of positive  samples for alachlor was 0.17 to 2.9 |ig/L.  The number of
samples analyzed, number of positive samples, mean, and detection limit were not reported.

3.1.6  Conclusion

Alachlor is a manufactured chemical that is the second most commonly used herbicide in the United
States. It is most often used for control of annual grasses and broadleaf weeds in crops, primarily on
corn, sorghum and soybeans. Recent statistics regarding use of alachlor indicate production and use are
considerable. Industrial releases of alachlor have been reported to  TRI since 1995 from five States.
Alachlor was also an analyte for the NAWQA occurrence studies.  In the NAWQA study, alachlor was
detected in ground and surface water; however, none of the median values exceeded the detection limit.
The Stage 2 analysis, based on the 16-State cross-section, estimated that zero percent of combined
ground water and surface water systems serving zero percent of the population exceeded the MCL of
0.002 mg/L. Based on this estimate, zero PWSs nationally are estimated to have levels greater than the
MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for alachlor from 5  States, including 3 of
16 cross-section States. According to information from USGS, all  of the  16 cross-section States use
alachlor, although for most States in light to moderate amounts.  Alachlor is used most heavily in the
Midwest, where there are six cross-section States. The cross-section should  adequately represent the
occurrence of alachlor on a national scale based upon the use, production, and release patterns of the 16-
State cross-section in relation to  the patterns observed for all 50  States.

Occurrence Summary and Use Support Document         124                                     March 2002

-------
3.1.7  References

Baker, D.B., K.A. Krieger, and J.V. Setzler. 1981. The Concentrations and Transport of Pesticides in
       Northwestern Ohio Rivers. Prepared by Water Quality Laboratory, Heidelberg College, Tiffin,
       Ohio, for Lake Erie Wastewater Management Study, U.S. Army Corps of Engineers, Buffalo,
       New York.

Baker, D.B.  1983. Herbicide Contamination inMunicipal Water Supplies of Northwestern Ohio.  Draft
       Final Report. Water Quality Laboratory, Heidelberg College, Tiffin, OH.

Baker, J.L., and T.A. Austin. 1983.  Impact of Agricultural Drainage Wells on Groundwater Quality.
       Completion Report 1981-1983.  Departments of Agricultural Engineering and Civil Engineering,
       Iowa State University. EPA Grant No. G007228010.

Blomquist, J.D., J.M. Denis, J.L. Cowles, J.A.  Hetrick, R.D. Jones, andN.B. Birchfield.  2001.
       Pesticides in Selected Water-Supply Reservoirs and Finished Drinking Water, 1999-2000:
       Summary of Results from a Pilot Monitoring Survey. U.S. Geological Survey Open-File Report
       01-456. 65pp.

Boland, P.A.  1981. National Screening Program for Organics in Drinking Water. Part II. Data.
       Prepared by SRI International, Menlo Park, California, for Office of Drinking Water, U.S.
       Environmental Protection Agency, Washington, DC. EPA Contract No. 68-01-4666.

Cohen, S.Z., C. Eiden, and M.N. Lorder.  1986. Evaluation  of Pesticides in Groundwater.  W.Y. Garner,
       R.C. Honeycutt, H.N. Nigg (eds.).  American Chemical Society, Washington, DC.  ACS
       Symposium Series 315.

Datta, P.R. No date. Memorandum: Review of six documents regarding monitoring of pesticides in
       Northwestern Ohio rivers. Sent to: D.J. Severn, Chief- Exposure Assessment Branch, HED.
       Washington, DC:  Office of Pesticides and Toxic Substances, USEPA.

Dudley, D.R., and J.R. Karr. 1980. Pesticides and PCB residues in the Black Creek watershed, Allen
       County, Indiana, 1977-1978. Pestic. Monit. J. v.13, no. 4, pp.155-157.

EXTOXNET.  2001. Pesticide Information Profile: Alachlor.  Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://pmep.cce.cornell.edu/profiles/extoxnet/24d-captan/alachlor-ext.html, last updated March 1,
       2001.

Gilliom, R.J., O.K. Mueller, and L.H. Nowell.  1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589, last updated October 9, 1998.

Holden, P.W.  1986. Pesticides and Groundwater Quality. Issues and Problems in Four States.
       Prepared for the Board of Agriculture, National Research  Council. Washington, DC: National
       Academy Press.
Occurrence Summary and Use Support Document          125                                     March 2002

-------
Keith, L.H., H.W. Garrison, F.R. Allen, etal. 1976. Identification of organic compounds in drinking
       water from 13 U.S. cities. In: Identification and Analysis of Organic Pollutants in Water. Ann
       Arbor, MI: Ann Arbor Sciences Publishers, Inc. pp. 329-373.

Kelley, R., and M. Wnuk. 1986. Little Synthetic Organic Compound Municipal Well Sampling Survey.
       Des Moines, IA: Iowa Department of Water, Air, and Waste,.

Konasevich, D., W. Traversy, and H. Zar. 1978. Status Report on Organic and Heavy Metal
       Contaminants in the Lakes Erie, Michigan, Huron and Superior Basins.  Great Lakes Water
       Quality Board,  p. 37.

Leung, SY.Th., R.V. Bulkley, and J.J. Richard.  1982.  Pesticide accumulation in a new impoundment in
       Iowa.  Water Res. Bull. v. 18, no. 3, pp. 485-493.

Monsanto Agricultural Company. 1986. Information to support the registration of Lasso herbicides.
       EAP Reg. Nos.  524-285, 524-296, 524-314, 524-329, 524-341, 524-344. Alachlor in raw and
       finished drinking water derived from surface sources from 24 community water systems located
       in regions of extensive Lasso use. Compiled by S.R. Muench, Monsanto Agricultural Company,
       St. Louis, MO.

Schepers, J.S., E.J. Vavricka, and G.E. Schuman.  1980. Agricultural runoff during a drought period. J.
       WPCF  v. 52, no. 4.

Spalding, R.F., G.A. Junk, and J.J. Richard.  1980.  Pestic. Monit. J.  v. 14, no. 2, pp. 70-73.

State of Maryland.  1983. Summary Report: Results of a Maryland ground water herbicide survey. Fall
       1983. Office of Environmental Programs.  Department of Health and Mental Hygiene.

Thelin, Gail P., and Leonard P. Gianessi.  2000. Method for Estimating Pesticide Use for County Areas
       of the Conterminous United States. U.S. Geological Survey Open-File Report 00-250.
62 pp. Available on the Internet at: http://water.wr.usgs.gov/pnsp/rep/ofr00250/ofr00250.pdf

USEPA.  1989. Draft Final Report on the Occurrence and Human Exposure to Pesticides in Drinking
       Water, Food, and Air in the United States of America.  Office of Drinking Water, USEPA.
       September, 1989.

USEPA.  1998. Registration Eligibility Decision (RED): Alachlor.  EPA Report  738-R-98-020.
       Washington, DC: Office of Prevention, Pesticides, and Toxic Substances, USEPA. Available on
       the Internet at: http://www.epa.gov/oppsrrdl/REDs/0063red.pdf

USEPA.  1998. R.E.D. Facts: Alachlor.  EPA Report 738-F-98-018.  Washington, DC: Office of
       Prevention, Pesticides, and Toxic Substances, USEPA.  12 pp.  Available on the Internet at:
       http://www.epa.gov/oppsrrdl/REDs/factsheets/0063fact.pdf

USEPA.  2000. TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May  5, 2000.

USEPA.  2001. National Primary Drinking  Water Regulations - Consumer Factsheet on: Alachlor.
       Office of Ground Water and Drinking Water, USEPA.  Available on the Internet at
       http://www.epa.gov/safewater/dwh/c-soc/alachlor.html, last updated April 12, 2001.

Occurrence Summary and Use Support Document          126                                    March 2002

-------
USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 1998a. Annual Use Maps.  Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated March 20, 1998.

USGS. 1998b. Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA). PROVISIONAL DATA - SUBJECT TO
       REVISION. Available on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/, last updated
       October 9, 1998.

USGS. 1998c. Sources & limitations of data used to produce maps of annual pesticide use. Reston,
       VA: United States Geological Survey. Available on the Internet at:
       http://water.wr.usgs.gov/pnsp/use92/mapex.html (last updated March 20, 1998).
Occurrence Summary and Use Support Document         127                                    March 2002

-------
3.2    Bis(2-ethylhexyl)phthalate (DEHP)
Table of Contents

3.2.1 Introduction, Use and Production  	  129
3.2.2 Environmental Release  	  130
3.2.3 Ambient Occurrence  	  131
3.2.4 Drinking Water Occurrence Based on the 16-State Cross-Section	  133
3.2.5 Additional Drinking Water Occurrence Data  	  137
3.2.6 Conclusion	  138
3.2.7 References  	  139
Tables and Figures

Table 3.2-1:  DEHP Manufacturers and Processors by State  	  129

Table 3.2-2:  Environmental Releases (in pounds) for DEHP in the United States, 1988-1999	  131

Table 3.2-3:  Stage 1 DEHP Occurrence Based on 16-State Cross-Section - Systems	  133

Table 3.2-4:  Stage 1 DEHP Occurrence Based on 16-State Cross-Section - Population	  134

Table 3.2-5:  Stage 2 Estimated DEHP Occurrence Based on 16-State Cross-Section - Systems  ....  135

Table 3.2-6:  Stage 2 Estimated DEHP Occurrence Based on 16-State Cross-Section - Population  . .  136

Table 3.2-7:  Estimated National DEHP Occurrence - Systems and Population Served  	  137
Occurrence Summary and Use Support Document         128                                    March 2002

-------
3.2.1  Introduction, Use and Production

Bis(2-ethylhexyl)phthalate (also known as di(2-ethylhexyl)phthalate, or DEHP) is the most commonly
used chemical of a group of related synthetic organic chemicals called phthalates or phthalic acid esters.
DEHP is a colorless liquid with almost no odor.  It does not evaporate easily, and little will be present in
the air even near sources of production. It dissolves more easily in materials such as gasoline, than it
does in water. Trade names for DEHP are Platinol DOP, Octoil, Silicol 150, Bisoflex 81, and Eviplast 80
(ATSDR, 2000).

DEHP is a manufactured chemical that makes plastic more flexible. It is one of several plasticizers in
polyvinyl chloride (PVC) resins used to fabricate flexible vinyl products (NSC, 2001). At least 95% of
DEHP produced is used as a plasticizer for polyvinylchloride (PVC) and other polymers including
rubber, cellulose and styrene. PVC is used in many consumer products such as wall coverings,
tablecloths, floor tiles, furniture upholstery, shower curtains, garden hoses, swimming pool liners,
rainwear, baby pants, dolls, toys, shoes, automobile upholstery and tops. A number of packaging
materials and tubing used in the production of foods and beverages are PVC contaminated with phthalic
acid esters, primarily DEHP (USEPA, 2001). These materials include packaging film and sheet,
sheathing for wire and cable, medical tubing, and blood storage bags.

Although DEHP does have uses as a non-plasticizer, they are very minor and relatively unimportant. It
can be used as a solvent in erasable ink, as an acaricide in orchards, as a component of dielectric fluids in
electrical capacitors, to detect leaks in respirators, and as an inert ingredient in pesticide products, in
cosmetics, in vacuum pump oil. DEHP can also be used in testing air filtration systems.  In some cases,
its use has been  discontinued because of concerns about health effects. Many toy manufacturers have
stopped using DEHP, and it is also no longer used for teethers and rattlers or in plastic food wrap
products (ATSDR, 2000).

DEHP is produced as part of a group of chemicals that are collectively known as phthalate esters, and, as
such, no production value for any single one of them is recorded. DEHP is included in a subgroup called
dioctyl phthalates, which also includes di-ethylhexyl phthalate, diisooctyl phthalate, and di-n-octyl
phthalate.  Collective production amounts of dioctyl phthalates are as follows:  1990, 309 million pounds;
1994, 258 million pounds; 1995, 280 million pounds; 1996, 280 million pounds; 1997, 287 million
pounds, and 1998, 285 million pounds. Production of DEHP in coming years is expected to grow at a
slower rate than the economy because of limited growth in the PVC market.  Demand for DEHP will also
continue to decrease due to health concerns (ATSDR, 2000).

Table 3.2-1 shows the number of facilities in each State that manufacture and process DEHP, the
intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 2000).
Table 3.2-1:  DEHP Manufacturers and Processors by State
State"
AL
AR
CA
CO
CT
DE
FL
Number of facilities
4
6
15
1
4
1
3
Range of maximum amounts on site in
pounds'"
1,000-99,999
100-999,999
1,000-999,999
1,000-99,999
1,000-99,999
1,000-9,999
1,000-99,999

8,11,13
8,9,12,13
2,3,8,9,12
8,9
8,9,10,12
8
8,9
Activities and uses0







Occurrence Summary and Use Support Document         129                                    March 2002

-------
State"
GA
IA
IL
IN
KS
KY
LA
MA
MD
MI
MN
MO
MS
NC
NE
NH
NJ
NV
NY
OH
OK
PA
PR
RI
SC
SD
TN
TX
UT
VA
VT
WA
WI
wv
Number of facilities
8
1
19
8
4
3
2
12
6
11
9
12
7
17
3
4
11
9
11
31
3
16
8
3
7
1
13
12
1
4
1
2
9
2
Range of maximum amounts on site in
pounds'"
1,000-99,999
10,000-99,999
100-999,999
1,000-99,999
100-99,999
100-99,999
1,000-99,999
100-999,999
100-9,999,999
100-99,999,999
1,000-99,999
1,000-999,999
1,000-999,999
100-999,999
10,000-99,999
1,000-99,999
1,000-999,999
1,000-999,999
1,000-9,999,999
1,000-999,999
1,000-999,999
1,000-9,999,999
10,000-999,999
10,000-999,999
100-999,999
1,000-9,999
1,000-9,999,999
1,000-999,999
1,000-9,999
10,000-999,999
10,000-99,999
100-9,999
1,000-999,999
1,000-9,999
Activities and uses0
8,9,10
9
2,3,7,8,9,11
7,8,9
8,9,13
8
7,8
1,6,8,9,10,11,13
1,4,8,13
8,9,12,13
7,8,9
7,8,9,13
8,9
1,2,3,5,8,9,11
8,9
8,9,13
2,3,8,9,11
8
1,3,5,7,8,9,13
2,3,7,8,9,12
7,9
1,4,8,9,12,13
8,9
8,9
8,9,12,13
9
1,3,4,5,6,7,8,9,10,11,12
2,3,8,9,10,13
2,5,8,12
8,9
8
8
8,9
8
aPost office State abbreviations used
bRange represents maximum amounts on site reported by facilities in each State
cActivities/Uses:
1. Produce                 8. Formulation component
2. Import                  9. Article component
3. On-site use/processing         10. Repackaging
4. Sale/distribution             11. Chemical processing aid
5. Byproduct                        12. Manufacturing aid
6. Impurity                 13. Ancillary/other uses
7. Reactant

Source: ATSDR, 2000 compilation O/TRI972000 data
3.2.2 Environmental Release

DEHP is listed as a Toxics Release Inventory (TRI) chemical. Table 3.2-2 illustrates the environmental
releases for DEHP from 1988 to 1999.  (There are only DEHP data for these years.) Air emissions
constitute most of the on-site releases, with a relatively steady decrease over the years. Other on-site
releases, such as surface water releases and releases to land (such as spills or leaks within the boundaries
of the reporting facility) do not display any discernable trend from 1988 to 1999. Underground injections
decreased from 1988 to!993, and have remained at zero ever since. Off-site releases (including metals or
metal compounds transferred off-site) have generally fallen over the years (with the exception of a brief
spike in 1995) from almost 4 million pounds to under 1 million pounds. The decrease in off-site releases,
as well as the decrease in air emissions, have contributed to decreases the total on- and off-site releases
of DEHP in recent years.  These TRI data for DEHP were reported from 44 States, with 33 States
reporting every year (USEPA, 2000). Of the 44 States, 14 are included in the 16-State cross-section
Occurrence Summary and Use Support Document
130
March 2002

-------
(used for analyses of DEHP occurrence in drinking water; see Section 3.2.4). (For a map of the 16-State
cross-section, see Figure 1.3-1.)
Table 3.2-2: Environmental Releases (in pounds) for DEHP in the United States, 1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
226,926
215,583
240,523
447,429
504,167
462,795
571,014
911,486
1,161,162
1,413,760
1,173,021
1,217,329
Surface Water
Discharges
2,629
669
595
363
921
967
1,248
954
3,849
2,424
2,987
2,781
Underground
Injection
0
0
0
0
0
0
0
35
370
265
600
3,091
Releases
to Land
4,685
24,184
71,009
59,612
19,705
5,308
7,192
5,357
87,625
19,551
25,937
20,748
Off-Site Releases
917,024
1,121,544
1,087,976
1,946,238
3,041,389
1,988,809
2,894,667
2,816,258
2,955,417
3,236,012
3,723,521
3,630,612
Total On- &
Off-site
Releases
1,151,264
1,361,980
1,400,103
2,453,642
3,566,182
2,457,879
3,474,121
3,734,090
4,208,423
4,672,012
4,926,066
4,874,561
 Source: USEPA, 2000b
3.2.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers.  There are few available data on the occurrence of DEHP in ambient waters of the United
States. The most comprehensive and nationally consistent data describing ambient water quality in the
United States are being produced through the  United States Geological Survey's (USGS) National Water
Quality Assessment (NAWQA) program. However, national NAWQA data, as well as NPDES and
NURP data, are currently unavailable for DEHP.

3.2.3.1 Additional Ambient Occurrence Data

A summary document entitled Occurrence and Exposure of Phthalate Esters in Public Drinking Water
Supplies" (Wade Miller, 1989), was previously prepared for past USEPA assessments  of DEHP. A
number of studies reviewed in that document provided data on concentrations of DEHP in water other
than drinking water. These studies, conducted on both regional and national levels, are summarized in
the following section. The following information is taken directly from "Occurrence and Exposure of
Phthalate Esters in Public Drinking Water Supplies" (Wade Miller, 1989).

3.2.3.1.1  Ground Water Sources

DEHP was detected in ground waters at infiltration sites of secondary effluents at Ft. Devens, MA;
Boulder, CO; Lubbock, TX; and Phoenix, AZ. The maximum reported concentration for DEHP was 1.40
M-g/L (USEPA, 1987, as cited in Wade Miller, 1989).
Occurrence Summary and Use Support Document
131
March 2002

-------
Several phthalate esters have been detected in ground waters from landfill sites nationwide. DEHP was
detected at a maximum concentration of 100.0 |ig/L at a landfill site in New Castle County, DE (USEPA,
1987, as cited in Wade Miller, 1989).

The USEPA's computerized water quality data base known as STORET was devised to assist Federal and
State institutions meet objectives of Public Law 92-500 to maintain and enhance the physical, chemical,
and biological quality of the nation's ambient waterways by providing for the collection and
dissemination of basic water quality data (Staples et al., 1985, as cited in Wade Miller, 1989).  Data are
collected by States, EPA regional offices, and other government agencies and are maintained in the
STORET system.

Before presenting a summary of the ambient water data in STORET, it  is  important to note that there are
significant limitations in using the data base to estimate representative concentrations of a contaminant
such as hexachlorobenzene. Data entered into STORET are gathered from an array of studies conducted
for various purposes. Analyses are conducted in a number of different laboratories employing different
methodologies with a range of detection limits.  In many cases, detection limits are not reported, making
the reliability of the data highly questionable. Where detection limits have been reported, STORET
assigns the detection  limit value to those observations reported as not detected. This can lead to errors in
interpretation and overestimation of concentrations in cases in which there is a preponderance of
nondetectable values. Additionally, a few high values can inflate mean values and result in large
standard deviations relative to the means  (Staples et al., 1985, as cited in Wade Miller, 1989).  Very high
values may not be correct, as they may reflect sample contamination or analytical error and can
significantly distort assessment of average concentrations.  Staples et al. (1985, as cited in Wade Miller,
1989) also notes that the use of data collected prior to the 1980s is not recommended, since such data was
obtained using less sensitive laboratory techniques and quality assurance procedures were not yet
mandated for the data entered into the system.

The STORET water quality data base provides information on the occurrence of contaminants at ambient
water stations in U.S. waterways. A summary of this information was obtained for DEHP in ambient
waters. Ambient sites include streams, lakes, ponds, wells, reservoirs, canals, estuaries, and oceans.
While the preponderance of data were collected from surface water sources, the number of samples
collected from ground water wells, relative to the total number of samples collected from all ambient
sites combined, is unspecified (Staples  et al., 1985, as cited in Wade Miller, 1989).  Staples et al. (1985,
as cited in Wade Miller, 1989) have summarized data from the 1980's only; that is, data from  1980
through 1983.  This was done based on the number of data points and the likelihood that better quality
assurance practices have been employed in more recent years.  In the absence of sophisticated statistical
analyses to eliminate improbable data, median values have been reported. The median value is sensitive
to extreme values, and reflects a measure of central tendency more accurately than the mean value in the
presence  of these extreme values (Staples et al.,  1985, as cited in Wade Miller,  1989).

The median concentration of DEHP was  10.0 |ig/L, with 24 percent of 901 observations reported as
detectable. Detection limits and other sampling information were not given for any of the contaminants.

3.2.3.1.2  Surface Water Sources

Phthalate esters have been detected in surface waters throughout the United States.  DEHP was detected
in Delaware River water two miles downstream from a Philadelphia wastewater treatment plant at a
concentration of 1.0 |ig/L (USEPA, 1987, as cited in Wade Miller, 1989). DEHP was detected in water
from Galveston Bay, TX at a mean concentration of 0.6 |ig/L, and in Mississippi River water at a
(tentatively) maximum concentration of 600.0 |ig/L (USEPA, 1987, as cited in Wade Miller, 1989).

Occurrence Summary and Use Support Document          132                                     March 2002

-------
DEHP has also been detected in U.S. industrial water basins at concentrations ranging from 1.0 to 85.0
[ig/L (USPHS, 1987, as cited in Wade Miller, 1989).

3.2.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of DEHP occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States. The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date.  These data were evaluated
relative to several concentration thresholds of interest:  0.006 mg/L; 0.003 mg/L; and 0.0006 mg/L.

Thirteen of the sixteen cross-section State data sets contained occurrence data for DEHP. (No data were
reported from New Jersey, New Mexico, or Oregon.) These data represent more than 41,000 analytical
results from approximately 9,400 PWSs during the period from 1984 to 1998 (with most analytical
results from 1992 to 1997).  The number of sample results and PWSs vary by State, although the State
data sets have been reviewed and checked to  ensure adequacy of coverage and completeness. The overall
modal detection limit for DEHP in the 16 cross-section States is equal to 0.0006 mg/L.  (For details
regarding the 16-State  cross-section, please refer to Section 1.3.5 of this report.)

3.2.4.1 Stage 1 Analysis Occurrence Findings

Table 3.2-3 illustrates the Stage 1 analysis of DEHP occurrence in drinking  water for the public water
systems in the 16-State cross-section relative to three thresholds:  0.006 mg/L (the current MCL), 0.003
mg/L, and 0.0006 mg/L (the modal MRL). A total of 207 (approximately 2.20% of) ground water and
surface water PWSs had analytical results exceeding the MCL (0.006 mg/L); 3.81% of systems (359
systems) had results exceeding 0.003 mg/L. When evaluated relative to a threshold of 0.0006 mg/L, the
percentage of systems tripled to 11.7% (1,106 systems).

Approximately 2.07%  of ground water systems (178 systems) had any analytical results greater than the
MCL (0.006 mg/L). About 3.55% of ground water systems (305 systems) had results above 0.003 mg/L.
The percentage of ground water systems with at least one result greater than  0.0006 mg/L was equal to
10.7% (921 systems).

A total of 29 (3.51% of) surface water systems had results greater than 0.006 mg/L.  Fifty-four (6.53%
of) surface water systems had at least one analytical result greater than 0.003 mg/L.  Approximately
22.4% of surface water systems (185 systems) had results exceeding 0.0006  mg/L.
Table 3.2-3:  Stage 1 DEHP Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.006
0.003
0.0006
Percent of Systems
Exceeding Threshold
2.07%
3.55%
10.7%
Number of Systems
Exceeding Threshold
178
305
921

Surface Water
0.006
0.003
3.51%
6.53%
29
54
Occurrence Summary and Use Support Document
133
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0006
Percent of Systems
Exceeding Threshold
22.4%
Number of Systems
Exceeding Threshold
185

Combined Ground &
Surface Water
0.006
0.003
0.0006
2.20%
3.81%
11.7%
207
359
1,106
Table 3.2-4 illustrates DEHP occurrence in the 16 cross-section States by PWS population served.
Approximately 3.20% of the  16-State population (over 2.5 million people) was served by PWSs with at
least one analytical result of DEHP greater than the MCL. Approximately 5.5 million people (7.06%)
were served by systems with an exceedance of 0.003 mg/L. Over 19 million people (24.6%) were served
by systems with at least one analytical result greater than 0.0006 mg/L.

The percentage of population served by ground water systems with analytical results greater than 0.006
mg/L was  equal to 3.46% (about 1.1 million people). When evaluated relative to 0.003 mg/L and 0.0006
mg/L, the percent of population exposed was equal to 8.51% (approximately 2.7 million people) and
19.5% (over 6 million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.006 mg/L was
equal to 3.01% (almost  1.4 million people). Approximately 6.05% of the population served by surface
water systems (almost 2.8 million people) were exposed to DEHP concentrations greater than 0.003
mg/L. When evaluated relative to 0.0006 mg/L, the percent of population exposed was equal to 28.2%
(over 13 million people).
Table 3.2-4:  Stage 1 DEHP Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.006
0.003
0.0006
Percent of Population
Served by Systems
Exceeding Threshold
3.46%
8.51%
19.5%
Total Population Served by
Systems Exceeding
Threshold
1,112,500
2,735,100
6,265,300

Surface Water
0.006
0.003
0.0006
3.01%
6.05%
28.2%
1,388,900
2,792,100
13,028,100

Combined Ground &
Surface Water
0.006
0.003
0.0006
3.20%
7.06%
24.6%
2,501,300
5,527,200
19,293,400
Occurrence Summary and Use Support Document
134
March 2002

-------
3.2.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.2-5 and 3.2-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

About 0.256% (approximately 24) of ground water and surface water PWSs had estimated mean
concentrations of DEHP greater than the MCL (0.006 mg/L). Approximately 89 systems (0.948%) in the
16 States had mean concentrations of DEHP exceeding 0.003 mg/L. When evaluated relative to 0.0006
mg/L, the percentage of systems exceeding the threshold increased by a factor often to 10.2% (an
estimated 957 systems).

Approximately 23 ground water systems in the  16 States (0.263%) have an estimated mean concentration
of DEHP greater than 0.006 mg/L.  An estimated 0.944% of ground water systems (about 81 systems)
had mean concentration values greater than 0.003 mg/L.  A total of 846 PWSs (9.85%) had estimated
mean concentration values of DEHP greater than 0.0006 mg/L.
Only 2 (about 0.184% of) surface water systems in the 16 States had estimated mean concentrations
greater than the MCL.  Eight (0.988%  of) surface water systems had estimated mean concentration
values greater than 0.003 mg/L. Approximately 13.3% (about 110 systems) of surface water systems in
the  16 States had estimated mean concentration values of DEHP greater than 0.0006 mg/L.
Table 3.2-5:  Stage 2 Estimated DEHP Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.006
0.003
0.0006
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.263%
0.944%
9.85%
Range
0.151% - 0.396%
0.698% - 1.23%
8.93% - 10.8%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
23
81
846
Range
13-34
60 - 106
767 - 928

Surface Water
0.006
0.003
0.0006
0.184%
0.988%
13.3%
0.000% - 0.484%
0.363% -1.69%
10. 8% -15. 7%
2
8
110
0-4
3-14
89-130

Combined Ground
& Surface Water
0.006
0.003
0.0006
0.256%
0.948%
10.2%
0.149% -0.393%
0.690% -1.24%
9. 22% -11.1%
24
89
957
14-37
65-117
868-1,043
Occurrence Summary and Use Support Document
March 2002

-------
Reviewing DEHP occurrence by PWS population served (Table 3.2-6) shows that approximately 0.119%
of population served by all PWSs in the 16 cross-section States (an estimate of approximately 93,400
people) was served by systems with estimated mean concentrations of DEHP above 0.006 mg/L. The
percentage of population served by PWSs in the 16 States with levels of DEHP above 0.003 mg/L and
0.0006 mg/L were 0.652% (an estimated 510,100 people) and 9.82% (almost 7.7 million people),
respectively.

When the percent of population served by ground water systems was evaluated relative to a threshold of
0.006 mg/L, 0.003 mg/L, and 0.0006 mg/L, the percentage of population exposed in the 16  cross-section
States was equal to 0.189% (an estimated 60,600 people), 0.676% (an estimated 217,400 people) and
7.22% (an estimated 2.3 million people), respectively.

The percentage of population served by surface water systems with levels above 0.006 mg/L was equal to
0.0712% (an estimated 32,900 people), and the percentage of population served with levels above 0.003
mg/L was  0.634% (an estimated 292,700 people nationally).  The percentage of the population served by
surface water systems with levels above 0.0006 mg/L was 11.6% (almost 5.4 million people).
Table 3.2-6:  Stage 2 Estimated DEHP Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.006
0.003
0.0006
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.189%
0.676%
7.22%
Range
0.0215% -0.463%
0.326% -1.1 9%
5.45% -9. 92%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
60,600
217,400
2,320,700
Range
6,900 - 148,700
104,600 - 382,400
1,751,000-3,187,300

Surface Water
0.006
0.003
0.0006
0.0712%
0.634%
11.6%
0.000% - 0.426%
0.0312% -3.00%
5. 81% -18.2%
32,900
292,700
5,363,700
0 - 196,600
14,400-1,386,600
2,679,600 - 8,410,200

Combined Ground
& Surface Water
0.006
0.003
0.0006
0.119%
0.652%
9.82%
0.0186% -0.363%
0.220% - 2.02%
6. 20% -13. 9%
93,400
510,100
7,686,000
14,600 - 284,500
172,200-1,580,700
4,852,600-10,851,400
3.2.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 166 PWSs serving approximately 254,100 people nationally could be exposed to
DEHP concentrations above 0.006 mg/L.  About 616 systems serving almost 1.4 million people
nationally had estimated mean concentrations greater than 0.003 mg/L. Approximately 6,607 systems
serving almost 21 million people served nationally were estimated to have DEHP concentrations greater
than 0.0006 mg/L. (See Section  1.4 for a description of how Stage 2 16-State estimates are extrapolated
to national values.)
Occurrence Summary and Use Support Document
136
March 2002

-------
For ground water systems, an estimated 156 PWSs serving about 161,500 people nationally had mean
concentrations greater than 0.006 mg/L.  Approximately 561 systems serving about 579,600 people
nationally had estimated mean concentration values that exceeded 0.003 mg/L. About 5,856 ground
water systems serving almost 6.2 million people nationally had estimated mean concentrations greater
than 0.0006 mg/L.

Approximately 10 surface water systems serving 90,600 people nationally were estimated to have mean
concentrations of DEHP above 0.006 mg/L. An estimated 55 surface water systems serving 807,500
people had estimated mean concentrations greater than 0.003 mg/L.  An estimated 745 surface water
systems serving almost  15 million people had mean concentrations greater than 0.0006 mg/L.
Table 3.2-7:  Estimated National DEHP Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.006
0.003
0.0006
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
156
561
5,856
Range
90-235
415-733
5,307 - 6,420
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
161,500
579,600
6,187,900
Range
18,400-396,400
278,900-1,019,600
4,668,800 - 8,498,800

Surface Water
0.006
0.003
0.0006
10
55
745
0-27
20-95
601 - 879
90,600
807,500
14,795,300
0 - 542,400
39,700 - 3,824,900
7,391,300-23,198,900

Combined Ground
& Surface Water
0.006
0.003
0.0006
166
616
6,607
97 - 256
449 - 808
5,993-7,199
254,100
1,387,700
20,911,000
39,700 - 774,100
468,600 - 4,300,600
13,202,200-29,522,900
3.2.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of DEHP in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Exposure of Phthalate Esters in Public Drinking Water Supplies" (Wade Miller, 1989),
is presented in following section.  This variety of regional studies and information are presented
regarding levels of DEHP in drinking water.  (No national studies were included.) Note that none of the
studies presented in the following section provide the quantitative analytical results or comprehensive
coverage that would enable direct comparison to the occurrence findings estimated with the cross-section
occurrence data presented in Section 3.2.4. These additional studies, however, do enable a broader
assessment of the Stage 2 occurrence estimates presented for this Six-Year Review.  All the following
information in Section 3.2.5 is taken directly from "Occurrence and Exposure of Phthalate Esters in
Public Drinking Water Supplies" (Wade Miller, 1989).
Occurrence Summary and Use Support Document
137
March 2002

-------
3.2.5.1 Groundwater Studies

Thirty-nine public water wells in New York State were sampled and analyzed for the presence of
phthalate esters.  DEHP was detected at a maximum concentration of 170.0 |ig/L. The frequency of
occurrence for DEHP was 92 percent (USEPA, 1987, as cited in Wade Miller, 1989).

DEHP was detected in drinking water obtained from groundwater sources in Miami, FL at a
concentration of 30.0 |ig/L.  The number of positive detections was not reported (USEPA, 1987, as cited
in Wade Miller, 1989).  DEHP was detected in five small water systems (wells with fewer than 200
connections) in Santa Cruz County, CA at concentrations ranging from 8.7 to 102.0 |ig/L (California
Department of Health Services, Unpublished, as cited in Wade Miller, 1989).

3.2.5.2 Surface Water Studies

Water from three water treatment plants in New Orleans, LA was sampled and analyzed for the presence
of phthalate esters. DEHP was detected at concentrations ranging from 0.1 to 0.46 |ig/L. The number of
positive detections was not reported (USEPA, 1987, as cited in Wade Miller, 1989). Water from a water
treatment plant in Philadelphia, PA was also sampled and analyzed for the presence of phthalate esters.
DEHP was detected at a concentration of 0.6 |ig/L. The number of positive detections was not reported
(USEPA, 1987, as cited in Wade Miller,  1989).

Drinking water from Seattle, WA; Ottumwa, IA; Philadelphia, PA: and Cincinnati, OH was sampled and
analyzed for the presence of phthalate esters. DEHP was not detected in the drinking water of any of the
four cities surveyed. The number of positive detections was not reported for any of the contaminants in
any of the cities (USEPA, 1987, as cited in Wade Miller, 1989).

3.2.5.5 Unspecified Water Sources

Tap water concentrations of phthalate esters in Cincinnati, OH: Hartford, CT; Atlanta, GA; and  St.
Louis, MO were reported in USEPA (1987). DEHP was detected in Cincinnati at a concentration  of 16.5
l-ig/L. The water sources were unspecified.

3.2.6  Conclusion

DEHP is a manufactured chemical that is part of a group of chemicals known as phthalate esters.  Over
95% of DEHP is used as a plasticizer for PVC. Recent statistics regarding production and release  of
DEHP indicate production and use are steady, but have the possibility of decline. Industrial releases of
DEHP have been reported to TRI since 1988 from 44 States. DEHP was not analyzed in any available
ambient studies.  The Stage 2 analysis, based on the 16-State cross-section, estimated that approximately
0.256% of combined ground water and surface water systems serving 0.119% of the population  exceeded
the MCL of 0.006 mg/L. Based on this estimate, approximately  166 PWSs serving  approximately
254,100 people nationally are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, DEHP is manufactured and/or processed in 40 States and has TRI  releases
in 44 States. DEHP is manufactured and/or processed in 13 out of the 16 cross-section States and  has
TRI releases in 14 of the 16 cross-section States.  The cross-section should adequately represent the
occurrence of DEHP on a national scale based upon the use, production, and release patterns of the 16-
State cross-section in relation to the patterns observed for all 50 States.

Occurrence Summary and Use Support Document         138                                     March 2002

-------
3.2.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 2000. Toxicological Profile for
       Di(2-ethylhexyl)phthalate.  U.S. Department of Health and Human Services, Public Health
       Service.  252 pp. + Appendices. Available on the Internet at
       http://www.atsdr.cdc .gov/toxprofiles/tp9 .pdf.

California Department of Health Services. Unpublished. Final Report on a Monitoring Program for
       Organic Chemical Contamination of Small Public Water Systems in California. Summary
       Version.  Sacramento, CA.

National Safety Council (NSC). 2001. Di(2-ethylhexyl) Phthalate Chemical Backgrounder. Itasca, IL:
       National  Safety Council. Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?id=103&chempath=chemicals, accessed
       July 23, 2001.

Staples, C.A., A.F. Werner, and T.J. Hoogheem.  1985.  Assessment of Priority Pollutant Concentrations
       in the United States Using STORET Data Base. Environmental Toxicology and Chemistry,  v. 4,
       pp.131-142.

USEPA.  1987. Health and Environmental Effects Profile for Phthalic Acid Alkyl, Aryl, and Alkyl/Aryl
       Esters. Final Draft.  Cincinnati, OH: Environmental Criteria and Assessment Office, Office of
       Health and Environmental Assessment.  ECAO-CIN-P188.

USEPA.  2000. TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001. National Primary Drinking  Water Regulations -  Consumer Factsheet on:
       Di(2-ethylhexyl) Phthalate. Office of Ground Water and Drinking Water, USEPA. Available on
       the Internet at http://www.epa.gov/safewater/dwh/c-SOC/DEHP.html, last updated April 12,
       2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National  Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USPHS.  1987. Draft Toxicological Profile for Di(2-Ethylhexyl)  Phthalate. Atlanta, GA: Agency for
       Toxic Substances and Disease Registry.

Wade Miller Associates, Inc. 1989. Occurrence and Exposure Assessment of Phthalate Esters in Public
       Drinking Water Supplies - DRAFT. Draft report submitted to EPA for review July 10,  1989.
Occurrence Summary and Use Support Document         139                                    March 2002

-------
3.3    Carbofuran
Table of Contents

3.3.1 Introduction, Use and Production 	  141
3.3.2 Environmental Release  	  142
3.3.3 Ambient Occurrence 	  143
3.3.4 Drinking Water Occurrence Based on the 16-State Cross Section 	  147
3.3.5 Additional Drinking Water Occurrence Data 	  151
3.3.6 Conclusion	  152
3.3.7 References  	  152
Tables and Figures

Figure 3.3-1:  Carbofuran Estimated Annual Agricultural Use	  142

Table 3.3-1:  Environmental Releases (in pounds) for Carbofuran in the United States, 1995-1999 . .  143

Table 3.3-2:  Carbofuran Detections and Concentrations in Surface Water and Ground Water	  143

Table 3.3-3.  Results (|ig/L) from the USGS NAWQA Surface Water Monitoring Program	  145

Table 3.3-4.  Results from the USGS NAWQA Ground Water Monitoring Program (for all wells
       sampled)  	  145

Table 3.3-5.  Results from the USGS NAWQA Monitoring Program for Ground Water	  146

Table 3.3-3:  Stage 1 Carbofuran Occurrence Based on 16-State Cross-Section - Systems	  147

Table 3.3-4:  Stage 1 Carbofuran Occurrence Based on 16-State Cross-Section - Population	  148

Table 3.3-5:  Stage 2 Estimated Carbofuran Occurrence Based on  16-State Cross-Section -
       Systems	  149

Table 3.3-6:  Stage 2 Estimated Carbofuran Occurrence Based on  16-State Cross-Section -
       Population	  150

Table 3.3-7:  Estimated National Carbofuran Occurrence - Systems and Population Served	  150
Occurrence Summary and Use Support Document         140                                    March 2002

-------
3.3.1  Introduction, Use and Production

Carbofuran (2,3-dihydro-2,2-dimethyl-7-benzofuranyl methylcarbamate) is a white crystalline solid with
a slightly phenolic odor.  Carbofuran is available in liquid and granular formulations. Thermal
decomposition of Carbofuran may include toxic oxides of nitrogen.  Fires, and the runoff from fire
control, may produce irritating or poisonous gases. Some of its trade names are Furadan, Bay 70143,
Curaterr, D 1221, ENT 27164, Yaltox, Furacarb (EXTOXNET, 1998).

Carbofuran is a broad spectrum insecticide used primarily to exterminate insects, mites, nematodes and
rootworm on contact or after ingestion. Earlier uses were primarily on corn crops.  The greatest use of
carbofuran is against foliar pests on alfalfa and rice, with turf and grapes making up most of the
remainder.  It is sprayed directly onto soil and plants just after emergence to control beetles, nematodes
and rootworm. These insecticides work by blocking the normal functioning of cholinesterase, an
essential nervous system enzyme (EXTOXNET, 1998).

Glaze (1982, as cited in USEPA, 1989) estimated that in 1980 approximately 11 million pounds of
carbofuran active ingredient were available for domestic use and approximately 8 million pounds were
exported. Gianessi (1986, as cited in USEPA, 1989) estimated that of the approximately 30 million
pounds of carbofuran used nationwide from 1978 to 1982, 20 million pounds were applied to field corn
and 4 million pounds were applied to sorghum. The remaining production was used on a larger number
of crops,  including alfalfa, tobacco, peanuts, rice, sugarcane, potatoes, soybeans, sweet corn, cotton,
grapes and small grains (USEPA, 1989).

The National Center for Food and Agricultural Policy (NCFAP) estimates total annual agricultural use at
about 5.1 million pounds per year (based on the 1992 agricultural census with updates from 1993 and
1995). The NCFAP estimates, combined with Census of Agriculture data on crop distributions can be
used to estimate the geographic distribution of carbofuran use (USGS, 1997).

Carbofuran is used on about 30 crops ~ the largest markets (in terms of pounds of active ingredient) are
for use on corn, alfalfa, rice, sorghum, potatoes, and cotton; followed by grapes, sugarbeets, soybeans,
wheat, tobacco, sunflowers,  sweet corn, sugarcane, and peanuts (USGS, 1997).

Figure 3.3-1 shows the United States Geological Survey (USGS, 1998a) derived geographic distribution
of estimated average annual  carbofuran use in the United States for 1992. A breakdown of use by crop is
also included.  The USGS (1998a) estimates almost 5 million pounds of carbofuran active ingredient
were used in 1992.  The largest concentrations of carbofuran use are seen in the Midwest.  A comparison
of this use map with the map of the 16 cross-section States (Figure 1.3-1) shows that States across the
range of high of low carbofuran use are well represented in the cross-section.
Occurrence Summary and Use Support Document         141                                     March 2002

-------
Figure 3.3-1: Carbofuran Estimated Annual Agricultural Use
                                         CARBOFURAN
                                  ESTIMATED ANNUAL AGRICULTURAL USE
                   Average use of
                   Active Ingredient
                 Pounds per square mile
                   of county par year
                   D  No Estimated lisa
                   D   < 0.219
                   D 0.219-0.859
                   D 0.880-2.152
                   D 2.153-4.1B9
                   •   >= 4.200
Total
Crops Pounds Applied
com 2,480,218
alfalfa hay 999,611
sorghum 389,221
potatoes 297,440
rice 206,884
cotton 1 72, B3B
tobacco 1 55, 462
grapas 120, 132
sugar cans: sugar & saad 87, 498
sugar beets for sugar 77, 9S2
Percent
National Use
48.40
19.51
7.60
5.60
4.04
337
323
2.35
1.71
1.52
Source: USGS, 1998a
3.3.2 Environmental Release

Carbofuran is listed as a Toxics Release Inventory (TRI) chemical. Table 3.3-1 illustrates the
environmental releases for carbofuran from 1995 to 1999.  Air emissions constitute most of the on-site
releases. Carbofuran air emissions generally decreased over the years, with the exception of a sharp
increase in 1999. Surface water discharges also dramatically increased in 1999. However, before 1999,
surface water discharges were negligible. Underground injection and releases to land are less significant
on-site releases, with underground injections always equal to zero and the only releases to land occurring
in 1995. Off-site releases of carbofuran are minimal. The TRI data for carbofuran were reported from 10
States, with four reporting every year (USEPA, 2000).  Four of the 10 States with TRI data are included
in the 16-State cross-section (used for analyses of carbofuran occurrence in drinking water; see Section
3.3.4).  (For a map of the 16-State cross-section, see Figure 1.3-1.)
Occurrence Summary and Use Support Document
142
March 2002

-------
Table 3.3-1: Environmental Releases (in pounds) for Carbofuran in the United States, 1995-1999
Year
1999
1998
1997
1996
1995
On-Site Releases
Air Emissions
13,999
2,921
3,903
3,854
4,187
Surface Water
Discharges
112
1
1
1
2
Underground
Injection
-
-
-
-
-
Releases
to Land
-
-
-
-
250
Off-Site Releases
-
-
42
-
250
Total On- &
Off-site
Releases
14,111
2,922
3,946
3,855
4,689
 Source: USEPA, 2000
3.3.3 Ambient Occurrence

Carbofuran is an analyte for both surface and ground water NAWQA studies, with a method detection
limit (MDL) of 0.003 |ig/L. Additional information on analytical methods used in the NAWQA study
units, including method detection limits, are described by Gilliom and others (1998).

Carbofuran concentrations exceed the detection limit in all of the ground and surface sites. Detection
frequencies are consistently greater for surface water than for ground water, possibly because surface
waters are more sensitive to anthropogenic releases. High percentages of surface water samples exceed
detection frequencies of 0.01 and 0.05 |ig/L.  Within surface water sites, urban sites have consistently
lower percentage exceeding the detection frequencies then agricultural or integrator sites. Within ground
water sites, agricultural sites have slightly higher percentage exceeding the detection frequencies then
urban or major aquifer sites. The maximum concentration exceeds the detection limit in all ground and
surface water sites.  The 95th percentile values exceed the detection limit for all surface water sites except
urban sites, while none of the 95th percentile values exceed the detection limit for any ground water sites.
None of the median values exceed the detection limit in ground or surface water sites.
Table 3.3-2: Carbofuran Detections and Concentrations in Surface Water and Ground Water
                               Detection frequency
                                 (% of samples)
                    Concentration percentiles
                      (all samples; |J.g/L)

surface water
agricultural
urban
integrator
all sites
all samples

11.99%
2.75%
9.35%
7.96%
>0.01 us/L

10.79%
2.75%
7.72%
6.93%
> 0.05 us/L

5.49%
0.31%
2.85%
3.08%
10th


-------
                               Detection frequency                    Concentration percentiles
                                 (% of samples)                         (all samples; |J.g/L)
ground water
agricultural
urban
major aquifers
all sites

0.76%
0.66%
0.54%
0.50%

0.65%
0.66%
0.54%
0.46%

0.43%
0.33%
0.21%
0.26%


-------
percentiles shown will be biased high for commonly occurring conditions because more samples were
collected at sites where concentrations were high, or samples were collected more frequently during
periods of elevated concentrations. On the other hand, the values shown may be biased low because
sampling was not conducted during high-use periods.  The maximum concentrations shown in Table 3.3-
3 are the highest concentrations observed in all NAWQA stream samples. Table 3.3-3 should not be
presumed to be a statistically representative summary of the NAWQA pesticide results. Samples were
collected between 4/20/92 and 12/16/96. The Level of Detection (LOD) for carbofuran was 0.003 |ig/L.
Table 3.3-3.  Results (ng/L) from the USGS NAWQA Surface Water Monitoring Program
Land Use
Agricultural
Non-Agricultural
Sites
506
550
Samples
2,996
2,203
Detects
297
117
Range 1
9.70 - ND2
7.00 - ND
Mean
0.023
0.012
95th
Percentile
0.040
0.008
Median
ND
ND
1. Range, mean, median and 95th percentile are determined from all samples. Samples below the LOD were given a value one-half the LOD.
2. Below the LOD.
Table 3.3-4 summarizes data for every NAWQA ground water sample that was analyzed for carbofuran,
including newly drilled monitoring wells, production wells (such as domestic and public-supply wells),
springs, and tile drains. Although Table 3.3-4 provides a complete summary of all NAWQA results, it
should not be presumed to be a statistically representative summary of the NAWQA pesticide results.
The data in the table contain a variety of spatial and temporal biases for which corrections must be
applied before any reliable statistical summaries can be compiled. For example, many of the sites were
sampled more than once.  Failure to account for this would lead to an over-representation of these sites in
any statistical summary of chemistry data in which they were included.  Samples were collected between
8/19/92 and 11/15/96. The LOD for carbofuran was 0.003 |ig/L.
Table 3.3-4.  Results from the USGS NAWQA Ground Water Monitoring Program (for all wells
sampled)
Wells
2,550
Samples
3,024
Detects
15
Range'CM-g/L)
1.30-ND2
Mean
ND
95th Percentile
ND
Median
ND
1. Range, mean, median and 95th percentile are determined from all samples. Samples below the LOD were given a value one-half the LOD.
2. Below the LOD.
The USGS generated statistical summaries of the ground water data for three different settings: shallow
ground water in primarily agricultural areas, shallow ground water in primarily urban areas, and major
aquifers. The agricultural and urban land-use categories were represented by wells chosen or designed to
sample shallow, recently recharged (less than ten years old) ground water to determine the effects of
specific land uses on water quality. Sites comprising the "major aquifer" (an aquifer that is currently
being used as a source of drinking water) category had no such restrictions on land use or water age, and
thus, represent a broader mixture of land uses and ground water depths. The USGS followed the
following procedures to generate the relatively unbiased and comparable statistical summaries: tile
Occurrence Summary and Use Support Document
145
March 2002

-------
drains and springs were excluded to reduce the variability in site type; any well co-located with another
existing well was excluded; networks with fewer than 10 wells were excluded because they contained an
insufficient number of wells to be spatially representative of an area; wells that were included in more
than one type of network (e.g. a land-use study and an aquifer survey) were allowed to exist in both; one
sample from each well was selected.
Table 3.3-5. Results from the USGS NAWQA Monitoring Program for Ground Water
Setting/Land Use
Shallow
Ground Water
Urban
Agricultural
Major Aquifer
Wells
301
925
933
Samples
301
925
933
Detects
2
7
5
Range1
(llg/L)
0.058 -ND2
1.30 -ND
0.40 -ND
Mean
NR3
NR
NR
95th
Percentile
ND
ND
ND
Median
ND
ND
ND
1. Range, mean, median and 95th percentile are determined from all samples. Samples below the LOD were given a value one-half the LOD.
2. Below the LOD.
3. Not Reported.
3.3.3.2.2 Groundwater Sources

Unpublished data summarized from EPA registration files and reported in Cohen et al. (1984, as cited in
USEPA, 1989) indicated that concentrations of carbofuran were present in samples of ground water
collected from private wells in areas with sandy soils and water table aquifers in Wisconsin and New
York. Concentrations ranged from 1 to 50 |ig/L carbofuran. The number of samples, number of positive
samples, and detection limit were not reported.

Three studies were obtained examining ground water wells in California.  These studies were conducted
by the California Department of Food and Agriculture in 1981, and again in 1982, and an evaluation in
1984 during the California State Board's Toxics Special Project (Ramlit Associates, Inc., 1983; Holden,
1986; Cohen and Bowes, 1984, respectively, all as cited in USEPA, 1989).  Overall, approximately 30
counties were sampled for the occurrence of carbofuran.  Over 200 samples were collected from as many
sites, with only 2 proving positive. One concentration was 0.5 |ig/L; the  other was not reported. The
detection limit for the sample concentration given above was not reported.

3.3.3.2.3 Surface Water Sources

The Army Corps of Engineers sampled U.S. Geological Survey water stations along rivers of the Honey
Creek watershed in northwest Ohio during 1981 (Datta, no date, as cited  in USEPA, 1989). Carbofuran
was sampled for at 12 locations and a maximum concentration of 45 |ig/L was found. The number of
samples, number of positives, and detection limit were not reported.

Dudley  and Karr (1980, as cited in USEPA, 1989) presented data on levels of carbofuran in water, fish,
and sediment samples collected from a stream draining the Black Creek agricultural watershed  in Allen
County, Indiana, during 1977 to 1978. Although the detection limit and the number of water samples
collected and tested were not reported, none of the samples contained carbofuran in excess of the
detection limit.
Occurrence Summary and Use Support Document
146
March 2002

-------
Woodham et. al (1975, as cited in USEPA, 1989) presented monitoring data from a study to determine
whether significant pesticide accumulation had occurred in two counties in North Carolina.  Samples of
pond water collected both inside and outside the study area contained no carbofuran in excess of the
detection limit of 0.05 |ig/L.

3.3.3.2.4 Unidentified Sources

Water samples of unidentified sources were collected nationally from various studies and entered in the
U.S. EPA's STORET database (as reported in USEPA, 1984, as cited in USEPA, 1989) during 1979 to
1982. Of the 21 stations sampled for carbofuran, 11 had undetectable levels. However, 58 samples were
reported as positive. The total number of samples, detection limit(s), and range of positive values were
not reported.

3.3.4 Drinking Water Occurrence Based on the 16-State Cross Section

The analysis of carbofuran occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States. The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date.  These data were evaluated
relative to several concentration thresholds of interest: 0.04 mg/L; 0.007 mg/L; and 0.004 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
carbofuran. These data represent almost 52,000 analytical results from approximately 14,000 PWSs
during the period from 1983 to 1998 (with most analytical results from 1992 to 1997). The number of
sample results and PWSs vary by State, although the State data sets have been reviewed and checked to
ensure adequacy of coverage and completeness. The overall modal detection limit for carbofuran in the
16 cross-section States is equal to 0.0009 mg/L. (For details regarding the 16-State cross-section, please
refer to Section 1.3.5 of this report.)

3.3.4.1 Stage 1 Analysis Occurrence Findings

Table 3.3-3 illustrates the very low occurrence of carbofuran in public drinking water  in drinking water
for the public water systems in the 16-State cross-section. Zero systems had any analytical results
exceeding the MCL (0.04 mg/L); 0.00718% of systems (1 system) had results exceeding 0.007 mg/L; and
0.0215% of systems (3 systems) had results exceeding 0.004 mg/L.

Only 1 (0.00798% of) ground water system had any analytical  results greater than 0.007 mg/L. No
surface water systems had results exceeding 0.007 mg/L. About 0.0160% of ground water systems (2
systems) had results above 0.004 mg/L.  The percentage of surface water systems with at least one result
greater than 0.004 mg/L was equal to 0.0717% (1 system).
Table 3.3-3:  Stage 1 Carbofuran Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.04
0.007
0.004
Percent of Systems
Exceeding Threshold
0.000%
0.00798%
0.0160%
Number of Systems Exceeding
Threshold
0
1
2
Occurrence Summary and Use Support Document
147
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems Exceeding
Threshold

Surface Water
0.04
0.007
0.004
0.000%
0.000%
0.0717%
0
0
1

Combined Ground &
Surface Water
0.04
0.007
0.004
0.000%
0.00718%
0.0215%
0
1
3
Reviewing carbofuran occurrence in the 16 cross-section States by PWS population served (Table 3.3-4)
shows that approximately 0.000239% of the population (200 people) was served by PWSs with at least
one analytical result of carbofuran greater than 0.007 mg/L. A total of 2,100 (0.00218% of) people were
served by systems with an exceedance of 0.004 mg/L. As indicated above, no systems (therefore, no
population served by systems) had any analytical results greater than the MCL (0.04 mg/L).

The percentage of population served by ground water systems with analytical results greater than 0.007
mg/L was equal to 0.000576% (200 people).  When evaluated relative to 0.004 mg/L, the percent of
population by ground water systems exposed was equal to 0.00105% (400 people).  No surface water
systems had exceedances of 0.007 mg/L. The percentage of population served by surface water systems
with exceedances of 0.004 mg/L was equal to 0.00308% (1,700 people).
Table 3.3-4:  Stage 1 Carbofuran Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.04
0.007
0.004
Percent of Population Served
by Systems
Exceeding Threshold
0.000%
0.000576%
0.00105%
Total Population Served by
Systems Exceeding
Threshold
0
200
400

Surface Water
0.04
0.007
0.004
0.000%
0.000%
0.00308%
0
0
1,700

Combined Ground &
Surface Water
0.04
0.007
0.004
0.000%
0.000239%
0.00218%
0
200
2,100
Occurrence Summary and Use Support Document
148
March 2002

-------
3.3.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.3-5 and 3.3-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

No ground water or surface water PWSs in the 16 States had an estimated  mean concentration of
carbofuran exceeding 0.04 mg/L or 0.007 mg/L. The percentage of total ground and surface water
systems in the 16 States with estimated mean concentration values of carbofuran greater than 0.004 mg/L
was equal to 0.000115% (an estimate of less than one  system in the 16-State cross-section). The
percentage of ground water systems with estimated mean concentration values greater than 0.004 mg/L
was equal to 0.000112%. The percentage of surface water systems with estimated mean concentration
values greater than 0.004 mg/L was equal to 0.000144%.
Table 3.3-5:  Stage 2 Estimated Carbofuran Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.04
0.007
0.004
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000112%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.04
0.007
0.004
0.000%
0.000%
0.000144%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.04
0.007
0.004
0.000%
0.000%
0.000115%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
Reviewing carbofuran occurrence by PWS population served (Table 3.3-6) shows that approximately
0.00000461% of population served by all PWSs (less than 1  system in the 16 States) were potentially
exposed to carbofuran levels above 0.004 mg/L. The percentage of population served by ground water
systems potentially exposed to carbofuran levels above 0.004 mg/L was equal to 0.00000806%. The
percentage of population served by surface water systems with estimated mean concentration values
greater than 0.004 mg/L was equal to 0.00000217%. When evaluated relative to a threshold of 0.04
mg/L, and 0.007 mg/L, the percent of population exposed was equal to 0% for all  system types.
Occurrence Summary and Use Support Document
149
March 2002

-------
Table 3.3-6:  Stage 2 Estimated Carbofuran Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.04
0.007
0.004
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.00000%
0.00000%
0.00000806%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in
the 16 States Estimated to Exceed
Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.04
0.007
0.004
0.00000%
0.00000%
0.00000217%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.04
0.007
0.004
0.00000%
0.00000%
0.00000461%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
3.3.4.3 Estimated National Occurrence

As illustrated in Table 3.3-7, the Stage 2 analysis estimates zero systems serving none of the national
population have estimated mean concentration values of carbofuran greater than 0.04 mg/L, 0.007 mg/L,
or 0.004 mg/L. (See Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to
national values.)
Table 3.3-7: Estimated National Carbofuran Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.04
0.007
0.004
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.04
0.007
0.004
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
Occurrence Summary and Use Support Document
150
March 2002

-------
Source Water Type
Threshold
(mg/L)
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.04
0.007
0.004
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
3.3.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of carbofuran in drinking water are also
reviewed. Previously compiled occurrence carbofuran information, from an OGWDW summary
document entitled "Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air in
the United States of America" (USEPA, 1989), is presented in the following section.  This variety of
studies and information are presented regarding levels of carbofuran in drinking water, with the scope of
the reviewed studies ranging from national to regional. Note that none of the studies presented in the
following section provide the quantitative analytical results or comprehensive coverage that would
enable direct comparison to the occurrence findings estimated with the cross-section occurrence data
presented in Section 3.3.4.  These additional  studies, however, do enable a broader assessment of the
Stage 2 occurrence estimates presented for this Six-Year Review. The information presented in the
following section is taken directly from "Occurrence and Human Exposure to Pesticides in Drinking
Water, Food and Air in the United States of America" (USEPA, 1989).

3.3.5.1 Ground Water Sources - Regional Studies

In 1984, the Suffolk County Department of Health Services (Holden, 1986, as cited in USEPA, 1989)
examined drinking water wells in Long Island, New York, for various pesticides. The survey sampled
both public and private wells in close proximity to fields where carbofuran, aldicarb, and other pesticides
were used.  Of the 5,083 wells sampled, 1,535 contained detectable levels of carbofuran and 250 to 300
wells  contained levels greater than 15 |ig/L.  The maximum level reported was 65 |ig/L.

As part of a program of the Wisconsin Department of Natural Resources (Holden, 1986, as cited in
USEPA,  1989), drinking water wells were also analyzed for carbofuran during 1983-1984.  This study
examined wells suspected of contamination by both point and nonpoint sources.  Of 78 samples analyzed,
2 were positive, with a high concentration of 7 |ig/L. No other information was reported.

Ground water wells were sampled near Richmond, Rhode Island, as part of a cooperative project between
the U.S. Geological Survey and the Rhode Island Water Resources Board to identify potential drinking
water sources for future water supply use (Offutt, 1984, as cited in USEPA,  1989).  Eleven samples were
collected during 1984 from five locations and analyzed for carbofuran. Seven were positive, with a mean
concentration of 3.7 |ig/L (range = 2 to 7 M-g/L). The detection limit was 1.0 |ig/L.
Occurrence Summary and Use Support Document
151
March 2002

-------
3.3.6  Conclusion

Carbofuran is a broad spectrum carbamate pesticide used primarily to exterminate insects, mites,
nematodes and rootworm on contact or after ingestion. It is used against soil and foliar pests of field,
fruit, vegetable and forest crops. The USGS (1998a) estimates almost 5 million pounds of carbofuran
active ingredient were used in 1992. The largest concentrations of carbofuran use are seen in the
Midwest. Industrial releases of carbofuran have been reported to TRI since 1995 from  10 States, with
four reporting every year. Carbofuran was also an analyte for the NAWQA occurrence studies. In the
NAWQA study, carbofuran was detected in ground and surface water; however, none of the median
values exceeded the detection limit. The Stage 2 analysis, based on the 16-State cross-section, estimated
that zero percent of combined ground water and surface water systems serving zero percent of the
population exceeded the MCL of 0.04 mg/L. Based on this estimate, zero PWSs nationally  are estimated
to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. According to information from USGS, all of the 16 cross-section States use carbofuran,
although most States in light to moderate amounts.  Carbofuran is used most heavily in the Midwest and
California, covered by seven cross-section States. Nationally, TRI releases have been reported for
carbofuran from  10 States,  including 4 of 16 cross-section States.  The cross-section should adequately
represent the occurrence of carbofuran on a national scale based upon the use, production, and release
patterns of the 16-State cross-section in relation to the patterns observed for all 50 States.

3.3.7  References

Blomquist, J.D., J.M. Denis, J.L. Cowles, J.A. Hetrick, R.D. Jones, andN.B. Birchfield.  2001.
       Pesticides in Selected Water-Supply Reservoirs and Finished Drinking Water, 1999-2000:
       Summary of Results from a Pilot Monitoring Survey.  U.S. Geological Survey Open-File Report
       01-456.  65pp.

Cohen, S.Z., S.M. Creeger, R.F.  Carsel, and C.G. Enfield. 1984. Pesticide Contamination of
       Groundwaterfrom Agricultural Uses.  American Chemical Society Symposium Series Treatment
       and Disposal of Pesticide Wastes.

Cohen, D.B, and G.W. Bowes.  1984.  Water Quality and Pesticides: A  California Risk Assessment
       Program (Volume 1).  Sacramento, CA: State Water Resources Control Board, Toxic Substances
       Control Program.

Datta, P.R.  No date. Memorandum: Review of six documents regarding monitoring of pesticides in
       Northwestern Ohio rivers.  Sent to: D.J. Severn, Chief- Exposure Assessment Branch, HED.
       Washington, DC: Office of Pesticides and Toxic Substances, U.S. Environmental Protection
       Agency.

Dudley, D.R, and J.R. Karr. 1980.  Pesticides and PCB  residues in the Black Creek watershed, Allen
       County, Indiana, 1977-1978. Pestic. Monit. J. v. 13, no. 4, pp. 155-157.

EXTOXNET.  1998. Pesticide Information Profile: Carbofuran.  Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://pmep.cce .Cornell .edu/profiles/extoxnet/metiram-propoxur/carbofuran-ext.html, last
       updated March 1,2001.

Occurrence Summary and Use Support Document         152                                     March 2002

-------
Gianessi, L.P.  1986. A National Pesticide Usage Database. Prepared by Resources for the Future for
       U.S. Environmental Protection Agency, Office of Standards and Regulations, Washington, DC.
       Cooperative Agreement CR 811858-01-0.

Gilliom, R.J., D.K. Mueller, and L.H. Nowell.  1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589, last updated October 9, 1998.

Glaze, M.H. 1982.  Preliminary Quantitative Usage of Carbofuran. Washington, DC: Office of
       Pesticide Programs, U.S. Environmental Protection Agency.

Holden, P.W. 1986. Pesticides and Groundwater Quality.  Issues and Problems in Four States.
       Prepared for the Board of Agriculture, National Research Council. Washington, DC: National
       Academy Press.

JRB Associates.  1984. Occurrence of Carbofuran in Drinking Water, Food, and Air.  Prepared for and
       submitted to EPA on February 9, 1984.

Offutt, C.K. 1984.  Memorandum: Aldicarb in Rhode Island Groundwater. From: Carolyn K. Offutt,
       Chief- Environmental Processes and Guidelines  Section, Exposure Assessment Branch, HED.
       Sent to: Paul Lapsley, Chief- Special Review Branch, Registration Division.  Washington, DC:
       Office of Pesticides and Toxic Substances, U.S. Environmental Protection Agency.

Ramlit Associates, Inc.  1983. Groundwater Contamination by Pesticides: A California Assessment.
       Submitted to:  State Water Resources Control Board, Sacramento, California.  Submitted by:
       Ramlit Associates, Inc., Berkeley, California. Publication No. 83-4SP.

USEPA.  1984. Health and Environmental Effects Profile for Carbofuran. Prepared for Office of Solid
       Waste and Emergency Response, Washington, DC, by Experimental Criteria and Assessment
       Office, Cincinnati, OH.

USEPA.  1989. Draft Final Report on the Occurrence and Human Exposure to Pesticides in Drinking
       Water,  Food, and Air in the United States of America. Office of Drinking Water, USEPA.
       September, 1989.

USEPA.  2000. TRIExplorer: Trends. Available on the  Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001. EFED Risk Assessment for the Reregistration Eligibility Decision on Carbofuran,
       2/15/2001.  Environmental Fate and Effects Division, Office of Pesticide Programs, USEPA.
       February, 2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 1997. Carbofuran Occurrence and Distribution  in Surface and Ground Waters 1991-1997 -
       DRAFT.  Pesticides National Synthesis Project, USGS. November 25, 1997.
Occurrence Summary and Use Support Document         153                                    March 2002

-------
USGS.  1998a. Annual Use Maps.  Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated 3/20/1998.

USGS.  1998b. Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA). PROVISIONAL DATA - SUBJECT TO
       REVISION. Available on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/,  last updated
       October 9, 1998.

Woodham, D.W., M.C. Ganyard, C.A. Bond, and R.G. Reeves.  1975. Monitoring of agricultural
       pesticides in a cooperative pest management project in North Carolina, 1971, first year of study.
       In: Savage, E.P. (ed.). Environmental chemicals. Human and animal health. Proceedings of 4th
       Annual Conference, Colorado  State University, Fort Collins, CO.
Occurrence Summary and Use Support Document         154                                    March 2002

-------
3.4    Chlordane
Table of Contents

3.4.1  Introduction, Use and Production  	  156
3.4.2  Environmental Release  	  157
3.4.3  Ambient Occurrence 	  157
3.4.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  159
3.4.5  Additional Drinking Water Occurrence Data  	  163
3.4.6  Conclusion	  165
3.4.7  References  	  165
Tables and Figures

Table 3.4-1: Facilities that Manufacture or Process Chlordane  	  156

Table 3.4-2: Environmental Releases (in pounds) for Chlordane in the United States, 1988-1996 . . .  157

Table 3.4-3: Stage 1 Chlordane Occurrence Based on 16-State Cross-Section - Systems	  159

Table 3.4-4: Stage 1 Chlordane Occurrence Based on 16-State Cross-Section - Population	  160

Table 3.4-5: Stage 2 Estimated Chlordane Occurrence Based on 16-State Cross-Section -
       Systems	  161

Table 3.4-6: Stage 2 Estimated Chlordane Occurrence Based on 16-State Cross-Section -
       Population	  162

Table 3.4-7: Estimated National Chlordane Occurrence - Systems and Population Served  	  162
Occurrence Summary and Use Support Document          155                                     March 2002

-------
3.4.1  Introduction, Use and Production

Chlordane (1,2,4,5,6,7,8,8-octachloro-2,3,3a,4,7,7a-hexahydro-4,7-methanoindene) does not occur
naturally in the environment. It is a thick nonpolar liquid with a mild, irritating odor whose color ranges
from colorless to amber. "Technical chlordane" (the contaminant described throughout this report) is
actually not a single chemical. It is a mixture of pure chlordane and about 10 major related chemicals
and numerous other components. Some of the major components are trans-chlordane, cis-chlordane,
beta-chlordene, heptachlor, and trans-nonachlor. Chlordane does  not dissolve in water. Therefore, before
it can be used as a spray, it must  be placed in water with emulsifiers (soap-like substances), resulting in a
milky-looking mixture. In the environment, chlordane tends to bind to soil and to degrade slowly. Some
of its trade names are Octachlor and Velsicol 1068 (ATSDR, 1994).

Chlordane is a manufactured chemical that was used as a pesticide in the United States from 1948 to
1988 (ATSDR, 1994). It was used on corn, citrus, deciduous fruits and nuts, and  vegetables.  It was also
used on garden and ornamentals; lawns, turf, ditchbanks and roadsides. Chlordane was used to control a
variety of insect pests including parasitic roundworms and other nematodes, termites, cutworms,
chiggers, and leafhoppers (USEPA, 2001). In the mid 1970s the  use pattern for chlordane was as
follows: 35% used by pest control operators,  mostly on termites;  28% on agricultural crops, including
corn and citrus; 30% for home lawn and garden use;  and 7% on turf and ornamentals  (ATSDR, 1994).

On March 6, 1978, use of chlordane was suspended except for termite control and some application on
non-food plants.  In 1983, its use on non-food plants was also banned (ATSDR, 1994). Its use and
production were ultimately banned on April 14, 1988 because of concern over cancer risk, evidence of
human exposure and accumulation in body fat, persistence in the environment, and danger to wildlife
(NSC, 2001). Currently, its only permitted commercial use is for fire ant control in power transformers
(USEPA, 2001).

Prior to 1983 the estimated annual usage of chlordane was over 3.6 million pounds (USEPA, 2001).
Velsicol Chemical Company in Memphis, TN, is currently the only U.S. manufacturer of chlordane, and
it was the only domestic manufacturer when use of chlordane ceased in 1988 (ATSDR, 1994).

Table 3.4-1 shows the facilities that manufacture and process chlordane, the intended uses of the product,
and the range of maximum amounts derived from the Toxics Release Inventory (TRI) of EPA  (ATSDR,
1994).
Table 3.4-1:  Facilities that Manufacture or Process Chlordane3
Facility
Stennis Space Center
Velsicol Chemical Corp.
Location1"
Stennis Space Cen,
Memphis, TN
Range of maximum amounts on site in
pounds
0-99
100,000-999,999
Activities and uses
In ancillary or other uses
Produce; for sale/distribution
'Derived from TRI90 (1992)
'Post Office State abbreviations

Source: ATSDR, 1994 compilation O/TR190 1992 data
Occurrence Summary and Use Support Document
156
March 2002

-------
3.4.2  Environmental Release

Chlordane is listed as a Toxics Release Inventory (TRI) chemical. Table 3.4-2 illustrates the
environmental releases for chlordane from 1988 to 1996.  (There are only chlordane data for these years.)
Air emissions constitute most of the on-site releases, and emissions fluctuated from 1988 to 1993, before
decreasing from 1994 to  1996.  Surface water increases have gradually increased, from 1 pound in 1991
to 95 pounds in 1996.  The fluctuation in air emissions has been the major contributor to changes in
chlordane total on- and off-site releases in recent years. No releases to land (such as spills or leaks
within the boundaries of the reporting facility), or off-site releases (including metals or metal compounds
transferred off-site) were reported for chlordane. These TRI data for chlordane were reported from 10
States (USEPA, 2000). Five of the 16 cross-section Sates (used for analyses of chlordane occurrence in
drinking water; see Section 3.4.4) reported data for chlordane.  (For a map of the 16-State cross-section,
see Figure 1.3-1.)
Table 3.4-2:  Environmental Releases (in pounds) for Chlordane in the United States, 1988-1996
Year
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
660
823
1,300
51
1,713
1,427
4,422
3,753
2,698
Surface Water
Discharges
95
22
13
15
1
1
1
4
4
Underground
Injection
-
-
-
-
-
-
-
-
4,262
Releases
to Land
-
-
-
-
-
-
-
-
-
Off-Site Releases
-
-
-
-
-
-
-
-
-
Total On- &
Off-site
Releases
755
845
1,313
66
1,714
1,428
4,423
3,757
6,964
 Source: USEPA, 2000b
3.4.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers.  There are few available data on the occurrence of chlordane in ambient waters of the
United States. The most comprehensive and nationally consistent data describing ambient water quality
in the United  States are being produced through the United States Geological Survey's (USGS) National
Water Quality Assessment (NAWQA) program. However, national NAWQA data are currently
unavailable for chlordane.

3.4.3.1 Additional Ambient Occurrence Data

A summary document entitled "Occurrence and Human Exposure to Pesticides in Drinking Water, Food
and Air in the United States of America" (USEPA,  1989), was previously prepared for past USEPA
assessments of various pesticides.  In that review, two national studies and seven regional studies
analyzed concentrations of chlordane in water other than  drinking water. The following information is
taken directly from "Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air in
the United States of America" (USEPA, 1989).
Occurrence Summary and Use Support Document
157
March 2002

-------
3.4.3.1.1 Ground Water Sources - Regional Study

Tucker and Burke (1978, as cited in USEPA, 1989) reported the results of a Statewide groundwater
monitoring project conducted in New Jersey. For this study, samples of water were collected from 163
wells in nine counties. The analysis showed that samples of water from five wells contained levels of
chlordane in excess of the minimum reportable concentrations of 0.01 |ig/L. The range of positive values
was 0.01 to 0.02 |ig/L chlordane.

One California study of ground water wells found four samples, from as many counties, positive for
chlordane with a maximum value of 22 |ig/L. No other data were presented for this study. One of three
well  samples collected in the Barstow, California, area was found to have a chlordane concentration of
0.4 |ig/L by the Victorville Regional Water Quality Board in 1980 (Ramlit Associates, Inc., 1983, as
cited in USEPA, 1989).

3.4.3.1.2 Surface Water Sources - National Study

The National Pesticide Monitoring Network (Gilliom et al.,  1985, as cited in USEPA, 1989) examined
surface water samples from rivers nationwide during 1975-1980, and found no positive samples of
chlordane out of 2,943 samples analyzed from 177 locations. The detection limit was 0.15 |ig/L.

The National Surface Water Monitoring Program (Carey and Kutz, 1983, as cited in USEPA, 1989)
presented data on levels of chlordane in surface water samples collected throughout the United States
between 1976 and 1980.  Although no detection limit for chlordane was reported, 1.1 percent of the
samples analyzed had detectable concentrations of chlordane, with a maximum reported value of 0.23
l-ig/L. The number of samples taken was not reported. It is not known whether the samples were filtered
or unfiltered.

3.4.3.1.3 Surface Water Sources - Regional Studies

Truhlar and Reed (1976, as cited in USEPA, 1989) reported the results of analysis of water samples taken
from four streams in Pennsylvania. The streams drained four types of land use areas: forests, general
farms, orchards, and residential areas.  None of the 19 samples  collected and analyzed from April 1970 to
February 1971 contained chlordane in excess of the detection limit. The detection limit was not reported.

Benvenue et al. (1972b, as cited in USEPA, 1989) conducted extensive sampling of two canals on Oahu,
Hawaii, to determine the extent of organochlorine pesticide  contamination.  A total of nine samples was
collected from two canals and analyzed for residues of chlordane.  The analyses showed an average
concentration of chlordane of 0.007 |ig/L and a range of positive values of 0.003 to 0.018 |ig/L. The
number of positive samples and the detection limit were not reported.

Barks (1978, as cited in USEPA, 1989) presented the results of a USGS water quality study conducted
from April  1973 to July 1974 in the Ozark National Scenic Riverways, Missouri. During the study, 20
surface water samples were collected from 3 sites on the Current River and 1 site on Jacks Fork and
analyzed for pesticide content. None of the unfiltered samples  contained  concentrations of chlordane in
excess of the detection limit (no detection limit was reported).

Englande et al. (1978, as cited in USEPA, 1989) presented the results of extensive chemical analyses of
six Advanced Wastewater Treatment (AWT) plant effluents. The plants were located in California, the
District of Columbia, and Texas. A mean concentration of less than 0.039 |ig/L chlordane was identified
Occurrence Summary and Use Support Document          158                                     March 2002

-------
in one plant effluent; the other systems had no detectable levels of chlordane. The number of positive
samples and the detection limit were not reported.

3.4.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of chlordane occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States. The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest:  0.002 mg/L; 0.001 mg/L; and 0.0002 mg/L.

All sixteen cross-section State data sets, with the exception of Michigan, contained occurrence data for
chlordane.  These data represent more than 59,000 analytical results from approximately 13,000 PWSs
during  the period from 1984 to 1998 (with most analytical results from 1992 to 1997). The number of
sample results and PWSs vary by State, although the State data sets have been reviewed and checked to
ensure  adequacy of coverage and  completeness. The overall modal detection limit for chlordane in the
16 cross-section States is equal to 0.0002 mg/L.  (For details regarding the 16-State cross-section, please
refer to Section 1.3.5 of this report.)

3.4.4.1 Stage 1 Analysis Occurrence Findings

Table 3.4-3 illustrates the Stage 1 analysis of chlordane occurrence in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds:  0.002 mg/L (the current MCL), 0.001,
and 0.0002 mg/L (the modal MRL). A total of 2 (approximately  0.0152% of) ground water and surface
water PWSs had  analytical results exceeding the MCL (0.002 mg/L); 0.303% of systems (4 systems) had
results  exceeding 0.001 mg/L; and 0.106% of systems (14 systems) had results exceeding 0.0002 mg/L.

Approximately 0.0167% of ground water systems had any analytical results greater than 0.002 mg/L.
About 0.0337% of ground water systems had results above the MCL (0.002 mg/L).  The percentage of
ground water systems with at least one result greater than 0.0002 mg/L was equal to 0.118%. No surface
water systems had any analytical results greater than 0.002 mg/L, 0.001 mg/L, or 0.0002 mg/L.
Table 3.4-3:  Stage 1 Chlordane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0002
Percent of Systems
Exceeding Threshold
0.0167%
0.0337%
0.118%
Number of Systems
Exceeding Threshold
2
4
14

Surface Water
0.002
0.001
0.0002
0.000%
0.000%
0.000%
0
0
0
Occurrence Summary and Use Support Document
159
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems
Exceeding Threshold

Combined Ground &
Surface Water
0.002
0.001
0.0002
0.0152%
0.0303%
0.106%
2
4
14
Reviewing chlordane occurrence in the 16 cross-section States by PWS population served (Table 3.4-4)
shows that approximately 0.000477% of the population (about 500 people) was served by ground water
PWSs with at least one analytical result of chlordane greater than the MCL (0.002 mg/L). A total of 600
(0.000631% of) people in the 16-States were served by ground water systems with an exceedance of
0.001 mg/L. Approximately 0.0275% of the total 16-State population (26,800 people) was served by
ground water systems with at least one analytical result greater than 0.0002 mg/L. No surface water
systems exceeded any of the threshold levels.
Table 3.4-4:  Stage 1 Chlordane Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0002
Percent of Population
Served by Systems
Exceeding Threshold
0.00114%
0.00151%
0.0658%
Total Population Served
by Systems Exceeding
Threshold
500
600
26,800

Surface Water
0.002
0.001
0.0002
0.000%
0.000%
0.000%
0
0
0

Combined Ground &
Surface Water
0.002
0.001
0.0002
0.000477%
0.000631%
0.0275%
500
600
26,800
3.4.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.4-5 and 3.4-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).
Occurrence Summary and Use Support Document
160
March 2002

-------
No ground or surface water systems had estimated mean concentrations of chlordane greater than the
MCL (0.002 mg/L). Approximately 0.0000910% of systems (less than 1 system in the 16 States) had
estimated mean concentration values greater than 0.001 mg/L. Approximately 0.0363% of systems
(about 5 systems) had estimated mean concentrations greater than 0.0002 mg/L.

The percentage of ground water PWSs with an estimated mean concentration of chlordane exceeding
0.001 mg/L was equal to 0.000101% (less than 1 ground water system in the 16 States). Approximately
0.0403% of ground water systems (5 systems) had an estimated mean concentration of chlordane greater
than 0.0002 mg/L.  A very small proportion (0.000301%) of surface water systems had estimated mean
concentration values greater than 0.0002 mg/L.
Table 3.4-5:  Stage 2 Estimated Chlordane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0002
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000101%
0.0403%
Range
0.000% - 0.000%
0.000% - 0.000%
0.0253% - 0.0591%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
5
Range
0-0
0-0
3-7

Surface Water
0.002
0.001
0.0002
0.000%
0.000%
0.000301%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.002
0.001
0.0002
0.000%
0.0000910%
0.0363%
0.000% - 0.000%
0.000% - 0.000%
0.0228% -0.0531%
0
0
5
0-0
0-0
3-7
Table 3.4-6 illustrates chlordane occurrence by PWS population served.  Approximately 0.00000146% of
population served by all PWSs in the 16 States were potentially exposed to chlordane levels above 0.001
mg/L.  The percent of population exposed to levels of 0.0002 mg/L was 0.00147% (approximately 1,400
people in the 16 States). As stated above, no systems had estimated mean concentrations of chlordane
greater than the MCL (0.002 mg/L).

The percentage of population served by ground water systems in the 16 cross-section States with
estimated mean concentration values greater than 0.001 mg/L was equal to 0.00000350%.  When
evaluated relative to a threshold of 0.0002 mg/L, the percent of population exposed was equal to
0.00346% (1,400 people). The percentage of population served by surface water systems with estimated
mean concentrations of chlordane greater than 0.0002 mg/L was equal to 0.0000479%.
Occurrence Summary and Use Support Document
161
March 2002

-------
Table 3.4-6: Stage 2 Estimated Chlordane Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.00000350%
0.00346%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000945% -0.0166%
Total Population Served by Systems in
the 16 States Estimated to Exceed
Threshold
Best Estimate
0
0
1,400
Range
0-0
0-0
400 - 6,700

Surface Water
0.002
0.001
0.0002
0.000%
0.000%
0.0000479%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.002
0.001
0.0002
0.000%
0.00000146%
0.00147%
0.000% - 0.000%
0.000% - 0.000%
0.000394% -0.00701%
0
0
1,400
0-0
0-0
400 - 6,800
3.4.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 24 PWSs nationally serving approximately 3,100 people could be exposed to
chlordane concentrations above 0.0002 mg/L.  No systems and population served by systems nationally
were estimated to have mean concentrations greater than 0.002 mg/L or 0.001 mg/L. (See Section  1.4 for
a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 24 PWSs serving about 3,000 people nationally were expected to
have mean concentrations greater than 0.0002 mg/L.  Approximately 1 surface water systems serving less
than 100 people was estimated to have mean concentrations of chlordane above 0.0002 mg/L.
Table 3.4-7: Estimated National Chlordane Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.002
0.001
0.0002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
24
Range
0-0
0-0
15-35
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
3,000
Range
0-0
0-0
800 - 14,200

Surface Water
0.002
0.001
0
0
0-0
0-0
0
0
0-0
0-0
Occurrence Summary and Use Support Document
162
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
1
Range
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
<100
Range
0-0

Combined Ground
& Surface Water
0.002
0.001
0.0002
0
0
24
0-0
0-0
15-35
0
0
3,100
0-0
0-0
800 - 14,900
3.4.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of chlordane in drinking water are also
reviewed. Previously compiled occurrence chlordane information, from an OGWDW summary
document entitled "Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air in
the United States of America" (USEPA, 1989), is presented in following section. This variety of regional
studies and information are presented regarding levels of chlordane in drinking water. (No national
studies were included in the review.)  Note that none of the studies presented in the following section
provide the quantitative analytical results or comprehensive coverage that would enable direct
comparison to the occurrence findings estimated with the cross-section occurrence data presented in
Section 3.4.4. These additional studies, however, do enable a broader assessment of the Stage 2
occurrence estimates presented for this Six-Year Review. All the following information in Section 3.4.5
is taken directly from "Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air
in the United States of America" (USEPA, 1989).

3.4.5.1 Ground Water Sources

Irwin and Healy (1978,  as cited in USEPA, 1989) summarized data collected in  1976 during a water
quality reconnaissance of public water supplies in Florida.  None of the 100 ground water supplies
sampled, representative of the 5 aquifers in Florida, contained measurable levels of chlordane (no
detection limit was reported).

Less than 8 percent of samples from 96 locations utilizing the Floridian aquifer were found positive for
chlordane in a 1984 study by the Florida Department of Environmental Resources and the U.S.
Geological Survey (Holden, 1986, as  cited in USEPA, 1989). These supplies serve over 3 million
people. No other information from the study was reported.

No positive samples were found for chlordane in the following two studies; one involved the 1984-1985
sampling of 42 sites from 12 towns in Connecticut  and the other involved the analysis of 67 samples
from Long Island, New York, in 1984 (Waggoner,  1985; Holden, 1986, all as cited in USEPA, 1989).
The Connecticut drinking water wells serve a population of over 570,000, and the detection limit for that
study was 3.3 |ig/L. No other information about either study  was reported.

Two positive samples (one for alpha-chlordane and one for gamma-chlordane) out of 107 samples
analyzed were found for ground water wells in Idaho (Idaho Department of Health and Welfare, 1984, as
cited in USEPA, 1989). Monitoring for pesticides  in drinking water wells is not routinely done; the
sampling performed was in response to a particular contamination incident, not for any comprehensive
Occurrence Summary and Use Support Document
163
March 2002

-------
monitoring program.  A relatively low mean of 0.0002 |ig/L was reported for these two samples, with one
sample having a chlordane concentration of 0.038 |ig/L (no detection limit was reported). It appears that
"negative" samples, possibly assigned a detection limit value, were included in the calculation of the
mean.

Benvenue et al. (1972b, as cited in USEPA, 1989) conducted a study to determine the extent of
organochlorine pesticide contamination of drinking water in Hawaii.  A total of 45 finished drinking
water samples were collected from February 1971 to May 1971. Concentrations of chlordane were
detected in four of the samples, with a range of positive values between 0.0005 and 0.005 |ig/L, and an
average concentration of 0.001 |ig/L (no detection limit was reported).

3.4.5.2 Surface Water Sources

Irwin and Healy (1978, as cited in USEPA,  1989), summarizing data collected during a water quality
reconnaissance of public water supplies in Florida, reported that none of the 16 surface water supplies
sampled contained chlordane in excess of the detection limit.  No detection limit was reported.

In the New Orleans Water Supply Study conducted by USEPA Region VI (USEPA, 1975a, as cited in
USEPA,  1989), samples  of drinking water were analyzed for levels of halogenated organics.  Although
the number of positive samples was not reported, the concentrations of chlordane in three samples
analyzed ranged from "non-detected" to less than 0.1 |ig/L. No detection limit for the analyses was
reported.

To assemble a database which would reflect the status of Great Lakes drinking water quality, the
Canadian Public Health Association gathered data from October 1984 through August 1985 (Canadian
Public Health Association, 1986, as cited in USEPA, 1989). The data collected covered the period from
the mid 1970s to early 1985. The study was funded by the Health Protection Branch of Health and
Welfare Canada and the  Ontario Ministry of the Environment. A research team, appointed by the
Association, reviewed data on the quality of water at 31 representative Canadian and United States
communities and 24 offshore sites to evaluate the human health implications.

For each of the 31 communities, data consisted of: 1) background information on the community; 2)
treatment plant schematics and associated treatment process information; and 3) water quality data.
Water sample types included raw water (treatment plant intake), distribution water (treated water), and
tap water. Water quality data collected included general parameters (e.g., alkalinity, turbidity),
microbiological and radiological parameters, inorganic parameters, and organic parameters (including
volatiles, base/neutrals, pesticides and PCBs, and phenols and acids). For each parameter, the water
type, time period,  concentration (mean and range), number of samples, and detection limit are presented.

For most of the volatile organics, including  chlordane, the available data indicated that there were very
low levels of these contaminants in the raw, treated, or tap water. Most of the  values found were "not
detected" or near the detection limit (Canadian Public Health Association, 1986, as cited in USEPA,
1989).

3.4.5.3 Unspecified Sources

A Region V Survey (USEPA, 1975b, as cited in USEPA, 1989), conducted during the first three months
of 1975, assessed levels of organic chemicals in samples of raw and finished drinking water collected
from 83 utilities in Indiana, Illinois, Michigan, Minnesota, Ohio, and Wisconsin.  Samples were analyzed
Occurrence Summary and Use Support Document         164                                     March 2002

-------
for various pollutants including gamma-chlordane. One sample of finished drinking water contained
gamma-chlordane at a level of 0.004 |ig/L.  No limits of detection for pesticides were reported.

In a study to identify the sources of pollutants entering a sewage treatment plant, Levins et al. (1979, as
cited in USEPA, 1989) collected samples from two drinking water sources. Although no detection limit
was reported, the analyses detected no chlordane in any of the samples.

Schafer et al. (1969, as cited in USEPA, 1989) reported concentrations of chlordane up to 8 |ig/L for
finished drinking water, although no location(s) were given. No other information on this study was
reported.

3.4.6  Conclusion

Chlordane is a synthetic chemical that was used in the 1980s as a pesticide in agriculture and other areas.
Today, chlordane's only use is as fire ant control in power transformers. As  of 1992, chlordane was
produced solely by Velsicol Chemical Corp., and its production has greatly decreased since its
cancellation.  Industrial releases of chlordane were been reported to TRI from 1988 to 1996 from 10
States. Chlordane ambient occurrence was not analyzed in any available ambient studies. The Stage 2
analysis, based on the  16-State cross-section, estimated that zero percent of combined ground water and
surface water systems  serving zero percent of the population exceeded the MCL of 0.002 mg/L.  Based
on this estimate, zero PWSs nationally are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for chlordane  from 10  States, including 5
of 16 cross-section States. Chlordane use has been virtually eliminated throughout the U.S.  The only
producer of chlordane  is located in Tennessee, which is not one of the 16 cross-section States.  However,
the cross-section should adequately represent the occurrence of chlordane on a national scale based upon
the use, production, and release patterns of the 16-State cross-section in relation to the patterns observed
for all 50 States.

3.4.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1994. Toxicological Profile for
       Chlordane.  U.S.  Department of Health and Human Services, Public Health Service. 234 pp. +
       Appendices. Available on the Internet at http://atsdrl .atsdr.cdc.gov/toxprofiles/tp31 .pdf

Barks, J.H. 1978.  Water Quality in the OzarkNational Scenic Riverways, Missouri.  Washington, DC:
       U.S. Geological Survey, U.S. Department of Interior.  Geological Survey Water-Supply Paper
       2048.

Benvenue, A., J.N. Ogata, and J.W. Hylin.  1972b. Organochlorine pesticides in rainwater, Oahu,
       Hawaii, 1971-1972. Bull.  Environ. Contam.  Toxicol.  v. 8, no. 4, pp. 238-241.

Canadian Public Health Association.  1986. Comprehensive Survey of the Status of Great Lakes
       Drinking Water.  Prepared in cooperation with Health and Welfare Canada and the Ministry of
       the Environment. Ottawa, Canada: Canadian Public Health Association. August 1986.
Occurrence Summary and Use Support Document          165                                     March 2002

-------
Carey, A.E., and F.W. Kutz.  1983. Trends in Ambient Concentrations of Agrochemicals in Humans and
       the Environment of the United States. Presented at the International Conference on
       Environmental Hazards of Agrochemicals in Developing Countries, Alexandria, Egypt,
       November 8-12, 1983.

Englande, A.J., J.K. Smith, and J.N. English. 1978. Potable water quality of advanced wastewater
       treatment plant effluents. Prog. Wat. Tech.  v. 10, no. !/2, pp. 17-39.

Gilliom, R.J., R.B. Alexander, and R.A. Smith. 1985.  Pesticides in the Nation's Rivers, 1975-1980, and
       Implications for Future Monitoring.  U.S. Geological Survey Water-Supply Paper 2271. U.S.
       Government Printing Office, U.S. Department of Interior.

Holden, P.W. 1986. Pesticides and Groundwater Quality. Issues and Problems in Four States.
       Prepared for the Board of Agriculture, National Research Council. Washington, DC: National
       Academy  Press.

Idaho Department of Health and Welfare. 1984. Letter to Charles Berry, Office of Drinking Water, U.S.
       Environmental Protection Agency, summarizing data on groundwater contamination incidents.
       Sent by: Charles D. Brokopp, State Epidemiologist, Division of Health, Idaho Department of
       Health and Welfare.

Irwin, G.A., and H.G. Healy. 1978. Chemical and Physical Quality of Selected Public Water Supplies in
       Florida, August-September 1976. Tallahassee, FL: Water  Resources Division, U.S. Geological
       Survey. USGS/WRI 78-21.

Levins, P., J. Adams, P. Brenner, S. Coons, K. Thrun, and J. Varone. 1979. Sources of Toxic Pollutants
       Found in Influents to Sewage Treatment Plants.  IV. R.M. Clayton Drainage Basin, Atlanta
       report. Prepared by Arthur D. Little, Inc., for Office of Water Planning and Standards, U.S.
       Environmental Protection Agency, Washington, DC. EPA Contract No. 68-01-3857.

National Safety Council (NSC). 2001. Chlordane Chemical Backgrounder.  Itasca, IL: National Safety
       Council. Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?id=90&chempath=chemicals, accessed
       July 18, 2001.

Ramlit Associates, Inc.  1983. Groundwater Contamination by Pesticides: A California Assessment.
       Submitted to: State Water Resources Control Board, Sacramento, CA. Submitted by: Ramlit
       Associates, Inc., Berkeley, CA. Publication No. 83-4SP.

Schafer, M.L., et al.  1969. Pesticides in drinking water. Environ. Sci. Technol.  v. 3, p. 1261.

Truhlar, J.F., and L.A. Reed. 1976. Occurrence of pesticide residues in four streams draining different
       land-use areas in Pennsylvania, 1969-1971. Pestic. Monit. J. v. 10, no.3, pp.  101-110.

Tucker, R.K., and  T.A. Burke. 1978.  A Second Preliminary Report on the Findings of the State
       Groundwater Monitoring Project. New Jersey: Department of Environmental Protection.

USEPA.  1975a. Analytical Report: New Orleans Water Supply Study.  Region VI, EPA.
       EPA-906/9-75-003.
Occurrence Summary and Use Support Document          166                                     March 2002

-------
USEPA.  1975b. Preliminary Assessment of Suspected Carcinogens in Drinking Water. Washington,
       DC: Office of Toxic Substances, USEPA.

USEPA.  1989. Draft Final Report on the Occurrence and Human Exposure to Pesticides in Drinking
       Water, Food, and Air in the United States of America.  Office of Drinking Water, USEPA.
       September, 1989.

USEPA.  2000. TRIExplorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on: Chlordane.
       Office of Ground Water and Drinking Water, USEPA. Available on the Internet at
       http://www.epa.gov/safewater/dwh/c-SOC/chlordan.html, last updated April 12, 2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Waggoner, P.E.  1985.  Memorandum from the Connecticut Agricultural Experiment Station on
       Pesticides in  Connecticut Groundwater. P. Gough, (ed.). New Haven, CT: News of Science.
Occurrence Summary and Use Support Document         167                                    March 2002

-------
3.5    l,2-Dibromo-3-Chloropropane (DBCP)
Table of Contents

3.5.1 Introduction, Use and Production  	  169
3.5.2 Environmental Release  	  169
3.5.3 Ambient Occurrence  	  170
3.5.4 Drinking Water Occurrence Based on the 16-State Cross-Section	  171
3.5.5 Additional Drinking Water Occurrence Data 	  176
3.5.6 Conclusion	  177
3.5.7 References 	  177
Tables and Figures

Table 3.5-1:  Environmental Releases (in pounds) for l,2-Dibromo-3-Chloropropane in the
       United States, 1991-1992 	  170

Table 3.5-2:  Stage 1 DBCP Occurrence Based on 16-State Cross-Section - Systems	  171

Table 3.5-3:  Stage 1 DBCP Occurrence Based on 16-State Cross-Section - Population	  172

Table 3.5-4:  Stage 2 Estimated DBCP Occurrence Based on 16-State Cross-Section - Systems  ....  173

Table 3.5-5:  Stage 2 Estimated DBCP Occurrence Based on 16-State Cross-Section - Population . .  174

Table 3.5-6:  Estimated National DBCP Occurrence - Systems and Population Served   	  175
Occurrence Summary and Use Support Document         168                                   March 2002

-------
3.5.1  Introduction, Use and Production

l,2-Dibromo-3-chloropropane, also known as DBCP, is a manufactured chemical not found naturally in
the environment. It is a colorless liquid with a sharp smell that can be tasted in water at very low
concentrations.  Most DBCP that enters surface water evaporates into the air within several days or a
week. Some of what is spilled on or applied to soil moves through the soil into the groundwater, where it
may remain for a long time.  Trade names for pesticides containing DBCP are Nemagon, Nemafume,
Fumazone, Fumagon, andNemazon (ATSDR, 1992).

l,2-Dibromo-3-chloropropane is a manufactured chemical and is not found naturally in the environment
(ATSDR,  1992). It is used as an intermediate in the synthesis of organic chemicals, such as the
brominated flame retardanttris[(2,3-dibromopropyl)phosphate] (Verschueren, 1983, as cited in ATSDR,
1992). Until  1977, DBCP was extensively used as a soil fumigant and nematocide on over 40 different
crops  in the United States (Anonymous, 1988, as cited in ATSDR, 1992). The chemical was used to
protect field crops, vegetables, fruits and nuts, nursery and greenhouse crops, and turf from pests (NTP,
1985,  as cited in ATSDR, 1992).

From  1977 to 1979, EPA suspended registration of products containing DBCP except for use on
pineapples in Hawaii (Anonymous 1988; USEPA  1977, 1979; all as cited in ATSDR, 1992). In 1985,
EPA issued an intent to cancel all registrations for DBCP-containing pesticide products, including use on
pineapples. Subsequently, the use of existing stocks of DBCP on pineapples was prohibited (USEPA
1985a, 1985b, all as  cited in ATSDR,  1992).

Prior to the cancellation of pesticide uses, DBCP was used extensively; 9.8 million pounds of DBCP
were applied in 1974. In California, 831,000 pounds of the chemical were applied, mainly on grapes  and
tomatoes, during 1977 (NTP, 1985, as cited in ATSDR, 1992). The volume of DBCP applied to
pineapple fields in Hawaii between 1979  and  1985 was probably high, since during much of that time,
the chemical was the preferred fumigant for use on pineapple fields (Albrecht, 1987, as cited in ATSDR,
1992).

DBCP was first produced commercially in the United States in 1955. In 1969, U.S. production was 8.58
million pounds (IARC, 1979, as cited in ATSDR,  1992). Estimates of annual production during 1974 to
1975 ranged from 18 to 20 million pounds (IARC, 1979; NTP, 1985, all as cited in ATSDR,  1992).
DBCP is no longer commercially manufactured in the United States (Hawley,  1981; Sax and Lewis,
1987,  all as cited in ATSDR, 1992). RW. Greeff & Co., Inc., in Old Greenwich, Connecticut, is listed as
a current supplier of DBCP for domestic research purposes (OPD, 1989, as cited in ATSDR, 1992). It is
not known whether this supplier produces DBCP in the United States or imports the chemical.

Two companies were listed as producers of DBCP in 1977 (USEPA, 1989a, as cited in ATSDR, 1992).
The production volume of Columbia Organic  Chemicals Co., in Columbia, South Carolina, was listed as
less than 1,000 pounds. No production volume was listed for the other producer, Velsicol Chemical
Corp., in El Dorado, Arkansas. As DBCP is no longer used as a fumigant and nematocide in the United
States, it is likely that its current production volume in the United States, if any, is very low.

3.5.2  Environmental Release

l,2-Dibromo-3-chloropropane is listed as a Toxics Release Inventory (TRI) chemical. Table 3.5-1
illustrates the environmental releases for DBCP from 1991 and 1992.  (There are only DBCP data for
these years.) Air emissions are the only recorded releases of DBCP, and, given that the data is only for
two years, a trend is  not self-evident. No surface water discharges, underground injection, releases to

Occurrence Summary and Use Support Document         169                                    March 2002

-------
land (such as spills or leaks within the boundaries of the reporting facility), or off-site releases (including
metals or metal compounds transferred off-site) were reported for DBCP. These TRI data for DBCP
were reported in Ohio and Rhode Island, neither of which are included in the 16-State cross-section (used
for analyses of DBCP occurrence in drinking water; see Section 3.5.4). (For a map of the 16-State cross-
section, see Figure 1.3-1.)
Table 3.5-1: Environmental Releases (in pounds) for l,2-Dibromo-3-Chloropropane in the United
States, 1991-1992
Year
1992
1991
On-Site Releases
Air Emissions
294
580
Surface Water
Discharges
—
—
Underground
Injection
—
—
Releases
to Land
—
—
Off-Site Releases
—
—
Total On- &
Off-site
Releases
294
580
 Source: USEPA, 2000
3.5.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers.  There are few available data on the occurrence of DBCP in ambient waters of the United
States. The most comprehensive and nationally consistent data describing ambient water quality in the
United States are being produced through the United States Geological Survey's (USGS) National Water
Quality Assessment (NAWQA) program. However, national NAWQA data are currently unavailable for
DBCP. Additional studies of ambient data are summarized below.

3.5.3.1 Additional Ambient Occurrence Data

A summary document entitled ""Occurrence and Human Exposure to Pesticides in Drinking Water,
Food, and Air in the United States of America" (USEPA,  1989b)), was previously prepared for past
USEPA assessments of pesticides. Regional studies reviewed in that document were the only
information found that addressed levels of DBCP in water other than drinking water or water not
specified as drinking water. The following information is taken directly from "Occurrence and Human
Exposure to Pesticides in Drinking Water, Food, and Air in the United States of America" (USEPA,
1989b).

Several regional studies (Peoples et al., 1980; Nelson et al., 1981; Pinto, 1980; Cohen, 1981; Carter and
Riley, 1981; and Mink, 1981, all as cited in Cohen et al, 1984 and all as cited in USEPA, 1989b) and the
EPA pesticide registration files (as reported in Cohen et al., 1984 and as cited in USEPA, 1989b)
provided monitoring data on levels of DBCP in ground water.  Positive concentrations of DBCP ranging
from 0.02 |ig/L to 20 |ig/L were detected in samples of ground water in Hawaii, California, Arizona,
South Carolina, and Maryland. Areas with the highest concentrations were the San Joaquin Valley in
California and the region southwest of Phoenix, Arizona.  The number of samples analyzed, the number
of positive samples detected, and the detection limit were  not reported.

Holden (1986, as cited in USEPA, 1989b) reported that, in a massive survey of more than 8,000 public
and private wells in the San Joaquin Valley of California,  detectable levels of DBCP were found in
approximately 25% of the wells tested.  The highest level  of DBCP found in the survey was 1,240 |ig/L.
Occurrence Summary and Use Support Document
170
March 2002

-------
Samples of ground water collected from seven previously unsampled sites in Hawaii showed detectable
levels of DBCP in the range of 0.05 |ig/L to 0.5 |ig/L (USEPA, 1984, as cited in USEPA, 1989b).  Some
of the sites have been used as drinking water supplies. The number of samples tested, the number of
positive samples, and the detection limit were not reported.

3.5.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of DBCP occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest: 0.0002 mg/L; 0.0001 mg/L; and 0.00002 mg/L.

All sixteen cross-section State data sets, with the exception of Texas, contained occurrence data for 1,2-
dibromo-3-chloropropane. These data represent more than 98,000 analytical results from approximately
14,000 PWSs during the period from  1984 to 1998 (with most analytical results from 1992 to  1997). The
number of sample results and PWSs vary by State, although the State data sets have been reviewed and
checked to ensure adequacy of coverage and completeness. The overall modal detection limit for DBCP
in the 16 cross-section States is equal to 0.00002 mg/L. (For details regarding the 16-State cross-section,
please refer to Section  1.3.5 of this report.)

3.5.4.1 Stage 1 Analysis Occurrence Findings

Table 3.5-2 illustrates the Stage 1 analysis of DBCP occurrence in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds:  0.0002 mg/L (the current MCL),
0.0001, and 0.00002 mg/L (the modal MRL). A total of 128 (approximately 0.912% of) ground water
and surface water PWSs had analytical results exceeding the MCL (0.0002 mg/L); 1.08% of systems
(152 systems) had results exceeding 0.0001 mg/L. When evaluated relative to athreshold of 0.00002
mg/L, the  percentage of systems increased to 1.50% (211 systems).

Approximately 0.861% of ground water systems (112 systems) had any analytical results greater than the
MCL (0.0002 mg/L). About 1.02% of ground water systems (132 systems) had results above 0.0001
mg/L.  The percentage of ground water systems with at least one result greater than 0.00002 mg/L was
equal to 1.30% (169 systems).

A total of 16 (1.55% of) surface water systems had results greater than  0.0002 mg/L. Twenty (1.93% of)
surface water systems had at least one analytical result greater than  0.0001 mg/L. Approximately 4.06%
of surface water systems (42 systems) had results  exceeding 0.00002 mg/L.
Table 3.5-2:  Stage 1 DBCP Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Systems
Exceeding Threshold
0.861%
1.02%
1.30%
Number of Systems
Exceeding Threshold
112
132
169

Occurrence Summary and Use Support Document
111
March 2002

-------
Source Water Type
Surface Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Systems
Exceeding Threshold
1.55%
1.93%
4.06%
Number of Systems
Exceeding Threshold
16
20
42

Combined Ground &
Surface Water
0.0002
0.0001
0.00002
0.912%
1.08%
1.50%
128
152
211
Reviewing DBCP occurrence in the 16 cross-section States by PWS population served (Table 3.5-3)
shows that approximately 11.4% of the population served by PWSs in the 16 States (almost 10 million
people) was served by PWSs with at least one analytical result of DBCP greater than the MCL.
Approximately 10.3 million (11.8% of) people were served by systems with an exceedance of 0.0001
mg/L.  Over 11 million (13.0% of) people were served by systems with at least one analytical result
greater than 0.00002 mg/L.

The percentage of population served by ground water systems with at least one analytical result greater
than 0.0002 mg/L was equal to 3.41% (about 1.3 million people).  When evaluated relative to 0.0001
mg/L and 0.00002 mg/L, the percent of population exposed was equal to 3.78% (almost 1.5 million
people) and 5.36% (over 2 million people), respectively.

The percentages of population served by surface water systems with exceedances of all thresholds are
much greater than the ground water percentages of exceedance.  About 17.7% of the population (over 8.6
million people) was served by surface water systems with at least one analytical result greater than the
MCL. Approximately 18.1% of the population served by surface water systems (over 8.8 million people)
were exposed to DBCP concentrations greater than 0.0001 mg/L.  When evaluated relative to 0.00002
mg/L, the percent of population exposed was equal to 19.1% (almost 9.4 million people).
Table 3.5-3:  Stage 1 DBCP Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Population
Served by Systems
Exceeding Threshold
3.41%
3.78%
5.36%
Total Population Served by
Systems Exceeding
Threshold
1,322,000
1,466,500
2,080,200

Surface Water
0.0002
0.0001
0.00002
17.7%
18.1%
19.1%
8,638,200
8,843,800
9,352,000
Occurrence Summary and Use Support Document
172
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population
Served by Systems
Exceeding Threshold
Total Population Served by
Systems Exceeding
Threshold

Combined Ground &
Surface Water
0.0002
0.0001
0.00002
11.4%
11.8%
13.0%
9,960,300
10,310,200
11,432,200
3.5.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.5-4 and 3.5-5.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

About 1.41% (approximately 199) of ground water and surface water PWSs had estimated mean
concentrations of DBCP greater than the MCL (0.0002 mg/L). Approximately 273 (1.94% of) systems in
the 16 States had mean concentrations of DBCP exceeding 0.0001 mg/L.  When evaluated relative to
0.00002 mg/L, the percentage of systems exceeding the threshold increased to 4.09% (an estimated 574
systems).

Approximately 182 (1.40% of) ground water systems in the 16 States have an estimated mean
concentration of DBCP greater than 0.0002 mg/L. An estimated 1.91% of ground water systems (about
249 systems) had mean concentration values greater than 0.0001 mg/L. A total of 517 (3.97% of) PWSs
had estimated mean concentration values of DBCP greater than 0.00002 mg/L.

Only  17 (about 1.64% of) surface water systems in the 16 States had estimated mean concentrations
greater than the MCL. Approximately 24 (2.33% of) surface water systems had estimated  mean
concentration values greater than 0.0001 mg/L.  Fifty-seven (approximately 5.56% of) surface water
systems in the  16 States had estimated mean concentration values of DBCP greater than 0.00002 mg/L.
Table 3.5-4:  Stage 2 Estimated DBCP Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
1.40%
1.91%
3.97%
Range
1.19% -1.61%
1.65% -2. 18%
3.41% -4.52%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
182
249
517
Range
155-210
214-284
443 - 588
Occurrence Summary and Use Support Document
173
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
Range
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
Range

Surface Water
0.0002
0.0001
0.00002
1.64%
2.33%
5.56%
1.06% -2. 32%
1.55% -3. 19%
4.06% - 7.06%
17
24
57
11-24
16-33
42-73

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
1.41%
1.94%
4.09%
1.22% -1.65%
1.70% -2.21%
3. 53% -4.63%
199
273
574
171-231
238-310
496 - 650
Reviewing DBCP occurrence by PWS population served (Table 3.5-5) shows that approximately 2.60%
of population served by all PWSs in the 16 cross-section States (an estimate of approximately 2.3 million
people) were potentially exposed to DBCP levels above 0.0002 mg/L.  The percentage of population
served by PWSs in the 16 States with mean levels of DBCP above 0.0001 mg/L and 0.00002 mg/L were
3.22% (about 2.8 million people) and 5.75% (over 5 million people), respectively.

When the percent of population served by ground water systems in the 16 States was evaluated relative to
a threshold of 0.0002 mg/L, 0.0001 mg/L, and 0.00002 mg/L, the percentage of population exposed in the
16 cross-section States was equal to 3.27% (an estimated 1.27 million people), 4.03% (an estimated 1.56
million people) and 6.21% (an estimated 2.4 million people), respectively.

The percentage of population served by surface water systems in the 16 States with mean levels above
0.0002 mg/L was equal to 2.07% (over 1 million people), and the percentage of population served with
levels above 0.0001 mg/L was 2.58% (an estimated 1.26 million people). The percentage of the
population served by surface water systems with levels above 0.00002 mg/L was 5.39% (over 2.6 million
people).
Table 3.5-5:  Stage 2 Estimated DBCP Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
3.27%
4.03%
6.21%
Range
2.71% -4.20%
3.40% -5.38%
4.92% -9. 98%
Total Population Served by Systems in the 16
States Estimated to Exceed Threshold
Best Estimate
1,267,200
1,563,900
2,409,100
Range
1,050,000-1,629,900
1,318,400-2,086,000
1,907,200-3,869,100

Surface Water
0.0002
0.0001
0.00002
2.07%
2.58%
5.39%
1.51% -3. 97%
1.60% -5.24%
2.74% - 14.2%
1,011,100
1,264,600
2,637,300
738,000-1,941,900
782,100-2,565,400
1,340,000-6,929,800
Occurrence Summary and Use Support Document
174
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the 16
States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
2.60%
3.22%
5.75%
2. 11% -3. 77%
2.49% - 4.96%
3.99% -10.4%
2,278,300
2,828,300
5,046,100
1,853,700-3,307,300
2,182,700-4,353,900
3,503,800 - 9,149,900
3.5.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 920 PWSs nationally serving approximately 5.5 million people could be exposed
to DBCP concentrations above 0.0002 mg/L. About 1,263 systems serving almost 6.9 million people
were estimated to have mean concentrations greater than 0.0001 mg/L. Approximately 2,658 systems
serving over 12 million people nationally were estimated to have DBCP concentrations greater than
0.00002 mg/L. (See Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to
national values.)

For ground water systems, about 830 PWSs serving almost 2.8 million people nationally were estimated
to have mean concentrations greater than 0.0002 mg/L. Approximately 1,136 systems serving about 3.5
million people nationally were estimated to have a mean concentration value of DBCP that exceeded
0.0001 mg/L. About 2,360 ground water systems serving over 5.3 million people were estimated to have
mean concentrations greater than 0.00002 mg/L.

Approximately 92 surface water systems serving about 2.6 million people were estimated to have mean
concentrations of DBCP above 0.0002 mg/L. An estimated 130 surface water systems serving almost 3.3
million people were estimated to have mean concentrations greater than 0.0001 mg/L. About 311 surface
water systems serving almost 6.9 million people were estimated to have mean concentrations greater than
0.00002 mg/L.
Table 3.5-6:  Estimated National DBCP Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
830
1,136
2,360
Range
709 - 959
978 - 1,298
2,025 - 2,687
Total Population Served by Systems Nationally
Estimated to Exceed Threshold
Best Estimate
2,799,200
3,454,700
5,321,700
Range
2,319,400-3,600,300
2,912,300-4,608,000
4,213,000 - 8,546,700

Surface Water
0.0002
0.0001
0.00002
92
130
311
59-130
86 - 178
227 - 395
2,630,600
3,290,100
6,861,600
1,920,100 - 5,052,300
2,034,700 - 6,674,500
3,486,200-18,029,400
Occurrence Summary and Use Support Document
175
March 2002

-------
Source Water Type
Threshold
(mg/L)
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems Nationally
Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
920
1,263
2,658
792-1,070
1,102 - 1,436
2,297 - 3,010
5,531,800
6,867,400
12,252,200
4,500,900 - 8,030,400
5,299,600 - 10,571,600
8,507,500-22,216,800
3.5.5  Additional Drinking Water Occurrence Data

Several regional studies reviewed in the summary document entitled "Occurrence and Human Exposure
to Pesticides in Drinking Water, Food, and Air in the United States of America" (USEPA, 1989b)
addressed levels of DBCP in samples of drinking water obtained from ground water and surface water
sources. Note that the study presented in the following section does not provide the quantitative
analytical results or comprehensive coverage that would enable direct comparison to the occurrence
findings estimated with the cross-section occurrence data presented in Section 3.5.4.  This additional
study, however, does enable a broader assessment of the Stage 2 occurrence estimates presented for this
Six-Year Review. All the following information in Section 3.5.5 is taken directly from "Occurrence and
Human Exposure to Pesticides in Drinking Water, Food and Air in the United States of America"
(USEPA, 1989b).

3.5.5.1 Ground Water Sources

Carey and Kutz (1983, as cited in USEPA, 1989b) presented data collected in 1979 on levels of DBCP in
ground water samples collected in California, Arizona, Texas, South Carolina, and Alabama. Eight
positive ground water samples collected from California and South Carolina showed DBCP
concentrations ranging from 0.01 |ig/L to  10.8 |ig/L, and an average concentration of 2.8 |ig/L. Seven of
the positive samples were from private wells; one sample containing 0.01 |ig/L was from a municipal
well.  The number of ground water samples tested was not reported.

3.5.5.2 Surface Water Sources

Carey and Kutz (1983, as cited in USEPA, 1989b) also reported on levels of DBCP in samples of surface
water collected in California, Arizona, Texas, South Carolina, and Alabama in  1979. Two positive
samples collected from municipal water supplies in California showed DBCP concentrations of 0.09
l-ig/L and 0.1 |ig/L.  The number of surface water samples tested was not reported.

To assemble a database which would reflect the status of Great Lakes drinking  water quality, the
Canadian Public Health Association gathered data from October 1984 through August 1985 (Canadian
Public Health Association, 1986, as cited in  USEPA, 1989b). The data collected covered the period from
the mid 1970s to early 1985. The study was funded by the Health Protection Branch of Health and
Welfare Canada and the Ontario Ministry of the Environment. A research team, appointed by the
Association, reviewed data on the quality  of water at 31 representative Canadian and United States
communities and 24 offshore sites to evaluate the human health implications.
Occurrence Summary and Use Support Document
176
March 2002

-------
For each of the 31 communities, data consisted of: 1) background information on the community; 2)
treatment plant schematics and associated treatment process information; and 3) water quality data.
Water sample types included raw water (treatment plant intake), distribution water (treated water), and
tap water.  Water quality data collected included general parameters (e.g., alkalinity, turbidity),
microbiological and radiological parameters, inorganic parameters, and organic parameters (including
volatiles, base/neutrals, pesticides and PCBs, and phenols and acids). For each parameter, the water
type, time  period, concentration (mean and range), number of samples, and detection limit were recorded.

For most of the volatile organics, including DBCP, the available data indicated that these contaminants
did not occur at significant levels in the raw, treated, or tap water. Most of the values found were "not
detected" or near the detection limit (Canadian Public Health Association, 1986, as cited in USEPA,
1989b).

3.5.6  Conclusion

DBCP is a pesticide that was used and produced extensively in the United States from 1955 to 1977,
when its use was restricted to pineapples in Hawaii.  All uses were subsequently canceled in 1985.  Until
its restriction and cancellation, DBCP was extensively used as a soil fumigant and nematocide on over 40
different crops in the United  States. There are no recent statistics available regarding use or production
of DBCP,  and any production is probably minimal. Industrial releases of DBCP were reported to TRI in
1991 and 1992 from two States.  DBCP was not an analyte for any of the ambient occurrence studies.
The Stage 2 analysis, based on the 16-State cross-section, estimated that approximately  1.41% of
combined  ground water and surface water systems serving 2.60% of the population exceeded the MCL of
0.0002 mg/L. Based on this estimate, a national estimate of 920 PWSs, serving approximately 5,531,800
people, are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as  suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for DBCP from 2 States, not including any
of the cross-section States. The use of DBCP has been canceled nationwide, and any production of
DBCP is projected to be very minimal. The cross-section should adequately represent the occurrence of
DBCP on a national scale based upon the use, production, and release patterns of the 16-State cross-
section in relation to the patterns observed for all 50  States.

3.5.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1992. Toxicological Profile for 1,2-
       Dibromo-3-chloropropane. U.S. Department of Health and Human Services, Public Health
        Service.  152 pp. + Appendices. Available on the Internet at
       http://www.atsdr.cdc .gov/toxprofiles/tp3 6 .pdf

Albrecht, W.N.   1987.   Occupational exposure to 1,3-dichloropropene (Telone II@) in Hawaiian
       pineapple culture.  Archives of Environmental Health, v.  42, pp. 286-291.

Anonymous. 1988.  DBCP.  Reviews of Environmental Contamination and Toxicology, v. 104,
       pp. 73-91.

Canadian Public Health Association. 1986. Comprehensive Survey of the Status of Great Lakes
       Drinking Water. Prepared in cooperation with Health and Welfare Canada and the Ministry of
       the Environment. Ottawa, Canada: Canadian Public Health Association. August, 1986.

Occurrence Summary and Use Support Document         177                                     March 2002

-------
Carey, A.E., and F.W. Kutz. 1983. Trends in Ambient Concentrations of Agrochemicals in Humans and
       the Environment of the United States. Presented at the International Conference on
       Environmental Hazards of Agrochemicals in Developing Countries, Alexandria, Egypt,
       November 8-12, 1983.

Carter, G.E., and M.B.  Riley.  1981.  Pestic. Monit. J. v.  15, pp. 139-142.

Cohen, S.Z. 1981.  Summary Report-DBCP in Groundwater in the Southeast.  Washington, DC:
       Hazard Evaluation Division, U.S. Environmental Protection Agency.

Cohen, S.Z., S.M. Creeger, R.F. Carsel, and C.G. Enfield. 1984. Potential for Pesticide Contamination
       in Groundwater Resulting from Agricultural Uses. In: Treatment and Disposal of Pesticide
       Wastes.  R.F. Kruger and J.W. Seiber (eds.).  Washington,  DC: American Chemical Society.
       ACS Symposium Series No. 259: pp. 247-325.

Hawley, G.G. 1981. The Condensed Chemical Dictionary. lOthed. New York, NY: Van Nostrand
       ReinholdCo., p. 325.

Holden, P.W.  1986. Pesticides and Groundwater Quality.  Issues  and Problems in Four States.
       Prepared for the Board of Agriculture, National Research Council. Washington, DC: National
       Academy Press.

International Agency for Research on Cancer (IARC).  1979. Monograph on the evaluation of the
       carcinogenic risk of chemicals to humans. Vol. 20: Some halogenated hydrocarbons. Lyon,
       France: World  Health Organization, pp. 1-40, 83-96.

Mink, J.F. 1981. DBCP and EDB in soil and water at Kuhia, Oahu, Hawaii.  Prepared for Del Monte
       Corp., Honolulu, Hawaii.

National Toxicology Program (NTP). 1985. Fourth Annual Report on Carcinogens. National
       Toxicology Program, pp. 179-182.

Nelson, S., M. Iskander, M. Volz, S. Khalifa, and R Haberman. 1981. Sci. Total. Environ, v. 21,
       pp.35-40.

OPD. 1989. OPD Chemical Buyers Directory. 76th ed.  New York, NY: Schnell Publishing Co. Inc.,
       pp. 235, 696.

Peoples, S.A., K.T.  Maddy, W. Cusick, T. Jackson, C. Cooper, and A.S. Frederickson.  1980. Bull.
       Environ. Contam. Toxicol. v. 24, pp. 611-618.

Pinto, E.  1980.  Report of Groundwater Contamination Study in Wicomico, Maryland.  Wicomico
       County Health Department, State of Maryland (plus 1981-1982 addenda).

Sax, N.I., and R.J. Lewis. 1987.  Hawley's Condensed Chemical Dictionary.  10th Ed.  New York, NY:
       Van Nostrand Reinhold Co., p. 367.

USEPA.  1977.  U.S. Environmental Protection Agency: Southern Agricultural Insecticides, Inc., et al.
       Suspension order. Federal Register 42, pp. 57543-57548.
Occurrence Summary and Use Support Document         178                                    March 2002

-------
USEPA.  1979. U.S. Environmental Protection Agency: Dibromochloropropane (DBCP); Suspension
       order and notice of intent to cancel. Federal Register 44, pp. 65135-65176.

USEPA.  1984. Proposal for comment on intent to cancel use of dibromochloropropane in Hawaii.
       Federal Register 46, p. 19594.

USEPA.  1985a. U.S. Environmental Protection Agency: Dibromochloropropane; Intent to cancel
       registrations of pesticide products containing dibromochloropropane (DBCP). Federal Register
       50, pp. 1122-1123.

USEPA.  1985b. U.S. Environmental Protection Agency: Dibromochloropropane; Denial of use of
       existing stocks.  Federal Register 50, p. 46512.

USEPA.  1989a. Computer printout (CIS): 1977 Non-confidential production data from TSCA
       inventory. Washington, DC: Office of Pesticides and Toxic Substances, CID, USEPA.

USEPA.  1989b. Draft Final Report on the Occurrence and Human Exposure to Pesticides in Drinking
       Water, Food, and Air in the United States of America.  Office of Drinking Water, USEPA.
       September, 1989.

USEPA.  2000. TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Verschueren, K., ed. 1983. Handbook of Environmental Data on Organic Chemicals.  New York, NY:
       Van Nostrand Reinhold Co., pp. 464-465.
Occurrence Summary and Use Support Document          179                                    March 2002

-------
3.6    Diquat
Table of Contents

3.6.1  Introduction, Use and Production  	  181
3.6.2  Environmental Release  	  182
3.6.3  Ambient Occurrence 	  182
3.6.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  183
3.6.5  Additional Drinking Water Occurrence Data  	  187
3.6.6  Conclusion	  188
3.6.7  References  	  188
Tables and Figures

Figure 3.6-1:  Estimated Annual Agricultural Use for Diquat (1992)	  182

Table 3.6-1: Stage 1 Diquat Occurrence Based on 16-State Cross-Section - Systems  	  183

Table 3.6-2: Stage 1 Diquat Occurrence Based on 16-State Cross-Section - Population  	  184

Table 3.6-3: Stage 2 Estimated Diquat Occurrence Based on 16-State Cross-Section - Systems ....  185

Table 3.6-4: Stage 2 Estimated Diquat Occurrence Based on 16-State Cross-Section - Population .  .  186

Table 3.6-5: Estimated National Diquat Occurrence - Systems and Population Served 	  187
Occurrence Summary and Use Support Document          180                                     March 2002

-------
3.6.1  Introduction, Use and Production

Diquat [6,7-dihydrodipyrido (l,2-a:2',l'-c) pyrazinediium dibromide] is an organic solid of colorless or
yellow crystals.  It is a quick-acting contact herbicide, causing injury only to the parts of the plant to
which it is applied.  It is nonselective, meaning that it does not spare 'nontarget' plants from its
herbicidal effects. Diquat is referred to as a desiccant because it causes a leaf or an entire plant to dry out
quickly.  It is not residual, so it does not leave any trace of herbicide on or in plants, soil, or water
(EXTOXNET, 2001). Diquat is persistent, but essentially immobile, in the environment, indicating that
it will most likely be associated with the soil and sediment rather than water (USEPA, 200 Ib). A trade
name for diquat is Reglone.

Diquat is widely used as a herbicide, algicide, and plant growth  regulator (EXTOXNET, 2001). Usage in
1980  was estimated to be 200,000 pounds of active ingredient, and 1982 data indicates that diquat was
not produced domestically, but imports were nearly 835,000 pounds. Diquat has been in use since the
1950s to  control  both crop and aquatic weeds. It is used on potatoes; as an aid in harvesting cotton,
rapeseed and other oil seed crops; to wilt and dry out silage, standing hay, etc. for storage; a plant growth
regulator and sugar cane-flowering suppressant (USEPA, 200la).

As a herbicide/algicide, diquat is used to control broadleaf and grassy weeds in non-crop (including
residential) and aquatic areas. As a desiccant/defoliant, it is used in seed crops and potatoes.  Its largest
use is as a desiccant on potato crops (USEPA, 1995).

Figure 3.6-1 shows the United States Geological Survey (USGS, 1998a) derived geographic distribution
of estimated average annual diquat use in the United  States for 1992. A breakdown of use by crop is also
included.  The USGS (1998a) estimates approximately 230,000  pounds of diquat active ingredient were
used in 1992. The two largest concentrations of diquat use are seen in the West and the Northeast. Given
the amount of white space on the map, the possibility also exists that some States did not fully report
their diquat usage. While non-agricultural uses are not reflected here, and any sharp spatial differences
in use within a county are not well represented (USGS, 1998b),  existing data suggest that non-
agricultural use of diquat is minimal to non-existent (USEPA, 1995). A comparison of this use map with
the map of the 16 cross-section States (Figure  1.3-1)  shows that States across the range of high of low
diquat use are well represented in the cross-section.
Occurrence Summary and Use Support Document          181                                     March 2002

-------
Figure 3.6-1:  Estimated Annual Agricultural Use for Diquat (1992)
                                             DIQUAT
                                   ESTIMATED ANNUAL AGRICULTURAL USE
                      Average use of
                     Active Ingredient
                    Pounds per square mile
                     of county per year

                     D No Estimated lisa
                     D  < 0.003
                     D 0.003 - 0.020
                     D 0.021 -0.115
                     D 0.118-0.442
                     •  >= 0.443
Total
Crops Pounds Applied
potatoes
alfalfa hay
Held and grass seen
176.943
49,956
2. STB
Parcenl
National Use
77.01
S1.74
1.25
Source: USGS 1998a
3.6.2 Environmental Release

Diquat is released directly to the environment in its use as a herbicide, algicide, and plant growth
regulator and potentially during its manufacture, handling, and storage.  EPA estimated that 200,000
pounds of diquat active ingredient were used in the United States in 1980 (USEPA, 2001a). Diquat is not
listed as a Toxics Release Inventory (TRI) contaminant, so no TRI release records are maintained.
Therefore, the use of diquat (described in the previous section) may provide the primary indication of
where releases are most likely.  The areas of highest diquat use are in various potato growing regions of
the U.S., including Idaho and other Western States, as well as Maine and other locations in the Northeast.
In the Midwest it is most likely  used on alfalfa hay crops. These and other use areas are illustrated in
Figure 3.6-1.

3.6.3 Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers.  There are few available  data on the occurrence of diquat in ambient waters of the United
States. The most comprehensive and nationally consistent data describing ambient water quality in the
United States are being produced through the United States Geological Survey's (USGS) National Water
Quality Assessment (NAWQA) program. However, national NAWQA data are currently unavailable for
diquat.
Occurrence Summary and Use Support Document
182
March 2002

-------
A summary document entitled "Occurrence and Exposure Assessment of Diquat in Public Drinking
Water Supplies" (Wade Miler, 1989), was previously prepared for past USEPA assessments of diquat.
However, no national or regional studies were included on the occurrence of diquat in water other than
drinking water from ground water or surface water sources.

3.6.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of diquat occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date.  These data were evaluated
relative to several concentration thresholds of interest: 0.02 mg/L; 0.004 mg/L; and 0.002 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey and Texas, contained
occurrence data for diquat. These data represent more than 36,000 analytical results from approximately
9,000 PWSs during the period from 1984 to 1998 (with most analytical results from 1992 to 1997).  The
number of sample results and PWSs vary by State, although the State data sets have been reviewed and
checked to ensure adequacy of coverage and completeness. The overall modal detection limit for diquat
in the 16 cross-section States is equal to 0.0004 mg/L. (For details regarding the 16-State cross-section,
please refer to Section 1.3.5 of this report.)

3.6.4.1  Stage 1 Analysis Occurrence Findings

Table 3.6-1 illustrates the low occurrence of diquat in drinking water for the public water systems in the
16-State cross-section relative to three thresholds:  0.02 mg/L (the current MCL), 0.004 mg/L (the modal
MRL), and 0.002 mg/L.  A total of 2 (approximately 0.0218% of) ground water and surface water PWSs
had analytical results exceeding the MCL; 0.0655% of systems (6 systems) had results exceeding 0.004
mg/L; and 0.153% of systems (14 systems) had results exceeding 0.002 mg/L.

Approximately 0.0120% of ground water systems (1 system) had any analytical results greater than the
MCL. About 0.0600% of ground water systems (5 systems) had results above 0.004 mg/L. The
percentage of ground water systems with at least one result greater than 0.002 mg/L was equal to  0.120%
(10 systems).

Only  1 (0.122% of) surface water systems had results greater than the MCL. Also, approximately 1
(0.122% of) surface water systems had at least one analytical result greater than 0.004 mg/L.  Four
(0.487% of) surface water systems had results exceeding 0.002 mg/L.
Table 3.6-1:  Stage 1 Diquat Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.02
0.004
0.002
Percent of Systems
Exceeding Threshold
0.0120%
0.0600%
0.120%
Number of Systems
Exceeding Threshold
1
5
10
Occurrence Summary and Use Support Document
183
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems
Exceeding Threshold

Surface Water
0.02
0.004
0.002
0.122%
0.122%
0.487%
1
1
4

Combined Ground &
Surface Water
0.02
0.004
0.002
0.0218%
0.0655%
0.153%
2
6
14
Reviewing diquat occurrence in the 16 cross-section States by PWS population served (Table 3.6-2)
shows that approximately 0.208% of the population (about 153,400 people) was served by PWSs with at
least one analytical result of diquat greater than the MCL (0.02 mg/L).  Approximately 171,300 (0.233%
of) people were served by systems with an exceedance of 0.004 mg/L.  Over 337,000 (0.458% of) people
were served by systems with at least one analytical result greater than 0.002 mg/L.

The percentage of population served by ground water systems with analytical results greater than the
MCL was equal to 0.441% (approximately 142,000 people).  When evaluated relative to 0.004 mg/L and
0.002 mg/L, the percent of population exposed was equal to 0.497% (almost 160,000 people) and 0.664%
(almost 214,000 people), respectively.

The percentage of population served by surface water systems with exceedances of 0.02 mg/L and 0.004
mg/L was equal to 0.0275% (11,400 people). When evaluated relative to 0.002 mg/L, the percent of
population exposed was equal to 0.298% (over 123,000 people).
Table 3.6-2:  Stage 1 Diquat Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.02
0.004
0.002
Percent of Population Served
by Systems
Exceeding Threshold
0.441%
0.497%
0.664%
Total Population Served by
Systems Exceeding
Threshold
142,000
159,900
213,817

Surface Water
0.02
0.004
0.002
0.0275%
0.0275%
0.298%
11,400
11,400
123,300
Occurrence Summary and Use Support Document
184
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served
by Systems
Exceeding Threshold
Total Population Served by
Systems Exceeding
Threshold

Combined Ground &
Surface Water
0.02
0.004
0.002
0.208%
0.233%
0.458%
153,400
171,300
337,100
3.6.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.6-3 and 3.6-4.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

No ground water or surface water PWSs had an estimated mean concentration of diquat exceeding 0.02
mg/L. The percentage of ground and surface water systems with estimated mean concentration greater
that 0.004 mg/L and 0.002 mg/L is equal to 0.00096% (less than 1 system) and 0.0202% (2 systems),
respectively. Only 1 (0.00101% of) ground water PWS in the 16 States was estimated to have a mean
concentration greater than 0.004 mg/L and 2 (0.0204%) are estimated to have a mean concentration
greater than 0.002 mg/L.  The percentage of surface water systems with estimated mean concentration
values greater than 0.004 mg/L was equal to 0.000487% (less than 1 system in the 16 States). Only 1
(0.0173% of) surface water system in the 16 States had an estimated mean concentration value of diquat
greater than 0.002 mg/L.
Table 3.6-3:  Stage 2 Estimated Diquat Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.02
0.004
0.002
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.00101%
0.0204%
Range
0.000% - 0.000%
0.000% - 0.0120%
0.000% - 0.0600%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
1
2
Range
0-0
0-1
0-5

Surface Water
0.02
0.004
0.002
0.000%
0.000487%
0.0173%
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.122%
0
0
1
0-0
0-0
0-1
Occurrence Summary and Use Support Document
185
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
Range
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.02
0.004
0.002
0.000%
0.000961%
0.0202%
0.000% - 0.000%
0.000% -0.0109%
0.000% - 0.0546%
0
1
2
0-0
0-1
0-5
Reviewing diquat occurrence by PWS population served (Table 3.6-4) shows that approximately
0.000470% of population served by all PWSs (an estimate of approximately 300 people in the 16 States)
were potentially exposed to diquat levels above 0.004 mg/L.  When evaluated relative to a threshold of
0.002 mg/L, the percent of population exposed was equal to 0.0218% (approximately 16,300 people).
The percent of population exposed was equal to 0% for all system types when evaluated relative to the
MCL (0.02 mg/L).

The percentage of population served by ground water systems in the 16 States with estimated mean
concentration values greater than 0.004 mg/L was equal to 0.00101% (an estimate of approximately 300
people) and the percentage served by surface water systems in the 16 States was equal to 0.0000503%.
Approximately 14,600 (0.0453% of) people in the 16 cross-section States were served by ground water
systems with estimated mean concentration values greater than 0.002 mg/L.  The number of people in the
16 cross-section States served by surface water systems with estimated mean concentration values greater
than 0.002 mg/L was equal to about 1,500 (0.00356% of) people.
Table 3.6-4:  Stage 2 Estimated Diquat Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.020
0.004
0.002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.00101%
0.0453%
Range
0.000% - 0.000%
0.000% -0.00621%
0.000% - 0.404%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
300
14,600
Range
0-0
0 - 2,000
0 - 130,000

Surface Water
0.020
0.004
0.002
0.000%
0.0000503%
0.00356%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0275%
0
0
1,500
0-0
0-0
0-11,400

Combined Ground
& Surface Water
0.020
0.004
0.002
0.000%
0.000470%
0.0218%
0.000% - 0.000%
0.000% - 0.00295%
0.000% -0.177%
0
300
16,300
0-0
0 - 2,200
0 - 130,000
Occurrence Summary and Use Support Document
186
March 2002

-------
3.6.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, no systems had estimated mean concentration values of diquat greater than the MCL (0.02
mg/L). An estimated 1 system serving approximately 1,000 people had a mean concentration value of
diquat greater that 0.004 mg/L.  About 13 systems serving about 46,500 people nationally were estimated
to have mean diquat concentrations greater than 0.002 mg/L. (See Section 1.4 for a description of how
Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, approximately 1 PWSs serving about 900 people nationally had mean
concentrations greater than 0.004 mg/L.  An estimated 12 systems serving about 38,800 people nationally
had estimated mean concentration values that exceeded 0.002 mg/L.

Only 1 surface water system serving less than 100 people nationally was estimated to have mean
concentrations of diquat above 0.004 mg/L.  Approximately 1 surface water systems serving 4,500
people had estimated mean concentrations greater than 0.002 mg/L.
Table 3.6-5: Estimated National Diquat Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.02
0.004
0.002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
1
12
Range
0-0
0-7
0-36
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
900
38,800
Range
0-0
0 - 5,300
0 - 346,000

Surface Water
0.02
0.004
0.002
0
1
1
0-0
0-0
0-7
0
<100
4,500
0-0
0-0
0-35,100

Combined Ground
& Surface Water
0.02
0.004
0.002
0
1
13
0-0
0-7
0-35
0
1,000
46,500
0-0
0 - 6,300
0 - 376,200
3.6.5  Additional Drinking Water Occurrence Data

A summary document entitled "Occurrence and Exposure Assessment of Diquat in Public Drinking
Water Supplies" (Wade Miler, 1989), was previously prepared for past USEPA assessments of diquat.
However, no national or regional studies were included on the occurrence of diquat in drinking water
from ground water or surface water sources.
Occurrence Summary and Use Support Document
187
March 2002

-------
3.6.6  Conclusion

Diquat is widely used in agriculture and in non-agricultural applications as a herbicide, algicide, and
plant growth regulator.  In 1982, imports of diquat active ingredient were 835,000 pounds.  Information
from USGS indicates that use of diquat is most widespread in Western States and the Northeast, with
some scattered use in the Midwest. Diquat is not a TRI chemical, so there is no information available on
the environmental release of diquat. There are also no ambient data available.  The Stage 2 analysis,
based on the 16-State cross-section, estimated that zero percent of combined ground water and surface
water systems serving zero percent of the population exceeded the MCL of 0.02 mg/L. Based on this
estimate, zero PWSs nationally are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, according to information from USGS, approximately 25 States consumed
diquat for agricultural uses, including 8 of the 16 cross-section States. As there is no TRI or other data,
agricultural use seems to be the only basis for comparison between the 16 States and the nation. The
cross-section should adequately represent the occurrence of diquat on a national scale based upon the
use, production, and release patterns of the 16-State cross-section in relation to the patterns observed for
all 50 States.

3.6.7  References

EXTOXNET.  2001. Pesticide Information Profile: Diquat Bromide. Ithaca, NY:  Extension
       Toxicology Network, Pesticide Management Education Program. Available on the Internet at
       http://pmep.cce .Cornell .edu/profiles/extoxnet/dienochlor-glyphosate/diquat-ext.html, last updated
       March 1,2001.

USEPA.  1995. R.E.D. Facts: Diquat Bromide.  EPA Report 738-F-95-015. Washington, DC: Office of
       Prevention, Pesticides, and Toxic Substances, USEPA. 10 pp. Available on the Internet at:
       http://www.epa.gov/oppsrrdl/REDs/factsheets/0288fact.pdf

USEPA.  200 la. Consumer Factsheet on Diquat. Office of Ground Water and Drinking Water,
       USEPA. Available on the Internet at: http://www.epa.gov/OGWDW/dwh/c-SOC/diquat.html,
       last updated April 12, 2001.

USEPA.  200 Ib. Memorandum:  Tier I Drinking Water and Aquatic Ecological Exposure Assessments
       for Diquat Dibromide. From: James Breithaupt, Environmental Risk Branch II/EFED.
       Washington, D.C., October 2001.

USEPA.  2002. Occurrence Estimation Methodology and  Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS.  1998a. Annual Use Maps. Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated March 20, 1998.

USGS.  1998b.  Sources & Limitations of Data Used to Produce Maps of Annual Pesticide  Use.
       Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/mapex.html, last updated
       March 20, 1998.
Occurrence Summary and Use Support Document         188                                    March 2002

-------
Wade Miller Associates, Inc. 1989. Occurrence and Exposure Assessment ofDiquat in Public Drinking
        Water Supplies - DRAFT. Draft report submitted to EPA for review April 24, 1989.
Occurrence Summary and Use Support Document         189                                    March 2002

-------
3.7    Glyphosate
Table of Contents

3.7.1  Introduction, Use and Production  	  191
3.7.2  Environmental Release  	  192
3.7.3  Ambient Occurrence 	  192
3.7.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  193
3.7.5  Additional Drinking Water Occurrence Data  	  196
3.7.6  Conclusion	  197
3.7.7  References  	  197
Tables and Figures

Figure 3.7-1:  Estimated Annual Agricultural Use for Glyphosate (1992)  	  192

Table 3.7-1: Stage 1 Glyphosate Occurrence Based on 16-State Cross-Section - Systems	  193

Table 3.7-2: Stage 1 Glyphosate Occurrence Based on 16-State Cross-Section - Population	  194

Table 3.7-3: Stage 2 Estimated Glyphosate Occurrence Based on 16-State Cross-Section -
       Systems	  195

Table 3.7-4: Stage 2 Estimated Glyphosate Occurrence Based on 16-State Cross-Section -
       Population	  195

Table 3.7-5: Estimated National Glyphosate  Occurrence - Systems and Population Served	  196
Occurrence Summary and Use Support Document          190                                     March 2002

-------
3.7.1  Introduction, Use and Production

Glyphosate [N-(phosphonomethyl) glycine] is an organic solid of odorless white crystals that is usually
formulated as an isopropylamine salt. While it can be described as an organophosphorus compound,
glyphosate is not an organophosphate ester but a phosphanoglycine, and it does not inhibit cholinesterase
activity (EXTOXNET, 2001).  Some trade names for glyphosate include Roundup, Sting, Rodeo,
Tumbleweed, Sonic, Glycel, and Rondo (USEPA, 2001).

Glyphosate is a non-selective herbicide used on many food and non-food crops as well as non-crop areas
where total vegetation control is desired, such as roadsides (USEPA, 1993). It is useful on essentially all
annual and perennial plants, including grasses, sedges, broad-leaved weeds and woody plants
(EXTOXNET, 2001). When applied at lower rates, glyphosate serves as a plant growth regulator. The
most common uses include control of broadleaf weeks and grasses in: hay and pasture, soybeans, field
corn, ornamentals, lawns, turf, forest plantings, greenhouses, and rights-of-way (USEPA, 2001).

Although production data seems to be unavailable, glyphosate is one of the most commonly used
pesticides in the U.S. by volume. In 1990 usage was  estimated  at 11.6 million pounds, and it ranked
eleventh in volume among conventional pesticides in 1990 and  1991  (USEPA, 2001). Use from 1989 to
1991 ranged between 11.4 and 18.7 million pounds of glyphosate per year for anywhere from 13 to 20
million acres (USEPA, 1993).  Hay/pasture (20%), soybeans (20%), field corn (9%), and other
agricultural areas (20%) comprise 71% of the total acreage treated with glyphosate. Non-agricultural
areas (33%), soybeans (15%), hay/pasture (11%), and corn (8%) comprise 67% of the total pounds of ai
applied (USEPA, 1993).

Figure 3.7-1 shows the USGS (1998) derived geographic distribution of estimated average annual
glyphosate use in the United States for 1992.  The United States Geological Survey (USGS, 1998)
estimates approximately  16 million pounds of glyphosate active ingredient were used in 1992. These
estimates were derived using State-level data sets on pesticide use rates available from National Center
for Food and Agricultural Policy (NCFAP) combined with county-level data on harvested crop acreage
from the Census of Agriculture (Thelin and Gianessi, 2000). A breakdown of use by crop is also
included.  Soybeans account for the majority of usage (about 6.5 million pounds glyphosate), while
intermediate use can be found on several other crops as well (e.g., corn, all citrus, cotton, almonds,
grapes). Glyphosate use  appears to be geographically distributed across the U.S.  The largest
concentrations of glyphosate use are seen in the soybean production region of the Midwest and the citrus
growing areas in California and Florida.  A comparison of this use map  with the map of the  16 cross-
section States (Figure 1.3-1) shows that States across the range  of high of low glyphosate use are well
represented in the cross-section.
Occurrence Summary and Use Support Document          191                                     March 2002

-------
Figure 3.7-1:  Estimated Annual Agricultural Use for Glyphosate (1992)
                                          GLYPHOSATE
                                   ESTIMATED ANNUAL AGRICULTURAL USE
                       Average use of
                      Active Ingredient
                     Pounds per square mile
                      of county par year
                      D No Estimated Use
                      D  < 0.367
                      D 0.387-1.536
                      D 1.54O- 4.566
                      D 4.570- 12.337
                      •  >=12.33B
Crops
soybeans
com
all citrus
paatira
whaat and grains
cotton
almonds
yapas
sorghum
rice
Total
Pounds Applied
$482.564
3, 102, 793
1,139,922
991,773
953,321
702, 753
476,132
430,492
3CB, 532
203,008
Percem
National Use
40.48
19.34
7.11
8.13
&94
438
297
268
1.93
1.27
Source: USGS 1998
3.7.2  Environmental Release

Glyphosate is released directly to the environment when used as a herbicide and potentially during its
manufacture, handling and storage. Unfortunately, glyphosate is not listed as a Toxics Release Inventory
(TRI) contaminant, so no TRI release records are maintained. The use of glyphosate (described in the
previous section) may provide the primary indication of where releases are most likely. The areas of
highest glyphosate use are in the soybean and corn production regions of the Midwest and Mid-Atlantic,
and on the citrus production areas in the Pacific Northwest, California, and Florida. Many of the highest
use States are contained in the 16-State cross-section (i.e., California, Florida, Illinois, Indiana,
Kentucky,  Michigan, Oregon, and South Dakota).

3.7.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
often useful. In a drinking water context, ambient water is untreated source water residing in surface
waters and aquifers.  There  are few available data on the occurrence of glyphosate in ambient waters of
the United States. The most comprehensive and nationally consistent data describing ambient water
quality in the United States  are being produced through the United States Geological Survey's (USGS)
National Water Quality Assessment (NAWQA) program.  However, national NAWQA data are currently
unavailable for glyphosate.

Additional studies of ambient data are also unavailable. A summary document entitled ""Occurrence and
Exposure Assessment of Glyphosate in Public Drinking Water Supplies" (Wade Miler, 1989), was
previously prepared for past USEPA assessments of glyphosate. However, no national or regional
Occurrence Summary and Use Support Document
192
March 2002

-------
studies were included on the occurrence of glyphosate in water other than drinking water from ground
water or surface water sources.

3.7.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of glyphosate occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date.  These data were evaluated
relative to several concentration thresholds of interest: 0.7 mg/L; 0.06 mg/L; and 0.006 mg/L.

Thirteen of the sixteen cross-section State data sets contained occurrence data for glyphosate.  (There
were no glyphosate data from Nebraska, New Jersey or Texas.)  These data represent more than 33,000
analytical results from approximately 7,800 PWSs during the period from 1984 to 1998 (with most
analytical results from 1992 to 1997). The number of sample results and PWSs vary by State, although
the State data sets have been reviewed and checked to ensure adequacy of coverage and completeness.
The overall modal detection limit for glyphosate in the 16 cross-section States is equal to 0.006 mg/L.
(For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this report.)

3.7.4.1  Stage  1 Analysis Occurrence Findings

Table 3.7-1 illustrates the very low occurrence of glyphosate in drinking water for the public water
systems in the  16-State cross-section. No ground water or surface water PWSs in the 16 States had any
analytical results exceeding 0.7 mg/L or 0.06 mg/L. Only 0.0763% of total ground and surface water
systems (6 systems) had analytical results greater than 0.006 mg/L.  Five (approximately 0.0707% of)
ground water systems had analytical results greater than 0.006 mg/L. The percentage of surface water
systems with at least one analytical result greater than 0.006 mg/L was equal to 0.126% (1 system).
Table 3.7-1:  Stage 1 Glyphosate Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.7
0.06
0.006
Percent of Systems
Exceeding Threshold
0.000%
0.000%
0.0707%
Number of Systems
Exceeding Threshold
0
0
5

Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.126%
0
0
1

Combined Ground &
Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.0763%
0
0
6
Occurrence Summary and Use Support Document
193
March 2002

-------
Table 3.7-2 illustrates glyphosate occurrence in the 16 cross-section States based on the PWS population
served.  Approximately 0.0270% of the 16-State population was served by PWSs with analytical
detections of glyphosate greater than 0.006 mg/L (about 19,000 people). The percentage of population
served by ground water systems with at least one analytical result greater than 0.006 mg/L was equal to
0.0347% (almost 11,000 people). A total of 8,300 (0.0210% of) people served by surface water systems
were exposed to glyphosate levels greater than 0.006 mg/L. When evaluated relative to 0.7 mg/L and
0.06 mg/L, the percent of population exposed was equal to 0% for all system types.
Table 3.7-2:  Stage 1 Glyphosate Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.7
0.06
0.006
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.000%
0.0347%
Total Population Served
by Systems Exceeding
Threshold
0
0
10,700

Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.0210%
0
0
8,300

Combined Ground &
Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.0270%
0
0
19,000
3.7.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables  3.7-3 and 3.7-
4. The statistically generated best estimate values, as well as the ranges around the best estimate value,
are presented. (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.
For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology
and Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations
(USEPA, 2002)).

No ground water or surface water PWSs in the 16 cross-section States had an estimated mean
concentration of glyphosate exceeding 0.7 mg/L or 0.06 mg/L. Approximately 0.0000736% of ground
and surface water PWSs (less than 1  system in the 16-State cross-section) was estimated to have a mean
concentration greater than 0.006 mg/L. The percentage of ground water PWSs with estimated mean
concentrations greater than 0.006 mg/L was equal to approximately 0.0000849%.  No surface water
PWSs had estimated mean concentration exceeding any of the three specified concentration thresholds.
Occurrence Summary and Use Support Document
194
March 2002

-------
Table 3.7-3: Stage 2 Estimated Glyphosate Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.7
0.06
0.006
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0000849%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.0000763%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
Approximately 0.0000138% of population served by all PWSs in the 16 States was potentially exposed to
glyphosate levels above 0.006 mg/L.  The percentage of population served by ground water systems
relative to 0.006 mg/L was equal to 0.0000313%. (No surface water systems exceeded any of the
specified health thresholds.) When evaluated relative to a threshold of 0.7 mg/L and 0.06 mg/L, the
percent of population exposed was equal to 0% for all system types.
Table 3.7-4: Stage 2 Estimated Glyphosate Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.7
0.06
0.006
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0000313%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
Occurrence Summary and Use Support Document
195
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.7
0.06
0.006
0.000%
0.000%
0.0000138%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
3.7.4.3 Estimated National Occurrence

As illustrated in Table 3.7-5, the Stage 2 analysis estimates no PWSs nationally (therefore, no population
served) were estimated to have a mean concentration value of glyphosate greater than 0.7 mg/L, 0.06
mg/L, or 0.006 mg/L. (See Section 1.4 for a description of how Stage 2 16-State estimates are
extrapolated to national values.)
Table 3.7-5:  Estimated National Glyphosate Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.7
0.06
0.006
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.7
0.06
0.006
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.7
0.06
0.006
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
3.7.5  Additional Drinking Water Occurrence Data

Additional studies of ambient data are also unavailable. A summary document entitled "Occurrence and
Exposure Assessment of Glyphosate in Public Drinking Water Supplies" (Wade Miler, 1989), was
previously prepared for past USEPA assessments of glyphosate. However, no national or regional
studies were included on the occurrence of glyphosate in drinking water from ground water or surface
water sources.
Occurrence Summary and Use Support Document
196
March 2002

-------
3.7.6  Conclusion

Glyphosate is a non-selective herbicide used on many food and non-food crops as well as non-crop areas
such as roadsides. It is useful on essentially all annual and perennial plants, including grasses, sedges,
broad-leaved weeds and woody plants. Recent statistics regarding use of glyphosate indicate production
and use are steady, with glyphosate use concentrated in the Midwest, California, and Florida. Glyphosate
is not a TRI chemical, so there is no information available on the environmental release of glyphosate.
There are also no ambient data available. The Stage 2 analysis, based on the 16-State cross-section,
estimated that zero percent of combined ground water and surface water systems serving zero percent of
the population exceeded the MCL of 0.7 mg/L. Based on this estimate, zero PWSs nationally are
estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, according to information from USGS, all 50 States used glyphosate for
agricultural purposes.  The areas of most use are the Midwest, Florida, and California, represented by
eight cross-section States. As there is no TRI or other data, agricultural use seems to be the only basis
for comparison between the 16 States and the nation. The cross-section should adequately represent the
occurrence of glyphosate on a national scale based upon the use, production, and release patterns of the
16-State cross-section in relation to the patterns observed for all 50 States.

3.7.7  References

EXTOXNET. 2001. Pesticide Information Profile: Glyphosate. Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://pmep.cce.cornell.edu/profiles/extoxnet/dienochlor-glyphosate/glyphosate-ext.html,
       accessed July  18, 2001.

Thelin, G.P., and L.P.  Gianessi.  2000. Method for Estimating Pesticide Use for County Areas of the
       Conterminous United States.  U.S. Geological Survey Open-File Report 00-250.
       62 pp. Available on the Internet at: http://water.wr.usgs.gov/pnsp/rep/ofr00250/ofr00250.pdf

USEPA.  1993.  Registration Eligibility Decision (RED): Glyphosate.  EPA Report 738-R-93-014.
       Washington, DC:  Office of Prevention, Pesticides, and Toxic Substances, USEPA.  74 pp. +
       Appendices. Available on the Internet at:
       http://www.epa.gov/oppsrrdl/REDs/old_reds/glyphosate.pdf

USEPA. 2001.  National Primary Drinking Water Regulations - Consumer Factsheet on: Glyphosate.
       Office of Ground Water and Drinking Water, USEPA. Available on the Internet at
       http://www.epa.gov/safewater/dwh/c-SOC/glyphosa.html, last updated April 12, 2001.

USEPA. 2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS.  1998. Annual Use Maps. Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/, last
       updated March 20, 1998.

Wade Miller Associates, Inc. 1989.  Occurrence and Exposure Assessment of Glyphosate in Public
       Drinking Water Supplies - DRAFT. Draft report submitted to EPA for review April 12, 1989.

Occurrence Summary and Use Support Document         197                                     March 2002

-------
3.8    Heptachlor & Heptachlor Epoxide
Table of Contents

3.8.1  Introduction, Use and Production  	  199
3.8.2  Environmental Release  	  200
3.8.3  Ambient Occurrence 	  201
3.8.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  203
3.8.5  Additional Drinking Water Occurrence Data  	  211
3.8.6  Conclusion	  213
3.8.7  References  	  213
Tables and Figures

Table 3.8-1: Environmental Releases (in pounds) for Heptachlor in the United States,
       1988-1996	 201

Table 3.8-2: Stage 1 Heptachlor Occurrence Based on 16-State Cross-Section - Systems  	 203

Table 3.8-3: Stage 1 Heptachlor Occurrence Based on 16-State Cross-Section - Population	 204

Table 3.8-4: Stage 1 Heptachlor Epoxide Occurrence Based on 16-State Cross-Section -
       Systems	 205

Table 3.8-5: Stage 1 Heptachlor Epoxide Occurrence Based on 16-State Cross-Section -
       Population	 206

Table 3.8-6: Stage 2 Estimated Heptachlor Occurrence Based on 16-State Cross-Section -
       Systems	 207

Table 3.8-7: Stage 2 Estimated Heptachlor Occurrence Based on 16-State Cross-Section -
       Population	 207

Table 3.8-8: Stage 2 Estimated Heptachlor Epoxide Occurrence Based on 16-State
       Cross-Section - Systems	 208

Table 3.8-9: Stage 2 Estimated Heptachlor Epoxide Occurrence Based on 16-State
       Cross-Section - Population	 209

Table 3.8-10: Estimated National Heptachlor Occurrence - Systems and Population Served	 210

Table 3.8-11: Estimated National Heptachlor Epoxide Occurrence - Systems and
       Population Served	 210
Occurrence Summary and Use Support Document         198                                     March 2002

-------
3.8.1  Introduction, Use and Production

Heptachlor is a synthetic chemical that does not occur naturally.  It is both a breakdown product and a
component of the pesticide chlordane. Pure heptachlor is a white powder that smells like camphor
(mothballs). Technical-grade heptachlor is a tan powder and has a lower level of purity than pure
heptachlor. Technical-grade heptachlor was the form of heptachlor used most often as a pesticide.
Heptachlor does not burn easily, does not explode, nor does it dissolve easily in water (ATSDR, 1993).

Heptachlor epoxide is a biodegradation product of heptachlor.  It is not manufactured and was not used
as an insecticide like heptachlor. Heptachlor epoxide is made by bacteria in the environment. Like pure
heptachlor, heptachlor epoxide is a white powder that does not explode easily. Heptachlor epoxide is
frequently detected in areas of heptachlor use and is more likely to be found in the environment than
heptachlor. Animals and people also make heptachlor epoxide when heptachlor enters their bodies.
About 20% of heptachlor is changed within hours to heptachlor epoxide in the environment and in the
body (ATSDR, 1993).

Heptachlor is a chlorinated hydrocarbon insecticide, presently used in the United States only to control
fire ants, in buried, pad-mounted electric power transformers, and in underground cable television and
telephone cable boxes.  The end use product, a granular formulation packaged in small plastic bags, is
applied by pouring the contents of the plastic bag directly into a metal or concrete enclosure, which is
rarely opened again (USEPA, 1992).

Heptachlor is a persistent dermal insecticide with some fumigant action.  It is nonphytotoxic at
insecticidal concentrations (Worthing and Walker, 1987, as cited in ATSDR, 1993).  Heptachlor is
present as an impurity in chlordane; typically, hepatchlor constitutes 10 percent of technical grade
chlordane (USEPA, 1980, as cited in USEPA, 1989). Because chlordane  is applied by subsurface ground
injection, the potential for water contamination in some  areas may be high (USEPA,  1989). Heptachlor
was used extensively from 1953 to 1974 as a soil and seed treatment to protect corn,  small grains, and
sorghum from pests.  It was used to control ants, cutworms, maggots, termites, thrips, weevils, and
wireworms in both cultivated and uncultivated soils. Heptachlor was also used nonagriculturally during
this time period to control termites and household insects (USEPA, 1986; Worthington  and Walker,
1987, as cited in ATSDR, 1993).

Heptachlor is converted to heptachlor epoxide, and other degradation products, in the environment.
Heptachlor epoxide degrades more slowly and, as a result, is more persistent than heptachlor. Heptachlor
epoxide has been found in food crops grown in soils treated with heptachlor many years before (ATSDR,
1993).

In 1974, the EPA Administrator issued a notice of intent to suspend the registrations of certain pesticide
products containing heptachlor for subsurface control of termites and dipping of roots and tops of
nonfood plants (USEPA, 1983,  as cited in USEPA,  1989). EPA proposed cancellation of nearly all
registered uses of heptachlor because of its potential cancer risk and its persistence and bioaccumulation
throughout the food chain. On March 6, 1978, EPA issued a final cancellation order putting into effect
the terms of settlement for the cancellation proceedings  (USEPA, 1983, as cited in USEPA, 1989).
Under this cancellation order, an agreement was outlined to phase out the nontermiticide uses of
heptachlor over a 5-year period. All uses permitted during the phase-out period were restricted to
treatment by certified applicators or professional commercial seed treatment companies. The few uses
that were not canceled in 1974, treatment of field corn, seed (for corn, wheat, oats, barley, rye, and
sorghum), citrus, pineapple, and narcissus bulbs, were phased out gradually over a 5-year period ending
on July 1, 1983 (USEPA, 1986, as cited in ATSDR, 1993). Certain uses of heptachlor were specifically

Occurrence Summary and Use Support Document         199                                     March 2002

-------
exempted from EPA's suspension and cancellation actions because they were believed to result in
insignificant exposure and, consequently, insignificant risk.  Registrations were retained for subsurface
termite control, fire ant control in buried cable closures, and dipping of roots or tops of nonfood plants (a
use subsequently canceled voluntarily by the registrant, Velsicol Chemical Corporation) (USEPA,  1986,
as cited in ATSDR, 1993).

Approximately 30 percent (or 550,000 pounds) of the heptachlor used in 1971 was applied to commercial
and residential structures for protection against termites and to nurseries, lawns, and gardens. By 1974,
domestic use of heptachlor for termite control had increased to 1.4 million pounds (USEPA, 1976,  as
cited in USEPA, 1989).

In the early 1980s, heptachlor was used to control subterranean termites. Nearly  10 million pounds of
chlordane was applied in the United States in 1980; approximately 1 million pounds of heptachlor would
have been present as an impurity in chlordane in that year. EPA (USEPA, 1983,  as cited in USEPA,
1989) reported that the largest quantity of chlordane was initially distributed to EPA Region IV, which
includes Alabama, Georgia, Florida, Mississippi, North Carolina, South Carolina, Tennessee, and
Kentucky.

In 1983, between 1 and 2 million pounds of heptachlor was used in the United States for termite control
(USEPA, 1983, as cited in USEPA, 1989).  Between 0.75 and 1 million pounds of heptachlor were used
in 1986 (Kuch, 1986, as cited in USEPA, 1989).  Heptachlor has limited use as a single active ingredient
and usually is applied in combination with chlordane (USEPA, 1983, as cited in USEPA, 1989).  A
second approved use  of heptachlor is to control fire ants in sugarcane and pineapple fields in Hawaii.

In 1988, EPA prohibited the sale, distribution, and  shipment of existing  stocks of all canceled chlordane
and heptachlor products. Subsequently, virtually all uses of heptachlor products were voluntarily
canceled by the registrant, Velsicol Chemical Corporation (USEPA, 1990, as cited in ATSDR, 1993).
The only commercial use of heptachlor products still permitted is fire ant control  in power transformers.
Use of existing stocks of heptachlor-containing termicide products in the possession of homeowners is
also permitted (USEPA, 1990, as cited in ATSDR,  1993).

3.8.2  Environmental Release

Heptachlor is listed as a Toxics Release Inventory (TRI) chemical.  Table 3.8-1 illustrates the
environmental releases for heptachlor from 1988 to 1996. (Heptachlor data are only available for these
years.) Air emissions constitute most of the on-site releases, with a general trend to decrease over the
years.  Air emissions decreased drastically after 1988, which was the year EPA prohibited the sale,
distribution, and shipment of existing stocks of all canceled heptachlor products.  The decrease in air
emissions are solely responsible for decreases in heptachlor total on- and off-site  releases in recent years.
Minimal surface water discharges were reported. No underground injection, releases to land (such as
spills or leaks within the boundaries of the reporting facility), or off-site releases  (including metals  or
metal compounds transferred off-site) were reported for heptachlor. These TRI data for heptachlor were
reported from 5 States (Georgia, Illinois, Tennessee, Florida, and Mississippi), although no more than 2
States reported releases in a given year.  Only Tennessee reported releases after 1992 (USEPA, 2000).
Of the five States, Illinois and Florida are included in the 16-State cross-section (used for analyses  of
heptachlor occurrence in drinking water; see Section 3.8.4).  (For a map of the 16-State cross-section, see
Figure 1.3-1.)
Occurrence Summary and Use Support Document         200                                     March 2002

-------
Table 3.8-1:  Environmental Releases (in pounds) for Heptachlor in the United States,
1988-1996
Year
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
198
203
830
31
710
5
3,797
3,411
54,295
Surface Water
Discharges
5
6
3
2
1
0
1
2
2
Underground
Injection
-
-
-
-
-
-
-
-
-
Releases
to Land
-
-
-
-
-
-
-
-
-
Off-Site Releases
-
-
-
-
-
-
-
-
-
Total On- &
Off-site
Releases
203
209
833
33
711
5
3,798
3,413
54,297
 Source: USEPA, 2000
3.8.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers.  There are few available data on the occurrence of heptachlor in ambient waters of the
United States.  The most comprehensive and nationally consistent data describing ambient water quality
in the United States are being produced through the United States Geological Survey's (USGS) National
Water Quality Assessment (NAWQA) program.  However, national NAWQA data are currently
unavailable for heptachlor.

3.8.3.1 Additional Ambient Occurrence Data

Additional studies of ambient data are summarized below. A summary document entitled "Occurrence
and Human Exposure to Pesticides in Drinking Water, Food and Air in the United States of America"
(USEPA, 1989), was previously prepared for past USEPA assessments of various pesticides. Three
national studies are summarized for the occurrence of heptachlor and heptachlor epoxide in surface
waters.  Eight regional studies addressed levels of heptachlor and heptachlor epoxide in water other than
drinking water.  Two of the regional studies addressed levels of heptachlor and heptachlor epoxide in
ground water, and six studies provided data on levels of heptachlor and heptachlor epoxide in surface
water. The following information is taken directly from "Occurrence and Human Exposure to Pesticides
in Drinking Water, Food and Air in the United States of America" (USEPA, 1989).

3.8.3.1.1 Ground Water Sources

Ground water wells in 25 California counties were analyzed as part of the California State Board's
Toxics Special Project during 1984 (Cohen and Bowes, 1984, as cited in USEPA, 1989). Heptachlor was
found in three samples with a maximum concentration of 0.3 |ig/L. The other positive values, total
number of samples, and detection limit were not reported.

Tucker and Burke (1978, as cited in USEPA, 1989) presented data on levels of heptachlor and heptachlor
epoxide in samples of water collected from 163 wells, including private and public drinking water
supplies, industrial sites, and wells in the vicinity of landfills in nine New Jersey counties. The analysis
Occurrence Summary and Use Support Document
201
March 2002

-------
showed that samples of water from three supply wells contained heptachlor and samples from seven
wells contained heptachlor epoxide in excess of the minimum reportable concentration of 0.01 |ig/L.  The
maximum values for heptachlor and heptachlor epoxide were  1.0 and 0.6 |ig/L, respectively. The
maximum concentration for heptachlor epoxide occurred in a public drinking water supply well.

3.8.3.1.2 Surface Water Sources

National studies of surface water analyses for heptachlor and heptachlor epoxide were conducted by
USEPA (1976, as cited in USEPA, 1989) and Breidenbach et al. (1967, as cited in USEPA, 1989).  The
overall mean as reported for both pesticides was 0.0063 |ig/L. The minimum values determined for each
were 0.001 |ig/L and the maximum values were 0.035 |ig/L and 0.02 |ig/L for heptachlor and heptachlor
epoxide, respectively.  The number of positive values, total number of samples, and detection limit were
not reported.

The National Pesticide Monitoring Network examined rivers nationwide from 1975 to 1980 (Gilliom et
al., 1985, as cited in USEPA, 1989).  Heptachlor epoxide was reported as being detected in 9 out of 2,946
samples analyzed from 177 locations (detection limit = 0.01 |ig/L). The mean and range of values were
not reported.

Heptachlor and heptachlor epoxide were analyzed in rivers and streams in upstate New York
(Estabrooks, no date, as cited in USEPA, 1989) from 1982 to  1983. No samples for either pesticide were
found to be positive, out of 252 samples analyzed. The detection limit was 10.0 |ig/L.

Barks (1978, as cited in USEPA, 1989) presented the results of a USGS water quality study conducted
from April 1973 to July 1974 in the Ozark National Scenic Riverways, Missouri. During the study, 20
surface water samples were collected from 3 sites on the Current River and 1 site on Jacks Fork and
analyzed for pesticide content. The analysis of unfiltered samples found no heptachlor or heptachlor
epoxide in excess of the detection limit (the detection limit was not reported).

Englande et al. (1978, as cited in USEPA, 1989) presented the results of extensive chemical analysis of
six Advanced Wastewater Treatment (AWT) plant effluents. Four plants were located in California, and
one each in the District of Columbia and Texas. None of the 63 AWT effluent samples  contained
concentrations of heptachlor in excess of the detection limit (the detection limit was not reported).

Truhlar and Reed (1976,  as cited in USEPA, 1989) reported on water samples collected from four
streams in Pennsylvania and analyzed for chlorinated hydrocarbon pesticides during the period from
April 1970 to February 1971. The streams drained four different types of land use areas. Concentrations
of heptachlor and heptachlor epoxide were not detected in any of the 25 stream samples.

Schacht (1974, as cited in USEPA, 1989) presented the results of a study to determine the levels of
pesticides  in the surface water of Lake Michigan and its tributaries. During the period 1970 to 1972, a
total of 45 water samples was collected.  Concentrations of heptachlor epoxide ranging from "non-
detected" to 0.017 |ig/L were found in the samples.  The detection limit for heptachlor epoxide was
reported as less than 0.0002 |ig/L

Dappen (1974, as cited in USEPA, 1989) reported the results of a study to determine the pesticide
content of urban storm runoff in Nebraska.  Runoff samples were collected at three stations in a Nebraska
city during and after storms. A total of 80 samples was collected at the first station during 16 different
storms.  Concentrations of heptachlor and heptachlor epoxide  in samples from the first station ranged
from 0 to 0.059 |ig/L and  0 to 0.2  |ig/L, respectively. At the second sampling station, 55 samples were

Occurrence Summary and Use Support Document         202                                    March 2002

-------
collected during 9 different storms.  Concentrations of heptachlor and heptachlor epoxide in samples
from the second station ranged from 0 to 0.176 |ig/L and 0 to 0.194 |ig/L respectively. A total of 14
samples was collected at the third station during 3 storms. Concentrations of heptachlor in samples from
the third station ranged from 0 to 0.055 |ig/L. No heptachlor epoxide was detected. No detection limits
were reported for the study.

3.8.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of heptachlor and heptachlor epoxide occurrence presented in the following section is based
on State compliance monitoring  data from the 16 cross-section States.  The 16-State cross-section is the
largest and most comprehensive  compliance monitoring data set compiled by EPA to date. These data
were evaluated relative to several concentration thresholds of interest.  Heptachlor was reviewed relative
to 0.0004 mg/L; 0.0001 mg/L; and 0.00004 mg/L. Heptachlor epoxide was reviewed relative to 0.0002
mg/L; 0.0001; and 0.00002 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
heptachlor and heptachlor epoxide.  Heptachlor and heptachlor epoxide data each represent more than
57,000 analytical results from approximately 14,000 PWSs during the period from 1984 to 1998 (with
most analytical results from 1992 to 1997).  The number of sample results and PWSs vary by State,
although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness.  The overall modal detection limits for heptachlor and heptachlor epoxide in the 16 cross-
section States are equal to 0.00004 mg/L and 0.00002 mg/L, respectively. (For details regarding the 16-
State cross-section, please refer to Section 1.3.5 of this report.)

3.8.4.1  Stage 1 Analysis Occurrence Findings

Table 3.8-2 illustrates the low occurrence of heptachlor in drinking water for the public water systems in
the 16-State cross-section relative to three thresholds: 0.0004 mg/L (the current MCL), 0.0001 mg/L, and
0.00004 mg/L (the modal MRL). Only 1 ground water PWS (approximately 0.00702% of PWSs in the
16 States) had analytical results exceeding the MCL; 0.0562% of systems (8 systems) had results
exceeding 0.0001 mg/L; and 0.0702% of systems (10 systems) had results exceeding 0.00004 mg/L.

Approximately 0.00779% of ground water systems (1 system) had any analytical results greater than the
MCL.  About 0.0545% of ground water systems (7 systems) had results above 0.0001 mg/L. The
percentage of ground water systems with at least one result greater than 0.00004 mg/L was equal to
0.0624% (8 systems).

No surface water systems had results greater than the MCL. Only 1 (0.0709% of) surface water systems
had at least one analytical result  greater than 0.0001 mg/L.  Two (0.142% of) surface water systems had
results exceeding 0.00004 mg/L.
Table 3.8-2:  Stage 1 Heptachlor Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0004
0.0001
Percent of Systems
Exceeding Threshold
0.00779%
0.0545%
Number of Systems
Exceeding Threshold
1
7
Occurrence Summary and Use Support Document
203
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.00004
Percent of Systems
Exceeding Threshold
0.0624%
Number of Systems
Exceeding Threshold
8

Surface Water
0.0004
0.0001
0.00004
0.000%
0.0709%
0.142%
0
1
2

Combined Ground &
Surface Water
0.0004
0.0001
0.00004
0.00702%
0.0562%
0.0702%
1
8
10
Reviewing heptachlor occurrence in the 16 cross-section States by PWS population served (Table 3.8-3)
shows that approximately 0.000398% of the population (about 400 people) was served by PWSs with at
least one analytical result of heptachlor greater than the MCL (0.0004 mg/L). Approximately 160,400
(0.166% of) people were served by systems with an exceedance of 0.0001 mg/L.  Over 170,000 (0.176%
of) people were served by systems with at least one analytical result greater than 0.00004 mg/L.

The percentage of population served by ground water systems with analytical results greater than the
MCL was equal to 0.000945% (approximately 400 people). When evaluated relative to 0.0001 mg/L and
0.00004 mg/L, the percent of population exposed was equal to 0.247% (over 100,000 people) and
0.252% (almost 103,000 people), respectively.

The percent of population served by surface water systems exposed to heptachlor at levels greater than
0.0004 mg/L was equal to 0%. The percentage of population served by surface water systems with
exceedances of 0.0001 mg/L and 0.00004 mg/L was equal to 0.107% (60,000 people), and 0.121%
(67,800 people), respectively.
Table 3.8-3: Stage 1 Heptachlor Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0004
0.0001
0.00004
Percent of Population Served
by Systems
Exceeding Threshold
0.000945%
0.247%
0.252%
Total Population Served by
Systems Exceeding
Threshold
400
100,400
102,600

Surface Water
0.0004
0.0001
0.00004
0.000%
0.107%
0.121%
0
60,000
67,800

Occurrence Summary and Use Support Document
204
March 2002

-------
Source Water Type
Combined Ground &
Surface Water
Threshold
(mg/L)
0.0004
0.0001
0.00004
Percent of Population Served
by Systems
Exceeding Threshold
0.000398%
0.166%
0.176%
Total Population Served by
Systems Exceeding
Threshold
400
160,400
170,400
Table 3.8-4 illustrates the low occurrence of heptachlor epoxide in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds: 0.0002 mg/L (the current MCL),
0.0001 mg/L, and 0.00002 mg/L (the modal MRL). Four (approximately 0.0283% of) ground and
surface water PWS had analytical results exceeding 0.0002 mg/L and 0.0001 mg/L.  About 0.0566% of
systems (8 systems) had results exceeding 0.00002 mg/L.

Approximately 0.0314% of ground water systems (4 systems) had any analytical results greater than
0.0002 mg/L and 0.0001  mg/L. The percentage of ground water systems with at least one result greater
than 0.00002 mg/L was equal to 0.0550% (7 systems).

No surface water systems had results greater than 0.0002 mg/L and 0.0001 mg/L. Only 1  (0.0712% of)
surface water systems had at least one analytical result greater than 0.00002 mg/L.
Table 3.8-4:  Stage 1 Heptachlor Epoxide Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Systems
Exceeding Threshold
0.0314%
0.0314%
0.0550%
Number of Systems
Exceeding Threshold
4
4
7

Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.0712%
0
0
1

Combined Ground &
Surface Water
0.0002
0.0001
0.00002
0.0283%
0.0283%
0.0566%
4
4
8
Reviewing heptachlor epoxide occurrence in the 16 cross-section States by PWS population served
(Table 3.8-5) shows that approximately 0.0291% of the population (about 28,000 people) was served by
PWSs with at least one analytical result of heptachlor epoxide greater than the MCL (0.0002 mg/L), and
0.0001 mg/L. Approximately 91,600 (0.0952% of) people were served by systems with at least one
analytical result greater than 0.00002 mg/L.
Occurrence Summary and Use Support Document
205
March 2002

-------
The percentage of population served by ground water systems with analytical results greater than 0.0002
mg/L or 0.0001 mg/L was equal to 0.0695% (approximately 28,000 people). When evaluated relative to
0.00002 mg/L, the percent of population exposed was equal to 0.153% (almost 62,000 people).

The percent of population served by surface water systems exposed to heptachlor epoxide at levels
greater than 0.0002 mg/L and 0.0001 mg/L was equal to 0%. The percentage of population served by
surface water systems with exceedances of 0.00002 mg/L was equal to 0.0537% (30,000 people).
Table 3.8-5:  Stage 1 Heptachlor Epoxide Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Population Served
by Systems
Exceeding Threshold
0.0695%
0.0695%
0.153%
Total Population Served
by Systems Exceeding
Threshold
28,000
28,000
61,600

Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.0537%
0
0
30,000

Combined Ground &
Surface Water
0.0002
0.0001
0.00002
0.0291%
0.0291%
0.0952%
28,000
28,000
91,600
3.8.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings for heptachlor, based on the cross-section data, are presented in Tables
3.8-6 and 3.8-7. The statistically generated best estimate values, as well as the ranges around the best
estimate value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section
1.4 of this report.  For complete details regarding the Stage 2 analyses, please refer to Occurrence
Estimation Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking
Water Regulations (USEPA, 2002)).

No ground water or surface water PWSs had an estimated mean concentration of heptachlor exceeding
0.0004 mg/L. Approximately 0.0000140% of PWSs (less than 1 system in the 16-State cross-section)
were estimated to have a mean concentration greater than 0.0001 mg/L. Only 1 (0.00119% of) system
had estimated mean concentration values of heptachlor greater than 0.00004 mg/L.

The percentage of ground water systems with estimated mean concentration values of heptachlor greater
than 0.0001 mg/L was equal to 0.0000156% (less than 1 system in the  16 States).  Approximately
0.00131% of ground water PWSs (about 1 system in the 16 States) had estimated mean concentration
values greater than 0.00004 mg/L.  The percentage of surface water PWSs with estimated mean
concentration exceeding 0.00004 mg/L was equal to 0.000142% (less than 1 system).
Occurrence Summary and Use Support Document
206
March 2002

-------
Table 3.8-6: Stage 2 Estimated Heptachlor Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0004
0.0001
0.00004
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.0000156%
0.00131%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.00779%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
1
Range
0-0
0-0
0-1

Surface Water
0.0004
0.0001
0.00004
0.000%
0.000%
0.000142%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0004
0.0001
0.00004
0.000%
0.0000140%
0.00119%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.00702%
0
0
1
0-0
0-0
0-1
Reviewing heptachlor occurrence by PWS population served (Table 3.8-7) shows that approximately
0.000000242% of population served by all PWSs in the 16 States were potentially exposed to heptachlor
levels above 0.0001 mg/L. Approximately 0.000350% of the population served by PWSs in the 16 States
(about 300 people) was exposed to heptachlor at levels above 0.00004 mg/L.  When evaluated relative to
a threshold of 0.0004 mg/L, the percent of population exposed was equal to 0%.

The percentage of population served by ground water systems relative to 0.0001 mg/L was equal to
0.000000576%. Approximately 0.000784% of the population was served by ground water systems
(about 300 people) with estimated mean concentration values of heptachlor greater than 0.00004 mg/L.
The percentage of population served by surface water systems with estimated mean concentration values
greater than 0.00004 mg/L was equal to 0.0000341%.
Table 3.8-7: Stage 2 Estimated Heptachlor Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0004
0.0001
0.00004
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000000576%
0.000784%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.0113%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
300
Range
0-0
0-0
0 - 4,600
Occurrence Summary and Use Support Document
207
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Surface Water
0.0004
0.0001
0.00004
0.000%
0.000%
0.0000341%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0004
0.0001
0.00004
0.000%
0.000000242%
0.000350%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.00476%
0
0
300
0-0
0-0
0 - 4,600
The Stage 2 occurrence findings for heptachlor epoxide, based on the cross-section data, are presented in
Tables 3.8-8 and 3.8-9. No ground water or surface water PWSs had an estimated mean concentration of
heptachlor epoxide exceeding 0.0002 mg/L or 0.0001 mg/L.  Approximately 0.00528% of PWSs (about 1
system in the 16-State cross-section) were estimated to have a mean concentration greater than 0.00002
mg/L.

The percentage of ground water systems with estimated mean concentration values of heptachlor epoxide
greater than 0.00002 mg/L was  equal to 0.00583% (approximately 1  system in the 16 States).
Approximately 0.000285% of surface water PWSs (about 1 system in the 16 States) had estimated mean
concentration values greater than 0.00002 mg/L.
Table 3.8-8:  Stage 2 Estimated Heptachlor Epoxide Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.00583%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0236%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
1
Range
0-0
0-0
0-3

Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.000285%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
1
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.00528%
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.0212%
0
0
1
0-0
0-0
0-3
Occurrence Summary and Use Support Document
208
March 2002

-------
Table 3.8-9 illustrates heptachlor epoxide occurrence by PWS population served in the 16-State cross-
section. When evaluated relative to a threshold of 0.0002 mg/L and 0.0001 mg/L, the percent of
population exposed to heptachlor epoxide was equal to 0%. Approximately 0.00489% of the population
served by PWSs in the 16 States (about 4,700 people) was exposed to heptachlor epoxide at levels above
0.00002 mg/L.

The percentage  of population served by ground water systems relative to 0.00002 mg/L was equal to
0.0114% (approximately 4,600 people).  The percentage of population served by surface  water systems
with estimated mean concentration values of heptachlor epoxide greater than 0.00002 mg/L was equal to
0.000215% (about 100 people).
Table 3.8-9:  Stage 2 Estimated Heptachlor Epoxide Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0114%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0570%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
4,600
Range
0-0
0-0
0 - 23,000

Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.000215%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
100
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
0.000%
0.000%
0.00489%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0239%
0
0
4,700
0-0
0-0
0 - 23,000
3.8.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, zero systems nationally had estimated mean concentration values of heptachlor greater than the
MCL (0.0004 mg/L) and 0.0001 mg/L. An estimated 1 ground water system serving approximately 700
people was exposed to heptachlor concentrations greater than 0.00004 mg/L. (See Section 1.4 for a
description of how Stage 2 16-State estimates are extrapolated to national values.)
Occurrence Summary and Use Support Document
209
March 2002

-------
Table 3.8-10: Estimated National Heptachlor Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.0004
0.0001
0.00004
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
1
Range
0-0
0-0
0-5
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
700
Range
0-0
0-0
0 - 9,700

Surface Water
0.0004
0.0001
0.00004
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0004
0.0001
0.00004
0
0
1
0-0
0-0
0-5
0
0
700
0-0
0-0
0-10,100
The Stage 2 analysis estimated zero systems nationally with mean concentration values of heptachlor
epoxide greater than 0.0002 mg/L and 0.0001 mg/L. An estimated 3 ground water systems serving
approximately 9,800 people were exposed to heptachlor epoxide concentrations greater than 0.00002
mg/L. Approximately 1 surface water system serving 300 people nationally had estimated mean
concentration values of heptachlor epoxide greater than 0.00002 mg/L.
Table 3.8-11: Estimated National Heptachlor Epoxide Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.0002
0.0001
0.00002
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
3
Range
0-0
0-0
0-14
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
9,800
Range
0-0
0-0
0 - 48,800

Surface Water
0.0002
0.0001
0.00002
0
0
1
0-0
0-0
0-0
0
0
300
0-0
0-0
0-0

Combined Ground
& Surface Water
0.0002
0.0001
0.00002
0
0
3
0-0
0-0
0-14
0
0
10,400
0-0
0-0
0 - 50,800
Occurrence Summary and Use Support Document
210
March 2002

-------
3.8.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of heptachlor and heptachlor epoxide in
drinking water are also reviewed. Previously compiled occurrence heptachlor and heptachlor epoxide
information, from an OGWDW summary document entitled "Occurrence and Human Exposure to
Pesticides in Drinking Water, Food and Air in the United States of America" (USEPA, 1989), is
presented in following section.  This variety of studies and information are presented regarding levels of
heptachlor and heptachlor epoxide in drinking water, with the scope of the reviewed studies ranging from
national to regional. Note that none of the studies presented in the following section provide the
quantitative analytical results or comprehensive coverage that would enable direct comparison to the
occurrence findings estimated with the cross-section occurrence data presented in Section 3.8.4.  These
additional studies, however, do enable a broader assessment of the Stage 2 occurrence estimates
presented for this Six-Year Review. All the following information in Section 3.8.5 is taken directly from
"Occurrence and Human Exposure to Pesticides  in Drinking Water, Food and Air in the United States of
America" (USEPA, 1989).

3.8.5.1 Ground Water Sources

Twelve towns in Connecticut were sampled during 1984 to 1985 for heptachlor and heptachlor epoxide
by the Connecticut Agricultural Experiment Station, New Haven, Connecticut (Waggoner,  1985, as cited
in USEPA, 1989).  These towns, combined, serve a population of over 570,000 people.  Drinking water
wells  were sampled at 42 locations, and no samples were positive for either heptachlor or heptachlor
epoxide (detection limits were 0.2 |ig/L and 0.47 |ig/L, respectively).

Shallow drinking water wells in 10 counties in northwest Mississippi were analyzed from 1983 to 1984
during a Mississippi State University study on pesticide hazard assessment (MSU, 1984, as cited in
USEPA,  1989). No positive samples were found for either heptachlor or heptachlor epoxide out of 143
samples analyzed (detection limit = 0.001 |ig/L).

Similarly, no positive results were obtained from 67 samples analyzed in Long Island, New York, for
either heptachlor or heptachlor epoxide.  These results were from a 1984 study reported  by the Suffolk
County Department of Health Services (Holden, 1986, as cited in USEPA, 1989).  No detection limit(s)
was reported.

Drinking water supplies of the Floridian aquifer were analyzed at 96 locations in 1984 by the Florida
Department of Environmental Regulation and the U.S. Geological Survey (Holden, 1986, as cited in
USEPA,  1989). These supplies serve a combined population of over 3 million people.  Less than 8
percent of the samples were positive for any of the pesticides sampled for, including heptachlor and
heptachlor epoxide. No other information was reported.

Irwin  and Healy (1978, as cited in USEPA, 1989) summarized data collected in 1976 during a water
quality reconnaissance of public water supplies in Florida. None of the 100 water supplies  sampled
using  the five aquifers in Florida contained heptachlor and heptachlor epoxide in excess of the detection
limits. The detection limits were not reported.

Drinking water wells were sampled in Idaho for heptachlor epoxide by the Idaho Department of Health
and Welfare (1984, as cited in USEPA, 1989). One sample, out of 107, was positive, with a
concentration of 0.015 |ig/L (no detection limit was reported).
Occurrence Summary and Use Support Document         211                                     March 2002

-------
Sandhu et al. (1978, as cited in USEPA, 1989) summarized the results from a study conducted in two
rural counties in South Carolina. Water supplies sampled were outside incorporated municipalities.
Samples were taken randomly from wells where there was no pretreatment prior to use.  It was not noted
whether these supplies were public or private. Also, data were collected on drinking water from different
land use areas in each county (i.e., agricultural, forest, and residential). The number of samples taken
and the number of positive samples were not reported: however, the percent of samples exceeding the
detection limit of 0.01 |ig/L was reported by the county.

Results indicated that 62.5 percent of the samples taken in Chesterfield County had detectable levels of
heptachlor, and 42 percent had detectable levels of heptachlor epoxide. The sample means were 0.015
and 0.008 |ig/L for heptachlor and heptachlor epoxide, respectively. In Hampton County, 45.5 percent of
the samples contained heptachlor and 64 percent of the samples contained heptachlor epoxide in excess
of the detection limit. The sample means for the Hampton County samples were 0.009 and 0.018 |ig/L
for heptachlor and heptachlor epoxide, respectively.  (These percentages included samples taken from
forest land use areas.) Concentrations of heptachlor in drinking water samples from agricultural areas in
Chesterfield and Hampton counties ranged from not detected to 0.16 |ig/L with a mean of <0.01 |ig/L
(not detected).  For heptachlor epoxide, concentrations ranged from not detected to  0.09 |ig/L with a
mean of <0.01  |ig/L (not detected).  In residential areas in both counties, concentrations of heptachlor
and heptachlor epoxide in drinking water samples ranged from not detected to 0.045 |ig/L and not
detected to 0.01 |ig/L respectively.  Sample means were <0.01 |ig/L (not detected) in all cases.

Tucker and Burke (1978, as cited in USEPA,  1989) reported that heptachlor epoxide was detected at a
level of 0.6 |ig/L in water from a public drinking water supply well in  Camden County, New Jersey.

3.8.5.2 Surface Water Sources

Irwin and Healy (1978, as cited in USEPA, 1989) reported that none of 16 surface water supplies
sampled in Florida contained heptachlor or heptachlor epoxide in excess of the detection limits. The
detection limits were not reported.

To assemble a database which would reflect the status of Great Lakes  drinking water quality, the
Canadian Public Health Association gathered data from October 1984 through August 1985 (Canadian
Public Health Association, 1986, as cited in USEPA, 1989). The data collected covered the period from
the mid 1970s to early 1985.  The study was funded by the Health Protection Branch of Health and
Welfare Canada, and the Ontario Ministry of the Environment. A research team, appointed by the
Association, reviewed data on the quality of water at 31 representative Canadian and United States
communities and 24 offshore sites to evaluate the human health implications.

For each of the 31 communities, data consisted of: 1) background information on the community; 2)
treatment plant schematics and associated treatment process information; and 3) water quality data.
Water sample types included raw water (treatment plant intake), distribution water (treated water), and
tap water. Water quality data collected included general parameters (e.g., alkalinity, turbidity),
microbiological and radiological parameters, inorganic parameters, and organic parameters (including
volatiles, base/neutrals, pesticides and PCBs, and phenols and acids).  For each parameter, the water
type, time period,  concentration  (mean and range), number of samples, and detection limit were recorded.

For most of the synthetic organics, including heptachlor and heptachlor epoxide, the available data
indicated that there were very low levels of these contaminants in the raw, treated, or tap water. Most of
the values found were "not detected" or near the detection limit (Canadian Public Health Association,
1986, as cited in USEPA, 1989).

Occurrence Summary and Use Support Document         212                                     March 2002

-------
3.8.5.3 Unspecified Sources

Pellizzari et al. (1979, as cited in USEPA, 1989) presented monitoring data on halogenated hydrocarbons
in drinking water from the northern New Jersey area and the Buffalo/Niagara, New York area.  The
limits of detection ranged from 0.005 to 0.025 |ig/L for both studies. Of 22 drinking water samples taken
from the northern New Jersey area, two samples (9%) contained heptachlor at levels of 0.007 and 0.02
l-ig/L, and none contained heptachlor epoxide. None of the  16 drinking water samples from the
Buffalo/Niagara area were found to contain heptachlor  or heptachlor epoxide in excess of the detection
limits.

In a report on source identification of pollutants entering a sewage treatment plant, Levins et al. (1979, as
cited in USEPA, 1989) sampled two drinking water sources in a drainage basin in Georgia.  Although
detection limits were not reported, heptachlor and heptachlor epoxide were not detected in either of the
drinking water samples.

3.8.6  Conclusion

Heptachlor is synthetically produced and exists  at low levels in soil, water, air, and food. It was used as
an insecticide until 1988.  Recent statistics regarding environmental release of heptachlor indicate use
has rapidly declined since the late 1980s, when it was banned for sale, distribution, and shipment.
Reported industrial releases of heptachlor have been reported to TRI since 1988 in 5 States, with most
releases coming from Tennessee. Heptachlor ambient occurrence was not analyzed in any available
studies. Because heptachlor epoxide is a biodegradation product of heptachlor, it is frequently  detected
in areas of heptachlor use and is more likely to be found in the environment than heptachlor. However,
little production and use information is available on heptachlor epoxide.

The Stage 2 analysis of heptachlor, based on the 16-State cross-section, estimated that approximately
zero percent of combined ground water and surface water systems serving zero percent of the population
exceeded the MCL of 0.0004 mg/L. Based on this estimate, zero PWSs nationally are estimated to have
heptachlor levels greater than the MCL. The Stage 2 analysis of heptachlor epoxide, based on the 16-
State cross-section, estimated that approximately zero percent of combined ground water and surface
water systems serving zero percent of the population exceeded the MCL of 0.0002 mg/L. Based on this
estimate, zero PWSs nationally are  estimated to have heptachlor epoxide levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOc, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for heptachlor from 5 States, including 2
of 16 cross-section States.  Heptachlor use has been severely restricted, and its production has been
banned nationwide.  The cross-section should adequately represent the occurrence of heptachlor and
heptachlor epoxide on a national scale based upon the use, production, and release patterns of the 16-
State cross-section in relation to the patterns observed for all 50 States.

3.8.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1993. Toxicological Profile for
       Heptachlor and Heptachlor Epoxide. U.S. Department of Health and Human Services, Public
       Health Service. 131 pp. + Appendices.  Available on the Internet at:
       http://www.atsdr.cdc .gov/toxprofiles/tp 12 .pdf
Occurrence Summary and Use Support Document          213                                      March 2002

-------
Barks, J.H. 1978. Water Quality in the OzarkNational Scenic Riverways, Missouri. Washington, DC:
       U.S. Geological Survey, U.S. Department of Interior.  U.S. Geological Survey Water-Supply
       Paper 2048.

Breidenbach, A.W., et al.  1967'. Chlorinated hydrocarbon pesticides in major river basins, 1957-65.
       Pub. Health Rep. v. 82, p.  139.

Canadian Public Health Association. 1986.  Comprehensive Survey of the Status of Great Lakes
       Drinking Water.  Prepared in cooperation with Health and Welfare Canada and the Ministry of
       the Environment. Ottawa, Canada: Canadian Public Health Association. August 1986.

Cohen, D.B., and G.W. Bowes. 1984. Water Quality and Pesticides: A California Risk Assessment
       Program (Volume 1).  Sacramento, CA: State Water Resources Control Board, Toxic Substances
       Control Program.

Dappen, G.  1974. Pesticide Analysis from Urban Storm Runoff.  Prepared by University of Nebraska,
       Lincoln, Nebraska, for Office of Water Research and Technology, Nebraska We sleyan
       University. Project No. A-025-NEB(2).

Englande, A.J., J.K. Smith,  and J.N. English. 1978. Potable water quality of advanced wastewater
       treatment plant effluents. Prog. Wat. Tech.  v. 10, no. !/2, pp. 17-39.

Estabrooks, F.  No date. Memorandum of pesticide information in New York surface water.  From:
       Frank Estabrooks, U.S. Department of Environmental Conservation. Albany, NY.

Gilliom, R.J., R.B. Alexander, and RA. Smith.  1985. Pesticides in the Nation's Rivers, 1975-1980, and
       Implications for Future Monitoring. U.S. Geological Survey Water-Supply Paper 2271. U.S.
       Government Printing Office, U.S. Department of Interior.

Holden, P.W. 1986. Pesticides and Groundwater Quality. Issues and Problems in Four States.
       Prepared for the Board of Agriculture, National Research Council. Washington, DC: National
       Academy Press.

Idaho Department of Health and Welfare. 1984.  Letter to Charles Berry, Office of Drinking Water, U.S.
       Environmental Protection Agency, summarizing data on groundwater contamination incidents.
       Sent by: Charles D. Brokopp, State Epidemiologist, Division of Health, Idaho Department of
       Health and Welfare.

Irwin, G.A., and H.G. Healy.   1978. Chemical and Physical Quality of Selected Public Water Supplies in
       Florida, August-September 1976. Tallahassee, FL: Water Resources Division, U.S. Geological
       Survey. USGS/WRI  78-21.

Kuch, P.  1986. Letter to Frederic A. Zafran, SAIC, McLean, VA. October 28, 1986.

Levins, P., J. Adams, P. Brenner, S. Coons, K. Thrun, and J. Varone. 1979. Sources of Toxic Pollutants
       found in Influents to Sewage Treatment Plants.  IV. R.M. Clayton Drainage Basin, Atlanta
       report. Prepared by Arthur D. Little, Inc., for Office of Water Planning and Standards, U.S.
       Environmental Protection Agency, Washington, DC.  EPA Contract No. 68-01-3857.
Occurrence Summary and Use Support Document         214                                     March 2002

-------
Mississippi State University (MSU).  1984. Cooperative Agreement Progress Report: Mississippi
       Pesticide Hazard Assessment Project.  Pesticide Residue Monitoring of Drinking Water from
       Shallow Wells in the Mississippi Delta. Prepared by: Larry Lane, Principal Investigator,
       Mississippi State Chemical Laboratory, Mississippi State University

Pellizzari, E.D., M.D. Erickson, and R.A. Zweidinger. 1979. Formulation of a Preliminary Assessment
       ofHalogenated Organic Compounds in Man and Environmental Media. EPA Report 560-13-79-
       006. Prepared by Research Triangle Institute, Research Triangle Park, North Carolina, for Office
       of Toxic Substances, U.S. Environmental Protection Agency, Washington, DC.

Sandhu, S.S., W.J. Warren, and P. Nelson.  1978. Pesticidal residue in rural potable water. Journal of
       the American Water Works Association,  v. 70, no. 1, pp. 41-45.

Schacht, RA.  1974. Pesticides in the Illinois  Waters of Lake Michigan. EPA Report 660-3-74-002.
       Prepared by Illinois Environmental Protection Agency, Chicago, Illinois, for Office of Research
       and Development, Office of Ground Water and Drinking Water, U.S. Environmental Protection
       Agency, Washington, DC.

Truhlar, J.F., and L.A. Reed.  1976. Occurrence of pesticide residues in four streams draining different
       land-use areas in Pennsylvania, 1969-1971.  Pestic. Monit. J. v.  10, no. 3, pp. 101-110.

Tucker, R.K., and T.A. Burke. 1978.  A Second Preliminary Report on the Findings of the State
       Groundwater Monitoring Project. New Jersey: Department of Environmental Protection.

USEPA.  1976. EPA Actions  to Cancel and Suspend Uses ofChlordane and Heptachlor as Pesticides -
       Economic and Social Implications.  EPA Report 540-4-76-004.  Washington, DC: Office of
       Pesticide Programs, USEPA.

USEPA.  1980. Ambient Water Quality Criteria for Heptachlor. EPA Report 440-4-80-052.
       Washington, DC: Office of Water Regulations and Standards, USEPA.

USEPA.  1983. Analysis of the risks and benefits of seven chemicals used for subterranean termite
       control. EPA Report  540-9-83-005. Washington, DC: Office of Pesticide Programs, USEPA.

USEPA.  1986. Guidance for the Reregistration of Pesticide Products Containing Heptachlor as the
       Active Ingredient.  EPA Report 540-RS-87-018.  Washington, DC: Office of Pesticides and
       Toxic Substances, USEPA.

USEPA.  1989. Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air in the
       United States of America (Draft Final Report).  Office of Drinking Water, USEPA. September,
       1989.

USEPA.  1990. Suspended, Canceled, and Restricted Pesticides. Office of Pesticides and Toxic
       Substances, USEPA.

USEPA.  1992. R.E.D. Facts: Heptachlor.  EPA Report 738-F-OO-013.  Washington, DC: Office of
       Prevention, Pesticides, and Toxic Substances, USEPA. 4 pp. Available on the Internet at:
       http://www.epa.gov/oppsrrdl/REDs/factsheets/0175fact.pdf, last updated June 21, 2001.
Occurrence Summary and Use Support Document         215                                    March 2002

-------
USEPA.  2000.  TW Explorer'.Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Waggoner, P.E. 1985. Memorandum from the Connecticut Agricultural Experiment Station on
       Pesticides in Connecticut Groundwater. P. Gough, (ed.). New Haven, CT: News of Science.

Worthing, C.R., Walker, S.B., eds.  1987.  The Pesticide Manual: A World Compendium. 8th ed.
       Suffolk, Great Britain: The British Crop Protection Council, pp. 455-456.
Occurrence Summary and Use Support Document         216                                    March 2002

-------
3.9    Hexachlorobenzene
Table of Contents

3.9.1 Introduction, Use and Production 	  218
3.9.2 Environmental Release  	  219
3.9.3 Ambient Occurrence 	  219
3.9.4 Drinking Water Occurrence Based on the 16-State Cross-Section	  221
3.9.5 Additional Drinking Water Occurrence Data 	  225
3.9.6 Conclusion	  226
3.9.7 References  	  227
Tables and Figures

Table 3.9-1:  Hexachlorobenzene Manufacturers and Processors by State	  218

Table 3.9-2:  Environmental Releases (in pounds) for Hexachlorobenzene in the
       United States, 1988-1999  	  219

Table 3.9-3:  Stage 1 Hexachlorobenzene Occurrence Based on 16-State Cross-Section -
       Systems	  221

Table 3.9-4:  Stage 1 Hexachlorobenzene Occurrence Based on 16-State Cross-Section -
       Population	  222

Table 3.9-5:  Stage 2 Estimated Hexachlorobenzene Occurrence Based on 16-State
       Cross-Section - Systems	  223

Table 3.9-6:  Stage 2 Estimated Hexachlorobenzene Occurrence Based on 16-State
       Cross-Section - Population	  223

Table 3.9-7:  Estimated National Hexachlorobenzene Occurrence - Systems and
       Population Served	  224
Occurrence Summary and Use Support Document         217                                    March 2002

-------
3.9.1  Introduction, Use and Production

Hexachlorobenzene, also known as HCB, is a chlorinated hydrocarbon fungicide used as a seed
treatment, especially on wheat. It may be used with or without other seed treatments.
Hexachlorobenzene is a white crystalline solid that is not very soluble in water. It does not occur
naturally in the environment. It is formed as a by-product while making other chemicals, in the waste
streams of chloralkali and wood-preserving plants, and when burning municipal waste (ATSDR,  2000).

Currently, hexachlorobenzene is not manufactured as a commercial end product in the U.S. and also has
no commercial end uses. It does not occur naturally in the environment, and is mostly formed as a
byproduct in the manufacture of chlorinated solvents, in the waste streams of chloralkali and wood-
preserving plants, in fly ash, and in flue gas effluents from municipal incineration. Hexachlorobenzene
has not been commercially produced since the late 1970s. In 1975 about 3,200 pounds were produced,
and in 1984 between 7,700 and 25,350 pounds were produced as a byproduct in chemical production
(ATSDR, 2000).

Until  1984 hexachlorobenzene was widely used as a pesticide to protect the seeds of onions, sorghum,
wheat and other grains against fungus. In 1984 the last registered use of hexachlorobenzene as a
pesticide was voluntarily canceled.  It was also used to make fireworks, ammunition and synthetic
rubber; as a porosity controller in the manufacture of electrodes; as a chemical intermediate in dye
manufacturing; and as a wood preservative (ATSDR, 2000).

Table 3.9-1  shows the number of facilities in each State that manufacture and process
hexachlorobenzene, the intended uses of the product, and the range of maximum amounts derived from
the Toxics Release Inventory (TRI) of EPA (ATSDR, 2000).
Table 3.9-1: Hexachlorobenzene Manufacturers and Processors by State
State"
CA
IL
KS
KY
LA
NJ
TN
TX
Number of facilities
3
1
1
1
4
1
1
4
Range of maximum amounts on site in
100-99,999
1,000-9,999
100-199
10,000-99,999
0-99,999
1,000-9,999
1,000-9,999
1 000-999 999
Activities and uses0
1,4,5
13
13
1,3,7
1,2,3,5,7,8
13
1,6
1 561013
Tost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
1. Produce                 8. As a formulation component
2. Import                  9. As an article component
3. For on-site use/processing      10. For repackaging only
4. For sale/distribution          11. As a chemical processing aid
5. As a byproduct             12. As a manufacturing aid
6. As an impurity             13. Ancillary or other uses
7. As a reactant

Source: ATSDR, 2000 compilation ofTRI982000 data
Occurrence Summary and Use Support Document          218                                      March 2002

-------
3.9.2  Environmental Release

Hexachlorobenzene is listed as a Toxics Release Inventory (TRI) chemical.  Table 3.9-2 illustrates the
environmental releases for hexachlorobenzene from 1988 through 1999.  (There are only
hexachlorobenzene data for these years.) Although there is no detectable trend in virtually any category
of release for hexachlorobenzene, total on- and off-site releases have moderated and decreased in recent
years. Previous to 1995, levels had ranged from over 1 million pounds to around 30,000 pounds; recent
levels hover well under 20,000 pounds.  Off-site releases comprise the vast majority of total releases, and
air emissions are, in most years, the most significant source of on-site releases. Releases to land
remained at or near zero until just recently, and both surface water discharges and underground injection
normally contributed less than 1,000 pounds to the overall total. These TRI data for hexachlorobenzene
were reported from 11 States, with four States (Texas, Tennessee, Louisiana, and California) reporting
every year (USEPA, 2000).  Out of those 11 States, 6 are included in the 16-State cross section (used for
analyses  of hexachlorobenzene occurrence in drinking water; see Section 3.9.4). (For a map of the 16-
State cross-section, see Figure 1.3-1.)
Table 3.9-2:  Environmental Releases (in pounds) for Hexachlorobenzene in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
1,524
371
154
220
566
458
636
4,471
841
1,468
4,613
4,045
Surface Water
Discharges
7
4
276
274
6,458
269
476
227
111
124
338
4
Underground
Injection
--
--
139
717
480
204
548
794
60
220
710
410
Releases
to Land
23
96
0
0
0
0
0
0
1
--
0
0
Off-Site Releases
13,550
13,251
12,038
23,449
6,975
940,478
648,010
28,380
1,064,793
34,011
1,008,186
443,541
Total On- &
Off-site
Releases
15,104
13,722
12,607
24,660
14,479
941,409
649,670
33,872
1,065,806
35,823
1,013,847
448,000
 Source: USEPA, 2000
3.9.3  Ambient Occurrence

The most comprehensive and nationally consistent data describing ambient water quality in the United
States are being produced through the United States Geological Survey's (USGS) National Water Quality
Assessment (NAWQA) program. However, national NAWQA data, as well as NURP and NPDES data,
are currently unavailable for hexachlorobenzene.

3.9.3.1 Additional Ambient Occurrence Data

Additional studies of ambient data are summarized below. A summary document entitled "Occurrence
and Exposure Assessment of Hexachlorobenzene in Public Drinking Water Supplies" (Wade Miller,
Occurrence Summary and Use Support Document
219
March 2002

-------
1989), was previously prepared for past USEPA assessments of hexachlorobenzene. Several studies
included in that document have addressed concentrations of hexachlorobenzene in ambient surface water.
(No studies were found that addressed the occurrence of hexachlorobenzene in ambient ground water
sources.) The following information is taken directly from "Occurrence and Exposure Assessment of
Hexachlorobenzene in Public Drinking Water Supplies" (Wade Miller, 1989).

The USEPA's computerized water quality data base known as STORET was devised to assist Federal and
State institutions meet objectives of Public Law 92-500 to maintain and enhance the physical, chemical,
and biological quality of the nation's ambient waterways by providing for the collection and
dissemination of basic water quality data (Staples et al., 1985, as cited in Wade Miller, 1989).  Data are
collected by States, EPA regional offices, and other government agencies and are maintained in the
STORET system.

Before presenting a summary of the ambient water data in STORET, it is important to note that there are
significant limitations in using the data base to estimate representative concentrations of a contaminant
such as hexachlorobenzene.  Data entered into STORET are gathered from an array of studies conducted
for various purposes. Analyses are conducted in a number of different laboratories employing different
methodologies with a range of detection limits.  In many cases, detection limits are not reported, making
the reliability of the data highly questionable. Where detection limits have been reported, STORET
assigns the detection limit value to those observations reported as not detected. This can lead to errors in
interpretation and overestimation of concentrations in cases in which there is a preponderance of
nondetectable values. Additionally, a few high values can inflate mean values and result in large
standard deviations relative to the means  (Staples et al., 1985, as cited in Wade Miller, 1989).  Very high
values may not be correct, as they may reflect sample contamination or analytical error and can
significantly distort assessment of average concentrations.  Staples et al. (1985, as cited in Wade Miller,
1989) also notes that the use of data collected prior to the 1980s is not recommended, since such data was
obtained using less sensitive laboratory techniques and quality assurance procedures were not yet
mandated for the data entered into the system.

The STORET water quality data base provides information on the occurrence of contaminants at ambient
water stations in U.S. waterways. A summary of this information was obtained for hexachlorobenzene in
ambient waters. Ambient sites include streams, lakes, ponds, wells, reservoirs, canals, estuaries, and
oceans.  While the preponderance of data were collected from surface water sources, the number of
samples collected from ground water wells, relative to the total number of samples collected from all
ambient sites combined, is unspecified (Staples et al., 1985, as cited in Wade Miller, 1989). Staples et al.
(1985, as cited in Wade Miller, 1989) have  summarized data from the 1980's only; that is, data from 1980
through 1983. This was done based on the number of data points and the likelihood that better quality
assurance practices have been employed in  more recent years. In the absence of sophisticated statistical
analyses to eliminate improbable data, median values have been reported. The  median value is sensitive
to extreme values, and reflects a measure of central tendency more accurately than the mean value in the
presence of these extreme values (Staples et al., 1985, as cited in Wade Miller,  1989).

For a total of 1,786 observations from ambient water stations, the median concentration of
hexachlorobenzene was 0.020 |ig/L.  Of the total number of observations, 26.0% were reported as
detected. However, detection limits and other sampling information were not reported. Clark et al.
(1988, as cited in Wade Miller, 1989) reported that hexachlorobenzene concentrations ranging from
0.00001 to 0.00012  |ig/L were detected in water samples collected from Lakes Ontario and Michigan.
No other information was reported.
Occurrence Summary and Use Support Document         220                                     March 2002

-------
3.9.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of hexachlorobenzene occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative to several concentration thresholds of interest:  0.001 mg/L; 0.0005 mg/L; and 0.0001
mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
hexachlorobenzene. These data represent approximately 53,000 analytical results from more than 14,000
PWSs during the period from 1984 to 1998 (with most analytical results from 1992 to 1997). The
number of sample results and PWSs vary by State, although the State data sets have been reviewed and
checked to ensure adequacy of coverage and completeness. The overall modal detection limit for
hexachlorobenzene in the 16 cross-section States is equal to 0.0001 mg/L.  (For details regarding the 16-
State cross-section, please refer to Section 1.3.5 of this report.)

3.9.4.1 Stage 1 Analysis Occurrence Findings

Table 3.9-3 illustrates the low occurrence of hexachlorobenzene in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds:  0.001 mg/L (the current MCL), 0.0005
mg/L, and 0.0001 mg/L. One ground water PWSs (approximately 0.00714% of all PWSs in the 16
States) had at least one analytical result exceeding 0.001 mg/L, 0.0005 mg/L, and 0.0001 mg/L. No
surface water systems had any analytical results greater than any of the three thresholds.
Table 3.9-3:  Stage 1 Hexachlorobenzene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.001
0.0005
0.0001
Percent of Systems
Exceeding Threshold
0.00792%
0.00792%
0.00792%
Number of Systems
Exceeding Threshold
1
1
1

Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.000%
0
0
0

Combined Ground &
Surface Water
0.001
0.0005
0.0001
0.00714%
0.00714%
0.00714%
1
1
1
Occurrence Summary and Use Support Document
221
March 2002

-------
Reviewing hexachlorobenzene occurrence in the 16 cross-section States by PWS population served
(Table 3.9-4) shows that approximately 0.0399% of the population (about 37,500 people) was served by
one ground water PWS with at least one analytical result of hexachlorobenzene greater than 0.001 mg/L,
0.0005 mg/L, and 0.0001 mg/L.  No people served by surface water systems were exposed to
hexachlorobenzene at levels greater than any of the three thresholds.
Table 3.9-4:  Stage 1 Hexachlorobenzene Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.001
0.0005
0.0001
Percent of Population
Served by Systems
Exceeding Threshold
0.0951%
0.0951%
0.0951%
Total Population Served
by Systems Exceeding
Threshold
37,500
37,500
37,500

Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.000%
0
0
0

Combined Ground &
Surface Water
0.001
0.0005
0.0001
0.0399%
0.0399%
0.0399%
37,500
37,500
37,500
3.9.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.9-5 and 3.9-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations (USEPA,
2002)).

No ground water or surface water PWSs had an estimated mean concentration of hexachlorobenzene
exceeding 0.001 mg/L or 0.0005 mg/L. Approximately 1 (0.00307% of) ground water PWS and 1
(0.00101% of) surface water PWS in the 16 States were estimated to have a mean concentration greater
than 0.0001 mg/L.
Occurrence Summary and Use Support Document
222
March 2002

-------
Table 3.9-5: Stage 2 Estimated Hexachlorobenzene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.001
0.0005
0.0001
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.00307%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.0158%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
1
Range
0-0
0-0
0-2

Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.00101%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
1
0-0
0-0
0-0

Combined Ground
& Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.00287%
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.0143%
0
0
1
0-0
0-0
0-2
Reviewing hexachlorobenzene occurrence by PWS population served (Table 3.9-6) shows that
approximately 0.0176% of population served by all PWSs in the 16 States (an estimate of approximately
16,600 people) were potentially exposed to hexachlorobenzene levels above 0.0001 mg/L. The percent
of population exposed was equal to 0% for all system types when evaluated relative to 0.001 mg/L and
0.0005 mg/L.

The percentage of population served by ground water systems in the 16 States with estimated mean
concentration values greater than 0.0001 mg/L was equal to 0.0384% (an estimate of approximately
15,200 people).  Approximately 1,400 (0.00260% of) people in the 16 cross-section States were served
by surface water systems with estimated mean concentration values greater than 0.0001 mg/L.
Table 3.9-6: Stage 2 Estimated Hexachlorobenzene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.001
0.0005
0.0001
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0384%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.214%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
15,200
Range
0-0
0-0
0 - 84,300
Occurrence Summary and Use Support Document
223
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.00260%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
1,400
0-0
0-0
0-0

Combined Ground
& Surface Water
0.001
0.0005
0.0001
0.000%
0.000%
0.0176%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.0896%
0
0
16,600
0-0
0-0
0 - 84,300
3.9.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, no systems had estimated mean concentration values of hexachlorobenzene greater than 0.001
mg/L or 0.0005 mg/L. An estimated 2 systems nationally serving approximately 37,600 people were
expected to have estimated mean concentrations of hexachlorobenzene greater than 0.0001 mg/L.  (See
Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, approximately 2 PWSs serving about 32,900 people nationally were estimated
to have a mean concentration greater than 0.0001  mg/L.  Only 1 surface water systems serving
approximately 3,300 people was estimated to have a mean concentration of hexachlorobenzene above
0.0001 mg/L.
Table 3.9-7: Estimated National Hexachlorobenzene Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.001
0.0005
0.0001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
2
Range
0-0
0-0
0-9
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
32,900
Range
0-0
0-0
0-183,100

Surface Water
0.001
0.0005
0.0001
0
0
1
0-0
0-0
0-0
0
0
3,300
0-0
0-0
0-0

Combined Ground
& Surface Water
0.001
0.0005
0.0001
0
0
2
0-0
0-0
0-9
0
0
37,600
0-0
0-0
0-190,900
Occurrence Summary and Use Support Document
224
March 2002

-------
3.9.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of fluoride in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Exposure Assessment of Hexachlorobenzene in Public Drinking Water Supplies"
(Wade Miller, 1989), is presented in the following section. This variety of studies and information are
presented regarding levels of fluoride in drinking water nationally. (No regional studies were included.)
Note that none of the studies presented in the following section provide the quantitative analytical results
or comprehensive coverage that would enable direct comparison to the occurrence findings estimated
with the cross-section occurrence data presented in Section 3.9.4. These additional studies, however, do
enable a broader assessment of the Stage 2 occurrence estimates presented for this Six-Year Review. All
the following information in Section 3.9.5 is taken directly from "Occurrence and Exposure Assessment
of Hexachlorobenzene in Public Drinking Water Supplies" (Wade Miller, 1989).

3.9.5.1 Ground Water Sources - National Studies

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March 1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51  organic chemical contaminants, including hexachlorobenzene. Analyses were
made using gas  chromatography/mass spectrometry (Boland, 1981, as cited in Wade Miller, 1989).  The
NSP data extracted from Boland (1981, as cited in Wade Miller, 1989) showed that none of the 12
ground water supplies sampled contained detectable levels of hexachlorobenzene. The minimum
quantifiable concentration for hexachlorobenzene was 0.1 |ig/L.

3.9.5.2 Surface Water Sources - National Studies

The National Screening Program for Organics in Drinking Water (NSP) (see section 3.9.5.1) also
reported the occurrence of hexachlorobenzene in finished drinking water from surface water sources.
The NSP data extracted from Boland (1981, as cited in Wade Miller, 1989) showed that none of the 104
surface water supplies sampled contained detectable levels of hexachlorobenzene. The minimum
quantifiable concentration for hexachlorobenzene was 0.1 |ig/L.

3.9.5.3 Unspecified Water Sources

Several studies provided data on the occurrence of hexachlorobenzene in drinking water supplies but
failed to specify water sources. Drinking water supplies from 83 locations in USEPA Region V were
analyzed for various pesticides and other organic chemicals.  Hexachlorobenzene was detected in two
finished supplies at concentrations of 0.004 |ig/L and 0.006 |ig/L (USEPA, 1986, as cited in Wade
Miller, 1989).

In another study reported by USEPA (1986, as cited in Wade Miller, 1989), three samples of Lake
Ontario water were analyzed for the presence of hexachlorobenzene. Hexachlorobenzene was detected in
all three samples at concentrations ranging from 0.00006 to 0.0002 |ig/L, with a mean concentration of
0.0001 |ig/L.  In Dade County, FL, hexachlorobenzene was detected in 4 out of 10 drinking water
samples analyzed for the presence of hexachlorobenzene. The mean and maximum concentrations were
0.014 and 0.68 |ig/L, respectively.  The detection limit was approximately 0.006 |ig/L.
Occurrence Summary and Use Support Document         225                                    March 2002

-------
3.9.5.4 Ground Water and Surface Water Sources - STORET

The STORET system contains approximately 80 million pieces of data, including data for drinking water
from ground water and surface water sources. The limitations with these data are the same as described
for ambient water in Section 3.9.3.1.

With these limitations in mind, a summary of the most recently obtained data for drinking water from
ground water sources is presented here (USEPA, 1988, as cited in Wade Miller, 1989).  According to
STORET, there were a total of 1,053 observations for hexachlorobenzene in ground water from February
1978 to November 1987.  In 27 of these samples, hexachlorobenzene was not detected; however, for
these the detection limit values were assigned.  In the remaining 1,026 samples, hexachlorobenzene was
reported to have been detected, but known to have been present below the reporting value.  Based on the
assigned and reported values for all 1,053 samples, STORET reported  a mean value of 3.80 |ig/L and a
range of 0.5 to 10.0 |ig/L. The standard deviation for all observations was 1.90 |ig/L. Detection limits
and other sampling information were not reported.

The USEPA STORET database similarly contains data on the occurrence of hexachlorobenzene in
drinking water from surface water sources  (USEPA,  1988, as cited in Wade Miller,  1989).  Between
February 1978 and July 1987, there were 54 observations for hexachlorobenzene in surface water.  One
positive observation was reported in March 1978 at a value of 0.10 |ig/L. In 5 samples,
hexachlorobenzene was not detected, but were assigned to the detection limit values. In the remaining 48
observations, hexachlorobenzene was reported as being detected but was known to be present at levels
below the reported value. Based on the reported and assigned values for all 54 samples, STORET
reported an overall mean value of 7.17 |ig/L and a range of 0.1 to 10.0 |ig/L.  The standard deviation for
all observations was 3.03 |ig/L. Detection limits and other sampling information were not reported.

3.9.6  Conclusion

In summary, hexachlorobenzene is not manufactured as a commercial end product in the U.S. and also
has no commercial end uses. It does not occur naturally in the  environment, and is mostly formed as a
byproduct in the manufacture of chlorinated solvents, in the waste streams of chloralkali and wood-
preserving plants, in fly ash, and in flue gas effluents from municipal incineration.  Hexachlorobenzene
was at one time used as a pesticide, but that use was canceled in 1984.  Although data indicates that there
are manufacturers and processors  of hexachlorobenzene, it has not been commercially produced since the
1970s. Industrial releases of hexachlorobenzene have been reported to TRI from 11 States since 1988.
There is no ambient  occurrence survey data available for hexachlorobenzene. The  Stage 2 analysis of
hexachlorobenzene,  based on the  16-State cross-section, estimated that approximately zero percent of
combined ground water and surface water systems serving zero percent of the population exceeded the
MCL of 0.001 mg/L. Based on this estimate, zero PWSs nationally are estimated to have
hexachlorobenzene epoxide levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for hexachlorobenzene from 11 States,
including 6 of 16 cross-section States. The use and commercial production of hexachlorobenzene has
been eliminated nationwide. Five of the eight States listed in ATSDR  as manufacturers and/or
processors of hexachlorobenzene are cross-section States. The cross-section should adequately represent
the occurrence of hexachlorobenzene on a national scale based upon the use, production, and release
patterns of the 16-State cross-section in relation to the patterns observed for all 50 States.
Occurrence Summary and Use Support Document         226                                     March 2002

-------
3.9.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 2000. Toxicological Profile for
       Hexachlorobenzene.  U.S. Department of Health and Human Services, Public Health Service.
       308 pp. + Appendices. Available on the Internet at:
       http://www.atsdr.cdc .gov/toxprofiles/tp90 .pdf

Boland, P.A.  1981. National Screening Program for Organics in Drinking Water. Prepared by SRI
       International for the Office of Drinking Water, U.S. Environmental Protection Agency,
       Washington, DC.  Contract No. 68-01-4666.

Clark, T., K. Clark, S. Paterson, D. MacKay, and R.J. Norstrom.  1988.  Wildlife monitoring, modelling,
       and fugacity.  Environ. Sci. Technol v. 22, no. 2, pp. 120-127.

Staples, C.A., A.F. Werner, and T.J. Hoogheem.  1985. Assessment of priority pollutant concentrations
       in the United States using STORET database. Environ. Toxicol. Chem.  v. 4, pp. 131-142.

USEPA.  1986. Exposure Assessment for Hexachlorobenzene. Final Report. EPA Report 560-5-86-019.
       Washington, DC:  Office of Pesticides and Toxic Substances, USEPA.

USEPA.  1988. Computer printout of STORET water quality database.  Retrieval conducted March 23,
       1988 by Science Applications International Corporation. Data available through Office of Water
       Regulations and Standards, USEPA, Washington, DC.

USEPA.  2000. TRIExplorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National  Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005,  Office of
       Water, 55 pp.

Wade Miller Associates, Inc. 1989.  Occurrence and Exposure Assessment of Hexachlorobenzene in
       Public Drinking Water Supplies - DRAFT.  Draft report submitted to EPA for review May 15,
       1989.
Occurrence Summary and Use Support Document         227                                    March 2002

-------
3.10   Hexachlorocyclopentadiene
Table of Contents

3.10.1 Introduction, Use and Production  	 229
3.10.2 Environmental Release  	 229
3.10.3 Ambient Occurrence 	 230
3.10.4 Drinking Water Occurrence Based on the 16-State Cross-Section	 231
3.10.5 Additional Drinking Water Occurrence Data  	 235
3.10.6 Conclusion	 236
3.10.7 References  	 237
Tables and Figures

Table 3.10-1: Facilities that Manufacture or Process Hexachlorocyclopentadiene  	 229

Table 3.10-2: Environmental Releases (in pounds) for Hexachlorocyclopentadiene
       in the United States, 1988-1999  	 230

Table 3.10-3: Stage 1 Hexachlorocyclopentadiene Occurrence Based on 16-State
       Cross-Section - Systems	 232

Table 3.10-4: Stage 1 Hexachlorocyclopentadiene Occurrence Based on 16-State
       Cross-Section - Population	 233

Table 3.10-5: Stage 2 Estimated Hexachlorocyclopentadiene Occurrence Based on 16-State
       Cross-Section - Systems	 233

Table 3.10-6: Stage 2 Estimated Hexachlorocyclopentadiene Occurrence Based on 16-State
       Cross-Section - Population	 234

Table 3.10-7: Estimated National Hexachlorocyclopentadiene Occurrence - Systems and
       Population Served	 235
Occurrence Summary and Use Support Document         228                                    March 2002

-------
3.10.1  Introduction, Use and Production

Hexachlorocyclopentadiene, also known as HEX or HCCPD, is a manufactured chemical that does not
occur naturally. It is a light, lemon-yellow liquid that has a sharp musty odor.
Hexachlorocyclopentadiene easily evaporates into the air and its vapor looks like a blue haze. Most
hexachlorocyclopentadiene  in the environment results from releases during its production and disposal
(ATSDR, 1999).

The greatest use of hexachlorocyclopentadiene is as a raw material in manufacturing other chemicals,
including pesticides, flame retardants, resins, dyes, pharmaceuticals, plastics, etc.
Hexachlorocyclopentadiene has no end uses of its own (USEPA, 2001).

Hexachlorocyclopentadiene is a key intermediate in the production of chlorinated cyclodiene pesticides,
including aldrin, dieldrin, endrin, chlordane, hepachlor, kepone, endosulfan, pentac, isodrin, and mirex.
It is also used as an intermediate in the manufacture of flame retardants and, to a lesser extent, in the
manufacture of nonflammable resins, polyester resins, pharmaceuticals, unbreakable plastics, acids,
esters, ketones, fluorocarbons, and dyes. It has previously been used as a biocide (ATSDR, 1999).

The only current commercial producer of hexachlorocyclopentadiene is the Velsicol Chemical Company
in Memphis, TN.  Because production is limited to a single producer, information on the current
production volume of hexachlorocyclopentadiene is not available. Estimates of past production, based
on production volumes of chlorinated cyclodiene pesticides, indicate that production volume for
hexachlorocyclopentadiene  was about 50 million pounds per year in the  early 1970s.  Due to regulatory
restrictions on many of the organochlorine pesticides using hexachlorocyclopentadiene as a chemical
intermediate, production volume dropped to between 8 to 15 million pounds per year in the late 1970s. It
has also been estimated that 18 million pounds of hexachlorocyclopentadiene were produced in 1983, but
more recent data is not available (ATSDR, 1999).  Only two pesticides which use
hexachlorocyclopentadiene  as an intermediate, endosulfan and pentac, are currently registered for use in
the U.S.

Table 3.10-1 shows the facilities that manufacture or process hexachlorocyclopentadiene, the activities
and uses of the product, and the range of maximum amounts on site derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1999).
Table 3.10-1:  Facilities that Manufacture or Process Hexachlorocyclopentadiene
Facility
Occidental Chemical
Morton Intl. Inc.
Velsicol Chemical Corp.
Raof Pr,rp
Location"
Niagra Falls, NY
West Alexandria,
Memphis, TN
Rpniimnnt TV
Range of maximum amounts
on site in pounds
100,000-999,999
10,000-99,999
1,000,000-9,999,999
i n nnn.qq QQQ
Activities and uses
Reactant
Reactant
Produce, on-site use/processing, sale/distribution,
T?^nptnnt
Tost office State abbreviations used
Source: ATSDR, 1999 compilation of TRI961998 data
3.10.2 Environmental Release
Hexachlorocyclopentadiene is listed as a Toxics Release Inventory (TRI) chemical.  Table 3.10-2
illustrates the environmental releases for hexachlorocyclopentadiene from 1988 - 1999. (There are only


Occurrence Summary and Use Support Document         229                                      March 2002

-------
hexachlorocyclopentadiene data for these years.)  Air emissions, which constitute most of the on-site
releases, decreased substantially from 1989 through 1993, and then rose briefly before declining to their
lowest level on record in 1999.  Surface water discharges have been consistently at very low levels, and
have been at or near zero since 1992. Although underground injection has varied from 5 to 2,131
pounds, levels remained at 250 pounds from 1993 to 1998. No information on releases to land was
reported prior to 1998, and since then readings have shown no trend. Off-site releases have generally
decreased, from a high of 28,000 pounds in 1988 to a low of 567 pounds in 1998. Overall, total on- and
off-site releases have mostly mirrored the changes in air emissions, with mild fluctuations and a declining
trend. These TRI data for hexachlorocyclopentadiene were reported from nine States, with three States
(Texas,  Tennessee, and New York) reporting every year (USEPA, 2000). Four of the nine States are in
the 16-State cross-section (used for analyses of hexachlorocyclopentadiene occurrence in drinking water;
see Section 3.10.4). (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 3.10-2:  Environmental Releases (in pounds) for Hexachlorocyclopentadiene in the United
States, 1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
1,098
5,791
6,927
7,966
8,311
8,923
3,765
8,380
25,461
84,585
89,246
78,317
Surface Water
Discharges
1
0
3
0
6
1
1
0
23
10
6
6
Underground
Injection
—
250
250
250
250
250
250
5
5
5
250
2,131
Releases
to Land
0
5,520
—
—
—
—
—
—
—
—
—
—
Off-Site Releases
903
567
930
1,000
2,995
—
—
2,740
3,000
5,000
1,204
28,470
Total On- &
Off-site
Releases
2,002
12,128
8,110
9,216
11,562
9,174
4,016
11,125
28,489
89,600
90,706
108,924
 Source: USEPA, 2000
3.10.3 Ambient Occurrence

The most comprehensive and nationally consistent data describing ambient water quality in the United
States are being produced through the United States Geological Survey's (USGS) National Water Quality
Assessment (NAWQA) program. However, national NAWQA data, as well as NURP and NPDES data,
are currently unavailable for hexachlorocyclopentadiene.

3.10.3.1 Additional Ambient Occurrence Data

Additional studies of ambient surface water data are summarized below.  (No data were available on the
occurrence  of hexachlorocyclopentadiene in ground water sources.)  A summary document entitled
"Occurrence and Exposure Assessment of Hexachlorocyclopentadiene in Public Drinking Water
Supplies" (Wade Miller, 1989), was previously prepared for past USEPA assessments of
hexachlorocyclopentadiene. The following information is taken directly from that document.

The USEPA's computerized water quality data base known as STORET was devised to assist Federal and
State institutions meet objectives of Public Law 92-500 to maintain and enhance the physical, chemical,
Occurrence Summary and Use Support Document
230
March 2002

-------
and biological quality of the nation's ambient waterways by providing for the collection and
dissemination of basic water quality data (Staples et al., 1985, as cited in Wade Miller, 1989). Data are
collected by States, EPA regional offices, and other government agencies and are maintained in the
STORET system.

Before presenting a summary of the ambient water data in STORET, it is important to note that there are
significant limitations in using the data base to estimate representative concentrations of a contaminant
such as hexachlorocyclopentadiene. Data entered into STORET are gathered from an array of studies
conducted for various purposes. Analyses are conducted in a number of different laboratories employing
different methodologies with a range of detection limits.  In many cases, detection limits are not reported,
making the reliability of the data highly questionable. Where detection limits have been reported,
STORET assigns the detection limit value to those observations reported as not detected.  This can lead
to errors in interpretation and overestimation of concentrations in cases in which there is a preponderance
of nondetectable values. Additionally, a few high values can inflate mean values and result in large
standard deviations relative to the means (Staples et al., 1985, as cited in Wade Miller, 1989). Very high
values may not be correct, as they may reflect sample contamination or analytical error and can
significantly distort assessment of average concentrations. Staples et al. (1985, as cited in Wade Miller,
1989) also notes that the use of data collected prior to the 1980s is not recommended, since such data was
obtained using less sensitive laboratory techniques and quality assurance procedures  were not yet
mandated for the data entered into the system.

The STORET water quality data base provides information on the occurrence of contaminants at ambient
water stations in U.S. waterways. A summary of this information was obtained for
hexachlorocyclopentadiene in ambient waters. Ambient sites include streams, lakes,  ponds, wells,
reservoirs, canals, estuaries, and oceans.  While the preponderance of data were collected from surface
water sources, the number of samples collected from ground water wells, relative to the total number of
samples collected from all ambient sites combined, is unspecified (Staples et al., 1985, as cited in Wade
Miller, 1989). Staples et al. (1985, as cited in Wade Miller, 1989) have summarized  data from the 1980's
only; that is, data from  1980 through 1983.  This was done based on the number of data points and the
likelihood that better quality assurance practices have been employed in more recent  years.  In the
absence of sophisticated statistical analyses to eliminate improbable data, median values have been
reported. The median value is sensitive to extreme values, and reflects a measure of  central tendency
more accurately than the mean value in the presence  of these extreme values (Staples et al., 1985, as cited
in Wade Miller, 1989).

For a total of 854 observations from ambient water stations, the median concentration of
hexachlorocyclopentadiene was 10.0 |ig/L.  Of the total number of observations, 0.1  percent were
reported as detectable.  Detection limits and other sampling information were not reported.

3.10.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of hexachlorocyclopentadiene occurrence presented in the following section is based on
State compliance monitoring data from the 16 cross-section States.  The 16-State cross-section is the
largest and most comprehensive compliance monitoring data set compiled by EPA to date. These data
were evaluated relative to several concentration thresholds of interest:  0.05 mg/L; 0.04 mg/L; and 0.005
mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
hexachlorocyclopentadiene. These data represent more than 52,000 analytical results from
approximately 14,000 PWSs during the period from 1984 to 1998 (with most analytical results from 1992

Occurrence Summary and Use Support Document          231                                      March 2002

-------
to 1997). The number of sample results and PWSs vary by State, although the State data sets have been
reviewed and checked to ensure adequacy of coverage and completeness.  The overall modal detection
limit for hexachlorocyclopentadiene in the 16 cross-section States is equal to 0.005 mg/L. (For details
regarding the 16-State cross-section, please refer to Section  1.3.5 of this report.)

3.10.4.1 Stage 1 Analysis Occurrence Findings

Table 3.10-3 illustrates the very low occurrence of hexachlorocyclopentadiene in drinking water for the
public water systems in the  16-State cross-section. No ground water or surface water PWSs had any
analytical results exceeding the MCL  (0.05 mg/L) or 0.04 mg/L. Only 1 (0.0722% of) surface water
system had any analytical results greater than the modal detection limit (0.005  mg/L). No ground water
systems had analytical results greater  than any of the three thresholds.
Table 3.10-3: Stage 1 Hexachlorocyclopentadiene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.05
0.04
0.005
Percent of Systems
Exceeding Threshold
0.000%
0.000%
0.000%
Number of Systems
Exceeding Threshold
0
0
0

Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.0722%
0
0
1

Combined Ground &
Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.00718%
0
0
1
Reviewing hexachlorocyclopentadiene occurrence in the 16 cross-section States by PWS population
served (Table 3.10-4) shows that approximately 0.0296% of the total 16-State population was served by
PWSs with analytical detections of hexachlorocyclopentadiene greater than 0.005 mg/L (about 27,700
people, all served by one surface water system).
Occurrence Summary and Use Support Document
232
March 2002

-------
Table 3.10-4: Stage 1 Hexachlorocyclopentadiene Occurrence Based on 16-State Cross-Section
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.05
0.04
0.005
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.000%
0.000%
Total Population Served
by Systems Exceeding
Threshold
0
0
0

Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.0509%
0
0
27,700

Combined Ground &
Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.0296%
0
0
27,700
3.10.4.3  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.10-5 and
3.10-6. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations (USEPA, 2002)).

No ground water or surface water PWSs (therefore, no population served by systems) had an estimated
mean concentration of hexachlorocyclopentadiene exceeding 0.05 mg/L, 0.04 mg/L or 0.005 mg/L.
Table 3.10-5: Stage 2 Estimated Hexachlorocyclopentadiene Occurrence Based on 16-State Cross-
Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.05
0.04
0.005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Occurrence Summary and Use Support Document
233
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
Range
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
Range

Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
Table 3.10-6:  Stage 2 Estimated Hexachlorocyclopentadiene Occurrence Based on 16-State Cross-
Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.05
0.04
0.005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.05
0.04
0.005
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
3.10.4.3  Estimated National Occurrence

As illustrated in Table 3.10-7, the Stage 2 analysis estimated zero systems serving none of the national
population to have an estimated mean concentration values of hexachlorocyclopentadiene greater than
0.05 mg/L, 0.04 mg/L, or 0.005 mg/L.  (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)
Occurrence Summary and Use Support Document
234
March 2002

-------
Table 3.10-7:  Estimated National Hexachlorocyclopentadiene Occurrence - Systems and
Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.05
0.04
0.005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.05
0.04
0.005
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.05
0.04
0.005
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
3.10.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of fluoride in drinking water are also reviewed.
For a previous summary document entitled "Occurrence and Exposure Assessment of
Hexachlorocyclopentadiene in Public Drinking Water Supplies" (Wade Miller, 1989), an extensive
search of the scientific and regulatory literature was conducted in an effort to obtain information on the
occurrence of hexachlorocyclopentadiene in drinking water from ground water and surface water
sources. In addition, knowledgeable sources within the Office of Drinking Water were contacted.
Except for data obtained from the STORET data base, no national, regional or State studies were found
which analyzed the occurrence of hexachlorocyclopentadiene in drinking water supplies. The STORET
data are discussed below.  Note that STORET data do not provide the quantitative analytical results or
comprehensive coverage that would enable direct comparison to the occurrence findings estimated with
the cross-section occurrence data presented in Section 3.10.4.  These additional studies, however, do
enable a broader assessment of the Stage 2 occurrence estimates presented for this Six-Year Review. All
the following information in Section 3.1.5 is taken directly from "Occurrence and Exposure Assessment
of Hexachlorocyclopentadiene in Public Drinking Water Supplies" (Wade Miller, 1989).

3.10.5.1 Ground Water and Surface Water Sources  - STORET

The STORET system contains approximately 80 million pieces of data, including data for drinking water
from ground water and surface water sources.  The limitations with these  data are the same as described
for ambient water in Section 3.10.3.1.

With these limitations in mind, a summary of the most recently obtained data for drinking water from
ground water sources is presented here (USEPA, 1988,  as cited in Wade Miller, 1989). According to
STORET, there were  a total of 1,042 observations for hexachlorocyclopentadiene in ground water from
February 1978 to November 1987. In 27 of these samples, hexachlorocyclopentadiene was not detected;
Occurrence Summary and Use Support Document
March 2002

-------
however, for these the detection limit values were assigned. In the remaining 1,015 samples,
hexachlorocyclopentadiene was reported to have been detected, but known to have been present below
the reported value. Based on the assigned and reported values for all 1,042 samples, STORET reported a
mean value of 14.35 |ig/L and a range of 1.0 to 40.0 |ig/L.  The standard deviation for all observations
was 7.23 |ig/L. Detection limits and other sampling information were not reported.

STORET also contains data on the occurrence of hexachlorocyclopentadiene in drinking water from
surface water sources (USEPA, 1988, as cited in Wade Miller,  1989). Between February 1978 and July
1987, there were 62 observations for hexachlorocyclopentadiene in surface water. In 16 samples,
hexachlorocyclopentadiene was not detected, but were assigned the detection limit values. In the
remaining 46 observations hexachlorocyclopentadiene was reported as being detected but was known to
be present at levels below the reported values. Based on the reported and assigned values for all 62
samples, STORET reported an overall mean value of 8.95 |ig/L and a range of 4.0 to 30.0 |ig/L.  The
standard deviation for all observations was 4.61 |ig/L. Detection limits and other sampling information
were not reported.

3.10.5.2  National Estimates of the Occurrence of Hexachlorocyclopentadiene and Population
Exposure for Public Water Supplies

No Federal survey data on the occurrence of hexachlorocyclopentadiene in public drinking water
supplies are available. Information on the occurrence of hexachlorocyclopentadiene in public drinking
water supplies is available from the STORET data base; however, there are significant limitations
involved with utilizing these data.  These limitations are discussed in detail in Section 3.10.5.1.
Consequently, national estimates of occurrence and population exposure could not be made for
hexachlorocyclopentadiene.

3.10.6  Conclusion

Hexachlorocyclopentadiene is a manufactured chemical that does not occur naturally. Its greatest use is
as a raw material in manufacturing other chemicals, including pesticides, flame retardants, resins, dyes,
Pharmaceuticals, plastics, etc.  The only commercial producer of hexachlorocyclopentadiene  is Velsicol
Chemical Corp. in Memphis, Tennessee, and other production or processing of
hexachlorocyclopentadiene is relatively limited.  Industrial releases of hexachlorocyclopentadiene  have
been reported to TRI from nine States since  1988. Ambient occurrence data was not available for
hexachlorocyclopentadiene. The Stage 2 analysis, based on the 16-State cross-section, estimated that
zero percent of combined ground water and surface water systems serving zero percent of the population
exceeded the MCL of 0.05 mg/L. Based on this estimate, zero  PWSs nationally are estimated to have
hexachlorocyclopentadiene levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors.  Nationally, hexachlorocyclopentadiene is manufactured and/or processed in 4 States
and has TRI releases in 9 States. Hexachlorocyclopentadiene is manufactured and/or processed in 1 out
of the 16 cross-section States and has TRI releases in 4 of the 16 cross-section States. The cross-section
should adequately represent the occurrence of hexachlorocyclopentadiene on a national scale based upon
the use, production, and  release patterns of the 16-State cross-section in relation to the patterns observed
for all 50 States.
Occurrence Summary and Use Support Document         236                                      March 2002

-------
3.10.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 1999. Toxicological Profile for
       Hexachlorocyclopentadiene. U.S. Department of Health and Human Services, Public Health
       Service.  185 pp. + Appendices. Available on the Internet at:
       http://www.atsdr.cdc.gov/toxprofiles/tp 112.pdf

Staples, C.A., A.F. Werner, and T.J. Hoogheem.  1985.  Assessment of Priority Pollutant Concentrations
       in the United States Using STORET DataBase. Environ. Toxicol. Chem.  v. 4, pp. 131-142.

USEPA.  1988.  Computer Printout of STORET Water Quality Data Base. Retrieval conducted March
       23, 1988 by Science Applications International Corporation. Data available through Office of
       Water Regulations and Standards, USEPA, Washington, DC.

USEPA. 2000. TR1Explorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001.  National Primary Drinking Water Regulations - Consumer Factsheet on:
       Hexachlorocyclopentadiene. Office of Ground Water and Drinking Water, USEPA. Available
       on the Internet at: http://www.epa.gov/safewater/dwh/c-soc/hexachl2.html, last updated April 12,
       2001.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Wade Miller Associates, Inc. 1989. Occurrence and Exposure Assessment of
       Hexachlorocyclopentadiene in Public Drinking Water Supplies - DRAFT.  Draft report submitted
       to EPA for review May 15, 1989.
Occurrence Summary and Use Support Document         237                                    March 2002

-------
3.11   Oxamyl
Table of Contents

3.11.1  Introduction, Use and Production  	  239
3.11.2  Environmental Release  	  240
3.11.3  Ambient Occurrence 	  240
3.11.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  242
3.11.5  Additional Drinking Water Occurrence Data 	  246
3.11.6  Conclusion	  247
3.11.7  References  	  247
Tables and Figures

Figure 3.11-1:  Estimated Annual Agricultural Use for Oxamyl (1992)	  240

Table 3.11-1: Oxamyl Detections and Concentrations in Surface Water and Ground Water  	  241

Table 3.11-2: Stage 1 Oxamyl Occurrence Based on  16-State Cross-Section - Systems	  242

Table 3.11-3: Stage 1 Oxamyl Occurrence Based on  16-State Cross-Section - Population	  243

Table 3.11-4: Stage 2 Estimated Oxamyl Occurrence Based on 16-State Cross-Section -
       Systems	  244

Table 3.11-5: Stage 2 Estimated Oxamyl Occurrence Based on 16-State Cross-Section -
       Population	  245

Table 3.11-6: Estimated National Oxamyl Occurrence - Systems and Population Served 	  246
Occurrence Summary and Use Support Document         238                                    March 2002

-------
3.11.1  Introduction, Use and Production

Pure oxamyl, also known as Vydate (or methyl-N',N'-dimethyl-N-[methylcarbamoyl)oxy]-l-thio-
oxamimidate) is an off-white crystalline powder or a white crystalline solid with a slight sulfurous odor.
While oxamyl, a carbamate, is stable in solid form and in most solutions, it decomposes to innocuous
materials in natural waters and in soils. These insecticides work by blocking the normal functioning of
cholinesterase, an essential nervous system enzyme.  Aeration, sunlight, alkalinity, and higher
temperatures increase its rate of decomposition.  Toxic fumes of nitrogen oxides and sulfur oxides are
emitted if oxamyl is heated to decomposition. Oxamyl is noncorrosive. It is available in a variety of
formulations, such as a 24% water- soluble liquid and 10% granules (EXTOXNET, 1998).  Oxamyl is
used to kill and control a broad spectrum of insects, as an acaricide to control mites and ticks, and as a
nematicide against roundworms. Its action is both systemic and contact.  Oxamyl may be applied directly
on plants or on the soil surface (USEPA, 2001).

In 1982, U.S. production of oxamyl was reported to be 400,000 pounds (USEPA, 2001), but production
appears to have dropped off by the late 1980s (Wade Miller, 1989). A systemic and contact
insecticide/acaricide and nematocide, oxamyl is a restricted use pesticide used on apples, bananas,
carrots, celery, citrus, cotton, cucumbers, eggplants, garlic, ginger, muskmelon (including cantaloupe and
honeydew melon), onion (dry bulb), peanuts, pears, peppers, peppermint, pineapples, plantains, potatoes,
pumpkins, soybeans, spearmint, squash, sweet potatoes, tobacco, tomatoes, watermelons, and yams.
Oxamyl is also used on non-bearing apple, cherry, citrus, peach, pear, and tobacco (USEPA, 2001).

Recent national estimates of agricultural use for oxamyl are available.  The United States Geological
Survey (USGS, 1998a) estimates approximately  750,000 pounds of oxamyl active ingredient were used
in 1992, and were applied to roughly 1.6 million acres (USGS, 2000).  These estimates were derived
using State-level data sets on pesticide use rates available from National Center for Food and
Agricultural Policy (NCFAP) combined with county-level data on harvested crop acreage from the
Census of Agriculture  (Thelin and  Gianessi, 2000).  EPA estimates that on average approximately
800,000 pounds of oxamyl  active ingredient (ai)  are used per year (USEPA, 2000a). Cotton accounts for
the majority of usage (600,000 pounds oxamyl ai), while intermediate use can be found on several other
crops as well (e.g., apples, celery, potatoes, and tomatoes, see Figure 3.11-1). Although cotton accounts
for most of the oxamyl usage, it is used on only 7% of cotton produced annually in the United States
(USEPA, 2000a).  Cotton application is 1 to 2 times per season, usually at a rate of about 0.4 Ib oxamyl
ai per acre. When oxamyl is used on other crops, it is generally applied 1 to 3 times per season at
between 0.2 and 1.0 Ib ai per acre (the current label rates do allow for higher use rates on some crops)
(USEPA, 2000a).

Figure  3.11-1 shows the USGS (1998a) derived geographic distribution of estimated average annual
oxamyl use in the  United States for 1992. Oxamyl is used in cotton production as well as on a variety of
fruit and root crops, its use  is geographically distributed across the United States. The two largest
concentrations of oxamyl use are seen in the cotton production regions of the Texas panhandle and the
lower Mississippi  River Valley, including Alabama.  Texas ranks as one of the largest producers of
cotton, and therefore has the highest oxamyl usage (USEPA, 2000a).  Oxamyl's application on potato,
fruit, and vegetable production is apparent with high  oxamyl use identified in the Pacific Northwest,
California, and Florida (Figure 3.11-1). While non-agricultural uses are not reflected here (USGS,
1998b), existing data suggest that non-agricultural use of oxamyl is minimal to non-existent (USEPA,
2000b). In addition, the map does not provide any further resolution that county-level, obscuring any
intra-county usage differences (USGS, 1998b). A comparison of this use map with the map of the 16
cross-section States (Figure 1.3-1)  shows that States across the range of high of low oxamyl use are well
represented in the cross-section.

Occurrence Summary and Use Support Document         239                                    March 2002

-------
Figure 3.11-1:  Estimated Annual Agricultural Use for Oxamyl (1992)
                                            OXAMYL
                                   ESTIMATED ANNUAL AGRICULTURAL USE
                      Average use of
                      Active Ingredient
                    Pounds per square mile
                      of county par year
                      D  No Estimated Use
                      D   < 0.004
                      D 0.004-0.016
                      D 0.017-D.078
                      D 0.080 - D.467
                      •   >= 0.468
Crops
cotton
potatoes
apptes
mint
swart peppers
eatery
tomatoes
cucumbers
oriortG
rnslons
Total
Pounds Applied
477,086
75,635
72,319
31,554
27,967
23,599
21,537
12,269
11,075
9,378
Percent
National Use
58.62
9.29
& 86
7.58
344
2.90
2.65
1.61
1.3G
1.15
Source: USGS 1998a
3.11.2 Environmental Release

Oxamyl is released directly to the environment when used as an insecticide and potentially during its
manufacture, handling and storage. EPA estimated that 400,000 pounds of oxamyl were produced in the
U.S. in 1982 (USEPA, 2001). Oxamyl is not listed as a Toxics Release Inventory (TRI) contaminant, so
no TRI release records are maintained. Therefore, the use of oxamyl (described in the previous section)
may provide the primary indication of where releases are most likely. The areas of highest oxamyl use
are in the cotton production regions of the Texas panhandle, the lower Mississippi River Valley, and on
the fruit and vegetable production areas in the Pacific Northwest, California, and Florida. These and
other agricultural use areas are illustrated in Figure 3.11-1.

3.11.3 Ambient Occurrence

Oxamyl is an analyte for both surface and ground water NAWQA studies, with a method detection limit
(MDL) of 0.018 LLg/L. Additional information on analytical methods used in the NAWQA study units,
including method detection limits, are described by Gilliom and others (1998).

Oxamyl is an analyte for both surface and ground water NAWQA studies. Table 3.11-1 summarizes the
findings of USGS NAWQA sampling for oxamyl within the first 20 NAWQA study basins.  Oxamyl
concentrations in all of the surface water samples were below the detection limit in most sites, although it
was detected in at least one sample. No oxamyl was detected in any of the ground water NAWQA study
sites.
Occurrence Summary and Use Support Document
240
March 2002

-------
Table 3.11-1:  Oxamyl Detections and Concentrations in Surface Water and Ground Water
                              Detection frequency
                                (% of samples)
                                           Concentration percentiles
                                             (all samples; |J.g/L)
   surface -water

       agricultural

           urban

        integrator

          all sites
                     all samples   >0.01 ug/L    > 0.05 ug/L
 ND

 ND

 ND

0.03%
                                             median
95th
ND
ND
ND
0.03%

-------
3.11.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of oxamyl occurrence in drinking water presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date.  These data were
evaluated relative to several concentration thresholds of interest: 0.2 mg/L;  0.04 mg/L; 0.03 mg/L; 0.02
mg/L; and 0.007 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
oxamyl.  These data represent more than 47,000 analytical results from approximately 13,000 PWSs
during the period from 1984 to 1998 (with most analytical results from 1992 to 1997). The number of
sample results and PWSs vary by State, although the State data sets have been reviewed and checked to
ensure adequacy of coverage and completeness. The overall modal detection limit for oxamyl in the 16
cross-section States is equal to 0.002 mg/L.  (For details regarding the 16-State cross-section, please refer
to Section 1.3.5 of this report.)

3.11.4.1  Stage 1 Analysis Occurrence Findings

Table 3.11-2 illustrates the very low occurrence of oxamyl in drinking water for the public water systems
in the 16-State cross-section .  No ground water or surface water PWSs had any analytical results greater
than 0.2 mg/L, 0.04 mg/L, or 0.03 mg/L. About 0.00760% of total ground and surface water systems had
analytical results greater than 0.02 mg/L (approximately 1 system). Approximately  0.0304% of systems
(4 systems) had any analytical detections greater than 0.007 mg/L.  Only 1 (0.00848% of) ground water
system had any exceedances of 0.02 mg/L.  Four (0.0339% of) ground water systems had at least one
analytical result greater than 0.007 mg/L. No surface water systems had analytical results greater than
any of the thresholds.
Table 3.11-2:  Stage 1 Oxamyl Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.2
0.04
0.03
0.02
0.007
Percent of Systems
Exceeding Threshold
0.000%
0.000%
0.000%
0.00848%
0.0339%
Number of Systems
Exceeding Threshold
0
0
0
1
4

Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.000%
0
0
0
0
0
Occurrence Summary and Use Support Document
242
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems
Exceeding Threshold

Combined Ground &
Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.00760%
0.0304%
0
0
0
1
4
Reviewing oxamyl occurrence in the 16 cross-section States by PWS population served (Table 3.11-3)
shows that approximately 0.00162% of the population (about 1,500 people) was served by PWSs with
analytical detections of oxamyl greater than 0.02 mg/L.  The number of people exposed to oxamyl
drastically increased to about 51,300 (approximately 0.0555%) when evaluated relative to 0.007 mg/L.
Approximately 0.00391% of people served by ground water systems were exposed to oxamyl at levels
greater than 0.02 mg/L. The percentage of population served by ground water systems exposed to
oxamyl at levels greater than 0.007 mg/L was equal to 0.134%. When evaluated relative to 0.2 mg/L,
0.04 mg/L, or 0.03 mg/L, the percent of population exposed to oxamyl was equal to 0% for all system
types. No surface water systems exceeded any of the specified health thresholds.
Table 3.11-3:  Stage 1 Oxamyl Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.2
0.04
0.03
0.02
0.007
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.000%
0.000%
0.00391%
0.134%
Total Population Served by
Systems Exceeding
Threshold
0
0
0
1,500
51,300

Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.000%
0
0
0
0
0

Combined Ground &
Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.00162%
0.0555%
0
0
0
1,500
51,300
Occurrence Summary and Use Support Document
243
March 2002

-------
3.11.4.2  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.11-4 and
3.11-5. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations (USEPA, 2002)).

No ground water or surface water PWSs had an estimated mean concentration of oxamyl exceeding 0.2
mg/L, 0.04 mg/L, 0.03 mg/L, or 0.02 mg/L. Approximately 0.0000456% of all PWSs in the 16 States
(less than 1 system) had estimated mean concentration values of oxamyl greater than 0.007 mg/L. The
percentage of ground water PWSs in the 16 States estimated to have a mean concentration greater than
0.007 mg/L was equal to 0.0000509%.  No surface water PWSs had estimated mean concentration
exceeding any of the five  specified concentration thresholds.
Table 3.11-4:  Stage 2 Estimated Oxamyl Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.2
0.04
0.03
0.02
0.007
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000%
0.000%
0.0000509%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
0
0
Range
0-0
0-0
0-0
0-0
0-0

Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.0000456%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0
Occurrence Summary and Use Support Document
244
March 2002

-------
Reviewing oxamyl occurrence by PWS population served (Table 3.11-5) shows that approximately
0.0000323% of population served by all PWSs in the 16-State cross-section were potentially exposed to
oxamyl levels above 0.007 mg/L. The percentage of population served by ground water systems in the
16 States relative to 0.007 mg/L was equal to 0.0000777%.  (No surface water systems exceeded any of
the specified health thresholds.) When evaluated relative to a threshold of 0.2 mg/L, 0.04 mg/L, 0.03
mg/L, and 0.02 mg/L, the percent of population exposed was equal to 0% for all system types.
Table 3.11-5:  Stage 2 Estimated Oxamyl Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.2
0.04
0.03
0.02
0.007
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.00000%
0.00000%
0.00000%
0.00000%
0.0000777%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in tht
16 States Estimated to Exceed Threshold
Best Estimate
0
0
0
0
0
Range
0-0
0-0
0-0
0-0
0-0

Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.2
0.04
0.03
0.02
0.007
0.000%
0.000%
0.000%
0.000%
0.0000323%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0
3.11.4.3 Estimated National Occurrence

As illustrated in Table 3.11-6, the Stage 2 analysis estimates zero systems serving none of the national
population have estimated mean concentration values of oxamyl greater than 0.2 mg/L, 0.04 mg/L, 0.03
mg/L, or 0.02 mg/L. Approximately 1 ground water system serving less than 100 people nationally was
estimated to have a mean concentration value of oxamyl greater than 0.007 mg/L. (See Section 1.4 for a
description of how Stage 2 16-State estimates are extrapolated to national values.)
Occurrence Summary and Use Support Document
245
March 2002

-------
Table 3.11-6:  Estimated National Oxamyl Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.2
0.04
0.03
0.02
0.007
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
0
1
Range
0-0
0-0
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
0
<100
Range
0-0
0-0
0-0
0-0
0-0

Surface Water
0.2
0.04
0.03
0.02
0.007
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0
0
0
0
0
0
0-0
0-0
0-0
0-0
0-0

Combined Ground
& Surface Water
0.2
0.04
0.03
0.02
0.007
0
0
0
0
1
0-0
0-0
0-0
0-0
0-0
0
0
0
0
<100
0-0
0-0
0-0
0-0
0-0
3.11.5  Additional Drinking Water Occurrence Data

For previously prepared USEPA assessments of oxamyl, discussed in Section 3.11.3.1, a literature search
was conducted and knowledgeable sources within the Office of Water were contacted.  No national
studies were found on the occurrence of oxamyl in drinking water from ground water or surface water
sources nor were any regional or State studies found that addressed the occurrence of oxamyl in drinking
water from surface water sources (Wade Miller, 1989).  Although the studies reviewed include oxamyl
occurrence information and/or data, none of the studies discussed provide the quantitative analytical
results  or comprehensive coverage that would enable direct comparison to the occurrence findings
estimated with the cross-section occurrence data and presented in Section 3.11.4. All the following
information in Section 3.11.5 is taken directly from "Occurrence and Exposure Assessment of Oxamyl in
Public  Drinking Water Supplies" (Wade Miller, 1989).

One county-level study was identified.  Summary data were obtained from a drinking water well
pesticide sampling program conducted by Suffolk County Department of Health Services of
Massachusetts during 1980 to 1985. Oxamyl residues were detected at 0.001 to 0.049 mg/L in 2.9
percent (455) of the  15,535 samples tested. The average concentration was 0.01 mg/L. During the five
year period less than 0.1 percent (11) of the samples exceeded 0.05 mg/L (Dougherty, 1987, as cited in
Wade Miller, 1989). Additional summary data were obtained from a Massachusetts interagency
pesticide monitoring program of 341 public and private water sources in 27 communities. Of the 146
Occurrence Summary and Use Support Document
246
March 2002

-------
wells sampled for oxamyl, 1 was positive at the detection limit of 0.001 |ig/L (Dougherty, 1987, as cited
in Wade Miller, 1989).

3.11.6  Conclusion

A systemic and contact insecticide/acaricide and nematocide, oxamyl is a restricted use pesticide used on
apples, bananas, carrots, celery, citrus, cotton, cucumbers, eggplants, garlic, ginger, muskmelon
(including cantaloupe and honeydew melon), onion (dry bulb), peanuts, pears, peppers, peppermint,
pineapples, plantains, potatoes, pumpkins, soybeans, spearmint, squash, sweet potatoes, tobacco,
tomatoes, watermelons, and yams. Oxamyl is also used on non-bearing apple, cherry, citrus, peach, pear,
and tobacco.  Recent national estimates of agricultural use for oxamyl are available. The United States
Geological Survey estimates approximately 750,000 pounds of oxamyl active ingredient were used in
1992, and were applied to roughly 1.6 million acres. Oxamyl is not a TRI chemical, so there is no
information available on releases.  Oxamyl was an analyte for the NAWQA ambient occurrence studies.
In the NAWQA study, the median value of oxamyl for both ground water and surface water was less than
the detection limit. The Stage 2 analysis, based on the 16-State cross-section, estimated that zero percent
of combined ground water and surface water systems  serving zero percent of the population exceeded the
MCL of 0.2 mg/L. Based on this estimate, zero PWSs nationally are estimated to have oxamyl levels
greater than the MCL.

The  16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. According to information from USGS, 12 of the 16 cross-section States use oxamyl,
although most States in light to moderate amounts.  Oxamyl is used most heavily in California, Texas,
Florida, and the Northeast; these States/regions are represented by five of the cross-section States. As
there is no TRI or other data, agricultural use seems to be the only basis for comparison between the 16
States and the nation. In the case of oxamyl, the cross-section should adequately represent the
occurrence of oxamyl on a national scale based upon the use, production, and release patterns of the 16-
State cross-section in relation to the patterns observed for all 50 States.

3.11.7  References

Cohen S.Z., S.M. Creeger, R.F. Carsel, and C.G. Enfield.  1984.  Potential  Pesticide Contamination of
       Ground Water Resulting from Agricultural Uses.  In: Treatment and Disposal of Pesticide
       Wastes. Washington,  DC. R.F. Kruger and J.W. Seiber (eds.). ACS Symposium Series No. 259:
       pp. 247-325.

Dougherty, T.M.  1987. Oxamyl - EAB science chapter for the registration standard.  Memorandum from
       Therese M. Dougherty, Exposure Assessment Branch, to Dennis Edwards, Insecticide -
       Rodenticide Branch, Office of Pesticide Programs, U.S. Environmental Protection Agency,
       Arlington, VA. February 11, 1987.

EXTOXNET.  1996. Pesticide Information Profile: Oxamyl. Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://ace.ace.orst.edu/info/extoxnet/pips/oxamyl.htm, revised June, 1996.

Gilliom, R.J., O.K. Mueller, and L.H. Nowell.  1998.  Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study  Units, 1992-95.  U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589, last updated October 9, 1998.

Occurrence Summary and Use Support Document         247                                     March 2002

-------
Thelin, G.P., and L.P. Gianessi. 2000.  Method for Estimating Pesticide Use for County Areas of the
       Conterminous United States. U.S. Geological Survey Open-File Report 00-250.
       62 pp.  Available on the Internet at: http://water.wr.usgs.gov/pnsp/rep/ofr00250/ofr00250.pdf

USEPA.  2000a.  Interim Reregistration Eligibility Decision (IRED): Oxamyl. EPA Report 73 8-R-OO-
       015. Washington, DC: Office of Prevention, Pesticides, and Toxic Substances, USEPA. 135 pp.
       October 2000. Available on the Internet at: http://www.epa.gov/REDs/0253ired.pdf

USEPA.  2000b.  R.E.D. Facts: Oxamyl. EPA Report 738-F-OO-013. Washington, DC: Office of
       Prevention, Pesticides, and Toxic Substances, USEPA. 4 pp. November 2000. Available on the
       Internet at: http://www.epa.gov/oppsrrdl/REDs/factsheets/0253iredfact.pdf, last updated June
       21,2001.

USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on: OXAMYL
       (VYDATE). Available  on the Internet at: http://www.epa.gov/safewater/dwh/c-soc/oxamyl.html,
       last updated April 12, 2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS. 1998a. Annual Use Maps. Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated March 20, 1998.

USGS. 1998b. Sources & Limitations of Data Used to Produce Maps of Annual Pesticide Use.
       Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/mapex.html, last updated
       March 20, 1998.

USGS. 1998c. Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA).  PROVISIONAL DATA - SUBJECT TO
       REVISION. Available  on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/, last updated
       October 9, 1998.

USGS. 2000. PESTICIDES ANALYZED IN NAWQA SAMPLES: Use,  Chemical Analyses, and
       Water-Quality Criteria (PROVISIONAL DATA ~ SUBJECT TO REVISION). Available on the
       Internet at: http://water.wr.usgs.gov/pnsp/anstrat/, last updated August 20, 1999.

Wade Miller Associates, Inc.  1989. Occurrence and Exposure Assessment of Oxamyl in Public
       Drinking Water Supplies - DRAFT. Draft report submitted to EPA for review April 12, 1989.
Occurrence Summary and Use Support Document         248                                   March 2002

-------
3.12   Picloram
Table of Contents

3.12.1  Introduction, Use and Production  	  250
3.12.2  Environmental Release  	  251
3.12.3  Ambient Occurrence  	  252
3.12.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  253
3.12.5  Additional Drinking Water Occurrence Data  	  257
3.12.6  Conclusion	  258
3.12.7  References 	  258
Tables and Figures

Figure 3.12-1:  Picloram Estimated Annual Agricultural Use	  251

Table3.12-l: Environmental Releases (in pounds) for Picloram in the United States, 1995-1999 ...  252

Table 3.12-2: Picloram Detections and Concentrations in Surface Water and Ground Water	  252

Table 3.12-3: Stage 1 Picloram Occurrence Based on 16-State Cross-Section - Systems	  254

Table 3.12-4: Stage 1 Picloram Occurrence Based on 16-State Cross-Section - Population	  254

Table 3.12-5: Stage 2 Estimated Picloram Occurrence Based on 16-State Cross-Section -
       Systems	  255

Table 3.12-6: Stage 2 Estimated Picloram Occurrence Based on 16-State Cross-Section -
       Population	  256

Table 3.12-7: Estimated National Picloram Occurrence - Systems and Population Served	  256
Occurrence Summary and Use Support Document          249                                     March 2002

-------
3.12.1  Introduction, Use and Production

Picloram, (4-amino-3,5,6-trichloropicolinic acid) in the pyridine family of compounds, is a crystalline
organic solid with a chlorine-like odor. Picloram is formulated either as an acid (technical product) or as
a potassium salt. In salt form, it is used as a systemic herbicide. It is used for controlling annual weeds
on crops, and in combination with  2,4-D or 2,4,5-T against perennials on non-croplands for brush
control. It also controls a wide range of broad-leaved weeds excepting mustards (crucifers). Most
grasses are  resistant to picloram  so it is used in range management programs.  Picloram is used to control
bitterweed, knapweed, leafy spurge, locoweed, larkspur, mesquite, prickly pear, and snakeweed on
rangeland in the western States (USEPA, 2001).  Commercial products containing picloram have trade
names  including Grazon and Tordon.  The compound may be used in formulations with other herbicides
such as bromoxynil, atropine, diuron, 2,4-D, MCPA, triclorpyr, and atrazine among others. Picloram is
also compatible with fertilizers (EXTOXNET, 2001).

Picloram is a systemic herbicide used to control deeply rooted herbaceous weeds and woody plants in
rights-of-way, forestry, rangelands, pastures, and small grain crops (USEPA, 1995). It is applied in the
greatest amounts to pasture and rangeland, followed by forestry. Picloram acid is a manufacturing use
product with no end uses. Picloram products have no household or residential uses. All picloram
products  are classified as Restricted Use Pesticides, based on hazard to nontarget plants, and may be
applied only by or under the direct supervision of certified applicators (USEPA, 1995).

No information was found that suggested any natural production of picloram.  Picloram is produced
domestically for commercial markets by Dow Chemical USA, and by Union Carbide (Wade Miller,
1989).  Picloram is produced by chlorination of a-picoline (2-methylpyridine) and reaction with
ammonia, followed by hydrolysis of the intermediate 4-amino-3,5,6, a, a, a-hexachloropicoline (Kirk-
Othmer,  1980, as cited in Wade Miller, 1989). Domestic production of picloram was estimated to be
between 2.2 and 2.9 million pounds in 1981; approximately 1.4 to 1.9 million pounds were exported
(Schutte, 1982, as cited in Wade Miller, 1989).

Picloram is registered for use on pastureland and rangeland grasses in the following States: Alabama,
Arkansas, Colorado, Georgia, Kansas, Kentucky, Louisiana, Mississippi, Missouri, New Mexico, North
Carolina, Oklahoma, South Carolina, Tennessee, Texas, Virginia, and West Virginia (USEPA, 1981, as
cited in Wade Miller, 1989).  Picloram is registered for use on uncultivated agricultural areas, rights-of-
way, commercial and industrial premises, forests, and drainage ditch banks in the following States:
California,  Colorado, Hawaii, Idaho, Minnesota, Nebraska, Oregon, South Dakota, Utah, Washington,
West Virginia, and Wyoming (USEPA, 1981, as cited in Wade Miller, 1989).

Recent national estimates of agricultural use for picloram are available. The United States Geological
Survey (USGS) estimates approximately  1.7 million pounds of picloram active ingredient used for the
year 1992, with roughly 7 million acres treated (USGS, 1999).  These estimates were derived using State-
level data sets on pesticide use rates available from NCFAP combined with county-level data on
harvested crop acreage  from the  Census of Agriculture (CA) (Thelin and Gianessi, 2000).  Figure 3.12-1
shows  the geographic distribution of estimated average annual picloram use in the United States for 1992
(USGS, 1998a). A breakdown of use by crop is also included.  By far, the greatest amount of picloram is
used in pasture (92%).  A comparison of this use map with the map of the 16 cross-section States (Figure
1.3-1)  shows that States across the range of high of low picloram use are well represented in the cross-
section.
Occurrence Summary and Use Support Document         250                                     March 2002

-------
Figure 3.12-1: Picloram Estimated Annual Agricultural Use
                                            PICLORAM
                                   ESTIMATED ANNUAL AGRICULTURAL USE
                    Average use of
                    Active Ingredient
                  Pounds per square mile
                    of county par year
                    D  No Estimated Use
                    D   < 0.370
                    D  0.370-0.774
                    D  0.775 - 1.382
                    D  1.383-2.643
                    •   >= 2.644
Crops
pastj-a
other hay
wheat and grains
flax
oats
bar lay
Total
Pounds Applied
1,583,583
113,067
21,040
3,336
1,907
256
Percent
National Use
91.90
156
1.22
0.19
0.11
ft 01
Source: USGS, 1998a
3.12.2 Environmental Release

Picloram is listed as a Toxics Release Inventory (TPJ) chemical. Table 3.12-1 illustrates the
environmental releases for picloram from 1995 - 1999. (There are only picloram data for these years.)
Air emissions and surface water discharges are the sole contributors to total on- and off-site releases.
Between 1995 and 1999, the releases of air emissions fluctuated greatly, with a dramatic increase in 1997
and 1999. Surface water discharges were generally very low, with the exception of a huge peak in 1998
(almost 400,000 pounds). No underground injection, releases to land  (such as spills or leaks within the
boundaries of the reporting facility), or off-site releases (including metals or metal compounds
transferred off-site) were reported for picloram. These TPJ data for picloram were reported from only
Michigan and Texas (USEPA, 2000), both of which are  included in the 16-State cross-section (used for
analyses of picloram occurrence in drinking water; see Section 3.12.4). (For a map of the 16-State cross-
section, see Figure 1.3-1.)
Occurrence Summary and Use Support Document
251
March 2002

-------
Table 3.12-1: Environmental Releases (in pounds) for Picloram in the United States, 1995-1999
Year
1999
1998
1997
1996
1995
On-Site Releases
Air Emissions
2,800
460
2,900
522
220
Surface Water
Discharges
10
380,006
0
0
1
Underground
Injection
--
--
--
--
--
Releases
to Land
--
--
--
--
--
Off-Site Releases
--
--
--
--
--
Total On- &
Off-site
Releases
2,810
380,466
2,900
522
221
Source: USEPA, 2000
3.12.3 Ambient Occurrence

Picloram is an analyte for both surface and ground water NAWQA studies, with a method detection limit
(MDL) of 0.05 |ig/L. Table 3.12-2 summarizes the findings of USGS NAWQA sampling for picloram
within the first 20 NAWQA study basins. Picloram concentrations at all of the surface water sites were
below the detection limit in most sites, although it was detected in at least one sample. Also, picloram
concentrations at all of the ground water sites were  below the detection limit in most sites, although it
was detected in at least one sample.
Table 3.12-2: Picloram Detections and Concentrations in Surface Water and Ground Water
                               Detection frequency
                                 (% of samples)
                    Concentration percentiles
                       (all samples; |J.g/L)

surface water
agricultural
urban
integrator
all sites
ground water
agricultural
urban
major aquifers
all sites
all samples > 0.01 ug/L

ND
ND
ND
0.15%

0.11%
ND
ND
0.20%
> 0.05 ug/L

ND
ND
ND
0.12%

ND
ND
ND
0.20%
10th


-------
3.12.3.2 Additional Ambient Occurrence Data

A summary document entitled "Occurrence and Exposure Assessment of Picloram in Public Drinking
Water Supplies" (Wade Miller, 1989), was previously prepared for past USEPA assessments of picloram.
In that review, several national studies were included on the occurrence of picloram in both ground water
and surface water non-drinking water sources.  The following information is taken directly from
"Occurrence and Exposure Assessment of Picloram in Public Drinking Water Supplies" (Wade Miller,
1989).

3.12.3.2.1 Groundwater Sources

Picloram was detected in five out of 77 groundwater samples collected at 49 locations throughout the
United States. Picloram was detected in seven States. The 85th percentile of all nonzero samples was
1.0 |ig/l in groundwater sources. The maximum concentration found was 1.0 |ig/L (USEPA,  1987, as
cited in Wade Miller, 1989). No other sampling information was provided.

The Pesticides in Groundwater Data Base, developed by EPA's Offices of Pesticide Programs (OPP),
contains information derived from monitoring studies conducted by pesticide registrants, universities,
and government agencies (Williams et al., 1988, as cited in Wade Miller, 1989). The data are presented
in several categories, including all information collected to date (excluding data known to be of poor
quality and data that is from point source contamination); data derived from scientifically confirmed
agricultural use; and confirmed information attributed to known point sources and documented misuse.
Three States have confirmed detections of picloram in groundwater attributed to normal agricultural use
with a maximum concentration of 49 |ig/L and a median concentration of 1.4 |ig/L.

While extreme care was taken to confirm detections of pesticides in ground water, the reader is warned
that the studies investigated have not been limited to drinking water supplies. Some studies included
samples from many types of wells including observation and irrigation wells. In addition, many studies
have focused on shallow ground water which may not be representative of drinking water sources in the
area.  As such, the results cannot be used to estimates human exposure to pesticides in drinking water,
but do provide an additional assessment of possible picloram occurrence.

3.12.3.2.2 Surface Water Sources

Picloram was detected in 359 out of 653 surface water samples collected at 124 locations throughout the
United States. Picloram was detected in seven States. The 85th percentile of all nonzero samples was
0.13 |ig/L in surface water. The maximum concentration found was 4.6 |ig/L (USEPA, 1987, as cited in
Wade Miller, 1989). No other sampling information was provided.

3.12.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of picloram occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest:  1 mg/L; 0.5 mg/L; and 0.05 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
picloram. These data represent almost 46,000 analytical results from approximately 13,000 PWSs during
the period from 1983 to 1998 (with most analytical results from 1992 to 1997).  The number of sample
results and PWSs vary by State, although the State data sets have been reviewed and checked to ensure

Occurrence Summary and Use Support Document         253                                      March 2002

-------
adequacy of coverage and completeness. The overall modal detection limit for picloram in the 16 cross-
section States is equal to 0.0001 mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)

3.12.4.1 Stage 1 Analysis Occurrence Findings

Table 3.12-3 illustrates the very low occurrence of picloram in drinking water forthe public water
systems in the  16-State cross-section.  No ground water or surface water PWSs (therefore, no population
served by systems, as seen in Table 3.12-4) had any analytical results exceeding 1 mg/L, 0.5 mg/L, or
0.005 mg/L.
Table 3.12-3:  Stage 1 Picloram Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
1
0.5
0.05
Percent of Systems
Exceeding Threshold
0.000%
0.000%
0.000%
Number of Systems
Exceeding Threshold
0
0
0

Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0
0
0

Combined Ground &
Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0
0
0
Table 3.12-4:  Stage 1 Picloram Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
1
0.5
0.05
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.000%
0.000%
Total Population Served
by Systems Exceeding
Threshold
0
0
0

Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0
0
0

Occurrence Summary and Use Support Document
254
March 2002

-------
Source Water Type
Combined Ground &
Surface Water
Threshold
(mg/L)
1
0.5
0.05
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.000%
0.000%
Total Population Servet
by Systems Exceeding
Threshold
0
0
0
3.12.4.2  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.12-5 and
3.12-6. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented. (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations (USEPA, 2002)).

No ground water or surface water PWSs (therefore, no population served by systems) had an estimated
mean concentration of picloram exceeding 1 mg/L, 0.5 mg/L or 0.05 mg/L.
Table 3.12-5: Stage 2 Estimated Picloram Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
1
0.5
0.05
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
Occurrence Summary and Use Support Document
255
March 2002

-------
Table 3.12-6:  Stage 2 Estimated Picloram Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
1
0.5
0.05
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0

Combined Ground
& Surface Water
1
0.5
0.05
0.000%
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0
0-0
0-0
0-0
3.12.4.3 Estimated National Occurrence

As illustrated in Table 3.12-7, the Stage 2 analysis estimates zero systems serving zero people nationally
have estimated mean concentration values of picloram greater than 1 mg/L, 0.5 mg/L, or 0.05 mg/L.  (See
Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to national values.)
Table 3.12-7:  Estimated National Picloram Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
1
0.5
0.05
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
1
0.5
0.05
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
Occurrence Summary and Use Support Document
256
March 2002

-------
Source Water Type
Threshold
(mg/L)
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
1
0.5
0.05
0
0
0
0-0
0-0
0-0
0
0
0
0-0
0-0
0-0
3.12.5 Additional Drinking Water Occurrence Data

For a previous USEPA assessment of picloram occurrence in drinking water, a literature search was
conducted and knowledgeable sources within the Office of Water were contacted. No national studies
were found on the occurrence of picloram in drinking water from ground water or surface water sources,
nor were any regional or State studies found that addressed the occurrence of picloram in drinking water
from surface water sources (Wade Miller, 1989). However, some regional studies of ground water were
found. Note that none of the studies presented in the following section provide the quantitative analytical
results or comprehensive coverage that would enable direct comparison to the occurrence findings
estimated with the cross-section occurrence data presented in Section 3.12.4. These additional studies,
however, do enable a broader assessment of the Stage 2 occurrence estimates presented for this Six-Year
Review.  All the following information in Section 3.12.5 is taken directly from "Occurrence and
Exposure Assessment of Picloram in Public Drinking Water Supplies" (Wade Miller, 1989).

EPA has developed contamination incidence surveys, which were subsequently distributed to  States by
their Federal/State Toxicological and Regulatory Alliance Committee (FSTRAC). As of April 1988, the
contamination incidence surveys of 11 States have been returned and summarized by EPA, Region VII
(Schlachter, 1988, as cited in Wade Miller,  1989). These States include: California, Iowa,
Massachusetts, Michigan, Minnesota, Nebraska, Pennsylvania, Rhode Island,  South Dakota, Texas, and
Wisconsin. The summary provided limited information on the occurrence of picloram in public ground
water supplies. Picloram was only detected in South Dakota in two samples from 17 sites, though at
values of 520 and 8,300 |ig/L (mean = 4,410 M-g/L). No detection limit or other sampling information
was reported.

From July 1985 to June 1987, the Minnesota Departments of Health (MDH) and Agriculture (MDA)
conducted cooperative surveys of observation, irrigation, and drinking water wells for selected pesticides
(MDH and MDA, 1988, as cited in Wade Miller, 1989). These studies were designed  to provide
information on the  extent of agricultural pesticide contamination in Minnesota's ground water and
drinking water. In the first survey (MDA), picloram was not detected in any of the 100 samples taken.
The detection limit for the MDA survey  was 1.8 |ig/L. In the second survey (MDH), picloram was found
in five samples from three wells, out of 400 wells sampled, at a median concentration of 0.16 |ig/L (range
= 0.08-0.63 M-g/L).  The detection limit for the MDH survey was 0.04 |ig/L.  The summary report did not
distinguish between observation, irrigation, and drinking water wells.
Occurrence Summary and Use Support Document
257
March 2002

-------
3.12.6 Conclusion

Picloram is a systemic herbicide used to control deeply rooted herbaceous weeds and woody plants in
rights-of-way, forestry, rangelands, pastures, and small grain crops.  Picloram products are all restricted
use pesticides.  Approximately 1.7 million pounds of picloram were used in 1992 on pastureland and for
other agricultural applications.  The greatest concentrations of use are in Texas, New Mexico, and the
upper Midwest. Industrial releases of picloram have been reported to TRI from two States since 1995.
Picloram was also an analyte for the NAWQA occurrence studies. In the NAWQA study, picloram was
detected in ground and surface water; however, none of the median values exceeded the detection limit.
The Stage 2 analysis, based on the 16-State cross-section, estimated that zero percent of combined
ground water and surface water systems serving zero percent of the population exceeded the MCL of 0.5
mg/L.  Based on this estimate, zero PWSs nationally are estimated to have picloram levels greater than
the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, according to information from USGS,  14 States use picloram agriculturally,
including 5 of the cross-section States.  Picloram  use is heaviest in Texas, New Mexico, and Kansas, two
of which are cross-section States. Nationally, TRI releases  have been reported for picloram from 2
States, both of which are cross-section States. The cross-section should adequately represent the
occurrence of picloram on a national  scale based upon the use, production, and release patterns of the 16-
State cross-section in relation to the patterns observed for all 50 States.

3.12.7 References

Gilliom, R.J., D.K. Mueller, and L.H. Nowell.  1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological  Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589/, last updated October 9, 1998.

EXTOXNET.  2001. Pesticide Information Profile: Picloram. Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://ace.ace.orst.edu/info/extoxnet/pips/picloram.htm, revised March, 2001.

Kirk-Othmer.  1980.  Herbicides. In: Kirk-Othmer Encyclopedia of Chemical Technology.  Vol. 12, 3rd
       ed. New York, NY: John Wiley and Sons. pp. 297-351.

Minnesota Department of Health (MDH) and Minnesota Department of Agriculture (MDA). 1988.
       Pesticides and Groundwater: Surveys of Selected Minnesota Wells.  Prepared by Tomas G.
       Klaseus (MDH), Greg C. Buzicky (MDA), and Edward C. Schneider (MDH) for Legislative
       Commission on Minnesota Resources. February 1988.

Schlachter, J. 1988. Memorandum on Program Improvement Subcommittee, FSTRAC, Preparation for
       April Meeting. Sent by Jacquelyn D.  Schlachter, Chairman - Program Improvement
       Subcommittee, U.S. Environmental Protection Agency, Region VII, Kansas City, KS.  Sent to
       Dennis Alt, IA; Pam Bonrud, SD; Bill Lee, NE; Kathy Miller, MT; Jennifer Orme, HQ; Sheila
       Sullivan, EPA Region V; and Dan Wilson, WI.
Occurrence Summary and Use Support Document         258                                     March 2002

-------
Schutte, W.D.  1982. Preliminary Quantitative Usage Analysis of Picloram.  Economic Analysis
       Branch, Office of Pesticide Programs, U.S. Environmental Protection Agency, Arlington, VA.
       June, 1982.

Thelin, G.P., and L.P. Gianessi.  2000. Method for Estimating Pesticide Use for County Areas of the
       Conterminous United States. U.S. Geological Survey Open-File Report 00-250.
       62 pp.  Available on the Internet at: http://water.wr.usgs.gov/pnsp/rep/ofr00250/ofr00250.pdf

USEPA.  1981.  EPA Index to Pesticide Chemicals.  Picloram, Potassium, Salt. Washington, DC: Office
       of Pesticide Programs, USEPA.

USEPA.  1987.  Picloram Health Advisory. Washington, DC: Office of Drinking Water, USEPA.
       August, 1987.

USEPA.  1995.  Registration Eligibility Decision (RED): Picloram. EPA Report 738-R-95-019.
       Washington, DC: Office of Prevention, Pesticides, and Toxic Substances, USEPA. 283 pp.
       Available on the Internet at: http://www.epa.gov/oppsrrdl/REDs/, last updated June 21, 2001.

USEPA.  2000.  TRIExplorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USGS.  1998a. Annual Use Maps. Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated March 20, 1998.

USEPA.  2001.  National Primary Drinking Water Regulations - Consumer Factsheet on: PICLORAM.
       Available on the Internet at: http://www.epa.gov/safewater/dwh/c-soc/picloram.html, last
       updated April 12,2001.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

USGS.  1998b. Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA). PROVISIONAL DATA - SUBJECT TO
       REVISION. Available on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/ last updated
       October 9, 1998.

USGS.  1999. PESTICIDES ANALYZED IN NAWQA SAMPLES: Use,  Chemical Analyses, and
       Water-Quality Criteria (PROVISIONAL DATA ~ SUBJECT TO REVISION). Available on the
       Internet at: http://water.wr.usgs.gov/pnsp/anstrat/, last updated August 20, 1999.

Wade Miller Associates, Inc. 1989. Occurrence and Exposure Assessment of Picloram in Public
       Drinking Water Supplies.  Prepared for and submitted to EPA on April 24, 1989.

Williams, W.M., P.W. Holden, D.W. Parsons, and M.N. Lorber. 1988. Pesticides in groundwater Data
       Base. Washington, DC: Office of Pesticide Programs, U.S. Environmental Protection Agency.
Occurrence Summary and Use Support Document         259                                    March 2002

-------
3.13   Simazine
Table of Contents

3.13.1 Introduction, Use and Production  	 261
3.13.2 Environmental Release  	 262
3.13.3 Ambient Occurrence 	 263
3.13.4 Drinking Water Occurrence Based on the 16-State Cross-Section	 266
3.13.5 Additional Drinking Water Occurrence Data  	 271
3.13.6 Conclusion	 274
3.13.7 References 	 274
Tables and Figures

Figure 3.13-1:  Estimated Annual Agricultural Use for Simazine (1992)	 262

Table 3.13-1: Environmental Releases (in pounds) for Simazine in the United States, 1995-1999 . .  . 263

Table 3.13-2: Simazine Detections and Concentrations in Surface Water and Ground Water  	 263

Table 3.13-3: Stage 1 Simazine Occurrence Based on 16-State Cross-Section - Systems	 267

Table 3.13-4: Stage 1 Simazine Occurrence Based on 16-State Cross-Section - Population	 267

Table 3.13-5: Stage 2 Estimated Simazine Occurrence Based on 16-State Cross-Section -
       Systems	 268

Table 3.13-6: Stage 2 Estimated Simazine Occurrence Based on 16-State Cross-Section -
       Population	 269

Table 3.13-7: Estimated National Simazine  Occurrence - Systems and Population Served  	 270

Table 3.13-8: Frequency of Surface Water Contamination by Simazine - National
       Screening Program for Organics in Drinking Water  	 272
Occurrence Summary and Use Support Document         260                                     March 2002

-------
3.13.1  Introduction, Use and Production

Simazine [2-chloro-4,6-bis(ethylamino)-s-triazine], is an organic white solid.  It is available in wettable
powder, water dispersible granule, liquid, and granular formulations. It may also be soil-applied
(EXTOXNET, 1996).  Simazine is used on a variety of deep-rooted crops such as artichokes, asparagus,
berry crops, broad beans, citrus, etc., and on non-crop areas such as farm ponds and fish hatcheries.  Its
major use is on corn where it is often combined with AAtrex.  Other herbicides with which simazine is
combined include: paraquat, on apples, peaches; Roundup or Oust for non-crop use; Surflan on
Christmas trees; Dual on corn and ornamentals (USEPA, 2001).

Simazine is a selective triazine herbicide (EXTOXNET, 1996).  It is used as a pre-emergence herbicide
for control of broad-leaved and grassy weeds on a variety of deep-rooted crops such as artichokes,
asparagus, berry crops, broad beans, citrus, etc.,  and on non-crop areas.  Its major use is on corn
(USEPA, 2001).  At higher rates, simazine is used for nonselective weed control in industrial areas.
Before 1992, it was used to control submerged weeds and algae in large aquariums, farm ponds, fish
hatcheries, swimming pools, ornamental ponds, and cooling towers (EXTOXNET,  1996). In 1975, the
consumption pattern of simazine was approximately as follows: 49%, corn; 8%, citrus; 6%, deciduous
fruits; 5%, field crops; 3%, vegetables; 17%, industrial commercial uses; and  13%, aquatic uses (HSDB,
2001).

Two manufacturers of simazine are Drexel Chemical Company in Memphis, TN and Novartis
Corporation in Summit, NJ. Although production data is seemingly scarce, there is information on
consumption of simazine in the U.S. Agricultural use of simazine by year is approximately as follows: in
1989, 3.96 million pounds active ingredient (ai); 1982, 3.98 million pounds ai; 1976, 3.25 million pounds
ai; 1971, 1.74 million pounds ai; and 1966, 193,000 pounds ai (HSDB, 2001).

Recent national estimates of agricultural use for  simazine are available.  Figure 3.13-1 illustrates the
USGS (1998a) derived geographic distribution of estimated average annual simazine use in the United
States for 1992. USGS (1998a) estimates approximately 4.81 million pounds  of simazine active
ingredient were used in 1992. A breakdown of use by crop is also included. Corn accounts for the
majority of usage (2.35 million pounds simazine active ingredient), while intermediate use can be found
on several other crops as well (e.g., citrus, grapes, alfalfa, apples, peaches, almonds, etc.) (USGS, 1998a).
The two largest concentrations of simazine use are seen in the corn growing area of the Midwest and the
citrus areas in California and Florida, with some concentration also in the Mid-Atlantic States and
Washington State (Figure 3.13-1). Note that non-agricultural uses are not reflected here, and any sharp
spatial differences in use within a county may not be well represented (USGS, 1998b). A comparison of
this use map with the map of the 16 cross-section States (Figure 1.3-1) shows  that States across the range
of high of low simazine use are well represented in the cross-section.
Occurrence Summary and Use Support Document         261                                     March 2002

-------
Figure 3.13-1: Estimated Annual Agricultural Use for Simazine (1992)
                                            SIMAZINE
                                   ESTIMATED ANNUAL AGRICULTURAL USE
                     Average use of
                    Active Ingredient
                   Pounds per square mile
                    of county per year

                    D  No Estimated USB
                    D   < 0.037
                    D 0.037-0.179
                    D 0.180-0.886
                    D 0.887-3.575
                    •   >= 3.576
Crops
com
all citrus
yapas
alfalfa toy
apples
peaches
almonds
walnuts
pecans
asparagus
Total
Pounds Applied
2,349,911
1,1-10,321
481,199
333,372
114,112
81,784
59,940
39,719
32,490
31,155
Percern
National Use
43.85
23. 71
10.00
a 93
Z37
1.9
1.25
0.83
0.68
0.65
Source: USGS, 1998a
3.13.2 Environmental Release

Simazine is listed as a Toxics Release Inventory (TRI) chemical. Table 3.13-1 illustrates the
environmental releases for simazine from 1995 - 1999.  (There are only simazine data for these years.)
From 1995 to 1999, the releases of air emissions has generally decreased. Surface water discharges
generally hovered around 300 pounds.  No underground injection or releases to land (with the exception
of 5 pounds reported in 1995) were reported for simazine.  The amount of off-site releases (including
metals or metal compounds transferred off-site) has steadily decreased since 1995, with a dramatic
decrease in 1998. Decreases in air emissions and off-site releases have predominantly contributed to the
decreasing amount of total on- and off-site simazine releases.  These TRI data for simazine were reported
from 8 States.  Four of these 8 States reported TRI data every  year (USEPA, 2000). Two of the 8 States
(Alabama and Nebraska) are included in the 16-State cross-section (used for analyses of simazine
occurrence in drinking water; see Section 3.13.4). (For a map of the 16-State cross-section, see Figure
1.3-1.)
Occurrence Summary and Use Support Document
262
March 2002

-------
Table 3.13-1: Environmental Releases (in pounds) for Simazine in the United States, 1995-1999
Year
1999
1998
1997
1996
1995
On-Site Releases
Air Emissions
3,928
3,321
2,939
4,591
4,990
Surface Water
Discharges
385
332
348
93
232
Underground
Injection
—
—
0
0
-
Releases
to Land
0
0
0
0
5
Off-Site Releases
2,385
4,497
48,629
54,457
26,231
Total On- &
Off-site
Releases
6,698
8,150
51,916
59,141
31,458
 Source: USEPA, 2000
3.13.3 Ambient Occurrence

Simazine is an analyte for both surface and ground water NAWQA studies, with a method detection limit
(MDL) of 0.005 |ig/L. Additional information on analytical methods used in the NAWQA study units,
including method detection limits, are described by Gilliom and others (1998).

Table 3.13-2 summarizes the findings of USGS NAWQA sampling for simazine within the first 20
NAWQA study basins (USGS, 1998c).  Simazine concentrations at ground and surface sites exceed the
detection limit in all sites. High percentages of both ground and surface water samples exceed detection
frequencies of 0.01 and 0.05 |ig/L, although surface water sample percentages are notably higher than
ground water sample percentages. Within surface water sites, urban sites have consistently higher
percentage exceeding the detection frequencies then agricultural or integrator sites. Within ground water
sites, agricultural sites have consistently higher percentage exceeding the detection frequencies then
urban or integrator sites.  The 95th percentile value and maximum  concentration also exceed the detection
limit in all ground and surface water sites. The median concentration values exceed the detection limit
for all  surface water sites, while none of the median concentration values exceed the detection limit for
any ground water sites.
Table 3.13-2: Simazine Detections and Concentrations in Surface Water and Ground Water
                               Detection frequency
                                 (% of samples)
                    Concentration percentiles
                      (all samples; |J.g/L)

surface -water
agricultural
urban
integrator
all sites
all samples

61.74%
87.77%
82.93%
65.61%
>0.01 us/L

44.76%
80.73%
64.63%
53.86%
> 0.05 us/L

16.98%
40.67%
27.64%
25.35%
10th


-------
                              Detection frequency                     Concentration percentiles
                                (% of samples)                        (all samples; |J.g/L)
ground -water
agricultural
urban
major aquifers
all sites

22.38%
15.28%
6.32%
15.91%

13.30%
11.30%
3.00%
9.52%

4.54%
3.65%
0.75%
3.27%


-------
they may not be representative of drinking water sources in the area. For these reasons, use of these data
for the estimation of human exposure to pesticides in drinking water is not advised.

3.13.3.1.2 Groundwater Sources - Regional Studies

The U.S. Environmental Protection Agency has developed contamination incidence surveys, that have
been subsequently distributed to States by their Federal/State Toxicological and Regulatory Alliance
Committee (FSTRAC). As of April 1988, the contamination incidence surveys of 11 States had been
returned and summarized by EPA,  Region VII (Schlachter,  1988, as cited in Wade Miller, 1989). These
States include: California, Iowa, Massachusetts, Michigan, Minnesota, Nebraska, Pennsylvania, Rhode
Island, South Dakota, Texas and Wisconsin. The summary provided limited information on the
occurrence of simazine in ground water. Simazine was only detected in Wisconsin in four samples from
28 sites, at a range of 1.0-15.0 |ig/L. No other sampling information was reported. It is unclear whether
these detections are representative  of ambient water or drinking water.

In a recent California study on the movement of pesticides and herbicides to groundwater (Cohen et al,
1984, as cited in Wade Miller, 1989), collected samples to determine the levels of simazine in
groundwater. In analyses of samples collected from 166 wells, 6 wells were found to be contaminated
with simazine at levels between 0.5 |ig/L and 3.5 |ig/L. The limits of detection were not reported.

3.13.3.1.3 Surface Water Sources - National Studies

Simazine has been found in 877 of 5,067 surface water samples analyzed throughout the United States
(USEPA,  1987, as cited in Wade Miller, 1989).  Samples were collected at 472 surface water locations,
and simazine was detected in 22 States.  The 85th percentile of all nonzero samples was 2.18 |ig/L. The
maximum concentration found in surface water was 1,300 |ig/L.

Carey and Kutz (1983, as  cited in Wade Miller, 1989) summarized a study on water samples collected
and analyzed for pesticides and herbicides as part of the National Surface Water Monitoring Program.
During the collection period (1976 to 1980), 0.4 percent of the samples tested contained levels of
simazine in excess of the detection limit, with a maximum reported value of 1.13 |ig/L.  The detection
limits were not reported.

3.13.3.1.4 Surface Water Sources - Regional Studies

In a partial overview of activities performed by the International Joint Commission's Pollution from
Land Use Activities Reference Group (PLUARG), Baker (1987, as cited in Wade Miller, 1989) reported
on rural nonpoint pollution in three northwest Ohio rivers. From 1983  to 1985, mean simazine values
were not detected, 0.12 and 0.18 |ig/L, respectively. The number of samples taken for these three years
were 145, 183, and 188, respectively.  Peak observed concentrations for the three years were 0.01, 1.25,
and 0.98 |ig/L, all of which were from the Sandusky River.  No other sampling information was reported.

In a study on sediment, nutrient, and pesticide transport in lower Great Lakes tributaries, Baker (1986, as
cited in Wade Miller, 1989) included a summary of an investigation of nonpoint source pollution of
tributaries for Lake Erie and Lake Ontario for the years 1982-1985.  Biweekly samples were collected
each year between April 15 and August 15, when pesticide use was the greatest.  The range of time
weighted mean values reported for simazine was 0.0 - 0.84 |ig/L. The number of positive samples was
not reported; however, a total of 1,785 samples were taken from eight locations.  The highest observed
concentrations for simazine were 10.77 |ig/L in the Cuyahoga River in 1982 and 6.93 |ig/L in the
Maumee River in 1982. The  detection limit was reported to be 0.25 |ig/L.

Occurrence Summary and Use Support Document         265                                    March 2002

-------
Baker (1983b, as cited in Wade Miller, 1989) assessed the results of a pesticide monitoring program
conducted in Ohio. A total of 258 samples were collected from six streams of different watershed size
during the April and August 1982 sampling period to compare peak pesticide concentrations.  The range
of peak concentrations was 2.52 to 10.7 |ig/L of simazine, with an average peak value of 4.65 |ig/L.

In a subsequent related report, Baker (1983a, as cited in Wade Miller, 1989) reported the average
herbicide concentrations in samples collected from two rivers in Ohio from May 28 to July 27, 1983.  Of
the 23 surface water samples collected from each river, the average reported concentrations of simazine
were 0.26 and 0.44 |ig/L, respectively. The detection limits were not reported.

Baker et al. (1981, as cited in Wade Miller, 1989) assessed the results of a study to determine the
concentrations of various pesticides in Northwestern Ohio rivers.  Samples were collected from June
1981 to September 1981 during periods of peak pesticide usage and maximum pesticide export due to
large amounts of runoff. The authors reported that the  concentrations observed at the stream stations
were generally higher than those reported in other areas.  Low levels of simazine were detected in 100%
of the 135 surface water samples collected from nine streams.  The values ranged from 0.03 to 2.55 |ig/L
for simazine, with an average value of 0.60 |ig/L. The  detection limit for the analysis was not reported.

3.13.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of simazine occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section  States. The  16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest:  0.004 mg/L; 0.002 mg/L; and 0.001 mg/L.

All sixteen cross-section State data sets, with the exception of New Jersey, contained occurrence data for
simazine. These data represent more than 68,000 analytical results from approximately 15,000 PWSs
during the period from 1984 to  1998 (with most analytical results from 1992 to 1997). The number of
sample results and PWSs vary by State, although the State data sets have been reviewed and checked to
ensure adequacy of coverage and completeness. The overall modal detection limit for simazine in the 16
cross-section  States is equal to 0.001 mg/L. (For details regarding the 16-State cross-section, please refer
to Section 1.3.5 of this report.)

3.13.4.1 Stage 1 Analysis Occurrence Findings

Table 3.13-3 illustrates the low occurrence of simazine in drinking water for the public water systems in
the 16-State cross-section relative to three  thresholds:  0.004 mg/L (the current MCL), 0.002 mg/L, and
0.001 mg/L (the modal MRL).  A total of 8 (approximately 0.0550% of) ground water and surface water
PWSs had analytical results exceeding the MCL; 0.144% of systems (21 systems) had results exceeding
0.002 mg/L; and 0.358% of systems (52 systems) had results exceeding 0.001 mg/L.

Zero ground water systems had analytical results greater than the MCL.  Approximately 0.0152% of
ground water systems (2 systems) had results above 0.002 mg/L. The percentage of ground water
systems with at least one result greater than 0.001 mg/L was equal to 0.129% (17 systems).

The percentage of surface water systems with results greater than the MCL was equal to 0.573% (8
systems).  Approximately 19 (1.36% of) surface water  systems had at least one analytical result greater
than 0.002 mg/L.  Thirty-five (2.51% of) surface water systems had results exceeding 0.001 mg/L.
Occurrence Summary and Use Support Document         266                                     March 2002

-------
Table 3.13-3:  Stage 1 Simazine Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.004
0.002
0.001
Percent of Systems
Exceeding Threshold
0.000%
0.0152%
0.129%
Number of Systems
Exceeding Threshold
0
2
17

Surface Water
0.004
0.002
0.001
0.573%
1.36%
2.51%
8
19
35

Combined Ground &
Surface Water
0.004
0.002
0.001
0.0550%
0.144%
0.358%
8
21
52
Reviewing simazine occurrence in the 16 cross-section States by PWS population served (Table 3.13-4)
shows that approximately 0.0471% of the 16-State population (about 46,300 people) was served by
PWSs with at least one analytical result of simazine greater than the MCL (0.004 mg/L).  Approximately
113,000 (0.115% of) people were served by systems with an exceedance of 0.002 mg/L.  Over 1 million
(1.06% of) people were served by systems with at least one analytical result greater than 0.001 mg/L.

The percentage of population served by ground water systems in the 16 States with analytical results
greater than the MCL was equal to 0%. When evaluated relative to 0.002 mg/L and 0.001 mg/L, the
percent of population exposed was equal to 0.0150% (about 6,200 people) and 0.823% (almost 342,000
people), respectively.

The percentage of population served by surface water systems in the 16 States with exceedances of 0.004
mg/L and 0.002 mg/L was equal to 0.0817% (46,300 people) and 0.189% (106,800 people) respectively.
When evaluated relative to 0.001 mg/L, the percent  of population exposed was equal to 1.23% (over
697,000 people).
Table 3.13-4:  Stage 1 Simazine Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.004
0.002
0.001
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.0150%
0.823%
Total Population Served
by Systems
Exceeding Threshold
0
6,200
341,900

Occurrence Summary and Use Support Document
267
March 2002

-------
Source Water Type
Surface Water
Threshold
(mg/L)
0.004
0.002
0.001
Percent of Population
Served by Systems
Exceeding Threshold
0.0817%
0.189%
1.23%
Total Population Served
by Systems
Exceeding Threshold
46,300
106,800
697,100

Combined Ground &
Surface Water
0.004
0.002
0.001
0.0471%
0.115%
1.06%
46,300
113,100
1,039,000
3.13.4.2  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.13-5 and
3.13-6. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations (USEPA, 2002)).

The percentages of ground water and surface water PWSs with estimated mean concentrations of
simazine exceeding 0.004 mg/L, 0.002 mg/L, and 0.001 mg/L are 0.0000413%, 0.000509%, and
0.00992%, respectively.

Only 1 ground water PWS in the 16 States was estimated to have a mean concentration greater than 0.004
mg/L (0.0000305%), 0.002 mg/L (0.00289%) and 0.001 mg/L (0.00382%).  The percentage of surface
water systems with estimated mean concentration values greater than 0.004 mg/L was equal to
0.000143%. One surface water system in the  16 States was estimated to have a mean concentration value
of simazine greaterthan 0.002 mg/L (0.00258%) and 0.001 mg/L (0.0673%).
Table 3.13-5:  Stage 2 Estimated Simazine Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.004
0.002
0.001
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0000305%
0.00289%
0.00382%
Range
0.000% - 0.000%
0.000% -0.00761%
0.000% -0.0152%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
1
1
1
Range
0-0
0-1
0-2

Surface Water
0.004
0.002
0.001
0.000143%
0.00258%
0.0673%
0.000% - 0.000%
0.000% -0.0716%
0.000% -0.215%
0
1
1
0-0
0-1
0-3
Occurrence Summary and Use Support Document
268
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
Range
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.004
0.002
0.001
0.0000413%
0.000509%
0.00992%
0.000% - 0.000%
0.000% - 0.00688%
0.000% - 0.0275%
1
1
1
0-0
0-1
0-4
Reviewing simazine occurrence by PWS population served (Table 3.13-6) shows that approximately
0.000107% of population served by all PWSs in the 16 States (an estimate of approximately 100 people)
were potentially exposed to simazine levels above 0.004 mg/L. When evaluated relative to a threshold of
0.002 mg/L, the percent of population exposed was equal to 0.000897% (approximately 900 people).
The percent of potential population exposed equaled 0.0197% (over 19,000 people in the 16 States) when
evaluated relative to 0.001 mg/L.

The percentage of population served by ground water systems in the 16 States with estimated mean
concentration values greater than 0.004 mg/L was equal to 0.000238% (an estimate of approximately 100
people) and the percentage served by surface water systems in the 16 States was equal to  0.0000106%.
Approximately 700 (0.00178% of) people in the 16 cross-section States were served by ground water
systems with estimated mean concentration values greater than 0.002 mg/L. The number of people in the
16 cross-section States served by surface water systems with estimated mean concentration values greater
than 0.002 mg/L was equal to about 100 (0.000253% of) people.  An estimated 4,100 (0.00991% of)
people served by ground water systems in the in the 16 States were exposed to simazine at concentrations
greater than 0.001 mg/L. Approximately 15,200 (0.0268% of) people were served by surface water
systems with mean concentrations of simazine greater than 0.001 mg/L.
Table 3.13-6:  Stage 2 Estimated Simazine Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.004
0.002
0.001
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000238%
0.00178%
0.00991%
Range
0.000% - 0.000%
0.000% - 00733%
0.000% - 0.0668%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
100
700
4,100
Range
0-0
0 - 3,000
0 - 27,700

Surface Water
0.004
0.002
0.001
0.0000106%
0.000253%
0.0268%
0.000% - 0.000%
0.000% - 0.00382%
0.000% -0.147%
0
100
15,200
0-0
0 - 2,200
0 - 83,300
Occurrence Summary and Use Support Document
269
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.004
0.002
0.001
0.000107%
0.000897%
0.0197%
0.000% - 0.000%
0.000% -0.00310%
0.000% -0.105%
100
900
19,300
0-0
0 - 2,200
0 - 68,000
3.13.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, 1 system serving approximately 200 people nationally was estimated to have a mean
concentration value of simazine greater than 0.004 mg/L. Approximately 1,900 people served by an
estimated 1 system were potentially exposed to simazine concentrations greater than 0.002 mg/L. About
6 systems serving about 41,900 people nationally were estimated to have mean simazine concentrations
greater than 0.001 mg/L. (See Section 1.4 for a description of how Stage  2 16-State estimates are
extrapolated to national values.)

For ground water systems, approximately 1 PWSs serving about 200 people nationally was estimated to
have a mean concentration greater than 0.004 mg/L. One ground water system serving approximately
1,500 people was estimated to have mean concentration value of simazine greater than 0.002 mg/L. An
estimated 2 systems serving about 8,500 people nationally had estimated mean concentration values that
exceeded 0.001 mg/L.

Zero surface water systems had estimated mean concentration values of simazine greater than 0.004
mg/L.  Only 1 surface water systems serving about 300 people was estimated to have a mean
concentration of simazine above 0.002 mg/L. Approximately 4 surface water systems serving 34,200
people had estimated mean concentrations greater than 0.001 mg/L.
Table 3.13-7:  Estimated National Simazine Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.004
0.002
0.001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
1
1
2
Range
0-0
0-5
0-9
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
200
1,500
8,500
Range
0-0
0 - 6,300
0 - 57,200

Surface Water
0.004
0.002
0.001
0
1
4
0-0
0-4
0-12
0
300
34,200
0-0
0 - 4,900
0 - 187,300

Occurrence Summary and Use Support Document
270
March 2002

-------
Source Water Type
Combined Ground
& Surface Water
Threshold
(mg/L)
0.004
0.002
0.001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
1
1
6
Range
0-0
0-4
0-18
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
200
1,900
41,900
Range
0-0
0 - 6,600
0 - 223,900
3.13.5 Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of fluoride in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled Occurrence
and Exposure Assessment of Simazine in Public Drinking Water Supplies" (Wade Miller, 1989), is
presented in the following section. This variety of studies and information are presented regarding levels
of simazine in drinking water, with the scope of the reviewed studies ranging from national to regional.
Note that none of the studies presented in the following section provide the quantitative analytical results
or comprehensive coverage that would enable direct comparison to the occurrence findings estimated
with the cross-section occurrence data presented in Section 3.13.4.  All the following information in
Section 3.13.5 is taken directly from "Occurrence and Exposure Assessment of Simazine in Public
Drinking Water Supplies" (Wade Miller, 1989).

3.13.5.1  Groundwater Sources - National Studies

The National  Screening Program for Organics in Drinking Water (NSP) (Boland, 1981, as cited in Wade
Miller, 1989)  was conducted by Stanford Research Institute (SRI) from June  1977 to March 1981.
Finished drinking water samples, collected from 12 groundwater systems throughout the United States,
were analyzed for simazine. The  12 systems analyzed included 1 small system (serving 501-3,300
individuals), 1 medium system (serving 3,301-10,000 individuals),  1 large system (serving 10,001-
100,000 individuals), and 9 very large systems (serving greater than 100,000  individuals). Only 1 out of
the 12 systems sampled, a very large system, had a concentration of simazine in excess of the
quantification limit of 0.1 |ig/L. The value of the positive sample was 1.0 |ig/L.

3.13.5.2  Groundwater Sources - Regional Studies

The State of California conducted a monitoring program for organic chemical contamination  of wells
used by large  public water systems (California Department of Health Services,  1986, as cited in Wade
Miller, 1989). A total of 819 water systems were required to be monitored. Of the 5,650 wells that are
used as part of these systems, 2,947 (over 50%) were sampled. Simazine was found in only 26 wells, at a
mean  concentration of 0.8 |ig/L.  The highest value found was 2.0 |ig/L. The detection limit and other
sampling information were not reported.

In addition, the California Department of Health Services provided unpublished data on results of
sampling from small water systems  (i.e., systems with fewer than 200 service connections) (California
Department of Health Services, Unpublished, as cited in Wade Miller, 1989). A total of 304  systems and
608 samples were collected from 21 counties.  Simazine was only found in two samples from one well in
Merced County, at levels of 0.4 and 0.8 |ig/L. The detection limit was not reported.
Occurrence Summary and Use Support Document
271
March 2002

-------
From July 1985 to June 1987, the Minnesota Departments of Health (MDH) and Agriculture (MDA)
conducted cooperative surveys of observation, irrigation, and drinking water wells for selected pesticides
(MDH and MDA, 1988, as cited in Wade Miller, 1989). These studies were designed to provide
information on the extent of agricultural pesticide contamination in Minnesota's ground water and
drinking water.  In the first survey (MDA), simazine was detected in four samples from one well, out of
100 wells sampled, at a median concentration of 1.4 |ig/L (range = 0.49 - 2.58 |ig/L). The detection limit
was 0.08 |ig/L.  In the  second survey (MDH), simazine was not detected in any of the 400 samples taken.
The detection limit for the MDH survey was 0.3 |ig/L.  The summary report did not distinguish between
observation, irrigation, and drinking water wells.

3.13.5.3 Surface Water Sources - National Studies

The National Screening Program for Organics in Drinking Water (NSP) (Boland, 1981, as cited in Wade
Miller, 1989) also presented data on levels of simazine in drinking water obtained from surface water
systems. Finished drinking water samples, collected from 104 surface water systems throughout the
United States, were analyzed  for the presence of simazine. Of the 104 systems sampled, 13 systems (or
12%) contained simazine in excess of the quantification limit of 0.1 |ig/L (Table 3.13-8). The range of
positive values of simazine was 0.1 to 4.4 |ig/L.
Table 3.13-8:  Frequency of Surface Water Contamination by Simazine - National Screening
Program for Organics in Drinking Water
System Size
(Population
Served)
Unspecified
25-500
501-3,300
3,301-10,000
10,001-100,000
> 100,000
Total
Number of
Systems in
U.S.
-
3,937
3,817
1,679
1,552
217
11,202
Number of
Systems Samples
3
0
0
1
25
75
104
Number of
Positive Systems1
0
0
0
0
5
8
13
Percent of
Positive Systems
0
-
-
0
20
11
12
Range of Values
for Positive
Systems
-
-
-
-
0.0001 - 0.0044
0.0001 - 0.0023
0.0001 - 0.0044
1. Quantification limit = 0.0001 mg/L.
Source: Wade Miller 1989
3.13.5.4  Surface Water Sources - Regional Studies

Baker (1985, as cited in Wade Miller, 1989) reported on sampling conducted between 1980 and 1983
proximal to the Maumee River, Sundusky River, and Honey Creek. Finished tap water samples were
analyzed for the presence of simazine in Tiffin, Fremont, and Bowling Green, Ohio. The range of peak
concentrations reported was 0.13 - 1.9 |ig/L.  The highest value was from a sample in Tiffin, Ohio. No
other sampling information was reported.

Baker et al. (1981, as cited in Wade Miller, 1989) analyzed finished drinking water samples collected
from a water supply in Ohio between June 1980 and September 1981.  Of five finished drinking water
Occurrence Summary and Use Support Document
272
March 2002

-------
samples analyzed, all were found to contain simazine in the range of 0.026 to 0.883 |ig/L.  The detection
limit was not reported.

Baker (1983a, as cited in Wade Miller, 1989) analyzed samples collected from three water supplies in
Ohio in 1983 for simazine. The supplies obtained their raw water from two rivers that drained
agricultural areas. Average concentrations of simazine in the samples collected at each of the supplies
between May 28 and July 27, 1983, were 0.30 |ig/L (18 samples), 0.077 |ig/L (15 samples), and 0.19
l-ig/L (16 samples). Peak concentrations of simazine observed in 1983 were 0.63, 0.13, and 0.35 |ig/L,
respectively. The detection limit and the number of positive samples were not reported.

Pellizzari (1978, as cited in Wade Miller, 1989) summarized the results of the USEPA Region VI New
Orleans Water Supply Study. For the three samples analyzed for residues of simazine, a sample mean of
less than 0.1 |ig/L was determined.  The detection limit was not reported.

3.13.5.5  Groundwater and Surface Water Sources - STORET

The USEPA computerized water quality data base known as STORET was devised to assist Federal and
State institutions in meeting objectives of Public Law 92-500 to maintain and enhance the physical,
chemical, and biological quality of the nation's ambient waterways by providing for the collection and
dissemination of basic water quality data (Staples et al., 1985, as cited in Wade Miller, 1989). Data are
collected by States, EPA regional offices, and other government agencies and are maintained in the
STORET system. STORET now contains approximately 80 million pieces of data, including  data for
drinking water from ground water and surface water sources.

Before presenting a summary of the drinking water data in STORET, it is important to note that there are
significant limitations in using the data base to estimate representative concentrations  of a contaminant
such as simazine. Data entered into STORET are gathered from an array of studies conducted for various
purposes. Analyses are conducted in a number of different laboratories employing different
methodologies with a range of detection limits.  In many cases, detection limits are not reported, making
the reliability of the data highly questionable. Where detection limits have been reported, STORET
assigns the detection limit value to those observations reported as  not detected. This could lead to errors
in interpretation and overestimation of concentrations in cases in which most values are nondetectable.
Additionally, a few high values can inflate mean values and result in large standard deviations relative to
the means (Staples et al., 1985, as cited in Wade Miller, 1989).  Very high values may not be correct, as
they may reflect sample contamination or analytical error and can significantly distort assessment of
average concentrations. Staples et al.  (1985, as cited in Wade Miller, 1989) also notes that the use of
data collected prior to the 1980s is not recommended, since such data were obtained using less sensitive
laboratory techniques and quality assurance procedures were not yet mandated for the data entered into
the system.

With these limitations in mind, a summary of data for simazine is  presented for drinking water from
groundwater sources (USEPA, 1988, as cited in Wade Miller, 1989). According to STORET, there were
68 positive observations for simazine in groundwater from June 1984 to June 1987, with an overall mean
value of 0.75 |ig/L, and a range of 0.06 to 2.0 |ig/L. The standard deviation for these observations was
0.42 |ig/L. There were 112 samples reported as undetected, and assigned a detection limit value giving a
calculated mean value of 1.07 |ig/L, a range of 0.5 to 1.0 |ig/L, and a standard deviation of 0.32 |ig/L.
Forty-seven samples were reported as having the presence of simazine verified, but not quantified,
although a value of 0.5 |ig/L was given as a mean. Including these samples, the undetected samples, and
the detections known to be below the reported value, there were 1,677 observations for simazine in
groundwater from February 1984 to November 1987, with an overall mean value of 0.58 and a range of

Occurrence Summary and Use Support Document         273                                     March 2002

-------
0.03 to 10.0 |ig/L. The standard deviation for all observations was 0.94 |ig/L. Detection limits and other
sampling information were not given.

The USEPA's STORET data base similarly contains data on the occurrence of simazine in drinking
water from surface water sources (USEPA, 1988 , as cited in Wade Miller, 1989). There were three
positive observations for simazine in surface water during August 1986, with an overall mean value of
0.027 |ig/L and a range of 0.02 to 0.04 |ig/L. The standard deviation for these observations was 0.012
l-ig/L. One sample was reported as undetected, with a value of 0.001 |ig/L. Including the undetected
sample and the detections known to be below the reported value, there were seven observations for
simazine in surface water from August 1986 to July 1987, with an overall mean value of 0.037 |ig/L and
a range of 0.001 to 0.06 |ig/L. The standard deviation for all observations was 0.024 |ig/L. Detection
limits and other sampling information were not given.

3.13.6 Conclusion

Simazine is atriazine herbicide used to control deeply rooted herbaceous weeds.  Its major use is on
corn.  Approximately 4.8 million pounds of simazine were used in  1992 on corn and several other crops.
The greatest concentrations of simazine use are in the Midwest, Mid-Atlantic, California, and Florida.
Industrial releases of simazine have been reported to TRI from eight States since  1995.  Simazine was
also an analyte for the NAWQA occurrence studies. In the NAWQA study, simazine was detected in
ground and surface water.  The median concentration for simazine  in surface water was 0.012 |ig/L, and
for ground water, none of the median values exceeded the detection limit. The Stage 2 analysis, based on
the 16-State cross-section, estimated that approximately 0.0000413% of combined ground water and
surface water systems serving 0.000107% of the population exceeded the MCL of 0.004 mg/L.  Based on
this estimate,  zero PWSs nationally are estimated to have simazine levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, according to information from USGS, 44 States use simazine agriculturally,
including 14 of the 16 cross-section States. Most agricultural use is concentrated in the Midwest, Mid-
Atlantic, California, and Florida. The areas of high use are represented by six cross-section States with
extensive use of simazine. TRI releases have been reported for simazine from 8 States nationally,
including 2 of 16 cross-section States. The cross-section should adequately represent the occurrence  of
simazine on a national scale based upon the use, production, and release patterns  of the  16-State cross-
section in relation to the patterns observed for all 50 States.

3.13.7 References

Baker, D.B., K.A. Krieger, and J.V. Setzler. 1981. The Concentrations and Transport of Pesticides in
       Northwestern Ohio Rivers.  Prepared by Water Quality Laboratory, Heidelberg College, Tiffin,
       OH, for Lake Erie Wastewater Management Study, U.S. Army Corps of Engineers, Buffalo, NY.

Baker, D.B. 1983a. Herbicide Contamination in Municipal Water Supplies of Northwestern Ohio. Draft
       Final Report.  Prepared by Water Quality Laboratory, Heidelberg College, Tiffin, OH, for Great
       Lakes National Program Office, U.S. Environmental Protection Agency,  Chicago, IL.

Baker, D.B. 1983b. Studies of Sediment, Nutrient, and Pesticide Loading in Selected Lake Erie and
       Lake  Ontario Tributaries. Prepared by Water Quality Laboratory, Heidelberg College, Tiffin,
       OH, for U.S. Environmental Protection Agency, Region V, Chicago, IL.
Occurrence Summary and Use Support Document         274                                     March 2002

-------
Baker, D.B.  1985.  Regional Water Quality Impacts of Intensive Row-Crop Agriculture:  A Lake Erie
       Basin Case Study. Reprinted from the Journal of Soil and Water Conservation,  v. 40, no. 1, pp.
       125-132.

Baker, D.B.  1986.  Sediment, Nutrient and Pesticide Transport in Selected Lower Great Lakes
       Tributaries.  Prepared for Great Lakes National Program Office, U.S. Environmental Protection
       Agency, Chicago, IL. Grant Nos. R005727 and R005817.

Baker, D.B.  1987.  Overview of Rural Nonpoint Pollution in the Lake Erie Basin. In: Effects of
       Conservation Tillage on Groundwater Quality: Nitrates and Pesticides.  T.J. Logan et al. (eds.).
       Chelsea, MI: Lewis Publishers, Inc. pp. 65-91.

Blomquist, J.D., J.M. Denis, J.L. Cowles, J.A. Hetrick, RD. Jones, andN.B. Birchfield.  2001.
       Pesticides in Selected Water-Supply Reservoirs and Finished Drinking Water, 1999-2000:
       Summary of Results from a Pilot Monitoring Survey.  U.S. Geological Survey Open-File Report
       01-456. 65pp.

Boland, P.A.  1981. National Screening Program for Organics in Drinking Water. Part II.  Data.
       Prepared by SRI International, Menlo Park, CA, for U.S. Environmental Protection Agency,
       Office of Drinking Water, Washington, DC. EPA Contract. No. 68-01-4666.

California Department of Health Services.  1986. Final Report on a Monitoring Program for Organic
       Chemical Contamination of Large Public Water Systems in California. Summary Version.
       Sacramento, CA: Department of Health Services. April, 1986.

California Department of Health Services.  Unpublished. Final Report on a Monitoring Program for
       Organic Chemical Contamination of Small Public Water Systems in California.  Summary
       Version (tables). Sacramento, CA: Department of Health Services.

Carey, A.E., and F.W. Kutz. 1983. Trends in Ambient Concentrations ofAgrochemicals in Humans and
       the Environment of the United States.  Presented at the International Conference on
       Environmental Hazards ofAgrochemicals in Developing Countries, Alexandria,  Egypt,
       November 8-12, 1983.

Cohen, S.Z., S.M. Creeger, RF. Carsel, and C.G. Enfield. 1984. Potential for Pesticide Contamination
       in Groundwater Resulting from Agricultural Uses.  In: Treatment and Disposal of Pesticide
       Wastes. R.F. Kruger and J.W. Seiber (eds.). Washington, DC: American Chemical Society.
       ACS Symposium Series No. 259: pp. 247-325.

EXTOXNET.  1996. Pesticide Information Profile: Simazine. Ithaca, NY: Extension Toxicology
       Network, Pesticide Management Education Program. Available on the Internet at
       http://ace.ace.orst.edu/info/extoxnet/pips/simazine.htm, revised June, 1996.

Gilliom, R.J., O.K. Mueller,  and L.H. Nowell. 1998. Methods for comparing water-quality conditions
       among National Water-Quality Assessment Study Units, 1992-95. U.S. Geological Survey
       Open-File Report 97-589. Available on the Internet at:
       http://ca.water.usgs.gov/pnsp/rep/ofr97589, last updated October 9,  1998.

Hazardous Substances Data Bank (HSDB). 2001. Hazardous Substances Data Bank: Simazine.
       Available on the Internet at: http://toxnet.nlm.nih.gov/, last updated May  15, 2001.

Occurrence Summary and Use Support Document         275                                    March 2002

-------
Minnesota Department of Health (MDH) and Minnesota Department of Agriculture (MDA).  1988.
       Pesticides and Groundwater: Surveys of Selected Minnesota Wells.  Prepared by Thomas G.
       Klaseus (MDH), Greg C. Buzicky (MDA), and Edward C. Schneider (MDH) for Legislative
       Commission on Minnesota Resources. February, 1988.

Pellizzari, E.D.  1978. Preliminary Assessment ofHalogenated Organic  Compounds in Man and
       Environment Media. Monthly Technical Progress Report No. 5,  April 1-30, 1978.  Prepared by
       Research Triangle Institute, Research Triangle Park, NC, for Office of Toxic Substances, U.S.
       Environmental Protection Agency, Washington,  DC.  EPA Contract No. 68-01-4731.

Schlachter, J.  1988.  Memorandum on Program Improvement Subcommittee, FSTRAC, Preparation for
       April Meeting. Sent by: Jacquelyn D. Schlachter, Chairman - Program Improvement
       Subcommittee, U.S. Environmental Protection Agency, Region VII, Kansas City, KS. Sent to:
       Dennis Alt, IA; Pam Bonrud, SD; Bill Lee, NE; Kathy Miller, MT; Jennifer Orme, HQ; Sheila
       Sullivan, EPA Region V; and Dan Wilson, WI.

Staples, C.A., A.F. Werner, and T.J. Hoogheem. 1985. Assessment of Priority Pollutant Concentrations
       in the United States Using STORET Data Base.  In: Environmental Toxicology and Chemistry.
       v. 4, pp. 131-142. New York, New York: Pergamon Press, Ltd.

USEPA.  1987.  Health Advisory for Simazine. Washington, D.C: Office of Drinking Water, USEPA.
       August, 1987.

USEPA.  1988.  Computer Printout of STORET Water Quality Data Base Retrieval Conducted March
       23, 1988 by Science Applications International Corporation. Data available through the Office
       of Water Regulations and Standards.

USEPA.  2000.  TRIExplorer: Trends.  Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001.  National Primary Drinking Water Regulations - Consumer Factsheet on: SIMAZINE.
       Available on the Internet at: http://www.epa.gov/safewater/dwh/c-soc/simazine.html, last
       updated April 12,2001.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations -  DRAFT.  EPA Report/815-D-02-005, Office of
       Water,  55 pp.

USGS. 1998a.  Annual Use Maps. Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/,
       last updated March 20, 1998.

USGS. 1998b. Sources & Limitations of Data Used to Produce Maps of Annual Pesticide Use.
       Available on the Internet at: http://water.wr.usgs.gov/pnsp/use92/mapex.html, last updated
       March 20, 1998.

USGS. 1998c.  Pesticides in Surface and Ground Water of the United States: Summary of Results of the
       National Water Quality Assessment Program (NAWQA). PROVISIONAL DATA - SUBJECT TO
       REVISION. Available on the Internet at: http://water.wr.usgs.gov/pnsp/allsum/,  last updated
       July 22, 1998.
Occurrence Summary and Use Support Document        276                                    March 2002

-------
Wade Miller Associates, Inc.  1989.  Occurrence and Exposure Assessment ofSimazine in Public
       Drinking Water Supplies - DRAFT. Draft report submitted to EPA for review April 12, 1989.

Williams, W.M., P.W. Holden, D.W. Parsons, and M.N. Lorber. 1988. Pesticides in Groundwater Data
       Base. 1988 Interim report. Washington, DC: Office of Pesticide Programs, USEPA.
Occurrence Summary and Use Support Document         277                                    March 2002

-------
3.14   Toxaphene
Table of Contents

3.14.1 Introduction, Use and Production 	 279
3.14.2 Environmental Release  	 280
3.14.3 Ambient Occurrence 	 280
3.14.3 Drinking Water Occurrence Based on the 16-State Cross-Section	 281
3.14.4 Additional Drinking Water Occurrence Data 	 285
3.14.5 Conclusion	 287
3.14.6 References  	 288
Tables and Figures

Table 3.14-1: Stage 1 Toxaphene Occurrence Based on 16-State Cross-Section - Systems 	 282

Table 3.14-2: Stage 1 Toxaphene Occurrence Based on 16-State Cross-Section - Population	 283

Table 3.14-3: Stage 2 Estimated Toxaphene Occurrence Based on 16-State Cross-Section -
       Systems	 284

Table 3.14-4: Stage 2 Estimated Toxaphene Occurrence Based on 16-State Cross-Section -
       Population	 284

Table 3.14-5: Estimated National Toxaphene Occurrence - Systems and Population Served	 285
Occurrence Summary and Use Support Document         278                                    March 2002

-------
3.14.1  Introduction, Use and Production

Toxaphene is a manufactured insecticide containing over 670 chemicals. It does not occur naturally and
is usually found as a solid or gas. In its original form, toxaphene is a yellow to amber waxy solid that
smells like turpentine.  It does not burn and evaporates when in solid form or when mixed with liquids.
Toxaphene was one of the most heavily used insecticides in the United States until 1982, when it was
canceled for most uses; all uses were banned in 1990. It was used primarily in the southern United States
to control insect pests on cotton and other crops. It was also used to control insect pests on livestock and
to kill unwanted fish in lakes.  Toxaphene is also known as camphechlor, chlorocamphene,
polychlorocamphene, and chlorinated camphene (ATSDR, 1996).

Toxaphene was formerly used as a nonsystemic stomach and contact insecticide. It was used to control
many insects that thrived on cotton, corn, fruit, vegetables, and small grains and to control the Cussia
obtusifola soybean pest. The principal use was for pest control on cotton crops.  Toxaphene was also
used to control livestock ectoparasites such as lice, flies, ticks, mange, and scab mites. Its relatively low
toxicity to bees and its long-persisting insecticidal effect made it especially useful in the treatment of
flowering plants.  At one time toxaphene was also used in the United States to eradicate fish. In 1974, an
estimated 44.1 million pounds was used in the United States, and was distributed as follows: 85% on
cotton; 7%  on livestock and poultry; 5% on other field crops; 3% on soybeans; and less than 1% on
sorghum (ATSDR, 1996).

Through the early 1970s toxaphene or mixtures of toxaphene with rotenone were used widely in lakes
and streams by fish and game agencies to eliminate biologic communities that were considered
undesirable for sport fishing. Such uses of toxaphene by fish and game agencies have been discontinued
in the United States and Canada. Toxaphene use in this country has declined drastically since 1975, at
which time it was reported to be the most heavily used pesticide. The total amount consumed was
estimated to be 18.7 million pounds in 1980 and 10.8 million pounds in 1982 (ATSDR, 1996).

In November 1982, EPA canceled the registrations of toxaphene for most of its uses as a pesticide or
pesticide ingredient. In the period  following November 1982, its use was restricted to controlling scabies
on sheep and cattle; controlling grasshopper and army worm infestations on cotton, corn, and small
grains; controlling specific insects  on banana and pineapple crops in Puerto Rico and the U.S. Virgin
Islands; and for emergency use only (to be determined on a case-by-case basis by EPA).  Formulations
suitable for other purposes were to be sold or distributed until December 31, 1983, for use only on
registered sites. The distribution or sale of remaining stocks of toxaphene formulations was permitted
until December 31, 1986, for use on no-till corn, soybeans, and peanuts (to control sicklepod), and  dry
and southern peas, and to control emergency infestations. All registered uses of toxaphene mixtures in
the United States and any of its territories were canceled in 1990 (ATSDR, 1996).

In 1976, toxaphene was produced primarily by Hercules Incorporated, Wilmington, Delaware.
Production  during 1976 by three U.S. companies (Hercules Incorporated, Tenneco, and Vicksburg
Chemical Co., a division of Vertac) totaled 41.9 million pounds, which was a 29% decline from the
production  level of 59.5 million pounds in  1975 (ATSDR, 1996).

Total U.S. production in 1977 was  estimated to be 39.9 million pounds. The most recent production
volumes are from 1982, when it was estimated that 3.7 million pounds were produced in the United
States.  This represents a decline of more than 90% from  1972, when toxaphene was the most heavily
manufactured insecticide in the United States with a production volume of 46.0 million pounds (ATSDR,
1996).  By  1982, when EPA canceled most of its uses, consumption was reported at 12 million pounds
(USEPA, 2001).

Occurrence Summary and Use Support Document         279                                    March 2002

-------
3.14.2  Environmental Release

Toxaphene is released directly to the environment primarily in its use as an insecticide. Total U.S.
production in 1982 was estimated to be 3.7 million pounds (ATSDR, 1996). Toxaphene is not listed as a
Toxics Release Inventory (TRI) contaminant, so no TRI release records are maintained.  Therefore, the
use of toxaphene (described in the previous section) may provide the primary indication of where
releases are most likely.

3.14.3  Ambient Occurrence

To understand the presence of a chemical in the environment, an examination of ambient occurrence is
useful. In a drinking water context, ambient water is untreated source water residing in surface waters
and aquifers. There are few available data on the occurrence of toxaphene in ambient waters of the
United States. The most comprehensive and nationally consistent data describing ambient water quality
in the United States are being produced through the United States Geological Survey's (USGS) National
Water Quality Assessment (NAWQA) program. However, national NAWQA data are currently
unavailable for toxaphene.

3.14.3.1 Additional Ambient Occurrence Data

Additional studies of ambient data are  summarized below.  A summary document entitled "Occurrence
and Human Exposure to Pesticides in Drinking Water, Food and Air in the United States of America"
(USEPA, 1989), was previously prepared for past USEPA assessments of various pesticides. In that
review, one national and several regional studies were included on the occurrence of toxaphene in
surface water non-drinking water sources. Barbash and Resek (1996) also conducted reviews of
pesticides in ground water non-drinking water sources, including a national review and a compilation of
various monitoring studies on the occurrence of toxaphene. The following information for the surface
water studies only is taken directly from "Occurrence and Human Exposure to Pesticides in Drinking
Water, Food and Air in the  United States of America" (USEPA,  1989).

3.14.3.1.1 Surface Water  Sources - National Studies

The National Surface Water Monitoring Program presented data on levels of toxaphene in surface water
samples collected throughout the U.S. during the period 1975 to  1979 (USEPA, 1989).  During this time,
2,479 samples were collected and analyzed for toxaphene.  Although no detection limit was given for
toxaphene, 11 positive samples (or 0.4%) were found during testing.  The range of positive values was
0.0 to 1.65 |ig/L.

3.14.3.1.2 Surface Water  Sources - Regional Studies

Mattraw (1975, as cited in USEPA, 1989) did not  detect toxaphene in any surface water samples
collected and analyzed during an organochlorine residue survey in Florida. The number of samples
analyzed and the detection limit were not reported.

Toxaphene was detected  in 11 (or 55%) of 20 Mississippi Delta lakes sampled by Herring and Cotton
(1970, as cited in USEPA, 1989).  The maximum reported concentration was 1.92 |ig/L (the detection
limit was not reported).
Occurrence Summary and Use Support Document         280                                     March 2002

-------
3.14.3.1.3 Ground Water Sources - National Studies

The Pesticides in Ground Water Data Base (PGWDB), developed by EPA's Offices of Pesticide
Programs (OPP), contains information derived from monitoring studies conducted by pesticide
registrants, universities, and government agencies. The data are presented in several categories,
including all information collected to date (excluding data known to be of poor quality and data that is
from point source contamination); data derived from scientifically confirmed agricultural use; and
confirmed information attributed to known point sources and documented misuse. While extreme care
was taken to confirm detections of pesticides in ground water, the authors caution that the studies
investigated have not been limited to drinking water supplies and some studies included samples from
many types of wells including observation and irrigation wells.  In addition, many studies have focused
on shallow ground water which may not be  representative  of drinking water sources in the area. As such,
the results cannot be used to estimate human exposure to pesticides in drinking water, but do provide an
additional assessment of possible toxaphene occurrence. The PGWDB is the most comprehensive
summary available on pesticide occurrence  in ground water, including data from a total of 68,824 wells
sampled for pesticide analytes from 45 States (Barbash and Resek, 1996).

Results of the review of the PGWDB indicate that toxaphene was detected in 9 of the 4,273 wells
sampled. The maximum concentration reported was 0.0180 |ig/L. The minimum concentration reported
was 0.00115 |ig/L. The frequency of exceedances of the MCL (or Health Advisory  Level [HAL], if the
MCL unavailable) among sampled wells is 0.070% (Barbash and Resek, 1996).

3.14.3.1.4 Ground Water Sources - Regional Studies

Several monitoring studies, including results from various multistate, State and local monitoring studies
of pesticides, are reviewed in Barbash and Resek (1996). Results from these studies indicate that
toxaphene was detected in at least one study reviewed, although it was not possible to determine the
exact number of studies for which toxaphene was an analyte, or exactly how many studies did detect
toxaphene.  In the summary information reviewed, details were provided regarding the geographic
distribution and the total number of systems collected, although further detail is beyond the  scope of this
report.

3.14.3 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of toxaphene occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States.  The  16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest:  0.003 mg/L; 0.0015 mg/L; and 0.001 mg/L.

All sixteen cross-section State data sets, with the exception of Alabama, contained occurrence data for
toxaphene. These data represent over 52,000 analytical results from approximately 14,000 PWSs during
the period from 1984 to 1998 (with most analytical results from 1992 to 1997).  The number of sample
results and PWSs vary by State, although the State data sets have been reviewed and checked to ensure
adequacy of coverage and completeness. The overall modal detection limit for toxaphene in the 16 cross-
section States is equal to 0.001 mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)
Occurrence Summary and Use Support Document         281                                     March 2002

-------
3.14.3.1 Stage 1 Analysis Occurrence Findings

Table 3.14-1 illustrates the low occurrence of toxaphene in drinking water for the public water systems in
the 16-State cross-section relative to three thresholds:  0.003 mg/L (the current MCL), 0.0015 mg/L, and
0.001 mg/L (the modal MRL). Only 1 PWS (approximately 0.00724%) had analytical results exceeding
the MCL; 0.0290% of systems (4 systems) had results exceeding 0.0015 mg/L; and 0.0362% of systems
(5 systems) had results exceeding 0.001 mg/L.

No ground water systems had any analytical results greater than the MCL. Only 1 ground water system
(0.00806%) had results above 0.0015 mg/L or 0.001 mg/L.

One (0.0716% of) surface water systems had results greater than the MCL. Approximately 3 (0.215%
of) surface water systems had at least one analytical result greater than 0.0015 mg/L. Four (0.286% of)
surface water systems had results exceeding 0.001 mg/L.
Table 3.14-1:  Stage 1 Toxaphene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.003
0.0015
0.001
Percent of Systems
Exceeding Threshold
0.000%
0.00806%
0.00806%
Number of Systems
Exceeding Threshold
0
1
1

Surface Water
0.003
0.0015
0.001
0.0716%
0.215%
0.286%
1
3
4

Combined Ground &
Surface Water
0.003
0.0015
0.001
0.00724%
0.0290%
0.0362%
1
4
5
Reviewing toxaphene occurrence in the 16 cross-section States by PWS population served (Table 3.14-2)
shows that approximately 0.131% of the population (125,000 people) was served by PWSs with at least
one analytical result of toxaphene greater than the MCL (0.003 mg/L).  Approximately 150,900 (0.159%
of) people were served by systems with an exceedance of 0.0015 mg/L. Almost 206,000 (0.216% of)
people were served by systems with at least one analytical result greater than 0.001 mg/L.

The percentage of population served by ground water systems with analytical results greater than the
MCL was equal to 0%.  When evaluated relative to 0.0015 mg/L and 0.001 mg/L, the percent of
population exposed was equal to 0.0436% (approximately 17,500 people).

The percentage of population served by surface water systems with exceedances of the MCL (0.003
mg/L) was equal to 0.227% (125,000 people). When evaluated relative to 0.0015 mg/L and 0.001 mg/L,
Occurrence Summary and Use Support Document
282
March 2002

-------
the percent of population exposed was equal to 0.243% (about 133,400 people) and 0.343% (about
188,400 people), respectively.
Table 3.14-2: Stage 1 Toxaphene Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.003
0.0015
0.001
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.0436%
0.0436%
Total Population Served
by Systems Exceeding
Threshold
0
17,500
17,500

Surface Water
0.003
0.0015
0.001
0.227%
0.243%
0.343%
125,000
133,400
188,400

Combined Ground &
Surface Water
0.003
0.0015
0.001
0.131%
0.159%
0.216%
125,000
150,900
205,900
3.14.3.2  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 3.14-3 and
3.14-4. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations (USEPA, 2002)).

No ground water or surface water PWSs had an estimated mean concentration of toxaphene exceeding
0.003 mg/L, or 0.0015 mg/L.  Approximately 0.000145% of PWSs (approximately 1 system in the 16-
State cross-section) were estimated to have a mean concentration greater than 0.001 mg/L. The
percentage of ground water PWSs with estimated mean concentration exceeding 0.001 mg/L was equal to
0.0000967% (less than 1 system in the 16 States). Approximately 0.000573% of surface water systems
(about 1 system in the 16  States) had estimated mean concentration values of toxaphene greater than
0.001 mg/L.
Occurrence Summary and Use Support Document
283
March 2002

-------
Table 3.14-3: Stage 2 Estimated Toxaphene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.003
0.0015
0.001
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0000967%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.003
0.0015
0.001
0.000%
0.000%
0.000573%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
1
0-0
0-0
0-0

Combined Ground
& Surface Water
0.003
0.0015
0.001
0.000%
0.000%
0.000145%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
1
0-0
0-0
0-0
Reviewing toxaphene occurrence by PWS population served (Table 3.14-4) shows that approximately
0.000833% of population served by all PWSs in the 16 States (an estimated 800 people in the 16 States)
were potentially exposed to toxaphene levels above 0.001 mg/L.  The percentage of population served by
ground water systems relative to 0.001 mg/L was equal to 0.0000376%. Approximately 0.00141% of the
population served by surface water systems (about 800 people) were exposed to toxaphene at levels
greater than 0.001 mg/L. When evaluated relative to athreshold of 0.003 mg/L and 0.0015 mg/L, the
percent of population exposed was equal to 0% for all system types.
Table 3.14-4: Stage 2 Estimated Toxaphene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.003
0.0015
0.001
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
0.0000376%
Range
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.003
0.0015
0.001
0.000%
0.000%
0.00141%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
800
0-0
0-0
0-0
Occurrence Summary and Use Support Document
284
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.003
0.0015
0.001
0.000%
0.000%
0.000833%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
800
0-0
0-0
0-0
3.14.3.3 Estimated National Occurrence

As illustrated in Table 3.14-5, the Stage 2 analysis estimates zero systems serving zero people nationally
have estimated mean concentration values of toxaphene greater than 0.003 mg/L or 0.0015 mg/L.  Only 1
surface water system serving approximately 1,800 people nationally had an estimated mean concentration
value of toxaphene greater than 0.001 mg/L. (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)
Table 3.14-5:  Estimated National Toxaphene Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.003
0.0015
0.001
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
0
Range
0-0
0-0
0-0

Surface Water
0.003
0.0015
0.001
0
0
1
0-0
0-0
0-0
0
0
1,800
0-0
0-0
0-0

Combined Ground
& Surface Water
0.003
0.0015
0.001
0
0
1
0-0
0-0
0-0
0
0
1,800
0-0
0-0
0-0
3.14.4  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of toxaphene in drinking water are also
reviewed. Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence and Human Exposure to Pesticides in Drinking Water, Food and Air in the United States of
America" (USEPA, 1989), is presented in the following section. This variety of studies and information
are presented regarding levels of toxaphene in drinking water, with the scope of the reviewed studies
Occurrence Summary and Use Support Document
285
March 2002

-------
ranging from national to regional. Note that none of the studies presented in the following section
provide the quantitative analytical results or comprehensive coverage that would enable direct
comparison to the occurrence findings estimated with the cross-section occurrence data presented in
Section 3.14.4. These additional studies, however, do enable a broader assessment of the Stage 2
occurrence estimates presented for this Six-Year Review.  All the following information in Section
3.14.5 is taken directly from "Occurrence and Human Exposure to Pesticides in Drinking Water, Food
and Air in the United States of America" (USEPA, 1989).

3.14.4.1 Ground Water Sources - National Studies

The Federal Reporting Data System (FRDS, 1984, as cited in USEPA, 1989) provides information on
public water supplies found to be in violation of current Maximum Contaminant Levels (MCLs).  Data
are not available on the number of ground water systems in the U.S. that have monitoring requirements
for toxaphene; however, no violations of the current MCL of 5 |ig/L were reported for toxaphene during
the years 1979-1983.

The 1978 Rural Water Survey (USEPA, 1984, as cited in USEPA, 1989) involved the collection of
samples from 267 households (the majority using private water supplies) in rural locations throughout the
U.S. and analyses for toxaphene. Atotal of 71 public drinking water systems of varying sizes using
ground water were covered by the survey. None of the  samples from the 71 ground water systems
studied exceeded the minimum quantification limit of 0.17 |ig/L for toxaphene.

3.14.4.2 Ground Water Sources - Regional Studies

Irwin and Healy (1978, as cited in USEPA, 1989) summarized a study of data collected in 1976 during a
water quality reconnaissance of public water supplies in Florida. None of the 100 water supplies
sampled utilizing the five aquifers in Florida contained toxaphene in excess of the detection limit (the
detection limit was not reported).

Tucker and Burke (1978, as cited in USEPA 1989) presented the results of analyses of samples collected
from wells in nine New Jersey counties. None of the samples tested contained concentrations of
toxaphene in excess of the minimum reportable concentration of 0.06 |ig/L for this study.

3.14.4.3 Surface Water Sources - National Studies

Data were obtained from the Federal Reporting Data System (FRDS, 1984, as cited in USEPA, 1989) on
violations of the current MCL of 5 |ig/L for toxaphene.  The data were obtained for the years 1979 to
1983 and include information for all surface water systems in the United States. The analysis indicated
that none of the samples from the surface water systems studied contained toxaphene in excess of the
current MCL.

The 1978 Rural Water Survey (USEPA, 1984, as cited in USEPA, 1989) also presented data on drinking
water samples obtained from surface water systems of varying sizes.  None of the samples from the 21
public drinking water systems contained concentrations of toxaphene in excess of the minimum
quantification limit of 0.17 |ig/L.

3.14.4.4 Surface Water Sources - Regional Studies

Irwin and Healy (1978, as cited in USEPA, 1989), summarizing data collected during a water quality
reconnaissance of public water supplies in Florida, reported that none of the samples from the 16 surface

Occurrence Summary and Use Support Document         286                                    March 2002

-------
water supplies studied contained concentrations of toxaphene in excess of the detection limit. The
detection limit was not reported.

In a study on the effects of forest runoff on the quality of a public water supply in Oregon, Elliot (1979,
as cited in USEPA, 1989) observed a concentration of toxaphene of 3 |ig/L.

To assemble a database which would reflect the status of Great Lakes drinking water quality, the
Canadian Public Health Association gathered data from October 1984 through August 1985 (Canadian
Public Health Association, 1986, as cited in USEPA,  1989). The data collected covered the period from
the mid 1970s to early 1985. The study was funded by the Health Protection Branch of Health and
Welfare Canada and the Ontario Ministry of the Environment. A research team, appointed by the
Association, reviewed data on the quality of water at 31 representative Canadian and United States
communities and 24 offshore sites to evaluate the human health implications.

For each of the 31 communities, data consisted of:  1) background information on the community; 2)
treatment plant schematics and associated treatment process information; and 3) water quality data.
Water sample types included raw water (treatment plant intake), distribution water (treated water), and
tap water. Water quality data collected included general parameters (e.g., alkalinity, turbidity),
microbiological and radiological parameters, inorganic parameters, and organic parameters (including
volatiles, base/neutrals, pesticides and PCBs, and phenols and acids). For each parameter, the water
type, time period, concentration (mean, range), number of samples and detection limit are presented.

For most of the volatile organics, including toxaphene, the available data indicate that there were very
low levels of these contaminants in the raw, treated, or tap water.  Most of the values found were "not
detected" or near the detection limit (Canadian Public Health Association, 1986, as cited in USEPA,
1989).

3.14.4.5 Unspecified  Sources - National Study

In an EPA survey (USEPA, 1977, as cited in USEPA, 1989) of pesticide contamination in commercial
drinking water sampled during 1975 and  1976, 27 of 58 samples analyzed were found to be positive for
toxaphene. The levels of toxaphene were at or below 0.05 |ig/L. The detection limit was not reported.

3.14.4.6 Unspecified  Sources - Regional Study

In a report on source identification of pollutants entering a sewage treatment plant, Levins et al. (1979, as
cited in USEPA 1989) two drinking water sources in a drainage basin were tested in Georgia. Although
the detection limits were not reported, toxaphene was not detected in either of the samples tested.

3.14.5 Conclusion

Toxaphene was formerly used as a nonsystemic stomach and contact insecticide.  The principal use was
for pest control on cotton crops.  In the U.S., toxaphene use has declined drastically since 1975, at which
time it was reported to be the most heavily used pesticide. In November 1982, EPA canceled the
registrations of toxaphene for most of its uses as a pesticide or pesticide ingredient. All  registered uses
of toxaphene mixtures in the United States and any of its territories were canceled in 1990. Toxaphene is
not a TRI chemical, so there is no information available on releases of toxaphene.  There are also no
ambient data available. The Stage 2 analysis, based on the 16-State cross-section, estimated that zero
percent of combined ground water and surface water systems serving zero percent of the population
Occurrence Summary and Use Support Document         287                                     March 2002

-------
exceeded the MCL of 0.003 mg/L. Based on this estimate, zero PWSs nationally are estimated to have
levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. The cross-section should adequately represent the occurrence of toxaphene on a
national scale based upon the fact that toxaphene is no longer produced or used in any state and that there
are no documented releases for toxaphene.

3.14.6  References

Agency for Toxic Substances and Disease Registry (ATSDR). 1996. Toxicological Profile for
        Toxaphene.  U.S. Department of Health and Human Services, Public Health Service. 215 pp. +
        Appendices. Available on the Internet at: http://www.atsdr.cdc.gov/toxprofiles/tp94.pdf

Barbash, J. E., and E. A. Resek.  1996. Pesticides in Ground Water: Distribution, Trends, and
        Governing Factors. Volume two of Pesticides in the Hydrologic System. Chelsea, MI: Ann
        Arbor Press, Inc. 588 pp.

Canadian Public Health Association.  1986. Comprehensive Survey of the Status of Great Lakes
        Drinking Water. Prepared in cooperation with Health and Welfare Canada and the Ministry of
        the Environment. Ottawa, Canada: Canadian Public Health Association. August 1986.

Elliot, W.M. 1979. Portland's Public Water Supply - Pure and Simple.  Proceedings of the National
        Conference on Environmental Engineering. New York, NY: American Society of Civil
        Engineers.

Federal Reporting Data System (FRDS). 1984. Computer printout containing data on organic chemical
        MCL violations, FY 1979-1983. Washington, DC: USEPA. April 4, 1984.

Herring, J., and  D. Cotton.  1970.  Pesticide residues of 20 Mississippi delta lakes. Proceedings of the.
        4th Annual Conference of S.E. Assoc. Game Fish Comm. 482.

Irwin, G.A., and H.G. Healy.  1978. Chemical and Physical Quality of Selected Public Water Supplies in
        Florida, August-September 1976.  Tallahassee, FL: Water Resources Division, U.S. Geological
        Survey. USGS/WRI  78-21.

Levins, P., J. Adams, P. Brenner, S. Coons, K. Thrun, and J. Varone. 1979. Sources of Toxic Pollutants
        Found in Influents to  Sewage Treatment Plants. IV. R.M. Clayton Drainage Basin, Atlanta
        report.  Prepared by Arthur D. Little, Inc., for Office of Water Planning and Standards, U.S.
        Environmental Protection Agency, Washington, DC. EPA Contract No. 68-01-3857.

Mattraw, H.C.   1975. Occurrence of chlorinated hydrocarbon insecticides, southern Florida, 1968-1972.
       Pestic. Monit. J. v. 9, p. 106.

Tucker, R.K., and T.A. Burke.  1978. A Second Preliminary Report on the Findings of the State
        Groundwater Monitoring Project.  New Jersey: Department of Environmental Protection.

USEPA. 1977.  Toxaphene: Position document 1. Washington, DC: Special Pesticide Review Division,
        USEPA. EPA/SPRD-80/55.

Occurrence Summary and Use Support Document          288                                     March 2002

-------
USEPA.  1984. Rural Water Survey. Computer data provided by Department of Sociology, Cornell
       University, Ithaca, New York.

USEPA.  1989. Draft Final Report on the Occurrence and Human Exposure to Pesticides in Drinking
       Water, Food, and Air in the United States of America. Office of Drinking Water, USEPA.
       September, 1989. 202pp.

USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on: TOXAPHENE.
       Available on the Internet at: http://www.epa.gov/OGWDW/dwh/c-soc/toxaphen.html, last
       updated April 12,2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.
Occurrence Summary and Use Support Document         289                                   March 2002

-------
                                          [Intentionally blank]
Occurrence Summary and Use Support Document          290                                         March 2002

-------
4.0 VOLATILE ORGANIC CONTAMINANTS
Occurrence Summary and Use Support Document      291                          March 2002

-------
4.1    Benzene
Table of Contents

4.1.1  Introduction, Use and Production  	 293
4.1.2  Environmental Release  	 294
4.1.3  Ambient Occurrence 	 295
4.1.4  Drinking Water Occurrence Based on the 16-State Cross-Section	 296
4.1.5  Additional Drinking Water Occurrence Data 	 301
4.1.6  Conclusion	 307
4.1.7  References  	 307
Tables and Figures

Table 4.1-1:  Benzene Manufacturers and Processors by State	 294

Table 4.1-2:  Environmental Releases (in pounds) for Benzene in the United States, 1988-1999 .... 295

Table 4.1-3:  Stage 1 Benzene Occurrence Based on 16-State Cross-Section - Systems 	 297

Table 4.1-4:  Stage 1 Benzene Occurrence Based on 16-State Cross-Section - Population	 298

Table 4.1-5:  Stage 2 Estimated Benzene Occurrence Based on 16-State Cross-Section -
       Systems	 299

Table 4.1-6:  Stage 2 Estimated Benzene Occurrence Based on 16-State Cross-Section -
       Population	 300

Table 4.1-7:  Estimated National Benzene Occurrence - Systems and Population Served	 301
Occurrence Summary and Use Support Document         292                                    March 2002

-------
4.1.1  Introduction, Use and Production

Benzene, also known as benzol, has the chemical formula C6H6 and is a colorless liquid with a sweet
odor.  Benzene evaporates into air very quickly and dissolves slightly in water.  It is also highly
flammable. Benzene is found in the environment as a result of both anthropogenic and natural processes.
It is also a part of crude oil and gasoline and cigarette smoke. Benzene was first discovered and isolated
from coal tar in the 1800s. Today, benzene is made mostly from petroleum sources. Because of its wide
use, benzene ranks in the top 20 in production volume for chemicals produced in the United States
(ATSDR,  1997).

Benzene is produced naturally by volcanoes and forest fires and is present in many plants and animals. It
is also a major industrial chemical made from coal and oil (NSC, 2001).

The greatest use of benzene is as a building block for making plastics, rubber, resins and synthetic fabrics
like nylon and polyester (USEPA, 2001).  Benzene is also used in the following  manufacturing
processes: medicines, dyes, artificial leather, linoleum, oil cloth, pesticides, plastics and resins, PCB,
aviation fuel, detergents, flavors and perfumes, paints and coatings, airplane dope, varnishes, lacquers,
explosives and other organics.  It is used in photogravure printing and as a component of high octane
gasoline. In organic synthesis, it is used to make ethylbenzene, isopropyl-benzene, cyclohexane, aniline,
maleic anhydride and alkylbenzenes (NTP, 2001).

Benzene is also a component of gasoline since it occurs naturally in crude oil and it is a byproduct of oil
refining processes. The percentage by volume of benzene in unleaded gasoline is approximately 1-2%.
In the past, certain consumer products, such as some types of paint strippers and carpet glue, carburetor
cleaners, denatured alcohol, textured carpet liquid detergent, furniture wax, and rubber cement used in
tire patch kits and arts and crafts supplies contained small amounts of benzene (ATSDR, 1997).

The use of benzene in certain pesticides has been canceled. Under the Federal Insecticide, Fungicide,
and Rodenticide Act (FIFRA), the USEPA has established a voluntary cancellation of registered products
containing benzene as an active ingredient. Benzene was once used alone or in formulations to control
screwworms on animals and as an early fumigant for grain. Under the Food, Drug, and Cosmetics Act,
benzene is regulated as an indirect food additive and restricted for use only as a component of adhesives
used on articles intended for packaging, transport, or holding  foods (ATSDR, 1997).

More than 98% of the benzene produced in the U.S. is derived from the petrochemical and petroleum
refining industries. Catalytic reformat is the major source of benzene; between 1978 and 1981, catalytic
reformats accounted for approximately 44-50% of the total  U.S. benzene production. In 1994, benzene
was the  17th highest volume chemical produced in the US, with 14.7 billion pounds produced. Other
production data are as follows:  12.3 billion pounds in 1993, 12.0 billion pounds in 1992, and 11.5 billion
pounds in  1991 (ATSDR,  1997).

Table 4.1-1 shows the number of facilities in each State that manufacture and process benzene, the
intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1997).
Occurrence Summary and Use Support Document         293                                     March 2002

-------
Table 4.1-1:  Benzene Manufacturers and Processors by State
State"
AK
AL
AR
AZ
CA
CO
DE
GA
HI
IA
IL
IN
KS
KY
LA
MD
MI
MN
MO
MS
MT
NC
ND
NE
NJ
NM
NY
OH
OK
OR
PA
PR
SC
TN
TX
UT
VA
VT
WA
WI
wv
WY
Number of
facilities
4
18
4
2
35
2
4
6
2
1
26
18
10
12
40
3
28
3
7
6
4
3
1
1
12
4
5
25
6
1
21
5
3
9
94
8
5
1
6
4
8
5
Range of maximum amounts on site in
thousands of pounds'"
0-10,000
0-50,000
1-10,000
10-1,000
0-50,000
1,000-10,000
1-50,000
0-10,000
1,000-10,000
1,000-10,000
1-100,000
1-100,000
0-50,000
1-50,000
0-500,000
1-10,000
0-10,000
1-100,000
0-100
10-50,000
100-500,000
0-1,000
10,000-50,000
0-1
1-50,000
1-10,000
1-1,000
0-100,000
1-50,000
1,000-10,000
0-50,000
100-50,000
0-1,000
0-10,000
0-500,000
10-50,000
1-1,000
100,000-500,000
100-100,000
1-10,000
1-10,000
0-10,000
Activities and uses0
1,3,4,8,10
1,2,3,4,5,6,7,8,9,11,13
1,3,4,7,13
1,6,10,13
1,2,3,4,5,6,7,8,9,10,13
1,3,4,6,7
1,6,7,9,12,13
1,2,4,8,9,10,11
1,2,6
1,4
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,5,6,8,9,10,11,13
1,3,4,5,6,7,8,9,10,13
1,3,4,5,6,7,8,9,10,11,13
1,2,3,4,5,6,7,8,9,10,11,13
7,8,9
1,2,3,4,5,6,7,8,9,10,11,13
1,3,4,6,8,9
8,9,13
1,3,4,6,7,8
1,2,3,4,6,7,8,9
1,5,11
1,2,3,4,7
11
1,2,3,4,6,7,8,9,10,11,13
1,3,4,6,8,13
1,2,4,5,9,11,13
1,2,3,4,5,6,8,9,10,11,13
1,2,3,4,5,6,7,8,9,13
10
1,2,3,4,5,6,7,8,10,11,13
1,2,3,5,6,7,8
1,5,7,12
1,3,5,6,7,8,9,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
1,2,3,4,6,7,8,9,10,13
1,2,3,5,6,7,9,11
1,2,3,4,7
1,2,3,4,6,7,8
1,6,9,12,13
1,4,5,6,7,8,9,13
1,3,4,5,6,7,8,10
aPost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses:
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
8. As a formulation component
9. As a product component
10. For repackaging only
11. As a chemical processing aid
12. As a manufacturing aid
13. Ancillary or other uses
Source: AT SDR, 1997 compilation ofTRI93 1995 data
4.1.2  Environmental Release

Benzene is listed as a Toxics Release Inventory (TRI) chemical.  Table 4.1-2 illustrates the
environmental releases for benzene from 1988 to 1999.  (Benzene data are only available for these years.)
Air emissions constitute most of the on-site releases, with a steady decrease over the years, except for
Occurrence Summary and Use Support Document
                            294
March 2002

-------
small increases in 1997 and 1999. The decrease in air emissions has been the major contributor to
decreases in benzene total on- and off-site releases, as the release total declined every year but 1997.
Surface water discharges have also generally decreased, and averaged around 15,000 pounds for 1997-
1999. Underground injection declined until 1994, and has been rising ever since. Releases to land (such
as spills or leaks within the boundaries of the reporting facility) have varied between levels of 717,000
pounds and 18,000 pounds between 1988 and 1999, with no discernable trend.  Off-site releases
(including metals or metal compounds transferred off-site) also span a wide range of values, although
releases between 1988 and 1990 were greater than any releases since. These TRI data for benzene were
reported from 49 States (except Vermont), Puerto Rico, Washington B.C., the Virgin Islands, Guam,
American Samoa, and the Northern Mariana Islands (USEPA, 2000). Fifteen of the 16 cross-section
States (used for analyses of benzene occurrence in drinking water; see Section 4.1.4) reported releases of
benzene, the exception being Vermont. (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.1-2:  Environmental Releases (in pounds) for Benzene in the United States, 1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
7,287,778
7,239,578
8,777,297
8,219,965
9,410,086
9,853,271
11,362,776
13,236,803
19,032,803
25,846,649
27,422,786
32,341,184
Surface Water
Discharges
13,647
15,070
12,226
27,357
21,290
22,294
19,450
24,347
27,238
25,303
169,170
46,732
Underground
Injection
617,825
504,109
363,100
312,766
282,642
223,103
356,660
355,683
824,342
689,066
668,610
825,035
Releases
to Land
18,732
237,544
62,474
51,178
18,583
25,555
31,220
342,241
113,153
717,008
114,132
126,728
Off-Site Releases
147,994
185,440
83,476
66,246
71,391
203,685
94,171
93,486
140,989
264,685
241,518
396,880
Total On- &
Off-site
Releases
8,085,976
8,181,741
9,298,573
8,677,512
9,803,992
10,327,908
11,864,277
14,052,560
20,138,525
27,542,71 1
28,616,216
33,736,559
 Source: USEPA, 2000
4.1.3  Ambient Occurrence

Benzene was detected in 14 out of 398 wells (3.5%) in urban areas of the local, State, and federal data set
compiled by NAWQA.  The minimum and maximum concentrations detected were 0.2 |ig/L and 290
l-ig/L, respectively. The median value of detection concentrations was  1.0 |ig/L. Benzene was also
detected in 24 of the 2,504 wells (0.96%) with analysis in rural areas. The minimum and maximum
concentrations detected were 0.2 |ig/L and 73 |ig/L, respectively.  The median value of detection
concentrations was 0.7 |ig/L. These data (urban and rural) represent untreated ambient ground water of
the conterminous United States for the years 1985-1995 (Squillace et al., 1999).

Benzene was also an analyte in both the NURP and NPDES data. The NURP study found benzene in
urban runoff. The minimum and maximum concentrations detected were 3.5 |ig/L and 13 |ig/L,
respectively, with no mean value reported.  The use of the land from which the samples were taken was
unspecified. The NPDES related investigations analyzing urban and highway runoff detected benzene.
Occurrence Summary and Use Support Document
295
March 2002

-------
The maximum concentration of benzene in NPDES samples of stormwater was 0.8 |ig/L, and the
drinking water standard (MCL) is 5 |ig/L (Lopes and Dionne, 1998).

4.1.3.1 Additional Ambient Occurrence Data

A summary document entitled "Benzene: Occurrence in Drinking Water, Food, and Air" (JRB
Associates, 1983), was previously prepared for past USEPA assessments of benzene.  However, no
information on the ambient occurrence of benzene was included in that document. (The document did
include information regarding benzene occurrence in drinking water, which is discussed in Section 4.1.5
of this report.)

4.1.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of benzene occurrence presented in the following section is based on State compliance
monitoring data from the 16 cross-section States. The 16-State cross-section is the largest and most
comprehensive compliance monitoring data set compiled by EPA to date. These data were evaluated
relative to several concentration thresholds of interest: 0.005 mg/L; 0.0005 mg/L; and 0.0004 mg/L.

All sixteen cross-section State data sets contained occurrence data for benzene.  These data represent
more than 194,000 analytical results from approximately 23,000 PWSs during the period from 1984 to
1998 (with most analytical results from 1992 to 1997).  The number of sample results and PWSs vary by
State, although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness.  The overall modal detection limit for benzene in the 16 cross-section States is equal to
0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to Section 1.3.5 of this
report.)

4.1.4.1 Stage 1 Analysis Occurrence Findings

Table 4.1-3 illustrates the occurrence of benzene in drinking water for the public water systems in the 16-
State cross-section relative to three thresholds: 0.005 mg/L (the current MCL), 0.0005 mg/L (the modal
MRL), and 0.0004 mg/L.  Based on the 16-State cross-section data, a total of 44 (approximately 0.189%
of) ground water and surface water PWSs had analytical results exceeding the MCL; 1.02% of systems
(237 systems) had results exceeding 0.0005 mg/L; and 1.24% of systems (288 systems) had results
exceeding 0.0004 mg/L.

Approximately 0.185% of ground water systems (40  systems) had any analytical results greater than the
MCL. About 0.941% of ground water systems (204 systems) had results above 0.0005 mg/L. The
percentage of ground water systems with at least one result greater than  0.0004 mg/L was equal to 1.10%
(239 systems).

Only 4 (0.251% of) surface water systems had results greater than the MCL.  A total of 33 (2.07% of)
surface water systems had at least one analytical result greater than 0.0005 mg/L.  Forty-nine (3.07% of)
surface water systems had results exceeding 0.0004 mg/L.
Occurrence Summary and Use Support Document         296                                     March 2002

-------
Table 4.1-3:  Stage 1 Benzene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Systems
Exceeding Threshold
0.185%
0.941%
1.10%
Number of Systems
Exceeding Threshold
40
204
239

Surface Water
0.005
0.0005
0.0004
0.251%
2.07%
3.07%
4
33
49

Combined Ground &
Surface Water
0.005
0.0005
0.0004
0.189%
1.02%
1.24%
44
237
288
Reviewing benzene occurrence by PWS population served (Table 4.1-4) shows that approximately
0.581% (almost 575,000 people) of the 16-State population was served by PWSs with at least one
analytical result of benzene greater than the MCL (0.005 mg/L). The percentage of population exposed
dramatically increased when evaluated relative to 0.0005 mg/L and 0.0004 mg/L.  Approximately 12.4
million (11.2% of) people were served by systems with an exceedance of 0.0005 mg/L.  A total of about
13.2 million (11.9% of) people were served by systems with at least one analytical result greater than
0.0004 mg/L.

The percentage of population served by ground water systems with analytical results greater than the
MCL was equal to 0.980% (about 486,300 people). When evaluated relative to 0.0005 mg/L or 0.0004
mg/L, the percent of population exposed was equal to 6.20% (3,074,800 people) and 6.74%
(approximately 3.3 million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.005 mg/L was
equal to 0.144% (88,300 people). Approximately 15.2% of the population served by surface water
systems (about 9.3 million people) was exposed to benzene concentrations greater than 0.0005 mg/L.
When evaluated relative to 0.0004 mg/L, the percent of population exposed was equal to 16.0% (over 9.8
million people).
Occurrence Summary and Use Support Document
297
March 2002

-------
Table 4.1-4:  Stage 1 Benzene Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Population
Served by Systems
Exceeding Threshold
0.980%
6.20%
6.74%
Total Population Served by
Systems Exceeding
Threshold
486,300
3,074,800
3,346,200

Surface Water
0.005
0.0005
0.0004
0.144%
15.2%
16.0%
88,300
9,289,900
9,823,500

Combined Ground &
Surface Water
0.005
0.0005
0.0004
0.518%
11.2%
11.9%
574,700
12,364,700
13,169,800
4.1.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.1-5 and 4.1-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

A total of 7 (0.0312% of) ground water and surface water PWSs in the 16  States had an estimated mean
concentration of benzene exceeding 0.005 mg/L.  Approximately 65 (0.279% of) PWSs in the 16 States
had an estimated mean concentration exceeding 0.0005 mg/L, and 80  (0.343%) had an estimated mean
concentration exceeding 0.0004 mg/L.

An estimated 7 (0.0332% of) ground water PWSs in the 16 cross-section States had a mean concentration
greater than 0.005 mg/L, 61 (0.282%) had a mean concentration greater than 0.0005 mg/L, and 75
(0.344%) had a mean concentration greater than 0.0004 mg/L.  Approximately 1 (0.00363% of), 4
(0.237% of), and 5 (0.323% of) surface water PWSs in the 16 States had estimated mean concentrations
exceeding 0.005 mg/L, 0.0005 mg/L, and 0.0004 mg/L, respectively.
Occurrence Summary and Use Support Document
298
March 2002

-------
Table 4.1-5:  Stage 2 Estimated Benzene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0332%
0.282%
0.344%
Range
0.01 38% -0.0554%
0.198% -0.369%
0.249% - 0.452%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
7
61
75
Range
3-11
43-80
54-98

Surface Water
0.005
0.0005
0.0004
0.00363%
0.237%
0.323%
0.000% - 0.0627%
0.0627% -0.501%
0.125% -0.627%
1
4
5
0-1
1-8
2-10

Combined Ground
& Surface Water
0.005
0.0005
0.0004
0.0312%
0.279%
0.343%
0.0172% -0.0516%
0.198% -0.370%
0.245% - 0.456%
7
65
80
4-12
46-86
57-106
Reviewing benzene occurrence by PWS population served (Table 4.1-6) shows that approximately
0.00947% of population served by all PWSs in the 16 cross-section States (an estimate of about 10,500
people) was potentially exposed to benzene levels above 0.005 mg/L. The percentage of population
served by PWSs in the 16 States with levels of benzene above 0.0005 mg/L and 0.0004 mg/L were
0.260% (about 288,000 people) and 0.309% (over 342,000 people), respectively.

When the percent of population served by ground water systems was evaluated relative to a threshold of
0.005 mg/L,  0.0005 mg/L, and 0.0004 mg/L, the percentage of population exposed in the 16 cross-section
States was equal to 0.0204% (an estimated 10,100 people), 0.481% (an estimated 238,700 people) and
0.546% (an estimated 271,100 people),  respectively.

The percentage of population served by surface water systems with levels above 0.005 mg/L was equal to
0.000625% (an estimated 400 people in the 16 States), and the percentage of population served with
levels above  0.0005 mg/L was 0.0805% (over 49,000 people). The percentage of the population served
by surface water systems with levels above 0.0004 mg/L was 0.117% (about 71,400 people in the  16
States).
Occurrence Summary and Use Support Document
299
March 2002

-------
Table 4.1-6:  Stage 2 Estimated Benzene Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.0204%
0.481%
0.546%
Range
0.00407% - 0.0629%
0.340% -0.632%
0.399% -0.758%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
10,100
238,700
271,100
Range
2,000-31,200
168,900-313,500
197,900-376,000

Surface Water
0.005
0.0005
0.0004
0.000625%
0.0805%
0.117%
0.000% - 0.0220%
0.0194% -0.283%
0.0244% -0.451%
400
49,300
71,400
0-13,500
11,900-173,100
15,000-276,200

Combined Ground
& Surface Water
0.005
0.0005
0.0004
0.00947%
0.260%
0.309%
0.00188% -0.0300%
0.176% -0.389%
0.211% -0.514%
10,500
288,000
342,500
2,100 - 33,200
194,700-431,400
234,200 - 570,100
4.1.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 20 PWSs nationally serving approximately 20,200 people could be exposed to
benzene concentrations above 0.005 mg/L. About 181 systems serving 553,400 people nationally had
estimated mean concentrations greater than 0.0005 mg/L. Approximately 223 systems serving about
658,000 people nationally were estimated to have mean benzene concentrations greater than 0.0004
mg/L.  (See Section  1.4 for a description of how Stage 2 16-State estimates are extrapolated to national
values.)

For ground water systems, an estimated 20 PWSs serving about 17,500 people nationally had mean
concentrations greater than 0.005 mg/L.  Approximately 168 systems serving about 412,100 people
nationally had estimated mean concentration values that exceeded 0.0005 mg/L. About 204 ground
water systems serving almost 468,200 people had estimated mean concentrations greater than 0.0004
mg/L.

Approximately 1 surface water systems serving 800 people was estimated to have mean concentration of
benzene above 0.005 mg/L. About 13 surface water systems serving 102,500 people had estimated mean
concentrations greater than 0.0005 mg/L. An estimated 18 surface water systems serving approximately
148,500 people had mean concentrations greater than 0.0004 mg/L.
Occurrence Summary and Use Support Document
300
March 2002

-------
Table 4.1-7:  Estimated National Benzene Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
20
168
204
Range
8-33
118-219
148 - 269
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
17,500
412,100
479,900
Range
3,500 - 53,900
291,700-541,300
341,800 - 649,200

Surface Water
0.005
0.0005
0.0004
1
13
18
0-4
4-28
7-35
800
102,500
151,400
0 - 28,000
24,700 - 359,800
31,100-574,200

Combined Ground
& Surface Water
0.005
0.0005
0.0004
20
181
223
11-34
129 - 240
159-296
20,200
553,400
658,000
4,000 - 63,800
374,000 - 828,800
449,900 - 1,095,300
4.1.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of benzene in drinking water are also reviewed.
Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence of Benzene in Drinking Water, Food, and Air" (JRB Associates, 1983), is presented in this
section. This variety of studies and information are presented regarding levels of benzene in drinking
water, with the scope of the reviewed studies ranging from national to regional. Note that none of the
studies presented in the following section provide the quantitative analytical results or comprehensive
coverage that would enable direct comparison to the occurrence findings estimated with the cross-section
occurrence data presented in Section 4.1.4. These additional studies, however, do enable a broader
assessment of the Stage 2 occurrence estimates presented for this Six-Year Review. All the following
information in Section 4.1.5 is taken directly from "Occurrence of Benzene in Drinking Water, Food, and
Air" (JRB Associates, 1983).

4.1.5.1 Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Four national drinking water surveys provide data on benzene: the National Organic Monitoring Survey
(NOMS), the National Screening Program for Organics in Drinking Water (NSP), the 1978 Community
Water Supply Survey (CWSS), and the Groundwater Supply Survey (GWSS). The terms used in this
report are those used in the individual surveys, recognizing that they may not always correspond to strict
technical definitions.

The National Organic Monitoring Survey (NOMS) was conducted to identify contaminant sources, to
determine the frequency of occurrence of specific drinking water contaminants, and to provide data for
the establishment of maximum contaminant levels (MCL's) for various organic compounds in drinking
water (Brass et al., 1977, as cited in JRB Associates,  1983). The NOMS was conducted in three phases:
March-April 1976, May-July 1976, and November 1976-January 1977.  Finished drinking water samples
Occurrence Summary and Use Support Document
301
March 2002

-------
from 113 communities were analyzed for 21 different compounds. Of the 113 community supplies
sampled, 18 had groundwater sources, 91 had surface water sources, and 4 had a mixed
groundwater/surface water source.

The analytical results of the NOMS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water.  Additional information on the locations and source of the supplies,
and on the populations served by the supplies in the NOMS were provided by Wayne Mello (1983) at
EPA's Technical Support Division, Office of Drinking Water.  A single value for benzene was reported
for each supply studied in the NOMS.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March 1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic chemical contaminants. Data are available for benzene on finished
water samples from 12 groundwater and 100 surface water supplies. However, the data for three of the
groundwater supplies and 27 of the  surface water supplies were indicated by  Boland (1981, as cited in
JRB Associates, 1983) to be suspect due to potential interference by chlorodibromomethane and
trichloroethylene.

In the Community Water Supply Survey (CWSS), carried out in 1978, 106 surface water supplies, 330
groundwater supplies, and 16 supplies with mixed sources were examined for volatile organic chemical
contamination. Fourteen purgeable organic compounds were analyzed and total organic carbon levels
were determined. Samples were taken of raw, finished, and distribution water.  Only the  latter two types
of water are considered here. Data for benzene in finished and/or distribution samples were obtained
from a total of 289 groundwater and 97 surface water supplies.

The Groundwater Supply Survey (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al, 1983, as cited in JRB Associates,  1983). It was hoped that this study would
stimulate State efforts toward the detection and control of groundwater contamination and the
identification of potential chemical "hot spots." A total of 945  systems were sampled, of which 466 were
chosen at random.  The remaining 479 systems were chosen nonrandomly based on information from
States encouraged to identify locations believed to have a higher than normal probability  of VOC
contamination (e.g., locations near landfills or industrial activity). The file provided a single analytical
result for each supply sampled.  One sample of finished water was collected from each supply at a point
near the entrance to the distribution system.

Each of the drinking water surveys was evaluated with respect to the validity of the reported occurrence
data for a number of organic chemicals, including benzene. The evaluations  were carried out by
analyzing information about the  procedures used for collection and analysis of samples as well as the
quality control protocols used. The reported results for the compounds  studied in each survey were
judged with respect to their qualitative acceptability (i.e., was the substance correctly identified as
benzene), and their quantitative acceptability (i.e., are the reported measurements reliable). In the case of
benzene, a qualitatively acceptable but quantitatively unacceptable rating was given for data from the
NOMS, NSP and CWSS due to suspected biodegradation of the samples, which were held unrefrigerated
for prolonged periods before analysis (particularly the CWSS). Benzene values in excess of the
quantitation limit reported for some samples in these studies are qualitatively valid and can be taken as
minimum values, representative  of samples which probably originally contained benzene at higher
concentrations. In the case of the GWSS, all data were rated both quantitatively and qualitatively
acceptable.
Occurrence Summary and Use Support Document         302                                     March 2002

-------
4.1.5.2 Groundwater - Federal Surveys

The National Organic Monitoring Survey (NOMS), the National Screening Program for Organics in
Drinking Water (NSP), the Community Water Supply Survey (CWSS), and the Groundwater Supply
Survey (GWSS) all contain data concerning the levels of benzene in groundwater supplies from across
the country.

Eighteen groundwater systems were analyzed for benzene during Phase I of NOMS (March to April
1976); the Phase I data are not reported here because of sample contamination invalidating the results.
Eighteen groundwater systems were sampled during Phase II of the study (May to July 1976) with one
containing benzene at a concentration of 0.1 |ig/L.  Only one supply was analyzed during Phase III of the
study (November 1976 to January 1977) for benzene; none was observed. The minimum quantifiable
limits for benzene were 0.1-0.2 |ig/L in Phase II and 0.2 |ig/L in Phase III.

Twelve groundwater supplies were tested for benzene contamination in the NSP. Data for three of these
were indicated by Boland (1981, as cited in JRB Associates, 1983) to be suspect because of a potential
analytical interference. Of the remaining 9 systems, none were found to be contaminated with benzene
above the quantification limit of 0.1 |ig/L.

The 1978 CWSS provided information on benzene levels in 289 groundwater systems. Of these systems,
5 contained detectable levels of benzene, with values ranging from 0.51-21.8 |ig/L. The mean value was
5.0 |ig/L with a standard deviation of 9.3 |ig/L; the median value was 0.95 |ig/L.  The minimum
quantitation limits for benzene in the CWSS ranged from 0.50-1 |ig/L.

In the GWSS, 3 of the 456 randomly chosen water systems serving 25 or more individuals were
contaminated with benzene, at concentrations of 0.61, 3, and 15 |ig/L. Of the 473 nonrandom locations
sampled serving 25 or more individuals, 8 were contaminated with benzene, at concentrations between
0.5-12 |ig/L, the highest values being 1.6, 2.7, and  12 |ig/L. The average benzene level for the
nonrandom systems was 4.1 |ig/L with a standard deviation of 4.9 |ig/L; the median value was 2.2 |ig/L.
The minimum quantitation limit for benzene was 0.5  |ig/L.

4.1.5.3 Groundwater - State Data

Data supplied to the U.S. Environmental Protection Agency by Connecticut and Delaware, showed
relatively small amounts of benzene in the groundwater supplying one of four cities examined. Six
samples from this site had benzene concentrations ranging from nondetectable to 4.0 |ig/L, with an
average of 1.8 |ig/L. In Elkhart, Indiana, five samples had benzene concentrations ranging from 0.3-
3,500 |ig/L, with an average of 2,200 |ig/L.  Benzene was detected in drinking water samples from
Lunenberg, Massachusetts, where 10 samples had benzene levels from undetectable to 145 |ig/L (average
of 70 |ig/L). Finally, in Nassau County, New York, eight of 165 wells tested for benzene showed
positive results, but were not quantified.

4.1.5.4 Surface Water - Federal Surveys

The National Organic Monitoring Survey (NOMS), the National Screening Program for Organics in
Drinking Water (NSP), and the Community Water Supply Survey (CWSS) all contain data concerning
the levels of benzene in surface water supplies from across the country.

In Phase I of the National Organic Monitoring Survey (March to April 1976), water samples from 89
surface water systems were analyzed for benzene; the Phase I data are not reported here because of

Occurrence Summary and Use Support Document         303                                    March 2002

-------
sample contamination invalidating the results. Ninety-one systems were sampled during the second
phase of the survey and 5 were found to be contaminated, at levels ranging from 0.1-1.8 |ig/L. The
average benzene concentration among the positive Phase II samples was 0.5 |ig/L with a standard
deviation of 0.7 |ig/L; the median concentration was 0.3 |ig/L.  During the third phase of the NOMS
(November 1976 to January 1977), analyses revealed benzene contamination in 4 out of atotal of 14
systems sampled.  The contamination levels for these systems ranged from 0.1-1.5 |ig/L. The mean
concentration was 0.9 |ig/L with a standard deviation of 0.6 |ig/L; the median level was 1.0 |ig/L. The
minimum quantifiable limits for benzene ranged from 0.1-0.4 |ig/L in Phase II and 0.1-0.2  |ig/L in Phase
III.

Surface water samples from 100 drinking water systems were analyzed for benzene during the National
Screening Program (NSP) between June 1977 and March 1981. Data for 27 of these were  indicated by
Boland (1981, as cited in JRB Associates, 1983) to be suspect because of a potential analytical
interference. Of the remaining 73 systems, 23 contained detectable levels of benzene, ranging from 0.1-
1.8 |ig/L. The average concentration among the 23 positive systems was 0.2 |ig/L with a standard
deviation of 0.4 |ig/L; the median level was 0.1 |ig/L. The quantification limit for the NSP was 0.1 |ig/L.

One of the 97 surface water systems sampled during the Community Water Supply Survey (CWSS)
contained benzene, at a concentration of 0.56 |ig/L. The minimum quantitation limit for benzene in the
CWSS was 0.5 ^g/L.

4.1.5.5  Surface Water - State Data

Three States provided data to the EPA on levels of benzene in drinking water supplies derived from
surface water.  Connecticut and Delaware reported data for a total of four cities, none of which had any
positive samples.  In New York, one of two cities supplying surface water data had detectable benzene
concentrations.  Of five samples from Waterford, New York, four were positive for benzene, with levels
of 1.0-4.7 |ig/L (average of 2.2 |ig/L).

4.1.5.6  Projected National Occurrence of Benzene in Public Water Supplies

As reported in the JRB Associates (1983) report, public water systems in the United States fall into two
major categories with respect to water source (surface water and groundwater) and into five size
categories and twelve subcategories according to the number of individuals served. The JRB 1983 report
presented estimates of both the number of drinking water supplies nationally within each of the
source/size categories expected to have benzene present, and of the concentration of benzene expected to
be present in those supplies.

The key features of the methodology used and assumptions made to develop the national estimates are
summarized here.  The estimates are based on the data from the Federal surveys only.  The State data
were not included for several reasons.  Generally, these data are from a few States and were not
considered to be geographically representative. There was also a general lack of data on the population
served by systems measured, the type of water sampled, and the methodologies used to sample, identify,
and measure benzene.

The Federal survey data from the NOMS, NSP, CWSS, and GWSS were pooled together for developing
the national projections. It was assumed in combining these surveys that the resulting data base would be
representative of the nation's water supplies.  In the case of the GWSS data, both the random and
nonrandom samples were included in the projections because a statistical test of the GWSS data showed
Occurrence Summary and Use Support Document         304                                     March 2002

-------
no statistically significant difference in the frequency of occurrence of positive values or the mean of the
positive values between the random and nonrandom samples.

Ideally, adequate data would be available to develop the national projections separately for each of the
twelve system size categories within the groundwater and surface water groups; however, the available
data were too limited for this. JRB (1983) consolidated some of the size categories to have sufficient
data for developing the projections. In consolidating data from various size categories,  consideration was
given to the potential for there being statistically significant differences in the frequency of occurrence of
benzene as a function of system size.  The consolidation of size categories therefore involved a balancing
of the need to group size categories together to have an adequate data base for developing the national
projections against the need to treat size categories separately in order to preserve the influence of system
size as a determinant of contamination potential. The consolidation of size categories also took into
account EPA's classification of systems into the five major groups as very small (25-500), small (501-
3,300), medium (3,301-10,000), large (10,001-100,000), and very large (> 100,000) (Kuzmack, 1983, as
cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national level were tested and
an appropriate model selected.  In the case of benzene, the delta distribution was used for groundwater
and the multinominal method was used for surface water.  The frequency of contamination of
groundwater and surface water systems at various concentrations was determined for each consolidated
size category. For completing the national estimates, it was assumed that the frequency of contamination
observed for each consolidated category was directly applicable to each of the system sizes comprising it.

In the JRB Associates (1983) report, it is noted that some of the data used in computing the national
estimates are from samples held for a prolonged period of time prior to analysis, with possible
biodegradation of benzene.  Therefore, they concluded that these projections of national occurrence may
underestimate actual contaminant levels.

4.1.5.6.1 Groundwater Supplies

The combined benzene groundwater data from the NOMS, NSP, CWSS, and GWSS surveys are
available. JRB Associates (1983) also reported that data were available for a total of 1,223 supplies from
the combined surveys.  Of these,  17 supplies were reported to have benzene present, at concentrations
ranging from 0.13 |ig/L to 15 |ig/L. Based on the overall distribution of positive values and maximum
possible values for those supplies in which benzene was not found, 0.5 |ig/L was selected as the common
minimum quantifiable concentration for the combined survey data.  That is, quantitative projections are
made  of supplies at several concentration ranges > 0.5 |ig/L, while only a total number for supplies
expected to have either no benzene or levels below 0.5 |ig/L can be determined. Although some data
indicate the presence of benzene in groundwater supplies at levels < 0.5 |ig/L, it is not possible to
determine the proportion of supplies that have benzene present and the proportion that are actually free of
benzene contamination.

Of the 1,206 supplies reporting no benzene to be present,  1,202 were assumed to have maximum possible
levels of < 0.5 |ig/L based on the  minimum quantifiable concentrations reported for the  various surveys.
The other 4 supplies reporting no benzene to be present had maximum possible levels ranging from 0.6
|ig/L to 1.0 |ig/L.  It is assumed, based on the overall distribution of values, that benzene if present in
these 4 supplies is so at a concentration of < 0.5 |ig/L, although a rigorous, conservative argument could
be made for assuming a level equal to the maximum possible value.
Occurrence Summary and Use Support Document         305                                      March 2002

-------
Sixteen of the 1,223 supplies examined had measured values of benzene > 0.5 |ig/L. When the twelve
size categories were consolidated into the five major EPA groupings, there was no apparent trend in the
frequency of values > 0.5 |ig/L as a function of size:

Very small            0.3%   (1/362)
Small                 1.4%   (4/293)
Medium               1.6%   (3/186)
Large                 2.4%   (8/338)
Very large            0%     (0/44)
Overall               1.3%   (16/1,223)

A test for statistical significance revealed that at the 0.05 level, the difference among the five categories
was not significant; therefore, all supplies were grouped together for the analysis. As noted previously,
the frequency of occurrence of benzene at various concentrations was determined for the consolidated
group and then applied to the number of supplies nationally within each of the individual size categories
comprising the group.

No groundwater supplies  are estimated to have benzene present in any of the 10 |ig/L ranges  above 50
l-ig/L, but it can be estimated that there are two supplies with benzene present at levels exceeding 50
l-ig/L. This apparent discrepancy is due to the rounding of fractional numbers of supplies resulting from
the delta  distribution to the nearest integer.  That is, the fractional numbers of supplies estimated to have
benzene above 50 |ig/L were rounded to 0 within each individual 10 |ig/L range,  while the sum of all the
fractional numbers above 50 |ig/L was rounded to 2.  The cumulative estimates are probably more
descriptive of the national occurrence than are the estimates within each 10 |ig/L range. An estimated
635 groundwater supplies (range of 330-939), approximately 1.3% of the total groundwater supplies in
the United States, are expected to have benzene at levels of > 0.5 |ig/L; the remaining 47,823 supplies
have either no benzene or levels < 0.5 |ig/L.

It is estimated that 60 supplies (range of 0-154) are expected to have benzene levels  > 10 |ig/L; while 2
supplies (range of 0-23) are expected to have levels > 50 |ig/L. Most of the supplies with high benzene
levels are expected to be in the smaller size categories.  Although, as noted previously, the frequency of
benzene occurrence appears to be independent of system size, the number of systems affected nationally
is greater for the small sizes because there are many more small systems in existence.

4.1.5.6.2  Surface Water Systems

Data are available for a total of 223 surface water supplies.  Of these, 29 supplies were reported by JRB
Associates (1983) to have benzene present at concentrations ranging from 0.1 |ig/Lto 1.4 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which benzene was not found, 0.5 |ig/L was selected as the common minimum quantifiable concentration
for the combined survey data. That is, quantitative projections are made of supplies at several
concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either no
benzene or levels below 0.5 |ig/L can be determined. Although some data indicate the presence of
benzene in surface water supplies at levels < 0.5 |ig/L, it is not possible to determine the proportion that
have benzene present and the proportion that are free of benzene contamination.

Only 6 of the 223  supplies examined had measured values of benzene > 0.5 |ig/L. The twelve size
categories were consolidated into the five major EPA groupings, showing the following frequency of
occurrence of values > 0.5 |ig/L as a function of size:

Occurrence Summary and Use Support Document          306                                     March 2002

-------
Very small     0%    (0/18)
Small
Medium
Large
Very large
Overall
2.9% (1/34)
4.2% (1/24)
1.9% (1/52)
3.2% (3/95)
2.7% (6/223)
A test for statistical significance revealed that at the 0.05 level, there was no difference in the frequency
of occurrence of benzene at levels > 0.5 |ig/L among these groups.  Therefore, the data from all sizes
were consolidated to develop the national estimates. As noted previously, the frequency of occurrence of
benzene at various concentrations was determined for the consolidated group and then applied to the
number of supplies nationally within each of the individual size categories comprising it.

An estimated 301 surface water supplies (range of 64-537), approximately 2.7% of the total surface water
systems in the United States, are expected to have benzene at levels > 0.5 |ig/L; the remaining 10,901
supplies have either no benzene or levels < 0.5 |ig/L. It is estimated that no surface water supplies will
have levels > 5 |ig/L.

4.1.6  Conclusion

Benzene is a manufactured organic chemical that is used heavily in several industries and (in 1994) was
the 17th highest volume chemical produced in the U.S.  Most benzene is used as a building block for
making plastics, rubber, resins and synthetic fabrics like nylon and polyester. It is also used in the
manufacture of medicines, dyes, artificial leather, linoleum, oil cloth, pesticides, plastics and resins,
PCB, aviation fuel, other organics, and in organic synthesis.  Recent statistics regarding production and
use indicate that benzene is  a very abundant chemical.  Industrial releases of benzene have been reported
to TRI since 1988 in 49  States, Puerto Rico, Washington B.C., the Virgin Islands, Guam, American
Samoa, and the Northern Mariana Islands.  Benzene was also an analyte for the NAWQA, NURP, and
NPDES ambient occurrence studies. In the NAWQA study,  benzene was detected in 3.5% of urban wells
and 0.96% of rural wells, with median detection values of 1.0 |ig/L and 0.7 |ig/L, respectively.  The Stage
2 analysis, based on the  16-State cross-section, estimated that approximately  0.0313% of combined
ground water and surface water systems serving 0.0105% of the population had estimated mean
concentrations of benzene greater than the MCL of 0.005 mg/L. Based on this estimate, approximately
20 PWSs nationally serving approximately 22,500 people are expected to have estimated mean
concentrations of benzene greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC,  SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, benzene is manufactured  and/or processed in over 40 States and has TRI
releases in every state except for Vermont. Benzene is  manufactured and/or processed  in 14 out of the 16
cross-section States and has TRI releases in 15 of the 16 cross-section States. The cross-section should
adequately represent the occurrence of benzene on a national scale based upon the use,  production, and
release patterns of the  16-State cross-section in relation to the patterns observed for all 50 States.

4.1.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 1997. Toxicological Profile for Benzene.
       U.S. Department of Health and Human  Services, Public Health Service. 423 pp. + Appendices.
       Available on the Internet at http://www.atsdr.cdc.gov/toxprofiles/tp3.pdf
Occurrence Summary and Use Support Document          307                                     March 2002

-------
Boland, P.A.  1981. National Screening Program for Organics in Drinking Water.  Prepared by SRI
       International, Menlo Park, California, for Office of Drinking Water, U.S. Environmental
       Protection Agency, Washington, DC. EPA Contract No. 68-01-4666.

Brass, H.J., M.A. Feige, T. Halloran, J.W. Mello, D. Munch, and R.F. Thomas.  1977. The National
       Organic Monitoring Survey: A sampling and analysis for purgeable organic compounds.
       Drinking water quality enhancement through source protection. R.B. Pojasek (ed.).  Ann Arbor,
       MI: Ann Arbor Science, pp. 393-416.

JRB Associates.  1983. Benzene: Occurrence in Drinking Water, Food, and Air. Draft report submitted
       to EPA for review November 23, 1983.

Kuzmack, A.M.  1983. Memorandum: Characterization of the water supply industry (FY82).
       Washington, D.C.: Office of Water, U.S. Environmental Protection Agency. May 16, 1983.

Lopes, T.J. and S.G. Dionne. 1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater. U.S. Geological Survey Open-File Report 98-409.
       67pp.

Mello, Wayne.  1983. Personal communication between Wayne Mello, Technical Support Division,
       Office of Drinking Water, U.S. Environmental Protection Agency, and author of JRB Associates,
       1983, March 10, 1983.

National Safety  Council (NSC). 2001. Benzene Chemical Backgrounder.  Itasca, IL: National Safety
       Council. Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?id=83&chempath=chemicals, accessed
       July 23, 2001.

National Toxicology Program (NTP). 2001.  National Toxicology Program Health and Safety
       Information Sheet - Benzene.  Available on the Internet at
       http://ntp-server.niehs.nih.gov/htdocs/CHEM_H&S/NTP_Chem7/Radian71-43-2.html, last
       updated August 13,2001.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, R.M. Clawges, and J.S. Zogorski.  1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci. and Tech.  33(23):4176-4187.

USEPA. 2000. TRIExplorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001.  National Primary Drinking  Water Regulations - Consumer Factsheet on: Benzene.
       Office of Ground Water and Drinking Water, USEPA. Available on the Internet at
       http://www.epa.gov/safewater/dwh/c-voc/benzene.html, last updated April 12, 2001.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.
Occurrence Summary and Use Support Document         308                                    March 2002

-------
Westrick, J.J., J.W. Mello, and R.F. Thomas.  1983. The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data. EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         309                                     March 2002

-------
4.2    Carbon Tetrachloride
Table of Contents

4.2.1  Introduction, Use and Production  	 311
4.2.2  Environmental Release  	 312
4.2.3  Ambient Occurrence 	 313
4.2.4  Drinking Water Occurrence Based on the 16-State Cross-Section	 313
4.2.5  Additional Drinking Water Occurrence Data 	 318
4.2.6  Conclusion	 327
4.2.7  References  	 327
Tables and Figures

Table 4.2-1: Carbon Tetrachloride Manufacturers and Processors by State 	 312

Table 4.2-2: Environmental Releases (in pounds) for Carbon Tetrachloride
       in the United States, 1988-1999  	 313

Table 4.2-3: Stage 1 Carbon Tetrachloride Occurrence Based on 16-State Cross-Section -
       Systems	 314

Table 4.2-4: Stage 1 Carbon Tetrachloride Occurrence Based on 16-State Cross-Section -
       Population	 315

Table 4.2-5: Stage 2 Estimated Carbon Tetrachloride Occurrence Based on 16-State
       Cross-Section - Systems	 316

Table 4.2-6: Stage 2 Estimated Carbon Tetrachloride Occurrence Based on 16-State
       Cross-Section - Population	 317

Table 4.2-7: Estimated National Carbon Tetrachloride Occurrence - Systems and
       Population Served	 318
Occurrence Summary and Use Support Document         310                                     March 2002

-------
4.2.1  Introduction, Use and Production

Carbon tetrachloride (CC14) is a clear liquid that evaporates very easily. It has a sweet odor and does not
easily burn.  Most carbon tetrachloride that escapes to the environment is, therefore, found as a gas.
Carbon tetrachloride does not occur naturally but has been produced in large quantities to make
refrigeration fluid and propellants for aerosol cans (ATSDR, 1994).  Its major routes of entry to drinking
water are a consequence of its industrial production and use (JRB Associates, 1983). It is expected that
discharges to surface water during production and use, and  leaching into groundwater from wastes
deposited in landfills are primary causes of carbon tetrachloride contamination of drinking water (JRB
Associates, 1983).  Carbon tetrachloride is also known as carbona, carbon chloride, carbon tet, methane
tetrachloride, perchloromethane, tetrachloromethane, and benzinoform (ATSDR, 1994).

With few exceptions, production of carbon tetrachloride has been eliminated in developed nations,
including the U.S.  As an ozone-depleting substance (ODS), carbon tetrachloride falls under the auspices
of the Montreal Protocol, which mandated the cessation of production of carbon tetrachloride as of
January 1, 2000. The only exceptions to this rule are for "transformation" and "destruction".

Under the transformation exemption, the Protocol excludes from phase-out those ODSs that are used as
feedstocks during the manufacturing process. The rationale for this exemption is that no ODS is released
into the atmosphere after transformation, and thus it is unnecessary for the Protocol to prohibit such use.
To qualify for the exemption under international law, an ODS must be nearly 100% transformed into a
non-ODS product and any residual must be destroyed as provided under the destruction exemption
(McKenna and Cuneo, 2001).

Under the destruction exemption, controlled substances that are destroyed rather than emitted to the
atmosphere are not subject to the ban.  This exemption is based on the Protocol's definition of
"production" which is "the amount of controlled substances produced, minus the amount destroyed by
technologies approved by the Parties" (McKenna and Cuneo, 2001).

Prior to the enactment of the Montreal Protocol, use of carbon tetrachloride was already heavily
restricted. All consumer uses, such as use in the production of refrigeration fluid and propellants for
aerosols, as a pesticide, as a cleaning fluid, and in fire extinguishers, were prohibited, and its application
was limited to some industrial applications. The major use  of carbon tetrachloride in the early 1990s was
in the production of chlorofluorocarbons, such as dichlorodifluoromethane and trichlorofluoroemthane
(ATSDR, 1994).

The decrease in use of carbon tetrachloride has led to a similar decrease in the amount produced.
According to the latest figures available before  the ban on production, the U.S. produced between 573
and 761 million pounds each year from 1981-1988. Production was 413 million pounds in 1990 and 315
million pounds in 1991. Because of the continual phasing out of most uses of carbon tetrachloride,
production was expected to lessen to an even greater extent in the 1990s, although no data is available to
corroborate that assumption (ATSDR, 1994).

Table 4.2-1 shows the number of facilities in each State that manufactured and processed carbon
tetrachloride as of 1992, the intended uses of the product, and the range of maximum amounts derived
from the Toxics Release Inventory (TRI) of EPA (ATSDR, 1994).
Occurrence Summary and Use Support Document         311                                     March 2002

-------
Table 4.2-1:  Carbon Tetrachloride Manufacturers and Processors by State3
Stateb
AL
AR
CA
CO
DE
GA
IL
IN
KS
KY
LA
MD
MI
MN
MO
MS
MT
ND
NJ
NY
OH
OK
PA
TN
TX
VA
VI
WV
WY
Number of facilities
4
2
8
1
1
2
3
2
5
6(1)'
16
1
2
2
2
2
1
1
4
1(1)'
6
2
2
2
14
1
1
2
1
Range of maximum amounts on site in thousands of
pounds0
10-49,999
10-99
0-49,999
1-9
100-999
0-0.09
10-9,999
10-99
10-9,999
10-9,999
0-49,999
10-99
10-999
1-99
10-99
0.1-99
0.1-0.9
10-99
0-9,999
No Data
1-999
10-99
10-99
1-999
1-49,999
10-99
10-99
0.1-999
1-9
Activities and usesd
1,4,5,11
7,11
1,3,4,5,7,8,11,12,13
1,5,11
7
8,9,12
7,11
1,5,6,11
1,3,4,7,11
1,3,5,7,10,11,13
1,3,4,5,6,7,11,12,13
7
7,11
11,13
13
11,13
11
11
3,7,8,9,11
11
1,5,10,11,12,13
11
11
11
1,4,5,6,7,8,10,11,12,13
13
11
1,4,6
11
'Derived from TRI90 (1992)
bPost office State abbreviations used
cData in TRI are maximum amounts on site at each facility
dActivities/Uses include:
                        8. As a formulation component
                        9. As an article component
                        10. For repackaging only
                        11. As a chemical processing aid
                        12. As a manufacturing aid
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
                        13. Ancillary or other uses
eNumber of facilities reporting "no data" regarding maximum amount of the substance on site

Source: ATSDR, 1994 compilation O/TR190 1992 data
4.2.2 Environmental Release

Carbon tetrachloride is listed as a Toxics Release Inventory (TRI) chemical.  Table 4.2-2 illustrates the
environmental releases for carbon tetrachloride from 1988 to 1999.  (Carbon tetrachloride data are only
available for these years.)  Air emissions constitute the vast majority of the on-site releases, with a steady
decrease over the years.  Surface water discharges have generally decreased from 1988 to 1999, with the
exception of an increase in 1998.  Releases to land (such as spills or leaks within the boundaries of the
reporting facility) decreased from 1988 to  1993, remained at zero for three years, and then began
increasing in  1997. Off-site releases (including metals or metal compounds transferred off-site) have
generally decreased from 1988 to 1999. The decrease in air emissions, as well as underground injection,
have predominantly contributed to decreases in carbon tetrachloride total on- and off-site releases in
recent years.  These TRI data for carbon tetrachloride were reported from 38 States, Puerto Rico, and the
Virgin Islands (USEPA, 2000). Sixteen of the 38 States reported every year.  Twelve of the cross-section
Occurrence Summary and Use Support Document
                                                   312
March 2002

-------
States (used for analyses of carbon tetrachloride occurrence in drinking water; see Section 4.2.4) reported
releases. (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.2-2:  Environmental Releases (in pounds) for Carbon Tetrachloride in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
237,235
274,291
357,902
363,681
420,754
651,098
2,243,590
1,399,490
1,549,782
1,753,340
3,465,179
3,795,248
Surface Water
Discharges
84
2,586
315
215
717
1,223
1,453
2,444
2,849
4,718
15,656
15,627
Underground
Injection
27,548
23,163
32,958
44,515
53,966
12,654
34,332
45,984
42,475
31,557
122,030
98,050
Releases
to Land
938
1,679
135
0
0
0
79
333
2,157
1,005
1,616
14,759
Off-Site Releases
7,307
9,956
18,697
9,245
7,735
50,791
121,363
11,955
39,111
10,163
24,996
49,703
Total On- &
Off-site
Releases
273,112
311,675
410,007
417,656
483,172
715,766
2,400,817
1,460,206
1,636,374
1,800,783
3,629,477
3,973,387
 Source: USEPA, 2000
4.2.3  Ambient Occurrence

National NAWQA data, as well as NURP and NPDES data, are currently unavailable for carbon
tetrachloride. Additional studies of ambient data are also unavailable. A summary document entitled
"Occurrence of Carbon Tetrachloride in Drinking Water, Food and Air" (JRB Associates, 1983), was
previously prepared for past USEPA assessments of carbon tetrachloride. However, no information on
the ambient occurrence of carbon tetrachloride was included in that document. (The document did
include information regarding carbon tetrachloride occurrence in drinking water, which is discussed in
Section 4.2.5 of this report.)

4.2.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of carbon tetrachloride occurrence presented in the following section is based on  State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative to several concentration thresholds of interest:  0.005 mg/L; 0.0025 mg/L; and 0.0005
mg/L.

All sixteen cross-section State data sets contained occurrence data for carbon tetrachloride. These data
represent more than 182,000 analytical results from approximately 23,000 PWSs during the period from
1984 to 1998 (with most analytical results from 1992 to 1997).  The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to ensure adequacy of
coverage and completeness. The  overall modal detection limit for carbon tetrachloride in the 16 cross-
section States is equal to 0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)
Occurrence Summary and Use Support Document
313
March 2002

-------
4.2.4.1 Stage 1 Analysis Occurrence Findings

Table 4.2-3 illustrates the occurrence of carbon tetrachloride in drinking water for the public water
systems in the 16-State cross-section relative to three thresholds: 0.005 mg/L (the current MCL), 0.0025
mg/L, and 0.0005 mg/L (the modal MRL). Based on the 16-State cross-section data, a total of 47
(approximately 0.204% of) ground water and surface water PWSs had analytical results exceeding the
MCL; 0.386% of systems (89 systems) had results exceeding 0.0025 mg/L; and 1.56% of systems (359
systems) had results exceeding 0.0005 mg/L.

Approximately 0.205% of ground water systems (44 systems) had any analytical results greater than the
MCL. About 0.350% of ground water systems (75 systems) had results above 0.0025 mg/L. The
percentage of ground water systems with at least one result greater than 0.0005 mg/L was equal to 1.32%
(283 systems).

Only 3 (0.191% of) surface water systems had results greater than the MCL. A total of 14  (0.889% of)
surface water systems had at least one analytical result greater than 0.0025 mg/L. Seventy-six (4.83% of)
surface water systems had results exceeding 0.0005 mg/L.
Table 4.2-3:  Stage 1 Carbon Tetrachloride Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems
Exceeding Threshold
0.205%
0.350%
1.32%
Number of Systems
Exceeding Threshold
44
75
283

Surface Water
0.005
0.0025
0.0005
0.191%
0.889%
4.83%
3
14
76

Combined Ground &
Surface Water
0.005
0.0025
0.0005
0.204%
0.386%
1.56%
47
89
359
Reviewing carbon tetrachloride occurrence in the 16 cross-section States by PWS population served
(Table 4.2-4) shows that approximately 7.35% of the 16-State population (over 8 million people) was
served by PWSs with at least one analytical result of carbon tetrachloride greater than the MCL (0.005
mg/L). Approximately 9.2 million (8.32% of) people were served by systems with an exceedance of
0.0025 mg/L. Over 13.6 million (12.3% of) people were served by systems with at least one analytical
result greater than 0.0005 mg/L.

The percentage  of population served by ground water systems with analytical results greater than the
MCL was equal to 0.830% (about 411,200 people). When evaluated relative to 0.0025 mg/L and 0.0005
Occurrence Summary and Use Support Document
314
March 2002

-------
mg/L, the percent of population exposed was equal to 2.14% (1,060,200 people) and 7.64%
(approximately 3.8 million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.005 mg/L was
equal to 12.6% (over 7.7 million people). Approximately 13.3% of the population served by surface
water systems in the 16 States (almost 8.2 million people) was exposed to carbon tetrachloride
concentrations greater than 0.0025 mg/L. When evaluated relative to 0.0005 mg/L, the percent of
population exposed was equal to 16.1% (over 9.8 million people).
Table 4.2-4:  Stage 1 Carbon Tetrachloride Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population
Served by Systems
Exceeding Threshold
0.830%
2.14%
7.64%
Total Population Served
by Systems Exceeding
Threshold
411,200
1,060,200
3,783,200

Surface Water
0.005
0.0025
0.0005
12.6%
13.3%
16.1%
7,721,100
8,147,200
9,818,800

Combined Ground &
Surface Water
0.005
0.0025
0.0005
7.35%
8.32%
12.3%
8,132,300
9,207,300
13,602,000
4.2.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.2-5 and 4.2-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

A total of 4 (0.0159% of) ground water and surface water PWSs in the 16 States had an estimated mean
concentration of carbon tetrachloride exceeding 0.005 mg/L.  Approximately 6 (0.0256% of) PWSs in
the 16 States had an estimated mean concentration exceeding  0.0025 mg/L, and 48 (0.210%) had an
estimated mean concentration exceeding 0.0005 mg/L.

An estimated 4 (0.0171% of) ground water PWSs in the 16 cross-section States had a mean concentration
greater than 0.005 mg/L, 6 (0.0273%) had a mean concentration greater than 0.0025 mg/L, and 41
(0.193%) had a mean concentration greater than 0.0005 mg/L. Approximately 0.000381% (less than 1),
Occurrence Summary and Use Support Document
March 2002

-------
0.00267% (about 1), and 0.441% of (approximately 7 ) surface water PWSs in the 16 States had
estimated mean concentrations exceeding 0.005 mg/L, 0.0025 mg/L, and 0.0005 mg/L, respectively.
Table 4.2-5:  Stage 2 Estimated Carbon Tetrachloride Occurrence Based on 16-State Cross-Section
- Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0171%
0.0273%
0.193%
Range
0.00932% - 0.0233%
0.01 86% -0.0373%
0.140% -0.242%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
4
6
41

Surface Water
0.005
0.0025
0.0005
0.000381%
0.00267%
0.441%
0.000% - 0.000%
0.000% - 0.0635%
0.254% - 0.699%
0
1
7

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.0159%
0.0256%
0.210%
0.00869% -0.0217%
0.0174% -0.0347%
0.161% -0.265%
4
6
48
Range
2-5
4-8
30-52

0-0
0-1
4-11

2-5
4-8
37-61
Reviewing carbon tetrachloride occurrence by PWS population served (Table 4.2-6) shows that
approximately 0.0316% of population served by all PWSs in the 16 cross-section States (an estimate of
approximately 35,000 people) was potentially exposed to carbon tetrachloride levels above 0.005 mg/L.
The percentage of population served by PWSs in the 16 States with levels of carbon tetrachloride above
0.0025 mg/L was 0.0474% (approximately 52,500 people).  Approximately 7.37% of the population
served in the 16 States (over 8.1 million people) was exposed to carbon tetrachloride concentrations
above 0.0005 mg/L.

When the percent of population served by ground water systems was evaluated relative to a threshold of
0.005 mg/L, 0.0025 mg/L, and 0.0005 mg/L, the percentage of population exposed in the 16 cross-section
States was equal to 0.0705% (an estimated 34,900 people), 0.104% (an estimated 51,600 people) and
0.764% (an estimated 378,500 people), respectively.

The percentage of population served by surface water systems with levels above 0.005 mg/L was equal to
0.0000713%, and the percentage of population served with levels above 0.0025 mg/L was 0.00139%
(about 800 people). The percentage of the population served by surface water systems dramatically
increased to about 12.7% (about 7.8 million people in the 16 States) when evaluated relative to 0.0005
mg/L.
Occurrence Summary and Use Support Document
316
March 2002

-------
Table 4.2-6:  Stage 2 Estimated Carbon Tetrachloride Occurrence Based on 16-State Cross-Section
- Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.0705%
0.104%
0.764%
Range
0.00158% -0.0950%
0.0938% -0.191%
0.627% -1.05%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
34,900
51,600
378,500
Range
800 - 47,000
46,500 - 94,400
310,600-521,700

Surface Water
0.005
0.0025
0.0005
0.0000713%
0.00139%
12.7%
0.000% - 0.000%
0.000% -0.00104%
12.4% -13. 9%
0
800
7,779,500
0-0
0-600
7,571,900-8,487,800

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.0316%
0.0474%
7.37%
0.000707% - 0.0429%
0.0420% - 0.0865%
7.16% -8.00%
35,000
52,500
8,154,900
800 - 47,500
46,500 - 95,700
7,916,000 - 8,849,500
4.2.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 10 PWSs serving approximately 67,400 people nationally could be exposed to
carbon tetrachloride  concentrations above 0.005 mg/L. About 17 systems serving 101,100 people had
estimated mean concentrations greater than 0.0025 mg/L. Approximately 137 systems serving about 15.7
million people nationally were estimated to have mean carbon tetrachloride concentrations greater than
0.0005 mg/L.  (See Section 1.4 for a description of how  Stage 2  16-State estimates are extrapolated to
national values.)

For ground water systems, an estimated 10 PWSs serving about  60,400 people nationally had mean
concentrations greater than 0.005 mg/L. Approximately 16 systems serving about 89,300 people
nationally had estimated mean concentration values that exceeded 0.0025 mg/L. About 115 ground
water systems serving almost 654,600 people had estimated mean concentrations greater than 0.0005
mg/L.

Approximately 1 surface water system serving less than 100 people was estimated to have a mean
concentration of carbon tetrachloride above 0.005 mg/L.  One surface water system serving about 1,800
people nationally had an estimated mean concentration greater than 0.0025 mg/L.  An estimated 25
surface water systems serving over 16 million people had mean concentrations greater than 0.0005 mg/L.
Occurrence Summary and Use Support Document
317
March 2002

-------
Table 4.2-7:  Estimated National Carbon Tetrachloride Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
10
16
115
Range
6-14
11-22
83 - 144
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
60,400
89,300
654,600
Range
1,400-81,400
80,400 - 163,300
537,200 - 902,200

Surface Water
0.005
0.0025
0.0005
1
1
25
0-0
0-4
14-39
<100
1,800
16,221,400
0-0
0-1,300
15,788,500-17,698,400

Combined Ground
& Surface Water
0.005
0.0025
0.0005
10
17
137
6-14
11-23
105 - 172
67,400
101,100
15,705,100
1,500-91,400
89,500-184,300
15,245,000-17,042,800
4.2.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of carbon tetrachloride in drinking water are
also reviewed. Previously compiled occurrence information, from an OGWDW summary document
entitled "Occurrence of Carbon Tetrachloride in Drinking Water, Food, and Air" (JRB Associates, 1983),
is presented in the following section. This variety of studies and information are presented regarding
levels of carbon tetrachloride in drinking water, with the scope of the reviewed studies ranging from
national to regional. Note that none of the studies presented in the following section provide the
quantitative analytical results or comprehensive coverage that would enable direct comparison to the
occurrence findings estimated with the cross-section occurrence data presented in Section 4.2.4.  These
additional studies, however, do enable a broader assessment of the Stage 2 occurrence estimates
presented  for this Six-Year Review. All the following information in Section 4.2.5 is taken directly from
"Occurrence of Carbon Tetrachloride in Drinking Water, Food, and Air" (JRB Associates, 1983).

JRB Associates (1983) found three major types of data available that were potentially useful for
describing the  occurrence of carbon tetrachloride in the nation's public drinking water supplies.  First,
there are several Federal surveys in which a number of public water supplies from throughout the U.S.
were selected for analysis of chemical contamination, including carbon tetrachloride. Second, data are
available from State surveys and from State investigations of specific incidents of known or suspected
contamination of a supply. Third, there are miscellaneous published data which, like some of the State
data, tend to be from studies in response to suspected contamination  of specific sites. For accomplishing
the basic objectives of this study, namely to estimate the number of public water supplies nationally
within the various source and size categories contaminated with carbon tetrachloride, the distribution of
carbon tetrachloride concentrations in those supplies, and the number of individuals exposed to those
concentrations, it was  determined that the Federal survey data provides the most  suitable data  base. The
State and miscellaneous data tend to be poorly described with respect to the source and size categories of
the supplies examined and the sampling and analysis methods used for determining contaminant levels.
Occurrence Summary and Use Support Document
318
March 2002

-------
The lack of source and system size information precludes using the data for estimating levels in public
water supplies of similar characteristics. The absence of details on sampling and analysis methods
precludes evaluating those data for their qualitative and quantitative reliability. Also, because much of
the State and miscellaneous data are from investigations in response to incidents of known or suspected
contamination (e.g., spills), they were judged to be not representative of contaminant levels in the
nation's water supplies in general. Although they are not used with the Federal data for the purpose of
estimating contamination levels nationally, the available State and miscellaneous data are presented here
to provide some additional perspective on carbon tetrachloride occurrence in drinking water.

Data are presented only on drinking water samples taken from a consumer's tap (i.e., distribution water
samples) or on treated water samples taken at the water supply (i.e., finished water samples) because
these are considered to be most representative of the water consumed by the public. No data on raw (i.e.,
untreated) water are presented.  It is recognized that for some groundwater supplies where no treatment
of the water occurs, samples identified as raw may be representative of water consumed by the users of
the supply. However, it was generally not possible to differentiate between those groundwater supplies
that do and those that do not treat raw water from the available survey data.

4.2.5.1 Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Six Federal drinking water surveys provide data on carbon tetrachloride: the National Organics
Reconnaissance Survey (NORS), the National Organic Monitoring Survey (NOMS), the National
Screening Program for Organics in Drinking Water (NSP), the 1978 Community Water Supply Survey
(CWSS), the Rural Water Survey (RWS), and the Groundwater Supply Survey (GWSS).  The terms used
in this report are those used in the individual surveys, recognizing that they may not always correspond to
strict technical definitions.

The National Organics Reconnaissance Survey (NORS) was conducted in  1975 to determine the extent
of the presence of carbon tetrachloride, 1,2-dichloroethane, and four trihalomethanes in drinking water
supplies from 80 cities across the country (Symons et al., 1975, as cited in JRB Associates, 1983). The
effect of the water sources and treatment practices on the formation of these compounds were also
examined in NORS. Of the 80 supplies studied, 16 were indicated as having a groundwater source  and
64 as having a surface water source.  Symons et al. (1975, as cited in JRB Associates, 1983) did not
provide data on the population served by the supplies studied in the NORS; the populations served were
estimated by JRB based on information available from other sources for the supplies studied and from
census data for the locations of the supplies.

The National Organic Monitoring Survey (NOMS) was conducted to identify contaminant sources, to
determine the frequency of occurrence of specific drinking water contaminants, and to provide data for
the establishment of maximum contaminant levels (MCLs) for various organic compounds in drinking
water (Brass et al. 1977, as cited in JRB Associates, 1983). The NOMS was conducted in three phases:
March to April 1976, May to July 1976, and November 1976 to January 1977. Finished drinking water
samples from 113 communities were analyzed for 21 different compounds. Of the 113  community
supplies sampled, 18 had groundwater sources, 91 had surface water sources, and 4 had a mixed
groundwater/surface water source. For carbon tetrachloride, data are available for  18 groundwater
sources and  89 surface water sources.

The analytical results of the NOMS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water.  Additional information on the locations and source of the supplies,
and on the populations served by the supplies in the NOMS were provided by Wayne Mello (1983, as
Occurrence Summary and Use Support Document         319                                     March 2002

-------
cited in JRB Associates, 1983) at EPA's Technical Support Division, Office of Drinking Water. A single
value for carbon tetrachloride was reported for each supply studied in the NOMS.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March 1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic chemical contaminants. Data are available for carbon tetrachloride
on finished water samples from 12 groundwater and 106 surface water supplies.

In the Community Water Supply Survey (CWSS), carried out in 1978, 106 surface water supplies, 330
groundwater supplies, and 16 supplies with mixed sources were examined for volatile organic chemical
contamination.  Fourteen purgeable  organic compounds were analyzed and total organic carbon levels
were determined.  Samples were taken of raw, finished, and distribution water. Only the latter two types
of water are considered here. Data for carbon tetrachloride in finished and/or distribution samples were
obtained from a total of 316 groundwater and 105 surface water supplies.

The Rural Water Survey (RWS), conducted in 1978, was carried out in response to Section 3 of the Safe
Drinking Water Act, which mandated that EPA "conduct a survey of the quantity, quality, and
availability of rural drinking water supplies." Drinking water samples were collected for analysis of
inorganic chemicals, pesticides, and VOCs from 2,655 households throughout the United States located
in areas defined in the survey as rural. Of these, a total of 855 household samples were examined for
VOCs. The majority of these samples were obtained from households receiving water from private wells
or small supplies serving fewer than 25 people.  For carbon tetrachloride, data are available in the RWS
for 207 groundwater and 45 surface water supplies serving 25 or more people.

The RWS did not obtain data on the number of persons in each household served by the supplies.
However, data were obtained on the number of service connections at each supply. With the input of Dr.
Bruce Brower at Cornell University, who participated in the statistical analysis of the RWS for
parameters other than VOCs, the population  served by each supply was estimated from the average
number of persons per household (3.034) observed in the survey. A single value was reported for each
household; in some cases it was necessary to average two or three households obtaining water from the
same supply. Brass (1981, as cited in JRB Associates, 1983) cautions that the RWS water samples were
analyzed 6 to 27 months after collection and that degradation of some VOCs may have  occurred during
this holding period.

The Groundwater Supply Survey (GWSS)  was conducted from  December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al, 1983, as cited in JRB Associates, 1983).  It was hoped that this study would
stimulate State efforts toward the detection and control of groundwater contamination and the
identification of potential chemical "hot spots." A total of 945 systems were sampled, of which 466 were
chosen at random. The remaining 479 systems were chosen nonrandomly based on information from
States encouraged to identify locations believed to have a higher than normal probability of VOC
contamination (e.g., locations near landfills or industrial activity).  One sample of finished water was
collected from each  supply at a point near the entrance to the distribution system.

Each of the drinking water surveys was evaluated with respect to the validity of the reported occurrence
data for a number of organic chemicals, including carbon tetrachloride. The evaluations were carried out
by analyzing information about the procedures used for collection and analysis of samples as well as the
quality control protocols used. The  analyzed compounds dealt with in each study were assigned one of
three possible ratings: quantitatively acceptable, qualitatively acceptable (i.e., the substance measured
was carbon tetrachloride), and totally unacceptable. In the case  of carbon tetrachloride, a qualitatively

Occurrence Summary and Use Support Document         320                                     March 2002

-------
acceptable rating was given for data from the CWSS and RWS because of suspected biodegradation of
the samples, which were held unrefrigerated for prolonged periods before analysis, although it is
recognized that biodegradation of carbon tetrachloride is less likely to occur than it is for unsaturated and
aromatic VOCs.  Carbon tetrachloride values in excess of the quantitation limit reported for some
samples in these studies are qualitatively valid and can be taken as minimum values, representative of
samples which probably originally contained carbon tetrachloride at higher concentrations.  In the case of
the NORS, NOMS, NSP, and GWSS, all data were rated both quantitatively and qualitatively acceptable.

4.2.5.2 Groundwater - Federal Surveys

The National Organics Reconnaissance Survey (NORS), the National Organic Monitoring Survey
(NOMS), the National Screening Program for Organics in Drinking Water (NSP), the Community Water
Supply Survey (CWSS), the Rural Water Survey (RWS), and the Groundwater Supply Survey (GWSS)
all contain data concerning the levels of carbon tetrachloride in groundwater supplies from across the
country.

In the NORS, finished water samples were analyzed for 16 groundwater systems. Only one system was
reported to have carbon tetrachloride present, although the level of contamination was below the
minimum quantifiable concentration of 2 |ig/L for that analysis. All other groundwater supplies were
indicated as having "none found," with the minimum quantifiable concentration ranging from 1-2 |ig/L.

Eighteen groundwater systems were analyzed for carbon tetrachloride during Phase I of NOMS (March
to April 1976); only one system contained detectable carbon tetrachloride, at a concentration of 4.0 |ig/L.
None of the  18 systems resampled during Phase II of the study (May to July 1976) contained quantifiable
levels of carbon tetrachloride.  Samples analyzed during Phase III of the study (November 1976 to
January 1977) were negative for all 14 systems examined. The minimum quantifiable limits for carbon
tetrachloride ranged from 1.0-2.0 |ig/L in Phase 1, 0.2-0.4 |ig/L in Phase II, and 0.2-0.4 |ig/L in Phase III.

Twelve groundwater supplies were tested for carbon tetrachloride contamination in the NSP.  Of these  12
systems, two were found to be contaminated with carbon tetrachloride at levels of 0.2 and 0.5 |ig/L. The
quantification limit for carbon tetrachloride was 0.1 |ig/L.

The 1978 CWSS provided information on carbon tetrachloride levels in 316 groundwater systems.  Of
these systems, 6 contained detectable levels of carbon tetrachloride, with values ranging from 0.67-2.15
l-ig/L. The mean value was 1.4 |ig/L with a standard deviation of 0.50 |ig/L; the median value was 1.4
l-ig/L. The minimum quantitation limit for carbon tetrachloride in the CWSS was 0.50 |ig/L.

The RWS examined 207 groundwater supplies for carbon tetrachloride and found 2 to have levels above
the minimum quantification limit of 0.5 |ig/L. The two positive values were 0.8 to 0.88 |ig/L.

In the GWSS, 15 of the 456 randomly chosen water systems serving 25 or more individuals were
contaminated with carbon tetrachloride, at concentrations ranging from 0.21-16 |ig/L.  The three systems
with the highest values were contaminated at 1.7, 2.8, and 16 |ig/L. Ten of the  15 systems with
detections of carbon tetrachloride served populations in excess of 10,000 people.  The average for all
randomly chosen systems was 1.7 |ig/L with a standard deviation of 4.0 |ig/L; the median was 0.37 |ig/L.
Of the 473 nonrandom locations sampled serving 25 or more individuals,  15 were contaminated with
carbon tetrachloride, at concentrations between 0.21-15 |ig/L, the highest values being 1.8, 9.4, and 15
l-ig/L. Six of the 15 systems with detections of carbon tetrachloride served populations in excess of
10,000 people. The average carbon tetrachloride level for the nonrandom systems was 2.2 |ig/L with a
Occurrence Summary and Use Support Document         321                                     March 2002

-------
standard deviation of 4.2 |ig/L; the median value was 0.45 |ig/L. The minimum quantitation limit for
carbon tetrachloride was 0.2 |ig/L.

4.2.5.3 Groundwater - State Data

Eight States (California, Connecticut, Delaware, Indiana, Maine, Massachusetts, New Jersey, and New
York) provided the U.S. Environmental Protection Agency with information concerning carbon
tetrachloride contamination in groundwater supplies.  Analytical results for carbon tetrachloride in
samples from four locations in California ranged from undetectable to greater than 20 |ig/L.  Of 33
locations sampled in Connecticut, carbon tetrachloride levels ranged from undetectable to 66 |ig/L; 70 of
the 85 samples contained less than 5.0 |ig/L. Indiana provided data on eight samples from one location,
all with undetectable carbon tetrachloride. A total of 84 groundwater samples from 57 locations in
Maine and four samples from one location in Massachusetts proved to be free of detectable carbon
tetrachloride. Data on 235 samples from  12 unidentified counties in New Jersey indicated that 190 had
no detectable carbon tetrachloride; another 35 samples contained carbon tetrachloride at less than 1.0
l-ig/L. In another study, 18 locations in New Jersey reported carbon tetrachloride levels ranging from not
detectable to 863 |ig/L. Finally, 20 of 365 groundwater samples taken in New York contained detectable
carbon tetrachloride.

4.2.5.4 Groundwater - Miscellaneous Data

Region V of the U.S. Environmental  Protection Agency conducted a study in early  1975 in which
drinking water samples from 50  cities were analyzed for carbon tetrachloride (USEPA, 1975, as cited in
JRB Associates, 1983). Of 16 groundwater systems sampled, seven contained carbon tetrachloride, at
levels ranging from  1.0-13.0 |ig/L (averaging 4.6 M-g/L). In a survey by Bush et al. (1977, as cited in JRB
Associates, 1983) of New York State drinking water supplies, 7 of 11 groundwater samples from the
Buffalo area and Long Island contained carbon tetrachloride, at levels of less than 0.2 |ig/L.

4.2.5.5 Surface Water - Federal Surveys

The National Organics Reconnaissance Survey (NORS), the National Organic Monitoring Survey
(NOMS), the National Screening Program for Organics in Drinking Water (NSP), the Community Water
Supply Survey (CWSS), and the Rural Water Survey  (RWS) all contain  data concerning the levels of
carbon tetrachloride in surface water supplies from across the country.

In the NORS, finished water from 64 surface water systems were studied, of which nine were found to
have carbon tetrachloride present.  Five of the nine positive systems reported carbon tetrachloride to be
present but below the minimum quantifiable concentration of 2  |ig/L. Of those with quantifiable levels,
two were reported as 2 |ig/L and two as 3 |ig/L. The 55 negative surface water supplies were indicated as
having "none found," with the minimum quantifiable  concentration from 1-2 |ig/L.  It should be noted
that confirmatory  quantitative analyses were performed for 8 surface water supplies in NORS, using a
method able to quantify carbon tetrachloride at 0.05 |ig/L. Three of these eight supplies were originally
reported as having carbon tetrachloride present, but below the minimum quantifiable concentration of 2
l-ig/L; the presence of carbon tetrachloride was confirmed in all  three cases at concentrations ranging
from 0.4 to 0.8 |ig/L. Of the five supplies originally reported as having none found, four were observed
in the confirmatory analysis to have carbon tetrachloride present at levels ranging from 0.2 to 0.3 |ig/L.

In Phase I of the National Organic Monitoring Survey (March to April 1976), water samples from 89
surface water systems were analyzed for carbon tetrachloride. Of these 89 systems, only two were found
to contain carbon tetrachloride, at levels of 1.8 and 2.9 |ig/L.  Eighty-eight systems were sampled during

Occurrence Summary and Use Support Document          322                                     March 2002

-------
the second phase of the survey (May to July 1976) and 10 were found to be contaminated, at levels
ranging from 0.19-10 |ig/L, including the two systems identified as positive in the first phase of the
study.  The average carbon tetrachloride concentration among the positive Phase II samples was 2.4 |ig/L
with a standard deviation of 2.9 |ig/L; the median concentration was 1.4 |ig/L. During the third phase of
the NOMS (November 1976 to January 1977), analyses revealed carbon tetrachloride contamination in
11 out of a total of 87 systems.  The contamination levels for these  systems ranged from 0.2-29 |ig/L.
The mean concentration averaged 4.3 |ig/L with a standard deviation of 8.4 |ig/L; the median level was
1.3 |ig/L.  Only two systems had concentrations above 4.9 |ig/L.  The minimum quantifiable limits for
carbon tetrachloride ranged from 1.0-3.0 |ig/L in Phase I, 0.2-0.4 |ig/L in Phase II, and 0.2-0.4 |ig/L in
Phase III.

Surface water samples from 106 drinking water systems were analyzed for carbon tetrachloride during
the National Screening Program (NSP) between June 1977 and March 1981.  Of these, 36 systems
contained detectable levels of carbon tetrachloride, ranging from 0.1-30 |ig/L. Only two of these systems
were contaminated at levels greater than 1.8 |ig/L (8.4 and 30 |ig/L).  The average concentration among
the 36 positive systems was 1.6 |ig/L with a standard deviation of 5.1 |ig/L; the median level was 0.4
l-ig/L. The quantification limit for the NSP was 0.1 |ig/L.

Of the  105 surface water systems sampled during the Community Water Supply Survey (CWSS), three
contained quantifiable levels of carbon tetrachloride.  These three positive values were 0.52, 0.64, and
0.925.  The mean concentration averaged 0.70 with a standard deviation of 0.21 |ig/L and a median of
0.64 |ig/L. The minimum quantitation limit for carbon tetrachloride in the CWSS was 0.50 |ig/L.

The RWS examined drinking water from 45 surface water supplies; only 1  supply was found to have
carbon tetrachloride present above the minimum quantification limit of 0.5  |ig/L (0.8 |ig/L).

4.2.5.6 Surface Water - State Data

Only two  States, Connecticut and New York,  provided carbon tetrachloride data from surface water
sources. In 11 samples from 9 locations in Connecticut, carbon tetrachloride concentrations ranged from
undetectable to 1.5 |ig/L.  Of 23 samples from three locations in New York, only one was found to
contain detectable carbon tetrachloride (0.1 |ig/L).

4.2.5.7 Surface Water - Miscellaneous Data

In 34 surface water systems analyzed during the EPA Region V study, carbon tetrachloride levels  ranged
from 0.9-26.0 |ig/L (averaging 4.8  |ig/L) (USEPA, 1975, as cited in JRB Associates, 1983). In the Bush
et al. (1977, as cited in JRB Associates, 1983) study, 16 drinking water samples from surface water were
analyzed,  13 from Lake Erie and 3 from Niagara Falls, New York.  Seven of the samples from Lake Erie
were contaminated with carbon tetrachloride,  at 0.2-4.8  |ig/L (averaging 2.4 M-g/L). Two of the samples
from Niagara Falls were contaminated, with levels of 0.3 and 1.1 M-g/L.

4.2.5.8 Projected National Occurrence of Carbon  Tetrachloride in Public Water Supplies

As reported in the JRB Associates  (1983) report, public water systems fall into two major categories with
respect to  water source (surface water and groundwater) and into five size categories and twelve
subcategories according to the number of individuals served. The JRB (1983) report presented estimates
of both the number of drinking water supplies nationally within each of the source/size categories
expected to have carbon tetrachloride present, and of the concentration of carbon tetrachloride expected
to be present in those supplies.

Occurrence  Summary and Use Support Document         323                                     March 2002

-------
The key features of the methodology used and assumptions made to develop the national estimates are
summarized here. The estimates are based on the data from the Federal surveys only.  The State data and
miscellaneous information were not included for several reasons. Generally, these data are from a few
States and were not considered to be geographically representative. There was also a general lack of data
on the population served by systems measured, the type of water sampled, and the methodologies used to
sample, identify, and measure carbon tetrachloride.  Furthermore, since much of these data were
apparently obtained in response to incidents of recognized contamination problems, they may not be
representative of typical conditions existing nationally. However, while these data were not used for
computing the national projections, they do provide a valuable and necessary perspective for evaluating
those projections, especially with respect to the projected high levels of contamination of carbon
tetrachloride.

The Federal survey data from the NORS, NOMS, NSP, CWSS, RWS, and GWSS were pooled together
for developing the national projections. It was assumed in combining these surveys that the resulting
data base would be representative of the nation's water supplies.  In the case of the GWSS data, both the
random and nonrandom samples were included in the projections because a test of the GWSS data
showed no statistically significant difference in the frequency of occurrence  of positive values or the
mean of the  positive values of carbon tetrachloride.

Ideally, adequate data would be available to develop the national projections separately for each of the
twelve system size categories within the groundwater and surface water groups; however, the available
data were too limited for this. JRB (1983) consolidated some of the size categories to have sufficient
data for developing the projections. In consolidating data from various size categories, consideration was
given to the  potential for there being statistically significant differences in the frequency of occurrence of
carbon tetrachloride as a function of system size.  The consolidation of size categories therefore involved
a balancing of the need to group size categories together to have an adequate data base for developing the
national projections against the need to treat size categories separately in order to preserve the influence
of system size as a determinant of contamination potential.  The consolidation of size categories also took
into account EPA's classification of systems into the five major groups as very small (25-500), small
(501-3,300), medium (3,301-10,000), large (10,001-100,000), and very large (> 100,000) (Kuzmack,
1983, as cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national level were tested and
an appropriate model selected.  In the case of carbon tetrachloride, the multinominal method was used.
The frequency of contamination of groundwater and surface water systems at various concentrations was
determined for each consolidated size category. For completing the national estimates, it was assumed
that the frequency of contamination observed for each consolidated category was directly applicable to
each of the system sizes comprising it.

4.2.5.8.1 Groundwater Systems

Data are available for a total of 1,466 supplies from the combined surveys (NORS, NOMS, NSP, CWSS,
RWS, and GWSS).  Of these, 42 supplies were reported to have carbon tetrachloride present, at
concentrations ranging from 0.2 |ig/L to 16 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which carbon tetrachloride was not found, 0.5 |ig/L was  selected as the common minimum quantifiable
concentration for the combined survey data. That is, quantitative projections are made of supplies at
several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either
no carbon tetrachloride or levels below 0.5 |ig/L can be determined. Although some data indicate the

Occurrence Summary and Use Support Document          324                                     March 2002

-------
presence of carbon tetrachloride in groundwater supplies at levels < 0.5 |ig/L, it is not possible to
determine the proportion of supplies that have carbon tetrachloride present and the proportion that are
actually free of carbon tetrachloride contamination.

Of the 1,424 supplies reporting no carbon tetrachloride to be present, 1,402 were assumed to have
maximum possible levels of < 0.5 |ig/L based on the minimum quantifiable concentrations reported for
the various surveys (885 of these had maximum levels of < 0.2 M-g/L).  The other 22 supplies reporting no
carbon tetrachloride to be present had maximum possible levels ranging from approximately 0.6 |ig/L to
2 |ig/L. It is assumed, based on overall distribution of values, that carbon tetrachloride if present in the
22 supplies is so at a concentration of < 0.5 |ig/L, although a rigorous conservative argument could be
made for assuming a level equal to the maximum possible value.  The impact of this assumption is
notable both in terms of the national projection of groundwater systems above 0.5 |ig/L and in terms of
the population exposed.

Twenty-three of the 1,466 supplies examined had measured values of carbon tetrachloride  > 0.5 |ig/L.
When the twelve size categories were consolidated into the five major EPA groupings, there was an
apparent relationship between the frequency of values > 0.5 |ig/L and system size:
Very small
Small
Medium
Large
Very large
Overall
0.2%
1.9%
1.4%
1.9%
8.2%
1.6%
(1/423)
(8/417)
(3/214)
(7/363)
(4/49)
(23/1,466)
A test for statistical significance revealed that at the a = 0.05 level, a statistically significant difference
was found between the very small and small categories, as well as between the large and very large
categories.  No further consolidation of these five categories was done. As noted previously, the
frequency of occurrence of carbon tetrachloride at various concentrations was determined for the
consolidated groups and then applied to the number of supplies nationally within each of the size
categories comprising each group.

About 335 groundwater  supplies (range of 124-545), approximately 0.7% of the total groundwater
supplies in the United States, are expected to have carbon tetrachloride at levels of > 0.5 |ig/L; the
remaining 48,123 supplies have either no carbon tetrachloride or levels < 0.5 |ig/L.

It is estimated that 110 supplies (range of 0-277) are expected to have carbon tetrachloride levels > 5
l-ig/L, while  107 supplies (range of 0-273)  are expected to have levels > 10 |ig/L. Most of the supplies
with high carbon tetrachloride levels are expected to be in the smaller size categories.  Although, as noted
previously, the frequency of carbon tetrachloride occurrence is greater in the very large system size, the
number of systems affected nationally is greater for the small sizes because there are many more small
systems in existence.

It is interesting to note the impact on the national projections of the assumption made that the 22 supplies
with undetected but maximum potential values of 0.6-2 |ig/L had < 0.5 |ig/L.  Had it been assumed that
carbon tetrachloride was present in those supplies at their maximum possible values, the national
projection of supplies with carbon tetrachloride levels > 0.5 |ig/L would have increased to 417 (200-633)
with no differences in levels > 5 |ig/L. These differences in the 0.5-5 |ig/L range would be found in
systems serving > 500 people.
Occurrence Summary and Use Support Document          325                                      March 2002

-------
4.2.5.8.2 Surface Water Systems

Data are available for a total of 312 surface water supplies.  Of these, 51 supplies were reported by JRB
Associates (1983) to have carbon tetrachloride present at concentrations ranging from 0.1 |ig/L to 30
l-ig/L, although this high value is the only one reported above 10 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which carbon tetrachloride was not found, 0.5 |ig/L was selected as the common minimum quantifiable
concentration for the combined survey data. That is, quantitative projections are made of supplies at
several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either
no carbon tetrachloride or levels below 0.5 |ig/L can be determined. Although some data indicate the
presence of carbon tetrachloride in surface water supplies at levels < 0.5 |ig/L, it is not possible to
determine the proportion that have carbon tetrachloride present and the proportion that are free of carbon
tetrachloride contamination.

Of the 261 supplies reporting no carbon tetrachloride to be present, 186 had maximum possible levels of
< 0.5  |ig/L based on the minimum quantifiable concentrations reported for the various surveys. The other
75 supplies reporting no carbon tetrachloride to be present had maximum possible levels ranging from
0.53 |ig/L to 2 |ig/L. It is assumed, based on the overall distribution of values, that carbon tetrachloride if
present in these 75 supplies is so at a concentration of < 0.5 |ig/L, although a rigorous conservative
argument could be made for assuming a level equal to the maximum possible value.  As will be noted
further below, the difference between these alternatives is considerable for the estimate of the number of
surface water supplies with carbon tetrachloride > 0.5 |ig/L; also the impact on the estimated population
exposed to carbon tetrachloride at levels > 0.5 |ig/L in surface water supplies is very large.

Forty of the  312 supplies examined had measured values of carbon tetrachloride  > 0.5 |ig/L.  When the
twelve size categories were consolidated into the five major EPA groupings, there was a general trend  in
the frequency of values > 0.5 |ig/L as a function of size:
Very small
Small
Medium
Large
Very large
Overall
0% (0/20)
3.3% (2/60)
2.4% (1/42)
12.2% (9/74)
24.1% (28/116)
12.8% (40/312)
A test for statistical significance revealed that at the a = 0.05 level, the very small, small, and medium
groups were not different from one another and that the large and very large groups are not different;
however, the combined very small, small, and medium groups and the combined large and very large
groups are different.  These two consolidated categories were selected for developing the national
estimates:

                               Very small/small/medium (25-10,000)
                                    Large/very large (> 10,000)

As noted previously, the frequency of occurrence of carbon tetrachloride at various concentrations was
determined for the consolidated groups and then applied to the number of supplies nationally within each
of the size categories comprising each group.
Occurrence Summary and Use Support Document          326                                     March 2002

-------
About 575 surface water supplies (range of 299-850), approximately 5.10% of the total surface water
systems in the United States, are expected to have carbon tetrachloride at levels > 0.5 |ig/L; the
remaining 10,627 supplies have either no carbon tetrachloride or levels < 0.5 |ig/L. It is estimated that 26
surface water supplies (range of 0-55) will have levels > 5 |ig/L and 9 (range of 0-26) will have levels > 2
l-ig/L. None are projected to have levels above 30 |ig/L.

JRB Associates (1983) reported that a notable impact on the national projections by the assumption made
that the 75 supplies with undetected but maximum potential values of 0.53-2 |ig/L had < 0.5 |ig/L.  Had it
been assumed that carbon tetrachloride was present in these supplies at their maximum possible values,
the national projections of supplies with carbon tetrachloride  levels > 0.5 |ig/L would be 1,819 (range of
1,326-2,311) supplies. However, there would be no difference in the projected number of surface water
supplies with levels > 5 |ig/L.

4.2.6 Conclusion

With few exceptions, production of carbon tetrachloride has been eliminated in developed nations,
including the U.S. The major use of carbon tetrachloride in the early 1990s was in the production of
chlorofluorocarbons, such as dichlorodifluoromethane and trichlorofluoromethane. According to the
latest figures available before the production ban, the U.S. produced between 573 and 761 million
pounds each year from 1981 to 1988.  Industrial releases of carbon tetrachloride have been reported to
TRI since 1988 in 38 States, Puerto Rico, and the Virgin Islands. The Stage 2 analysis, based on the 16-
State cross-section, estimated that approximately 0.0159% of combined ground water and surface water
systems serving 0.0316% of the population had estimated mean concentrations of carbon tetrachloride
greater than the MCL of 0.005 mg/L.  Based on this estimate, approximately 10 PWSs nationally serving
about 67,400 people are expected to have estimated mean concentrations of carbon tetrachloride greater
than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors.  Nationally, TRI releases have been reported for carbon tetrachloride from 38 States,
including 12 of 16 cross-section States. Except for certain exemptions, carbon tetrachloride is no longer
produced or used in any State, including the cross-section States. The cross-section should adequately
represent the occurrence of carbon tetrachloride on a national scale based upon the use, production, and
release patterns of the  16-State cross-section in relation to the patterns observed for all 50 States.

4.2.7 References

Agency for Toxic Substances and Disease Registry (ATSDR).  1994.  Toxicological Profile for Carbon
       Tetrachloride.  U.S. Department of Health and Human Services, Public Health Service.  229 pp.
       + Appendices.  Available on the Internet at:  http://www.atsdr.cdc.gov/toxprofiles/tp30.pdf

Brass, H.J., M.A. Feige, T. Halloran, J.W. Mello, D. Munch,  and R.F. Thomas.  1977. The National
       Organic Monitoring Survey: A sampling and analysis for purgeable organic compounds.
       Drinking water quality enhancement through source protection.  R.B. Pojasek (ed.). Ann Arbor,
       MI: Ann Arbor Science, pp. 393-416.

Brass, H.J.  1981. Rural Water Survey organics data.  Memorandum of March 17, 1981 to Hugh Hanson,
       Chief, Science and Technology Branch, Office of Drinking Water, and David Schnare, Office of
       Drinking Water, U.S. Environmental Protection Agency, Washington, DC.
Occurrence Summary and Use Support Document         327                                     March 2002

-------
Bush, B., R.S. Narang, and S. Syrotynski.  1977. Screening for halo organics in New York State drinking
       water.  Bull Envir. Contam. Toxicol 18(4): October 1977.

JRB Associates.  1983. Occurrence of Carbon Tetrachloride in Drinking Water, Food and Air - DRAFT.
       Draft report submitted to EPA for review November 17, 1983.

Kuzmack, A.M.  1983. Memorandum: Characterization of the  water supply industry (FY82).
       Washington, DC: Office of Water, U.S. Environmental Protection Agency. May 16, 1983.

McKenna and Cuneo, L.L.P.  2001. Montreal Protocol to limit the use of the 'transformation' exemption
       from the  ODS Ban. McKenna and Cuneo, L.L.P. Client Bulletin. Available on the Internet at:
       http ://www .mckennacuneo.com/articles/archive/ENV04019970620 .html

Mello, Wayne. 1983. Personal communication between Wayne Mello, Technical Support Division,
       Office  of Drinking Water, U.S. Environmental Protection Agency, and author of JRB Associates,
       1983, March 10, 1983.

Symons, J.M.,  T.A. Bellar, and J.K. Carswell et al.  1975. National Organics Reconnaissance Survey for
       Halogenated Organics.  J. Amer. Water Works Assoc. 667(11): 634-647.

USEPA.  1975. Sources of organic, metal, and other inorganic parameter concentrations in selected
       Region Vdrinking water supplies. Chicago IL: Water Division and Surveillance and Analyses
       Division, USEPA.

USEPA. 2000.  TRIExplorer: Trends. Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National  Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Westrick, J.J., J.W. Mello, and R.F. Thomas.  1983.  The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data. EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         328                                    March 2002

-------
4.3    1,4-Dichlorobenzene
Table of Contents

4.3.1  Introduction, Use and Production  	  330
4.3.2  Environmental Release  	  331
4.3.3  Ambient Occurrence 	  332
4.3.4  Drinking Water Occurrence Based on the 16-State Cross-Section Data	  332
4.3.5  Additional Drinking Water Occurrence Data  	  337
4.3.6  Conclusion	  342
4.3.7  References  	  343
Tables and Figures

Table 4.3-1: Facilities that Manufacture or Process 1,4-Dichlorobenzene	  330

Table 4.3-2: Environmental Releases (in pounds) for 1,4-Dichlorobenzene
       in the United States, 1988-1999  	  331

Table 4.3-3: Stage 1 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section -
       Systems	  333

Table 4.3-4: Stage 1 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section -
       Population	  334

Table 4.3-5: Stage 2 Estimated 1,4-Dichlorobenzene Occurrence Based on 16-State
       Cross-Section  - Systems	  335

Table 4.3-6: Stage 2 Estimated 1,4-Dichlorobenzene Occurrence Based on 16-State
       Cross-Section  - Population	  335

Table 4.3-7: Estimated National 1,4-Dichlorobenzene Occurrence - Systems and
       Population Served	  336
Occurrence Summary and Use Support Document          329                                     March 2002

-------
4.3.1  Introduction, Use and Production

1,4-Dichlorobenzene (chemical formula C6H4C12) is usually called para-DCB or p-DCB; however, there
are about 20 additional names for it, including para crystals and paracide.  It is also called paramoth
because it is one of two chemicals commonly used to make mothballs. At room temperature, 1,4-
dichlorobenzene is a white solid with a strong odor that one would probably recognize as the smell of
mothballs.  Most of the  1,4-dichlorobenzene that is released to the general environment is present as a
vapor. 1,4-Dichlorobenzene does not occur naturally, but is produced by chemical companies to make
products for home use and other chemicals such as resins (ATSDR, 1998). 1,4-Dichlorobenzene can
burn, but not easily.  Its major routes of entry to drinking water are a consequence of industrial activity.
It is expected that discharges to the environment during production of the dichlorobenzenes and the use
of a variety of solvents, pesticides, and deodorants containing dichlorobenzene are primary causes of
dichlorobenzene contamination of drinking water (JRB Associates, 1983).

For over 20 years, 1,4-dichlorobenzene has been primarily used as a space deodorant for toilets and
refuse containers, and as a fumigant for control of moths, molds, and mildews. When exposed to air, it
slowly changes from a solid to a vapor. It is the vapor that acts as a deodorizer or insect killer. Most
people recognize the odor as the smell of mothballs, and can smell 1,4-dichlorobenzene in the air at very
low levels (ATSDR, 1998).

About half to a third of the 1,4-dichlorobenzene produced is used as a deodorant and fumigant, 37% is
exported, 27% is used to produce polyphenylene sulfide resin, and another 10% is used as an
intermediate in the production of other chemicals, such as 1,2,4-trichlorobenzene (ATSDR, 1998) and
2,5-dichloroaniline (NTP, 1991). 1,4-Dichlorobenzene is also used in the control of certain tree-boring
insects and ants; in the control of blue mold in tobacco seed beds (ATSDR, 1998); as a disintegrating
paste for molding concrete and stoneware; as a lubricant (NTP, 1991); and in the manufacture of plastics,
dyes, and pharmaceuticals (USEPA, 2001).

Estimates for the production of 1,4-dichlorobenzene in the U.S. are as follows: 1972, 77.2 million
pounds; 1975, 45.9 million pounds;  1977, 16-116 million pounds; 1981, 15 million pounds (ATSDR,
1998). (More recent data is unavailable.) The three primary manufacturers of 1,4-dichlorobenzene are
Monsanto Company in Sauget, Illinois; PPG Industries, Inc., in Natrium, West Virginia; and Standard
Chlorine of Delaware, Inc., in Delaware City, Delaware; with annual production capacities of 33, 36, and
75 million pounds, respectively (ATSDR, 1998).

Table 4.3-1 lists the facilities in each State that manufacture and process 1,4-dichlorobenzene, the
intended uses of the product, and the range of maximum amounts that are stored on site. The data were
derived from the Toxics Release Inventory (TRI) of EPA (ATSDR, 1998).
Table 4.3-1:  Facilities that Manufacture or Process 1,4-Dichlorobenzene
Facility
Bay State Sterling
Bay State Sterling
Caroline Solite Corp.
Coughlan Prods. Corp.
Coughlan Prods. Corp.
Coughlan Prods. Corp.
Crest Prods. Inc.
Dow Chemical Co.
Fortran Ind.
Location "
North Manchester, IN
Westborough, MA
Norwood, NC
Clifton, NJ
Wayne, NJ
Paterson, NJ
Oldsmar, FL
Plaquemine, LA
Wilmington, NC
Range of maximum
amount on site in pounds
1,000-9,999
1,000-9,999
1,000-9,999
10,000-99,999
10,000-99,999
10,000-99,999
10,000-99,999
1,000-9,999
100,000-999,999
Activities and uses
Manufacturing Aid
Manufacturing Aid
Ancillary/Other Use
Formulation Component
Formulation Component
Formulation Component
Formulation Component
Produce, Impurity
Reactant
Occurrence Summary and Use Support Document         330                                    March 2002

-------
Facility
Fresh Prods. Inc.
Fuller Brash Co.
Heartland Cement Co.
Hospital Specialty Co.
I. Schneid Inc.
Monsanto
Nipa Hardwicke Inc.
Phillips Chemical Co.
Phillips Research Center
PPG Ind. Inc.
Standard Chlorine of
Willert Home Prods.
Location "
Toledo, OH
Great Bend, KS
Independence, KS
Cleveland, OH
Atlanta, GA
Sauget, IL
Elgin, SC
Borger, TX
Bartlesville, OK
New Martinsville, WV
Delaware City, DE
Saint Louis, MO
Range of maximum
amount on site in pounds
10,000-99,999
100,000-999,999
100-999
100,000-999,999
10,000-99,999
1,000,000-9,999,999
10,000-99,999
100,000-999,999
1,000-9,999
1,000,000-9,999,999
10,000,000-49,999,999
100,000-999,999
Activities and uses
Import, Sale/Dist, Repackaging, Ancillary/Other
Reactant
Ancillary/Other Use
Article Component
Article Component, Repackaging
Produce, Sale/Distribution
Reactant
Reactant
Reactant
Produce, Sale/Distribution
Produce, On-site Use/Processing, Sale/Dist.,
Article Component
aPost office State abbreviations used
Source: ATSDR, 1998 compilation of TRI961998 data


4.3.2 Environmental Release

1,4-Dichlorobenzene is listed as a Toxics Release Inventory (TRI) chemical. Table 4.3-2 illustrates the
environmental releases for 1,4-dichlorobenzene from 1988 to 1999.  (1,4-Dichlorobenzene data are only
available for these years.) Air emissions constitute most of the on-site releases, with a steady decrease
that has moderated in recent years. The decrease in air emissions has been the primary contributor to
decreasing levels of total on- and off-site releases.  Surface water discharges have slightly decreased from
the high in 1989, although levels have been rising since 1995. Underground injection has remained
relatively constant, except for a considerable increase from  1997 to 1999. Releases to  land (such as spills
or leaks within the boundaries of the reporting facility) and off-site releases (including metals or metal
compounds transferred off-site) have no discernable trend in the amounts released, as they range from
almost zero to a few thousand pounds in any given year.  These TRI data for 1,4-dichlorobenzene were
reported from 22 States, with six States reporting every year (USEPA, 2000).  Nine of the 22 States are in
the 16-State cross-section (used for analyses of 1,4-dichlorobenzene occurrence in drinking water; see
Section 4.3.4).  (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.3-2: Environmental Releases (in pounds) for 1,4-Dichlorobenzene in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
On-Site Releases
Air Emissions
178,210
181,899
262,266
236,502
242,372
257,211
357,891
337,946
344,254
Surface Water
Discharges
1,880
1,706
1,728
1,881
1,287
1,595
1,265
2,021
2,146
Underground
Injection
7,300
3,100
2,000
2,000
-
2,000
2,000
2,000
2,000
Releases
to Land
1,370
460
1,960
480
3,100
1,100
1,112
622
420
Off-Site Releases
0
-
289
-
3,328
-
213
751
770
Total On- &
Off-site
Releases
188,760
187,165
268,243
240,863
250,087
261,906
362,481
343,340
349,590
Occurrence Summary and Use Support Document
March 2002

-------
Year
1990
1989
1988
On-Site Releases
Air Emissions
818,133
1,592,229
1,891,419
Surface Water
Discharges
3,912
6,621
6,153
Underground
Injection
255
250
4,000
Releases
to Land
38
250
1,300
Off-Site Releases
4,006
1
750
Total On- &
Off-site
Releases
826,344
1,599,351
1,903,622
 Source: USEPA, 2000
4.3.3  Ambient Occurrence

1,4-Dichlorobenzene was detected in 6 out of 330 wells (1.8%) in urban areas of the local, State, and
federal data set compiled by NAWQA. The minimum and maximum concentrations detected were 0.3
l-ig/L and 56 |ig/L, respectively.  The median value of detection concentrations was 1 M-g/L. 1,4-
Dichlorobenzene was also detected in 4 of the 2,434 wells (0.16%) with analysis in rural areas.  The
minimum and maximum concentrations detected were 0.8 |ig/L and 6.7 |ig/L, respectively. The median
value of detection concentrations was 1.3 |ig/L. These data (urban and rural) represent untreated ambient
ground water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).

4.3.3.1  Additional Ambient Occurrence Data

A summary document entitled "Dichlorobenzenes: Occurrence in Drinking Water, Food, and Air" (JRB
Associates, 1983), was previously prepared for past USEPA assessments of 1,4-dichlorobenzene.
However, no information on the ambient occurrence of 1,4-dichlorobenzene was included in that
document.  (The document did include information regarding 1,4-dichlorobenzene occurrence in drinking
water, which is discussed in Section 4.3.5  of this report.)

4.3.4  Drinking Water Occurrence Based on the 16-State Cross-Section Data

The analysis of 1,4-dichlorobenzene occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative to several concentration thresholds of interest: 0.075 mg/L; 0.005 mg/L; and 0.0005
mg/L.

Fourteen of the sixteen cross-section State data sets contained occurrence data for 1,4-dichlorobenzene.
(There were no 1,4-dichlorobenzene data from California or Montana.) These data represent more than
123,000 analytical results from approximately  19,000 PWSs during the period from 1984 to 1998 (with
most analytical results from 1992 to 1997). The number of sample results and PWSs vary by State,
although the State data sets have been reviewed and checked to ensure adequacy of coverage and
completeness.  The overall modal detection limit for 1,4-dichlorobenzene in the  16  cross-section States is
equal to 0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to Section  1.3.5 of
this report.)
Occurrence Summary and Use Support Document
March 2002

-------
4.3.4.1 Stage 1 Analysis Occurrence Findings

Table 4.3-3 illustrates the Stage 1 analysis of 1,4-dichlorobenzene in drinking water for the public water
systems in the 16-State cross-section. Based on the 16-State cross-section data, no ground water or
surface water PWSs had any analytical results exceeding the MCL (0.075 mg/L). Only 0.0844% of all
PWSs in the 16 States (a total of 16 systems) had any analytical results greater than 0.005 mg/L.
Approximately  1.32% of PWSs (250 systems) had analytical detections of 1,4-dichlorobenzene greater
than 0.0005 mg/L.

Approximately  0.0908% of ground water PWSs (16 systems) had any analytical results exceeding the
0.005 mg/L, compared to 0% of surface water systems.  Over 1.2% of ground water PWSs (217 systems)
had at least one analytical result greater than 0.0005 mg/L. About 2.45% of surface water systems (33
systems) had any analytical results of 1,4-dichlorobenzene greater than 0.0005 mg/L.
Table 4.3-3:  Stage 1 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.075
0.005
0.0005
Percent of Systems
Exceeding Threshold
0.000%
0.0908%
1.23%
Number of Systems
Exceeding Threshold
0
16
217

Surface Water
0.075
0.0050
0.0005
0.000%
0.000%
2.45%
0
0
33

Combined Ground &
Surface Water
0.075
0.005
0.0005
0.000%
0.0844%
1.32%
0
16
250
Reviewing 1,4-dichlorobenzene occurrence in the 16 cross-section States by PWS population served
(Table 4.3-4) shows that no ground water or surface water systems had any analytical detections greater
than the MCL (0.075 mg/L).  Approximately 0.230% of the population (almost 168,000 people) was
served by systems with at least one analytical results greater than 0.005 mg/L. Approximately 2.53% of
the population (over 1.8 million people) was served by PWSs with analytical detections of 1,4-
dichlorobenzene greater than 0.0005 mg/L.

The percentage  of population served by ground water systems in the 16 States with analytical results
greater than 0.005 mg/L was equal to 0.464% (about 168,000 people). When evaluated relative to 0.0005
mg/L, the percent of population exposed by ground water systems was equal to 3.16% (over 1.1 million
people).  The percentage of population served by surface water systems with exceedances of 0.0005
mg/L was equal to  1.90% (700,400 people).
Occurrence Summary and Use Support Document
March 2002

-------
Table 4.3-4:  Stage 1 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.075
0.005
0.0005
Percent of Population
Served by Systems
Exceeding Threshold
0.000%
0.464%
3.16%
Total Population Served
by Systems Exceeding
Threshold
0
167,800
1,143,000

Surface Water
0.075
0.005
0.0005
0.000%
0.000%
1.90%
0
0
700,400

Combined Ground &
Surface Water
0.075
0.005
0.0005
0.000%
0.230%
2.53%
0
167,800
1,843,400
4.3.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.3-5 and 4.3 -
6. The statistically generated best estimate values, as well as the ranges around the best estimate value,
are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.
For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology
and Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations -
DRAFT (USEPA, 2002)).

No ground water or surface water PWSs in the 16 States had an estimated mean concentration of 1,4-
dichlorobenzene exceeding 0.075 mg/L (the current MCL).  The percentage of PWS in the 16 States with
estimated mean concentration values of 1,4-dichlorobenzene greater than 0.005 mg/L was equal to
0.000253%. The percentage of PWSs with estimated mean concentrations exceeding 0.0005 mg/L (the
modal detection limit) was about 0.114% PWSs (22 systems) in the 16 States.

A greater proportion of ground water systems, as compared to surface water systems, were estimated to
exceed the modal detection limit. Approximately 0.000273% of ground water systems in the 16 States
had estimated mean concentrations greater than 0.005 mg/L, as compared to 0% of surface water
systems.  About 0.117% of ground water systems (an estimated 21 systems in the 16 States) had
estimated mean concentrations greater than 0.0005 mg/L. This compares with about 0.0746% of the
surface water systems (about 1 system) with estimated mean concentrations greater than 0.0005.
Occurrence Summary and Use Support Document
March 2002

-------
Table 4.3-5: Stage 2 Estimated 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.075
0.005
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000273%
0.117%
Range
0.000% - 0.000%
0.000% - 0.00568%
0.0738% -0.170%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
21
Range
0-0
0-1
13-30

Surface Water
0.075
0.005
0.0005
0.000%
0.000%
0.0746%
0.000% - 0.000%
0.000% - 0.000%
0.000% - 0.223%
0
0
1
0-0
0-0
0-3

Combined Ground
& Surface Water
0.075
0.005
0.0005
0.0000%
0.000253%
0.114%
0.000% - 0.000%
0.000% - 0.00527%
0.0738% -0.164%
0
0
22
0-0
0-1
14-31
Reviewing 1,4-dichlorobenzene occurrence by PWS population served (Table 4.3-6) shows that
approximately 0.00000206% of the population in the 16 States was served by PWSs with mean
concentrations greater than 0.005 mg/L. Approximately 0.0379% of population served by all PWSs in
the 16 States (an estimate of approximately 27,700 people) was potentially exposed to 1,4-
dichlorobenzene levels above 0.0005 mg/L.  When evaluated relative to a threshold of 0.075 mg/L, the
percent of population exposed was equal to 0%.

For ground water systems, about 0.00000416% of the population served by ground water systems in the
16 States was exposed to 1,4-dichlorobenzene levels above 0.005 mg/L.  An estimated 0.0505% of the
population (about  18,300 people) was served by systems in the 16 States whose mean concentration
value exceeded 0.0005 mg/L.

No surface water systems had mean concentrations that exceeded 0.005 mg/L. Approximately 0.0255%
of the 16-State population was served by surface water PWSs (about 9,400 people) with estimated mean
concentrations of 1,4-dichlorobenzene above 0.0005 mg/L.
Table 4.3-6: Stage 2 Estimated 1,4-Dichlorobenzene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.075
0.005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.00000416%
Range
0.000% - 0.000%
0.000% -0.0000691%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
Range
0-0
0-<100
Occurrence Summary and Use Support Document
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.0505%
Range
0.00580% -0.1 87%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
18,300
Range
2,100 - 67,600

Surface Water
0.075
0.005
0.0005
0.000%
0.000%
0.0255%
0.000% - 0.000%
0.000% - 0.000%
0.000% -0.193%
0
0
9,400
0-0
0-0
0 - 70,900

Combined Ground
& Surface Water
0.075
0.005
0.0005
0.000%
0.00000206%
0.0379%
0.000% - 0.000%
0.000% - 0.0000343%
0.00390% -0.167%
0
0
27,700
0-0
0-<100
2,800-121,800
4.3.4.3 Estimated National Occurrence

As illustrated in Table 4.3-7, the Stage 2 analysis estimated zero PWSs nationally were estimated to have
mean concentration values of 1,4-dichlorobenzene greater than 0.075 mg/L or 0.005 mg/L.
Approximately 74 systems serving about 80,700 people nationally were estimated to have mean 1,4-
dichlorobenzene concentrations greater than 0.0005 mg/L. An estimated 69 ground water PWSs serving
about 43,300 people nationally had mean concentrations greater than 0.0005 mg/L.  Approximately 4
surface water systems serving  32,400 people was estimated to have mean concentrations of 1,4-
dichlorobenzene above 0.0005 mg/L. (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)
Table 4.3-7:  Estimated National 1,4-Dichlorobenzene Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.075
0.005
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
69
Range
0-0
0-3
44-101
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
43,300
Range
0-0
0-<100
5,000 - 160,000

Surface Water
0.075
0.005
0.0005
0
0
4
0-0
0-0
0-12
0
0
32,400
0-0
0-0
0 - 245,400
Occurrence Summary and Use Support Document
March 2002

-------
Source Water Type
Threshold
(mg/L)
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.075
0.005
0.0005
0
0
74
0-0
0-3
48-106
0
0
80,700
0-0
0-<100
8,300 - 355,500
4.3.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of 1,4-dichlorobenzene in drinking water are
also reviewed. Previously compiled occurrence information on one of the three isomers of
dichlorobenzene, para-dichlorobenzene (1,4-dichlorobenzene), from an OGWDW summary document
entitled "Dichlorobenzenes:  Occurrence in Drinking Water, Food, and Air" (JRB Associates, 1983), is
presented in the following section.  This variety of studies and information are presented regarding levels
of 1,4-dichlorobenzene in drinking water, with the scope of the reviewed studies ranging from national to
regional. Note that none  of the studies presented in the following section provide the quantitative
analytical results or comprehensive coverage that would enable direct comparison to the occurrence
findings estimated with the cross-section occurrence data presented in Section 4.3.4. These additional
studies, however, do enable a broader assessment of the Stage 2 occurrence estimates presented for this
Six-Year Review. All the following information in Section 4.3.5 is taken directly from
"Dichlorobenzenes: Occurrence in Drinking Water, Food, and Air" (JRB Associates, 1983).

The JRB Associates (1983) report found three major types of data available that were potentially useful
for describing the occurrence of dichlorobenzene in the nation's public drinking water supplies.  First,
there are several Federal surveys in which a number of public water supplies from throughout the U.S.
were selected for analysis of chemical contamination, including dichlorobenzene. Second, data are
available from State surveys and from State  investigations of specific incidents of known or suspected
contamination of a supply. Third, there are miscellaneous published data which, like some of the State
data, tend to be from studies  in response to suspected contamination of specific sites. For accomplishing
the basic objectives of this study, namely to  estimate the number of public water supplies nationally
within the various source and size categories contaminated with dichlorobenzene, the distribution of
dichlorobenzene concentrations in those supplies, and the number of individuals exposed to those
concentrations, it was determined that the Federal survey data provides the most suitable data base.  The
State and miscellaneous data tend to be poorly described with respect to the source and size categories of
the supplies examined and the sampling and analysis methods used for determining contaminant levels.
The lack of source and  system size information precludes using the data for estimating levels in public
water supplies of similar characteristics. The absence of details on sampling and analysis methods
precludes evaluating those data for their qualitative and quantitative reliability. Also, because much of
the State and miscellaneous data are from investigations in response to incidents of known or suspected
contamination (e.g., spills), they were judged to be not representative of contaminant levels in the
nation's water supplies  in general.  Although they are not used with the Federal data for the purpose of
estimating contamination levels nationally, the available State and miscellaneous data are presented here
to provide some  additional perspective on dichlorobenzene occurrence in drinking water.
Occurrence Summary and Use Support Document
March 2002

-------
Data are presented only on drinking water samples taken from a consumer's tap (i.e., distribution water
samples) or on treated water samples taken at the water supply (i.e., finished water samples) because
these are considered to be most representative of the water consumed by the public. No data on raw (i.e.,
untreated) water are presented.  It is recognized that for some groundwater supplies where no treatment
of the water occurs, samples identified as raw may be representative of water consumed by the users of
the supply. However, it was generally not possible to differentiate between those groundwater supplies
that do and do not treat raw water from the available survey data.

4.3.5.1 Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Three Federal drinking water surveys provide data on dichlorobenzene: the National Organic Monitoring
Survey (NOMS), the National Screening Program for Organics in Drinking Water (NSP), and the
Groundwater Supply Survey (GWSS). The terms used in this report are those used in the individual
surveys, recognizing that they may not always correspond to strict technical definitions.

The National Organic Monitoring Survey (NOMS) was conducted to identify contaminant sources, to
determine the frequency of occurrence of specific drinking water contaminants, and to provide data for
the establishment of maximum contaminant levels  (MCLs) for various-organic compounds in drinking
water (Brass et al., 1977, as cited in JRB Associates, 1983).  The NOMS was conducted in three phases:
March-April 1976, May-July 1976, and November 1976-January 1977. Finished drinking water samples
from 113 communities were analyzed for 21 different compounds. Of the 113 community supplies
sampled, 18 had groundwater sources, 91 had surface water sources, and 4 had a mixed
groundwater/surface water source.

The analytical results of the NOMS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water. Additional information on the locations and source of the supplies,
and on the populations served by the supplies in the NOMS were provided by Wayne Mello (1983) at
EPA's Technical Support Division, Office of Drinking Water. A single value for dichlorobenzenes was
reported for each supply studied in the NOMS.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March  1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic chemical  contaminants.  Data are available  only for p-
dichlorobenzene (1,4-dichlorobenzene) on finished water samples from 12 groundwater and 103 surface
water supplies.

The Groundwater Supply  Survey (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al., 1983, as cited in JRB Associates, 1983). It was hoped that this study would
stimulate State efforts toward the detection and control of groundwater contamination and the
identification of potential chemical "hot spots." A total of 945 systems were sampled, of which 466 were
chosen at random.  The remaining 479 systems were chosen nonrandomly based on information from
States encouraged to identify locations believed to have a higher than normal probability of VOC
contamination (e.g., locations near landfills or industrial activity).

Each of the drinking water surveys was evaluated with respect to the validity of the reported occurrence
data for a number of organic chemicals, including dichlorobenzene.  The evaluations were carried out by
analyzing information about the procedures used for collection and analysis of samples as well as the
quality control protocols used. The analyzed compounds dealt with in each study were assigned one of
three possible ratings: quantitatively acceptable, qualitatively acceptable (i.e., the substance measured

Occurrence Summary and Use Support Document         338                                     March 2002

-------
was dichlorobenzene), and totally unacceptable. In the case of dichlorobenzene, a qualitatively
acceptable rating was given for data from the NOMS (Phases II and III) because of suspected
biodegradation of the samples, which were held unrefrigerated for prolonged periods before analysis.
Dichlorobenzene values in excess of the quantitation limit reported for some samples in these studies are
qualitatively valid and can be taken as minimum values, representative of samples which probably
originally contained dichlorobenzene at higher concentrations. In the case of the NOMS (Phase I), NSP,
and GWSS, all data were rated both quantitatively and qualitatively acceptable.

4.3.5.2 Groundwater - Federal Surveys

Three Federal surveys contain data concerning the levels of one or more of the three dichlorobenzene
isomers in groundwater supplies from across the country.  The National Organics Monitoring Survey
(NOMS) and Groundwater Supply Survey (GWSS) contain data on all three isomers.  The National
Screening Program for Organics in Drinking Water (NSP) contains data on p-dichlorobenzene only.

Eighteen groundwater systems were analyzed for p-dichlorobenzene during Phase I of NOMS (March to
April 1976), with none of the systems containing quantifiable levels.  These 18 systems were sampled
again during Phase II of the study (May to July 1976) for all three isomers of dichlorobenzene. Four
systems contained p-dichlorobenzene at levels ranging from 0.006-0.41 |ig/L. The average of the
positive values was 0.13 |ig/L with a standard deviation of 0.19 |ig/L; the median value was 0.04 |ig/L.
Two of the 17 systems analyzed for p-dichlorobenzene proved positive, with levels of 0.09 and 0.1 |ig/L.
The minimum quantifiable limit for p-dichlorobenzene was 1.0 |ig/L in Phase I and 0.005 |ig/L in Phases
II and III.

Twelve groundwater supplies were tested for p-dichlorobenzene contamination in the NSP. Of these 12
systems,  one was contaminated with  p-dichlorobenzene at a level of 0.5 |ig/L. The quantification limit
for p-dichlorobenzene was 0.1 |ig/L.

In the GWSS, 5 of the 456 randomly chosen water systems serving 25 or more individuals were positive
for p-dichlorobenzene, at concentrations ranging from 0.52-1.3|ig/L.  The three systems with the highest
values were contaminated at 0.66, 0.68, and 1.3 |ig/L. Of the 5 positive p-dichlorobenzene systems,  3
were from systems serving populations in excess of 10,000 people. The average for all randomly chosen
systems was 0.8 |ig/L with a standard deviation of 0.3 |ig/L; the median was 0.7 |ig/L.  Of the 473
nonrandom locations sampled serving 25 or more individuals, 4 were contaminated with p-di-
chlorobenzene at levels ranging from 0.7-0.9 |ig/L. All 4 positive p-dichlorobenzene samples were from
systems serving populations under 10,001 people. The average p-dichlorobenzene concentration for the
nonrandom systems was 0.8 |ig/L with a standard deviation of 0.09 |ig/L; the median value was 0.7 |ig/L.
The minimum quantitation limit for dichlorobenzene was 0.5 |ig/L.

4.3.5.3 Groundwater - State Data

Dichlorobenzene data (undifferentiated by isomer) on groundwater from 274 wells, supplied to the EPA
by New Jersey, contained only one positive value (between 1-10 |ig/L).

Data from California, Indiana, and New Jersey revealed a low incidence of p-dichlorobenzene
contamination. Groundwater supplies from four locations in California (unspecified water type) were
sampled with only one containing a quantifiable amount of p-dichlorobenzene (0.4 |ig/L from Morada).
Analytical results from seven groundwater samples from Elkhart, Indiana proved negative for p-
dichlorobenzene. One hundred seventy-five groundwater samples were taken from New Jersey with only
one positive value between the detection limit and 9.9 |ig/L (unquantified).

Occurrence Summary and Use Support Document         339                                     March 2002

-------
4.3.5.4 Surface Water - Federal Surveys

Two Federal studies, the National Organic Monitoring Survey (NOMS) and the National Screening
Program for Organics in Drinking Water (NSP), contain data concerning dichlorobenzene levels in
surface water supplies from across the country. The NOMS contains information on all three isomers of
dichlorobenzene, while the NSP provides data for p-dichlorobenzene only.

In Phase I of the National Organic Monitoring Survey (March to April 1976), water samples from 89
surface water systems were analyzed only for p-dichlorobenzene. Of these 89 systems, only two were
found to contain p-dichlorobenzene, at levels of 1.0 and 3.0 |ig/L. Ninety-one systems were sampled
during the second phase  of the survey (May to July 1976) and sixteen of the 91 systems were
contaminated with p-dichlorobenzene, at levels ranging from 0.007-1.6 |ig/L, including the three systems
identified as positive in the first phase of the study. The average p-dichlorobenzene concentration among
the positive Phase II samples was 0.14 |ig/L with a standard deviation of 0.39 |ig/L; the median
concentration was 0.03 |ig/L. During the third phase of the NOMS (November 1976 to January 1977),
analyses revealed contamination by all three isomers of dichlorobenzene.  Of the 89 systems analyzed, 27
systems showed p-dichlorobenzene contamination, at levels ranging from 0.01-0.75 |ig/L. The average
concentration among the 27 positive systems was 0.06 |ig/L ±0.14; the median value was 0.02 |ig/L. For
p-dichlorobenzene, the minimum quantifiable limit was 1.0 |ig/L in Phase I, 0.1 |ig/L in Phase II, and
0.005 |ig/L in Phase III.

Surface water samples from 106 drinking water systems were analyzed for p-dichlorobenzene during the
National Screening Program (NSP) between June 1977 and March 1981.  Of these, 5 systems contained
detectable levels of p-dichlorobenzene, ranging from 0.1-0.9 |ig/L. Only two of these systems were
contaminated at levels greater than 0.1 |ig/L (0.2 and 0.9 |ig/L). The average concentration among the 5
positive systems was 0.3 |ig/L with a standard deviation of 0.3 |ig/L; the median level was 0.15 |ig/L.
The quantification limit for the NSP was 0.1 |ig/L.

4.3.5.5 Surface Water - State Data

The levels of dichlorobenzene in surface water systems at Niagara Falls were monitored. Eight surface
water samples were analyzed for p-dichlorobenzene with four containing quantities ranging from 0.01 to
0.18 |ig/L, the average being  0.05 |ig/L. Of five surface water samples analyzed for dichlorobenzene
undifferentiated as to isomer, three had concentrations of 0.02-0.26 |ig/L (average of 0.16 |ig/L for
positive samples).

4.3.5.6 Surface Water - Miscellaneous Data

Drinking water samples from surface water sources, collected from three cities on Lake Ontario, were
analyzed for the three isomers of dichlorobenzene. Positive results were found for each isomer. In the
case of p-dichlorobenzene, concentrations ranged from 0.008-0.020 |ig/L, with an average of 0.013 |ig/L
(Oliver and Nicol, 1982, as cited in JRB Associates, 1983).

4.3.5.7 Projected National Occurrence of Dichlorobenzene in Public Water Supplies

As reported in the JRB Associates (1983) report, public water systems fall into two major categories with
respect to water source (surface water and groundwater) and into five size categories and twelve
subcategories according to the number of individuals served.  The JRB Associates (1983) report
presented estimates of both the number of drinking water supplies nationally within each of the
Occurrence Summary and Use Support Document         340                                     March 2002

-------
source/size categories expected to have dichlorobenzene present, and of the concentration of
dichlorobenzene expected to be present in those supplies.

The key features of the methodology used and assumptions made to develop the national estimates are
summarized here. The estimates are based on the data from the Federal surveys only.  The State data and
miscellaneous information were not included for several reasons. Generally, these data are from a few
States and were not considered to be geographically representative. There was also a general lack of data
on the population served by systems measured, the type of water sampled, and the methodologies used to
sample, identify, and measure dichlorobenzene.

The Federal survey data from the NOMS, NSP, and GWSS were pooled together for developing the
national projections.  It was assumed in combining these surveys that the resulting data base would be
representative of the nation's water supplies.  In the case of the GWSS data, both the random and
nonrandom samples were included in the projections because statistical test of the GWSS data showed
that neither the frequency of occurrence of positive values nor the mean of the positive values for any of
the three dichlorobenzene isomers was significantly different between the two samples.

Ideally, adequate data would be available to develop the national projections separately for each of the
twelve system size categories within the groundwater and surface water groups; however, the available
data were too limited for this.  JRB Associates (1983) consolidated some of the size categories to have
sufficient data for developing the projections.  In consolidating data from various size categories,
consideration was given to the potential for there being statistically significant differences in the
frequency of occurrence of dichlorobenzene as a function of system size. The consolidation of size
categories therefore involved a balancing of the need to group size categories together to have an
adequate data base for developing the national projections against the need to treat size categories
separately in order to preserve the influence of system size as a determinant of contamination potential.
The consolidation of size  categories also took into account EPA's classification of systems into the five
major groups as very small (25500), small (501-3,300), medium (3,301-10,000), large  (10,001-100,000),
and very large (> 100,000) (Kuzmack, 1983, as cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national level were tested and
an appropriate model selected.  In the case of the dichlorobenzenes, the multinomial method was used.
The frequency of contamination of groundwater and surface water systems at various concentrations was
determined for each consolidated size category. For completing the national estimates, it was assumed
that the frequency of contamination observed for each consolidated category was directly applicable to
each of the system sizes comprising it.

In the JRB Associates (1983) report, it is noted some of the data used in computing the national estimates
are from samples held for a prolonged period of time prior to analysis, with possible biodegradation of
the dichlorobenzenes. Therefore, these projections of national occurrence may underestimate actual
contaminant levels.

4.3.5.7.1  Groundwater Supplies

Fourteen of 948 groundwater supplies sampled were found to have p-dichlorobenzene present at levels
ranging from 0.34 to 1.3 |ig/L.  Based on the overall distribution of positive values and maximum
possible values for those supplies in which the dichlorobenzenes were not found, 0.5 |ig/L was  selected
as the common minimum quantifiable concentration for the combined survey data. That is, quantitative
projections are made of supplies at several concentration ranges > 0.5 |ig/L, while only a total number for
supplies expected to have either no dichlorobenzene or levels below  0.5 |ig/L can be determined.

Occurrence Summary and Use Support Document        341                                      March 2002

-------
Although some data indicate the presence of dichlorobenzene (specifically p-dichlorobenzene) in
groundwater supplies at levels < 0.5 |ig/L, it is not possible to determine the proportion of supplies that
have dichlorobenzene present and the proportion that are actually free of dichlorobenzene contamination.

The survey data was combined across all system sizes for each of the dichlorobenzene isomers in order to
develop the national estimates. The frequency of occurrence of each of the dichlorobenzenes that was
determined on the basis of data for all size supplies was then applied to each of the separate size
categories to complete the national estimates.

An estimated 510 groundwater supplies (range of 197-822), approximately 1.1% of the total groundwater
supplies in the United States, are expected to have p-dichlorobenzene levels of > 0.5 |ig/L; the remaining
47,948 supplies have either no p-dichlorobenzene or levels < 0.5 |ig/L. No groundwater supplies are
expected to have p-dichlorobenzene at levels above 5 |ig/L.

4.3.5.7.2 Surface Water Systems

P-dichlorobenzene was observed in 38 of 150 supplies sampled. The concentrations ranged from 0.1 to
1.6 |ig/L.  Based on the overall distribution of positive values and maximum possible values for those
supplies in which the dichlorobenzenes were not found, 0.5 |ig/L was selected as the common minimum
quantifiable concentration for the combined survey data. That is, quantitative projections are made of
supplies at several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to
have either no dichlorobenzene or levels below 0.5 |ig/L can be determined. Although some data
indicate the presence of dichlorobenzene in surface water supplies at levels < 0.5 |ig/L, it is not possible
to determine the proportion that have dichlorobenzene present and the proportion that are free of
dichlorobenzene contamination.

Twelve surface water supplies (range of 0-34), approximately 0.1% of the total surface water supplies in
the United States, are expected to have p-dichlorobenzene  levels of > 0.5 |ig/L; the remaining 11,190
supplies have either no p-dichlorobenzene or levels < 0.5 |ig/L. No surface water supplies are expected
to have p-dichlorobenzene at levels > 5 |ig/L.

4.3.6  Conclusion

1,4-Dichlorobenzene is primarily used as a space deodorant for toilets and refuse containers, and as a
fumigant for control of moths, molds, and mildews. According to the latest  data available, production of
1,4-dichlorobenzene was 78.8 million pounds in 1994, and production has generally increased overtime.
Industrial releases of 1,4-dichlorobenzene have been reported to TRI since 1988 from 22 States. 1,4-
Dichlorobenzene was also an analyte for the NAWQA ambient occurrence studies. In the NAWQA
study, 1,4-dichlorobenzene was detected in 1.8% of urban  wells and 0.16% of rural wells, with median
detection values of 1.0 |ig/L  and 1.3 |ig/L, respectively.  The Stage 2 analysis, based on the 16-State
cross-section, estimated that  0% of combined ground water and surface water systems serving 0% of the
population exceeded the MCL of 0.075 mg/L.  Based on this estimate, no PWSs nationally are expected
to have estimated mean concentrations of 1,4-dichlorobenzene greater than MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, 1,4-dichlorobenzene is manufactured and/or processed in 16 States and has
TRI releases in 22 States.  1,4-Dichlorobenzene is manufactured and/or processed in 6 out of the 16
cross-section States and has TRI releases in 9 of the 16 cross-section States. The cross-section should
adequately represent the occurrence of 1,4-dichlorobenzene on a national scale based upon the use,

Occurrence Summary and Use Support Document          342                                      March 2002

-------
production, and release patterns of the 16-State cross-section in relation to the patterns observed for all
50 States.

4.3.7  References

Agency for Toxic Substances and Disease Registry (ATSDR). 1998. Toxicological Profile for 1,4-
       Dichlorobenzene.  U.S. Department of Health and Human Services, Public Health Service. 253
       pp. + Appendices. Available on the Internet at http://www.atsdr.cdc.gov/toxprofiles/tplO.pdf

Brass, H.J., M.A. Feige, T. Halloran, J.W. Mello, D. Munch, and R.F. Thomas.  1977. The National
       Organic Monitoring Survey: A sampling and analysis for purgeable organic compounds.
       Drinking water quality enhancement through source protection.  R.B. Pojasek (ed.). Ann Arbor,
       MI: Ann Arbor Science, pp. 393-416.

JRB Associates.  1983. Dichlorobenzenes: Occurrence in Drinking Water, Food, and Air. Draft report
       submitted to EPA  for review November 17, 1983.

Kuzmack, A.M.  1983. Memorandum: Characterization of the water supply industry (FY82).
       Washington, D.C.: Office of Water, USEPA. May 16, 1983.

Mello, Wayne. 1983. Personal communication between Wayne Mello, Technical Support Division,
       Office of Drinking Water, USEPA, and author of JRB Associates, 1983, March  10, 1983.

National Toxicology Program (NTP). 1991. National Toxicology Program Health and Safety
       Information Sheet - 1,4-Dichlorobenzene.  Available on the Internet at
       http://ntp-db.niehs.nih.gov/NTP_Reports/NTP_Chem_H&S/NTP_Cheml/Radianl06-46-7.txt

Oliver, E.G., K.D. Nicol.  1982.  Chlorobenzenes in sediments, water, and selected fish from Lakes
       Superior, Huron, Erie, and Ontario. Envir. Sci. Technol.  16:532-536.

Squillace,  P.J., M.J. Moran, W.W. Lapham, C.V. Price, RM.  Clawges, and J.S. Zogorski. 1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci.  and Tech. 33(23):4176-4187.

USEPA. 2000. TEJExplorer: Trends. Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA. 2001. National Primary Drinking Water Regulations -  Consumer Factsheet on: Para-
       dichlorobenzene.  Office of Ground Water and Drinking Water, USEPA. Available on the
       Internet at http://www.epa.gov/safewater/dwh/c-voc/p-dichlo.html (Last updated 04/12/2001)

USEPA. 2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Westrick, J.J., J.W. Mello, and R.F. Thomas.  1983.  The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data. EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         343                                    March 2002

-------
4.4    1,2-Dichloroethane
Table of Contents

4.4.1  Introduction, Use and Production  	  345
4.4.2  Environmental Release  	  347
4.4.3  Ambient Occurrence  	  347
4.4.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  348
4.4.5  Additional Drinking Water Occurrence Data  	  353
4.4.6  Conclusion	  362
4.4.7  References 	  362
List of Tables

Table 4.4-1: 1,2-Dichloroethane Manufacturers and Processors by State 	  345

Table 4.4-2: United States Production of 1,2-Dichloroethane 	  346

Table 4.4-3: Environmental Releases (in pounds) for 1,2-Dichloroethane
       in the United States, 1988-1999 	  347

Table 4.4-4: Stage 1  1,2-Dichloroethane Occurrence Based on 16-State Cross-Section -
       Systems	  349

Table 4.4-5: Stage 1  1,2-Dichloroethane Occurrence Based on 16-State Cross-Section -
       Population	  349

Table 4.4-6: Stage 2 Estimated 1,2-Dichloroethane Occurrence Based on  16-State
       Cross-Section - Systems	  351

Table 4.4-7: Stage 2 Estimated 1,2-Dichloroethane Occurrence Based on  16-State
       Cross-Section - Population	  352

Table 4.4-8: Estimated National 1,2-Dichloroethane Occurrence - Systems and
       Population Served	  353
Occurrence Summary and Use Support Document         344                                     March 2002

-------
4.4.1  Introduction, Use and Production

1,2-Dichloroethane is a clear, manufactured liquid with the chemical formula C2H4C12. It evaporates
quickly at room temperature and has a pleasant smell and a sweet taste.  1,2-Dichloroethane burns with a
smoky flame (ATSDR, 1999). The major routes of entry of 1,2-dichloroethane to drinking water are a
consequence of industrial activity. It is expected that discharges to surface water and leaching from solid
waste are the primary causes of 1,2-dichloroethane contamination of drinking water. Most discharges of
1,2-dichloroethane ultimately reach the atmosphere because of its high volatility (JRB Associates, 1983).
Other names for 1,2-dichloroethane are ethylene dichloride, dichloroethane, and EDC.

Although large amounts of 1,2-dichloroethane are produced today, most of it is used as a chemical
intermediate for vinyl chloride (ATSDR, 1999). This is, by far, the dominant use of 1,2-dichloroethane.
(Ninety-eight percent of the 1,2-dichloroethane produced is produced for this use.)  It is used in smaller
amounts to synthesize other organic compounds such as vinylidene chloride, 1,1,1-trichloroethane,
trichloroethylene, tetrachloroethylene, aziridines, ethylene diamines, and various chlorinated solvents. It
can also be added to leaded gasoline to remove lead (ATSDR, 1999).

Some uses of 1,2-dichloroethane have been discontinued, especially in many consumer products.  It is no
longer used in varnish and finish removers, soaps and scouring compounds, metal degreasers, ore
flotation, or paints, coatings, and adhesives; or in organic synthesis for extraction and cleaning purposes.
It was also formerly used as a grain, household,  and soil fumigant (ATSDR, 1999).

Most of the 1,2-dichloroethane produced is used captively by the manufacturers.  In 1986 only 6% of the
1,2-dichloroethane produced was sold on the open market, and the most recent information indicates that
about 85% of total 1,2-dichloroethane produced is used captively. Production totals for past years are as
follows: 1984, 7.3 billion pounds; 1985, 12.1 billion pounds; 1986, 12.9 billion pounds; 1990,  13.8
billion pounds;  1992, 15.2 billion pounds; 1993, 17.9 billion pounds; and 1994, 16.8 billion pounds
(ATSDR,  1999).

Table 4.4-1 shows the number of facilities in each State that manufacture and process 1,2-dichloroethane,
the intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA. Table 4.4-2 shows the company names and locations of facilities that produce
1,2-dichloroethane and their annual capacities as of February 1, 1998 (ATSDR, 1999).
Table 4.4-1:  1,2-Dichloroethane Manufacturers and Processors by State
State"
AL
AR
CA
HI
IA
IL
IN
KS
KY
LA
MI
MO
MS
NC
NJ
NM
Number of facilities
2
3
2
1
2
3
3
4
3
15
2
2
1
3
2
1
Range of maximum amounts on site in
pounds'1
1,000-99,999
10,000-999,999
10,000-99,000
10,000-99,999
1,000-9,999
0-99,999
10,000-999,999
100-999,999
10,000-49,999,999
1,000-499,999,999
1,000-9,999
100,000-9,999,999
100,000-999,999
1,000-999,999
10,000-99,999
1,000-9,999
Activities and uses0
11
7,11
8
8,11
1,6
7,11
11,12
2,4,10,11,13
1,3,5,7,10
1,3,4,5,6,7,10,
1,5,7,12,13
1,3,5,7,8,11
7,11
11,13
10,11,13
11









11,13






Occurrence Summary and Use Support Document         345                                     March 2002

-------
State"
NY
PA
PR
SC
TX
Number of facilities
1
6
2
9
19
Range of maximum amounts on site in
pounds'"
10,000-99,999
10,000-999,999
0-99,999
10,000-999,999
0-499,999,999
Activities and uses0
11
1,5,11,13
11,13
7,11
1,2,3,4,5,6,7,8,10,11,13
aPost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
8. As a formulation component
9. As a product component
10. For repackaging
11. As a chemical processing aid
12. As a manufacturing aid
13. Ancillary or other uses
Source: AT SDR, 1999 compilation of TRI961999 data
Table 4.4-2:  United States Production of l,2-Dichloroethanea'b
Manufacturer
Borden Chemicals and Plastics
CONDEA Vista Company
Dow Chemical USA
Formosa Plastics Corporation USA
Geon Company
Georgia Gulf Corporation
Occidental Chemical Corporation
Electrochemicals and Proprietary Products
Division
Electrochemicals
Oxymar
PHH Monomers
PPG Industries, Inc. Chemicals Group
Vulcan Materials Company
Vulcan Chemicals Division
Westlake Monomers Corporation
TOTAL
Location
Geismar, LA
Lake Charles, LA
Freeport, TX
Plaquemine, LA
Baton Rouge, LA
Point Comfort, TX
LaPorte, TX
Plaquemine, LA
Convent, LA
Deer Park, TX
Ingleside, TX

Ingleside, TX
Lake Charles, LA
Lake Charles, LA
Geismar, LA
Calvert City, KY

Annual capacity (millions of pounds)
745
1,400
4,500
2,300
525
1,900
4,000
1,760
1,500
1,950
3,300

3,000
1,400
1,600
500
1,950
32,330
aDerived from Anonymous 1998
'Estimates as of February 1, 1998

Source: ATSDR, 1999
Occurrence Summary and Use Support Document
                                 346
March 2002

-------
4.4.2  Environmental Release

1,2-Dichloroethane is listed as a Toxics Release Inventory (TRI) chemical.  Table 4.4-3 illustrates the
environmental releases of 1,2-dichloroethane from 1988 - 1999.  (1,2-Dichloroethane data are only
available for these years.) Air emissions constitute most of the on-site releases, with a steady decrease
over the years, except for a minor upturn in 1990. Surface water discharges have fluctuated with a
decreasing trend, as releases have remained below 10,000 pounds since 1993. Underground injection has
also generally declined, from a high of almost 1.5 million pounds in 1988 to about 1,200 pounds in 1999.
Releases to land (such as spills or leaks within the boundaries of the reporting facility) have ranged from
near zero to over 7,000 pounds, with no apparent trend.  Off-site releases (including metals or metal
compounds transferred off-site) decreased from 1988-1991, but have since risen, with by far the greatest
release reported in 1999.  The decreases in air emissions, underground injection, and surface water
discharges have contributed to an overall decreasing trend of total 1,2-dichloroethane on- and off-site
releases. These TRI data for 1,2-dichloroethane were reported from 37 States and Puerto Rico, with  18
States reporting every year (USEPA, 2000). Of the 37 States that reported, 11 are in the 16 State cross-
section (used for analyses of 1,2-dichloroethane occurrence in drinking water; see Section 4.4.4).  (For a
map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.4-3:  Environmental Releases (in pounds) for 1,2-Dichloroethane in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
545,225
708,117
910,890
1,051,183
1,292,842
1,930,617
2,382,308
3,307,692
4,089,523
5,609,152
4,251,586
4,615,179
Surface Water
Discharges
833
2,337
1,826
1,848
5,194
7,501
9,871
12,760
26,264
49,513
225,824
40,527
Underground
Injection
1,171
2,178
4,549
5,126
24,339
34,296
5,198
85,750
6,334
826,672
1,046,661
1,452,084
Releases
to Land
2,983
886
27
250
256
15
303
1,858
7,051
7,351
714
2,166
Off-Site Releases
679,749
162,677
120,476
91,249
23,671
75,642
61,675
20,530
6,789
33,279
110,085
166,131
Total On- &
Off-site
Releases
1,229,961
876,195
1,037,768
1,149,656
1,346,302
2,048,071
2,459,355
3,428,590
4,135,961
6,525,967
5,634,870
6,276,087
 Source: USEPA, 2000
4.4.3  Ambient Occurrence

The local, State, and federal data set compiled by NAWQA reports that 1,2-dichloroethane was detected
in 7 out of 351 urban wells (2.0%). The minimum and maximum concentrations detected were 0.2 |ig/L
and 3 |ig/L, respectively. The median value of detection concentrations was 0.5 |ig/L.  1,2-
Dichloroethane was also detected in 18  of the rural 2,539 wells (0.71%) analyzed. The minimum and
maximum concentrations detected were 0.2 |ig/L and 2.8 |ig/L, respectively. The median value of
detection concentrations was 0.55 |ig/L. These data (urban and rural) represent untreated ambient ground
water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).
Occurrence Summary and Use Support Document
347
March 2002

-------
1,2-Dichloropropane was also an analyte in the NURP data.  The NURP study found 1,2-dichloroethane
in urban runoff (Lopes and Dionne, 1998). The minimum concentration detected was not reported, and
the maximum concentration detected was 4 |ig/L, with no mean value reported. The use of the land from
which the samples were taken was unspecified.

4.4.3.1 Additional Ambient Occurrence Data

A summary document entitled "Occurrence of 1,2-Dichloroethane in Drinking Water, Food, and Air"
(JRB Associates, 1983), was previously prepared for past USEPA assessments of 1,2-dichloroethane.
However, no information on the ambient occurrence of 1,2-dichloroethane was included in that
document. (The document did include information regarding 1,2-dichloroethane occurrence in drinking
water, which is discussed in Section 4.4.5 of this report.)

4.4.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of 1,2-dichloroethane occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date.  These data were
evaluated relative to several concentration thresholds of interest: 0.0005 mg/L; 0.0025 mg/L;  and 0.005
mg/L.

All sixteen cross-section State data sets contained occurrence data for 1,2-dichloroethane. These data
represent more than 180,000 analytical results from approximately 23,000 PWSs during the period from
1984 to 1998 (with most analytical results from 1992 to  1997).  The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to ensure adequacy of
coverage and completeness. The overall modal detection limit for 1,2-dichloroethane in the 16 cross-
section States is equal to 0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)

4.4.4.1 Stage 1 Analysis Occurrence Findings

Table 4.4-4 illustrates the occurrence of in drinking water for the public water systems in the  16-State
cross-section relative to three thresholds: 0.005 mg/L (the current MCL), 0.0025 mg/L, and 0.0005 mg/L
(the modal MRL). Based on the 16-State cross-section data, less than one percent of all ground water
and surface water systems had any threshold exceedances. A total of 29 (approximately 0.126% of)
PWSs had analytical results exceeding the MCL; 0.239% of systems (55 systems) had results  exceeding
0.0025 mg/L; and 0.977% of systems (225 systems) had results exceeding 0.0005 mg/L.

Approximately 0.112% of ground water systems (24 systems) had any analytical results greater than the
MCL.  About 0.214% of ground water systems (46 systems) had results above 0.0025 mg/L.  The
percentage of ground water systems with at least one result greater than 0.0005 mg/L was equal to
0.909% (195 systems).

Only 5 (0.317% of) surface water systems had results greater than the MCL.  A total of 9 (0.571% of)
surface water systems had at least one analytical result greater than 0.0025 mg/L. Thirty (1.91% of)
surface water systems had results exceeding 0.0005 mg/L.
Occurrence Summary and Use Support Document         348                                     March 2002

-------
Table 4.4-4:  Stage 1 1,2-Dichloroethane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems
Exceeding Threshold
0.112%
0.214%
0.909%

Surface Water
0.005
0.0025
0.0005
0.317%
0.571%
1.91%

Combined Ground &
Surface Water
0.005
0.0025
0.0005
0.126%
0.239%
0.977%
Number of Systems
Exceeding Threshold
24
46
195

5
9
30

29
55
225
Reviewing 1,2-dichloroethane occurrence in the 16 cross-section States by PWS population served
(Table 4.4-5) shows that approximately 8.40% of the population (over 9 million people) was served by
PWSs (ground and surface water systems) with at least one analytical result of 1,2-dichloroethane greater
than the MCL. Approximately 8.98% of the population (almost 10 million people) was exposed to 1,2-
dichloroethane concentrations greater than 0.0025 mg/L. The percentage of population served by all
PWSs with at least one analytical result of 1,2-dichloroethane greater than 0.0005 mg/L was equal to
11.2% (over 12 million people).

The percentage of population served by surface water systems with threshold exceedances was
considerably greater than the percentage of population served by ground water systems with threshold
exceedances. Approximately 0.789% of the population served by ground water systems (391,200
people) had at least one analytical result of 1,2-dichloroethane greater than 0.005 mg/L, compared to
14.6% of the population served by surface water systems (over  8.9 million people) with results greater
than 0.005 mg/L. When evaluated relative to 0.0025 mg/L, the  percent of population exposed by ground
water systems was  1.30% (641,600 people) and the percent of population exposed by surface water
systems was 15.2% (over 9.3 million people). Approximately 4.64% of the population was served by
ground water systems (over 2.3 million people) with analytical detections greater than 0.0005 mg/L, as
compared to 16.5% of the population served by surface water systems (about 10.1 million people) with
analytical detections greater than 0.0005 mg/L.
Table 4.4-5:  Stage 1 1,2-Dichloroethane Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
Percent of Population
Served by Systems
Exceeding Threshold
0.789%
Total Population Served
by Systems Exceeding
Threshold
391,200
Occurrence Summary and Use Support Document
349
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0025
0.0005
Percent of Population
Served by Systems
Exceeding Threshold
1.30%
4.64%
Total Population Servet
by Systems Exceeding
Threshold
641,600
2,300,500

Surface Water
0.005
0.0025
0.0005
14.6%
15.2%
16.5%
8,916,000
9,311,400
10,113,500

Combined Ground &
Surface Water
0.005
0.0025
0.0005
8.40%
8.98%
11.2%
9,307,200
9,953,000
12,414,000
4.4.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.4-6 and 4.4-7.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Only 1 PWS in the 16 States (approximately 0.00479% of all  systems in the 16 States) had an estimated
mean concentration of 1,2-dichloroethane exceeding 0.005 mg/L.  Four (about 0.0175% of) PWSs were
estimated to have mean concentrations greater than 0.0025 mg/L.  Approximately 30 (0.132% of) PWSs
had estimated mean concentrations of 1,2-dichloroethane greater than 0.0005 mg/L.

The estimated number of ground water PWSs in the 16 States with mean concentrations exceeding the
thresholds of 0.005 mg/L, 0.0025 mg/L, and 0.0005 mg/L was 1 (about 0.00514%), 4 (about 0.0187%),
and 29 (about 0.135%), respectively. No surface water PWSs had estimated mean concentration
exceeding 0.005 mg/L.  One (approximately 0.000762% of) surface water PWSs in the 16 States was
estimated to have a mean concentration greater than 0.0025 mg/L. The percentage of surface water
PWSs in the 16 States with estimated mean  concentration exceeding 0.0005 mg/L was equal to 0.0974%
(approximately 2 systems).
Occurrence Summary and Use Support Document
350
March 2002

-------
Table 4.4-6:  Stage 2 Estimated 1,2-Dichloroethane Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.00514%
0.0187%
0.135%
Range
0.000% -0.0140%
0.00932% - 0.0326%
0.0932% -0.177%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
1
4
29
Range
0-3
2-7
20-38

Surface Water
0.005
0.0025
0.0005
0.000%
0.000762%
0.0974%
0.000% - 0.000%
0.000% - 0.000%
0.0635% -0.191%
0
1
2
0-0
0-0
1 -3

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.00479%
0.0175%
0.132%
0.000% -0.0130%
0.00868% - 0.0304%
0.0955% -0.174%
1
4
30
0-3
2-7
22-40
Reviewing 1,2-dichloroethane occurrence by PWS population served (Table 4.4-7) shows that
approximately 0.0003341% of population served by all PWSs in the 16 States (an estimate of
approximately 400 people) were potentially exposed to 1,2-dichloroethane levels above 0.005 mg/L.
About 39,400 (approximately 0.0357% of) people and about 370,300 (0.334% of) people served by the
combined total of surface and ground water systems were potentially exposed to levels above 0.0025
mg/L and 0.0005 mg/L, respectively.

The percentage of population served by ground water systems in the 16 States with levels greater than
0.005 mg/L was approximately 0.000740% (over 400 people). When evaluated relative to thresholds of
0.0025 mg/L and 0.0005 mg/L, the population exposed was about 37,700 (0.0760%) and 189,200
(0.382%), respectively.

No surface water systems had estimated mean concentrations of 1,2-dichloroethane greater than 0.005
mg/L.  About 0.00297% (approximately  1,800 people) of the population served by surface water in the
16 States were potentially exposed to levels above  0.0025 mg/L. When evaluated relative to thresholds
of 0.0005 mg/L, the population exposed was equal to 0.296% (about 181,000 people).
Occurrence Summary and Use Support Document
351
March 2002

-------
Table 4.4-7:  Stage 2 Estimated 1,2-Dichloroethane Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000740%
0.0760%
0.382%
Range
0.000% -0.00155%
0.00107% -0.1 18%
0.212% -0.616%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
400
37,700
189,200
Range
0-800
500 - 58,400
105,100-305,000

Surface Water
0.005
0.0025
0.0005
0.000%
0.00297%
0.296%
0.000% - 0.000%
0.000% - 0.000%
0.246% - 0.548%
0
1,800
181,000
0-0
0-0
150,800-335,500

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.000331%
0.0357%
0.334%
0.000% - 0.000692%
0.000478% - 0.0542%
0.231% -0.5 18%
400
39,400
370,300
0-800
500 - 60,000
255,900 - 573,600
4.4.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 3 PWSs nationally serving approximately 700 people could be exposed to 1,2-
dichloroethane concentrations above 0.005 mg/L.  About 11 systems serving almost 76,000 people had
estimated mean concentrations greater than 0.0025 mg/L. Approximately 86 systems serving about
711,900 people nationally were estimated to have mean 1,2-dichloroethane concentrations greater than
0.0005 mg/L. (See Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to
national values.)

For ground water systems, an estimated 3 PWSs serving about 600 people nationally had mean
concentrations greater than 0.005 mg/L. Approximately 11 systems serving about 65,200 people
nationally had estimated mean concentration values that exceeded 0.0025 mg/L. About 80 ground water
systems serving almost 327,200 people had estimated mean concentrations greater than 0.0005 mg/L.

Zero surface water systems had estimated mean concentrations of 1,2-dichloroethane  above 0.005 mg/L.
About  1 surface water systems serving 3,800 people had estimated mean concentrations greater than
0.0025 mg/L. An estimated 5  surface water systems serving approximately 376,300 people had mean
concentrations greater than 0.0005 mg/L.
Occurrence Summary and Use Support Document
352
March 2002

-------
Table 4.4-8:  Estimated National 1,2-Dichloroethane Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
3
11
80
Range
0-8
6-19
55-105
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
600
65,200
327,200
Range
0 - 1,300
900-101,000
181,700-527,400

Surface Water
0.005
0.0025
0.0005
0
1
5
0-0
0-0
4-11
0
3,800
376,300
0-0
0-0
313,600-697,500

Combined Ground
& Surface Water
0.005
0.0025
0.0005
3
11
86
0-8
6-20
62-113
700
75,900
711,900
0 - 1,500
1,000-115,400
492,000-1,102,700
4.4.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of 1,2-dichloroethane in drinking water are also
reviewed. Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence of 1,2-Dichloroethane in Drinking Water, Food, and Air" (JRB Associates, 1983), is
presented in the following section.  This variety of studies and information are presented regarding levels
of 1,2-dichloroethane in drinking water, with the scope of the reviewed studies ranging from national to
regional. Note that none of the studies presented in the following section provide the quantitative
analytical results or comprehensive coverage that would enable direct comparison to the occurrence
findings estimated with the cross-section occurrence data presented in Section 4.4.4. These additional
studies, however, do  enable a broader assessment of the Stage 2 occurrence estimates presented for this
Six-Year Review. All the following information in Section 4.4.5 is taken directly from "Occurrence of
1,2-Dichloroethane in Drinking Water, Food, and Air" (JRB Associates, 1983).

JRB Associates (1983) found two major types of data available that were potentially useful for
describing the occurrence of 1,2-dichloroethane in the nation's public drinking water supplies.  First,
there are several Federal surveys in which a number of public water supplies from throughout the U.S.
were selected for analysis of chemical contamination, including 1,2-dichloroethane. Second, data are
available from State surveys and from State investigations of specific incidents of known or suspected
contamination of a supply. For accomplishing the basic objectives of this study, namely to estimate the
number of public water supplies nationally within the various source and size categories contaminated
with 1,2-dichloroethane, the distribution of 1,2-dichloroethane concentrations in those supplies, and the
number of individuals, exposed to those concentrations, it was determined that the Federal survey data
provides the most suitable data base. The State  data tend to be poorly described with respect to the
source and size categories of the supplies examined and the sampling and analysis methods used for
determining contaminant levels.  The lack of source and system size information precludes using the data
for estimating levels in public water supplies of similar characteristics. The absence of details on
sampling and analysis methods precludes evaluating those data for their qualitative and quantitative
Occurrence Summary and Use Support Document
353
March 2002

-------
reliability. Although they are not used with the Federal data for the purpose of estimating contamination
levels nationally, the available State data are presented here to provide some additional perspective on
1,2-dichloroethane occurrence in drinking water.

Data are presented only on drinking water samples taken from a consumer's tap (i.e., distribution water
samples) or on treated water samples taken at the water supply (i.e., finished water samples) because
these are considered to be most representative of the water consumed by the public. No data on raw (i.e.,
untreated) water are presented.  It is recognized that for some groundwater supplies where no treatment
of the water occurs,  samples identified as raw may be representative of water consumed by the users of
the supply. However, it was generally not possible to differentiate between those groundwater supplies
that do and those that do not treat raw water from the available survey data.

4.4.5.1 Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Six Federal drinking water surveys provide data on 1,2-dichloroethane: the National Organics
Reconnaissance Survey (NORS), the National Organic Monitoring Survey (NOMS), the National
Screening Program for Organics in Drinking Water (NSP), the 1978 Community Water Supply Survey
(CWSS), the Rural Water Survey (RWS), and the Groundwater  Supply Survey (GWSS). The terms used
in this report are those used in the individual surveys, recognizing that they may not always correspond to
strict technical definitions.

The National Organics Reconnaissance Survey (NORS) was conducted in 1975 to determine the extent
of the presence of 1,2-dichloroethane, carbon tetrachloride, and four trihalomethanes in drinking water
supplies from 80 cities across the country (Symons et al., 1975, as cited in JRB Associates, 1983).  The
effect of the water source and treatment practices on the formation of these compounds were also
examined in the NORS. Of the 80 supplies studied, 16 were indicated as having a groundwater source
and 64 as having a surface water source. Symons et al. (1975, as cited in JRB Associates, 1983) did not
provide data on the population served by the supplies studied in NORS; the populations served were
estimated by JRB based on information available from other sources for the supplies studied and from
census data for the locations of the supplies.

The National Organic Monitoring Survey (NOMS) was conducted to identify contaminant sources, to
determine the frequency of occurrence of specific drinking water contaminants, and to provide data for
the establishment of maximum contaminant levels  (MCL's) for various organic compounds in drinking
water (Brass et al., 1977, as cited in JRB Associates, 1983).  The NOMS was conducted in three phases:
March-April 1976, May-July 1976, and November 1976-January 1977. Finished drinking water samples
from 113 communities were analyzed for 21 different compounds. Of the 113 community supplies
sampled, 18 had groundwater sources, 91 had surface water sources, and 4 had a mixed
groundwater/surface water source. For 1,2-dichloroethane, 18 groundwater supplies were examined in
Phases I and II, and  14 groundwater supplies in Phase III; for surface water, 88, 91, and 87 supplies were
analyzed for 1,2-dichloroethane in Phases I, II, and III, respectively.

The analytical results of the NOMS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water. Additional information on the locations and source of the supplies,
and on the populations served by the supplies in the NOMS were provided by Wayne Mello (1983, as
cited in JRB Associates, 1983) at EPA's Technical Support Division, Office of Drinking Water.  A single
value for 1,2-dichloroethane was reported for each supply studied in the NOMS.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March  1981, examined both raw and finished drinking water samples from 166 water

Occurrence Summary and Use Support Document         354                                     March 2002

-------
systems in 33 States for 51 organic chemical contaminants. Data are available for 1,2-dichloroethane on
finished water samples from 12 groundwater and 103 surface water supplies.

In the Community Water Supply Survey (CWSS), carried out in 1978, 106 surface water supplies, 330
groundwater supplies, and 16 supplies with mixed sources were examined for volatile organic chemical
contamination. Samples were taken of raw, finished, and distribution water.  Only the latter two types of
water are considered here.  Data for 1,2-dichloroethane in finished and/or distribution samples were
obtained from a total of 315 groundwater and 104 surface water supplies.

The Rural Water Survey (RWS), conducted in 1978, was carried out in response to Section 3 of the Safe
Drinking Water Act, which mandated that EPA "conduct a survey of the quantity, quality, and
availability of rural drinking water supplies." Drinking water samples were collected for analysis of
inorganic chemicals, pesticides, and VOCs  from 2,655 households throughout the United States located
in areas defined in the survey as rural. Of these, a total of 855 household samples were examined for
VOCs.  The majority of these samples were obtained from households receiving water from private wells
or small supplies serving fewer than 25 people. For 1,2-dichloroethane, data are available in the RWS
for 206 groundwater and 35 surface water supplies serving 25 or more people.

The RWS did not obtain data on the number of persons in each household served by the supplies.
However, data were obtained on the number of service connections at each supply.  With the input of Dr.
Bruce Brower at Cornell University, who participated in the statistical analysis of the RWS for
parameters  other than VOCs, the population served by each supply was estimated from the average
number of persons per household (3.034) observed in the survey. A single value was reported for each
household;  in some cases it was necessary to average two or three households obtaining water from the
same supply.  Brass (1981, as cited in JRB Associates, 1983) cautions that the RWS water samples were
analyzed 6 to 27 months after collection and that degradation of some VOCs may have occurred during
this holding period.

The Groundwater Supply Survey  (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al, 1983, as cited in JRB Associates, 1983). It was hoped that this study would
stimulate State efforts toward the  detection  and control of groundwater contamination and the
identification of potential chemical "hot spots." A total of 944  systems were sampled for 1,2-
dichloroethane, of which 466 were chosen at random.  The remaining 478 systems were chosen non
randomly based on information from States encouraged to identify locations believed to have a higher
than normal probability of VOC contamination (e.g., locations near landfills or industrial activity). The
file provided a single analytical result for each supply sampled. One sample of finished water was
collected from each supply at a point near the entrance to the distribution system.

Each of the drinking water surveys was evaluated with respect to the validity of the reported occurrence
data for a number of organic chemicals, including 1,2-dichloroethane.  The evaluations were carried out
by analyzing information about the procedures used for collection and analysis of samples as well as the
quality control protocols used. The analyzed compounds dealt with in each study were assigned one of
three possible ratings: quantitatively acceptable, qualitatively acceptable (i.e., the substance measured
was 1,2-dichloroethane), and totally unacceptable. In the case of 1,2-dichloroethane, a qualitatively
acceptable rating was given for data from the CWSS and RWS because of suspected biodegradation of
the samples, which were held unrefrigerated for prolonged periods before analysis.  1,2-Dichloroethane
values in excess of the quantitation limit reported for some samples in these studies are qualitatively
valid and can be taken as minimum values,  representative of samples which probably originally
Occurrence Summary and Use Support Document         355                                     March 2002

-------
contained 1,2-dichloroethane at higher concentrations. In the case of the NORS, NOMS, NSP, and
GWSS, all data were rated both quantitatively and qualitatively acceptable.

4.4.5.2 Groundwater - Federal Surveys

The National Organics Reconnaissance Survey (NORS), the National Organic Monitoring Survey
(NOMS), the National Screening Program for Organics in Drinking Water (NSP), the Community Water
Supply Survey (CWSS), the Rural Water Survey (RWS), and the Groundwater Supply Survey (GWSS)
all contain data concerning the levels of 1,2-dichloroethane in groundwater supplies from across the
country.

In the NORS, finished water samples were taken from 16 groundwater systems from across the country.
Of these, 4 supplies were reported to have 1,2-dichloroethane present, although the level of
contamination in these 4 supplies was below the minimum quantifiable concentration of 0.2 |ig/L for
those analyses.  All other groundwater supplies were indicated as having "none found," with the
minimum quantifiable concentration ranging from 0.2 to  0.4 |ig/L.

Eighteen groundwater systems were analyzed for 1,2-dichloroethane during Phase I of NOMS (March to
April 1976), with none of the systems containing quantifiable levels.  These 18 systems were sampled
again during Phase II of the study (May to July 1976) with one containing 1,2-dichloroethane at 0.2 |ig/L.
Samples analyzed during Phase III of the study (November 1976 to January 1977) proved negative for all
of the 14 systems examined.  The minimum quantifiable  limits for 1,2-dichloroethane ranged from 1-2
|ig/L in Phase I, 0.05-1 |ig/L in Phase II, and 0.05-1 |ig/L in Phase III.

Twelve groundwater supplies were tested for 1,2-dichloroethane contamination in the NSP.  Of these 12
systems, one was found to be contaminated with  1,2-dichloroethane at 0.2 |ig/L. The quantification limit
for 1,2-dichloroethane was 0.1 |ig/L.

The 1978 CWSS provided information on 1,2-dichloroethane levels in 315 groundwater systems.  Of
these systems, 4 contained detectable levels of 1,2-dichloroethane, with values ranging from 0.56-1.1
l-ig/L. The mean value was 0.9 |ig/L with a standard deviation of 0.3 |ig/L; the median value was 0.9
l-ig/L. The minimum quantitation limit for 1,2-dichloroethane in the CWSS was 0.5 |ig/L.

The RWS examined 206 groundwater supplies for 1,2-dichloroethane and found none to have levels
above the minimum quantification limit of 0.5 |ig/L.

In the GWSS, 3 of the 456 randomly chosen water systems serving 25 or more individuals were
contaminated with 1,2-dichloroethane, at concentrations  of 0.53, 0.57, and 0.95 |ig/L. Of the 472
nonrandom locations sampled serving 25 or more individuals, 7 were contaminated with 1,2-
dichloroethane, at concentrations between 1.1-9.8 |ig/L, the highest values being 2.9, 3.4, and 9.8 |ig/L.
Of the 7 positive samples, 4 were from systems serving populations in excess of 10,000 people.  The
average 1,2-dichloroethane level for the nonrandom systems was 3.4 |ig/L with a standard deviation of
2.9 |ig/L; the median value was 2.5  |ig/L. The minimum quantitation limit for 1,2-dichloroethane was
0.2 \igfL.

4.4.5.3 Groundwater - State Data

Six States (California, Connecticut, Delaware, Indiana, Massachusetts, and New Jersey) provided EPA
with information concerning 1,2-dichloroethane contamination in groundwater supplies. Analytical
results for samples from four locations in California ranged from undetectable to 21 |ig/L. Samples from

Occurrence Summary and Use Support Document         356                                     March 2002

-------
two locations in Connecticut gave 1,2-dichloroethane readings of not detectable and 7.8 |ig/L. Delaware
reported data from three locations, with four samples having no detectable 1,2-dichloroethane. Data on
26 samples from two cities in Indiana showed 17 samples with undetectable 1,2-dichloroethane and nine
samples with concentrations ranging from 19-2,100 |ig/L (averaging 624 |ig/L). In data from
Massachusetts, there was no detectable 1,2-dichloroethane in five groundwater samples from two cities,
while levels ranging from 10.1-19.1 |ig/L (averaging 14.7 |ig/L) were reported in samples taken from two
other cities. In New Jersey data, 405 samples showed no detectable 1,2-dichloroethane and 26 additional
samples were contaminated at 0. 1-100 |ig/L (ranges were the only data reported for groups of samples, so
no averages can be taken).

4.4.5.4 Surface Water - Federal Surveys

The National Organics  Reconnaissance Survey (NORS), the National Organic Monitoring Survey
(NOMS), the National Screening  Program for Organics in Drinking Water (NSP), the Community Water
Supply Survey (CWSS), and the Rural Water Survey (RWS) all contain data concerning the levels of 1,2-
dichloroethane in  surface water supplies from across the country.

In the NORS, finished water from 64 surface water systems were studied, 22 of which were found to
have 1,2-dichloroethane present.  Sixteen of the 22 positive systems reported  1,2-dichloroethane to be
present, but below the minimum quantifiable concentration range of 0.2 to 0.4 |ig/L. Of six systems with
quantifiable levels, the  concentrations ranged from 0.2-6 |ig/L.  The 42 negative surface water systems
were indicated as  having "none found," with the minimum quantifiable concentration ranging from 0.2 to
0.4 |ig/L. It should be noted that confirmatory quantitative analyses were performed for 8 surface water
supplies in  NORS, using a method able to quantify 1,2-dichloroethane at 0.1 |ig/L. Three of these 8
supplies were originally reported  as having 1,2-dichloroethane present, but below the minimum
quantifiable concentration of 0.2 |ig/L in 2 cases and 0.4 |ig/L in the third case; 1,2-dichloroethane was
not observed in any of these 3 systems in the confirmatory analysis. Of the 5  supplies originally reported
as having "none found," one was  observed in the confirmatory analysis to have 1,2-dichloroethane
present at 0.2 |ig/L.

In Phase I of the National Organic Monitoring Survey (March to April  1976), water samples from 88
surface water systems were analyzed for 1,2-dichloroethane.  Of these 88 systems, only one was found to
contain 1,2-dichloroethane, at 2 |ig/L. Ninety-one systems were sampled during the second phase of the
survey (May to July 1979) and one was found to be contaminated at 1.8 |ig/L. During the third phase of
the NOMS  (November 1976 to January 1977), analyses revealed 1,2-dichloroethane contamination in
one out of a total of 87  systems, at 1.2 |ig/L. The minimum quantifiable limits for 1,2-dichloroethane
ranged from 1-3 |ig/L in Phase I, 0.05-1 |ig/L in Phase II, and 0.05-2 |ig/L in Phase III.

Surface water samples from 106 drinking water systems were analyzed for 1,2-dichloroethane during the
National Screening Program (NSP) between June 1977 and March 1981. Of these, only one system
contained a detectable level of 1,2-dichloroethane, at 4.8 |ig/L.  The quantification limit for the NSP was
0.1
None of the 104 surface water systems sampled during the Community Water Supply Survey (CWSS)
contained quantifiable levels of 1,2-dichloroethane.  The minimum quantitation limit for 1,2-
dichloroethane in the CWSS was 0.5 |ig/L.

The RWS examined drinking water from 35 surface water supplies; only one supply was found to have
1,2-dichloroethane present above the minimum quantification limit range of 0.5-2.0 |ig/L. The
concentration of 1,2-dichloroethane observed in that supply was 19  |ig/L.

Occurrence Summary and Use Support Document         357                                      March 2002

-------
4.4.5.5  Surface Water - State Data

The only State supplied surface water information on 1,2-dichloroethane was from New York.  One
sample from Poughkeepsie assayed for 1,2-dichloroethane was positive at 5.9 |ig/L.

4.4.5.6  Projected National Occurrence of 1,2-Dichloroethane in Public Water Supplies

As reported in the JRB Associates (1983) report, public water systems fall into two major categories with
respect to water source (surface water and groundwater) and into five size categories and twelve
subcategories according to the number of individuals served. The JRB Associates (1983) report
presented estimates of both the number of drinking water supplies nationally within each of the
source/size categories expected to have 1,2-dichloroethane present, and of the concentration of 1,2-
dichloroethane expected to be present in those supplies.

The key features of the methodology used and assumptions made to develop the national estimates are
summarized here. The estimates  are based on the data from the Federal surveys only.  The State data
were not included for several reasons. These data are from a few States and were not considered to be
geographically representative.  There was  also a general lack of data on the population served by systems
measured, the type of water sampled, and the methodologies used to sample, identify,  and measure 1,2-
dichloroethane.

The Federal survey data from the NORS, NOMS, NSP, CWSS, RWS, and GWSS (random only) were
pooled together for developing the national projections.  It was assumed in combining these surveys that
the resulting data base would be representative of the nation's water supplies. In the case of the GWSS
data, only the randomly selected samples were included in the projections because a statistical test of the
GWSS data showed that the mean of the positive values  for 1,2-dichloroethane  was  significantly higher
in the nonrandom portion. Therefore, these data were not included in the national projections to avoid
biasing the results. Ideally, adequate data would be available to develop the national projections
separately for each of the twelve  system size categories within the groundwater and surface water groups;
however, the available data were  too limited for this.  It was, therefore, necessary to consolidate some of
the size  categories to have sufficient data for developing the projections. In consolidating data from
various  size categories, consideration was  given to the potential for there being  statistically significant
differences in the frequency of occurrence of 1,2-dichloroethane as a function of system size.  The
consolidation of size categories therefore involved a balancing of the need to  group  size categories
together to have an adequate data base for developing the national projections against the need to treat
size categories separately in order to preserve the influence of system size as a determinant of
contamination potential. The consolidation of size categories also took into account EPA's classification
of systems  into the five major groups as very small (25-500), small (501-3,300), medium (3,301-10,000),
large (10,001-100,000), and very large (>  100,000) (Kuzmack, 1983, as cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national  level were tested and
an appropriate model selected.  In the case of 1,2-dichloroethane, the multinominal method was used.
The frequency of contamination of groundwater and surface water systems at various concentrations was
determined for each consolidated size category. For completing the national estimates, it was assumed
that the  frequency of contamination observed for each consolidated category was directly applicable to
each of the system sizes comprising it.

In the JRB  Associates (1983) report, it is noted that some of the data used in computing the national
estimates are from samples held for a prolonged period of time prior to analysis, with possible
Occurrence Summary and Use Support Document          358                                      March 2002

-------
biodegradation of 1,2-dichloroethane. Therefore, these projections of national occurrence may
underestimate actual contaminant levels.

4.4.5.6.1 Groundwater Supplies

JRB Associates (1983) reported that data were available for a total of 1,001 supplies from the combined
surveys (NORS, NOMS, NSP, CWSS, RWS, and GWSS (random)). Of these, 12 supplies were reported
to have 1,2-dichloroethane present, at concentrations ranging from 0.2 |ig/L to 1.1 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which  1,2-dichloroethane was not found, 0.5 |ig/L was selected as the common minimum quantifiable
concentration for the combined survey data. An estimate is made of supplies having concentrations > 0.5
l-ig/L, while only a total number for supplies expected to have either no 1,2-dichloroethane or levels
below 0.5 |ig/L can be determined. Although some data indicate the presence of 1,2-dichloroethane in
groundwater supplies at levels < 0.5 |ig/L, it is not possible to determine the proportion of supplies that
have 1,2-dichloroethane present and the proportion that are actually free of 1,2-dichloroethane
contamination.

Of the  989 supplies reporting no 1,2-dichloroethane to be present, 968 were assumed to have maximum
possible levels of < 0.5 |ig/L based on the minimum quantifiable concentrations reported for the various
surveys. The other 21 supplies reporting no 1,2-dichloroethane to be present had maximum possible
levels ranging from approximately 0.53  |ig/L to 1.5 |ig/L.  It is assumed, based on the overall distribution
of values, that 1,2-dichloroethane if present in these 21 supplies is so at a concentration of < 0.5 |ig/L,
although a rigorous, conservative argument could be made for assuming a level equal to the maximum
possible value. The impact of this assumption is considerable both in terms of the national projection of
groundwater systems above 0.5 |ig/L, and in terms  of the population exposed.

Nine of the 1,001 supplies examined had measured values of 1,2-dichloroethane > 0.5  |ig/L. When the
twelve size categories were consolidated into the five major EPA groupings, there was an apparent rela-
tionship between the frequency of values > 0.5 |ig/L and system size:
Very small
Small
Medium
Large
Very large
Overall
0%
0.7%
0%
2.7%
3.0%
0.9%
(0/352)
(2/290)
(0/100)
(6/225)
(1/34)
(9/1,001)
A test for statistical significance revealed that at the a = 0.05 level, the difference among the very small,
small, and medium categories was not significant; nor was the difference between the large and very
large categories. However, the combined very small, small, and medium categories were different from
the combined large and very large categories. Two consolidated categories were, therefore, selected for
developing the national estimates:

                               Very small/small/medium (< 10,000)
                                   Large/Very large (> 10,000)

As noted previously, the frequency of occurrence of 1,2-dichloroethane at various concentrations was
determined for the consolidated groups and then applied to the number of supplies nationally within each
of the size categories comprising each group.

Occurrence Summary and Use Support Document         359                                      March 2002

-------
About 161 groundwater supplies (range of 0-337), approximately 0.3% of the total groundwater supplies
in the United States, are expected to have 1,2-dichloroethane at levels of > 0.5 |ig/L; the remaining
48,297 supplies have either no 1,2-dichloroethane or levels < 0.5 |ig/L.  It is estimated that no
groundwater supplies will have levels exceeding 5 |ig/L.

It is interesting to note the impact on the national projections of the assumption made that the 21 supplies
with undetected but maximum potential values of 0.53-1.5 |ig/L had < 0.5 |ig/L. Had it been assumed
that 1,2-dichloroethane was present in those supplies at their maximum possible values, the national
projection of supplies with 1,2-dichloroethane levels of 0.5-5 |ig/L would have increased to 791 (381-
1,200) with no differences in levels > 5 |ig/L.  These differences  would be found primarily in systems
serving < 2,500 people.

It is also  interesting to compare these national estimates to estimates that include the nonrandom portion
of the GWSS. Inclusion of the nonrandom data results in an estimated 258 supplies exceeding 0.5 |ig/L.
Of these, 3 supplies are estimated to exceed 5  |ig/L, based on one observation in the 25,001-50,000
category. None, however, are expected to have levels > 10 |ig/L. These data indicate that, while the
projections for levels > 5  |ig/L based on the random data only may underestimate actual conditions, the
error is very small.

4.4.5.6.2 Surface Water

Data are  available for a total of 301 surface water supplies. Of these, 25 supplies were reported by JRB
Associates (1983) to have 1,2-dichloroethane present at concentrations ranging from 0.2 |ig/L to 19 |ig/L,
although this high value is the only one reported above 4.8 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which 1,2-dichloroethane was not found, 0.5 |ig/L was selected as the common minimum quantifiable
concentration for the combined survey data. That is, quantitative projections are made of supplies at
several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either
no 1,2-dichloroethane or levels below 0.5 |ig/L can be determined. Although some data indicate the
presence of 1,2-dichloroethane in surface water supplies at levels < 0.5  |ig/L, it is not possible to
determine the proportion that have  1,2-dichloroethane present and the proportion that are free of 1,2-
dichloroethane contamination.

Of the 276 supplies reporting no  1,2-dichloroethane to be present, 129 had maximum possible levels of <
0.5 |ig/L based on the minimum quantifiable concentrations reported for the various surveys. The other
147 supplies reporting no 1,2-dichloroethane to be present had maximum possible levels ranging from
approximately 0.52 |ig/L  to 2 |ig/L. It is assumed, based on the overall  distribution of values, that 1,2-
dichloroethane if present  in these 147 supplies is  so at a concentration of < 0.5 |ig/L, although a rigorous
conservative argument could be made for assuming a level equal to the maximum possible value. As will
be noted further below, the difference between these alternatives is very large for both the estimate of the
number of surface water supplies with 1,2-dichloroethane  > 0.5 |ig/L and for the estimated population
exposed to 1,2-dichloroethane at levels > 0.5 |ig/L in surface water supplies.

Eighteen of the 301 supplies examined had measured values of 1,2,-dichloroethane > 0.5 |ig/L.  When the
twelve size categories were consolidated into the  five major EPA groupings, there was an apparent
relationship between the frequency of values > 0.5 |ig/L and system size:

Very smal      10%    (0/20)
Small          0%    (0/55)

Occurrence Summary and Use Support Document         360                                     March 2002

-------
Medium       5.0%   (2/40)
Large          5.6%   (4/71)
Very large     10.4% (12/115)
Overall        6.0%   (18/301)

A test for statistical significance revealed that at the a = 0.05 level, the very small, small, and medium
groups were not different from one another and that the large and very large groups are not different;
however, the combined very small, small, and medium groups and the combined large and very large
groups are different. These two consolidated categories were selected for developing the national
estimates:

                               Very small/small/medium (< 10,000)
                                    Large/very large (> 10,000)

As noted previously, the frequency of occurrence of 1 ,2-dichloroethane at various concentrations was
determined for the consolidated groups and then applied to the number of supplies nationally within each
of the size categories comprising each group.

About 317 surface water supplies (range of 82-551), approximately 2.8% of the total surface water
systems in the United States, are expected to have 1,2-dichloroethane at levels > 0.5 |ig/L; the remaining
10,885 supplies have either no 1,2-dichloroethane or levels < 0.5  |ig/L. It is estimated that 82 surface
water supplies (range of 0-241) will have levels > 10  |ig/L and none are projected to have levels above 20
There was a notable impact on the national projections by the assumption made that the 147 supplies
with undetected but maximum potential values of 0.52-2 |ig/L had < 0.5 |ig/L.  Had it been assumed that
1,2-dichloroethane was present in these supplies at their maximum possible values, the national
projections of supplies with 1,2-dichloroethane levels >0.5 |ig/L would be 5,993 (range of 5,125-6,860)
supplies. However, there would be no difference in the projected number of surface water supplies with
levels > 5 |ig/L.

4.4.5.7 Drinking Water Intake - Estimated Population Exposed

The values given here were obtained using Federal Reporting Data Systems data on populations served
by primary water supply systems and the estimated number of these water systems that contain a given
level of 1,2-dichloroethane. An estimated 12,232,000 individuals (5.7% of the population of
214,419,000 using public water supplies) are exposed to levels of 1,2-dichloroethane in drinking water at
or above 0.5 |ig/L, while 143,000 individuals (0.1%) are exposed to levels above 5 |ig/L.  It is estimated
that no individuals are exposed to levels greater than 20 |ig/L.  Of the approximately 12 million people
exposed to  levels ranging from 0.5 to 5 |ig/L, 1 1 million (90%) obtain water from  surface water supplies.
All exposure to 1,2-dichloroethane in drinking water at levels above 5 |ig/L is expected to be from
surface water sources.

The previous section also presented estimates of the number of public water supplies exceeding various
concentrations when the GWSS nonrandom  data was included in the analysis.  Had the GWSS
nonrandom data been included, an increased estimate in the population exposed to 1,2-dichloroethane at
levels below 10 |ig/L would be seen, notably an increase of 104,000 individuals at levels between 5 and
10 i/L.
Occurrence Summary and Use Support Document         361                                      March 2002

-------
Also presented in the previous section were estimates of the number of public water supplies exceeding
various concentrations using an alternative assumption for 168 supplies in which no 1,2-dichloroethane
was observed, but which may have contained concentrations exceeding 0.5  |ig/L owing to relatively high
minimum quantifiable levels in the analysis of those supplies.  Had it been assumed that the 168 supplies
sampled had been contaminated at the maximum quantifiable level, the national projections would have
indicated that 82 million individuals were exposed to levels at or above 0.5  |ig/L. This population
represents an increase of about 70 million individuals over the estimate used here that assumes the
supplies do not have contamination at the maximum quantifiable level but rather at a level less than 0.5
l-ig/L. Note, however, that no change occurs in the estimated population exposed to levels exceeding 5
p-g/L.

4.4.6  Conclusion

1,2-Dichloroethane is used primarily to manufacture other organic compounds.  Ninety-eight percent of
the 1,2-dichloroethane produced in the U.S. is used to make vinyl chloride.  Most 1,2-dichloroethane
produced in the United States is used captively by the manufacturers.  Data from the 1990s indicates that
production is increasing, with almost 17 billion pounds of 1,2-dichloroethane being produced in 1994.
1,2-Dichloroethane is also a TRI chemical. Industrial releases of 1,2-dichloroethane have been reported
from 1988-1999 in 37 States and Puerto Rico. 1,2-Dichloroethane was an analyte for the NAWQA and
NURP ambient occurrence studies. In the NAWQA study, 1,2-dichloroethane was detected in 2.0% of
urban wells and 0.71% of rural wells, with median detection values of 0.5 |ig/L and 0.55 |ig/L,
respectively.  In the Stage 2 analysis of 16-State occurrence of 1,2-dichloroethane,  0.00479% of
combined ground water and surface water systems serving 0.000331% of the population exceeded the
MCL of 0.005 mg/L. Nationally, 3 ground water and/or surface water systems (serving approximately
700 people) are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, 1,2-dichloroethane is manufactured and/or produced in 21 States and has
TRI releases in 37 States. 1,2-dichloroethane is manufactured and/or processed in  10 out of the 16 cross-
section States and has TRI releases in 11 of the 16 cross-section States. The cross-section should
adequately represent occurrence of 1,2-dichloroethane on a national scale based upon the use,
production, and release patterns in the 16-State cross-section in relation to the patterns observed for all
50 States.

4.4.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1999.  Toxicological Profile for 1,2-
       Dichloroethane.  U.S. Department of Health and Human Services, Public Health Service. 239
       pp. + Appendices. Available on the Internet at http://www.atsdr.cdc.gov/toxprofiles/tp38.pdf

Anonymous.  1998. Chemical profile: Ethylene dichloride. Chem Mark  Rep. February 16, 1998.

Brass, H.J., M.A. Feige, T. Halloran, J.W. Mello, D. Munch, and R.F. Thomas.  1977.  The National
       Organic Monitoring Survey: A sampling and analysis for purgeable organic compounds.
       Drinking water quality enhancement through source protection.  R.B. Pojasek (ed.). Ann Arbor,
       MI: Ann Arbor Science, pp. 393-416.
Occurrence Summary and Use Support Document         362                                      March 2002

-------
Brass, H.J.  1981.  Rural Water Survey organics data.  Memorandum of March 17, 1981 to Hugh Hanson,
       Chief, Science and Technology Branch, Office of Drinking Water, and David Schnare, Office of
       Drinking Water, USEPA, Washington, DC.

JRB Associates. 1983. Occurrence of 1,2-Dichloroethane in Drinking Water Food and Air. Draft
       report submitted to EPA for review November 18, 1983.

Kuzmack, A.M. 1983. Memorandum: Characterization of the water supply industry (FY82).
       Washington, D.C.: Office of Water, USEPA.  May 16, 1983.

Lopes, T.J. and S.G. Dionne.  1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater. U.S. Geological Survey Open-File Report 98-409.
       67pp.

Mello, Wayne. 1983. Personal communication between Wayne Mello, Technical Support Division,
       Office of Drinking Water, USEPA, and author of JRB Associates, 1983, March 10,  1983.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, RM. Clawges, and J.S. Zogorski. 1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci. and Tech.  33(23):4176-4187.

Symons, J.M., T.A. Bellar, J.K. Carswell, et al. 1975. National Organics Reconnaissance Survey for
       Halogrenated Organics.  Journal of the American Water Works Association.  667(11): 634-647.

USEPA. 2000. TRIExplorer: Trends.  Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Westrick, J.J., J.W. Mello, and R.F. Thomas. 1983. The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data. EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         363                                    March 2002

-------
4.5    1,1-Dichloroethylene
Table of Contents

4.5.1  Introduction, Use and Production  	 365
4.5.2  Environmental Release  	 366
4.5.3  Ambient Occurrence  	 367
4.5.4  Drinking Water Occurrence Based on the 16-State Cross-Section	 367
4.5.5  Additional Drinking Water Occurrence Data  	 373
4.5.6  Conclusion	 378
4.5.7  References 	 378
Tables and Figures

Table 4.5-1: Facilities that Manufacture or Process 1,1-Dichloroethylene  	 365

Table 4.5-2: Environmental Releases (in pounds) for 1,1-Dichloroethylene
       in the United States, 1988-1999 	 366

Table 4.5-3: Stage 1  1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section -
       Systems	 368

Table 4.5-4: Stage 1  1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section -
       Population	 369

Table 4.5-5: Stage 2 Estimated 1,1-Dichloroethylene Occurrence Based on 16-State
       Cross-Section - Systems	 370

Table 4.5-6: Stage 2 Estimated 1,1-Dichloroethylene Occurrence Based on 16-State
       Cross-Section - Population	 371

Table 4.5-7: Estimated National 1,1-Dichloroethylene  Occurrence - Systems and
       Population Served	 372
Occurrence Summary and Use Support Document         364                                     March 2002

-------
4.5.1  Introduction, Use and Production

1,1-Dichloroethylene (chemical formula C2H2C12) is a man-made chemical known that is never found
naturally in the environment. It is a colorless liquid that evaporates quickly at room temperature, has a
mild sweet smell, and burns quickly. 1,1-Dichloroethylene can be found in landfills as the result of the
breakdown of polyvinylidene chloride products. Its major routes of entry to drinking water are a
consequence of industrial activity.  Quantities of 1,1-dichloroethylene in water may result from industrial
discharge, atmospheric fallout, or the release of quantities remaining as impurities in products (JRB
Associates, 1983).  Other names for 1,1-dichloroethylene include vinylidene chloride, 1,1-
dichloroethene, and DCE.

Although 1,1-dichloroethylene is manufactured in large quantities, most of it is used to make other
substances or products such as polyvinylidene chloride (ATSDR, 1994).  Single molecules of 1,1-
dichloroethylene are used in the production of polyvinylidene chloride copolymers and as intermediates
for captive organic chemical synthesis and (ATSDR, 1994). Ninety-six percent of the 1,1-
dichloroethylene produced is used in the production of copolymers with vinyl chloride or acrylonitrile
(USEPA, 2001).  The polymers, which have been commercially important since their introduction in the
early  1940s, are used extensively in many types of flexible packing materials, as flame retardant coatings
for fiber and carpet backing, and in piping, coating for steel pipes, and adhesive applications.  The major
application of polyvinylidene chloride copolymers is the production of flexible films for food packaging
(SARAN and VELON wraps). At one time, SARAN wrap was found to contain up to 30 parts per
million (ppm) 1,1-dichloroethylene. Currently, the plastic packaging films can contain no more than 10
ppm 1,1-dichloroethylene (ATSDR, 1994).

The two major producers of 1,1-dichloroethylene are Dow Chemical and Pittsburgh Paint and Glass
(PPG) Industries. Production capacity in 1977  was 270 million pounds/year, but by 1985, production
capacity had declined to 178 million pounds/year.  In 1989, production was estimated at 230 million
pounds (ATSDR, 1994).

Table 4.5-1 shows the number of facilities in each State that manufacture or process 1,1-
dichloroethylene, the activities and uses of the product, and the range of maximum amounts on site
derived from the Toxics Release Inventory (TRI) of EPA (ATSDR, 1994).
Table 4.5-1:  Facilities that Manufacture or Process l,l-Dichloroethylenea
Facility
3M
Monsanto Co. Chemical
Eastman Kodak Co.
Dow Chemical Dalton Site
Morton International Inc.
BF Goodrich Co. Louisville
W.R. Grace & Co.
Marine Shale Processors Inc.
Vulcan Materials Co.
Chemical Div.
Dow Chemical Co. Louisiana
PPG Industries Inc.
Dow Chemical USA Midland
Rhone-Poulenc Inc. Walsh
Allied-Signal Inc. Elizabeth
Location1"
Decatur, AL
Decatur, AL
Windsor, CO
Dalton, GA
Ringwood, IL
Louisville, KY
Owensboro, KY
Amelia, LA
Geismar, LA
Plaquemine, LA
Westlake, LA
Midland, MI
Gastonia, NC
Elizabeth, NJ
Range of maximum
amounts on site in
1,000-9,999
100,000-999,999
1,000-9,999
0-99
1,000,000-9,999,999
10,000-99,999
100,000-999,999
10,000-99,999
10,000-99,999
1,000-9,999
10,000,000-49,999,999
10,000,000-49,999,999
10,000-99,999
10,000-99,999
Activities and
As a reactant
As a reactant
As a reactant
As a reactant
As a reactant
As a reactant
As a reactant
As a reactant
Produce; as a
uses




byproduct; as an impurity; as a reactant
Produce; as a byproduct; as a reactant; in ancillary or
Produce; for sale/distribution; as a byproduct; as an
As a reactant; in ancillary or other uses
As a reactant
As a reactant
Occurrence Summary and Use Support Document
365
March 2002

-------
Facility
Du Pont Parlin Plant
Allied Signal Inc.
Eastman Kodak Co. Kodak
Gencorp Polymer Products
Occidental Chemical Corp.
VCM Plant
Dow Chemical Co. Texas
Operations
Occidental Chemical Co.
Corpus Christi Plant
Hercules Inc.
Arco Chemical Co.
Location11
Parlin, NJ
Buffalo, NY
Rochester, NY
Mogadore, OH
Deer Park, TX
Freeport, TX
Gregory, TX
Covington, VA
South Charleston,
Range of maximum
amounts on site in
1,000-9,999
10,000,000-49,999,999
1,000-9,999
100,000-999,999
10,000-99,999
1,000,000-9,999,999
100-999
100,000-999,999
100,000-999,999
Activities and uses
As a reactant
As a reactant
As a reactant
As a reactant
Produce; as a byproduct
Produce; for sale/distribution; as a byproduct; as an
impurity; in repackaging; as a processing aid; in
Produce; as a byproduct
As a reactant
As a reactant
'Derived from TRI91 (1993)
bPost Office State abbreviations used

Source: ATSDR, 1993 compilation O/TR191 1993 data
4.5.2 Environmental Release

1,1-Dichloroethylene is listed as a Toxics Release Inventory (TRI) chemical (as vinylidene chloride).
Table 4.5-2 illustrates the environmental releases for 1,1-dichloroethylene from 1988 - 1999 (1,1-
dichloroethylene data are only available for these years). Air emissions constitute most of the on-site
releases, with a downward trend in emissions. The decrease in air emissions has contributed to decreases
in 1,1-dichloroethylene total on- and off-site releases in recent years. Releases to land (such as spills or
leaks within the  boundaries of the reporting facility) decreased significantly from 1988-1991, and they
have been at or near zero from 1994-1999.  Releases from surface water discharges and underground
injection have fluctuated over the years with no discernable trends; also, underground injections data are
incomplete because they were not monitored from 1991-1996. Since 1989, off-site releases (including
metals or metal compounds transferred off-site) have fluctuated, from zero to over 6,000 pounds, with no
apparent trend.  These TRI data for 1,1-dichloroethylene were reported from 19 States and Puerto Rico,
with 11 States reporting every year (USEPA, 2000).  Of the  19 States that reported,  10 are included in the
16 State cross-section (used for analyses of 1,1-dichloroethylene occurrence in drinking water; see
Section 4.5.4). (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.5-2: Environmental Releases (in pounds) for 1,1-Dichloroethylene in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
On-Site Releases
Air Emissions
155,891
179,391
199,243
192,815
193,550
165,743
204,814
253,920
Surface Water
Discharges
132
311
662
466
642
215
192
1,306
Underground
Injection
99
218
323
-
-
-
-
-
Releases
to Land
0
0
-
1
0
0
20
14
Off-Site Releases
8
3
104
33
260
2,031
1
0
Total On- &
Off-site
Releases
156,130
179,923
200,332
193,315
194,452
167,989
205,027
255,240
Occurrence Summary and Use Support Document
366
March 2002

-------
Year
1991
1990
1989
1988
On-Site Releases
Air Emissions
287,640
305,686
222,738
296,353
Surface Water
Discharges
794
251
2,691
3,462
Underground
Injection
--
150
720
170
Releases
to Land
15
463
540
429
Off-Site Releases
7
12
6,307
44,281
Total On- &
Off-site
Releases
288,456
306,562
232,996
344,695
 Source: USEPA, 2000
4.5.3  Ambient Occurrence

The local, State, and federal data set compiled by NAWQA reports that 1,1-dichloroethylene was
detected in 12 out of 397 urban wells (3.0%). The minimum and maximum concentrations detected were
0.2 |ig/L and 11 M-g/L, respectively. The median value of detection concentrations was 0.9 |ig/L.  1,1-
Dichloroethylene was also detected in 7 of the 2,414 rural wells (0.29%) analyzed The minimum and
maximum concentrations detected were 0.4 |ig/L and 39 |ig/L, respectively. The median value of
detection concentrations was 1.0 |ig/L. These data (urban and rural) represent untreated ambient ground
water of the conterminous United States for the years 1985-1995 (Squillace et al.,  1999).

1,1-Dichloroethylene was also an analyte in the NURP data. The NURP study found 1,1-
dichloroethylene in urban runoff (Lopes and Dionne, 1998). The minimum and maximum concentrations
detected were 1.5  |ig/L and 4 |ig/L, respectively, with no mean value reported.  The use of the land from
which the samples were taken was unspecified.

4.5.3.1 Additional Ambient Occurrence Data

A summary document entitled "Occurrence of 1,1-Dichloroethylene in Drinking Water, Food, and Air"
(JRB Associates, 1983), was previously prepared for past USEPA assessments of 1,1-dichloroethylene.
However, no information on the ambient occurrence of 1,1-dichloroethylene was included in that
document. (The document did include information regarding 1,1-dichloroethylene occurrence in
drinking water, which is discussed in Section 4.5.5 of this report.)

4.5.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of 1,1-dichloroethylene occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative  to several concentration thresholds of interest: 0.007 mg/L; 0.005 mg/L; 0.001 mg/L;
and 0.0005 mg/L.

All sixteen cross-section State data sets contained occurrence data for 1,1-dichloroethylene. These data
represent more than 146,000 analytical results from approximately 19,000 PWSs during the period from
1984 to 1998 (with most analytical results from  1992 to 1997).  The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to ensure adequacy of
coverage and completeness. The overall modal detection limit for 1,1-dichloroethylene in the 16 cross-
section States is equal to 0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)
Occurrence Summary and Use Support Document
367
March 2002

-------
4.5.4.1 Stage 1 Analysis Occurrence Findings

Table 4.5-3 illustrates the occurrence of 1,1-dichloroethylene in drinking water for the public water
systems in the 16-State cross-section.  Based on the 16-State cross-section data, 0.236% of ground water
or surface water PWSs (45 systems) had analytical results of 1,1-dichloroethylene exceeding the MCL
(0.007 mg/L). Approximately 0.309% of PWSs (59 systems) had at least one analytical result greater
than 0.005 mg/L.  The percentage of PWSs with results greater than 0.001 mg/L was equal to 0.869%
(166 systems). Approximately 1.18% of PWSs (225 systems) in the 16 States had at least one analytical
result greater than 0.0005 mg/L.

Less than 1% of ground water systems had analytical results greater than any of the four thresholds.
Approximately 0.211% of ground water systems (37 systems) had results exceeding the MCL.  The
percentage of ground water systems in the 16 States with at least one analytical result greater than 0.005
mg/L was equal to 0.273% (48 systems).  Approximately 0.819% of ground water systems (144 systems),
and 0.944% of ground water systems (166 systems) had any analytical results exceeding 0.001 mg/L, and
0.0005 mg/L, respectively.

For surface water PWSs, 0.525% (8 systems) had analytical results exceeding 0.007 mg/L.
Approximately 0.721% of surface water systems (11 systems) had results exceeding 0.005 mg/L. The
percentage of surface water systems with at least one analytical result greater than 0.001 mg/L was equal
to 1.44% (22 systems). Approximately 1.90% of surface water systems (29 systems) had analytical
results of 1,1-dichloroethylene 0.0005 mg/L.
Table 4.5-3:  Stage 1 1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.007
0.005
0.001
0.0005
Percent of Systems
Exceeding Threshold
0.211%
0.273%
0.819%
0.944%
Number of Systems
Exceeding Threshold
37
48
144
166

Surface Water
0.007
0.005
0.001
0.0005
0.525%
0.721%
1.44%
1.90%
8
11
22
29

Combined Ground &
Surface Water
0.007
0.005
0.001
0.0005
0.236%
0.309%
0.869%
1.18%
45
59
166
225
Occurrence Summary and Use Support Document
368
March 2002

-------
Reviewing 1,1-dichloroethylene occurrence in the 16 cross-section States by PWS population served
(Table 4.5-4) shows that approximately 8.69% of the population (almost 9.3 million people) was served
by PWSs with analytical detections of 1,1-dichloroethylene greater than the MCL (0.007 mg/L).
Approximately 9.44% of the population (over 10 million people) was served by PWSs with at least one
analytical result greater than 0.005 mg/L. The percentage of population served by systems with results
greater than 0.001 mg/L was equal to 12.8% (about 13.6 million people). Approximately 14.4% of the
population (about 15.3 million people) was exposed to levels of 1,1-dichloroethylene greater than 0.0005
mg/1.

A much greater proportion of population was served by surface water systems with threshold
exceedances, as compared to ground water systems with threshold exceedances.  The percentage of
population served by ground water systems and surface water systems with analytical detections of 1,1-
dichloroethylene greater than the MCL was equal to 2.53% (almost 1.2  million people), and 13.3% (over
8 million people), respectively. Relative to 0.005 mg/L, 3.42% of the ground water population (over 1.5
million people) was served by systems exceeding that limit, whereas 13.9% of the surface water
population (approximately 8.5 million people) was served by systems exceeding 0.005 mg/L. In
addition, 7.75% of the population was served by ground water systems with results that exceeded 0.001
mg/L, while 16.5% of the  population was served  by surface water systems with results exceeding 0.001
mg/L. Approximately 9.30% of the population served by ground water systems was exposed to levels of
1,1-dichloroethylene greater than 0.0005 mg/L, compared to 18.2% of the population served by surface
water systems exposed to  levels greater than 0.0005 mg/L.
Table 4.5-4:  Stage 1 1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.007
0.005
0.001
0.0005
Percent of Population
Served by Systems
Exceeding Threshold
2.53%
3.42%
7.75%
9.30%
Total Population Served by
Systems Exceeding Threshold
1,152,600
1,556,600
3,528,900
4,237,200

Surface Water
0.007
0.005
0.001
0.0005
13.3%
13.9%
16.5%
18.2%
8,114,700
8,506,200
10,088,600
11,098,300

Combined Ground &
Surface Water
0.007
0.005
0.001
0.0005
8.69%
9.44%
12.8%
14.4%
9,267,400
10,062,800
13,617,500
15,335,500
Occurrence Summary and Use Support Document
369
March 2002

-------
4.5.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.5-5 and 4.5-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented. (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Seventeen (0.0911% of) systems nationally were estimated to have a mean concentration greater than
0.007 mg/L.  Twenty-one (0.111% of) systems nationally were estimated to exceed a mean concentration
of 0.005 mg/L.  Approximately 67 (0.350% of)  systems were estimated to exceed 0.001 mg/L, while 118
(0.616% of) systems had estimated mean concentrations exceeding 0.0005 mg/L.

Approximately  15 (0.0875% of) ground water systems had estimated mean concentrations of 1,1-
dichloroethylene greater than 0.007 mg/L. An estimated 19 (0.109% of) ground water systems had
estimated mean concentrations greater than 0.005 mg/L.  Approximately 62 (0.351% of) ground water
systems and 108 (0.613% of) ground water systems had estimated mean concentrations exceeding 0.001
mg/L and 0.0005 mg/L, respectively.

Only 2 (0.132% of) surface water systems in the 16 States was estimated to have a mean concentration
greater than 0.007 mg/L.  Two (0.134% of) surface water systems had an estimated mean concentration
greater than 0.005 mg/L.  Approximately 5 (0.329% of) surface water systems had estimated mean
concentrations of 1,1-dichloroethylene greater than 0.001 mg/L.  Ten (about 0.657% of) surface water
systems had estimated mean concentration values greater than 0.0005 mg/L.
Table 4.5-5:  Stage 2 Estimated 1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.007
0.005
0.001
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0875%
0.1090%
0.351%
0.613%
Range
0.0626%-0.120%
0.0740%-0.148%
0.273%-0.433%
0.507%-0.723%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
15
19
62
108
Range
11-21
13-26
48-76
89 - 127

Surface Water
0.007
0.005
0.001
0.0005
0.132%
0.134%
0.329%
0.657%
0.131% -0.1311%
0.131% -0.197%
0.262% - 0.459%
0.459% - 0.92%
2
2
5
10
2-2
2-3
4-7
7-14

Combined Ground
& Surface Water
0.007
0.005
0.0911%
0.1110%
0.0681% -0.120%
0.0786% -0.147%
17
21
13-23
15-28
Occurrence Summary and Use Support Document
370
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.001
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.350%
0.616%
Range
0.272% - 0.424%
0.513% -0.717%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
67
118
Range
52-81
98-137
Reviewing 1,1-dichloroethylene occurrence by PWS population served (Table 4.5-6) shows that
approximately 0.331% of population served by all PWSs in the 16 States (an estimate of approximately
352,700 people) was potentially exposed to 1,1-dichloroethylene levels above 0.007 mg/L.
Approximately 0.340% of the total population served by PWSs in the 16 States (exposure to an estimated
362,900 people) was exposed to levels of 1,1-dichloroethylene exceeding the limit of 0.005 mg/L, and
0.898% of the population served by PWSs in the 16 States (exposure to an estimated 956,800 people)
was potentially exposed to levels exceeding 0.001 mg/L. The percentage of population exposed
substantially increased to 8.58% (over 9.1 million people) when evaluated relative to 0.0005 mg/L.

When the ground water PWSs were evaluated relative to a thresholds of 0.007 mg/L, 0.005 mg/L, 0.001
mg/L, and 0.0005 mg/L, the percentages of population exposed in the 16 States were equal to 0.398% (an
estimated 181,200 people), 0.417% (an estimated 190,000 people), 0.973% (an estimated 443,000
people), and 2.08% (an estimated 945,000 people), respectively.

The percentage of population served by surface water systems in the 16 States that exceeded the limit of
0.007 mg/L was 0.281% (exposure to an estimated 171,400 people). The percentage that exceeded 0.005
mg/L was 0.283% (exposure to an estimated 172,900 people in the 16 States).  Approximately 0.841% of
the population served by surface water systems was potentially exposed to 1,1-dichloroethylene levels
greater than 0.001 mg/L (exposure to an estimated 513,800 people).  The percentage of surface water
systems that exceeded the limit of 0.0005 mg/L was  13.4% (exposure to an estimated 8.2 million people
in the 16 States).
Table 4.5-6:  Stage 2 Estimated 1,1-Dichloroethylene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.007
0.005
0.001
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.398%
0.417%
0.973%
2.08%
Range
0.370% - 0.459%
0.385% - 0.479%
0.727% -1.435%
1.593% -2.68%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
181,200
190,000
443,000
945,000
Range
168,600-209,000
175,400-218,300
331,100-653,900
725,900-1,222,200

Surface Water
0.007
0.005
0.001
0.281%
0.283%
0.841%
0.280% - 0.280%
0.280% -0.291%
0.611%- 1.762%
171,400
172,900
513,800
171,700-171,700
171,700 - 178,300
374,400-1,079,100
Occurrence Summary and Use Support Document
371
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
13.4%
Range
13.0% -14.5%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
8,195,300
Range
7,986,100-8,861,900

Combined Ground
& Surface Water
0.007
0.005
0.001
0.0005
0.331%
0.340%
0.898%
8.58%
0.319%-0.357%
0.325%-0.380%
0.673%-1.326%
8.24%-9.19%
352,700
362,900
956,800
9,141,600
340,400-381,100
347,200 - 405,500
719,000-1,416,300
8,797,000-9,811,700
4.5.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 59 PWSs nationally serving approximately 704,600 people could be exposed to
1,1-dichloroethylene concentrations above 0.007 mg/L.  About 72 systems serving almost 725,100 people
had estimated mean concentrations greater than 0.005 mg/L. Approximately 227 systems serving over
1.9 million people nationally were estimated to have mean 1,1-dichloroethylene concentrations greater
than 0.001 mg/L. A total of 401 systems serving over 18 million people had estimated mean
concentrations greater than 0.0005 mg/L.  (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)

For ground water systems, an estimated 52 PWSs serving about 341,000 people nationally had mean
concentrations greater than 0.007  mg/L. Approximately 65 systems serving about 357,500 people
nationally had estimated mean concentration values that exceeded 0.005 mg/L.  About 209 ground water
systems serving almost 833,500 people had estimated mean concentrations greater than 0.001 mg/L. An
estimated 364 ground water PWSs nationally exposed over 1.7 million people to concentrations of 1,1-
dichloroethylene greater than 0.0005 mg/L.

Approximately 1 surface water system serving 357,400 people was estimated to have a mean
concentration of 1,1-dichloroethylene above 0.007 mg/L. About 7 surface water systems serving 360,500
people had estimated mean concentrations greater than 0.005 mg/L. An estimated 18 surface water
systems serving approximately 1.1 million people had mean concentrations greater than 0.001 mg/L.
Approximately 73 surface water PWSs serving over 17 million people had estimated mean
concentrations of 1,1-dichloroethylene greater than 0.0005 mg/L.
Table 4.5-7:  Estimated National 1,1-Dichloroethylene Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.007
0.005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
52
65
Range
37-71
44-88
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
341,000
357,500
Range
317,100-392,900
329,800-410,500
Occurrence Summary and Use Support Document
372
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.001
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
209
364
Range
162-257
301-430
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
833,500
1,777,900
Range
622,600-1,229,500
1,364,900-2,298,000

Surface Water
0.007
0.005
0.001
0.0005
1
7
18
37
7-7
7-11
15-26
26-51
357,400
360,500
1,071,200
17,087,200
357,000 - 357,000
357,000 - 370,600
778,500 - 2,243,500
16,603,400-18,424,100

Combined Ground
& Surface Water
0.007
0.005
0.001
0.0005
59
72
227
401
44-78
51-95
177-276
334 - 467
704,600
725,100
1,911,700
18,265,500
678,900 - 760,000
692,500 - 808,600
1,433,800-2,824,500
17,543,400-19,566,900
4.5.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of 1,1-dichloroethylene in drinking water are
also reviewed. Previously compiled occurrence information, from an OGWDW summary document
entitled "Occurrence of 1,1-Dichloroethylene in Drinking Water, Food, and Air" (JRB Associates, 1983),
is presented in the following section.  This variety of studies and information are presented regarding
levels of 1,1-dichloroethylene in drinking water, with the scope of the reviewed studies ranging from
national to regional.  Note that none of the studies presented in the following section provide the
quantitative analytical results or comprehensive coverage that would enable direct comparison to the
occurrence findings estimated with the cross-section occurrence data presented in Section 4.5.4.  These
additional studies, however, do enable a broader assessment of the Stage 2 occurrence estimates
presented for this Six-Year Review. All the following information in Section 4.5.5 is taken directly from
"Occurrence of 1,1-Dichloroethylene in Drinking Water, Food, and Air" (JRB Associates, 1983).

JRB Associates (1983) found two major types of data available that are potentially useful for describing
the occurrence of 1,1-dichloroethylene in the nation's public drinking water supplies. First, there are
several Federal surveys in which a number of public water supplies from throughout the U.S. were
selected for analysis of chemical contamination, including 1,1-dichloroethylene. Second, data are
available from State surveys and from State  investigations of specific incidents of known or suspected
contamination of a supply. For accomplishing the basic objectives of this study, namely to estimate the
number of public water supplies nationally within the various source and size categories contaminated
with 1,1-dichloroethylene, the distribution of 1,1-dichloroethylene concentrations in those supplies, and
the number of individuals exposed to those concentrations, it was determined that the Federal survey data
provides the most suitable data base.  The State data tend to be poorly described with respect to the
source and size categories of the supplies examined and the sampling and analysis methods used  for
determining contaminant levels.  The lack of source and system size information precludes using the data
for estimating levels in public water supplies of similar characteristics.  The absence of details on
sampling and analysis methods precludes evaluating those data for their qualitative and quantitative
Occurrence Summary and Use Support Document
373
March 2002

-------
reliability.  Also, because much of the State data are from investigations in response to incidents of
known or suspected contamination (e.g., spills), they were judged to be not representative of contaminant
levels in the nation's water supplies in general. Although they are not used with the Federal data for the
purpose of estimating contamination levels nationally, the available State data are presented here to
provide some additional perspective on 1,1-dichloroethylene occurrence in drinking water.

Data are presented only on drinking water samples taken from a consumer's tap (i.e., distribution water
samples) or on treated water samples taken at the water supply (i.e., finished water samples) because
these are considered to be most representative of the water consumed by the public. No data on raw (i.e.,
untreated) water are presented. It is recognized that for some groundwater supplies where no treatment
of the water occurs, samples identified as raw may be representative of water consumed by the users of
the supply. However, it was generally not possible to differentiate between those groundwater supplies
that do and those that do not treat raw water from the available survey data

4.5.5.1  Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Two Federal drinking water surveys provide data on 1,1-dichloroethylene: the National Screening
Program for Organics in Drinking Water (NSP) and the Groundwater Supply Survey (GWSS). The terms
used in this report are those used in the individual surveys, recognizing that they may not always
correspond to strict technical definitions.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March 1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic  chemical contaminants.  Data are available for 1,1-dichloroethylene
on finished water samples from 12 groundwater and 103 surface water supplies.

The Groundwater Supply Survey (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al. 1983, as cited in JRB Associates, 1983). It was hoped that this study would
stimulate State efforts toward the detection and control of groundwater contamination and the
identification of potential chemical  "hot spots." A total of 945 systems were sampled, of which 466 were
chosen at random.  The remaining 479 systems were chosen nonrandomly based on information from
States encouraged to identify locations believed to have a higher than normal probability of VOC
contamination (e.g., locations near landfills or industrial activity).  The file provided a single analytical
result for each supply sampled. One sample of finished water was collected from each supply at a point
near the entrance to the distribution system.

Each of the drinking water surveys was evaluated with respect to the validity of the reported occurrence
data for a number of organic chemicals, including 1,1-dichloroethylene.  The  evaluations were carried
out by analyzing information about the procedures used for collection and analysis of samples as well as
the quality control protocols used. The analyzed compounds dealt with in each study were assigned one
of three possible ratings: quantitatively acceptable, qualitatively acceptable (i.e., the substance measured
was 1,1-dichloroethylene), and totally unacceptable. In the case of 1,1-dichloroethylene, a qualitatively
acceptable rating was given for data from the NSP because of suspected biodegradation of the samples,
which were held unrefrigerated for prolonged periods before analysis. 1,1-Dichloroethylene values in
excess of the quantitation limit reported for some samples in these  studies are qualitatively valid and can
be taken as minimum values, representative of samples which  probably originally contained 1,1-
dichloroethylene at higher concentrations. In the case  of the GWSS, all data were rated both
quantitatively and qualitatively acceptable.

Occurrence Summary and Use Support Document         374                                     March 2002

-------
4.5.5.2 Groundwater - Federal Surveys

The National Screening Program for Organics in Drinking Water (NSP) and the Groundwater Supply
Survey (GWSS) both contain data concerning the levels of 1,1-dichloroethylene in groundwater supplies
from across the country.

Twelve groundwater supplies were tested for, 1,1-dichloroethylene contamination in the NSP.  Of these
12 systems, only one was found to be contaminated with 1,1-dichloroethylene at a level of 0.2 |ig/L. The
quantification limit for 1,1-dichloroethylene was 0.1 |ig/L.

In the GWSS, 9 of the 456 randomly chosen water systems serving 25 or more individuals were
contaminated with 1,1-dichloroethylene, at concentrations ranging from 0.22-6.3 |ig/L. The three
systems with the highest values were contaminated at 2.1, 2.2, and 6.3 |ig/L.  Of the 9 positive systems, 5
were from systems serving populations in excess of 10,000 people. The average for all randomly chosen
systems was 1.4 |ig/L with a standard deviation of 2.0 |ig/L; the median was 0.28 |ig/L.  Of the 473
nonrandom locations sampled serving 25 or more individuals, 15 were contaminated with 1,1-
dichloroethylene, at concentrations between 0.22-3 |ig/L, the  highest values being 0.04, 1.2, and 3 |ig/L.
Of the 15 positive samples,  10 were from systems serving populations in excess of 10,000 people. The
average 1,1-dichloroethylene level for the nonrandom systems was 0.59 |ig/L with a standard deviation of
0.71 |ig/L; the median value was 0.35 |ig/L. The minimum quantitation limit for 1,1-dichloroethylene
was 0.2 |ig/L.

4.5.5.3 Groundwater - State Data

Three States (California, Massachusetts, and New Jersey) provided the USEPA with information
concerning 1,1-dichloroethylene contamination in groundwater supplies. Analytical results for samples
from three locations in California ranged from undetectable to 50 |ig/L.
1,1-Dichloroethylene levels ranged from undetectable to 261  |ig/L in 22 samples from six Massachusetts
cities. New Jersey provided data from  19 samples from Fair Lawn and Mahwah;  12 of the samples
contained undetectable  1,1-dichloroethylene while the other seven samples ranged from 2.7-3.5 |ig/L at
Mahwah and 0.9-27 |ig/L at Fair Lawn.

4.5.5.4 Surface Water - Federal Surveys

Only one federal survey, the National Screening Program (NSP), contains data concerning 1,1-
dichloroethylene levels in surface water supplies. During this survey,  106 drinking water systems were
analyzed for 1,1-dichloroethylene between June 1977 and March 1981. Of these, two systems contained
detectable levels of 1,1-dichloroethylene, with concentrations of 0.2 and 0.51 |ig/L.

4.5.5.5 Surface Water - State Data

Data from two surface water samples were reported from Niagara Falls, New York. Of the two finished
water samples, one contained no detectable 1,1-dichloroethylene.  The other sample was contaminated
with 1,1-dichloroethylene at 0.22 |ig/L.

4.5.5.6 Projected National Occurrence of 1,1-Dichloroethylene in Public Water Supplies

As reported by the JRB Associates (1983) report, public water systems fall into two major categories
with respect to water source (surface water and groundwater) and into five size categories and twelve

Occurrence Summary and Use Support Document         375                                     March 2002

-------
subcategories according to the number of individuals served. The JRB Associates (1983) report
presented estimates of both the number of drinking water supplies nationally within each of the
source/size categories expected to have 1,1-dichloroethylene present, and of the concentration of 1,1-
dichloroethylene expected to be present in those supplies.

The key features of the methodology used and assumptions made to develop the national estimates are
summarized here. The estimates are based on the data from the Federal surveys only.  The State data
were not included for several reasons.  Generally, these data are from a few States and were not
considered to be geographically representative. There was also a general lack of data on the population
served by systems measured, the type of water sampled, and the methodologies used to sample, identify,
and measure 1,1 -dichloroethylene.

The Federal survey data from the NSP and GWSS were pooled together for developing the national
projections. It was assumed in combining these surveys that the resulting data base would be
representative of the nation's water supplies.  In the case of the GWSS data, both the random and
nonrandom samples were included in the projections because a statistical test of the GWSS data showed
no statistically significant difference in the frequency of occurrence of positive values or the mean of the
positive values of vinyl chloride between the random and nonrandom samples.

Ideally, adequate survey data would be available to develop the national projections separately for each
of the twelve system size categories within the groundwater and surface water groups; however, the
available data were too limited for this. JRB (1983) consolidated some of the size categories to have
sufficient data for developing the projections. In consolidating data from various size categories,
consideration was given to the potential for there being statistically significant differences in the
frequency of occurrence of 1,1-dichloroethylene as a function of system size. The consolidation of size
categories therefore involved a balancing of the need to group size categories together to have an
adequate data base for developing the national projections against the need to treat size categories
separately in order to preserve the influence of system size as a determinant of contamination potential.
The consolidation of size categories also took into account EPA's classification of systems into the five
major groups as very small (25-500), small (501-3,300), medium (3,301-10,000), large  (10,001-100,000),
and very large (> 100,000) (Kuzmack, 1983, as cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national level were tested and
an appropriate model selected.  In the case of 1,1-dichloroethylene, the multinominal method was used.
The frequency of contamination of groundwater and surface water systems at various concentrations was
determined for each consolidated size category. For completing the national estimates, it was assumed
that the frequency of contamination observed for each consolidated category was directly applicable to
each of the system sizes comprising it.

In the JRB  Associates (1983) report, it is noted that some of the data used in computing the national
estimates are from  samples held for a prolonged period of time prior to analysis, with possible
biodegradation of 1,1-dichloroethylene. Therefore, these projections of national occurrence may
underestimate actual contaminant levels.

4.5.5.6.1 Groundwater Supplies

JRB Associates (1983) reported data that were available for a total of 938 supplies from the combined
surveys.  Of these,  25 supplies were reported to have 1,1-dichloroethylene present, at concentrations
ranging from 0.2 |ig/L to 6.3 |ig/L.  Based on the overall distribution of positive values  and maximum

Occurrence Summary and Use Support Document         376                                      March 2002

-------
possible values for those supplies in which 1,1-dichloroethylene was not found, 0.2 |ig/L was selected as
the common minimum quantifiable concentration for the combined survey data. That is, quantitative
projections are made of supplies at several concentration ranges > 0.2 |ig/L, while only a total number for
supplies expected to have either no 1,1-dichloroethylene or levels below 0.2 |ig/L can be determined.

When the twelve size categories were consolidated into the five major EPA groupings, there was an
apparent trend in the frequency of values > 0.2 |ig/L as a function of size:
Very small
Small
Medium
Large
Very large
Overall
0.9%
2.0%
2.6%
4.5%
2.6%
2.7%
(2/230)
(4/203)
(4/154)
(14/312)
(1/39)
(25/938)
A test for statistical significance revealed that at the a = 0.05 level, the difference among the very small,
small, and medium categories was not significant; nor was the difference between the large and very
large size categories. However, the combined very small, small, and medium categories and the
combined large and very large categories were found to be different. Therefore, two consolidated
categories were selected for developing the national estimates:

                               Very small/small/medium (25-10,000)
                                   Large/very large (> 10,000)

As noted previously, the frequency of occurrence of 1,1-dichloroethylene at various concentrations was
determined for the two consolidated groups and then applied to the number of supplies nationally within
each of the size categories comprising each group.

Approximately 1.8% of the total groundwater supplies in the United States (858 supplies, with a range of
366-1,349), are expected to have 1,1-dichloroethylene at levels of > 0.2 |ig/L; the remaining 47,600
supplies have either no 1,1-dichloroethylene or levels < 0.2 |ig/L.  It is estimated that 81 supplies (range
of 0-237) are expected to have 1,1-dichloroethylene levels > 5 |ig/L; while no supplies are expected to
have levels > 10 |ig/L.  Most of the supplies with 1,1-dichloroethylene levels > 0.2 |ig/L are expected to
be in the smaller size categories.  Although, as noted previously, the frequency of 1,1-dichloroethylene
occurrence appears to increase with increasing system size, the number of systems affected nationally is
greater for the small sizes because there are many more small systems in existence.

4.5.5.6.2 Surface Water Systems

Data are available for a total of 103 surface water supplies.  Of these, 2 supplies were reported by JRB
Associates  (1983) to have  1,1-dichloroethylene present at concentrations of 0.2 |ig/L and 0.5 |ig/L.  All
but one of the  103 surface  water supplies sampled in the NSP fall in the large and very large size
categories (i.e., serving > 10,000 people). Consequently, it was not possible to evaluate the frequency of
occurrence of 1,1-dichloroethylene as a function of system size for the very small, small, and medium
size categories. In the large size category (serving 10,001-100,000), none of the 19 supplies were found
to have 1,1-dichloroethylene present, while 2 of 83 in the very large category (serving > 100,000) had
1,1-dichloroethylene present.  The difference in the frequency of occurrence between the large and very
large groups was not statistically significant at the a = 0.05 level.  For the purpose of the national
estimates, the groundwater supplies were consolidated into two groups:

Occurrence Summary and Use Support Document         377                                      March 2002

-------
                               Very small/small/medium (25-10,000)
                                    Large/very large (> 10,000)

Again, the frequency of occurrence of 1,1-dichloroethylene at various concentrations was determined for
the two consolidated groups and then applied to the number of supplies nationally within each of the size
categories comprising each group.

Thirty-five surface water supplies (range of 0-81), approximately 0.3% of the total surface water systems
in the United States, are expected to have 1,1-dichloroethylene at levels > 0.2 |ig/L; the remaining 11,167
supplies have either no 1,1-dichloroethylene or levels < 0.2 |ig/L.  It is estimated that no surface water
supplies will have levels > 5 |ig/L.  Note that all of the supplies with levels in the 0.2-5 |ig/L range are in
the large and very large category. The estimate of no occurrence in the smaller categories  is based on
only one sample and is probably not a reliable estimate.

4.5.6 Conclusion

1,1-Dichloroethylene does not occur naturally, but it is found as the result of the breakdown of
polyvinylidene chloride products in landfills. Virtually all 1,1-dichloroethylene produced  is used in the
production of copolymers with vinyl chloride or acrylonitrile, and only a small percentage  (4%) is used
as chemical intermediates. The primary producers of 1,1-dichloroethylene  are Dow Chemical and
Pittsburgh Paint and Glass (PPG) Industries, and the most recent data available suggests that production
is relatively stable.  1,1-Dichloroethylene is also a TRI chemical. Under the TRI program, industrial
releases of 1,1-dichloroethylene have been recorded since 1988 in 19 States and Puerto Rico. 1,1-
Dichloroethylene was an analyte  for the NAWQA and NURP ambient occurrence studies.  In the
NAWQA study, 1,1-dichloroethylene was detected in 3.0% of urban wells  and 0.29% of rural wells, with
median detection values of 0.9 |ig/L and  1.0 |ig/L, respectively.  In the Stage 2 analysis of 16-State
occurrence of 1,1-dichloroethylene, 0.0144% of combined ground water and surface water systems
serving 0.0135% of the population exceeded the MCL of 0.007 mg/L.  Nationally, 9 ground water and
surface water systems combined (serving approximately 28,800  people) are estimated to have levels
greater than the MCL.

The  16-State cross-section was designed to be nationally representative based upon VOC,  SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally,  1,1-dichloroethylene is manufactured and/or processed in 14 States and has
TRI  releases have been reported in  19 States.  1,1-Dichloroethylene is manufactured and/or processed in
6 out of the 16 cross-section States  and has TRI releases in 10 of the 16 cross-section States. The cross-
section should adequately represent occurrence of 1,1-dichloroethylene on  a national scale based upon
the use, production, and release patterns of the 16-State cross-section in relation to the patterns observed
for all  50 States.

4.5.7 References

Agency for Toxic Substances  and Disease Registry (ATSDR). 1994. Toxicological Profile for 1,1-
       Dichloroethene. U.S. Department of Health  and Human Services, Public Health Service.  174
       pp. + Appendices. Available on the Internet at http://www.atsdr.cdc.gov/toxprofiles/tp39.pdf

JRB Associates. 1983. Occurrence of 1,1-Dichloroethylene in Drinking Water, Food and Air - DRAFT.
       Draft report submitted to EPA for review December 18, 1983.
Occurrence Summary and Use Support Document          378                                     March 2002

-------
Kuzmack, A.M. 1983. Memorandum: Characterization of the water supply industry (FY82).
       Washington, D.C.: Office of Water, USEPA. May 16, 1983.

Lopes, T.J. and S.G. Dionne.  1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater. U.S. Geological Survey Open-File Report 98-409. 67
       pp.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, R.M. Clawges, and J.S. Zogorski.  1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci. and Tech. 33(23):4176-4187.

USEPA. 2000. TRIExplorer:  Trends.  Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001.  National Primary Drinking Water Regulations - Consumer Factsheet on: 1,1-
       Dichloroethylene. Office of Ground Water and Drinking Water, USEPA. Available on the
       Internet at http://www.epa.gov/safewater/dwh/c-voc/! 1-dichl.html (Last updated 04/12/2001)

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Westrick, J.J., J.W. Mello, and R.F. Thomas. 1983. The Ground Water Supply Survey summary of
       volatile  organic contaminant occurrence data.  EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         379                                    March 2002

-------
4.6    Dichloromethane
Table of Contents

4.6.1 Introduction, Use and Production 	  381
4.6.2 Environmental Release  	  382
4.6.3 Ambient Occurrence 	  383
4.6.4 Drinking Water Occurrence Based on the 16-State Cross-Section	  384
4.6.5 Additional Drinking Water Occurrence Data 	  389
4.6.6 Conclusion	  389
4.6.7 References  	  390
Tables and Figures

Table 4.6-1:  Dichloromethane Manufacturers and Processors by State	  381

Table 4.6-2:  Environmental Releases (in pounds) for Dichloromethane
       in the United States, 1988-1999  	  383

Table 4.6-3:  Stage 1 Dichloromethane Occurrence Based on 16-State Cross-Section -
       Systems	  384

Table 4.6-4:  Stage 1 Dichloromethane Occurrence Based on 16-State Cross-Section -
       Population	  385

Table 4.6-5:  Stage 2 Estimated Dichloromethane Occurrence Based on 16-State
       Cross-Section - Systems	  387

Table 4.6-6:  Stage 2 Estimated Dichloromethane Occurrence Based on 16-State
       Cross-Section - Population	  388

Table 4.6-7:  Estimated National Dichloromethane Occurrence - Systems and Population Served . . .  389
Occurrence Summary and Use Support Document         380                                    March 2002

-------
4.6.1  Introduction, Use and Production

Dichloromethane, also known as methylene chloride, has the chemical formula CH2C12.  It is a colorless
liquid that has a mild sweet odor, evaporates easily, and does not burn easily.  Dichloromethane does not
appear to occur naturally in the environment. It is made from methane gas or wood alcohol.  Most of the
dichloromethane released to the environment results from its use as an end product by various industries
and the use of aerosol products and paint removers in the home (ATSDR, 2000).

The breakdown of dichloromethane use is as follows: paint strippers and removers, 25%; propellant in
aerosols, 25%; as a process solvent in the manufacture of drugs, pharmaceuticals, and film coatings,
20%; as a metal cleaning and finishing solvent, 10%; in electronics manufacturing, 10%; and as an agent
in urethane foam blowing, 10% (ATSDR, 2000). Dichloromethane has been the primary substitute for
Chlorofluorocarbon 11 as a flexible polyurethane foam blowing agent (Chemexpo, 2001).

Dichloromethane is also used as a solvent in the production of polycarbonate resins and triacetate fibers,
in film processing, in ink formulation, and as an extraction solvent for spice oleoresin, caffeine and hops.
It is registered in the U.S. as an insecticide for commodity fumigation of strawberries, citrus fruits, and a
variety of grains (NSC, 2001).

Many consumer products may also contain dichloromethane. These include spray shoe polish, water
repellant/protectors, spot removers, wood floor and panel cleaners, contact cement, superglue, spray
adhesives, adhesive removers, silicone lubricants, specialized electronic cleaners, wood stains, paint
thinners, aerosol rust removers,  and glass frosting/artificial snow (NSC, 2001). Concerns over health and
environmental issues have led to a decrease in use of dichloromethane in some products, such as
consumer aerosol products and as a decaffeinator.  Its use in hair sprays was banned in 1989  (ATSDR,
2000).

Production of dichloromethane increased steadily through the 1970s and early 1980s, with a peak
production volume of about 620 million pounds in  1984. Demand has declined in the years since the
early 1980s, primarily because manufacturers have moved towards water-based aerosol systems in
anticipation of further regulation of dichloromethane. In 1994, the most recent year for which
information is available, production of dichloromethane was 403 million pounds. Dichloromethane is
produced in a number of States, and the major manufacturers are Dow Chemical in Freeport, TX and
Plaquemine, LA and Vulcan Materials Company in Geismar, LA and Wichita, KS (ATSDR,  2000).

Table 4.6-1 shows the number of facilities in each State that manufacture and process dichloromethane,
the intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 2000).
Table 4.6-1: Dichloromethane Manufacturers and Processors by State
State"
AL
AR
AZ
CA
CO
CT
DE
FL
Number of facilities
12
7
5
23
4
8
9
12
Range of maximum amounts on site in
thousands of pounds'"
100-999,999
1,000-999,999
100-999,999
0-9,999,999
1,000-999,999
1,000-99,999
10,000-99,999
100-999,999
Activities and uses0
1,4,7,8,10,11,12,13
2,3,7,10,11,12,13
8,10,11,12,13
2,3,4,8,9,10,11,12,13
8,10,11,13
2,3,10,11,12,13
8,11
8,10,11,12,13
Occurrence Summary and Use Support Document          381                                     March 2002

-------
 State"
                    Number of facilities
                Range of maximum amounts on site in
                thousands of pounds'"
                                                                         Activities and uses0
GA
IA
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
NE
NH
NJ
NM
NV
NY
OH
OK
OR
PA
PR
RI
SC
TN
TX
UT
VA
WA
WI
wv
18
10
23
21
8
9
8
10
5
1
17
8
13
11
1
17
3
3
19
2
3
14
17
9
6
17
12
2
14
15
27
6
10
10
14
5
0-999,999
100-999,999
0-9,999,999
100-9,999,999
100-999,999
100-99,999
1,000-9,999,999
1,000-999,999
1,000-99,999
1,000-9,999
100-999,999
100-99,999
1,000-999,999
100-999,999
1,000-9,999
100-999,999
1,000-999,999
1,000-99,999
0-999,999
1,000-99,999
100-9,999
1,000-9,999,999
100-9,999,999
100-999,999
1,000-999,999
1,000-999,999
1,000-9,999,999
10,000-99,999
100-999,999
100-9,999,999
0-49,999,999
1,000-999,999
1,000-999,999
0-99,999
1,000-999,999
0-9,999,999
2,3,8,9,10,11,12,13
8,10,11,12,13
1,2,3,5,7,8,9,10,11,12,13
1,2,3,5,8,9,10,11,12,13
1,2,3,4,8,10,11,12,13
1,5,7,8,10,11,13
1,4,5,6,7,8,10,12,13
2,3,8,9,10,11,12,13
8,10,12,13
13
1,2,3,4,5,7,8,10,11,12,13
2,3,8,11,12,13
4,8,10,11,12,13
1,2,3,8,11,12,13
12
1,2,4,5,7,8,10,11,12,13
11,13
8,12,13
1,2,3,4,7,8,9,10,11,12,13
8,12
1,2,3,8,11,13
2,3,4,7,8,9,10,11,12,13
2,3,7,8,10,11,12,13
8,10,11,13
2,4,8,10,11,12,13
8,9,10,11,12,13
1,2,5,6,8,9,10,11,12,13
10,13
7,8,10,11,12,13
8,9,10,11,12,13
1,2,3,4,5,6,7,8,9,10,11,12,13
8,10,11,12,13
8,10,11,12,13
1,6,7,10,11,12,13
1,4,5,8,10,11,12,13
8,11,13
aPost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a byproduct
6. As an impurity
7. As a reactant
8. As a formulation component
9. As an article component
10. For repackaging only
11. As a chemical processing aid
12. As a manufacturing aid
13. Ancillary or other uses
Source: AT SDR, 2000 compilation ofTRI982000 data
4.6.2 Environmental Release

Dichloromethane is listed as a Toxics Release Inventory (TRI) chemical. Table 4.6-2 illustrates the
environmental releases for dichloromethane from 1988 - 1999. (Dichloromethane data are only
available for these years.) Air emissions constitute the vast majority of the on-site releases, with a
substantial decrease over the years.  Surface water discharges have also generally declined, with a slight
upturn since 1997.  Underground injection fluctuated between about 1 and 2 million pounds from 1988-
1995, and releases have steadily decreased since. Releases to  land (such as spills or leaks within the
boundaries of the reporting facility) have fluctuated and show  no apparent trend.  Off-site releases
(including metals or metal compounds transferred off-site) have declined from the releases of 1988-1991,
and numbers since have remained between about 150,000 and 300,000 pounds. The decrease in air
Occurrence Summary and Use Support Document
                           382
March 2002

-------
emissions has contributed to a considerable decrease in dichloromethane total on- and off-site releases
over the course of monitoring. These TRI data for dichloromethane were reported from 46 States and
Puerto Rico. Forty-one States and Puerto Rico reported TRI data every year (USEPA, 2000). Of the 46
States, 15 are included in the 16 State cross-section (used for analyses of dichloromethane occurrence in
drinking water; see Section 4.6.4).  (For a map of the 16-State cross-section, see Figure 1.3-1.)
Table 4.6-2: Environmental Releases (in pounds) for Dichloromethane in the United States, 1988-
1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
35,556,474
40,302,462
48,315,635
54,143,761
58,305,923
63,714,204
65,499,529
75,011,911
84,224,528
101,000,812
125,604,691
129,124,529
Surface Water
Discharges
12,056
15,492
9,493
10,064
28,620
52,289
62,909
233,786
99,220
194,764
227,025
349,960
Underground
Injection
107,386
456,962
528,030
749,507
1,140,335
960,942
956,098
1,183,867
1,317,706
850,018
1,937,469
1,478,833
Releases
to Land
8,344
173,592
11,180
4,957
2,064
41,645
69,467
77,208
96,559
21,024
15,894
157,156
Off-Site Releases
154,374
267,633
230,405
168,475
180,137
314,976
144,479
209,916
502,795
1,001,707
1,530,916
7,806,328
Total On- &
Off-site
Releases
35,838,634
41,216,141
49,094,743
55,076,764
59,657,079
65,084,056
66,732,482
76,716,688
86,240,808
103,068,325
129,315,995
138,916,806
 Source: USEPA, 2000
4.6.3  Ambient Occurrence

Dichloromethane was detected in 8 out of 336 wells (2.4%) in urban areas of the local, State, and federal
data set compiled by NAWQA. The minimum and maximum concentrations detected were 0.2 |ig/L and
1.5 |ig/L, respectively. The median value of detection concentrations was 0.45 |ig/L.  Dichloromethane
was also detected in 20 of the 2,345 wells (0.86%) with analysis in rural areas. The minimum and
maximum concentrations detected were 0.2 |ig/L and 4 |ig/L, respectively. The median value of
detection concentrations was 0.45 |ig/L.  These data (urban and rural) represent untreated ambient ground
water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).

Dichloromethane was also an analyte in both the NURP and NPDES programs (Lopes and Dionne,
1998). A comparison of the results of the NURP and NPDES studies found that the frequency of
detection of dichloromethane in the NURP data was 11%, whereas for the NPDES data it was 7%. The
NURP study found dichloromethane in urban runoff. The minimum and maximum concentrations
detected were 4 |ig/L and 14.5 |ig/L, respectively, with no mean value reported. The NPDES related
investigations analyzing urban and highway runoff detected dichloromethane.  The minimum and
maximum concentrations detected were <0.2 |ig/L and 13 |ig/L, respectively, with no mean value
reported. The use of the land from which both sets of samples were taken was  unspecified.

4.6.3.1 Additional Ambient Occurrence Data

Additional studies of ambient data are unavailable for dichloromethane.
Occurrence Summary and Use Support Document
383
March 2002

-------
4.6.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of dichloromethane occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date.  These data were
evaluated relative to several concentration thresholds of interest: 0.005 mg/L; 0.0025 mg/L; 0.0005
mg/L; and 0.00025 mg/L.

All sixteen cross-section State data sets, with the exception of Montana, contained occurrence data for
dichloromethane. These data represent more than 170,000 analytical results from approximately 22,000
PWSs during the period from 1984 to  1998  (with most analytical results from 1992 to  1997). The
number of sample results and PWSs vary by State, although the State data sets have been reviewed and
checked to ensure adequacy of coverage and completeness.  The overall modal detection limit for
dichloromethane in the 16 cross-section States is equal to 0.0005 mg/L. (For details regarding the 16-
State cross-section, please refer to Section 1.3.5 of this report.)

4.6.4.1 Stage 1 Analysis Occurrence Findings

Table 4.6-3 illustrates the occurrence of dichloromethane in drinking water for the public water systems
in the 16-State cross-section. About 0.669% of all PWSs in the 16 States (144 systems)  had analytical
results of dichloromethane exceeding the MCL (0.005 mg/L). Approximately 1.72% of PWSs (370
systems) had at least one analytical result greater than 0.0025 mg/L. The percentage of PWSs with
results greater than 0.0005 mg/L was equal to 7.72% (1,663 systems). Approximately 8.51% of PWSs in
the 16 States (1,832 systems) had at least one analytical  result greater than  0.00025 mg/L.

Approximately 0.569% of ground water systems (114 systems) had results  exceeding the MCL.  The
percentage of ground water systems in the 16 States with at least one analytical result greater than 0.0025
mg/L was equal to 1.46% (292 systems). Approximately 7.15% of ground  water systems (1,432
systems), and 7.93% of ground water systems (1,587 systems) had any analytical results exceeding
0.0005 mg/L, and 0.00025 mg/L, respectively.

For surface water PWSs, 1.99% (30 systems) had analytical results exceeding 0.005 mg/L.
Approximately 5.16% of surface water systems (78 systems) had results exceeding 0.0025 mg/L. The
percentage of surface water systems with at least one analytical result greater than 0.0005 mg/L was
equal to 15.3% (231 systems). Approximately 16.2% of surface water systems (245 systems) had
analytical results of dichloromethane 0.00025 mg/L.
Table 4.6-3:  Stage 1 Dichloromethane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
0.00025
Percent of Systems
Exceeding Threshold
0.569%
1.46%
7.15%
7.93%
Number of Systems
Exceeding Threshold
114
292
1,432
1,587
Occurrence Summary and Use Support Document
384
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems
Exceeding Threshold

Surface Water
0.005
0.0025
0.0005
0.00025
1.99%
5.16%
15.3%
16.2%
30
78
231
245

Combined Ground &
Surface Water
0.005
0.0025
0.0005
0.00025
0.669%
1.72%
7.72%
8.51%
144
370
1,663
1,832
Reviewing dichloromethane occurrence by PWS population served (Table 4.6-4) shows that
approximately 9.50% of the 16-State population (about 10.5 million people) was served by PWSs with
analytical detections of dichloromethane greater than the MCL.  Approximately 17.7% of the population
(almost 19.5 million people) was served by PWSs with at least one analytical result greater than 0.0025
mg/L.  The percentage of population served by systems with results greater than 0.0005 mg/L was equal
to 27.6% (over 30 million people). Approximately 28.5% of the population (over 31 million people) was
exposed to levels of dichloromethane greater than 0.00025 mg/L.

A much greater proportion of the 16-State population was served by surface water systems with threshold
exceedances, as compared to ground water systems with threshold exceedances.  The percentage of
population served by ground water systems and surface water systems with analytical detections of
dichloromethane greater than the MCL was equal to 3.91% (almost 2 million people), and 14.0%
(approximately 8.5 million people), respectively. Relative to 0.0025 mg/L, 6.16% of the population
(about 3 million people) was served by ground water systems with threshold exceedances, whereas
26.9% of the population (over  16 million people) was served by surface water systems with threshold
exceedances. In addition, 16.9% of the population was served by ground water systems with results that
exceeded 0.0005 mg/L, while 36.2% of the population was served by surface water systems with results
exceeding 0.0005  mg/L.  Approximately 17.5% of the population served by ground water systems was
exposed to levels of dichloromethane greater than 0.00025 mg/L, compared to 37.3% of the population
served by surface  water systems exposed to levels greater than 0.00025 mg/L.
Table 4.6-4:  Stage 1 Dichloromethane Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population Served
by Systems
Exceeding Threshold
3.91%
6.16%
16.9%
Total Population Served by
Systems Exceeding
Threshold
1,916,200
3,019,600
8,315,100
Occurrence Summary and Use Support Document
385
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.00025
Percent of Population Served
by Systems
Exceeding Threshold
17.5%
Total Population Served by
Systems Exceeding
Threshold
8,599,000

Surface Water
0.005
0.0025
0.0005
0.00025
14.0%
26.9%
36.2%
37.3%
8,548,400
16,434,700
22,125,500
22,769,400

Combined Ground &
Surface Water
0.005
0.0025
0.0005
0.00025
9.50%
17.7%
27.6%
28.5%
10,464,600
19,454,400
30,440,600
31,368,400
4.6.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.6-5 and 4.6-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Dichloromethane occurs in a small proportion of PWSs when evaluated relative to the MCL.  Only 3
(about 0.0131% of) ground water and surface water PWSs had an estimated mean concentration of
dichloromethane exceeding 0.005 mg/L, the MCL. Approximately 13 (0.0601% of) systems were
estimated to have mean concentrations greater than 0.0025 mg/L. An estimated 348 (1.62% of) systems
had estimated mean concentration values greater than 0.0005 mg/L. About  1,067 (4.96% of) systems in
the 16 States were estimated to have mean concentrations greater than 0.00025 mg/L.

For ground water systems, 3 (about 0.0130% of) systems were estimated to have a mean concentration
greater than 0.005 mg/L. About  11 (0.0564% of) ground water systems in the 16 States had an estimated
mean concentration exceeding 0.0025 mg/L.  An estimated 303 (1.51% of) ground water systems in the
16 States had an estimated mean  concentration above 0.0005 mg/L, and 954 (4.77% of) ground water
systems had an estimated mean concentration greater than 0.00025 mg/L.

Only  1 surface water system in the 16-State cross-section had an estimated mean concentration value of
dichloromethane greater than 0.005 mg/L.  Approximately 2 (about 0.133% of) surface water systems
had estimated mean concentrations greater than 0.0025 mg/L. In addition, 44 (2.94% of) surface water
systems, and 113 (7.50% of) surface water systems were estimated to have mean concentrations above
0.0005 mg/L, and 0.00025 mg/L, respectively.
Occurrence Summary and Use Support Document
386
March 2002

-------
Table 4.6-5:  Stage 2 Estimated Dichloromethane Occurrence Based on 16-State Cross-Section -
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
0.00025
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.0130%
0.0546%
1.51%
4.77%
Range
0.00500% - 0.0300%
0.0250% - 0.0849%
1.31% -1.73%
4. 36% -5. 19%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
3
11
303
954
Range
1-6
5-17
262 - 346
872-1,038

Surface Water
0.005
0.0025
0.0005
0.00025
0.0142%
0.133%
2.94%
7.50%
0.000% - 0.0662%
0.000% -0.331%
2.25% -3.84%
6.22% - 8.93%
1
2
44
113
0-1
0-5
34-58
94-135

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.00025
0.0131%
0.0601%
1.62%
4.96%
0.00465% - 0.0279%
0.0325% - 0.0929%
1.41% -1.84%
4. 54% -5. 37%
3
13
348
1,067
1-6
7-20
304 - 395
977-1,157
Reviewing dichloromethane occurrence in the 16 cross-section States by PWS population served (Table
4.6-6) shows that approximately 0.119% of the 16-State population (over 131,000 people) was served by
PWSs with estimated mean concentrations of dichloromethane greater than the 0.005 mg/L.
Approximately 1.36% of the population (just over 1.5 million people) was served by PWSs with an
estimated mean concentration greater than 0.0025 mg/L. The percentage of population served by systems
with estimated mean concentrations of dichloromethane greater than  0.0005 mg/L was equal to 6.59%
(about 7.3 million people). Approximately 9.40% of the population (over 10 million people) was
potentially exposed to levels of dichloromethane greater than 0.00025 mg/1.

Although fewer surface water systems than ground water systems had estimated mean concentration
values greater than the thresholds, a much larger population was served by surface water systems with
threshold exceedances because surface water systems tend to serve much larger populations. The
percentage of the 16-State population served by ground water systems and surface water systems with
estimated mean concentrations of dichloromethane greater than the MCL was equal to 0.00203%
(approximately 1,000 people), and 0.213% (over 130,000 people), respectively. Relative to 0.0025 mg/L,
0.0189% of the population (about 9,300 people) was served by ground water systems with estimated
mean concentrations exceeding that limit, whereas 2.44% of the population (almost 1.5 million people)
was served by surface water systems with estimated mean concentrations exceeding 0.0025 mg/L. In
addition, 1.06% of the population served by ground water systems was potentially exposed to levels of
dichloromethane that exceeded 0.0005 mg/L, while  11.0% of the population was served by surface water
systems with estimated mean concentrations of dichloromethane exceeding 0.0005 mg/L. Approximately
3.56% of the population served by ground water systems was potentially exposed to levels of
dichloromethane greater than 0.00025 mg/L, compared to 14.1% of the population served by surface
water systems potentially exposed to levels greater than 0.00025 mg/L.
Occurrence Summary and Use Support Document
387
March 2002

-------
Table 4.6-6:  Stage 2 Estimated Dichloromethane Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
0.00025
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.00203%
0.0189%
1.06%
3.56%
Range
0.000357% -0.0103%
0.00242% -0.0815%
0.720% -1.91%
2.85% -4.83%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
1,000
9,300
518,100
1,748,000
Range
200 - 5,000
1,200-40,000
353,000 - 938,000
1,396,700-2,368,600

Surface Water
0.005
0.0025
0.0005
0.00025
0.213%
2.44%
11.0%
14.1%
0.000% - 0.449%
0.000% - 8.98%
9.93% -12.3%
12.7% - 16.2%
130,100
1,490,500
6,737,800
8,600,900
0 - 274,500
0 - 5,483,700
6,063,400-7,531,900
7,739,600 - 9,895,900

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.00025
0.119%
1.36%
6.59%
9.40%
0.000159% -0.250%
0.00221% -4.99%
5. 92% -7. 34%
8.52% - 10.6%
131,200
1,500,200
7,256,400
10,350,400
200 - 275,300
2,400 - 5,494,100
6,518,400-8,089,100
9,383,400-11,642,400
4.6.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 9 PWSs nationally serving approximately 253,700 people could be exposed to
dichloromethane concentrations above 0.005 mg/L.  About 39 systems serving almost 3 million people
nationally were estimated to have mean concentrations greater than 0.0025 mg/L.  Approximately 1,050
systems serving over 14 million people nationally were estimated to have mean dichloromethane
concentrations greater than 0.0005 mg/L.  A total of 3,224 systems serving over 20 million people were
estimated to have mean concentrations greater than 0.00025 mg/L. (See Section 1.4 for a description of
how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 8 PWSs serving about 1,700 people nationally had mean
concentrations greater than 0.005 mg/L.  Approximately 32 systems serving about 16,200 people
nationally were estimated to have mean concentration values that exceeded 0.0025 mg/L. About 900
ground water systems serving almost 905,200 people were estimated to have mean concentrations greater
than 0.0005 mg/L. An estimated 2,833 ground water PWSs nationally could potentially expose over 3
million people to concentrations of dichloromethane greater than 0.00025 mg/L.

Approximately 1 surface water system serving 271,200 people nationally was estimated to have mean
concentrations of dichloromethane above 0.005 mg/L. About 7 surface water systems serving
approximately 3.1 million people were estimated to  have mean concentrations greater than 0.0025 mg/L.
An estimated 164 surface water systems serving approximately 14 million people had mean
Occurrence Summary and Use Support Document
388
March 2002

-------
concentrations greater than 0.0005 mg/L.  Approximately 419 surface water PWSs serving almost 18
million people were estimated to have mean concentrations of dichloromethane greater than 0.00025
mg/L.
Table 4.6-7:  Estimated National Dichloromethane Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
0.00025
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
8
32
900
2,833
Range
3-18
15-50
778 - 1,027
2,589 - 3,082
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
1,700
16,200
904,800
3,052,800
Range
300 - 8,800
2,100 - 69,800
616,600-1,638,200
2,439,400-4,136,700

Surface Water
0.005
0.0025
0.0005
0.00025
1
7
164
419
0-4
0-18
126-215
348 - 499
271,200
3,106,800
14,044,100
17,927,600
0 - 572,200
0-11,430,100
12,638,400-15,699,400
16,132,300-20,626,900

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.00025
9
39
1,050
3,224
3-18
21 -60
918-1,193
2,951-3,495
253,700
2,901,200
14,033,000
20,016,400
300 - 532,300
4,700 - 10,624,800
12,605,800-15,643,300
18,146,200-22,515,000
4.6.5  Additional Drinking Water Occurrence Data

Additional studies of drinking water occurrence data are unavailable for dichloromethane.

4.6.6  Conclusion

Dichloromethane is used in paint strippers and removers, as a propellant in aerosols, as a process solvent
in the manufacture of drugs, pharmaceuticals, and film coatings, as a metal cleaning and finishing
solvent, in electronics manufacturing, and as an agent in urethane foam blowing.  It is also used in many
consumer products.  Production of dichloromethane has declined since the early 1980s, and the most
recent data indicates that 403  million pounds of dichloromethane were produced in  1994.
Dichloromethane is also a TPJ chemical. Industrial releases of dichloromethane have occurred since
1988 in 46 States and Puerto Rico. Dichloromethane was an analyte for the NAWQA, NURP, and
NPDES ambient occurrence studies. In the NAWQA study, dichloromethane was detected in 3.0% of
urban wells and 0.29% of rural wells, with a median detection value of 0.45 |ig/L for all wells. In the
Stage 2 analysis of 16-State occurrence of dichloromethane, 0.0131% of combined ground water and
surface water systems serving 0.119% of the population exceeded the  MCL of 0.005 mg/L. Nationally, 9
Occurrence Summary and Use Support Document
389
March 2002

-------
ground water and surface water systems combined (serving approximately 253,700 people) are estimated
to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, dichloromethane is manufactured and/or processed in 43 States and has
TRI releases in 46 States.  Dichloromethane is manufactured and/or processed in 14 out of the 16 cross-
section States and has TRI releases in 15 of the 16 cross-section States. The cross-section should
adequately represent the occurrence of dichloromethane on a national scale based upon the use,
production, and release patterns of the 16-State cross-section in relation to the patterns observed for all
50 States.

4.6.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  2000. Toxicological Profile for
       Methylene Chloride. U.S. Department of Health and Human Services, Public Health Service.
       268 pp. + Appendices. Available on the Internet at
       http://www.atsdr.cdc .gov/toxprofiles/tp 14 .pdf..

Chemexpo. 2001. Methylene Chloride Chemical Profile.  Available on the Internet at
       http://www.chemexpo.com/news/PROFILE001009.cfm.

Lopes, T.J. and  S.G. Dionne. 1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater. U.S. Geological Survey Open-File  Report 98-409. 67
       pp.

National Safety Council (NSC). 2001. Dichloromethane Chemical Backgrounder. Itasca, IL: National
       Safety Council. Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?id=104&chempath=chemicals, accessed
       July 23, 2001.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, R.M. Clawges, and J.S. Zogorski. 1999.
       Volatile organic compounds in untreated ambient groundwater of the United States,  1985-1995.
       Env. Sci. and Tech.  33(23):4176-4187.

USEPA.  2000. TRI Explorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.
Occurrence Summary and Use Support Document          390                                     March 2002

-------
4.7    1,2-Dichloropropane
Table of Contents

4.7.1  Introduction, Use and Production  	  392
4.7.2  Environmental Release  	  393
4.7.3  Ambient Occurrence 	  393
4.7.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  395
4.7.5  Additional Drinking Water Occurrence Data  	  400
4.7.6  Conclusion	  402
4.7.7  References  	  402
Tables and Figures

Table 4.7-1: Environmental Releases (in pounds) for 1,2-Dichloropropane
       in the United States, 1988-1999  	  393

Table 4.7-2: Stage 1 1,2-Dichloropropane Occurrence Based on 16-State Cross-Section - Systems  .  396

Table 4.7-3: Stage 1 1,2-Dichloropropane Based on 16-State Cross-Section - Population	  396

Table 4.7-4: Stage 2 Estimated 1,2-Dichloropropane Occurrence Based on 16-State
       Cross-Section - Systems	  397

Table 4.7-5: Stage 2 Estimated 1,2-Dichloropropane Occurrence Based on 16-State
       Cross-Section - Population	  398

Table 4.7-6: Estimated National 1,2-Dichloropropane Occurrence - Systems and
       Population Served	  399
Occurrence Summary and Use Support Document         391                                     March 2002

-------
4.7.1  Introduction, Use and Production

1,2-Dichloropropane (chemical formula C3H6C12) is a colorless liquid that has a chloroform-like odor and
evaporates quickly at room temperature.  It is a man-made chemical, whose releases are almost all
anthropogenic in nature.  1,2-Dichloropropane is now used in the United States only in research and
industry. 1,2-Dichloropropane is also known as propylene dichloride and DCP.

Based on 1982 production data supplied by Dow Chemical (USEPA, 1986, as cited in ATSDR, 1989), it
has been estimated that over 95% of the isolated  1,2-dichloropropane manufactured by Dow was used
on-site as a captive intermediate in the production of perchloroethylene and other chlorinated products by
their 'per-tet' process (USEPA, 1986; Dow Chem. Co., 1983, all as cited in ATSDR, 1989).
Approximately 3 million pounds per year of 1,2-dichloropropane were marketed by Dow Chemical in
1982 for minor uses as an industrial solvent for oils, fats, resins, waxes, and rubber, in ion exchange
manufacture, in toluene diisocyanate (TDI) production, in photographic film manufacture, for paper
coating, and for petroleum catalyst regeneration (HSDB, 1988; IARC, 1986; USEPA, 1986, all as cited in
ATSDR, 1989). Outside of its use as a chemical intermediate, Dow Chemical Company's use pattern for
1,2-dichloropropane in 1982 was as follows: 41% in ion exchange manufacturing, 34% in toluene
diisocyanate (TDI) production, 19% in photographic film production, 4% in paper coating, and 2% in
petroleum catalyst regeneration (Dow Chem. Co., 1983, as cited in ATSDR, 1989).

Production for sale, as opposed to internal consumption by manufacturers, was greatly curtailed in the
early  1980s. By 1983, 1,2-dichloropropane was no longer sold for consumer use in paint strippers, paint
varnish, and furniture finish removers (USEPA, 1986; Dow Chem. Co., 1983, all as cited in ATSDR,
1989). By the end of 1983, its use as a solvent for film production was to be phased out in favor of 1,1,1-
trichloroethane (Dow, 1983, as cited in ATSDR,  1989). According to comments submitted by Dow to
ATSDR (USEPA, 1989, as cited in ATSDR,  1989), the phaseout of use of 1,2-dichloropropane as a
solvent for film production had not occurred as of June, 1989, although it is still planned. They further
stated that the use of 34% of 1,2-dichloropropane in TDI production has now been  discontinued.

Before the early 1980s, it was used in farming as a soil fumigant and was found in some paint strippers,
varnishes, and furniture finish removers (ATSDR, 1989). An estimated 20 million pounds/year of
dichloropropane were produced as a by-product in a mixture marketed as a soil fumigant which was used
in the cultivation of a variety of crops, including  citrus fruits, pineapple, soy beans, cotton, tomatoes, and
potatoes (IARC, 1986; HSDB, 1988, all as cited in ATSDR, 1989). Dow has discontinued production of
soil fumigants containing 1,2-dichloropropane, and pesticidal formulations containing this chemical are
no longer available in the U.S. (Meister, 1987, as cited in ATSDR, 1989). Other uses of 1,2-
dichloropropane include use as an intermediate in the synthesis of carbon tetrachloride, lead scavenger in
gasoline, textile stain remover, oil and paraffin extractant, scouring compound, and metal degreasing
agent, especially prior to electroplating (IARC, 1986, as cited in ATSDR, 1989).

1,2-Dichloropropane was produced by Columbia Organics in Cassatt, SC, Dow Chemical in Freeport, TX
and Dow Chemical in Plaquemine, LA (SRI,  1988; USITC, 1987, all as cited in ATSDR, 1989);
however, Dow Chemical Company was the only manufacturer of the isolated chemical in the United
States (USEPA, 1986, as cited in ATSDR, 1989), and Dow discontinued the production of 1,2-
dichloropropane in 1991  (USEPA, 2001). The total output of 1,2-dichloropropane by U.S. manufacturers
remained relatively stable until 1984 when a major manufacturer, Mannsville Chemical Products
Corporation, discontinued production.  Production of 1,2-dichloropropane in  1980 was 77 million pounds
(USEPA, 2001), and domestic production volume during 1984 was 59.8 million pounds (IARC, 1986, as
cited in ATSDR, 1989).

Occurrence Summary and Use Support Document         392                                    March 2002

-------
4.7.2  Environmental Release

1,2-Dichloropropane is listed as a Toxics Release Inventory (TRI) chemical.  Table 4.7-1 illustrates the
environmental releases for 1,2-dichloropropane from 1988 - 1999. (1,2-Dichloropropane data are only
available for these years.) Air emissions constitute most of the on-site releases, with a relatively steady
decrease over the years. Surface water discharges decreased from a high of over 23,000 pounds in 1988
to under 4,000 pounds in 1994, but have since fluctuated between 1,000-9,000 pounds.  The significant
decrease in air emissions, as well as a decrease in surface water discharges, has contributed to decreases
in 1,2-dichloropropane total on- and off-site releases from 1988-1999. Underground injection has not
been monitored for most of the years, and otherwise registers zero, with the exception of a measurement
in 1994. Releases to land (such as spills or leaks within the boundaries of the reporting facility) show no
apparent trend, ranging from zero to over 3,000 pounds. However, releases to land have been 150
pounds or less since 1993. Off-site releases (including metals or metal compounds transferred off-site)
vary between under 300 pounds to over 12,000 pounds, displaying no discernable trend. These TRI data
for  1,2-dichloropropane were reported from 15 States, with five States reporting every year (USEPA,
2000). Five of the 16 cross-section States (used for analyses of 1,2-dichloropropane occurrence in
drinking water; see Section 4.7.4) reported releases of 1,2-dichloropropane.  (For a map of the 16-State
cross-section, see Figure 1.3-1.)
Table 4.7-1:  Environmental Releases (in pounds) for 1,2-Dichloropropane in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
249,656
298,150
378,454
514,428
616,470
709,547
577,439
619,917
772,543
1,038,614
1,280,220
1,395,304
Surface Water
Discharges
9,242
1,122
2,609
1,855
4,344
3,609
4,749
6,755
6,570
10,453
11,577
23,785
Underground
Injection
-
-
-
-
-
215
0
-
-
-
0
-
Releases
to Land
30
32
30
150
20
12
19
1,206
0
300
5
3,400
Off-Site Releases
6,856
267
12,375
5,337
1,371
699
567
1,952
2,073
1,639
1,446
1,131
Total On- &
Off-site
Releases
265,784
299,571
393,468
521,770
622,205
714,082
582,774
629,830
781,186
1,051,006
1,293,248
1,423,620
 Source: USEPA, 2000
4.7.3  Ambient Occurrence

1,2-Dichloropropane was detected in 3 out of 359 wells (0.84%) in urban areas of the local, State, and
federal data set compiled by NAWQA.  The minimum and maximum concentrations detected were 0.2
l-ig/L and 1.3 |ig/L, respectively. The median value of detection concentrations was 0.2 |ig/L. 1,2-
Dichloropropane was also detected in 23 of the 2,508 wells (0.92%) analyzed in rural areas.  The
minimum and maximum concentrations detected were 0.2 |ig/L and 7.5 |ig/L, respectively.  The median
value of detection concentrations was 0.5 |ig/L. These urban and rural data represent untreated ambient
ground water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).
Occurrence Summary and Use Support Document
393
March 2002

-------
1,2-Dichloropropane was also an analyte in the NPDES data. NPDES-related investigations analyzing
urban and highway runoff detected 1,2-dichloropropane (Lopes and Dionne, 1998).  The minimum
concentration detected was not reported, and the maximum concentration detected was 3 |ig/L, with no
mean value reported.  The use of the land from which the samples were taken was unspecified.

4.7.3.1 Additional Ambient Occurrence Data

A summary document entitled "Occurrence of Synthetic Organic Chemicals in Drinking Water, Food,
and Air" (USEPA, 1987), was previously prepared for past USEPA assessments of 1,2-dichloropropane.
Several studies were included that addressed levels of 1,2-dichloropropane in water other than drinking
water. The following information is taken directly from "Occurrence of Synthetic Organic Chemicals in
Drinking Water, Food, and Air" (USEPA, 1987).

4.7.3.1.1 Ground Water Sources

Five State studies from California were obtained, including a study of various shallow and deep wells in
Kern County, California (near Bakersfield) for DCP  (Cohen and Bowes,  1984, as cited in USEPA, 1987).
Seventeen positive samples were reported with a range of concentrations from 0.1 to 7.9 |ig/L. No other
information on the study was reported.

Another State study of wells in Lathrop, California (San Joaquin County), near the Occidental Chemical
Company, was conducted in 1979, and again in 1983 (Cohen and Bowes, 1984, as cited in USEPA,
1987). In 1979, seven of 14 samples from seven locations were positive for DCP with a range of
concentrations from 0.2 to 5.0 |ig/L. The mean concentration and detection limit were not reported.
Analyzed subsequently in 1983, none of the samples from the seven locations had detectable levels of
DCP.

Ground water wells in 25 counties from California were analyzed as part of a toxics special project by
the California Water Resources Control Board (Cohen and Bowes, 1984; Holden,  1986, all as cited in
USEPA, 1987). Positive results for DCP were found in 68 out of 266  samples taken from as many sites.
A high value of 1,200 |ig/L was reported for a sample from Crescent City, California. No further
information was presented in either report on this study.

Ground water from 24 feet below surface level was sampled in 1982 as part of the California State Board
Toxic Substances Control Program (Cohen and Bowes,  1984, as cited in USEPA,  1987). One sample was
reported to have a concentration of 4.6 |ig/L. No other information was reported on this study.

Ground water well samples from California, Maryland, New York, and Washington were reported by the
USEPA Office of Pesticide Programs (Cohen et al., 1986, as cited in USEPA, 1987) to show typical
positive concentrations for DCP in the range of 1.0 to 50 |ig/L. No further information was reported.

4.7.3.1.2 Surface Water Sources

Urban stormwater runoff in Eugene, Oregon, was sampled as part of USEPA's Nationwide Urban Runoff
Program (Cole et al.  1984, as cited in USEPA, 1987). Of the 15 sites sampled, only 1 of 86 samples was
positive for DCP, at a concentration of 3.0 |ig/L. The detection limit was  not reported.

Surface water from rivers in Delaware, Ohio, and West Virginia, as well  as 14 other U.S. river basins,
were sampled during various studies reported in USEPA (1985, as cited in USEPA, 1987). In total, more

Occurrence Summary and Use Support Document         394                                    March 2002

-------
than 8 samples from over 207 sites tested positive, with some reported levels between "trace" and 2.0
l-ig/L. No other specific information was reported.

4.7.3.1.3 Unidentified Sources

Water samples have been compiled by USEPA's STORET data base from national studies without
specification as to type of source (USEPA, 1985, as cited in USEPA, 1987). For a total of 22,670
"observations," DCP concentrations ranged from 0 to 25,000 |ig/L, with a mean concentration of 10.0
l-ig/L. Whether "observations" referred to positive samples or to the total samples collected was
unspecified. No other information on the studies was reported.

4.7.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of 1,2-dichloropropane occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative to several concentration thresholds  of interest: 0.005 mg/L; 0.0005 mg/L;  and 0.0004
mg/L.

All sixteen cross-section State data sets contained occurrence data for 1,2-dichloropropane. These data
represent more than 180,000 analytical results from approximately 22,000 PWSs during the period from
1984 to 1998 (with most analytical results from 1992  to 1997).  The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to ensure adequacy of
coverage and completeness. The overall modal detection limit for 1,2-dichloropropane in the 16 cross-
section States is equal  to 0.0005 mg/L.  (For details regarding the 16-State  cross-section, please refer to
Section 1.3.5 of this report.)

4.7.4.1  Stage 1 Analysis Occurrence Findings

Table 4.7-2 illustrates the occurrence of in drinking water for the public water systems in the 16-State
cross-section relative to three thresholds: 0.005 mg/L (the current MCL), 0.0005 mg/L (the modal
MRL), and 0.0004 mg/L. According to the  16-State cross-section data, a total of 15 (approximately
0.0682% of) ground water and surface water PWSs had analytical results exceeding the MCL; 0.432% of
systems (95 systems) had results exceeding 0.0005 mg/L;  and 0.637% of systems (140 systems) had
results exceeding 0.0004 mg/L.

Approximately 0.0588% of ground water systems (12 systems) had any analytical results greater than the
MCL. About 0.402%  of ground water systems (82 systems) had results above 0.0005 mg/L.  The
percentage of ground water systems with at least one result greater than 0.0004 mg/L was equal to
0.539% (110 systems).

Approximately 3 (0.190% of) surface water systems had results greater than the MCL. A total of 13
(0.824% of) surface water systems had at least one analytical result greater than 0.0005 mg/L. Thirty
(1.90% of) surface water systems had results exceeding 0.0004 mg/L.
Occurrence Summary and Use Support Document         395                                     March 2002

-------
Table 4.7-2:  Stage 1 1,2-Dichloropropane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Systems
Exceeding Threshold
0.0588%
0.402%
0.539%
Number of Systems
Exceeding Threshold
12
82
110

Surface Water
0.005
0.0005
0.0004
0.190%
0.824%
1.90%
3
13
30

Combined Ground &
Surface Water
0.005
0.0005
0.0004
0.0682%
0.432%
0.637%
15
95
140
Reviewing 1,2-dichloropropane occurrence in the 16 cross-section States by PWS population served
(Table 4.7-3) shows that approximately 1.18% of the 16-State population (over 1.3 million people) was
served by PWSs with at least one analytical result of 1,2-dichloropropane greater than the MCL (0.005
mg/L). Over 11 million (10.1% of) people were served by systems with an exceedance of 0.0005 mg/L.
A total of about 12 million (10.9% of) people were served by systems with at least one analytical result
greater than 0.0004 mg/L.

The percentage of population served by ground water systems in the 16 States with analytical results
greater than the MCL was equal to 2.22% (approximately 1 million people). When evaluated relative to
1, 2-dicloropropane detections of 0.0005 mg/L or 0.0004 mg/L, the percent of population exposed was
equal to 3.48% (over 1.7 million people) and 3.80% (almost 1.9 million people), respectively.

The percentage of population served by surface water systems with exceedances of 0.005 mg/L was
equal to 0.347% (approximately 212,200 people). Approximately 15.4% of the population served by
surface water systems (about 9.4 million people)  was exposed to 1,2-dichloropropane concentrations
greater than 0.0005 mg/L.  When evaluated relative to 0.0004 mg/L, the  percent of population exposed
was equal to 16.7% (over 10 million people).
Table 4.7-3:  Stage 1 1,2-Dichloropropane Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Percent of Population Served
by Systems
Exceeding Threshold
2.22%
3.48%
Total Population Served by
Systems Exceeding
Threshold
1,091,000
1,715,000
Occurrence Summary and Use Support Document
396
March 2002

-------
Source Water Type

Threshold
(mg/L)
0.0004
Percent of Population Served
by Systems
Exceeding Threshold
3.80%
Total Population Served by
Systems Exceeding
Threshold
1,871,100

Surface Water
0.005
0.0005
0.0004
0.347%
15.4%
16.7%
212,200
9,443,700
10,216,500

Combined Ground &
Surface Water
0.005
0.0005
0.0004
1.18%
10.1%
10.9%
1,303,200
11,158,700
12,087,600
4.7.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.7-4 and 4.7-5.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented. (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report.  For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

One (0.00358% of) PWS in the 16 States had an estimated mean concentration of 1,2-dichloropropane
exceeding 0.005 mg/L, while 8 (0.0382% of) systems in the 16 States had an estimated mean
concentration exceeding 0.0005 mg/L and 11 (0.0506% of) systems had an estimated mean concentration
exceeding 0.0004 mg/L.

Zero ground water PWSs in the 16 States were estimated to have a mean concentration greater than 0.005
mg/L.  Seven (0.0347% of) ground water PWSs were estimated to exceed a mean concentration of
0.0005 mg/L and 10 (0.0472% of) systems are estimated to exceed a mean concentration of 0.0004 mg/L.
For surface water PWSs in the 16 States, 0.0176%, 0.0837%, and 0.0947% had estimated mean
concentrations exceeding 0.005 mg/L, 0.0005 mg/L, and 0.0004 mg/L, respectively.

Table 4.7-4:  Stage 2 Estimated 1,2-Dichloropropane Occurrence Based on 16-State Cross-Section -
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Systems Estimated to Exceed
Threshold
Best Estimate
0.00250%
0.0347%
0.0472%
Range
0.000% - 0.00980%
0.0196% -0.0539%
0.0245% - 0.0686%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
7
10
Range
0-2
4-11
5-14

Occurrence Summary and Use Support Document
397
March 2002

-------
Source Water Type
Surface Water
Threshold
fmg/T;»
0.005
0.0005
0.0004
Percent of Systems Estimated to Exceed
Threshold
0.0176%
0.0837%
0.0947%
0.000% - 0.0634%
0.0634% -0.127%
0.0634% -0.1 90%
Number of Systems in the 16 States
Estimated to Exceed Threshold
1
1
1
0-1
1-2
1 -3

Combined Ground
& Surface Water
0.005
0.0005
0.0004
0.00358%
0.0382%
0.0506%
0.000% -0.00910%
0.0227% - 0.0546%
0.03 18% -0.0728%
1
8
11
0-2
5-12
7-16
Reviewing 1,2-dichloropropane occurrence by PWS population served (Table 4.7-6) shows that
approximately 0.0358% of population served by all PWSs in the 16 States (an estimate of approximately
39,500 people) was potentially exposed to 1,2-dichloropropane levels above 0.005 mg/L. For all PWSs,
an estimated 0.137% of population served (an estimate of over 151,000 people served in the 16 States)
was exposed to levels above 0.0005 mg/L and 0.150% (an estimated 165,700 people in the  16-State
cross-section) was exposed to levels above 0.0004 mg/L.

When the population exposed by ground water systems in the 16 States was evaluated relative to
thresholds of 0.005 mg/L, 0.0005 mg/L, and 0.0004 mg/L, the percentages of population exposed were
equal to 0.0000401%, 0.0167% (8,200 people), and 0.0446% (22,000 people), respectively.

 The percentage of population served by  surface water systems in the 16 States with levels above 0.005
mg/L was 0.0645% (an estimated 39,500  people in the 16 States), while the population served with levels
above 0.0005 mg/L and 0.0004 mg/L was 0.234% (an estimated 143,200 people in the 16 States) and
0.235% (an estimate of 143,700 people in the 16 States), respectively.
Table 4.7-5:  Stage 2 Estimated 1,2-Dichloropropane Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.0000401%
0.0167%
0.0446%
Range
0.000% -0.000136%
0.00205% - 0.0649%
0.00466% - 0.0962%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
8,200
22,000
Range
0-100
1,000-32,000
2,300 - 47,400

Surface Water
0.005
0.0005
0.0004
0.0645%
0.234%
0.235%
0.000% - 0.232%
0.232% - 0.235%
0.232% - 0.250%
39,500
143,200
143,700
0 - 142,000
142,000 - 143,500
142,000-153,000
Occurrence Summary and Use Support Document
398
March 2002

-------
Source Water Type
Threshold
Percent of Population Served by Systems
Estimated to Exceed Threshold
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold

Combined Ground
& Surface Water
0.005
0.0005
0.0004
0.0358%
0.137%
0.150%
0.000% -0.129%
0.1 30% -0.1 58%
0.131% -0.179%
39,500
151,400
165,700
0 - 142,000
143,000 - 174,700
145,000-197,400
4.7.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) with mean 1,2-
dichloropropane occurrence exceeding each threshold, an estimated 2 PWSs nationally, serving over
76,000 people total could be exposed to 1,2-dichloropropane concentrations above 0.005 mg/L. About
25 systems, serving a total of about 292,000 people, had estimated mean concentrations greater than
0.0005 mg/L. Approximately 33 systems, serving a total of about 319,500 people nationally, were
estimated to have mean 1,2-dichloropropane concentrations greater than 0.0004 mg/L. (See Section 1.4
for a description of how Stage  2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 1 PWS in the nation, serving less than 100 people, had a mean
concentration greater than 0.005 mg/L.  Approximately 21 systems serving  14,300 people nationally had
estimated mean concentration values that exceeded 0.0005 mg/L. About 28 ground water systems
serving about 38,200 people had estimated mean concentrations greater than 0.0004 mg/L.

Although fewer surface water systems than ground water systems nationally had estimated mean
concentration values greater than the thresholds, a much larger population was served by surface water
systems with threshold exceedances because surface water systems tend to serve much larger
populations. Approximately 1  surface water system, serving about 82,100 people, was estimated to have
mean concentrations of 1,2-dichloropropane above 0.005 mg/L.  About 5 surface water systems serving
almost 298,000 people had estimated mean concentrations greater than 0.0005 mg/L.  An estimated 5
surface water systems, serving approximately 299,000 people, had mean concentrations greater than
0.0004 mg/L.
Table 4.7-6:  Estimated National 1,2-Dichloropropane Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
0.0004
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
1
21
28
Range
0-6
12-32
15-41
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
<100
14,300
38,200
Range
0-100
1,800-55,600
4,000 - 82,400

Surface Water
0.005
0.0005
1
5
0-4
4-7
82,100
297,900
0 - 295,500
295,500 - 298,600
Occurrence Summary and Use Support Document
399
March 2002

-------
Source Water Type

Threshold
(rng/T )
0.0004
Total Number of Systems Nationally
Estimated to Exceed Threshold
5
4-11
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
299,000
295,500-318,300

Combined Ground
& Surface Water
0.005
0.0005
0.0004
2
25
33
0-6
15-35
21-47
76,200
292,000
319,500
0 - 273,900
275,800 - 337,000
279,700 - 380,600
4.7.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of 1,2-dichloropropane in drinking water are
also reviewed. Previously compiled occurrence information on 1,2-dichloropropane, from an OGWDW
summary document entitled "Occurrence of Synthetic Organic Chemicals in Drinking Water, Food, and
Air" (USEPA, 1987), is presented in this section. This variety of studies and information are presented
regarding levels of 1,2-dichloropropane in drinking water, with the scope of the reviewed studies ranging
from national to regional. Note that none of the studies presented in the following section provide the
quantitative analytical results or comprehensive coverage that would enable direct comparison to the
occurrence findings estimated with the cross-section occurrence data presented in Section 4.7.4. These
additional studies, however, do enable a broader assessment of the Stage 2 occurrence estimates
presented for this Six-Year Review. All the following information in Section 4.7.5 is taken directly from
"Occurrence of Synthetic Organic Chemicals in Drinking Water, Food, and Air" (USEPA, 1987).

4.7.5.1 Groundwater - National Studies

The Ground Water Supply Survey was conducted from December 1980 to December 1981 to develop
data on the occurrence of volatile organic chemicals in the nation's ground water supplies (USEPA, 1983;
Westrick et al., 1983, all as cited in USEPA, 1987). The GWSS involved the sampling of 945 water
supply- systems, of which 466 were selected randomly while the remaining 479 systems were chosen by
the States as having a high potential for contamination. DCP was observed in 6 of the 466 randomly
selected systems. One of the six positive values obtained came from a supply serving fewer than 10,000
persons; the positive value was 0.75 |ig/L. The other five positive values, found in supplies serving
greater than 10,000 people, had a median value of 0.86 |ig/L (range = 0.48 to 21 |ig/L). In the nonrandom
portion of the GWSS, DCP was detected in 7 of 479 systems sampled. Three positives were from systems
serving fewer than  10,000 people, and had values of 0.51,  1.2,  and 1.4 |ig/L. The four positive values
from supplies serving more than 10,000 people had a median value of 0.7 |ig/L (range =  0.21 to 18 |ig/L).
The quantitation limit for both random and nonrandom systems was 0.2 |ig/L.

The American Water Works Association (AWWA) surveyed water utilities nationwide that were
monitoring for unregulated contaminants including 1,2-dichloropropane in raw and finished water from
ground and surface water sources. Of the 114 utilities monitoring for  1,2-dichloropropane, 3 utilities
whose primary water source is ground water detected 1,2-dichloropropane in 1 of 1,382 raw water
samples and 5 of 5 finished water samples. Concentrations ranged from less than 1.0 to 4.0 in finished
water. The concentration detected in raw water was 3.0 |ig/L. No detection limits were reported
(AWWA, 1988, as cited in USEPA, 1987). The American Waterworks Association advises caution in
utilizing this data as it appears there are discrepancies in the units reported by some utilities
(Achtermann, 1990, as cited in USEPA,  1987).
Occurrence Summary and Use Support Document
400
March 2002

-------
4.7.5.2 Groundwater - Regional Studies

Twelve towns in Connecticut were examined by the Connecticut Agricultural Experiment Station in New
Haven, Connecticut (Waggoner, 1985, as cited in USEPA, 1987). Drinking water wells representing the
major drinking water sources for those 12 towns were analyzed during 1984-1985. The population served
by the wells sampled exceeded 570,000 persons. None of the wells, at a total of 42 locations, were
positive for DCP. The detection limit was 0.05 |ig/L.

Community ground water wells were sampled during 1982 in Visalia (Tulare County) and Reedley
(Fresno County), California, as part of the USEPA National Groundwater Monitoring Program (Cohen
and Bowes, 1984, as cited in USEPA, 1987). DCP was found in three samples, at a mean concentration
of 9.93 |ig/L (range = 1.0 to 25.9 |ig/L). The total number of samples, number of locations, and detection
limit were not reported.

Domestic and irrigation wells were sampled for DCP on Long Island, New York, by the Suffolk County
Department of Health Services (Holden, 1986, as cited in USEPA, 1987). Out of 1,000 samples taken,
about 500 were contaminated with DCP at concentrations ranging from 30 to 300 |ig/L. More specific
information was not reported.

Community wells in Del Norte County, California,  were sampled in 1983 for DCP (Cohen and Bowes,
1984, as cited in USEPA, 1987). Twenty-five of 37 samples were found positive, with a maximum
concentration of greater than 10.0 |ig/L. No other information on this study was reported.

Private domestic wells from three other California counties were analyzed as part of the California State
Board Toxic Substances Control Program during 1982 (Cohen and Bowes, 1984, as cited in USEPA,
1987). DCP was found in 12 out of 95 samples, with a range of concentrations between 0.4 and 16 |ig/L.
The mean concentration and detection limit were not reported.

4.7.5.3 Surface Water - National Study

The American Water Works Association (AWWA) surveyed water utilities nationwide that were
monitoring for unregulated  contaminants including  1,2-dichloropropane in raw and finished water from
ground and surface water sources. Of the 114 utilities monitoring for 1,2-dichloropropane, 10 utilities
whose primary water source is surface water detected 1,2-dichloropropane in 793 of 11,394 raw water
samples and 257 of 885 finished water samples. Concentrations ranged from 0.1 |ig/L to 15.0 |ig/L in raw
water samples and less than 0.1 |ig/L to 9.0 |ig/L in finished water samples. No detection limits were
reported (AWWA,  1988, as cited in USEPA, 1987). The American Waterworks Association advises
caution in utilizing this data as it appears there are discrepancies in the units reported by some utilities
(Achtermann, 1990, as cited in USEPA, 1987).

4.7.5.4 Surface Water - Regional Studies

To assemble a database  which would reflect the status of Great Lakes drinking water quality, the
Canadian Public Health Association gathered data from October 1984 through August 1985. The data
collected covered the period from the mid 1970s to  early 1985. A research team, appointed by the
Association, reviewed data  on the quality of water at 31 representative Canadian and United States
communities and 24 offshore sites to evaluate the human health implications.
Occurrence Summary and Use Support Document         401                                    March 2002

-------
For each of the 31 communities, data consisted of: 1) background information on the community; 2)
treatment plant schematics and associated treatment process information; and 3) water quality data.
Water sample types included raw water (treatment plant intake), distribution water (treated water), and
tap water. Water quality data collected included general parameters (e.g., alkalinity, turbidity),
microbiological and radiological parameters, inorganic parameters, and organic parameters (including
volatiles, base/neutrals, pesticides and PCBs, and phenols and acids). For each parameter, the water type,
time period, concentration (mean, range), number of samples, and detection limit were recorded.

For most of the volatile organics, including DCP, the available data indicated that these contaminants did
not occur at significant levels in the raw, treated, or tap  water. Most of the values found were "not
detected" or near the detection limit (Canadian Public Health Association, 1986, as cited in USEPA,
1987).

4.7.5.5 Unidentified Sources

Drinking water samples were collected and analyzed  in the Love Canal area, New York, during 1980 by
Barkley et al. (1980, as cited in USEPA, 1987). The only information reported was that one sample
contained a DCP concentration of 1,200 |ig/L (detection limit not given).

4.7.6 Conclusion

1,2-Dichloropropane was mostly  used captively by manufacturers as a chemical intermediate in the
production of chlorinated products.  At one time, it was also used in various consumer products such as
paint and finish removers and was found in soil fumigants as a processing byproduct. Domestic
production of 1,2-dichloropropane as an isolated chemical ceased in 1991. 1,2-Dichloropropane is also a
TRI  chemical. Industrial releases of 1,2-dichloropropane have occurred since 1988 in 15 States.  1,2-
Dichloropropane was an analyte for the NAWQA and NPDES ambient occurrence studies. In the
NAWQA study, 1,2-dichloropropane was detected in 0.84% of urban wells and 0.92% of rural wells,
with median detection values of 0.2 |ig/L and 0.5 |ig/L,  respectively. In the Stage 2 analysis of 16-State
occurrence of 1,2-dichloropropane, 0.00358% of combined ground water and surface water systems
serving 0.0358% of the population exceeded the MCL of 0.005 mg/L. Nationally, 2 ground water and
surface water systems combined (serving approximately 76,200 people) are estimated to have levels
greater than the MCL.

The  16-State cross-section was designed to be nationally representative based upon VOC, SOC, and  IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, TRI releases have been reported for 1,2-dichloropropane from 15 States,
including 5 of 16 cross-section States.  As an isolated chemical,  1,2-Dichloropropane is no longer
produced in any State. The cross-section should adequately represent the occurrence of 1,2-
dichloropropane on a national scale based upon the use, production, and release patterns of the 16-State
cross-section in relation to the patterns observed for all  50 States.

4.7.7 References

Agency for Toxic Substances and Disease Registry (ATSDR).  1989.  Toxicological Profile for 1,2-
       Dichloropropane. U.S. Department of Health and Human Services,  Public Health Service.  118
       pp. + Appendices. Available on the Internet at  http://www.atsdr.cdc.gov/toxprofiles/tpl34.pdf.
Occurrence Summary and Use Support Document         402                                      March 2002

-------
Dow Chem Co.  1983. Letter summarizing the EPA/DOW Chemical Co. meeting regarding the
       manufacture ofl,2-dichloropropane. OTS Section 4 submission by Dow Chemical Co.
       Document No. 40-8367072. Microfiche No, OTS 0511731.

Hazardous Substances Data Bank (HSDB).  1988. Hazard Substances Data Bank.
       On-line: July 27, 1988.

International Agency for Research on Cancer (IARC). 1986. IARC Monograph on the Evaluation of
       Carcinogenic Risks to Humans: Some halogenated hydrocarbon and pesticide exposure, v.  41,
       pp.131-147.

Lopes, T.J. and S.G. Dionne. 1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater.  U.S. Geological Survey Open-File Report 98-409.
       67pp.

Meister, R.T. 1987. Farm chemicals handbook '87. Willoughby, OH: Meister Publishing Co.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, RM. Clawges, and J.S. Zogorski.  1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env.  Sci. and Tech. 33(23):4176-4187.

SRI. 1988. Directory of Chemical Producers:  United States of America, p. 921.

USEPA. 1986.  Toxic substances; 1,2-dichloropropane; testing requirements. Final rule. Fed Reg
       51:32079-32087.

USEPA. 1987.  Occurrence of Synthetic Organic Chemicals in Drinking Water, Food, and Air. Office
       of Drinking Water, USEPA. July 1987.

USEPA. 1989.  Comments of the Dow Chemical Company on ATSDR's Toxicology Profile for 1,2-
       Dichloride.  Submitted to ATSDR May 1, 1989.

USEPA. 2000. TRIExplorer: Trends.  Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA. 2001.  Consumer Factsheet on: 1,2-Dichloropropane. Available on the Internet at:
       http://www.epa.gov/safewater/dwh/c-voc/12-dich3.html, last updated April 12, 2001.

USEPA. 2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.

United States International Trade Commission (USITC).  1987. Synthetic Organic Chemicals: United
       States Production and Sales, 1986.  USITC Pub 2009. pp. 233; 238.

Still need references cited in USEPA, 1987
Occurrence Summary and Use Support Document         403                                    March 2002

-------
4.8    Tetrachloroethylene
Table of Contents

4.8.1  Introduction, Use and Production  	 405
4.8.2  Environmental Release  	 407
4.8.3  Ambient Occurrence  	 407
4.8.4  Drinking Water Occurrence Based on the 16-State Cross-Section	 408
4.8.5  Additional Drinking Water Occurrence Data  	 412
4.8.6  Conclusion	 420
4.8.7  References 	 421
Tables and Figures

Table 4.8-1: Tetrachloroethylene Manufacturers and Processors by State  	 406

Table 4.8-2: Environmental Releases (in pounds) for Tetrachloroethylene
       in the United States, 1988-1999  	 407

Table 4.8-3: Stage 1 Tetrachloroethylene Occurrence Based on 16-State Cross-Section -
       Systems	 409

Table 4.8-4: Stage 1 Tetrachloroethylene Occurrence Based on 16-State Cross-Section -
       Population	 409

Table 4.8-5: Stage 2 Estimated Tetrachloroethylene Occurrence Based on 16-State
       Cross-Section - Systems	 410

Table 4.8-6: Stage 2 Estimated Tetrachloroethylene Occurrence Based on 16-State
       Cross-Section - Population	 411

Table 4.8-7: Estimated National Tetrachloroethylene Occurrence - Systems and
       Population Served	 412
Occurrence Summary and Use Support Document         404                                     March 2002

-------
4.8.1  Introduction, Use and Production

Tetrachloroethylene is a manufactured organic compound with the chemical formula C2C14. It is a
nonflammable liquid at room temperature, evaporates easily into the air, and has a sharp, sweet odor.
The major routes of entry of tetrachloroethylene to drinking water are a consequence of its production
and industrial use.  It is expected that discharges into surface water would volatilize into the atmosphere
fairly rapidly.  It is expected that discharges during production and use, and leaching into groundwater
from wastes deposited in landfills are primary causes of tetrachloroethylene contamination of drinking
water. Additionally, tetrachloroethylene may enter drinking water through the use of vinyl-lined A/C
pipe (JRB Associates,  1983). Other names for tetrachloroethylene include 1,1,2,2,-tetrachloroethylene,
perchloroethylene, PCE, perc, tetrachloroethene, perclene, and perchlor (ATSDR, 1997).

Tetrachloroethylene is a commercially important chlorinated hydrocarbon solvent and chemical
intermediate in the production of chlorofluorocarbons. It is used as a dry cleaning and textile-processing
solvent and for vapor degreasing in metal-cleaning operations. Tetrachloroethylene was first
commercially produced in the U.S. in  1925.  There has been an overall decline in production of about
50% between 1983 and 1993, from 547 to 271 million pounds (ATSDR, 1997).  In 1986, production of
tetrachloroethylene totaled 405  million pounds (USEPA, 2001) and 1991 production has been reported as
310 million pounds. The major reasons for the decline in production are  solvent recycling and reduced
demand for chlorofluorocarbons (USEPA, 1994).

The 1995 Directory of Chemical Producers in the U.S. lists three major manufacturers of
tetrachloroethylene with a total annual capacity of 490 million pounds.  Tetrachloroethylene is also
produced naturally by  various temperate and subtropical marine macroalgae at the rate of 0.0026-8.2 ng/g
fresh weight/hour, a potentially significant amount in the global chlorine budget (ATSDR, 1997).

It is widely used for dry cleaning of fabrics and for metal-degreasing operations and is also  used as a
starting material (building block) for making other chemicals and in some consumer products (ATSDR,
1997). In 1995, the estimated end-use breakdown for tetrachloroethylene was as follows: 55% as a
chemical intermediate, 25% for metal cleaning and vapor degreasing, 15% for dry cleaning and textile
processing, and 5% miscellaneous (ATSDR, 1997).  In small amounts, tetrachloroethylene is used in
rubber coatings, solvent soaps, printing inks, adhesives and glues, sealants, polishes, lubricants, and
pesticides (NSC, 2001).

Other uses of tetrachloroethylene include use in veterinary medicine as an anthelmintic; as a fumigant for
insects and rodents; as a vermifuge; as a heat-transfer medium; in copying machines; in the  manufacture
of paint removers; and in removing soot from industrial boilers (NTP, 1991). Formerly
tetrachloroethylene was used as an anthelmintic against hookworms, intestinal flukes, and nematodes; but
it has  since been supplanted by drugs that are less toxic and easier to administer (ATSDR, 1997).

Table 4.8-1 shows the  number of facilities in each State that manufacture and process
tetrachloroethylene, the intended uses of the product, and the range of maximum amounts derived from
the Toxics Release Inventory (TRI) of EPA (ATSDR, 1997).
Occurrence Summary and Use Support Document         405                                     March 2002

-------
Table 4.8-1:  Tetrachloroethylene Manufacturers and Processors by State
State"
AL
AR
AZ
CA
CO
CT
FL
GA
IA
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
MT
NC
NE
NH
NJ
NY
OH
OK
OR
PA
PR
SC
TN
TX
UT
VA
VT
WA
WI
WY
Number of facilities
5
4
2
63
2
16
11
12
7
35
26
8
10
16
7
2
1
10
11
12
5
2
20
4
4
6
20
48
8
1
18
2
12
8
25
3
7
1
1
12
1
Range of maximum amounts on
site in thousands of pounds'1
1-100
1-100
1-10
0-10,000
1-100
1-1,000
1-1,000
0-1,000
0-1,000
0-10,000
0-100
1-10,000
1-1,000
0-50,000
1-1,000
1-10
1-10
1-1,000
1-100
1-1,000
10-1,000
1-100
0-1,000
1-100
1-100
0-100
0-1,000
0-1,000
1-1,000
10-100
0-10,000
0-10
1-100
1-1,000
1-50,000
1-100
0-1,000
100-1,000
1-10
1-100
0-1
Activities and uses0
2,3,4,10,13
2,3,8,12,13
13
2,3,4,7,8,9,10,11,12,13
12,13
11,12,13
8,11,12,13
2,3,5,8,10,11,13
11,12,13
2,3,4,8,10,11,12,13
2,3,8,10,11,12,13
1,3,4,7,8,11,12,13
1,3,7,8,11,12,13
1,3,4,5,6,7,8,9,11,12,13
8,10,12,13
12,13
12
7,8,10,11,12,13
2,3,11,12,13
8,11,12,13
8,11,12,13
11,13
2,3,5,7,8,11,12,13
12,13
13
8,11,12,13
8,10,11,12,13
2,3,8,10,11,12,13
11,12,13
11
8,10,11,12,13
2,5,8,13
8,12,13
8,11,12,13
1,2,3,4,5,6,7,8,10,11,12,13
11,13
2,3,9,11,13
12
12
7,8,10,12,13
11.13
Tost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
1. Produce
2. Import
3. For on-site use/processing
4. For sale/distribution
5. As a by-product
6. As an impurity
7. As a reactant
8. As a formulation component
9. As an article component
10. For repackaging only
11. As a chemical processing aid
12. As a manufacturing aid
13. Ancillary or other uses
Source: AT SDR, 1997 compilation ofTRI93 1995 data
Occurrence Summary and Use Support Document
                                406
March 2002

-------
4.8.2 Environmental Release

Tetrachloroethylene is listed as a Toxics Release Inventory (TRI) chemical.  Table 4.8-2 illustrates the
environmental releases for tetrachloroethylene from 1988 - 1999. (Tetrachloroethylene data are only
available for these years.)  Air emissions constitute the vast majority of the on-site releases, with a steady
decrease over the years. The decrease in air emissions, as well as surface water discharges and a general
decrease in off-site emissions (including metals or metal compounds transferred off-site), have
contributed to decreases in tetrachloroethylene total on- and off-site releases in recent years. Releases to
land (such as spills or leaks within the boundaries of the reporting facility) display no particular trend,
and underground injection has fluctuated between 5,000 and 20,000 pounds, with the exception of high
levels of releases in 1988-1989. These TRI data for tetrachloroethylene were reported from 48 States
(with the exceptions of Alaska and Wyoming), Puerto Rico, and the Virgin Islands (USEPA, 2000).
Thirty-one of the 48 States reported every year.  All 16 of the cross-section States (used for analyses of
tetrachloroethylene occurrence in drinking water; see Section 4.8.4) reported releases of
tetrachloroethylene. (For a map of the 16-State cross-section, see Figure  1.3-1.)
Table 4.8-2: Environmental Releases (in pounds) for Tetrachloroethylene in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
3,648,732
5,463,597
7,213,469
7,995,845
9,674,185
10,826,150
11,357,864
12,816,529
17,384,881
23,000,633
27,813,294
36,124,485
Surface Water
Discharges
1,793
1,490
2,282
1,561
2,407
3,877
10,157
10,322
7,453
21,510
53,940
33,314
Underground
Injection
8,897
5,916
15,118
13,436
20,481
4,051
15,041
12,780
14,005
11,012
50,000
72,250
Releases
to Land
19,885
2,992
5,074
5,472
6
4,349
618,026
9,754
23,309
1,260
10,791
82,144
Off-Site Releases
27,966
130,927
29,228
23,212
78,953
80,255
56,340
113,324
115,933
796,846
1,044,249
1,385,378
Total On- &
Off-site
Releases
3,707,273
5,604,922
7,265,171
8,039,526
9,776,032
10,918,682
12,057,428
12,962,709
17,545,581
23,831,261
28,972,274
37,697,571
 Source: USEPA, 2000
4.8.3 Ambient Occurrence

Tetrachloroethylene was detected in 67 out of 403 wells (16.6%) in urban areas of the local, State, and
federal data set compiled by NAWQA. The minimum and maximum concentrations detected were 0.2
l-ig/L and 260 |ig/L, respectively. The median value of detection concentrations was 1.1 M-g/L.
Tetrachloroethylene was also detected in 64 of the 2,526 wells (2.5%) with analysis in rural areas. The
minimum and maximum concentrations detected were 0.2 |ig/L and 250 |ig/L, respectively. The median
value of detection concentrations was 0.8 |ig/L.  These data (urban and rural) represent untreated ambient
ground water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).
Occurrence Summary and Use Support Document
407
March 2002

-------
Tetrachloroethylene was also an analyte in both the NURP and NPDES data (Lopes and Dionne, 1998).
In a comparison of the two data sets, the frequency of detection of tetrachloroethylene for NURP was
5%, while the frequency of detection for NPDES was 8%. The NURP study found tetrachloroethylene in
urban runoff. The minimum and maximum concentrations detected were 4.5 |ig/L and 43 |ig/L,
respectively, with no mean value reported. The NPDES related investigations analyzing urban and
highway runoff detected tetrachloroethylene. The minimum and maximum concentrations detected were
<0.2 |ig/L and 42 |ig/L, respectively, with no mean value reported.  The use of the land from which the
samples were taken was unspecified.

4.8.3.1 Additional Ambient Occurrence Data

A summary document, entitled "Occurrence of Tetrachloroethylene in Drinking Water, Food and Air"
(JRB Associates, 1983), was previously prepared for past USEPA assessments of tetrachloroethylene.
However, no information on the ambient occurrence of tetrachloroethylene was included in that
document. (The document did include information regarding tetrachloroethylene occurrence in drinking
water, which is discussed in Section 4.8.5 of this report.)

4.8.4 Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of tetrachloroethylene occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to  date.  These data were
evaluated relative to two concentration thresholds of interest: 0.005 mg/L; and 0.0005 mg/L.

All sixteen cross-section State data sets contained occurrence data for tetrachloroethylene. These data
represent more than 195,000 analytical results from approximately 22,000 PWSs during the period from
1984 to 1998 (with most analytical results from 1992 to 1997). The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to  ensure adequacy of
coverage and completeness. The overall modal detection limit for tetrachloroethylene in the 16 cross-
section States is equal to 0.0005  mg/L. (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)

4.8.4.1  Stage 1 Analysis Occurrence Findings

Table 4.8-3 illustrates the occurrence of tetrachloroethylene in drinking water for the public water
systems in the 16-State cross-section.  The percentage of total ground and surface water PWSs with any
analytical results exceeding the MCL (0.005 mg/L) was equal to 0.778% (a total of 174 systems).
Approximately 2.89% of total ground  and surface water systems (646 PWSs) had any analytical results
greater than the modal detection limit  (0.0005 mg/L).

A greater proportion of surface water  systems, as compared to ground water systems, exceeded each
threshold. Yet, a greater number of ground water systems exceeded each threshold, as compared to
surface water systems.  Approximately 0.692% of ground water PWSs (144 systems) had any analytical
results exceeding the MCL, compared to 1.91% of surface water systems (30 systems). The percentage
of ground water PWSs with any analytical results exceeding 0.0005 mg/L was equal to about 2.72% (565
ground water systems). Close to 5.2% of surface water PWSs (a total of 81 systems) had at least one
analytical result greater than 0.0005 mg/L.
Occurrence Summary and Use Support Document         408                                     March 2002

-------
Table 4.8-3:  Stage 1 Tetrachloroethylene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Percent of Systems
Exceeding Threshold
0.692%
2.72%
Number of Systems
Exceeding Threshold
144
565

Surface Water
0.005
0.0005
1.91%
5.17%
30
81

Combined Ground &
Surface Water
0.005
0.0005
0.778%
2.89%
174
646
Reviewing tetrachloroethylene occurrence by PWS population served (Table 4.8-4) shows approximately
13.5% of the total 16-State cross-section population (almost 15 million people) was served by PWSs with
at least one analytical result greater than the MCL.  Approximately 22.9% of the population (over 25
million people) was served by systems with analytical results greater than 0.0005 mg/L.

About 7.13% of the population served by ground water systems (over 3.5 million people) was exposed to
tetrachloroethylene levels above the MCL. Approximately 19.5% of the population (almost 10 million
people) was served by ground water systems with at least one analytical result greater than 0.0005 mg/L.

Approximately 18.6% of the population (over 11 million people) was served by surface water PWSs with
at least one analytical results greater than the MCL. About 25.7% of the population (almost 16 million
people) was served by surface water PWSs with any analytical results greater than 0.0005 mg/L.
Table 4.8-4:  Stage 1 Tetrachloroethylene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Percent of Population Served
by Systems
Exceeding Threshold
7.13%
19.5%
Total Population Served
by Systems Exceeding
Threshold
3,524,300
9,633,800

Surface Water
0.005
0.0005
18.6%
25.7%
11,390,000
15,717,500

Combined Ground &
Surface Water
0.005
0.0005
13.5%
22.9%
14,914,300
25,351,400
Occurrence Summary and Use Support Document
409
March 2002

-------
4.8.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.8-5 and 4.8-6.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

Approximately 45 (0.202% of all) ground water and surface water PWSs in the 16 States were estimated
to have mean concentrations of tetrachloroethylene above 0.005 mg/L. The percentage of PWSs with
estimated mean concentrations exceeding 0.0005  mg/L was about 1.62% of PWSs in the 16 States (362
systems).

Similar to the Stage 1 analysis, Stage 2 analysis estimated a greater proportion of surface water systems,
as compared to ground water systems, exceeding  each threshold, but a greater number of ground water
systems exceeding  each threshold.  Approximately 40 (0.194% of) ground water systems in the 16 States
had estimated mean concentrations of tetrachloroethylene above 0.005 mg/L, compared to approximately
5 (0.317% of) surface water systems in the 16 States.  The estimated mean concentration values for
approximately 324  (1.56%) ground water PWSs in the 16 States exceeded 0.0005 mg/L. About 39
(2.48% of) surface  water systems in the 16 States had estimated mean concentrations exceeding the
modal detection limit.
Table 4.8-5:  Stage 2 Estimated Tetrachloroethylene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.194%
1.56%
Range
0.154% -0.231%
1.43% -1.68%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
40
324
Range
32-48
297 - 349

Surface Water
0.005
0.0005
0.317%
2.48%
0.191% -0.447%
2.17% -2.87%
5
39
3-7
34-45

Combined Ground
& Surface Water
0.005
0.0005
0.202%
1.62%
0.166% -0.237%
1.50% -1.74%
45
362
37-53
335 - 390
Reviewing tetrachloroethylene occurrence by PWS population served (Table 4.8-6) shows that over
757,000 (0.685% of) people in the 16 States were served by systems with mean tetrachloroethylene
concentrations above 0.005 mg/L. When evaluated relative to a threshold of 0.0005 mg/L, the percent of
population exposed increased significantly to about 12.2% (approximately 13.5 million people served in
the 16 States).
Occurrence Summary and Use Support Document
410
March 2002

-------
Of the 16-State cross-section population served by ground water, approximately 0.833% of the
population was served by PWSs (almost 412,000 people served in the 16 States) with estimated mean
concentrations of tetrachloroethylene above 0.005 mg/L.  An estimated 3.6 million (7.36% of) people
served by ground water systems in the 16 States were exposed to tetrachloroethylene above 0.0005 mg/L.

Approximately 0.565% of the population (345,600 people in the 16 States) was served by surface water
PWSs with estimated mean concentrations greater than 0.005 mg/L. The percentage of population served
by surface water PWSs with estimated mean concentrations greater than 0.0005 mg/L increased
dramatically to approximately 16.2% (almost 10 million people in the 16 States).
Table 4.8-6:  Stage 2 Estimated Tetrachloroethylene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.833%
7.36%
Range
0.642% -1.10%
6.72% - 8.32%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
411,600
3,636,800
Range
317,200-544,500
3,319,200-4,112,100

Surface Water
0.005
0.0005
0.565%
16.2%
0.218% -0.777%
15.2% -18.6%
345,600
9,875,900
133,500-475,400
9,288,800-11,392,400

Combined Ground
& Surface Water
0.005
0.0005
0.685%
12.2%
0.470% - 0.873%
11. 5% -13. 7%
757,200
13,510,200
519,700-965,100
12,747,300-15,102,200
4.8.4.3 Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 132 PWSs serving almost 1.5 million people nationally could be exposed to
tetrachloroethylene concentrations above 0.005 mg/L.  About 1,053 systems serving over 26 million
people nationally had estimated mean concentrations greater than 0.0005 mg/L.  (See Section 1.4 for a
description of how Stage 2 16-State estimates are extrapolated to national values.)

For ground water systems, an estimated 115 PWSs serving about 713,800 people nationally had mean
concentrations greater than 0.005 mg/L. Approximately 925 systems serving about 6.3 million people
nationally had  estimated mean concentration values that exceeded 0.0005 mg/L.

Approximately 18 surface water systems serving about 719,500 people nationally were estimated to have
mean concentrations of tetrachloroethylene above 0.005 mg/L. About 139 surface water systems serving
almost 21 million people had estimated mean concentrations greater than 0.0005 mg/L.
Occurrence Summary and Use Support Document
411
March 2002

-------
Table 4.8-7:  Estimated National Tetrachloroethylene Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
115
925
Range
91-137
849 - 997
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
713,800
6,307,000

Surface Water
0.005
0.0005
18
139
11-25
121 -161
719,500
20,563,200

Combined Ground
& Surface Water
0.005
0.0005
132
1,053
108-154
974-1,134
1,458,900
26,029,600
Range
550,200 - 944,200
5,756,100-7,131,300

278,000 - 989,800
19,340,900-23,720,900

1,001,400-1,859,300
24,559,800 - 29,096,900
4.8.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of tetrachloroethylene in drinking water are also
reviewed. Previously compiled occurrence information, from an OGWDW summary document entitled
"Occurrence of Tetrachloroethylene in Drinking Water, Food, and Air" (JRB Associates, 1983), is
presented in this section.  This variety of studies and information are presented regarding levels of
tetrachloroethylene in drinking water, with the scope of the reviewed studies ranging from national to
regional. Note that none of the studies presented in the following section provide the quantitative
analytical results or comprehensive coverage that would enable direct comparison to the occurrence
findings estimated with the cross-section occurrence data presented in Section 4.8.4. These additional
studies, however, do enable a broader assessment of the Stage 2 occurrence estimates presented for this
Six-Year Review. All the following information in Section 4.8.5 is taken directly from "Occurrence of
Tetrachloroethylene in Drinking Water, Food, and  Air" (JRB Associates, 1983).

JRB Associates (1983) found two major types of data available that are potentially useful for describing
the occurrence of tetrachloroethylene in the nation's public drinking water supplies. First, there are
several Federal surveys in which a number of public water supplies from throughout the U.S. were
selected for analysis of chemical  contamination, including tetrachloroethylene. Second, data are
available from State surveys and  from State investigations of specific incidents of known or suspected
contamination of a supply. For accomplishing the  basic objectives of this study, namely to estimate the
number of public water supplies nationally within the various  source and size categories contaminated
with tetrachloroethylene, the distribution of tetrachloroethylene concentrations in those supplies, and the
number of individuals exposed to those concentrations, it was determined that the Federal survey data
provides the most suitable data base. The State data tend to be poorly described with respect to the
source and  size categories of the  supplies examined and the sampling and analysis methods used for
determining contaminant levels.  The lack of source and system size information precludes using the data
for estimating levels in public water supplies of similar characteristics. The absence of details on
sampling and analysis methods precludes evaluating those data for their qualitative and quantitative
reliability.  Also, because much of the State data are from investigations in response to incidents of
known or suspected contamination (e.g., spills), they were judged to be not representative of contaminant
Occurrence Summary and Use Support Document
412
March 2002

-------
levels in the nation's water supplies in general. Although they are not used with the Federal data for the
purpose of estimating contamination levels nationally, the available State data are presented here to
provide some additional perspective on tetrachloroethylene occurrence in drinking water.

Data are presented only on drinking water samples taken from a consumer's tap (i.e., distribution water
samples) or on treated water samples taken at the water supply (i.e., finished water samples) because
these are considered to be most representative of the water consumed by the public. No data on raw (i.e.,
untreated) water are presented. It is recognized that for some groundwater supplies where no treatment
of the water occurs, samples identified as raw may be representative of water consumed by the users of
the supply. However, it was generally not possible to differentiate between those groundwater supplies
that do and those that do not treat raw  water from the available survey data.

4.8.5.1  Overview and Quality Assurance Assessment of Federal Drinking Water Surveys

Five Federal drinking water surveys provide data on tetrachloroethylene: the National Organic
Monitoring Survey (NOMS), the National Screening Program for Organics in Drinking Water (NSP), the
1978 Community Water Supply Survey (CWSS), the Rural Water Survey (RWS), and the Groundwater
Supply Survey (GWSS). The terms used in this report are those used in the individual surveys,
recognizing that they may not always correspond to strict technical definitions.

The National Organic Monitoring Survey (NOMS) was conducted to identify contaminant sources, to
determine the frequency of occurrence of specific drinking water contaminants, and to provide data for
the establishment of maximum contaminant levels (MCL's) for various organic compounds in drinking
water (Brass et al., 1977, as cited in JRB Associates, 1983).  The NOMS was conducted in three phases:
March-April 1976, May-July 1976, and November 1976-January  1977. For tetrachloroethylene,
qualitative results are available for Phase II and quantitative results for Phase III.  Tetrachloroethylene
was not analyzed for in Phase I. Finished drinking water samples from 113 communities  were analyzed
for 21 different compounds. Of the 113 community supplies sampled, 18 had groundwater sources, 91
had surface water sources, and 4 had a mixed  groundwater/surface water source.  In Phase III, only 14
groundwater supplies had data available for tetrachloroethylene.

The analytical results of the NOMS were made available in printed form by EPA's Technical Support
Division, Office of Drinking Water. Additional information on the locations and  source of the supplies,
and on the populations served by the supplies in the NOMS were provided by Wayne Mello (1983, as
cited in JRB Associates, 1983) at EPA's Technical Support Division, Office of Drinking Water. A single
value for tetrachloroethylene was  reported for each supply studied in the NOMS.

The National Screening Program for Organics in Drinking Water (NSP), conducted by SRI International
from June 1977 to March  1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic chemical contaminants.  Data are available for tetrachloroethylene on
finished water samples from 12 groundwater and 106 surface water supplies.

In the Community Water Supply Survey (CWSS), carried out in 1978, 106 surface water  supplies, 330
groundwater supplies, and 16 supplies with mixed sources were examined for volatile organic chemical
contamination.  Samples were taken of raw, finished, and distribution water.  Only the latter two types of
water are considered here. Data for tetrachloroethylene in finished and/or distribution samples were
obtained from a total of 316 groundwater and 105 surface water supplies.
Occurrence Summary and Use Support Document         413                                     March 2002

-------
The Rural Water Survey (RWS), conducted in 1978, was carried out in response to Section 3 of the Safe
Drinking Water Act, which mandated that EPA "conduct a survey of the quantity, quality, and
availability of rural drinking water supplies." Drinking water samples were collected for analysis of
inorganic chemicals, pesticides, and VOCs from 2,655 households throughout the United States located
in areas defined in the survey as rural.  Of these, a total of 855 household samples were examined for
VOCs.  The majority of these samples were obtained from households receiving water from private wells
or small supplies serving fewer than 25 people. For tetrachloroethylene, data are available in the RWS
for 207 groundwater and 45 surface water supplies serving 25 or more people.

The RWS did not obtain data on the number of persons in each household served by the supplies.
However, data were obtained on the number of service connections at each supply. With the input of Dr.
Bruce Brower at Cornell University, who participated in the statistical analysis of the RWS for
parameters other than VOCs, the population served by each supply was estimated from the average
number of persons per household (3.034) observed in the survey.  A single value was reported for each
household; in some cases it was necessary to average two or three households obtaining water from the
same supply.  Brass (1981, as cited in JRB Associates, 1983) cautions that the RWS water samples were
analyzed 6 to 27 months after collection and that degradation of some VOCs may have occurred during
this holding period.

The Groundwater Supply Survey  (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals in the nation's groundwater
supplies (Westrick et al, 1983, as cited in JRB Associates, 1983). It was hoped that this study would
stimulate State efforts toward the  detection and control of groundwater contamination and the
identification of potential chemical "hot spots." A total of 945 systems were sampled, of which 466 were
chosen at random. The remaining 479 systems were chosen non-randomly based on information from
States encouraged to identify locations  believed to have a higher than normal probability of VOC
contamination (e.g., locations near landfills or industrial activity). The file provided a single analytical
result for each supply sampled. One sample of finished water was collected from each supply at a point
near the entrance to the distribution system.

Each of the drinking water surveys was evaluated  with respect to the validity of the reported occurrence
data for a number of organic chemicals, including  tetrachloroethylene.  The evaluations were carried out
by analyzing information about the procedures used for collection and analysis of samples as well as the
quality control protocols used.  The analyzed compounds dealt with in each study were assigned one of
three possible ratings: quantitatively acceptable, qualitatively acceptable (i.e., the substance measured
was tetrachloroethylene), and totally unacceptable. In the case of tetrachloroethylene, a qualitatively
acceptable rating was given for data from the NOMS, NSP, CWSS, and RWS because of suspected
biodegradation of the samples, which were held unrefrigerated for prolonged periods before analysis
(particularly CWSS and RWS). Tetrachloroethylene values in excess of the quantitation limit reported
for some samples in these studies are qualitatively valid and can be taken as minimum values,
representative of samples which probably originally contained tetrachloroethylene at higher
concentrations.  In the case of the GWSS, all data  were rated both quantitatively and qualitatively
acceptable.

4.8.5.2  Groundwater - Federal Surveys

The National Organic Monitoring Survey (NOMS), the National Screening Program for Organics in
Drinking Water (NSP), the Community Water Supply Survey (CWSS), the Rural Water Survey (RWS),
Occurrence Summary and Use Support Document         414                                     March 2002

-------
and the Groundwater Supply Survey (GWSS) all contain data concerning the levels of
tetrachloroethylene in groundwater supplies from across the country.

Tetrachloroethylene was not analyzed for in Phase I of NOMS (March to April 1976). In NOMS Phase
II (May to July 1976), a qualitative assessment that tetrachloroethylene was present in 6 of the 18
groundwater supplies sampled was made on the basis of the presence of a peak on the gas chromatograph
at the appropriate retention time for tetrachloroethylene. As indicated previously, a rigorous analysis of
all samples was not made, and some  samples with smaller tetrachloroethylene peaks may have been
overlooked. Samples analyzed during Phase III of the study (November 1976 to January 1977) proved
positive for two of the 14 systems for which data are available, with levels of 1.1  and 3.1 |ig/L. The
minimum quantifiable limit for tetrachloroethylene was 0.2 |ig/L in Phase III.

Twelve groundwater supplies were tested for tetrachloroethylene contamination in the NSP. Of these 12
systems, three were found to be contaminated with tetrachloroethylene at levels of 0.1, 0.2, and 0.7 |ig/L.
The quantification limit for tetrachloroethylene was 0.1 |ig/L.

The 1978 CWSS provided information on tetrachloroethylene levels in 316 groundwater systems. Of
these  systems, 18 contained detectable levels of tetrachloroethylene, with values ranging from 0.51-15.3
l-ig/L. The two highest values were 3.1 |ig/L and 15.3 |ig/L; all other values were less than 3.0 |ig/L.  The
mean value was 2.1 |ig/L with a standard deviation of 3.4 |ig/L; the median value was 1.2 |ig/L.  The
minimum quantitation limit for tetrachloroethylene in the CWSS was 0.5  |ig/L.

The RWS examined 207 groundwater supplies for tetrachloroethylene and found 7 to have levels above
the minimum quantification limit range of 0.5-1.5 |ig/L. The range of positive values was 1.0 to 8.2
l-ig/L. The mean of the positive values was  3.2 |ig/L with a standard deviation of 2.5 |ig/L; the median
concentration was 2.1 |ig/L. In the GWSS,  34 of the 456 randomly chosen water systems serving 25  or
more  individuals were contaminated with tetrachloroethylene, at concentrations ranging from 0.21-23
l-ig/L. The three systems with the highest values were contaminated at 2.9, 5.9, and 23 |ig/L.  Twenty-one
of the 34 positive systems served populations in excess of 10,000 people.  The average for all randomly
chosen systems was 1.5 |ig/L with a  standard deviation of 4 |ig/L; the median was 0.5 |ig/L. Of the 473
nonrandom locations sampled serving 25 or more individuals, 43 were contaminated with
tetrachloroethylene, at concentrations between 0.22-69 |ig/L, the highest values being 20, 21, and 69
l-ig/L. Of the 43 positive samples, 18 were from systems serving populations in excess of 10,000 people.
The average tetrachloroethylene level for the nonrandom systems was 4.7 |ig/L with a standard deviation
of 11  M-g/L; the median value was 0.73 |ig/L. The minimum quantitation limit for tetrachloroethylene
was 0.2 |ig/L.

4.8.5.3 Groundwater - State Data

Seven States (California, Connecticut, Delaware, Indiana, Massachusetts, New Jersey, and New York)
provided the USEPA with information concerning tetrachloroethylene contamination in groundwater
supplies. Analytical results for 407 samples from 16 areas in California revealed 257 positive samples
with tetrachloroethylene concentrations from 0.03 to "greater than  20" |ig/L. Monitoring data from 55
locations in Connecticut revealed a wide  range of contamination levels. Fifteen cities were supplied with
water containing no detectable tetrachloroethylene.  Of the 40 systems with detectable
tetrachloroethylene, 20 had levels of 2.5 |ig/L or less, 13 had levels between 2.5-10 |ig/L, and seven were
contaminated at 10-640 |ig/L. Data from Delaware indicated that of eight samples taken from finished
groundwater supplies in two counties, five contained tetrachloroethylene at 0.79-7.1 |ig/L.  Information
supplied by Indiana showed that eight samples from one system were free of detectable

Occurrence Summary and Use Support Document         415                                      March 2002

-------
tetrachloroethylene contamination.  Of five systems sampled in Massachusetts, four were free of
detectable contamination while another had between 24-49 |ig/L. Data from New Jersey showed a large
variation in the concentration of measured tetrachloroethylene contamination, ranging from undetectable
to 870 |ig/L. Finally, of 421 samples from Nassau County, New York, 67 reported tetrachloroethylene to
be present qualitatively.

4.8.5.4 Surface Water - Federal Surveys

The National Organic Monitoring Survey (NOMS), the National Screening Program for Organics in
Drinking Water (NSP), the Community Water Supply Survey (CWSS), and the Rural Water Survey
(RWS) all contain data concerning the levels of tetrachloroethylene in surface water supplies from across
the country.

Tetrachloroethylene was not analyzed for in Phase I of NOMS. In NOMS Phase II (May to July 1976), a
qualitative assessment that tetrachloroethylene was present in 40 of 91 surface water supplies sampled
(and in one of the mixed source supplies using surface water ,during Phase II) was made on the basis of
the presence of apeak on the gas chromatograph at the appropriate retention time for tetrachloroethylene.
As indicated previously, a rigorous analysis of all samples was not made, and some samples with smaller
tetrachloroethylene peaks may have been overlooked. During Phase III of NOMS (November 1976 to
January 1977), analyses revealed tetrachloroethylene  contamination in 7 out of a total of 87 systems.
The contamination levels for these systems ranged from 0.2-0.82 |ig/L. The mean concentration was 0.4
|ig/L with a standard deviation of 0.2 |ig/L; the median level was 0.45 |ig/L.  The minimum quantifiable
limit for tetrachloroethylene was 0.2 |ig/L in Phase III.

Surface water samples from 106 drinking water systems were analyzed for tetrachloroethylene during the
National Screening Program (NSP) between June 1977 and March 1981.  Of these,  16 systems contained
detectable levels of tetrachloroethylene, ranging from 0.1-3.2 |ig/L. Only two of these systems were
contaminated at levels greater than 1.0 |ig/L (1.1  and  3.2 |ig/L).  The average concentration among the 16
positive systems was 0.6 |ig/L with a standard deviation of 0.8 |ig/L; the median level was 0.4 |ig/L. The
quantification limit for the NSP was 0.1 |ig/L.

Of the 105 surface water systems  sampled during the  Community Water Supply Survey (CWSS), 2
contained quantifiable levels of tetrachloroethylene. The positive values were 0.5 and 1.7 |ig/L.  The
minimum quantitation limit for tetrachloroethylene in the CWSS was 0.5  |ig/L. The RWS examined
drinking water from 45 surface water supplies; none were found to have tetrachloroethylene present
above the minimum quantification limit range of 0.5-5 |ig/L.

4.8.5.5 Surface Water - State Data

Three States provided the USEPA with data on tetrachloroethylene from surface water sources. Of 12
cities sampled in Connecticut, 10 had tetrachloroethylene contamination,  with one mean concentration
higher than 10 |ig/L (at 31 |ig/L).  Two of three locations sampled in New York had detectable
tetrachloroethylene ranging from 0.49 to 2.8 |ig/L.  One location in California was sampled twice,
revealing no detectable tetrachloroethylene.
Occurrence Summary and Use Support Document         416                                     March 2002

-------
4.8.5.6 Projected National Occurrence of Tetrachloroethylene in Public Water Supplies

As reported in the JRB Associates study, public water systems fall into two major categories with respect
to water source (surface water and groundwater) and into five size categories and twelve subcategories
according to the number of individuals served. The JRB Associates (1983) report presented estimates of
both the number of drinking water supplies nationally within each of the source/size categories expected
to have tetrachloroethylene present, and of the concentration of tetrachloroethylene expected to be
present in those supplies.

The key features of the methodology used and assumptions made to develop the national estimates are
summarized here. The estimates are based on the data from the Federal surveys only.  The State data
were not included for several reasons.  Generally, these data are from a few States (e.g., California and
Connecticut provide most of the data) and were not considered to be geographically representative.
There was also a general lack of data on the population served by systems measured, the type of water
sampled, and the methodologies used to sample, identify, and measure tetrachloroethylene.  Furthermore,
since much of these data were apparently obtained in response to incidents of recognized contamination
problems, they may not be representative of typical conditions existing nationally.  However, while these
data were not used for computing the national projections, they do provide a valuable and necessary
perspective for evaluating those projections, especially with respect to the projected high levels of
contamination of tetrachloroethylene.

The Federal survey data from the NOMS, NSP, CWSS, RWS, and GWSS were pooled together for
developing the national projections. It was assumed in combining these surveys that the resulting data
base would be representative of the nation's water supplies. In the case of the GWSS data, both the
random and nonrandom samples were included in the projections because a statistical test of the GWSS
data showed that neither the frequency of occurrence of positive values nor the mean of the positive
values for tetrachloroethylene was significantly different in the two samples.

Ideally, adequate data would be available to develop the national projections  separately for each of the
twelve system size categories within the groundwater and surface water groups; however, the available
data were too limited for this. JRB Associates (1983) consolidated some of the size categories to have
sufficient data for developing the projections.  In consolidating data from various size categories,
consideration was given to the potential for there being statistically significant differences in the
frequency of occurrence of tetrachloroethylene as a function of system size.  The consolidation of size
categories therefore involved a balancing of the need to group size categories together to have an
adequate data base for developing the national projections against the need to treat size categories
separately in order to preserve the influence of system size as a determinant of contamination potential.
The consolidation of size categories also took into account EPA's classification of systems into the five
major groups as very small (25-500), small (501-3,300), medium (3,301-10,000), large (10,001-100,000),
and very large (>  100,000) (Kuzmack, 1983, as cited in JRB Associates, 1983).

Once the data were consolidated, statistical models for extrapolating to the national level were tested and
an appropriate model selected.  In the case of tetrachloroethylene, the multinominal method was used.
The frequency of contamination of groundwater and surface water systems at various concentrations was
determined for each consolidated size category.  For completing the national estimates, it was assumed
that the frequency of contamination observed for each consolidated category was directly applicable to
each of the  system sizes comprising it.
Occurrence Summary and Use Support Document          417                                     March 2002

-------
In the JRB Associates (1983) report, it is noted that some of the data used in computing the national
estimates are from samples held for a prolonged period of time prior to analysis, with possible
biodegradation of tetrachloroethylene.  Therefore, these projections of national occurrence may
underestimate  actual contaminant levels.

4.8.5.6.1 Groundwater Supplies

JRB Associates (1983) reported that data were available for a total of 1,457 supplies from the combined
surveys.  Of these, 103 supplies were reported to have tetrachloroethylene present, at concentrations
ranging from 0.20 |ig/L to 69 |ig/L.

Based on the overall distribution of positive values and maximum possible values for those supplies in
which tetrachloroethylene was not found, 0.5 |ig/L was selected as the common minimum quantifiable
concentration for the combined survey data. That is, quantitative projections are made of supplies at
several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either
no tetrachloroethylene or levels below 0.5 |ig/L can be determined. Although some data indicate the
presence of tetrachloroethylene in groundwater supplies at levels < 0.5 |ig/L, it is not possible to
determine the proportion of supplies that have tetrachloroethylene present and the proportion that are
actually free of tetrachloroethylene contamination.

Of the  1,354 supplies reporting no tetrachloroethylene to be present, 1,244 were assumed to have
maximum possible levels of < 0.5 |ig/L based on the minimum quantifiable concentrations reported for
the various surveys (857 of these had maximum levels of < 0.2 |ig/L). The other 110 supplies reporting
no tetrachloroethylene to be present had maximum possible levels ranging from 1.0 |ig/L to 1.5 |ig/L. It
is assumed, based on the overall distribution of values, that tetrachloroethylene if present in these 110
supplies is so at a concentration of < 0.5 |ig/L, although a rigorous, conservative argument could be made
for assuming a level equal to the maximum possible value. The impact of this assumption is considerable
both in terms of the national projection of groundwater systems between 0.5 and 5 |ig/L and the size of
the population exposed to that concentration range.

The data indicate that 68 of 1,457 supplies examined had measured values of tetrachloroethylene > 0.5
l-ig/L. When the twelve size categories were consolidated into the five major EPA groupings, there was
an apparent general trend in the frequency of values > 0.5 |ig/L as a function of size:
Very small
Small
Medium
Large
Very large
Overall
3.6% (15/423)
2.2% (9/416)
8.5% (18/213)
5.6% (20/358 )
12.8% (6/47)
4.7% (68/1,457)
A test for statistical significance revealed that, at the a = 0.05 level, the difference between the very
small and small categories was not significant; nor was the difference between the large and very large
groups. The medium category showed a statistically significant difference from the combined very small
and small categories. Three consolidated categories were, therefore, selected for developing the national
estimates:

                                   Very small/small (25-3,300)
                                      Medium (3,301-10,000)

Occurrence Summary and Use Support Document         418                                      March 2002

-------
                                    Large/very large (> 10,000)

As noted previously, the frequency of occurrence of tetrachloroethylene at various concentrations was
determined for the consolidated groups and then applied to the number of supplies nationally within each
of the size categories comprising each group.

An estimated 1,557 groundwater supplies (range of 1,047-2,066), approximately 3.2% of the total
groundwater supplies in the United States, are expected to have tetrachloroethylene at levels of > 0.5
l-ig/L; the remaining 46,901 supplies have either no tetrachloroethylene or levels < 0.5 |ig/L.  It is
estimated that 325 supplies (range of 89-560) are expected to have tetrachloroethylene levels > 5 |ig/L;
while only 3 supplies (range of 0-187) are expected to have levels > 30 |ig/L.  None are expected at
levels above 70 |ig/L. Most of the supplies with high tetrachloroethylene levels are expected to be in the
smaller size categories. Although, as noted previously, the frequency of tetrachloroethylene occurrence
increases with increasing system size, the number of systems affected nationally is greater for the small
sizes because there are many more small systems in existence.

It is interesting to note the impact on the national projections of the assumption made that the 1 10
supplies with undetected but maximum potential values of 1.0-1.5 |ig/L had < 0.5 |ig/L. Had it been
assumed that tetrachloroethylene was present in those supplies at their maximum possible values, the
national projection of supplies with tetrachloroethylene  levels > 0.5  |ig/L would have increased to 6,278
(5,258-7,297) with no differences in levels > 5 |ig/L. These  differences in the 0.5-5 |ig/L range would be
found predominantly in systems serving < 10,000 people.

4.8.5.6.2 Surface Water Systems

Data are available for a total of 296 surface water supplies. Of these, 24 supplies were reported by JRB
Associates (1983) to have tetrachloroethylene present at concentrations ranging from 0.1 |ig/L to 3.2
Based on the overall distribution of positive values and maximum possible values for those supplies in
which tetrachloroethylene was not found, 0.5 |ig/L was selected as the common minimum quantifiable
concentration for the combined survey data. That is, quantitative projections are  made of supplies at
several concentration ranges > 0.5 |ig/L, while only a total number for supplies expected to have either
no tetrachloroethylene or levels below 0.5 |ig/L can be determined.  Although some data indicate the
presence of tetrachloroethylene in surface water supplies at levels < 0.5 |ig/L, it is not possible to
determine the proportion that have tetrachloroethylene present and the proportion that are free of
tetrachloroethylene contamination .

Of the 272 supplies reporting no tetrachloroethylene to be present, 243 had maximum possible levels of <
0.5 |ig/L based on the minimum quantifiable concentrations reported for the various surveys. The other
29 supplies reporting no tetrachloroethylene to be present had maximum possible levels ranging from 1.0
|ig/L to 1.5 |ig/L. It is assumed, based on the overall distribution of values, that tetrachloroethylene if
present in these 29 supplies is so at a concentration of < 0.5 |ig/L, although a rigorous conservative
argument could be made for assuming a level equal to the maximum possible value.  As will be noted
further below, the difference between these alternatives is considerable for the estimate of the number of
surface water supplies with tetrachloroethylene at 0.5-5 |ig/L; also, the impact on the  estimated
population exposed to tetrachloroethylene at levels of 0.5-5 |ig/L in  surface water supplies is very large.
Occurrence Summary and Use Support Document         419                                      March 2002

-------
The data also indicates that 11 of the 296 supplies examined had measured values of tetrachloroethylene
>0.5 |ig/L. When the twelve size categories are consolidated into the five major EPA groupings, the
frequency of values >  0.5 |ig/L as a function of size are:
Very small
Small
Medium
Large
Very large
Overall
0%
1.7%
0%
6.1%
5.3%
3.7%
(0/20)
(1/58)
(0/39)
(4/66)
(6/113)
(11/296)
A test for statistical significance revealed that, at the a = 0.05 level, the very small, small, and medium
groups were not different from one another and that the large and very large groups are not different;
however, the combined very small, small, and medium groups and the combined large and very large
groups are different.  These two consolidated categories were selected for developing the national
estimates:

                               Very small/small/medium (25-10,000)
                                    Large/very large (> 10,000)

As noted previously, the frequency of occurrence of tetrachloroethylene at various concentrations was
determined for the consolidated groups and then applied to the number of supplies nationally within each
of the size categories comprising each group.

About 180 surface water supplies (range of 13-346), approximately 1.6% of the total surface water
systems in the United States, are expected to have tetrachloroethylene at levels between 0.5  and 5 |ig/L;
the remaining 11,022 supplies have either no tetrachloroethylene or levels < 0.5 |ig/L. It is estimated that
no surface water supplies will have levels > 5 |ig/L.

There was a notable impact on the national projections by the assumption made that the 29 supplies with
undetected but maximum potential values of 1.0-1.7 |ig/L had < 0.5 |ig/L.  Had it been assumed that
tetrachloroethylene was present in these supplies at their maximum possible values, the national
projections of supplies with tetrachloroethylene levels of 0.5-5 |ig/L would be 1,953 (range of 1,282-
2,623) supplies.  However, there would still be no surface water supplies expected to have levels > 5
p-g/L.

4.8.6 Conclusion

 Tetrachloroethylene is a commercially important chlorinated hydrocarbon solvent and chemical
intermediate in the production of chlorofluorocarbons. It is used as a dry cleaning and textile-processing
solvent and for vapor degreasing in metal-cleaning operations.  Production of tetrachloroethylene has
decreased in recent years, although it is still widely produced and processed.  Tetrachloroethylene is also
a TRI chemical.  Industrial releases of tetrachloroethylene have occurred since 1988 in 48 States, Puerto
Rico, and the Virgin Islands. Tetrachloroethylene was an analyte for the NAWQA, NURP,  and NPDES
ambient occurrence studies. In the NAWQA study, tetrachloroethylene was detected in 16.6% of urban
wells and 2.5% of rural wells, with median detection values of 1.1 |ig/L and 0.8 |ig/L, respectively.  The
frequency of detection of tetrachloroethylene for NURP was 5%, while the  frequency of detection for
NPDES was 8%.  In the Stage 2 analysis of 16-State occurrence of tetrachloroethylene, 0.202% of
combined ground water and surface water systems serving 0.685% of the population exceeded the MCL

Occurrence Summary and Use Support Document          420                                     March  2002

-------
of 0.005 mg/L. Nationally, 132 ground water and surface water systems combined (serving
approximately 1,458,900 people) are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, tetrachloroethylene is manufactured and/or processed in 40 States and has
TRI releases in 48 States.  Tetrachloroethylene is manufactured and/or processed in 14 out of the 16
cross-section States and has TRI releases in all of the 16 cross-section States.  The cross-section should
adequately represent the occurrence of tetrachloroethylene on a national scale based upon the use,
production, and release patterns of the 16-State cross-section in relation to the patterns observed for all
50 States.

4.8.7  References

Agency for Toxic Substances and Disease Registry (ATSDR).  1997.  Toxicological Profile for
       Tetrachloroethylene. U.S. Department of Health and Human Services, Public Health Service.
       278 pp. + Appendices. Available on the Internet at:
       http://www.atsdr.cdc .gov/toxprofiles/tp 18 .pdf.

Brass, H.J., M.A. Feige, T. Halloran, J.W. Mello, D. Munch, and R.F. Thomas.  1977.  The National
       Organic Monitoring Survey: A sampling and analysis for purgeable organic compounds.
       Drinking water quality enhancement through source protection. R.B. Pojasek (ed.). Ann Arbor,
       MI: Ann Arbor Science, pp. 393-416.

Brass, H.J. 1981. Rural Water Survey organics data.  Memorandum of March 17, 1981 to Hugh Hanson,
       Chief, Science and Technology Branch, Office of Drinking Water, and David Schnare, Office of
       Drinking Water , USEPA, Washington, DC.

JRB Associates.  1983.  Occurrence of Tetrachloroethylene in Drinking Water, Food and Air - DRAFT.
       Draft report submitted to EPA for review June 22,  1983.

Kuzmack, A.M.  1983. Memorandum: Characterization of the water supply industry (FY82).
       Washington, D.C.: Office of Water, USEPA. May 16, 1983.

Lopes, T.J. and S.G. Dionne. 1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater. U.S. Geological Survey Open-File Report 98-409.
       67pp.

Mello, Wayne. 1983. Personal communication between Wayne Mello, Technical Support Division,
       Office of Drinking Water, USEPA, and author of JRB Associates, 1983, March 10, 1983.

National Safety Council (NSC). 2001.  Tetrachloroethylene Chemical Backgrounder.  Itasca, IL:
       National Safety Council.  Available on the Internet at:
       http://www.crossroads.nsc.org/ChemicalTemplate.cfm?chempath=chemicals&id=137, accessed
       July 17, 2001.

National Toxicology Program (NTP).  1991.  National Toxicology Program Health and Safety
       Information Sheet - Tetrachloroethylene.  Available on the Internet at
       http://ntp-db.niehs.nih.gov/NTP_Reports/NTP_Chem_H&S/NTP_Cheml/Radianl27-18-4.txt.

Occurrence Summary and Use Support Document         421                                     March 2002

-------
Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, R.M. Clawges, and J.S. Zogorski.  1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci. and Tech. 33(23):4176-4187.

USEPA.  1994. R.E.D. Facts: Tetrachloroethylene.  EPA Report/749-F-94-020. Washington, DC:
       Office of Prevention, Pesticides, and Toxic Substances. Available on the Internet at:
       http ://www .epa.gov/opptintr/chemfact/f_perchl .txt

USEPA. 2000. TRIExplorer: Trends. Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA.  2001. National Primary Drinking Water Regulations - Consumer Factsheet on:
       Tetrachloroethylene.  Office of Ground Water and Drinking Water, USEPA. Available on the
       Internet at: http://www.epa.gov/safewater/dwh/c-voc/tetrachl.html, Last updated April 12, 2001.

USEPA.  2002. Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Westrick, J.J., J.W. Mello, and R.F. Thomas.  1983.  The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data.  EPA Technical Support Division, Office of
       Drinking Water, Cincinnati, Ohio.
Occurrence Summary and Use Support Document         422                                     March 2002

-------
4.9    I,l92-Trichloroethane
Table of Contents

4.9.1  Introduction, Use and Production  	  424
4.9.2  Environmental Release  	  424
4.9.3  Ambient Occurrence 	  425
4.9.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  426
4.9.5  Additional Drinking Water Occurrence Data  	  429
4.9.6  Conclusion	  433
4.9.7  References 	  433
Tables and Figures

Table 4.9-1: Environmental Releases (in pounds) for 1,1,2-Trichloroethane
       in the United States, 1988-1999  	  425

Table 4.9-2: Stage 1  1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section -
       Systems	  427

Table 4.9-3: Stage 1  1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section -
       Population	  427

Table 4.9-4: Stage 2 Estimated 1,1,2-Trichloroethane Occurrence Based on 16-State
       Cross-Section - Systems	  428

Table 4.9-5: Stage 2 Estimated 1,1,2-Trichloroethane Occurrence Based on 16-State
       Cross-Section - Population	  428

Table 4.9-6: Estimated National 1,1,2-Trichloroethane Occurrence - Systems and
       Population Served	  429
Occurrence Summary and Use Support Document          423                                     March 2002

-------
4.9.1  Introduction, Use and Production

1,1,2-Trichloroethane (chemical formula C2H3C13) is a colorless, sweet-smelling liquid. It does not burn
easily, but does boil at a higher temperature than water.  1,1,2-Trichloroethane is also known as ethane
trichloride, p-trichloroethane, vinyl trichloride, and 1,2,2-trichloroethane (ATSDR, 1989).

The principal use of 1,1,2-trichloroethane is as a chemical intermediate in the production of 1,1-
dichloroethylene (also known as vinylidene chloride), which is in turn used to make synthetic fibers and
plastic wraps (ATSDR, 1989; USEPA, 2001).  1,1,2-Trichloroethane has limited use as a solvent, where
its high solvency is useful. It may also be used as a solvent for fats, oils, waxes, and resins. Information
indicates that 1,1,2-trichloroethane has been sold for use in consumer products, but it is not known for
which products, nor how extensive the use. A Dow Chemical spokesman, as cited in ATSDR, stated that
the company had no knowledge of any consumer uses of 1,1,2-trichloroethane and that the company
screens potential customers to determine how they intend to use it (ATSDR, 1989).

1,1,2-Trichloroethane is produced exclusively by Dow Chemical in Freeport, TX, and  Olin Corporation
in Seward, IL. No recent information on production volumes is available.  Dow Chemical was formerly
the sole manufacturer of 1,1,2-trichloroethane, and therefore manufacturing volumes from that time
period are proprietary information and are unavailable.  The only estimates of production of 1,1,2-
trichloroethane are based upon the production of 1,1-dichloroethylene.  In  1974, the production volume
of 1,1,2-trichloroethane was estimated at 124 million pounds (USEPA, 2001), and in 1979, 412 million
pounds (ATSDR, 1989).

It should be noted that ATSDR indicates that information on 1,1,2-trichloroethane is wholly inadequate
and does not provide sufficient information to determine how widespread this chemical is, or what
potential exists for exposure in the general population. It is emphasized that information is especially
needed on the commercial uses of 1,1,2-trichloroethane and what types of consumer products, if any,
contain it. That information is requisite in determining exposure to 1,1,2-trichloroethane, and for
determining which groups in the population are occupationally, or generally, exposed (ATSDR, 1989).

4.9.2  Environmental Release

1,1,2-Trichloroethane is listed as a Toxics Release Inventory (TRI) chemical. Table 4.9-1 illustrates the
environmental releases for 1,1,2-trichloroethane from 1988 - 1999. (1,1,2-Trichloroethane data are only
available for these years.) Air emissions constitute the vast majority of the on-site releases, with a
relatively steady decrease over the years. The decrease in air emissions, as well as surface water
discharges, have contributed to decreases in 1,1,2-trichloroethane total on- and off-site releases over the
years. Except for releases in 1989-1991, no underground injection was reported for 1,1,2-
trichloroethane. Releases to land (such as spills or leaks within the boundaries of the reporting facility),
except for an upturn in 1999, decreased to at or near zero from 1992-1998.  Off-site  releases (including
metals or metal compounds transferred off-site) have also declined from large amounts recorded from
1988-1991, as releases since then have, with one exception, remained below 600 pounds.  These TRI data
for 1,1,2-trichloroethane were reported from 29 States, with six States reporting every year (USEPA,
2000). Twelve of the 16 cross-section States (used for analyses of 1,1,2-trichloroethane occurrence in
drinking water; see Section 4.9.4) reported release of 1,1,2-trichloroethane. (For a map of the 16-State
cross-section, see Figure 1.3-1.)
Occurrence Summary and Use Support Document         424                                      March 2002

-------
Table 4.9-1:  Environmental Releases (in pounds) for 1,1,2-Trichloroethane in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
198,539
279,470
296,348
339,055
280,352
310,112
315,152
562,085
527,866
588,464
744,618
1,741,442
Surface Water
Discharges
925
540
621
516
870
914
2,030
1,163
1,382
2,231
6,395
5,303
Underground
Injection
0
0
0
0
0
0
0
0
2
1,091
2,090
0
Releases
to Land
123
1
0
16
0
0
5
7
256
265
130
89
Off-Site Releases
91
1,203
141
114
113
166
592
219
8,580
25,498
75,990
19,810
Total On- &
Off-site
Releases
199,678
281,214
297,110
339,701
281,335
311,192
317,779
563,474
538,086
617,549
829,223
1,766,644
 Source: USEPA, 2000
4.9.3  Ambient Occurrence

1,1,2-Trichloroethane was not detected in rural or urban wells in the local, State, and federal data set
compiled by NAWQA. These data represent untreated ambient ground water of the conterminous United
States for the years 1985-1995 (Squillace et al., 1999).

The NURP study found 1,1,2-trichloroethane in urban runoff (Lopes and Dionne, 1998). The minimum
and maximum concentrations detected were 2 |ig/L and 3 |ig/L, respectively, with no mean value
reported. The use of the land from which the samples were taken was unspecified.

4.9.3.1 Additional Ambient Occurrence Data

A summary document entitled "Estimated National Occurrence and Exposure Assessment of 1,1,2-
Trichloroethane in Public Drinking Water Supplies" (Wade Miller, 1989), was previously prepared for
past USEPA assessments of 1,1,2-trichloroethane.  Various information was presented regarding levels
of 1,1,2-trichloroethane in water other than drinking water. The following information is taken directly
from "Estimated National Occurrence and Exposure Assessment of 1,1,2-Trichloroethane in Public
Drinking Water Supplies" (Wade Miller,  1989).

4.9.3.1.1 Groundwater Sources

In an effort to obtain information on the occurrence of 1,1,2-TCE in water other than drinking water from
ground water sources,  an extensive search of the literature was conducted. In addition, knowledgeable
sources within the Office of Drinking Water were contacted. No data are available on the occurrence of
1,1,2-TCE in non-drinking ground water sources.
Occurrence Summary and Use Support Document
425
March 2002

-------
4.9.3.1.2 Surface Water Sources

The STORET water quality data base provides information on the occurrence of contaminants at ambient
water stations in U.S. waterways. A summary of this information was obtained for 1,1,2-TCE in ambient
waters.  Ambient sites include streams, lakes, ponds, wells, reservoirs, canals, estuaries, and oceans.
Since the preponderance of data were collected from surface water sources, the data are presented in this
section. However, the number of samples collected from ground water wells, relative to the total number
of samples collected from all ambient sites combined, is unspecified (Staples et al. 1985, as cited in
Wade Miller, 1989).  The limitations with these data are the same as those described for drinking water
(see Section 4.9.5.6). Staples et al. (1985, as cited in Wade Miller, 1989) have summarized data from the
1980s only; that is, data from 1980 through 1983. This was done based on the number of data points and
the likelihood that better quality assurance practices have been employed in more recent years. In the
absence of sophisticated statistical analyses to eliminate improbable data, median values have been
reported. The median value is less sensitive to extreme values, and reflects a measure of central tendency
more accurately than the mean value in the presence of these extreme values (Staples et al. 1985, as cited
in Wade Miller, 1989).

For a total of 1,047 observations from ambient water stations, the median concentration of 1,1,2-TCE
was < 5.0 |ig/L. Of the total number of observations, two percent were reported as detectable. Detection
limits and other sampling information were not reported.

4.9.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of 1,1,2-trichloroethane occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date. These data were
evaluated relative to two concentration thresholds of interest: 0.005 mg/L; and 0.003 mg/L.

All sixteen cross-section State data sets contained occurrence data for 1,1,2-trichloroethane.  These data
represent approximately 174,000 analytical results from over 22,000 PWSs during the period from 1984
to 1998 (with most analytical results from 1992 to 1997). The number of sample results and PWSs vary
by State, although the State data sets have been reviewed and checked to ensure adequacy of coverage
and completeness.  The overall modal detection limit for 1,1,2-trichloroethane in the 16 cross-section
States is equal to 0.0005 mg/L. (For details regarding the 16-State cross-section, please refer to Section
1.3.5 of this report.)

4.9.4.1  Stage 1 Analysis Occurrence Findings

Table 4.9-2 illustrates the low occurrence  of in drinking water for the public water systems in the 16-
State cross-section relative to two thresholds: 0.005 mg/L (the current MCL), and 0.003 mg/L. Based on
the 16-State cross-section data, a total of 9 (approximately 0.0404% of) ground water and surface water
PWSs had analytical results exceeding the MCL. Twice as many (0.0808% of) systems had results
exceeding 0.003 mg/L.

Approximately  0.0385% of ground water systems (8 systems) had any analytical results greater than the
MCL.  Only 1 (0.0655% of) surface water system had results greater than the MCL.  About 0.0626% of
ground water systems (13 systems) had results above 0.003 mg/L.  A total of 5 (0.328% of) surface water
systems had at least one analytical result greater than 0.003 mg/L.
Occurrence Summary and Use Support Document         426                                     March 2002

-------
Table 4.9-2:  Stage 1 1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.003
Percent of Systems
Exceeding Threshold
0.0385%
0.0626%
Number of Systems
Exceeding Threshold
8
13

Surface Water
0.005
0.003
0.0655%
0.328%
1
5

Combined Ground &
Surface Water
0.005
0.003
0.0404%
0.0808%
9
18
Reviewing 1,1,2-trichloroethane occurrence in the 16 cross-section States by PWS population served
(Table 4.9-3) shows that approximately 6.97% of the population (almost 7.7 million people) was served
by PWSs with at least one analytical result of 1,1,2-trichloroethane greater than the MCL (0.005 mg/L).
Approximately 8.0 million (7.21% of) people were served by systems with an exceedance of 0.003 mg/L.

A much larger percentage of population served by surface water systems had MCL exceedances as
compared to the population served by ground water systems. Approximately 0.556% of people served by
ground water systems in the 16 States (about 274,900 people) were exposed to levels of 1,1,2-
trichloroethane greater than 0.005 mg/L. The percentage of population served by surface water systems
with exceedances of 0.005 mg/L was equal to 12.2% (approximately 7.4 million people). When
evaluated relative to 0.003 mg/L, the percent of population served by ground water exposed was equal to
0.627% (309,700 people). Approximately 12.6% of the population served by surface water systems
(almost 7.7 million people) was exposed to 1,1,2-trichloroethane concentrations greater than 0.003 mg/L.
Table 4.9-3:  Stage 1 1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.003
Percent of Population
Served by Systems
Exceeding Threshold
0.556%
0.627%
Total Population Served
by Systems Exceeding
Threshold
274,900
309,700

Surface Water
0.005
0.003
12.2%
12.6%
7,415,200
7,652,100

Combined Ground &
Surface Water
0.005
0.003
6.97%
7.21%
7,690,100
7,961,800
Occurrence Summary and Use Support Document
427
March 2002

-------
4.9.4.2 Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.9-4 and 4.9-5.
The statistically generated best estimate values, as well as the ranges around the best estimate value, are
presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this report. For
complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation Methodology and
Occurrence Findings for Six-Year Review of National Primary Drinking Water Regulations - DRAFT
(USEPA, 2002)).

No ground water or surface water PWSs (therefore, no population served by systems) had an estimated
mean concentration of 1,1,2-trichloroethane exceeding 0.005 mg/L or 0.003 mg/L.
Table 4.9-4:  Stage 2 Estimated 1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section
- Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.003
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
0
0
Range
0-0
0-0

Surface Water
0.005
0.003
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0-0
0-0

Combined Ground
& Surface Water
0.005
0.003
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0-0
0-0
Table 4.9-5:  Stage 2 Estimated 1,1,2-Trichloroethane Occurrence Based on 16-State Cross-Section
- Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.003
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
0.000%
0.000%
Range
0.000% - 0.000%
0.000% - 0.000%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
0
0
Range
0-0
0-0

Surface Water
0.005
0.003
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0-0
0-0
Occurrence Summary and Use Support Document
428
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
Range
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
Range

Combined Ground
& Surface Water
0.005
0.003
0.000%
0.000%
0.000% - 0.000%
0.000% - 0.000%
0
0
0-0
0-0
4.9.4.3 Estimated National Occurrence

As illustrated in Table 4.9-6, the Stage 2 analysis estimates zero systems serving zero people nationally
have estimated mean concentration values of 1,1,2-trichloroethane greater than 0.005 mg/L, or 0.003
mg/L.  (See Section 1.4 for a description of how Stage 2 16-State estimates are extrapolated to national
values.)
Table 4.9-6:  Estimated National 1,1,2-Trichloroethane Occurrence - Systems and Population
Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.003
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
0
0
Range
0-0
0-0
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
0
0
Range
0-0
0-0

Surface Water
0.005
0.003
0
0
0-0
0-0
0
0
0-0
0-0

Combined Ground
& Surface Water
0.005
0.003
0
0
0-0
0-0
0
0
0-0
0-0
4.9.5  Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of 1,1,2-trichloroethane in drinking water are
also reviewed. Previously compiled occurrence information, from an OGWDW summary document
entitled "Estimated National Occurrence and Exposure Assessment of 1,1,2-Trichloroethane in Public
Drinking Water Supplies" (Wade Miller, 1989), is presented in this section.  This variety of studies and
information are presented regarding levels of 1,1,2-trichloroethane in drinking water, with the scope of
the reviewed studies ranging from national to regional. Note that none of the studies presented in the
following section provide the quantitative analytical results or comprehensive coverage that would
enable direct comparison to the occurrence findings estimated with the cross-section occurrence data
presented in Section 4.9.4. These additional studies, however, do enable a broader assessment of the
Stage 2 occurrence estimates presented for this Six-Year Review. All the following information in
Occurrence Summary and Use Support Document
429
March 2002

-------
Section 4.9.5 is taken directly from "Estimated National Occurrence and Exposure Assessment of 1,1,2-
Trichloroethane in Public Drinking Water Supplies" (Wade Miller,  1989).

4.9.5.1 Groundwater Sources - National Studies

The Ground Water Supply Survey (GWSS) was conducted from December 1980 to December 1981 to
develop additional data on the occurrence of volatile organic chemicals (VOCs) in the Nation's ground
water supplies (Westrick et al. 1983, as cited in Wade Miller, 1989). A total of 929 finished drinking
water supplies serving 25 or more individuals were sampled. Four hundred and fifty-six systems were
chosen at random, while the remaining 473 systems were chosen non-randomly based on information
from States encouraged to identify locations believed to have a higher than normal probability of VOC
contamination (e.g., locations near landfills or industrial activity). None of the supplies sampled
contained 1,1,2-TCE in excess of the minimum quantifiable concentration of 0.5 |ig/L.

The National Screening Program for organics in drinking water (NSP), conducted by SRI International
from June 1977 to March 1981, examined both raw and finished drinking water samples from 166 water
systems in 33 States for 51 organic chemical contaminants, including 1,1,2-TCE. Analyses were made
using gas chromatography with electrolytic conductivity detection.  In this procedure,
chlorodibromomethane and 1,1,2-TCE coelute and are not distinguishable.  Reported levels should be
considered "as either compound or a combination of the two with maximum concentrations as indicated."
In summarizing the results of this survey for 1,1,2-TCE, SAIC has presented a worst case assessment by
assuming all of the coelute to be 1,1,2-TCE. The values reported are thus taken to be maximum
(potential) contamination levels in the sampled finished drinking water supplies.

The NSP data extracted from Boland (1981, as cited in Wade Miller, 1989)  indicated that for ground
water systems  (5 samples), concentrations ranged from  0.2 to 12.0 |ig/L (mean = 3.82 |ig/L; median = 0.3
l-ig/L). The minimum quantifiable concentration for 1,1,2-TCE was 0.1 |ig/L.

4.9.5.2 Groundwater Sources - Regional  Studies

The State of California's Department of Health Services (DHS) analyzed samples from 2,947 drinking
water wells used by large public water systems (having 200 connections) between 1984 and  1985.  1,1,2-
TCE was detected in four of the wells sampled with concentrations ranging  from 0.7 to 1.1 |ig/L. The
median concentration was 1.0 |ig/L (California Department of Health Services, 1986, as cited in Wade
Miller, 1989).  The California DHS also reported that one drinking water well used by  a small public
water system (< 200 connections) tested positive twice for 1,1,2-TCE at concentrations of 3.2 |ig/L and
1.1 |ig/L (California Department of Health Services, unpublished, as cited in Wade Miller, 1989). In an
earlier study, the State of California (1980, as cited in Wade Miller, 1989) reported that 1,1,2-TCE was
detected in two of four wells (1 well sampled at 4 separate California Water Utilities), both at
concentrations of 0.2 |ig/L. The other two samples were reported as less than the detection limit, which
was not provided.

The State of Minnesota's Department of Health (MDH) sampled 1,801 community water supply wells for
VOCs between 1982 and 1985. 1,1,2-TCE was detected in one well at a median and maximum
concentration of 0.3 |ig/L. The minimum quantifiable concentration was 0.2 |ig/L (Minnesota
Department of Health, 1985, as cited in Wade  Miller, 1989). One hundred and seven additional wells
were tested after the 1985 MDH survey was completed, but 1,1,2-TCE was  not detected in any of them
(detection limit not specified) (FSTRAC, 1988, as cited in Wade Miller, 1989).
Occurrence Summary and Use Support Document          430                                     March 2002

-------
Four hundred and fifty-nine actual or potential drinking water wells in Nebraska were tested for the
presence of 1,1,2-TCE. One well tested positive, with multiple samples showing concentrations ranging
from 2.3 to 32.1 |ig/L. The average concentration of 1,1,2-TCE was 13.7 |ig/L (FSTRAC, 1988, as cited
in Wade Miller, 1989).

The results of a Statewide New Jersey ground water quality survey revealed that through March 1981,
two percent of the 1,118 well water samples collected contained 1,1,2-TCE at concentrations exceeding
the minimum reportable quantity of 1.0 |ig/L. Although the maximum concentration reported was 31.1
l-ig/L, no other sample exceeded a concentration of 10.0 |ig/L (Tucker, 1981, as cited in Wade Miller,
1989).  USEPA (1981, as cited in Wade Miller,  1989) reported that of 399 ground water samples
collected in New Jersey, 203 samples had concentrations of < 1.0 |ig/L, 141 samples had concentrations
ranging from 1.0 to 10.0 |ig/L, 55 samples had concentrations ranging from 10.0 to 1,000.0 |ig/L, and one
sample had a concentration > 1,000.0 |ig/L.

In 1978, 372 drinking water wells in Nassau County, Long Island, New York were tested for the presence
of 1,1,2-TCE. Thirteen percent of the wells tested positive, with a maximum concentration of 300.0 |ig/L
detected (USEPA, 1981, as cited in Wade Miller, 1989).

4.9.5.3  Surface Water Sources - National Studies

The National Screening Program for Organics in Drinking Water (NSP) (see Section 4.9.5.1) also
contained information on the occurrence of 1,1,2-TCE in finished drinking water from surface water
sources. The NSP data extracted from Boland (1981, as cited in Wade Miller, 1989) indicated that, for
surface water systems (82 samples), concentrations ranged from 0.1 to 52.0 |ig/L (mean = 1.92 |ig/L;
median = 0.55 M-g/L).  The  minimum quantifiable concentration for 1,1,2-TCE was 0.1 |ig/L.

4.9.5.4  Surface Water Sources - Regional Studies

In an effort to obtain information from regional and State surveys on the occurrence of 1,1,2-TCE in
drinking water from surface water sources, an extensive literature search was conducted.  In addition,
knowledgeable sources within the Office of Drinking Water were contacted. No regional survey data are
available on the occurrence of 1,1,2-TCE in surface water supplies.

4.9.5.5  Unspecified Water Sources

The results of two studies which analyzed the quality of finished water at three Puerto Rico water utilities
were available as data base output.  Samples were collected during September of 1980 and August of
1981. None of the finished water, from any of the utilities sampled, contained detectable levels of 1,1,2-
TCE.  The detection limit,  sampling and analytical methodology employed, and water source were
unspecified (Puerto Rico, 1980 and 1981, as cited in Wade Miller, 1989).

The results of a study which analyzed the quality of finished water at ten New York water utilities were
available as data base output.  Samples were collected during November and December of 1980. The
results of this study revealed that none of the finished water from any of the utilities sampled contained
detectable levels of 1,1,2-TCE. The detection limit,  sampling and analytical methodology employed, and
water source were unspecified (State of New York, 1980, as cited in Wade Miller,  1989).

The results of a study which analyzed the quality of finished water at fourteen New Jersey water utilities
were available as data base output.  Samples were collected during July of 1980. The detection limit for

Occurrence Summary and Use Support Document         431                                     March 2002

-------
1,1,2-TCE was reported at 20.0 |ig/L for this study. None of the finished water collected at any of the
utilities sampled exceeded this value.  The sampling and analytical methodology employed and the water
source were unspecified (State of New Jersey, 1980, as cited in Wade Miller, 1989). The results of
another study which analyzed the quality of finished water at twelve New Jersey water utilities were also
available as data base output.  Samples were collected in 1981. The results of this study revealed that
none of the finished water, from any of the utilities sampled, contained detectable levels of 1,1,2-TCE.
The detection limit, sampling and analytical methodology employed, and water source were unspecified
(State of New Jersey, 1981, as cited in Wade Miller, 1989).

4.9.5.6 Occurrence in Groundwater and Surface Water Sources - STORET

The EPA computerized water quality data base known as STORET was devised to assist Federal and
State institutions in meeting objectives of Public Law 92-500 to maintain and enhance the physical,
chemical, and biological quality of the Nation's ambient waterways by providing for the  collection and
dissemination of basic water quality data (Staples et al., 1985,  as cited in Wade Miller, 1989).  Data are
collected by States, EPA regional offices, and other government agencies and are maintained in the
STORET system.  STORET contains approximately 80 million data entries including data on drinking
water from ground water and surface water sources.

Before presenting a summary of the drinking water data in STORET, it is important to note that there are
significant limitations in using this data base to estimate representative concentrations of a contaminant
such as 1,1,2-TCE. Data entered into STORET are gathered from an array of studies  conducted for
various purposes.  Analyses are conducted in a number of different laboratories employing different
methodologies with a range of detection limits.  In many cases, detection limits are not reported, making
the reliability of the data highly questionable. In cases in which the detection limits have been reported,
STORET assigns the detection limit value for those observations reported as not detected. This can lead
to errors in interpretation and overestimation of concentrations in a particular medium. Additionally, a
few high values can inflate mean values and result in large standard deviations relative to the means
(Staples et al., 1985, as cited in Wade Miller, 1989). Very high values may not be correct, as they may
reflect sample contamination or analytical error and can significantly distort assessment of average
concentrations.  Staples et al. (1985, as cited in Wade Miller, 1989) also notes that use of data collected
prior to the 1980s is not recommended since some data were obtained using less sensitive laboratory
techniques than are currently available and since quality assurance procedures were not yet mandated for
the data entered into the system.

With these  limitations in mind, a summary of the most recently obtained data for 1,1,2-TCE is presented
for drinking water from ground water sources (USEPA, 1988,  as cited in Wade Miller, 1989). According
to STORET, there were 16 positive observations for 1,1,2-TCE in ground water from December 1984 to
March 1987, with an overall mean value of 1.97 |ig/L and a range of 0.1 to 5.8 |ig/L.  The standard
deviation for these observations was 2.06 |ig/L.  There were 237 samples reported as undetected and
assigned detection limit values, resulting  in a mean value of 0.76 |ig/L, a range of 0.1  to 1.0 |ig/L and a
standard deviation of 0.26 |ig/L. Including the undetected samples, the 10,069 observations with actual
values less than the reported values, and the 3 detections that were verified but not quantified, there were
a total of 10,325 observations for 1,1,2-TCE in ground water from February 1978 to November 1987,
with an overall mean value of 0.42 |ig/L and a range of 0.01 to 10.0 |ig/L.  The standard deviation for all
observations was 0.35  |ig/L.  Detection limits and other sampling information were not reported.

The STORET data base similarly contains data on the occurrence of 1,1,2-TCE in drinking water from
surface water sources (USEPA, 1988, as cited in Wade Miller, 1989).  According to STORET, there were

Occurrence Summary and Use Support Document          432                                     March 2002

-------
89 observations for 1,1,2-TCE in surface water from February 1978 to July 1987, with an overall mean
value of 5.51 |ig/L and a range of 0.5 to 10.0 |ig/L.  The standard deviation for all observations was 4.44
l-ig/L. Detection limits and other sampling information were not reported. It is important to note that of
the 89 total observations for 1,1,2-TCE, 40 observations were reported as undetected and assigned
detection limit values resulting in a mean of 2.07 |ig/L (range = 1.0 - 10.0 |ig/L) and a standard deviation
of 2.58 |ig/L. The remaining 49 observations had actual values less than the reported values.

4.9.5.7  National Estimate of the Occurrence of 1,1,2-Trichloroethane and Population Exposure for
Public Water Supplies

Although Federal survey data on the occurrence of 1,1,2-trichloroethane in public drinking water
supplies are available, there are few detectable concentration values. Information on the occurrence of
1,1,2-TCE in public drinking water supplies is available in STORET; however, as discussed above, there
are significant limitations involved with utilizing these data.  The available State data are usually poorly
described with respect to the source and size categories of the supplies examined and the sampling and
analysis methods used for determining contaminant levels. Consequently, national estimates of
occurrence and population exposure  were not made for 1,1,2-TCE.

4.9.6 Conclusion

The primary use of 1,1,2-trichloroethane is as a chemical intermediate in the production of 1,1-
dichloroethylene. It is also used as a solvent, and it is unknown whether or not it may be present in any
consumer products. Virtually no production data about 1,1,2-trichloroethane is available due to
proprietary concerns of Dow Chemical. 1,1,2-Trichloroethane is also a  TRI chemical.  Industrial releases
of 1,1,2-trichloroethane have occurred in 29 States since 1988.  1,1,2-Trichloroethane was an analyte for
the NURP ambient occurrence studies. In the NURP study, 1,1,2-trichloroethane was detected at a
maximum concentration of 3 |ig/L and a minimum concentration of 2 |ig/L, with no median value
reported. In the Stage 2 analysis of 16-State occurrence of 1,1,2-trichloroethane, zero percent of
combined ground water and surface water systems serving zero percent  of the population exceeded the
MCL of 0.005 mg/L. Nationally, zero ground water and surface water systems combined (serving
approximately zero people) are estimated to have levels greater than the MCL.

The 16-State cross-section was designed to be nationally representative  based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors.  All production information for 1,1,2-trichloroethane is unavailable for comparison due
to proprietary restrictions. Nationally, 29 States have reported TRI releases of 1,1,2-trichloroethane,
including 12 of the 16 cross-section States. The cross-section should adequately represent the occurrence
of 1,1,2-trichloroethane on a national scale based upon the use, production, and release  patterns of the
16-State cross-section in relation to the patterns observed for all 50 States.

4.9.7 References

Agency for Toxic Substances and Disease Registry (ATSDR).  1989. Toxicological Profile for 1,1,2-
        Trichloroethane.  U.S. Department of Health and Human Services, Public Health Service.  108
        pp. + Appendices. Available on the Internet at http://atsdrl.atsdr.cdc.gov/toxprofiles/tpl48.pdf

Boland, P.A. 1981. National Screening Program for Organics in Drinking Water. Prepared by SRI
        International, Menlo Park, California, for Office of Drinking Water, USEPA, Washington, DC.
        EPA Contract No. 68-01-4666.

Occurrence Summary and Use Support Document          433                                      March 2002

-------
California Department of Health Services. 1986. Final Report on a Monitoring Program for Organic
       Chemical Contamination of Large Public Water Systems in California.  Summary Version.
       Sacramento, CA.

California Department of Health Services. Unpublished.  Final Report on a Monitoring Program for
       Organic Chemical Contamination of Small Public Water Systems in California. Summary
       Version.  Sacramento, CA.

FSTRAC. 1988.  Federal/State Toxicological and Regulatory Alliance Committee. FSTRAC Survey
       Results -  April 1988.  Survey results obtained from Ralph Langemeier, Chief of Drinking Water
       Branch, USEPA, Region VII.

Lopes, T.J. and S.G. Dionne.  1998.  A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban  Stormwater. U.S. Geological Survey Open-File Report 98-409.
       67pp.

Minnesota Department of Health.  1985.  Volatile Organic Survey of Community Water Supplies. Report
       to the Legislative Commission on Minnesota Resources.  Minneapolis, MN.

Puerto Rico. 1980. Results of Puerto Rico Water Supply Study for 1,1,2-Trichloroethane. Data obtained
       from Arthur Perler, Branch Chief, Science and Technology Branch, Office of Drinking Water,
       USEPA, Washington, D.C.

Puerto Rico. 1981. Results of Puerto Rico Water Supply Study for 1,1,2-Trichloroethane. Data obtained
       from Arthur Perler, Branch Chief, Science and Technology Branch, Office of Drinking Water,
       USEPA, Washington, D.C.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, RM. Clawges, and J.S. Zogorski. 1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci.  and Tech. 33(23):4176-4187.

Staples, C.A., A.F. Werner, and T.J.  Hoogheem.  1985. Assessment of Priority Pollutant Concentrations
       in the United States Using STORET Data Base. Environmental Toxicology and Chemistry,
       4:131-142.

State of California.  1980.  Memorandum: Analysis of California Department of Health Services Well
       Water Samples for 1,1,2-Trichloroethane. Data obtained from Arthur Perler, Branch Chief,
       Science and Technology Branch, Office of Drinking Water, USEPA, Washington, D.C.  1/30/80.

State of New Jersey. 1980. Results of New Jersey Water Supply Sampling for 1,1,2-Trichloroethane.
       Report date 9/24/80. Data obtained from Arthur Perler, Branch Chief, Science and Technology
       Branch, Office of Drinking Water, USEPA, Washington, D.C.

State of New Jersey. 1981. Results of New Jersey Water Supply Sampling for 1,1,2-Trichloroethane.
       Report date 10/8/81. Data obtained from Arthur Perler, Branch Chief, Science and Technology
       Branch, Office of Drinking Water, USEPA, Washington, D.C.
Occurrence Summary and Use Support Document         434                                    March 2002

-------
State of New York. 1980. Results of New York Water Supply Sampling for 1,1,2-Trichloroethane.
       Report date 2/26/81. Data obtained from Arthur Perler, Branch Chief, Science and Technology
       Branch, Office of Drinking Water, USEPA, Washington, D.C.

Tucker, R.K.  1981. Groundwater Quality in New Jersey: An Investigation of Toxic Contaminants. State
       of New Jersey Department of Environmental Protection. Office of Cancer and Toxic Substances
       Research, Trenton, NJ.

USEPA.  1981.  USEPA. An Exposure and Risk Assessment for Trichloroethanes. Final Draft Report.
       Office of Water Regulations and Standards, Office of Water, Washington, D.C. (Revised July
       1982).

USEPA.  1988.  Computer Printout of STORET Water Quality Data Base. Retrieval conducted March
       23, 1988 by Science Applications International Corporation. Data available through Office of
       Water Regulations and Standards, Washington, D.C.

USEPA. 2000. TRIExplorer: Trends. Available on the Internet at:
http://www.epa.gov/triexplorer/trends.htm, last updated May 5,  2000.

USEPA.  2001.  National Primary Drinking  Water Regulations  - Consumer Factsheet on: 1,1,2-
       Trichloroethane. Office of Ground Water and Drinking Water, USEPA. Available on the
       Internet at http://www.epa.gov/safewater/dwh/c-voc/! 12-tric.html (Last updated 04/12/2001)

USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT.  EPA Report/815-D-02-005, Office of
       Water, 55 pp.

Wade Miller Associates, Inc. 1989.  Estimated National Occurrence and Exposure Assessment of 1,1,2-
       Trichloroethane in Public Drinking Water Supplies. Prepared for and submitted to EPA on
       September 29, 1989.

Westrick, J.J., J.W. Mello, and R.F. Thomas. 1983.  The Ground Water Supply Survey summary of
       volatile organic contaminant occurrence data. EPA Technical Support Division, Office of
       Drinking Water, Cincinnati,  Ohio.
Occurrence Summary and Use Support Document         435                                    March 2002

-------
4.10   Trichloroethylene
Table of Contents

4.10.1  Introduction, Use and Production  	  437
4.10.2  Environmental Release  	  438
4.10.3  Ambient Occurrence  	  439
4.10.4  Drinking Water Occurrence Based on the 16-State Cross-Section	  440
4.10.5  Additional Drinking Water Occurrence Data  	  444
4.10.6  Conclusion	  447
4.10.7  References 	  447
Tables and Figures

Table 4.10-1: Facilities that Manufacture or Process Trichloroethylene  	  437

Table 4.10-2: Environmental Releases (in pounds) for Trichloroethylene
       in the United States, 1988-1999  	  439

Table 4.10-3: Stage 1 Trichloroethylene Occurrence Based on 16-State Cross-Section -
       Systems	  440

Table 4.10-4: Stage 1 Trichloroethylene Occurrence Based on 16-State Cross-Section -
       Population	  441

Table 4.10-5: Stage 2 Estimated Trichloroethylene Occurrence Based on 16-State
       Cross-Section - Systems	  442

Table 4.10-6: Stage 2 Estimated Trichloroethylene Occurrence Based on 16-State
       Cross-Section - Population	  443

Table 4.10-7: Estimated National Trichloroethylene Occurrence - Systems and
       Population Served	  444
Occurrence Summary and Use Support Document          436                                     March 2002

-------
4.10.1  Introduction, Use and Production

Trichloroethylene (chemical formula C2HC13) is a nonflammable, colorless liquid at room temperature
that has a somewhat sweet odor, and a sweet, burning taste. It is also known as acetylene trichloride, 1-
chloro-2,2-dichloroethylene, l,l-dichloro-2-chloroethylene, ethylene trichloride, trichloride, TCE, 1,1,2-
trichloroethylene, and trichloroethene. Trichloroethylene is also known as Triclene, Vitran, and by other
trade names in industry (ATSDR, 1997).

The end use pattern of trichloroethylene in the U.S. in 1987 was estimated as follows: vapor degreasing
of fabricated metal parts, 80%; chemical intermediates, 5%; miscellaneous uses, 5%; and exports, 10%
(Spectrum Laboratories, 2001). Trichloroethylene is an excellent extraction solvent for greases, oils,
fats, waxes, and tars and is used by the textile processing industry to scour cotton, wool, and other
fabrics. As a general solvent, or as a component of solvent blends, trichloroethylene is used with
adhesives, lubricants, paints, varnishes, paint strippers, pesticides, and cold metal cleaners (ATSDR,
1997).

Trichloroethylene is used as a chemical intermediate for the production of polyvinyl chloride,
Pharmaceuticals, polychlorinated aliphatics, flame retardant chemicals, and insecticides.
Trichloroethylene is used as a refrigerant for low-temperature heat transfer, and in the aerospace industry
for flushing liquid oxygen. Some consumer products that contain trichloroethylene are typewriter
correction fluids, paint removers and strippers, adhesives, spot removers, and rug cleaning fluids
(ATSDR, 1997).

Use of trichloroethylene has declined due to its toxicity (NTP, 2001). It had been used as a general and
obstetrical anesthetic; a grain fumigant; a skin, wound, and surgical disinfectant; pet food additive, and
extractant of spice oleoresins in food and of caffeine for the purpose of making decaffeinated coffee.
These uses were banned by the U.S. Food and Drug Administration in 1977 (ATSDR, 1997).

The only U.S. manufacturers of trichloroethylene are Dow Chemical in Freeport, TX, and PPG Industries
in Lake Charles, LA, with a combined annual production capacity of 320 million pounds. Because there
are currently only two manufacturers, recent production data is not available.  In previous years,
production volumes of trichloroethylene  have been reported as follows:  1978, 299 million pounds; 1979,
319 million pounds; 1980, 266 million pounds; 1981, 258 million pounds; and 1982, 200 million pounds.
U.S. demand for trichloroethylene was estimated at 235 million pounds in 1983, 180 million pounds in
1985, and 170 million pounds in 1986 (ATSDR, 1997).

Table 4.10-1 shows the number of facilities in each State that manufacture and process trichloroethylene,
the intended uses of the product, and the range of maximum amounts derived from the Toxics Release
Inventory (TRI) of EPA (ATSDR, 1997).
Table 4.10-1:  Facilities that Manufacture or Process Trichloroethylene
State"
AL
AR
AZ
CA
CO
CT
Number of facilities
12
11
4
5
1
15
Range of maximum amounts on
site in thousands of pounds'1
1-1,000
1-1,000
1-10
1-10,000
10-100
1-100
Activities and uses0
2,3,7,8,12,13
8,9,11,12,13
12,13
1,5,8,10,11,13
12
11,12,13
Occurrence Summary and Use Support Document         437                                     March 2002

-------
State"
DE
FL
GA
IA
IL
IN
KS
KY
LA
MA
MD
ME
MI
MN
MO
MS
NC
NE
NH
NJ
NY
OH
OK
OR
PA
PR
RI
SC
SD
TN
TX
VA
VT
WA
WI
Number of facilities
1
14
12
5
99
50
12
17
11
36
4
2
41
28
28
6
19
10
4
11
55
55
5
7
54
1
5
12
2
14
34
12
3
9
42
Range of maximum amounts on
site in thousands of pounds'"
100-1,000
0-100
1-1,000
1-10
0-100,000
0-10,000
0-1,000
1-50,000
1-10,000
0-1,000
1-100
10-100
1-1,000
0-100
0-1,000
1-1,000
0-100
1-100
1-100
1-1,000
0-1,000
0-1,000
1-100
0-1,000
1-1,000
10-100
0-100
1-1,000
10-100
1-1,000
1-50,000
1-1,000
1-100
0-100
0-1,000
Activities and uses0
13
11,12,13
8,12,13
12,13
1,4,8,9,10,11,12,13
7,8,10,11,12,13
8,10,11,12,13
1,3,7,10,12,13
1,3,4,5,6,7,8,12,13
8,10,11,12,13
2,3,12,13
13
2,3,8,10,11,12,13
11,12,13
2,3,8,10,12,13
12,13
10,11,12,13
11,12,13
11,13
8,12,13
3,7,10,11,12,13
8,10,11,12,13
2,3,13
11,12,13
3,8,10,11,12,13
13
8,12,13
7,12,13
13
2,3,7,11,12,13
1,3,4,5,6,7,8,10,12,13
12,13
11,13
12,13
8,10,11,12,13
aPost office State abbreviations used
bData in TRI are maximum amounts on site at each facility
cActivities/Uses include:
1. Produce                  8. As a formulation component
2. Import                   9. As a product component
3. For on-site use/processing       10. For repackaging only
4. For sale/distribution           11. As a chemical processing aid
5. As a byproduct              12. As a manufacturing aid
6. As an impurity              13. Ancillary or other uses
7. As a reactant

Source: ATSDR, 1997 compilation O/TR193 1995 data
4.10.2  Environmental Release

Trichloroethylene is listed as a Toxics Release Inventory (TRI) chemical. Table 4.10-2 illustrates the
environmental releases for trichloroethylene from 1988 - 1999.  (Trichloroethylene  data are only
available for these years.)  Air emissions constitute the vast majority of the on-site releases, with a steady
decrease over the years.  The decrease in air emissions, as well as surface water discharges and off-site
releases (including metals or metal compounds transferred off-site), have contributed to decreases in
trichloroethylene total on- and off-site releases in the years covered by the TRI data. Underground
injection and releases to land (such as spills or leaks within the boundaries of the reporting facility) have
fluctuated over the years, but usually do not make a very significant contribution to  overall releases.
These TRI data for trichloroethylene were reported from 46 States and Puerto Rico. (No data were
reported from Wyoming, North Dakota, Hawaii, or Alaska). Of the 46 States, 37 reported every year

Occurrence Summary and Use Support Document          438                                         March 2002

-------
(USEPA, 2000). All of 16 cross-section States (used for analyses of trichloroethylene occurrence in
drinking water; see Section 4.10.4) reported TRI data for trichloroethylene. (For a map of the 16-State
cross-section, see Figure 1.3-1.)
Table 4.10-2:  Environmental Releases (in pounds) for Trichloroethylene in the United States,
1988-1999
Year
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
On-Site Releases
Air Emissions
10,510,064
13,137,700
18,143,968
21,880,624
26,265,512
30,945,977
31,007,030
30,838,983
36,356,277
40,028,932
49,798,528
55,943,736
Surface Water
Discharges
1,043
867
563
541
1,477
1,671
5,220
8,606
12,784
14,285
15,849
13,801
Underground
Injection
0
588
986
1,291
550
288
460
466
800
805
390
390
Releases
to Land
138,522
800
3,975
9,740
3,577
4,070
8,212
20,726
62,991
12,554
8,686
21,186
Off-Site Releases
115,587
98,774
182,423
89,527
74,145
96,312
233,561
248,714
115,973
753,864
1,250,933
1,466,469
Total On- &
Off-site
Releases
10,765,216
13,238,729
18,331,915
21,981,723
26,345,261
31,048,318
31,254,483
31,117,495
36,548,825
40,810,440
51,074,386
57,445,582
 Source: USEPA, 2000
4.10.3 Ambient Occurrence

Trichloroethylene was detected in 47 out of 405 wells (11.6%) in urban areas of the local, State, and
federal data set compiled by NAWQA. The minimum and maximum concentrations detected were 0.2
l-ig/L and 80 |ig/L, respectively. The median value of detection concentrations was 0.9 |ig/L.
Trichloroethylene was also detected in 41 of the 2,542 wells (1.6%) with analysis in rural areas. The
minimum and maximum concentrations detected were 0.2 |ig/L and 63 |ig/L, respectively.  The median
value of detection concentrations was 0.6 |ig/L. These data (urban and rural) represent untreated ambient
ground water of the conterminous United States for the years 1985-1995 (Squillace et al., 1999).

Trichloroethylene was also an analyte in the NURP study. The NURP study found trichloroethylene in
urban runoff (Lopes and Dionne, 1998).  The minimum and maximum concentrations detected were 0.3
l-ig/L and 10 |ig/L, respectively, with no mean value reported.  The use of the land from which the
samples were taken was unspecified.

4.10.3.1  Additional Ambient Occurrence Data

A summary document entitled "Sources, Emission and Exposure for Trichloroethylene (TCE) and
Related Chemicals" (USEPA, 2001), was previously prepared for past USEPA assessments of
trichloroethylene. Various studies and information are presented regarding levels of TCE in water, both
ambient and drinking. These studies are summarized in Section 4.10.5.
Occurrence Summary and Use Support Document
439
March 2002

-------
4.10.4  Drinking Water Occurrence Based on the 16-State Cross-Section

The analysis of trichloroethylene occurrence presented in the following section is based on State
compliance monitoring data from the 16 cross-section States. The 16-State cross-section is the largest
and most comprehensive compliance monitoring data set compiled by EPA to date.  These data were
evaluated relative to several concentration thresholds of interest: 0.005 mg/L; 0.0025 mg/L; and 0.0005
mg/L.

All sixteen cross-section State data sets contained occurrence data for trichloroethylene. These data
represent more than 201,000 analytical results from approximately 23,000 PWSs during the period from
1984 to 1998 (with most analytical results from 1992 to 1997).  The number of sample results and PWSs
vary by State, although the State data sets have been reviewed and checked to ensure adequacy of
coverage and completeness. The overall modal detection limit for trichloroethylene in the 16 cross-
section States is equal to 0.0005 mg/L.  (For details regarding the 16-State cross-section, please refer to
Section 1.3.5 of this report.)

4.10.4.1  Stage 1 Analysis Occurrence  Findings

Table 4.10-3 illustrates the occurrence of trichloroethylene in drinking water for the public water systems
in the 16-State cross-section relative to three thresholds: 0.005 mg/L (the current MCL), 0.0025 mg/L,
and 0.0005 mg/L (the modal MRL).  A total of 149 (approximately 0.647% of) ground water and surface
water PWSs had analytical results exceeding the MCL; 1.02% (234 systems) of systems in the 16 States
had results exceeding 0.0025 mg/L; and  2.24% of systems (516 systems) had results exceeding 0.0005
mg/L.

Approximately 0.596% (128 systems) of ground water systems had any analytical results greater than the
MCL.  About 0.951% (204 systems) of ground water systems had results above 0.0025 mg/L. The
percentage of ground water systems with at least one result greater than 0.0005 mg/L was equal to 2.12%
(455 systems).

Approximately 21 (1.33% of) surface water systems had results greater than the MCL.  A total of 30
(1.91% of) surface water systems had at least one analytical result greater than 0.0025 mg/L. Sixty-one
(3.88% of) surface water systems had results exceeding 0.0005 mg/L.
Table 4.10-3:  Stage 1 Trichloroethylene Occurrence Based on 16-State Cross-Section - Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems
Exceeding Threshold
0.596%
0.951%
2.12%
Number of Systems
Exceeding Threshold
128
204
455

Surface Water
0.005
0.0025
0.0005
1.33%
1.91%
3.88%
21
30
61
Occurrence Summary and Use Support Document
440
March 2002

-------
Source Water Type
Threshold
(mg/L)
Percent of Systems
Exceeding Threshold
Number of Systems
Exceeding Threshold

Combined Ground &
Surface Water
0.005
0.0025
0.0005
0.647%
1.02%
2.24%
149
234
516
Reviewing trichloroethylene occurrence by PWS population served (Table 4.10-4) shows that
approximately 13.3% of the population served in the 16 States (almost 15 million people) was served by
PWSs with at least one analytical result of trichloroethylene greater than the MCL (0.005 mg/L). Over
17 million (15.6% of) people in the 16 States were served by systems with an exceedance of 0.0025
mg/L.  A total of about 25.6 million (23.1% of) people were served by systems with at least one
analytical result greater than 0.0005 mg/L.

The percentage of 16-State population served by ground water systems with analytical results greater
than the MCL was equal to 8.01% (almost 4 million people).  When evaluated relative to 0.0025 mg/L or
0.0005 mg/L, the percent of population exposed was equal to 11.2% (over 5.5 million people) and 18.6%
(approximately 9.2 million people), respectively.

The percentage of 16-State population served by surface water systems with exceedances of 0.005 mg/L
was equal to 17.5% (approximately 10.7 million people).  Approximately 19.2% of the population served
by surface water systems (about 11.7 million people) was exposed to trichloroethylene concentrations
greater than 0.0025 mg/L. When evaluated relative to  0.0005 mg/L, the percent of population exposed
was equal to 26.8% (over 16 million people).
Table 4.10-4:  Stage 1 Trichloroethylene Occurrence Based on 16-State Cross-Section - Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population
Served by Systems
Exceeding Threshold
8.01%
11.2%
18.6%
Total Population Served by
Systems Exceeding
Threshold
3,971,100
5,539,500
9,219,500

Surface Water
0.005
0.0025
0.0005
17.5%
19.2%
26.8%
10,710,100
11,713,300
16,357,500

Combined Ground &
Surface Water
0.005
0.0025
0.0005
13.3%
15.6%
23.1%
14,681,200
17,252,900
25,577,000
Occurrence Summary and Use Support Document
441
March 2002

-------
4.10.4.2  Stage 2 Analysis Occurrence Findings

The Stage 2 occurrence findings, based on the cross-section data, are presented in Tables 4.10-5 and
4.10-6. The statistically generated best estimate values, as well as the ranges around the best estimate
value, are presented.  (For a review of the Stage 2 analytical approach, please refer to Section 1.4 of this
report. For complete details regarding the Stage 2 analyses, please refer to Occurrence Estimation
Methodology and Occurrence Findings for Six-Year Review of National Primary Drinking Water
Regulations - DRAFT (USEPA, 2002)).

A combined total of 54 (0.236% of) ground water and surface water PWSs in the 16 States had an
estimated mean concentration of trichloroethylene exceeding 0.005 mg/L, while 103 (0.445% of) systems
in the 16 States had an estimated mean concentration exceeding 0.0025 mg/L and 299 (1.30% of)
systems had an estimated mean concentration exceeding 0.0005 mg/L.

About 47 (0.221%  of) ground water PWSs in the 16 States were estimated to have a mean concentration
greater than 0.005 mg/L.  Additionally, 90 (0.420% of) ground water PWSs were estimated to exceed a
mean concentration of 0.0025 mg/L and 272  (1.27% of) systems were estimated to exceed a mean
concentration of 0.0005 mg/L. For surface water PWSs in the  16 States, 7 (0.450%), 13 (0.795%), and
27 (1.71%) had estimated mean concentrations exceeding 0.005 mg/L, 0.0025 mg/L, and 0.0005 mg/L,
respectively.
Table 4.10-5:  Stage 2 Estimated Trichloroethylene Occurrence Based on 16-State Cross-Section
Systems
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Systems Estimated
to Exceed Threshold
Best Estimate
0.221%
0.420%
1.27%
Range
0.1 86% -0.256%
0.368% - 0.466%
1.17% -1.39%
Number of Systems in the 16 States
Estimated to Exceed Threshold
Best Estimate
47
90
272
Range
40-55
79 - 100
252 - 299

Surface Water
0.005
0.0025
0.0005
0.450%
0.795%
1.71%
0.381% -0.572%
0.699% - 0.953%
1.46% -2.03%
7
13
27
6-9
11-15
23-32

Combined Ground
& Surface Water
0.005
0.0025
0.0005
0.236%
0.445%
1.30%
0.204% - 0.274%
0.399% -0.491%
1.21% -1.42%
54
103
299
47-63
92-113
279 - 326
Reviewing trichloroethylene occurrence by PWS population served (Table 4.10-6) shows that
approximately 8.19% of population served by all PWSs in the 16 States (an estimate of over 9 million
people) were potentially exposed to trichloroethylene levels above 0.005 mg/L. For all PWSs in the 16
States, an estimated 9.17% of population served (an estimate of over 10 million people served in the 16
Occurrence Summary and Use Support Document
442
March 2002

-------
States) was exposed to levels above 0.0025 mg/L, and 12.5% (an estimated 13.8 million people in the 16-
State cross-section) was exposed to levels above 0.0005 mg/L.

When the population exposed for ground water systems was evaluated relative to thresholds of 0.005
mg/L, 0.0025 mg/L, and 0.0005 mg/L, the percentages of population exposed were equal to 1.94% (an
estimated 962,700 people), 3.30% (an estimate of over 1.6 million people), and 7.90% (an estimated 3.9
million people), respectively.

 The percentage of population served by surface water systems in the 16 States with levels above 0.005
mg/L was 13.3% (an estimated 8 million people), while the population served with levels above 0.0025
mg/L and 0.0005 mg/L was 13.9% (an estimated 8.5 million people) and 16.2% (an estimate of almost
9.9 million people), respectively.
Table 4.10-6:  Stage 2 Estimated Trichloroethylene Occurrence Based on 16-State Cross-Section -
Population
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Percent of Population Served by Systems
Estimated to Exceed Threshold
Best Estimate
1.94%
3.30%
7.90%
Range
1.56% -2.33%
2. 91% -3. 82%
7.26% - 8.69%
Total Population Served by Systems in the
16 States Estimated to Exceed Threshold
Best Estimate
962,700
1,634,100
3,914,900
Range
772,500 - 1,153,500
1,442,900-1,890,800
3,597,800 - 4,305,300

Surface Water
0.005
0.0025
0.0005
13.3%
13.9%
16.2%
13.0% -13. 7%
13. 8% -14. 5%
15.0% -18.6%
8,097,000
8,512,300
9,886,200
7,950,500 - 8,377,900
8,408,400 - 8,823,700
9,165,600-11,351,700

Combined Ground
& Surface Water
0.005
0.0025
0.0005
8.19%
9.17%
12.5%
7.98% -8.41%
8. 95% -9. 53%
11.7% -13.8%
9,062,500
10,147,600
13,804,500
8,828,000 - 9,302,500
9,894,300 - 10,535,900
12,952,800 - 15,242,500
4.10.4.3  Estimated National Occurrence

Based on the Stage 2 estimated percent of systems (and population served by systems) exceeding each
threshold, an estimated 154 PWSs nationally, serving over 17 million people, could be exposed to
trichloroethylene concentrations above 0.005 mg/L.  About 290 systems, serving almost 20 million
people nationally, had estimated mean concentrations greater than 0.0025 mg/L. Approximately 844
systems,  serving about 26.6 million people nationally, were estimated to have mean trichloroethylene
concentrations greater than 0.0005 mg/L. (See Section 1.4 for a description of how Stage 2 16-State
estimates are extrapolated to national values.)

For ground water systems, an estimated 131 PWSs, serving about 1.7 million people nationally, had
mean concentrations greater than 0.005 mg/L. Approximately 250 systems, serving over 2.8 million
Occurrence Summary and Use Support Document
443
March 2002

-------
people nationally, had estimated mean concentration values that exceeded 0.0025 mg/L.  About 754
ground water systems, serving almost 6.8 million people, had estimated mean concentrations greater than
0.0005 mg/L.

Although fewer surface water systems than ground water systems nationally had estimated mean
concentration values greater than the thresholds, a much larger population was served by surface water
systems with threshold exceedances because surface water systems tend to serve much larger
populations. Approximately 25 surface water systems, serving almost 17 million people, were estimated
to have mean concentrations of trichloroethylene above 0.005 mg/L. About 44 surface water systems,
serving almost 18 million people, had estimated mean concentrations greater than 0.0025 mg/L. An
estimated 95 surface water systems, serving approximately 20.6 million people, had mean concentrations
greater than 0.0005 mg/L.
Table 4.10-7:  Estimated National Trichloroethylene Occurrence - Systems and Population Served
Source Water Type
Ground Water
Threshold
(mg/L)
0.005
0.0025
0.0005
Total Number of Systems Nationally
Estimated to Exceed Threshold
Best Estimate
131
250
754
Range
111-152
219-277
698 - 828
Total Population Served by Systems
Nationally Estimated to Exceed Threshold
Best Estimate
1,664,800
2,825,800
6,769,700
Range
1,335,800-1,994,700
2,495,100 - 3,269,600
6,221,300 - 7,444,900

Surface Water
0.005
0.0025
0.0005
25
44
95
21 -32
39-53
82-114
16,883,500
17,749,300
20,614,200
16,577,900 - 17,469,200
17,532,900 - 18,398,700
19,111,700-23,670,000

Combined Ground
& Surface Water
0.005
0.0025
0.0005
154
290
844
133-178
260-319
788 - 920
17,451,800
19,541,400
26,583,400
17,000,200 - 17,914,000
19,053,600-20,289,000
24,943,300 - 29,352,500
4.10.5 Additional Drinking Water Occurrence Data

Several additional data sources regarding the occurrence of trichloroethylene in drinking water are also
reviewed. Previously compiled occurrence information, from an OGWDW summary document entitled
"Sources, Emission and Exposure for Trichloroethylene (TCE) and Related Chemicals" (USEPA, 2001),
is presented in this section. This variety of studies and information are presented regarding levels of
trichloroethylene in drinking water, with the scope of the reviewed studies ranging from national to
regional. Note that none of the studies presented in the following section provide the quantitative
analytical results or comprehensive coverage that would enable direct comparison to the occurrence
findings estimated with the cross-section occurrence data presented in Section 4.10.4. These additional
studies, however, do enable a broader assessment of the Stage 2 occurrence estimates presented for this
Six-Year Review. All the following information in Section 4.10.5 is taken directly from "Sources,
Emission and Exposure for Trichloroethylene (TCE)  and Related Chemicals" (USEPA, 2001).
Occurrence Summary and Use Support Document
444
March 2002

-------
4.10.5.1 Levels of Trichloroethylene (TCE) in Various Water Sources

According to IARC (1995, as cited in USEPA, 2001), the reported median concentrations of TCE in
1983-84 were 0.5 |ig/L in industrial effluents and 0.1 |ig/L in ambient water. ATSDR (1997, as cited in
USEPA, 2001) has reported that TCE is the most frequently reported organic contaminant in
groundwater and the one present in the highest concentration in a summary of ground water analyses
reported in 1982. It has been estimated that somewhere between 9 and 34 percent of the drinking water
supply sources tested in the U.S. may have some trichloroethylene contamination. This estimate is based
on available Federal  and State surveys (ATSDR, 1997, as cited in USEPA, 2001). Results from an
analysis of the EPA STORET Data Base (1980-1982) showed that TCE was detected in 28 percent of
9,295 surface water reporting stations nationwide (ATSDR, 1997, as cited in USEPA, 2001).

More recently, the USEPA Office of Ground Water and Drinking Water reposed that most water supplies
are in compliance with the maximum contaminant level (maximum contaminant level (MCL), 5 M-g/L),
and that only 407 samples out of many thousands taken from community and other water supplies
throughout the country over the past 11 years (1987-1997) have exceeded the MCL limit for TCE
(USEPA, 1998, as cited in USEPA, 2001).

TCE concentrations in ground water have been measured extensively in California.  The data were
derived from a survey of large water utilities (i.e. utilities with more than 200 service connections). The
survey was conducted by the California Department of Health Services (DHS, 1986, as cited in USEPA,
2001). From January 1984 through December 1985, wells in 819 water systems were sampled for organic
chemical contamination. The water systems use a total of 5,550 wells, 2,947 of which were sampled.
TCE was found in 187 wells, at concentrations up to 440 |ig/L, with a median concentration of 3.0 |ig/L.
Generally, the most contaminated wells and the wells with the highest concentrations were found in the
heavily urbanized areas of the State. Los Angeles County registered the greatest number of contaminated
wells (149).

4.10.5.2 Human Exposure and Population Estimates

4.10.5.2.1  General U.S. Population

Because of the pervasiveness of TCE in the environment, most people are exposed to it through ingestion
of drinking water, inhalation of ambient air, or ingestion of food (ATSDR, 1997, as cited in USEPA,
2001). Contamination of drinking water with TCE varies according to location and with the drinking
water source (whether source is surface water or groundwater). TCE readily volatilizes from water and
inhalation  of indoor air may be a major route of exposure in homes with contaminated water supply
(ATSDR,  1997, as cited in USEPA, 2001).

The California survey of large water utilities in 1984 found a median concentration  of 3.0 |ig/L (DHS,
1986). Using this value and a 2 L/day water consumption rate yields an estimate of 6 |ig/day. This  is
consistent  with ATSDR (ATSDR, 1997, as cited in USEPA, 2001) which reported an average daily water
intake for the general population of 2 to 20 |ig/day.

The use of ambient air data to estimate inhalation exposure does not account for possible differences
between contaminant levels in indoor vs. outdoor air. TCE readily volatilizes from water and indoor
inhalation  exposure may be comparable or greater than ingestion exposures in homes where the water
supply contains TCE (ATSDR, 1997; Andelman, et al.,  1985; Giardino, et al., 1992; Andelman et al.,
1986a; Andelman et al., 1986b, as cited in USEPA,  2001). For example, in two homes using well water

Occurrence Summary and Use Support Document         445                                    March 2002

-------
with TCE levels averaging 22 to 128 |ig/L, the TCE levels in bathroom air ranged from <0.5 to 40 mg/m3
when the shower was run less than 30 minutes (Andelman et al., 1985, as cited in USEPA, 2001). In one
study, the transfer of TCE from shower water to air had a mean efficiency of 61% (independent of water
temperature); it was concluded that a 10-minute shower in TCE-contaminated water could result in a
daily inhalation exposure comparable to that expected from drinking TCE-contaminated tap water
(ATSDR, 1997, as cited in USEPA, 2001).  TCE in bathing water can also cause dermal exposure. A
modeling study has suggested that a significant fraction of the total dose associated with exposure to
volatile organics in drinking water results from dermal absorption (Brown, et al., 1984, as cited in
USEPA, 2001).

Pharmacokinetic modeling can be used to gain further understanding of general population exposure.
Clewell et al. (1995,  as cited in USEPA, 2001) developed a physiologically based pharmacokinetic model
for TCE that can be used to estimate the long-term average ingested dose that would result in a measured
blood concentration, assuming no other TCE exposure. This dose can be converted to a TCE water
concentration assuming an ingestion rate such as 2 L/day. This model was applied to the range of TCE
levels in blood as measured in NHANES 111. The TCE environmental concentrations modeled from
blood levels exceeded the range of measured values for air and water: modeled mean concentration in
drinking water was 59.5 |ig/L (measured range was trace to 50 |ig/L) and the modeled mean air
concentration was 4.2 |ig/m3 (measured range was for 0.01 to 3.9 |ig/m3) . This implies that neither
inhalation nor water ingestion dominate exposure; rather both contribute to the total exposure. Exposure
estimates derived from blood cannot distinguish among exposure routes and sources. It is generally
believed that TCE exposure occurs primarily via water consumption and air inhalation, but it is
impossible to use the blood data to directly estimate how much of the total exposure is attributable to
each. A wide range of combinations of exposures from air and water could have produced the measured
blood levels. As noted earlier, most water supplies have TCE levels under the MCL of 5 |ig/L. The
modeling suggests that exposure at the MCL would correspond to a very low blood level. This implies
that the TCE exposure via the air and other nonwater pathways may generally be more important than
water ingestion.

4.10.5.2.2 Extent of General Population Exposure

Because of the pervasiveness of TCE in the environment, most people are likely to have some exposure
via one or more of the following pathways: ingestion of drinking water, inhalation of ambient air, or
ingestion of food (ATSDR, 1997, as cited in USEPA, 2001). As noted earlier, the NHANES survey
suggests that about 10% of the population has detectable levels of TCE in their blood. The exposures in
these individuals may be higher than those in others in the general population, as a result of a number of
factors. Some members of the general population may have increased TCE exposure via their drinking
water. The extent of TCE exposure via drinking water is difficult to estimate; but the following
discussion provide some perspective on this issue.

TCE is the most frequently reported organic contaminant in ground water (ATSDR, 1997, as cited in
USEPA, 2001). Ninety-three percent of the public water systems in the United  States obtain water from
groundwater (USEPA,  1995, as cited in USEPA, 2001), and between 9% and 34% of the drinking water
supply sources tested in the United States may have some TCE contamination (ATSDR, 1997, as cited in
USEPA, 2001). Although commonly detected in water supplies, the levels are generally low, since, as
discussed earlier, MCL violations for TCE in public water supplies are relatively rare for any extended
period (USEPA, 1998,  as cited in USEPA, 2001). Private wells,  however, are often not closely
monitored, and, if located near TCE disposal/contamination sites where leaching occurs, may have
undetected contamination levels. About 10% of Americans (27 million people) obtain water from sources

Occurrence Summary and Use Support Document         446                                    March 2002

-------
other than public water systems, primarily private wells (USEPA, 1995, as cited in USEPA, 2001). TCE
is a common contaminant at Superfund sites. It has been identified in at least 861 of the 1,428 hazardous
waste sites proposed for inclusion on the EPA National Priorities List (NPL) (ATSDR, 1997, as cited in
USEPA, 2001). Studies have shown that many people live near these sites: 41 million people live less
than 4 miles from one or more of the nation's NPL sites, and, on average, 3,325 people live within 1 mile
of any given NPL site (ATSDR, 1996, as cited in USEPA, 2001). Thus, although exact estimates cannot
be made, many people are probably exposed to TCE via drinking water from private wells. It is not
known how often such exposures would be above the MCL.

4.10.6 Conclusion

Trichloroethylene is primarily used for vapor degreasing of fabricated metal parts.  It is also used as a
chemical intermediate and as a product for export. The use of trichloroethylene has also declined due to
its toxicity. Recent statistics regarding production are not available, but according  to TRI data,
trichloroethylene is widely manufactured and processed in substantial amounts. Trichloroethylene is also
a TRI chemical.  Industrial releases of trichloroethylene have occurred since 1988 in 46 States and Puerto
Rico. Trichloroethylene was an analyte for the NAWQA and NURP ambient occurrence studies. In the
NAWQA study, trichloroethylene was detected in 11.6% of urban wells and 1.6% of rural wells, with
median detection values of 0.9 |ig/L and 0.6 |ig/L, respectively. In the Stage 2 analysis of 16-State
occurrence of trichloroethylene, 0.236% of combined ground water and surface water systems serving
8.19% of the population exceeded the MCL of 0.005 mg/L. Nationally, 154 ground water and surface
water systems combined (serving approximately 17,451,800 people) are estimated to have levels greater
than the MCL.

The  16-State cross-section was designed to be nationally representative based upon VOC, SOC, and IOC
pollution potentials as suggested by considerations of manufacturing, agriculture, and geographic
diversity factors. Nationally, trichloroethylene is manufactured and/or processed in 40 States  and has
TRI releases in 46 States.  Trichloroethylene is manufactured  and/or processed in 14 out of the 16 cross-
section States and has TRI releases in all of the 16 cross-section States. The cross-section should
adequately represent the occurrence of trichloroethylene on a national scale based upon the use,
production, and release patterns of the 16-State cross-section in relation to the patterns observed for all
50 States.

4.10.7 References

Agency for Toxic Substances and Disease Registry (ATSDR). 1996. Biennial  Report to Congress (1991
       and 1992).  Atlanta, GA: U.S. Department of Health and Human Services,  Centers for Disease
       Control and Prevention, ATSDR. Internet site: www.dhhs.gov.

Agency for Toxic Substances and Disease Registry (ATSDR). 1997. Toxicological Profile for
       Trichloroethylene. U.S. Department of Health and Human Services, Public Health Service. 298
       pp. + Appendices. Available on the Internet at http://www.atsdr.cdc.gov/toxprofiles/tpl9.pdf

Andelman, J.B. 1985. Human exposure to volatile halogenated organic chemicals in indoor and outdoor
       air. Environ Health Per sped 62:313-318.

Andelman, J., A. Couch, W.W. Thurston. 1986a. Inhalation exposures in indoor air to trichloroethylene
       from shower water.  Chapter 16.  Environmental epidemiology. Kopfler, F.C. and G.C. Craun
       (eds.). Chelsea, MI: Lewis Publishers.

Occurrence Summary and Use Support Document         447                                     March 2002

-------
Andelman, J.B., S.M. Meyers, L.C. Wilder.  1986b.  Volatilization of organic chemicals from indoor uses
       of water. Chemicals in the environment. Lester, J.N., R. Perry, and R.M. Sterritt (eds.).
       London: Selper Ltd.

Brown, H.S., D.R. Bishop, C.A. Rowan.  1984.  The role of skin absorption as a route of exposure for
       volatile organic compounds in drinking water . Am. J. Publ. Health 74(5): 479-484.

Clewell, H.J., P.R. Gentry, J.M. Gearhart et al. 1995. Considering pharmacokinetic and mechanistic
       information in cancer risk assessments for environmental contaminants: examples with vinyl
       chloride and trichloroethylene. Chemosphere. 31(1): 2561-2578.

DHS. 1986. Organic chemical contamination of large public water systems in California. Sacramento,
       CA: California Department of Health Services.

Giardino, N.J., E. Gumerman, N.A. Esmen, et al. 1992. Shower volatilization exposures in homes using
       tap water contaminated with TCE. J. Exposure Analysis Environ Epidemiol (Suppl. 1): 147-158.

International Agency for Research on Cancer (IARC). 1995. IARC monographs on the evaluation of
       carcinogenic risks to humans: dry cleaning,  some chlorinated solvents and other industrial
       chemicals. Vol.63. Lyon, France: IARC.

Lopes, T.J. and S.G.  Dionne.  1998. A Review of Semivolatile and Volatile Organic Compounds in
       Highway Runoff and Urban Stormwater.  U.S. Geological Survey Open-File Report 98-409.
       67pp.

National Toxicology Program (NTP).  2001. National Toxicology Program Health and Safety
       Information Sheet - Trichloroethylene. Available on the Internet at
       http://ntp-server.niehs.nih.gov/htdocs/CHEM_H&S/NTP_Chem77Radian79-01-6.html, accessed
       July 23, 2001.

Spectrum Laboratories.  2001. Chemical Fact Sheet - Trichloroethylene. Available on the Internet at
       http://www.speclab.com/compound/c79016.htm, accessed July 23, 2001.

Squillace, P.J., M.J. Moran, W.W. Lapham, C.V. Price, RM. Clawges, and J.S. Zogorski. 1999.
       Volatile organic compounds in untreated ambient groundwater of the United States, 1985-1995.
       Env. Sci. and Tech. 33(23):4176-4187.

USEPA.  1995. The  National Public Water Systems  Supervision Program.  The FY1994 Compliance
       Report. EPA 812-R-94-001.  Washington, DC: Office of Water, USEPA.

USEPA.  1998. EPA data file. Office of Groundwater and Drinking Water. Washington, D.C.: U.S.
       Environmental Protection Agency.

USEPA. 2000. TR1Explorer: Trends. Available on the Internet at:
       http://www.epa.gov/triexplorer/trends.htm, last updated May 5, 2000.

USEPA. 2001. Sources, Emissions and Exposure for Trichloroethylene (TCE) and Related Chemicals.
       Office of Research and Development, USEPA. March 2001.
Occurrence Summary and Use Support Document         448                                    March 2002

-------
USEPA.  2002.  Occurrence Estimation Methodology and Occurrence Findings for Six-Year Review of
       National Primary Drinking Water Regulations - DRAFT. EPA Report/815-D-02-005, Office of
       Water, 55 pp.
Occurrence Summary and Use Support Document         449                                    March 2002

-------
                                          [Intentionally blank]
Occurrence Summary and Use Support Document          450                                         March 2002

-------
                            Abbreviations and Acronyms
>MRL        - analytical results greater than the Minimum Reporting Limit (or Level)
1,1,2-TCE     - 1,1,2-trichloroethane

ai             - active ingredient
AMC         - annualized mean concentration
ARP          - Acetochlor Monitoring Partnership
ARP-GWMP   - Acetochlor Monitoring Partnership Ground water Monitoring Program
ATSDR       - Agency for Toxic Substances and Disease Registry
AWT         - Advanced Wastewater Treatment
AWWA       - American Water Works Association
CA           - Census of Agriculture
CAS          - Chemical Abstract Service
CASRN       - Chemical Abstract Service Registry Number
CDC          - Centers for Disease Control and Prevention
CERCLA      - Comprehensive Environmental Response, Compensation & Liability Act
CMR         - Chemical Monitoring Reform
CWS          - Community Water System
CWSS        - Community Water Supply Survey

DBCP        - l,2-Dibromo-3-chloropropane
DCE          - dichloroethylene
DCP          - dichloropropane
DEHP        - di(2-ethylhexyl)phthalate
DHS          - Department of Health Services, California

EFED         - Environmental Fate and Effects Division
EPA          - Environmental Protection Agency
EPCRA       - Emergency Planning and Community Right-to-Know Act

FIFRA - Federal Insecticide, Fungicide, and Rodenticide Act
FR           - Federal Register
FRDS         - Federal Reporting Data Systems
FSTRAC      - Federal/State Toxicological and Regulatory Alliance Committee

g/mol         - grams per mole
GW          - ground water
GWSS - Groundwater Supply Survey
HAL
HSDB
- Health Advisory Level
- Hazardous Substances Data Bank
IARC
IOC
- International Agency for Research on Cancer
- inorganic compound
LLMV - lower limit of method validation
LOD          - level of detection
Occurrence Summary and Use Support Document
                              451
March 2002

-------
MCL         - Maximum Contaminant Level
MCLG - Maximum Contaminant Level Goal
MDA         - Minnesota Department of Agriculture
MDH         - Minnesota Department of Health
MDL         - Method Detection Limit
mg/L         - milligrams per liter
MRL         - Minimum Reporting Level

NAWQA      - National Water Quality Assessment Program
NAWQA      - National Alachlor Water Well Survey
NCFAP       - National Center for Food and Agricultural Policy
NHANES      - National Health and Nutrition Examination Survey
NIRS         - National Inorganics and Radionuclides Survey
NLM         - National Library of Medicine
NOMS - National Organic Monitoring Survey
NORS        - National Organics Reconnaissance Survey
NPS          - National Pesticide Survey
NPDES       - National Pollution Discharge Elimination System
NPDWR      - National Primary Drinking Water Regulations
NPL          - National Priorities List
NSC          - National Safety Council
NSDWR      - National Secondary Drinking Water Regulation
NSP          - National Screening Program for organics in drinking water
NTNCWS     - Non-Transient Non-Community Water System
NTP          - National Toxicology Program
NURP        - National Urban Runoff Program

ODS          - ozone depleting substance
ODW         - Office of Drinking Water
OGWDW      - Office of Ground Water and Drinking Water
OL           - Optimum Level
OPP          - Office of Pesticide Programs
OSHA        - Occupational Safety and Health Administration

PCBs         - polychlorinated biphenyls
PEL          - permissible exposure limit
PGWDB      - Pesticides in Ground Water Data Base
pH           - the negative log of the concentration of H+ ions
PLUARG      - Pollution from Land Use Activities Reference Group
PPG          - Pittsburgh Paint and Glass Industries
PVC          - polyvinyl chloride
PWS          - Public Water System
PWSID       - Public Water System Identifier

QA           - quality assurance
QC           - quality control

RCRA        - Resource Conservation and Recovery Act
RED          - Reregistration Eligibility Decision
RWS         - Rural Water Survey
Occurrence Summary and Use Support Document
452
March 2002

-------
RY

SAIC
SARA Title III
SDWA
SDWIS/FED
SMCL
SMSA
soc
SRI
STORET
SW

TCE
TDI
TRI
TSD
TWMC
- Reporting Year

- Science Applications International Corporation
 Superfund Amendments and Reauthorization Act
- Safe Drinking Water Act
- Safe Drinking Water Information System/Federal version
- Secondary Maximum Contaminant Level
- standard metropolitan statistical areas
- synthetic organic compound
- formerly the Stanford Research Institute
- STOrage and RETrieval
- surface water

- trichloroethylene
- toluene diisocyanate
- Toxics Release Inventory
- Technical Support Division (ODW)
- time weighted mean concentration
USEPA       - United States Environmental Protection Agency
USGS         - United States Geological Survey
USITC - United States International Trade Commission
USPHS       - United States Public Health Service
voc

WHO
- volatile organic compound

- World Health Organization

- micrograms per liter
Occurrence Summary and Use Support Document
                              453
March 2002

-------