Sorting And Screening of Potential Drinking Water Contaminants: New and Existing Chemicals Under the Toxic Substances Control Act
Charles M. Auer, Christopher Blunck, Flora Chow, and David R. Williams
This paper reviews the chemical assessment and testing programs under the U.S. Toxic Substances Control Act (TSCA) of 1976 and considers elements relevant to screening and assessment of new and existing industrial chemicals as drinking water contaminants. TSCA established a number of new requirements and authorities for identifying and controlling existing and potential toxic chemical risks to human health and the environment. The TSCA programs are implemented by the U. S. Environmental Protection Agency's (EPA) Office of Pollution Prevention and Toxics. Generally, TSCA gives EPA the authority to gather certain kinds of basic data relevant to determining chemical risks from those who manufacture and process chemicals. The law also enables EPA to require companies to test chemicals for toxic effects and requires the agency to review most new chemicals before they are manufactured commercially. To prevent unreasonable risks, EPA may select from a broad range of control actions under TSCA, from requiring hazard-warning labels to outright bans on manufacture and/or use. EPA may regulate a chemical's risks at any stage of its life cycle: at manufacture, processing, distribution in commerce, use, or disposal. TSCA coverage does not extend to certain product categories, including tobacco, munitions, food additives, drugs, cosmetics, and pesticides.
Section 5 of TSCA provides EPA with the authority to regulate new chemical substances (i.e., those not on the TSCA inventory) of existing chemicals) prior to their commercial manufacture. Anyone who plans to manufacture or import a new chemical substance for a commercial purpose is required to provide the EPA with a premanufacture notice (PMN) at least 90 days prior to the activity.
Section 5 of TSCA thus gives EPA the role of gatekeeper between the laboratory and the commercial marketplace. EPA reviews the new chemical to determine whether its manufacture, processing, distribution in commerce, use, or disposal "may present an unreasonable risk" to human health or the environment or cause exposures of concern. The program's assessment includes exposures
and risks to workers, consumers, and the general population (including from drinking water or fish consumption) as well as risks to wildlife. From its inception in 1979 the program has reviewed over 30,000 new chemical notices (currently about 2,000 a year) and taken action to control or require testing on approximately 10 percent of the new chemicals notified.
PMN submissions require all available data on chemical identify, production volume, byproducts, use, environmental release, disposal practices, and human exposure. TSCA does not require the submission of "base set" testing with the notice, although the submitter is required to provide all health and environmental data in the possession of the submitter.
EPA assesses the potential risks to humans or the environment of each new substance based on data submitted with the notice, other information available to the agency, and exposure and release modeling. As noted, TSCA does not require prior testing of new chemicals and, as a consequence, approximately 50 percent of submissions contain no test data of any type. When submitted, health test data rarely extend beyond acute studies (~40 percent) or genotoxicity tests (~15 percent; Auer and Gould, 1987). In the absence of test data, EPA bases its health and environmental toxicity review of new substances primarily on structure activity relationships (SARs)(Auer and Gould 1987; Auer et al., 1990, 1994, 1995; Leeman et al. 1995).
Almost 90 percent of the PMNs submitted to the program complete the review process without being restricted or regulated in any way. Ten percent of the PMNs, however, are regulated by EPA, either under a TSCA Section 5(e) consent order or by a Significant New Use Rule (SNUR).
Section 5(e) Consent Orders
EPA may determine that activities involving the new substance "may present an unreasonable risk of injury to health or the environment" (referred to as a risk-based finding) and that the information available is insufficient to permit a "reasoned evaluation" of the new chemical. When EPA makes these two findings it acts under Section 5(e) of TSCA to regulate the new substance and can control uses and/or releases until test data or other information sufficient to adequately assess the potential risks become available. Section 5(e) consent orders have specified a variety of control measures, including protective equipment, use limitations, process restrictions, labeling requirements, and limits on environmental release. In other instances EPA may determine that a new substance will be produced in substantial quantities (currently set at 100,000 kg/year) and there is or may be significant or substantial human or environmental exposure to the substance (referred to as an exposure-based finding) and that the
available information is insufficient. Consent orders issued to address exposure-based concerns include testing requirements similar to the Minimum Premarket Dataset (MPD) used in the European Union (see Auer and Gould, 1987, for a description).
Section 5(e) orders apply only to the submitter of the PMN. When a PMN notifier submits a Notice of Commencement of Manufacture to EPA, EPA adds the former new substance to the TSCA inventory. When a substance is listed on the inventory, other persons are able to manufacture, import, or process it without EPA review and without any restrictions imposed by a Section 5(e) order. Under TSCA Section 5(a)(2), EPA may determine by rule that a use of a chemical substance is a "significant new use." Once EPA determines that a use of a chemical substance is a significant new use, TSCA requires persons to submit a Significant New Use Notice (SNUN) to EPA at least 90 days before they manufacture, import, or process the chemical substance for that use. Thus, EPA can use its SNUR authority to extend limitations in Section 5(e) orders to companies beyond the original submitter. After receiving and reviewing a SNUN, EPA has the option of either permitting the new use or acting to regulate the new submitter's activities.
Since 1979, EPA has reviewed over 32,000 new chemical substances and taken action to prevent potential risks to people and the environment on nearly 3,000. PMN actions based on health concerns for exposures from contaminated drinking water are uncommon (typically fewer than five such cases per year (Becky Jones and Flora Chow, OPPT, personal communication, 1998).
EU/U.S. Structure Activity Relationship/Minimum Premarketing Dataset Study
As noted above, EPA does not receive any test data with most PMNs. When data are submitted, they often do not go beyond acute toxicity endpoints. Quantitative structure activity relationships (QSARs) are the technique EPA uses to carry out preliminary hazard assessments of new chemicals in the absence of test data. These QSARs are predictive methods that estimate the properties of a chemical (e.g., melting point, vapor pressure, toxicity, ecotoxicity), on the basis of its structure and test data on analogous chemicals. A joint U.S./EU study was initiated to evaluate the predictive power of the QSAR by applying QSAR methods to chemicals for which MPDs were already available in the EU and then
comparing the properties predicted by QSAR with the properties observed from MPD testing (OECD, 1994). Analysis of the results of this study showed that the SAR approach was relatively successful in identifying chemicals of concern. Nevertheless, the study concluded, the approach could be improved by selectively incorporating specific testing schemes into the process. Such a focused effort would provide valuable data while not presenting large overall cost implications.
Persistent/Bioaccumulative/Toxic (PBT) New Chemicals
EPA recently proposed a new chemical category for PBT new chemicals (EPA, 1998a) that possess characteristics of persistence, bioaccumulation, and toxicity that cause such chemicals to present potentially significant hazards and risks. The proposed category is intended to alert manufacturers to EPA's concerns and facilitate assessment and risk management of such new chemicals. The policy statement includes boundary conditions, such as environmental half lives (>2 months and >6 months) and bioaccumulation factors (>1,000 and >5,000) that would determine inclusion in the category, recommended testing to address PBT concerns, and EPA's regulatory strategy for chemicals meeting the category requirements.
EPA's TSCA inventory currently contains over 70,000 chemicals, many of which are produced or imported at low or negligible volumes, while others are polymers that, because of their size (e.g., high molecular weight) and other characteristics, are unlikely to present significant risk concerns. Excluding low-volume chemicals (~25,000 chemicals produced in amounts less that 10,000 pounds per year) and polymers (which tend to be poorly absorbed by organisms and typically exhibit low toxicity), there are ~15,000 nonpolymeric chemicals that are produced at levels above 10,000 pounds per year. This 15,000 chemicals subset has been identified as being the broad focus of EPA's existing chemical testing and assessment program with the primary focus being on the ~3,000 U.S. high production volume (HPV) chemicals that are produced/imported at levels above 1 million pounds per year.
Existing Chemical Testing
Section 2 of TSCA states that "it is the policy of the United States that adequate data be developed with respect to the effect of chemical substances and mixtures on health and the environment and that development of such data be the responsibility of those who manufacture [and import] and those who process such chemicals and mixtures." Section 4 of TSCA gives EPA the authority to issue rules to require chemical producers, importers, and processors to conduct specified testing for health and environmental effects as well as chemical fate and exposure.
EPA must make certain Statutory "findings" to require chemical testing under a TSCA Section 4 Test Rule, as follows. The chemical may present an unreasonable risk; and/or is produced in substantial quantities and enters the environment in substantial quantities, or there is or may be substantial or signifier human exposure; and existing data are inadequate; and testing is necessary. As a result of litigation in the 1980s, EPA and other stakeholders developed an alternative to the TSCA Section 4 rule-making process that involves the negotiation of formal TSCA Section 4 enforceable consent agreements. EPA also looks to voluntary industry testing initiatives to supplement its formal chemical testing program. A primary example of such a voluntary initiative is the Screening Information DataSet (SIDS) program, an important program focused on cooperative voluntary testing of international HPV chemicals. The program is operated under the auspices of the Organization for Economic Cooperation and Development (OECD). Thus, EPA relies on a mix of TSCA Section 4 test rules, ECAs, orders, and voluntary testing agreements. Since 1979, approximately 550 chemicals have been the subject of final testing actions. Virtually all of the actions taken to date have involved U.S. HPV chemicals and more than 50 percent of the testing actions have been developed since 1990. Other recent developments related to existing chemical testing are discussed below.
TSCA Section 4 Final Test Rule for EPA's Office of Drinking Water
On November 10, 1993, OPPT published a final TSCA Section 4 Test Rule (58 FR 59667) covering four chemicals of interest to the Office of Drinking Water (ODW) in EPA's Office of Water. The chemicals subject to this rule (chloroethane, 1,3,5-trimethylbenzene, 1,1-dichloroethane, and 1,1,2,2-tetrachloroethane) were unregulated drinking water contaminants for which ODW needed data in order to develop 1-day, 10-day, and long term/lifetime health advisories. The required testing included 14- and 90-day oral toxicity studies in rats on each of the subject chemicals.
In consultation with ODW and following receipt of adequate data from studies conducted by the National Toxicology Program or conducted by industry as a result of settlement of lawsuits challenging the test rule, OPPT1 revised the
Testing included in the OECD Screening Information DataSet (SIDS). The SIDS test battery includes the following basic screening endpoints: physical-chemical properties, environmental fate, ecotoxicity, acute toxicity, genetic toxicity, repeat-dose toxicity, and developmental and reproductive toxicity, which are listed in Section 2.2 of the Screening Information DataSet Manual of the OECD Programme on the Co-operative Investigation of High Production Volume Chemicals, published in July 1997. This manual (also called the SIDS Manual) is available at www.epa.gov/opptintr/sids/sidsman.htm or can be obtained as hard copy from the OECD Environment Directorate, Environmental Health and Safety Division; 2, rue Andre-Pascal F-75775; Paris Cedex 16, France; Tel: 3314-4524 9844. Specific information on the SIDS test protocols can be found at:
final test rule to remove (1) the 14- and 90-day testing requirements for chloroethane and 1,1-dichloroethane and (2) the 90-day subchronic testing requirement for 1,1,2,2-tetrachloroethane. The remaining studies have been completed and the results forwarded to ODW for review and disposition.
Existing Chemical Assessment
Historically, the existing chemical assessment effort focused on screening efforts to identify potential problem chemicals that were then subjected to increasingly detailed reviews to determine their hazards, exposures, and risks. Recent developments have occurred that have fundamentally changed the focus of this EPA program.
Chemical Hazard Data Availability Study
In an Earth Day 1998 announcement, Vice President Gore and EPA Administrator Carol Browner committed EPA to testing initiatives aimed at strengthening the public's right and ability to know about the potential health and environmental risks from HPV existing chemicals. The announcement on the testing of HPV chemicals is a direct result of a recent EPA (1998a) analysis of the public availability of basic testing and screening information on chemicals produced or imported at a rate of more than 1 million pounds per year. EPA researched public information sources for data contained in the internationally agreed-upon Screening and Information DataSet (SIDS), considered the minimum set of tests that can allow an informed screening-level evaluation of a chemical's hazards. The study found that of the 3,000 U.S. HPV chemicals a full set of SIDS testing was publicly available for only 7 percent of the chemicals and that no SIDS data were available for 43 percent of the chemicals. The report also considered specific subsets of chemicals, including EPA's Toxics Release Inventory-listed chemicals, those with occupational exposure standards, and consumer chemicals. A similar analysis could be conducted for chemicals found in drinking water.
This recent EPA study is similar in several respects to a 1984 National Research Council (NRC) report. The NRC report considered the availability of
www.oecd.org/ehs/hpv.htm. The tests needed for each of the six screening endpoints are (a) physical/chemical property tests: melting point, boiling point, vapor pressure, n-octanol/water partition coefficient (shake flask method), water solubility; (b) environmental fate tests: photodegradation, hydrolysis stability in water, transport/distribution, inherent biodegradation; (c) ecotoxicity tests: acute toxicity to fish, acute toxicity to Daphnia, toxicity to aquatic plants (algae), chronic toxicity to Daphnia, when appropriate, terrestrial organism toxicity, when appropriate; (d) mammalian acute toxicity test: acute inhalation or acute oral or acute dermal test (if testing is needed, oral is default route except for gases); (e) mammalian genotoxicity tests: gene mutation (e.g., ames salmonella), chromosomal aberrations (in vive mouse micronucleus preferred if testing is needed); and (f) mammalian repeated dose/reproductive/developmental effects: combined repeated-dose with reproductive/developmental toxicity screen (45-day exposure) or repeated dose oral toxicity (28-day) and first generation reproductive toxicity test.
test data for various groups of chemicals, including industrial chemicals. It concluded that "minimal" toxicity information (defined as availability of any one or more of the following test types included in the NRC study: acute toxicity, subchronic toxicity, reproductive/developmental toxicity, mutagenicity) was available for only 22 percent of HPV chemicals. The EPA study found, in comparison, that while there is much work to be done, some progress has been made in improving our understanding of chemical hazards.
HPV Challenge Program
On October 9, 1998, the Vice President, EPA Administrator Browner, the Chemical Manufacturers Association, the American Petroleum Institute, and the Environmental Defense Fund made a joint announcement of a cooperative program to test 2,800 U.S. HPV chemicals and invited the chemicals industry to participate in a voluntary HPV Challenge Program to provide the public with basic screening data on the high-volume chemicals they produce. Under the HPV Challenge Program, companies can commit to voluntarily develop complete SIDS test datasets for their chemicals by the end of 2004. In addition, EPA will, as necessary, propose and finalize TSCA Section 4 test rules to require SIDS testing of HPV chemicals not handled under the Challenge Program. EPA is taking actions to secure these data so that individuals and communities can better evaluate the chemical hazards and risks they face.
Because of Vice President Gore's initiative, the United States expects to have complete SIDS test results on all U.S. HPV chemicals by 2005. It is also hoped that the U.S. action will spur other countries to step up their rate of testing under the closely related OECD SIDS effort.
Toxics Release Inventory
EPA requires annual reports of toxic chemical releases and transfers under Section 313 of the Emergency Planning and Community Right-to-Know Act. These reports provide the public with information on the releases of 600 listed toxic chemicals and chemical categories in their communities. Additionally, the information is used by EPA to support the development of EPA regulations and programs. Generally, facilities must report the quantities of both routine and accidental releases of listed toxic chemicals, including releases to the atmosphere, surface water, and land (Title HI of the Superfund Amendments and Reauthorization Act, 1986).
Approaches To Sorting And Screening Industrial Chemicals As Drinking Water Contaminants
EPA currently attempts to identify potential drinking water contaminants (both ground- and surface water) as part of its new chemicals review program. The potential hazards are generally identified using SARs, while exposure is estimated using a combination of measured and estimated physical-chemical properties, estimates of the amount released, estimates of removal in treatment processes, fate and transport in the environment, and the extent of dilution by the receiving environmental medium. Releases to surface water can occur from manufacture, processing, use (including industrial, commercial, and consumer), and disposal. The primary fate and transport processes that are considered in estimating removal during treatment include biodegradation, sorption, volatilization, and hydrolysis. For rivers and streams, complete mixing can be assumed and a simple dilution model can be used to estimate surface water concentrations. If the specific site of release and the point of entry into a drinking water treatment facility are known, the analysis can account for the additional dilution that can occur before entry into the drinking water treatment facility. Estimates of potential dose rates from ingestion of surface water are developed using surface water concentrations based on mean stream flow values multiplied by the daily drinking water ingestion rate and the annual frequency of ingestion (EPA, 1998b).
Potential carcinogenicity is the most frequently identified risk involving drinking water. The potential carcinogenicity of a PMN chemical is generally identified based on structural or functional analogy to known carcinogens and consideration of absorption and metabolic potential; consideration of mechanisms is also important when information is available (for more information on the approach, see Auer and Gould, 1987, and Auer et al., 1990). Using carcinogenicity test results on the analog, the data are modeled to produce a unit risk factor that can then be coupled with estimates of exposure to estimate drinking water risks. Carcinogenicity risk estimates of greater than 10-6 are generally considered significant for general population exposure.
As noted earlier, identification of new chemicals presenting significant concerns about risk based on drinking water exposures occurs at most several times per year (i.e., typically less than five such cases are identified each year out of over 2,000 new chemical notifications).
As for new chemicals, existing chemical hazards can also be screened using available data and SARs to identify chemicals of potential concern. Several large compilations of SAP, estimates for the health effects of industrial chemicals have been developed by EPA, although most are several years old (Becky Jones, OPPT, personal communication, 1998). However, in light of the HPV Challenge Program and the likelihood that basic screening test data will be available for almost 3,000 HPV industrial chemicals by 2005, the best strategy
may be to wait for the development and reporting of this information to support identification of industrial chemical drinking water contaminants warranting more detailed assessment.
In addition, emissions and release data contained in the toxic release inventory can be analyzed to identify chemicals presenting potential concerns owing to releases into drinking water sources.
This paper describes major components of EPA's program to assess, test, and manage the risk of new and existing industrial chemicals under TSCA. The paper also explores possible approaches to sorting and screening these chemicals to identify possible candidates that present drinking water concerns.
More information about TSCA and other EPA activities and programs can be found on EPA's Internet home page on the World Wide Web at http://www.epa.gov.
Auer, C. M. and D. H. Gould. 1987. Carcinogenicity assessment and the role of structure activity relationship (SAP,) analysis under TSCA Section 5. Environment Carcinogen Research C5(I):29-71.
Auer, C. M., J. V. Nabholz, and K. P. Baetcke. 1990. Mode of action and the assessment of chemical hazards in the presence of limited data: Use of structure-activity relationships (SAR) under TSCA Section 5. Environmental Health Perspectives 87:183-197.
Auer, C. M., M. Zeeman, J. V. Nabholz and R. G. Clements. 1994. SAR—The U.S. regulatory perspective. SAR & QSAR Environmental Research 2(1-2): 29-38.
EPA (U.S. Environmental Protection Agency). 1998a. Chemical Hazard Data Availability Study. Available on the Internet at http://www.epa.gov/opptintr/chemtest/hazchem.htm.
EPA. 1998b. Guidelines for Completing Initial Review Exposure Report for New Chemical Substances. Washington, D.C.: OPPT/EAB.
NRC (National Research Council). 1984. Toxicity Testing: Strategies to Determine Needs and Priorities. Washington, D.C.: National Academy Press.
OECD (Organization for Economic Cooperation and Development). 1994. EPA/EC Joint Project on the Evaluation of (Quantitative) Structure Activity Relationships. OECD Environment Monographs No. 88. Paris: OECD.
Zeeman, M., C. M. Auer, R. G. Clements, J. V. Nabholz, and R. S. Boethling. 1995. U.S. EPA regulatory perspectives on the use of QSAR for new and existing chemical evaluations. SAR and QSAR Environmental Research 3(3):179-201.