Setting Priorities for Drinking Water Contaminants (1999)

The establishment of laws, advisories, and standards to protect public drinking water supplies has been a major endeavor at the federal, state, and local government levels for decades. The provision of safe drinking water has been one of the major triumphs of twentieth century U.S. public health practice and has been a major factor in the improvement of the health status of U.S. communities.

For most Americans, in this century turning the tap has been an act of faith in the generally safe character of our public water supplies. The increasing consumption of bottled waters, whose cost is hundreds of times the cost of tap water, however, suggests that the public has begun to question that faith. At the same time, despite the presence of several layers of regulatory protection, many sources of raw and finished public drinking water in the United States contain chemical, microbiological, and even radiological contaminants at detectable levels (Neal, 1985). Some of these contaminants pose risks not only via ingestion of the contaminated water (which can be avoided by drinking bottled water) but also via dermal contact with the water or inhalation of vapors while showering. A recent National Research Council (NRC) report, Safe Water From Every Tap (NRC, 1997), noted that 23.5 percent of all U.S. community public water systems violated microbiological standards under the Safe Drinking Water Act (SDWA) at least once between October 1992 and January 1995 and that 1.3 percent violated chemical standards. The frequent presence of such contaminants, as well as documented outbreaks of waterborne disease and the many other outbreaks thought to go undetected, are a clear reminder that unprotected and contaminated drinking water can still pose health risks to the population.

When Congress amended the SDWA in 1996, it required the U.S. Environ-

mental Protection Agency (EPA) to prepare periodically a list of unregulated contaminants¹ to assist in priority-setting efforts for the agency's drinking water program. Subsequently, the first Drinking Water Contaminant Candidate List (CCL) was published in final form in the Federal Register (63 FR 10274) on March 2, 1998 (EPA, 1998a). Under the authority of the SDWA Amendments of 1996, and at the direction of Congress, EPA asked the NRC to recommend criteria for prioritizing CCL contaminants into categories for future regulatory action. EPA will need to classify contaminants into those ready for a rule-making decision; those ready for guidance development, including health advisories;² those needing additional occurrence³ data; and those that are priorities for additional health effects or other research.

Under the SDWA amendments, EPA must determine whether or not to regulate at least five contaminants on the CCL by August 2001 (three and one-half years following publication of the first CCL). To support these decisions and meet the statutory language of the SDWA amendments, EPA must determine whether the regulation of each contaminant on the CCL would provide a "meaningful opportunity" to reduce health risk for persons served by public water systems (EPA, 1998b). EPA must also consider risk to sensitive subpopulations such as infants and the elderly (EPA, 1997a). To help meet these requirements, EPA requested recommendations from the NRC for a process or criteria to help select contaminants from the CCL for regulatory determinations.

This report addresses EPA's request. It was prepared by the NRC's Committee on Drinking Water Contaminants, appointed in 1998 in response to EPA's solicitation. The committee consists of 14 volunteer experts in water treatment engineering, toxicology, public health, epidemiology, water and analytical chemistry, risk assessment, risk communication, public water system operations, and microbiology. Members convened twice over a six-month period to develop this report. The group incorporated input from a wide range of stakeholders and EPA personnel concerned about the health, economic, and technological implications of contaminants included on the CCL. The committee also sought the input of representatives of the National Drinking Water Advisory Council's (NDWAC's)

Working Group on Occurrence and Contaminant Selection. This group was integral to the identification and selection of potential drinking water contaminants for inclusion on the 1998 CCL (EPA, 1997a).

This chapter provides a brief overview of the historical development of drinking water regulations, especially the SDWA and its 1996 amendments. It also summarizes the development, purpose, and requirements of the CCL and other related SDWA provisions. Chapter 2 responds to EPA's request for the committee to review and summarize several schemes for prioritizing chemical contaminants according to risk. Chapter 3 briefly summarizes methods used to evaluate risks posed by microbiological contaminants in drinking water. Chapter 4 describes how the 1998 CCL was developed. Lastly, Chapter 5 recommends a decision framework and provides general guidelines for evaluating contaminant-related data for the selection of CCL contaminants for future regulatory action.

The essential benefits of filtration and chlorination of potable water supplies were established in the United States by the beginning of World War I (NRC, 1977). The primary reason for water purification was to protect public health from typhoid fever and other waterborne diseases.

Since 1920, waterborne disease outbreaks have been investigated and reported in a national database, which is operated jointly by EPA and the Centers for Disease Control and Prevention (CDC). According to this database, from 1920 to 1970 there were 1,085 waterborne disease outbreaks in the United States (see Figure 1-1). From 1971 to 1992, more than 164,000 individuals were reported ill during 684 documented waterborne outbreaks (Craun, 1991; Herwaldt et al., 1992; Moore et al., 1993). Interestingly, the average number of outbreaks for the first 50 years was 21 per year, and for the last 22 years the average has been 31 per year. Thus, despite improvements in water treatment, documented outbreaks continue to occur. A portion of this reported increase may be because of improved monitoring, detection, and reporting methods. Nevertheless, epidemiologists generally agree that these reported occurrences represent only a fraction of total waterborne disease outbreaks. The national waterborne disease database depends on detection, investigation, and complete reporting by individual states. However, state data are known to be uneven in quality and do not provide information about undetected or undocumented outbreaks. Furthermore, reporting is voluntary. For example, some states support active surveillance and investigation of waterborne disease outbreaks, while others rely on case reports provided by local physicians and public health officers regarding clusters of illness.

Until passage of the original SDWA in 1974, there was no enforceable provision in federal law to protect the public from hazardous chemical substances in drinking water. Prior to the SDWA, the only enforceable federal drinking

Figure 1-1

Causes of waterborne disease outbreaks reported to the CDC.

SOURCES: Adapted from Craun, 1991; Herwaldt et al., 1992; Moore et al., 1993.

water standards were directed at waterborne pathogens in water supplies utilized by interstate carriers such as buses, trains, airplanes, and ships (Viessman and Hammer, 1985; Dzurik, 1990; NRC, 1997). These standards were originally promulgated under the Public Health Service Act of 1912 (PHSA). While the PHSA did include recommended guidelines for drinking water contaminants unrelated to communicable diseases, they were not enforceable (Dzurik, 1990; NRC, 1997).

the purpose of the original SDWA was to ensure that public water supply systems⁴ meet national primary drinking water regulations for contaminants to protect public health. The SDWA also established a joint federal-state system for ensuring compliance with federal standards. Since its enactment, the SDWA has been significantly amended twice: first in 1986 and, most recently, in 1996. This report is principally concerned with requirements newly established in the SDWA Amendments of 1996.

Under the SDWA, setting a national primary drinking water regulation for a chemical contaminant is a two-step process (Gibson et al., 1997). First, a nonenforceable maximum contaminant level goal (MCLG) is determined. This criterion represents the level in drinking water that would result in "no known or anticipated adverse effect on health" with a margin of safety. Second, MCLGs serve as the target for setting either enforceable national primary drinking water standards, known as maximum contaminant levels (MCLs), or treatment techniques, if contaminant monitoring is not feasible. In general, MCLs are set as close to the MCLG as feasible, depending on risk management considerations (e.g., EPA determines that the cost of a standard at the MCLG is not justified by the benefits) (EPA, 1996a).

For microbiological contaminants, philosophically the original SDWA established a zero tolerance for disease-causing organisms as the health goal (i.e., the MCLG is set at zero). However, treatment performance techniques, rather than specific allowable concentration of pathogens, historically have served as the basis for regulating microbial contaminants. Historically, water supply regulators assumed that all waters carried some level of harmful organisms that could be treated generically, with levels of fecal coliform bacteria (which are generally harmless) serving as an overall measure of the performance of the treatment system. The SDWA Amendments of 1986 and 1996 both required some modifications to this historical approach. The 1986 amendments required development of the Surface Water Treatment Rule to optimize filtration and disinfection of surface waters to protect against microorganisms, such as Giardia , that resist treatment and those, such as Legionella (cause of Legionnaires' disease) that can grow in the water distribution system. Strong source water protection programs in some cases can supplement or supplant filtration requirements of this rule. The 1996 amendments require that methods be considered to protect the population from exposure to recently recognized waterborne pathogens, such as Cryptosporidium.

The NRC helped EPA establish the first set of national primary drinking water regulations for individual contaminants and contaminant classes under the original SDWA (NRC, 1977). The resulting NRC report Drinking Water and Health dealt with standards for chemical, microbiological, particulate, and radionuclide drinking water contaminants. NRC is now being asked to assist EPA in its task of identifying and selecting contaminants for future regulatory attention through the mechanism of the 1998 CCL. The EPA originally requested that this study focus primarily on chemical contaminants. After preliminary deliberations, however, the committee decided to pay equal attention to both chemical and microbiological contaminants.

The first CCL identification and selection process used separate approaches

for microbiological and chemical contaminants. The CCL comprises 60 contaminants and contaminant classes, including 50 chemicals and chemical groups and 10 microbiological contaminants and groups of microbes, as listed in Table 1-1 (EPA, 1997a).

With the exception of sulfate (included as a special case), the CCL includes contaminants that are not currently subject to any proposed or promulgated primary drinking water regulation, but are known or anticipated to occur in public water systems and may require regulation under the SDWA. Thus, the 1998 CCL is intended to be the primary source of priority contaminants for future regulatory actions by EPA's drinking water program until the next CCL is published in 2003 (see Figure 1-2).

As indicated in Figure 1-2, future CCL development will be closely interrelated with two other drinking water programs established by the SDWA Amendments of 1996: the National Drinking Water Contaminant Occurrence Database (NCOD) and the Unregulated Contaminant Monitoring Regulation (UCMR) (EPA, 1998b). Both of these programs, as well as the CCL, are currently the responsibility of EPA's Office of Ground Water and Drinking Water.

The purpose of the NCOD is to store quality-controlled data on the occurrence of regulated and unregulated drinking water contaminants (EPA, 1998b). When operational, the NCOD is expected to provide the basis for identifying drinking water contaminants that may be included on future CCLs and to support EPA's decisions about whether to regulate contaminants in the future. It is also expected to be used in the periodic review of existing contaminant regulations and monitoring requirements. The NCOD is currently under development and, by law, will need to be completed by August 1999. EPA has requested input from the public, states, and the scientific community regarding the NCOD's design, structure, and use (AWWA, 1997; EPA, 1997b).

The 1996 SDWA Amendments also direct EPA to develop regulations for monitoring selected unregulated contaminants (the UCMR) by August 1999 and every five years thereafter (EPA, 1998b). Unregulated contaminant monitoring is currently described under existing SDWA regulations (Title 40, Code of Federal Regulations, Part 141). However, the 1996 amendments require (1) development of a new list, which cannot exceed a total of 30 contaminants, (2) use of a representative sample of public water systems serving 10,000 or fewer people, (3) placement of data in the NCOD (when operational), and (4) consumer notification of monitoring results. Perhaps most significantly, the SDWA Amendments of 1996 require EPA to design the UCMR for use in developing future CCLs, making decisions about whether to regulate a contaminant, and promulgating subsequent regulations (EPA, 1997c). Contaminants from the CCL categorized as requiring additional occurrence data will also provide the primary

source of contaminants selected for inclusion in future UCMRs. EPA is currently requesting input from the public, states, and the scientific community on options for developing the UCMR.

Prior to the development of the first CCL and the direct involvement of stakeholders, states, and the NRC, EPA began work on a conceptual, risk-based approach for identifying unregulated chemical and microbiological drinking water contaminants. The identified agents were those known or anticipated to occur in public drinking water supply systems and to have the potential to affect human health (EPA, 1996b). This approach was called the Contaminant Identification Method (CIM) and was intended to identify and classify contaminants into several possible regulatory and nonregulatory categories. These categories included contaminants to be placed on the CCL (for future regulatory determinations), those requiring further toxicological research, those recommended for monitoring, those needing health advisory development or other guidance, and those for which no action is required. The CIM also was to be used periodically to reevaluate currently regulated contaminants.

Figure 1-2

Time line and interaction of selected major regulatory requirements of the SDWA  Amendments of 1996. SOURCE: Adapted from EPA, 1997b.

Under the CIM, the potential adverse effects, occurrence, production, use, and release of contaminants and other related factors were to be evaluated to assist in setting priorities for chemical and microbiological contaminants. The use of risk-based priorities was intended to use limited resources efficiently and to address the most important public health threats. Because of constraints associated with meeting the legislatively mandated publication deadline of February 1998 for the CCL, the CIM was not completed in time to be implemented in preparing the first CCL, and development has been postponed pending input on the CCL process from the NRC and NDWAC. Chapter 4 further discusses the CIM and the process that EPA used to prepare the draft and final 1998 CCL.

By congressional intent, the current and future CCLs will serve as cornerstones of EPA's future drinking water program. In making future regulatory decisions, section 1412(b)(3)(A)(i) of the amended SDWA requires EPA to use the "best available, peer-reviewed science and supporting studies conducted in accordance with sound and objective scientific practices."

It is disquieting to many nonscientists that scientific experts representing different interests can disagree markedly. There is an implicit assumption that

disagreement among scientists should be rare because science is capable of objective, if not always experimental, verification. In fact, however, differences of opinion are common in science, although the arguments are spread out over many research papers and long time spans and are usually couched in careful, if not polite, language. In a regulatory proceeding, by contrast, time and space are compressed and nuances of language are erased. However, the underlying disagreements exist outside the regulatory arena as well as inside of it.

While disagreements in science are commonplace, they usually center on the applications of scientific reasoning and judgment and the conclusions drawn from such applications, not on disagreements over whether the scientist has used scientifically accepted methodology and reasoning. Epidemiological and toxicological studies—the raw material for scientific judgments of health risks caused by drinking water contamination—are like picture puzzle pieces. Depending on a particular scientist's assessment of a study's validity, the piece may be seen as clear and well defined or as fuzzy and indefinite. Depending upon that same scientist's judgment of a study's relevance, the piece may be deemed as central to the picture, a small piece on the periphery, or not part of the picture at all. The integration of (often-conflicting) epidemiological and toxicological studies in regulatory decision making is more fully discussed in Chapter 5.

The raw materials provide the puzzle pieces, but the parts do not often fit together smoothly or without gaps. Each registers different aspects of the total picture, with results that show only a portion of the whole. Placing a scientific study within a coherent picture requires the use of critical thinking, including evaluation of the part played by bias, chance, and real effect, together and separately, and judgments on what generalizations are valid. In such a complex process and with practical matters of consequence at stake, it is not surprising that differences of opinion develop and are magnified by the regulatory process. But even when so magnified, such disagreements are not artifacts of that process but are essential features of science as it is routinely practiced. The resulting disagreements are not usually evidence of flawed scientific reasoning or methodology.

This report, consequently, takes the position that scientific disagreements are the norm and do not signal a deviation from sound science. These disagreements may be based on values other than strictly scientific ones, however, this does not mean that the sides of the debate are not based on sound science. Indeed, it is not unusual for scientists to disagree on the application of sound science to public policy issues. Any scheme that affects the provision of public water is likely to engender legitimate scientific disagreement. The report also recognizes that identifying and agreeing on what is sound science is itself a difficult and error-prone enterprise. It therefore makes no recommendations on what "soundness" entails, letting the accepted mechanisms of peer regard, peer review, and scientists' habits of critical thinking continue to serve as the ultimate arbiters.

Similarly, the committee purposely declined to define what should be con-

strued as "sufficient data," believing this is a matter of judgment that will vary with context. Any scientifically-based decision process will depend critically on the available data. It seems paradoxical that when data are sparse they are often consistent and coherent (for example, when produced by one or a few laboratories) but when data become more abundant "data gaps" appear as the problem is examined by different methods and from different perspectives. This is a natural evolution, but it makes it difficult to stipulate what should be considered "sufficient data'' for a particular decision process.

Efficient and practical provision of safe drinking water to communities of varying sizes and widely differing sources and qualities of raw water is a challenging task. Any regulation affecting this complex patchwork of local, regional, public, and private systems will almost certainly have effects beyond the intended ones. It is beyond the scope of this report to consider all the possible ramifications of its recommendations on those individuals and organizations charged with future identification and regulation of contaminants. Section 1412(b)(1)(A)(i) of the amended SDWA explicitly directs the EPA administrator to identify "contaminant[s] [that] may have an adverse effect on the health of persons." Section 1412(b)(1)(C) specifies that EPA must focus on contaminants that pose the "greatest public health concern." Therefore, in framing this report the committee has chosen to adopt an explicit public health perspective, rather than any of a number of other possible perspectives (e.g., enterprise centered, economic development, or legal). The report should be read with this qualification in mind.

AWWA (American Water Works Association). 1997. Workshop Report: Mission Definition for the National Contaminant Occurrence Database (NCOD). Preliminary Draft Report. Prepared by M. M. Frey and J. S. Rosen.

Craun, G. F. 1991. Causes of waterborne outbreaks in the United States. Water Science and Technology 24:17-20.

Dzurik, A. A. 1990. Water Resources Planning. Savage, Maryland: Rowman & Littlefield, Publishers.

EPA (Environmental Protection Agency). 1994. The National Public Water System Supervision Program: FY 1993 Compliance Report. EPA/812/R/94/001. Washington, D.C.: EPA, Office of Water.

EPA. 1996a. Safe Drinking Water Act Amendments of 1996: General Guide to Provisions. EPA/810/S/96/001. Washington, D.C.: EPA, Office of Water.

EPA. 1996b. The Conceptual Approach for Contaminant Identification (Working Draft). EPA/812/D/96/001. Washington, D.C.: EPA, Office of Ground Water and Drinking Water.

EPA. 1997a. Announcement of the Draft Drinking Water Contaminant Candidate List; Notice. Federal Register 62(193): 52194-52219.

EPA. 1997b. Meeting Summary: EPA National Drinking Water Contaminant Occurrence Data Base. Washington, D.C.: RESOLVE, Inc., Contract # 68-W4-0001 prepared for EPA, Office of Ground Water and Drinking Water.

EPA. 1997c. Options for Developing the Unregulated Contaminant Monitoring Regulation: Background Document (Working Draft). EPA/815/D/97/003. Washington, D.C.: EPA, Office of Ground Water and Drinking Water.

EPA. 1998a. Announcement of the Drinking Water Contaminant Candidate List; Notice. Federal Register 63(40):10274-10287.

EPA. 1998b. Drinking Water Contaminant List. EPA/815/F/98/002. Washington, D.C.: EPA, Office of Ground Water and Drinking Water.

EPA. 1998c. Definition of a Public Water System in SDWA Section 1401(4) as Amended by the 1996 SDWA Amendments. Federal Register 63(150):41939.

Gibson, M. C., S. M. deMonsabert, and J. Orme-Zavaleta. 1997. Comparison of noncancer risk assessment approaches for use in deriving drinking water criteria. Regulatory Toxicology and Pharmacology 26:243-256.

Herwaldt, B. L., G. F. Craun, S. L. Stokes, and D. D. Juranek. 1992. Outbreaks of waterborne disease in the United States: 1989-90. Journal of the American Water Works Association 84:129-135.

Moore, A. C., B. L. Herwaldt, G. F. Craun, R. L. Calderon, A. K. Highsmith, and D. D. Juranek. 1993. Surveillance for waterborne disease outbreaks-United States, 1991-1992. Morbidity and Mortality Weekly Reporter 42:1-22.

Neal, R. A. 1985. Chemicals and safe drinking water: National and international perspective. Pp. 1-8 (ch. 1) in Safe Drinking Water: The Impact of Chemicals on a Limited Resource, R. G. Rice, ed. Chelsea, Mich: Lewis Publishers, Inc.

NRC (National Research Council). 1977. Drinking Water and Health. Washington, D.C.: National Academy Press.

NRC. 1997. Safe Water from Every Tap: Improving Water Service to Small Communities. Washington, D.C.: National Academy Press.

Viessman, W., and M. J. Hammer. 1985. Water management: Environmental considerations. Pp. 18-26 (ch. 2) in Water Supply and Pollution Control (fourth edition). New York: Harper & Row, Publishers.