Status of Science and Policies for Ensuring the Protection of Source Water and Drinking Water*
The system that was put in place for delivering safe and adequate supplies of drinking water has been in existence for more than 100 years. During the ensuing century, the United States has experienced a surge in population growth, which is projected to increase until 2050. This surge represents a shift of population from sparsely populated rural areas to more densely populated urban areas and greater demands on water for multiple needs—recreation, drinking water, industrial use, and agriculture. The policy framework within which we live today was put in place prior to the challenges of our current modern day life and its suitability for the task before us—delivering safe, clean, and adequate supplies of drinking water to people—is being questioned. Lynn Goldman, professor at the Bloomberg School of Public Health, challenged speakers and participants to consider whether we are using the right paradigms and if we should continue to patch, repair, and expand the existing system or whether a new paradogma is required?
ARE THE CURRENT POLICIES ABLE TO MEET CURRENT AND FUTURE CHALLENGES?
There are many factors to consider in making risk management decisions about contaminants in drinking water, including the legal mandates and requirements, the control options, exposures, dose-response relationships, costs and benefits, laboratory methods, and agency policies, said
The policy framework within which we live today was put in place prior to the challenges of our current modern day life and its suitability for the task before us—delivering safe, clean, and adequate supplies of drinking water to people—is being questioned.
Frederick W.Pontius, president of Pontius Water Consultants, Inc. Such decision making cannot be based on science alone. More commonly, it is a blend of science and policy. It is a mixture of different lines of reasoning, different facts, different assumptions, and different judgments made by people with different perspectives. Trac ing the Safe Drinking Water Act (SDWA) from its authorization in 1974 through its various reauthorizations, Pontius posed six questions that underlie the interface of science and policy:
1. How should contaminants be selected for regulation? In 1974, when the SDWA was passed, the U.S. Environmental Protection Agency (EPA) was given discretion—general authority—to regulate contaminants. This resulted in a few regulations, such as those for trihalomethanes, but the pace was slow. So in 1986, Congress mandated regulation of 83 contaminants, whether they needed regulation or not, Pontius noted. Given this large number of required regulations, the contaminants were divided into several phases, each in turn regulating a subset of the 83. There also was a requirement in 1986 to regulate 25 contaminants every 3 years.
Inevitably, the number of contaminants regulated increased (see Figure 1.1). In the early 1990s, policy makers realized that continuing this regulatory pace—principally because of data, but also because of sheer resources—would make it impossible to meet the goals outlined in the 1986 amendment. Thus, in 1996, the law was amended again to mandate future contaminant regulation with contaminants selected from the Drinking Water Contaminant Candidate List (DWCCL). Monitoring of unregulated contaminants was required in order to collect data to determine those that posed the greatest risk. Therefore, in the current selection process EPA considers the available data and makes a determination whether or not to regulate at least five contaminants every 5 years. Meanwhile, the standards for the 83 contaminants that were regulated as a result of the 1986 reauthorization have been retained.
The EPA also applies the concept of “meaningful opportunity for risk reduction,” which was specified in the 1996 law. The agency may now conclude—especially in those cases where exposure occurs mostly through several routes (such as food, air, and water)—regulation is not warranted for numerous contaminants on the first DWCCL because there is no meaningful opportunity for risk reduction. Under such reasoning, we take a step toward focusing our resources on those contaminants that pose the greatest risk.
The issues that we are struggling with involve data gaps. Sorting through the large number of potential contaminants to identity those that pose the greatest risk is a real challenge.
The issues that we are struggling with involve data gaps, observed Pontius. Sorting through the large number of potential contaminants to identify those that pose the greatest risk is a real challenge. Currently, a Drinking Water Advisory Council committee is considering the recommendations of a National Academies report on selecting contaminants. This committee is sorting through the universe of potential contaminants to
develop a preliminary contaminant list. From the preliminary list, a final contaminant list will have to be developed by EPA using a transparent, scientifically sound process.
2. How should our health goals be established? In 1974, the SDWA specified the use of recommended maximum contaminant levels (RMCLs) as health goals; they were renamed Maximum Contaminant Level Goals (MCLGs) in 1986 and rendered nonenforceable, although they were based on health effects assessments with the findings expressed in criteria documents. In 1996, the health effects assessment was given specificity in the statute and a greater focus was placed on sensitive populations.
The policy for regulating a known carcinogen with an MCLG of zero was established in the 1970s and is generally applicable today. In a legal challenge regarding the agency’s regulation of chloroform, the court ruled in favor of the petitioner, concluding that EPA had not used the best science. EPA is now in the throes of revising its cancer risk guidelines and a nonzero MCLG has been proposed for chloroform.
Non-cancer effects typically are based on calculations involving a reference dose (an allowable daily intake). However, its replacement with a benchmark dose (a no-observed-adverse-effect level that is the highest dosage administered that does not produce toxic effects) is something that has been considered for quite some time, said Pontius. A policy change in that direction might allow science to more consistently drive health effects assessments.
3. How should MCLs be established? The SDWA was modified in 1986 to require that maximum contaminant levels (MCLs) be set as close to the health goal as feasible, and “feasible” was defined as the use of the best technology (treatment techniques or other means). In 1996, this led to a situation, in terms of regulating those mandatory 83 contaminants, in which the MCL would be zero for known or probable carcinogens in drinking water. More pragmatically, if a treatment technology existed that could lower the contaminant level to a new threshold—a practical quantitation limit (PQL)—then the MCL would be set not at zero but at that PQL.
In 1996, more flexibility was given to EPA to consider offsetting risks and risk tradeoffs in setting MCLs in drinking water regulations, noted Pontius. The administrator must determine whether the benefits justify the costs, and if not, the regulatory limit can be adjusted to the point at which they do. We are learning, with drinking water standards
and other issues, the ramifications of proceeding in one direction when we do not know what it is ultimately going to cost. We need to consider the cost of regulation before we commit to a regulatory direction, said Pontius.
One of the main challenges for the future is whether to regulate contaminants one at a time, regulate groups of contaminants, regulate surrogates, or pursue some other approach. This is complicated by the fact that many possible contaminants could occur. Another challenge involves the regulation of small water supply systems: should they be given special consideration in terms of affordability? In both areas, additional work needs to be done.
4. What constitutes the best science? Science was not explicitly addressed in either the original statute of 1974 or its reauthorization of 1986. In petitions that challenged EPA’s science in particular rules, the agency relied on “deference.” That is, the court generally will not rule that a regulation was arbitrary and capricious unless something is obviously incorrect about it. Deference is given by the courts to the agency’s judgments.
Since 1996, the statute requires EPA to use the best available peer-reviewed science in existence at the time of regulation. Yet even in meeting this standard much could still be said in terms of different perspectives, different lines of reasoning, and differences of opinions. A stakeholder always has the option to challenge a rule through a process called judicial review, but deference is given to EPA by courts, noted Pontius.
Future challenges for maintaining the use of the best science include filling data gaps, ensuring high-quality peer reviews, and applying “fair-minded thinking” to integrate differing or conflicting lines of reasoning among all stakeholders, regardless of their relative advantage. Also, the implications of underlying assumptions and presumptions must be transparent, and EPA must be willing to change its policies when justified.
5. What is the role of source protection? Source protection was contained in the first Safe Drinking Water Act in 1974 in the form of an underground injection control program. In 1986, the wellhead protection program was added, and in 1996, additional emphasis was placed on source protection through the source water assessment and source water petition programs. Certain aspects of Clean Water Act programs, such as TMDLs (total maximum daily loads), also have a direct application to source protection, Pontius pointed out.
6. How can compliance be ensured? In 1974 the primary focus was on technical assistance and establishing new state programs. There was
no federal funding for water systems, and variance and exemption provisions were included in the original law. Over the years, the emphasis on compliance also has evolved with subsequent reauthorizations. Regulatory compliance training for water systems currently is weak and must be improved.
In 1986, enforcement provisions were strengthened and a compliance period was set for 18 months. With the realization that 18 months was too short for many systems to obtain necessary financing, in 1996, compliance periods were extended to 3 years, plus 2 additional years if a system required capital improvements.
As progress is made regulatory agencies must realize that training and technical assistance are important. Although issues involving affordability for small systems can often be addressed, at least temporarily through variances and exemptions, the nation needs a better implementation of strategy for small water systems. Pontius noted that many of the current provisions embodied in the SDWA grew out of our prior failures and that if some provisions in the law are not working well, there is room for creative thinking. Also, when science advances, regulatory policies and practices must be adjusted accordingly.
ARE RECENT ADVANCES IN SCIENCE AND TECHNOLOGY ABLE TO MEET THE HEALTH CHALLENGES OF PROVIDING SAFE DRINKING WATER?
Answering the question whether science and technology are adequately providing safe drinking water requires understanding the risks that drinking water may carry, noted Jeffrey Griffiths of Tufts University School of Medicine. Many of these risks are related to the population, which is not only growing in size but changing in its characteristics—particularly with respect to enhanced sensitivity to waterborne contaminants. Thus, at the same time that water must be reused given the growing demand, it also must be purer than ever. Meanwhile, the changing activities and increasingly concentrated locations both of people and of industries have resulted in significant levels of new and emerging contaminants. These, together with agents already well established in the environment, represent approximately three million potential chemical contaminants—that calls for paradigm shifts in the ways scientists think about these issues, noted Griffiths.
People are living longer, and the U.S. population is not only growing—with 275 million in the year 2000—but aging, noted Griffiths. Although there is increase in the number of young children, there also will be significant increases in the age 55+ group and, more specifically, among those 85 and older. One reason, he pointed out, is that people are living longer, including those with chronic diseases, such as diabetes (see Figure 1.2). Individuals with diabetes, asthma, or chronic heart conditions, for example, and even those with HIV are surviving much longer than in the past.
One effect of the increasing population is that the number of highly sensitive individuals and the consequent demand for very safe drinking water will increase. People are especially susceptible to infections or chemical contaminants in infancy, during pregnancy, when they undergo various medical treatments, and when they become elderly. What this means, according to Griffiths, is that at some time all individuals will be in a susceptibility group.
Geographic Concentration of People and Industries
Confounding the problem of increasing population is the increasing density of the population in urban areas. Griffiths observed that in some areas of the country, demand for water will outstrip supply—for example, in the Boston-Washington corridor. On the West coast of the United States, this situation is exacerbated because of drought conditions that further limit the availability of water. The end result is that we will have to reuse water, he predicted, even as the source waters are initially laden with waste from humans and animals. As industries such as meat production become more concentrated, land is paved over or otherwise made impermeable, and more contaminants are washed into local waters. In terms of health, this means that infectious diseases can be transmitted more easily, noted Griffiths.
Multiple barriers have formed the cornerstone for ensuring safe drinking water. Different barriers, such as watershed protection, filtration, and disinfection, have represented critical and tremendous advancements in public health by nearly eliminating diseases such as leptospirosis, cholera, and typhoid from the United States. Newly emerging diseases are providing new challenges because some of these diseases are resistant to conventional treatment, humans as well as animals are involved in their spread, and a tiny inoculant can infect a number of people, cautioned Griffiths (see Box 1.1).
Newly emerging diseases are providing new challenges because some of these diseases are resistant to conventional treatment, humans as well as animals are involved in their spread, and a tiny inoculant can infect a number of people.
SOURCE: Griffiths (unpublished from presentation).
In terms of providing safe drinking water, 99 percent removal of these pathogens can still allow for transmission of disease. Cryptosporidium is a good example, having caused the worst waterborne disease outbreak in U.S. history, with more than 400,000 cases of illness in 1993 and a number of deaths in the susceptible population—individuals with HIV or children with cancer. The outbreak occurred because of a failure of one filtration plant. Even though the disinfection process was still used, it was not effective against Cryptosporidium.
Another example is hepatitis E virus, which kills 15–25 percent of pregnant women who are infected. This pathogen is on the horizon as a serious potential health problem and is currently found in sewer systems in Spain and Washington, D.C., among other locations. Researchers have suggested that it may have made the jump from animals to humans, given the similarities between the strains in humans and swine.
As health officials confront emerging diseases such as SARS (severe acute respiratory syndrome), hepatitis E, and so forth, the public health burden may be great if one of these turns out to be a waterborne disease. Griffiths questions whether scientists and public health officials know what to look for and whether the technology available will diminish the threat. He noted that we know in part what to look for because of knowledge of the classical disease-causing agents and some of the new agents.
One limitation of the technology used for monitoring is the use of culture methods. The low concentration and the inability of some pathogens to replicate on an agar plate continue to be problems. Sensitive and specific monitoring is needed to detect pathogens, chemical compounds, and acts of bioterrorism, said Griffiths. Moreover, the system should be
inexpensive, ubiquitous, every jursidiction should have one, no matter how rich or poor the community—and easily shared online.
For example, many people are talking about gene chips, which are used in concert with methods to amplify genetic material. Griffiths predicted that we also will have water detection chips, which will have 50,000 or 100,000 sections of nucleic acid materials to monitor the water for chemicals and toxins. Similar results also may be achieved with technologies such as optical sensors—which also could function as chemical indicators in identifying waterborne chemicals—and biosensors sensitive enough to detect one anthrax spore or one Cryptosporidium.
Meanwhile, Griffiths observed, it is important to continue monitoring for classical or known threats, but to remove all pathogens that get into drinking water, including the unidentified pathogens that cause some 80 percent of the outbreaks of waterborne disease. In addition, it is important to not leave any chemical traces and to remove naturally present but harmful chemicals such as arsenic—all while using less water and dealing with the presence of sewage and industrial waste. The challenges here are immense when it comes to health.
Conventional technology worked in 1910 and is still working today for many communities with normal populations, said Griffiths. However, it clearly has its limitations, especially for susceptible populations. Advanced technologies that can neutralize pathogens with ultraviolet radiation or pull them out with membrane technologies do exist, but they are not affordable by many systems. So we need technological advances, some of which may simply drive down the costs of present systems, though others will have to be of a different generation.
In effect, these treatment technologies should be inclusive, said Griffiths. They should monitor and eliminate across the spectrum of toxins and chemicals. They are necessary because it will be very difficult to come up with narrowly focused new treatment technologies that address one contaminant at a time—there are just too many of them and some will remain unknown. It is better, he concluded, for us to come up with solutions that essentially eliminate all of these risks at once.