Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Setting Environmental Standards for Hazardous Waste Sites: A Break from the Past or a Continuum? RICHARD M. DOWD Keynote addresses are an environmental hazard; they con- stantly present the danger of boring the audience into somnolence before the real proceedings get under way. After ~ examined the program for this symposium and read some of the thoughtful papers prepared for our discussion, ~ became particularly appre- hensive about triggering such a reaction this afternoon. Certainly, as a practicing scientist, ~ would be much more comfortable re- porting to you on, for instance, my findings on a comparison of recent ground water investigations involving the effects of well casing materials on monitoring results. But the more ~ thought about the general direction of this program's papers, in light of my own experiences with standard setting in various environmental media and regulatory programs over the past 15 years, the more ~ began to believe that our discussions here actually represent one point in a continuum. The standard-setting art has evolved over these 15 years, but we still face an uncanny sense of merely running in place when we look at some of the issues that still bedevil us. ~ would like to describe some milestones in the evolution of this art and one principle in particular that ~ believe our experience has demonstrated to be bedrock. Let me start by trying to put the setting of standards for hazardous waste sites into something of a historical context. would be the first to label my brief account of standard setting as a revisionist history of the evolution of environmental rule making 13
14 HAZARDOUS WASTE SITE MANAGEMENT over the past 17-odd years. Like many people, I would date the modern era of environmental regulation as starting around 1970 with the passage of the National Environmental Policy Act and the Clean Air Act. ~ further believe, as many of you may, that there are some differences in the way we did things routinely before 1977 and the way we have done things routinely since that time. In those early years of modern environmental regulation the emphasis was on what we tend to call conventional pollutants, those chemicals that were then known to make up the bulk of wastes disposed of from stacks or pipes and discharged into the air or water. In that time, the approach to setting standards for the most part used the concept of thresholds. The concept was useful, particularly in dealing with ways to regulate these bulk contaminants that cause or contribute to the types of human illness that were generally associated with these conventional pollutants. This concept of threshold levels below which these substances were "safe" was built into the language of the legislation and in turn drove the regulations. In both of the 1970s air and water laws, we thought we knew enough to determine the danger levels and to establish a threshold above which damage occurred to people and to ecological systems and below which damage did not occur. We also believed, and the legislation iterated, that beyond that threshold we needed to provide a certain margin of safety. With such a margin, if there were a mistake in our estimate of the threshold or if the standard were slightly exceeded, we would still be out of the damage range. During these years, some of the most intensive debate occurred over how large these margins of safety ought to be and what they should take into account. Even as recently as 1980, when debate was raging over the revision of the photochemical oxidant (ozone) air standard, a major part of that argument was the size of the margin of safety, as well as where the threshold was. In general, in setting those early standards the threshold concept had value in addressing health concerns, but it did not routinely allow ecological or other values, including aesthetic concerns, to be addressed. For example, visibility was regarded as a secondary value, as was materials damage, and these concerns were therefore termed "welfare" effects and addressed through the so-called "secondary standards." And although the legislation mandated the setting of these secondary standards, it put little emphasis on achieving
SETTING ENVIRONMENTAL STANDARDS 15 or enforcing them. As a result, many of the secondary standards have essentially been scrapped or have just not been enforced. Yet part of this neglect or downgrading is a result, it seems to me, of our inadequate knowledge in those areas. We had very little baseline data with which to work and a very limited understanding of some basic ecological processes. Our monitoring systems were almost nil and some of us might argue that they have not been all that much improved today and our modeling capabilities were in their infancy. In only a few cases did we know how to judge effects on ecological systems. We did know, for example, that dissolved oxygen affects aquatic life, and the water law set a goal of "swimmable, fishable" water quality. Thus, there was a dis- solved oxygen standard, with a threshold that was set to prevent fish from dying. But dealing with other waterborne pollutants proved far more troublesome, and, in fact, the committees writ- ing the 1972 legislation essentially threw up their hands over the difficulty of directly relating the concentrations of contaminants in water to human health or to ecological values. Essentially, the approach adopted in the law mandated a technology-based stan- dard. Thereafter, attempts were made to relate such standards to water quality parameters on a quantitative basis. Most standards, then, in this first wave of environmental reg- ulation, focused on the stack or on the outfall. They reflected attempts to set a number based on a threshold and required en- vironmental managers to control the stack or outfall emissions to meet the numbers. Other environmental legislation followed a similar approach. For example, the federal pesticide statute that made EPA, rather than the U.S. Department of Agriculture, re- sponsible for setting acceptable limits on pesticide residues in food incorporated this threshold concept under the term "tolerance." Even during this phase of regulatory clevelopment, however, there were elements that foreshadowed the regulatory approach of the post-1977 period. The nascent cancer policy that EPA first published in 1975 was a preview of coming attempts to quan- tify, weigh, and balance a process that has since evolved into what we could basically call the "postthreshold concept" age of regulation. More and more, we began to deal with what we gener- ically refer to as tonics. This focus on substances different in nature, concentration, and behavior from the conventional bulk pollutants has itself been driven, in many cases, by advances in analytical chemistry and the improvement in the sensitivity of our
16 HAZARDOUS WASTE SITE MANAGEMENT methods, developments that have driven detection limits lower and lower. The shift in focus has also been fueled by changes in biolog- ical experimentation: the number of long-term studies on rodents and other mammals to determine whether or not substances are carcinogenic has burgeoned since the early 1970s. We began to collect a data base that previously had been nonexistent, and this new body of information has led to major changes in the way we look at the issues and, consequently, to an evolution in standard setting. The first important legislative embodiment of this change was reflected in the passage of the Toxic Substances Control Act in 1976. This legislation was spurred by public concern over the discovery of suites of chemicals that were present often in un- known levels in soil, air, and water across the country in the mid-1970s. Not only were the concentrations of these chemicals often unknown; in many cases, their effects were also a question. And because so many of these substances were so widely prevalent and so integral a part of our industrialized society, concern about them led to a substantial increase in the use of different forms of quantitative risk assessment. In particular, when dealing with known or suspected carcinogens, risk assessments became an in- dispensable tool because of the (appropriate) consensus that there may not be thresholds in these instances; that risk may exist at any level, no matter how low; and that there is no strong basis for assuming with a few exceptions- that the human body re- covers from exposure to environmental carcinogens. If, therefore, the possibility of risk cannot be allowed to drive standards to zero because existence of the carcinogens in question is deemed necessary or inevitable then the only alternative is either to de- termine some nonzero value for acceptable risk or at least a level at which the benefits of such an exposure or risk outweigh the costs. This conclusion has led to growing efforts to focus on long-term chronic effects and a corresponding increase in research on and the application of quantitative risk assessment. In recent years since 198~public concerns over hazardous wastes have overtaken the earlier focus on conventional pollutants and moved to an almost exclusive preoccupation with the presence of toxic chemicals and the possibility that even trace concentra- tions can cause long-term chronic health damage. Such fears have, to some degree, led to a call for restrictions even on minute con
SETTING ENVIRONAD3NTAL STANDARDS 17 centrations that may, in fact, be irrelevant to human exposure and therefore to human health risk. To some extent, demand for the development of such stringent standards reflects a worthy premise: that already bad situations should not be made worse and should, in principle at least, be cleaned up. And there is where the rub sets in for the standard maker. Obviously, we can take a worst-case analysis, apply it to an unrealistic extreme, and set a standard designed either to prevent or to "curer it. But before we say that this is a bad idea-or a good one, for that matter-we should consider some of the elements involved In setting any standard. Theoretically, at least, the standard setter must, at a minimum, take three categories of knowledge into account: 1. the effects of whatever substance is of concern, whether these be carcinogenic, mutagenic, acute or chronic, Tong or short term; and the levels at which these effects can occur, either the reference dose or threshold, or a zero level; 2. the concentrations at which the substance is present in whatever media are of concern (in the case of hazardous wastes, these media are generally ground water and soils); and 3. how the material is transported and transformed as it moves to and from its site. As we look at our knowledge base regarding contamination at hazardous waste sites, we find that, in comparison to our base for air and water standard setting in the pre-1977 period, we know very much less than was then the case in ahnost all of these categories, particularly ground water. Our intellectual capital is thin. If, for the sake of discussion, one accepts my broad revisionist history of regulatory standard setting, can any useful compar- isons be drawn from it about the differences between then and now in formulating sensible standards? Today we are dealing with much smaller quantities or concentrations of substances than in previous times; we have much less knowledge about effects at these small levels than we previously had about effects of higher concentrations of other pollutants; and we are tending to Took toward chronic and long-term effects rather than acute, short- term effects. One other notable difference should be explicitly
18 HAZARDOUS WASTE SITE MANAGEMENT mentioned: we are often routinely attempting to deal with multiple chemicals, in multiple media, through multiple routes of transport and exposure. Even a cursory review of the papers prepared for this symposium highlights these characteristics of contemporary standard setting. ~ do not intend in these remarks to address the details of how standards should be set. Many of the colloquium's participants have studied this issue in far more depth than ~ have and have been developing promising methodologies to generate sensible values. But ~ have recently been considering some of the principles that must underlie any standard-setting efforts if the results are to mean or achieve anything in the real world, and it is that thinking that ~ would like to share here. For any standard to be valid, ~ suggest that it must have an objective, a meaningful range of application, and a realistic means or measure for verifying its success. These criteria apply, ~ believe, to standards in any field, whether it be (for example) education, consumer safety, or environmental protection. It is this third criterion of verification and its underlying assumption of what ~ call a "commitment to truth" that is the focus of the rest of my remarks. In dealing with standards designed to protect health, there is necessarily a concern with ensuring adequacy, and conservatism is often invoked as a necessary principle. Indeed, the objective of a health standard ought to include the provision of a desired level of protection, and, accordingly, margins of safety and ranges of application could be larger or smaller, depending on the degree of protection deemed adequate. But where conservatism cannot be invoked as a principle, in my view, is in the area of measurement, of verification. Here, in trying to assess truth, we cannot abort to wear either rose-colored or dark glasses. Without a commitment to unadorned truth on the part of all involved in standard setting or evaluation, standards will end up as mere exercises that do not encourage increased knowledge which itself could lead to the reevaluation of existing standards and perhaps therefore not even to improved protection. In this regard, we can learn something from some of the pre- ceding efforts in standard setting in the more traditional areas. Here, too, we were initially compelled to act, to select numerical values and limits, in the face of incomplete knowledge about pollu- tant ejects, levels, and transport routes and mechanisms. Because
SETTING ENVIRONMENTAL STANDARDS 19 we could not use inadequate knowledge as an excuse for delay, we were forced to develop predictive tools for example, models to assist in relating emissions at the stack level to atmospheric con- ditions some distance away, and eventually to concentrations that people outside property lines would breathe. Thus, in the air quality area, the need to establish ambient air quality standards- both to satisfy the legislative control scheme and to incorporate our then-current knowledge of thresholds led inevitably to a need to develop good air quality dispersion models based on and en- couraging an understanding of the transport and transformation of chemicals in the atmosphere. Despite the fact that the application of air dispersion mod- eling, as it has developed, reflects a tendency to require the use of models that predict the worst cases the most severe meteorol- ogy coupled with the highest emissions the truth of the models themselves can be tested, and these worst cases can be verified. Monitoring can determine whether the models predict accurately or inaccurately. There have been considerable debate and contro- versy over whether or not one of the criteria for mode! validity that is often used- consistency between accepted models and newly de- veloped models is necessary. But in any event, there has never been any question that the models need to be logically consistent, that they need to be tested and verified, and that there ought to be a real relationship between predictions and actual atmospheric conditions in the world about us. Likewise, in determining human exposure and therefore risk monitoring must be representative of the contamination that exists where people breathe. In fact, some monitoring programs have been challenged, successfully, on the basis that they do not realistically test the air to which people are exposed. Among other things, this commitment to truth has led to a substantial improvement in our knowledge base of the physics and chemistry of the atmosphere and the movement and transforma- tion of pollutants. Of course, we do not know everything, but we have begun the process of knowing in a much more vigorous way than we did in 1970. The lessons learned in the often difficult process of develop- ing air quality models to verify standards can be applied in the case of hazardous waste standard setting. It is my concern that if standards are set in this area without ensuring that the measures adopted for their validation incorporate a comrn~tment to truth
20 HAZARDOUS WASTE SITE MANAGEMENT sinner to that we are building into verification processes for other standards, the resulting lack of relationship to real-worId condi- tions will make it difficult for scientists and engineers to contribute to sensible decisionmaking in a regulatory agency. Without devel- oping appropriate methodologies be they for modeling or for monitoring systems to formulate and test our hazardous waste standard setting, our efforts will not embody the commitment to truth, to verification, to testing against the real world that ~ am advocating. In areas like hazardous wastes, in which we are faced with inadequate knowledge, tight timetables for taking action, and heightened public concern that we act to protect human health adequately, it is tempting to respond by establishing worst-case conditions, however unrealistic, and regulating "against" them. Take, as one example, the attempt to establish a methodology for Relisting wastes at a hazardous site. EPA devised a methodol- ogy and model that assumed that 100 percent of a given chem- ical for a site would leach into the ground water. At the same time, because the mode!-rightly-dealt with multimedia effects, the mode! makers wanted to incorporate any volatilization of the chemical into the air. The mode} they devised simultaneously as- sumed that 100 percent of the chemical leached into the ground water, and 100 percent of the chemical also volatilized into the air. The results of these two exercises were to be compared against existing air and water standards. But this simultaneous 100 per- cent behavior is not logically consistent; it cannot reflect the real world, it does not incorporate a commitment to truth, and it will not lead to any improvement in science or in technology to deal with hazardous waste sites. The assumptions are so excessively conservative as to be logically impossible. My argument here is not against conservatism; ~ repeat my be- lief that conservatism is an appropriate policy in standard setting and in mode! development. It is an acceptable, even necessary, part of the political process to establish a level or a standard or a number that is more protective than the minimum. But that is not the same as constructing a mode] that, because of its conser- vatism, because of the illogical assumptions that are built into it, can never be validated under real-worId conditions. In my view, there is never justification for not holding to the principle of using the real world as our ultimate measure. We may not always satisfy the test of absolute consistency with real-worId
SETTING ENVIRONMENTAL STANDARDS 21 conditions, either because of our lack of knowledge or through flaws in our standard-setting processes. Those processes must, however, allow for correction. Only a problem so overwhelming that mitiga- tion cannot wait would, in my view, justify our abandoning efforts to make standards conform to reality. Does the seriousness of the dangers posed by hazardous waste sites satisfy this criterion for abandoning "realistic" standard set- ting? Ibelieve not. In support of this belief, T would like to cite a newly completed evaluation by EPA of the relative risks associated with 31 different environmental hazards in terms of four concerns: cancer risks, noncancer risks, ecological effects, and welfare effects. The relative ranking for the risks associated with hazardous waste sites was low to medium, pretty much right in the middle of EPA programs. Although details of this ranking could be debated, it seems clear that hazardous waste sites are not an overwhelming risk compared with other areas of concern. ~ do not mean to suggest that a commitment to truth needs to stand In the way of standard setting. ~ am fully aware that political pressure and public concern often require, properly, action by reg- ulatory agencies even in the absence of full knowledge. However, the absence of knowledge and an understanding of uncertainties is not the same as ignoring knowledge. Our understanding of the movement of trace chemicals (some- times in quantities that make "trace" a misnomer) will undoubt- edly increase. As long as we are committed to using the best of that understanding in our standard-setting processes, as long as we are prepared to revise our view of the world as we learn, the standard-setting process will be healthy and defensible. There are many incentives to improved understanding: in- creased regulatory attention, public concern, and large potential costs (environmental, human health, and economic). But in all of this we cannot delude ourselves: if we do not see the "real" world, we cannot solve "real" problems. As Polonius, that ultimate bore, provoker of drowsiness" and purveyor of truth says in Hamlet, "To shine own self be true, and it must follow, as the night the day, thou canst not then be false to any man.'